’r

..

A .A ' 1w 2‘:
.1... 4%. ‘
V X'

«a
A.
1.. A
'-.w
a. ..

.

.—

w

boivv ‘0
Jams" .

J?

. ...~. .
M mfg“. r-M v
rn~,,‘.:'.‘.’.ro

.' ma.“
—\ «1953.?»-
..z.:,, .

33‘

-A “'M

”if?

m 3...”!

ﬂ" . ...:..‘ "gun--

It"

' Er“
JVIW

M

4.1 A

.-;

$1.. w
.I 3.}!

5
5s».
4)

 

 

TH E873

I? W‘W

LIBRARY
Michigan State
University

This is to certify that the
thesis entitled

Epidemiology of Brucellosis in Livestock and Humans in
Mongolia

presented by
Amanda Elizabeth Fine

has been accepted towards fulﬁllment
of the requirements for the

Master of degree in Epidemiology

 

 

Science
WW

Unajor Professor’s Signature
W8 / 25/ 0 AL

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 czﬁlFlC/Dateﬁjndd-pis

 

EPIDEMIOLOGY OF BRUCELLOSIS IN LIVESTOCK AND HUMANS IN
MONGOLIA

By

Amanda Elizabeth Fine

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Epidemiology

2004

ABSTRACT

EPIDEMIOLOGY OF BRUCELLOSIS IN LIVESTOCK AND HUMANS IN
MONGOLIA

By

Amanda Elizabeth Fine

The focus of this thesis is a study of the epidemiology of brucellosis in livestock
and humans in the Khangai mountain region of Bayankhongor Aimag (Province),
Mongolia. The study was cross-sectional in design and conducted at the herd-level using
a stratiﬁed random sample (n=105) of the 5,062 cattle (predominantly yak) herds in the
region. The brucellosis status of the cattle herds was determined by using a combination
of the Brucellosis Ring Test (BRT), on milk, and the Brucellosis Card Test (CARD), on
serum, to detect Brucella antibodies. Livestock management practices, the occurrence of
brucellosis-like symptoms in humans who herded the livestock, human/livestock contact,
and practices of livestock product consumption were recorded with the use of two
standardized questionnaires conducted as in-person interviews.

The estimated prevalence of brucellosis-seropostive cattle herds in the region was
53%; +/-9.5% (BRT or CARD-positive) and/or 24.3%; +/- 8.1% (BRT and CARD-
positive). Increasing herd size and the occurrence of abortions in cattle were associated
with brucellosis-seropositive herd status. The estimated prevalence of the occurrence of
brucellosis-like symptoms in the human herder population was 24%. The practices of
consuming raw milk, preparing dairy products from raw milk, and keeping newborn
livestock in human living quarters, were associated with the occurrence of symptoms

suggestive of brucellosis in the human herder population.

This thesis is dedicated to the livestock herders of Bayankhongor Aimag (Province),
Mongolia, in appreciation of their generous hospitality, and in admiration of their skill

and dedication to life on the open steppe.

iii

ACKNOWLEDGMENTS

This study would not have been possible without the cooperation and support of
many individuals, organizations, and institutions. I am grateful to all who provided their
time, expertise, guidance, and companionship during the course of this project. In
particular, I would like to acknowledge: Mongolia’s Chief Veterinarian, Dr.
Ulaambayar, and epidemiologist, Dr. Dorjsuren, of the State Veterinary Service,
Mongolian Ministry of Food and Agriculture, for the permission to perform this study
and for providing the background on past and current brucellosis control policies in
Mongolia; the Bayankhongor Aimag (Province) veterinarians, Dr. Batarj av and Dr.
Mondon-Gadis, for their assistance with project coordination; the soum (county)
veterinarians of Gurvanbulag (Mr. Serjpurev), Zag (Mr. Narmandakh), Jargalant (Ms.
Nanjiid), Galuut (Ms. Batnasan), Erdenstogt (Mr. Dalantai), and Shargaljuut (Mr.
Tolgach), for their expert guidance, and the many hours spent with us in the ﬁeld.

The Agricultural Cooperative Development IntemationaWolunteers in Overseas
Cooperative Assistance (ACDI/VOCA)l Project ofﬁce in Ulaanbaatar, Mongolia,
provided logistical and ﬁnancial support for the study. I would like to thank
ACDI/VOCA project director, Ms. Jeanne Bartholomew, Bayankhongor Aimag
(Province) ACDI/VOCA ofﬁce manager, Mr. Shagdar, and ACID/VOCA volunteer, Dr.
Earl Thompson. Brucellosis testing kits and reagents were provided by the United States
Department of Agriculture, Animal and Plant Health Inspection Service, National

Veterinary Services Laboratory, Ames, Iowa, USA. The Mongolian State Central

 

' ACDI/VOCA is a private, non-proﬁt organization that promotes broad-based economic growth in
developing countries. ACDI/VOCA’s work in Mongolia was in the agricultural sector and funded by the
United States Agency for International Development.

iv

Veterinary Diagnostic and Sanitary Laboratory volunteered their assistance with
additional laboratory testing.

I would like to thank the faculty, students and staff of the Epidemiology
Department, Large Animal Clinical Sciences Department, and Population Medicine
Center at Michigan State University, for providing technical support and encouragement
to write up this study. In particular I would like to thank: my advisor, Dr. John B.
Kaneene for his patience, gentle guidance, technical expertise, and willingness to take on
this extra responsibility; and the members of my thesis committee, Dr. C. A. Reddy and
Dr. Mathew J. Reeves, for their careful reading of this thesis and expert advice.

I would like to express my great appreciation for the effort, time and skill
dedicated to this study by the project ﬁeld team: Dr. Andrea Mikolon, for her statistical,
laboratory, and veterinary expertise; Dr. J antsansengee Baigalmaa, ACDI/VOCA
veterinary project assistant, for her invaluable contribution as translator, interviewer, and
medical professional; Mr. Nyamj av for his long hours of driving, equipment repair, and
restraining of rambunctious cattle; Ms. Jessie Fine, for her support and companionship.

I would like to acknowledge every member of the participating herder households
who took the time out of their busy lives to answer our questions, provide us with milk
samples, collect their cattle for testing, and who opened their homes to our team.
Without their cooperation and interest, none of this would have been possible.

Finally, a special thanks to my parents, James and Deborah Fine, for my early
introduction to the world beyond the borders of the United States, and for their support

and encouragement every step of the way.

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. x

LIST OF FIGURES ......................................................................................................... xiv

INTRODUCTION ............................................................................................................... 1
MONGOLIA: A BRIEF HISTORY ........................................................................ 1
BAYANKHONGOR AIMAG (PROVINCE) ......................................................... 4
STUDY RATIONALE ............................................................................................ 6
PROBLEM STATEMENT ...................................................................................... 7
OBJECTIVES .......................................................................................................... 8
OVERVIEW ............................................................................................................ 9

Chapter 1

EPIDEMIOLOGY OF BRUCELLOSIS IN MONGOLIA (1950-2002):

A REVIEW ....................................................................................................................... 11
INTRODUCTION ................................................................................................. l 1
BRUCELLOSIS RESEARCH IN MONGOLIA ................................................... l3
BRUCELLOSIS SURVEILLANCE IN MONGOLIA ......................................... 15

Brucellosis Testing ..................................................................................... 17
Livestock Brucellosis ................................................................................. 18
Human Brucellosis ..................................................................................... 22
BRUCELLOSIS CONTROL IN MONGOLIA ..................................................... 26
Historical Account .......................................................................... 26
Present and Planned Brucellosis Control Programs ................................... 30
CONCLUSIONS AND AREAS FOR FUTURE RESEARCH ............................. 31
Chapter 2

THE PREVALENCE OF BRUCELLOSIS-SEROPOSIT IVE CATTLE/Y AK HERDS
IN THE KHAN GAI MOUNTAIN REGION OF BAYANKHONGOR

AIMAG (PROVINCE), MONGOLIA ............................................................................... 39
STRUCTURED ABSTRACT ............................................................................... 39
INTRODUCTION ................................................................................................. 41
HYPOTHESES ...................................................................................................... 45
MATERIALS AND METHODS ........................................................................... 45

Study Site: Bayankhongor Aimag (Province), Mongolia .......................... 45
Study Design and Sampling Frame ............................................................ 46
Sample Size Calculation ............................................................................ 49
Herd Selection ............................................................................................ 50
Data Collection Team and Informed Consent ........................................... 51
Specimen Collection and Testing .............................................................. 51
Case Deﬁnitions ......................................................................................... 55
Herd Management Survey ......................................................................... 55

vi

Head of Household Survey ........................................................................ 56

DATA ANALYSIS ................................................................................................ 56
Herd-Level Data ......................................................................................... 56
Individual Animal-Level Data ................................................................... 57
Concordance Measurement ........................................................................ 58
Questionnaire Data ..................................................................................... 58

RESULTS .............................................................................................................. 58
General Characteristics of Livestock Herds .............................................. 58
Prevalence and Distribution of Brucellosis-Seropositive Cattle Herds ..... 60
Concordance of Herd-Level Brucellosis Testing ....................................... 63
Individual-Animal Brucellosis-Seroprevalence ......................................... 63
Human-Livestock Interaction .................................................................... 64
Human Health Status and Brucellosis Awareness ..................................... 65

DISCUSSION ........................................................................................................ 67
Brucellosis Prevalence and Distribution... ................................................. 67
Livestock Management .............................................................................. 71
Human/Livestock Interaction and Human Health ..................................... 72
Herder Knowledge of Brucellosis .............................................................. 73

CONCLUSION ...................................................................................................... 74

Chapter 3

MANAGEMENT-RELATED RISK FACTORS FOR BRUCELLOSIS IN
CATTLE/Y AK HERDS IN THE KHAN GAI MOUNTAIN REGION OF

BAYANKHONGOR AIMAG (PROVINCE), MONGOLIA ........................................... 84
STRUCTURED ABSTRACT ............................................................................... 84
INTRODUCTION ................................................................................................. 86
HYPOTHESES ...................................................................................................... 89
MATERIALS AND METHODS ........................................................................... 89

Study Design and Herd Selection .............................................................. 89

Data Collection Team and Herd Location ................................................. 9O

Brucellosis Testing of Cattle Herd ............................................................. 90

Case Deﬁnitions ......................................................................................... 91
Collection of Data Relating to Possible Risk Factors for Brucellosis-

Seropositive Cattle Herds .............................................................. 92

DATA ANALYSIS ................................................................................................ 92

Univariate Analysis of Herd and Individual-Level Data ........................... 92

Multi-Variable Analysis of Herd-Level Data ............................................ 94

RESULTS .............................................................................................................. 94

Description of Sample Population ............................................................. 94

Descriptive and Univariate Analyses of Herd Management Risk Factors 95

Descriptive and Univariate Analyses of Disease Indicators ...................... 97

Multi-Variable Analysis of Management Risk Factors and Disease
Indicators Associated with Herd Brucellosis-Seropositive Status.97
DISCUSSION ........................................................................................................ 99

Distribution of Brucellosis-seropositive Cattle Herds ............................... 99

vii

Livestock Management Risk Factors for Cattle Herd Brucellosis

Seropositivity ................................................................................. 99

Cattle Brucellosis Disease Indicators ....................................................... 101
Sero-Prevalence and Risk Factors Associated with Disease

in Individual Cattle ...................................................................... 103

Limitations ............................................................................................... 103

CONCLUSION...........................L ........................................................................ 105

Chapter 4

THE OCCURRENCE OF CLINICAL SIGNS SUGGESTIVE OF HUMAN
BRUCELLOSIS AND ASSOCIATED RISK FACTORS IN LIVESTOCK HERDERS
IN THE KHAN GAI MOUNTAIN REGION OF BAYANKHONGOR AIMAG

(PROVINCE), MONGOLIA ........................................................................................... 117
STRUCTURED ABSTRACT ............................................................................. 1 17
INTRODUCTION ........................................................................ 119
HYPOTHESES .................................................................................................... 122
MATERIALS AND METHODS ......................................................................... 122

Study Design ............................................................................................ 122
Herder Household Selection .................................................................... 123
Data Collection Team and Herd Location ............................................... 123

Head of Herder Household Interviews: Health Status and Livestock
Contact/Product Consumption Data ............................................ 124
Case Deﬁnitions............... ........................................................................ 125
DATA ANALYSIS .............................................................................................. 125

Univariate Analysis of Risk Factors Associated with Clinical Signs
Suggestive of Brucellosis and Reported Brucellosis

Diagnosis, in Livestock Herders .................................................. 125
Multi-variable Analysis of Risk Factors for the Occurrence of Clinical
Signs Suggestive of Brucellosis in Livestock Herders ................ 126
RESULTS ............................................................................................................ 127
Description of Sample Population ........................................................... 127
Univariate Analyses: History of Brucellosis Diagnosis ........................... 131
Univariate Analysis: Reporting of Clinical Signs and Symptoms
Suggestive of Acute and Chronic Brucellosis ............................. 132

Multi-Variable Analysis of Livestock Product Consumption and
Livestock Herding Practices and the Occurrence of Brucellosis-

Like Symptoms in the Herder Population .................................... 135
DISCUSSION ...................................................................................................... 136
Hypotheses ............................................................................................... 136
Prevalence of Reported Symptoms Suggestive of Brucellosis and
History of Brucellosis Diagnosis in Herder Population ............... 137
Cattle Herd Brucellosis Status and Human Brucellosis-Like Disease ..... 138
Limitations ............................................................................................... 143
CONCLUSION .................................................................................................... 144

viii

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS .................................... 160

Summary .............................................................................................................. 160
Conclusions .......................................................................................................... 162
Recommendations ................................................................................................ 163

APPENDIX A. Map of 19 Aimags (Provinces) and 2 Cities (Ulaanbaatar and

Darkhan-Uul of Mongolia. Inset: Mongolia in Central Asia. ............... 167
APPENDIX B. Livestock Management Questionnaire .................................................. 169
APPENDIX C. Human Health and Livestock Contact/Product Consumption

Questionnaire ......................................................................................... 1 72
LITERATURE CITED .................................................................................................... 175

ix

Table 1.1.

Table 2.1.

Table 2.2.

Table 2.3.

Table 2.4.

Table 2.5.

Table 2.6.

Table 2.7.

Table 2.8.

LIST OF TABLES

Summary of Published Reports of Brucellosis Prevalence in Livestock and
Human Populations in Mongolia by Year and Source ............................... 35

The 5,062 Herder Household Sampling Frame in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia, and the Relative
Proportion of Herder Households from Each Stratum ............................... 76

The Number and Percent of Herder Households Sampled in Each Stratum
and the Overall Proportion of Participation ............................................... 77

The Prevalence of Brucellosis Ring Test (BRT)-Positive Cattle Herds in
East Stratum and the Adjusted Overall Estimated Prevalence and 95%
Conﬁdence Intervals of Brucellosis-Seropositive Cattle Herds in the
Khangai Mountain Region of Bayankhongor Aimag (Province),

Mongolia .................................................................................................... 78

A Comparison of the Adjusted Estimates of the Prevalence of Brucellosis-
Seropositive Cattle Herds in the Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia, Based on the Results of the
USDA Brucellosis Ring Test (BRT), the USDA Brucellosis Card Test
(CARD), the CARD and BRT (Both Positive), the CARD and the
Mongolian Complement Fixation Test (CFT) in Series and the CARD or
BRT (Either Positive) ................................................................................ 79

The Degree of Concordance (Kappa Statistic) Between the Brucellosis
Ring Test (BRT) and the Brucellosis Card Test (CARD) as a Herd-Level
Brucellosis Testing Method ....................................................................... 80

Individual-Animal Brucel‘losis-Seroprevalence in Cattle Tested in the
Khangai Mountain Region of Bayankhongor Aimag (Province), Mongolia,
with the USDA Brucellosis Card Test (CARD) and conﬁrmed with the
Mongolian Complement Fixation Test (CFT) ........................................... 81

The Reported History of Brucellosis Diagnosis, Testing and Treatment in
the Surveyed Heads of Herder Households in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia ............................ 82

Reported Occurrence of Clinical Signs and Symptoms Suggestive of
Acute, Chronic, and General Brucellosis in the 100 Surveyed Heads of
Herder Households in the Khangai Mountain Region of Bayankhongor,
Mongolia .................................................................................................... 83

Table 3.1.

Table 3.2.

Table 3.3.

Table 3.4.

Table 3.5.

Table 3.6.

Table 3.7.

Univariable Analyses of Continuous Livestock Management Related Risk
Factors for Herd Brucellosis Status of Herds in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia, Tested in 2000,
using the BRT and CARD test [N=103 (BRT or CARD); N=96 (BRT and
CARD); Positive Herds=54 (BRT or CARD); Positive Herds=25 (BRT
and CARD)] ............................................................................................. 107

Univariate Analysis of Herd Management Related Categorical Risk
Factors for Cattle Herd Brucellosis Status of Herds in the Khangai
Mountain Region of Bayankhongor Aimag (Province), Mongolia, Tested
in 2000, Using the BRT and CARD test [N=103 (BRT or CARD); N=96
(BRT and CARD); Positive Herds=54 (BRT or CARD); Positive
Herds=25 (BRT and CARD) ................................................................... 109

Univariate Analysis (XZ-Test) of Cattle Breed, Cattle Age and Cattle Age
Class (X2 Test for Trend) as Risk Factors for Individual Bovine
brucellosis Sero—Positivity in a Population of Cattle in the Khangai
Mountain Region of Bayankhongor Aimag (Province), Mongolia, Tested
in 2000, Using the CARD Test (N= 516; Positive Catt1e=71) ................ 111

Univariate Analysis of Disease Indicators for Cattle Herd Brucellosis
Status of Herds in the Khangai mountain Region of Bayankhongor Aimag
(Province), Mongolia, Tested in 2000, Using the BRT and CARD Test
[N=103 (BRT or CARD); N=96 (BRT and CARD); Positive Herds=54
(BRT or CARD); Positive Herds=25 (BRT and CARD)] ....................... 112

Univariable Analyses of Continuous Disease Indicators for Herd
Brucellosis Status of Herds in the Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia, Tested in 2000, using the
BRT and CARD test [N=103 (BRT or CARD); N=96 (BRT and CARD);
Positive Herds=54 (BRT or CARD); Positive Herds=25 (BRT and
CARD)] .................................................................................................... 113

Risk Factors in the Full Logistic Regression Model of Livestock
Management and Disease-Related Risk Factors for Cattle Herd Brucellosis
Sero-Positive Status in the Khangai Mountain Region of Bayankhongor
Aimag (Province), Mongolia, Tested in 2000, using the BRT, CARD and
CFT test [N=103 (BRT or CARD); N=96 (BRT and CARD); Positive
Herds=54 (BRT or CARD); Positive Herds=25 (BRT and CARD)] ...... 114

Final Reduced Logistic Regression Model of Livestock Management and
Disease-Related Risk Factors for Cattle Herd Brucellosis Sero-Positive
Status in the Khangai Mountain Region of Bayankhongor Aimag
(Province), Mongolia, Tested in 2000, with the USDA Brucellosis Milk
Ring and Card Test .................................................................................. 116

xi

Table 4.1.

Table 4.2.

Table 4.3.

Table 4.4.

Table 4.5.

Table 4.6.

Table 4.7.

Prevalence of Clinical Signs and Symptoms of General Illness
Experienced Within the Previous 12 Months and Reported by the 100
Individuals Surveyed in the Khangai Mountain Region of Bayankhongor
Aimag (Province), Mongolia, in August 2000 ......................................... 146

Univariate Analysis of Reported Brucellosis Diagnosis in Humans and the
Reported Occurrence of Clinical Signs and Symptoms Typical of Acute
and Chronic Brucellosis in the 100 Individuals Surveyed in the Khangai
Mountain Region of Bayankhongor Aimag (Province), Mongolia, in
August, 2000. N=100; 16% “Yes” for Typical Acute Brucellosis; 20%
“Yes” for Typical Chronic Brucellosis .................................................... 147

Univariate Analysis of Cattle Herd Brucellosis Status and Reported
Brucellosis Diagnosis in the 100 Individuals Surveyed in the Khangai
Mountain Region of Bayankhongor Aimag (Province), Mongolia, in
August, 2000. N=100; 26.80 “Yes” for “Ever” Diagnosed with
Brucellosis; 14.43% “Yes” for “Recent-w/in 12 months” Diagnosis with
Brucellosis ................................................................................................ 148

Univariate Analysis of Cattle Herd Brucellosis Status and the Occurrence
of Clinical Signs and Symptoms Typical of Acute and Chronic Brucellosis
in the 100 Herders Surveyed in the Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia, in August, 2000. N=100;
16% “Yes” for Typical Acute Brucellosis; 20% “Yes” for Typical Chronic
Brucellosis ................................................................................................ 149

Univariate Analysis of Potential Human Behavior and Occupation-Related
Risk Factors and the Occurrence of Clinical Signs and Symptoms Typical
of Acute and Chronic Brucellosis in the 100 Individuals Surveyed in the
Khangai Mountain Region of Bayankhongor Aimag (Province), Mongolia,
in August, 2000. N=100; 16% “Yes” for Typical Acute Brucellosis; 20%
“Yes” for Typical Chronic Brucellosis .................................................... 150

Univariate Analysis (T-Test) of the Association Between Herder Age and
the Occurrence of Clinical Signs and Symptoms Typical of Acute and
Chronic Brucellosis in the 100 Individuals Surveyed in the Khangai
Mountain Region of Bayankhongor Aimag (Province), Mongolia, in
August, 2000 ............................................................................................ 153

Univariate Analysis of the Association Between Gender and the
Occurrence of Clinical Signs and Symptoms Typical of Acute and Chronic
Brucellosis in the 100 Individuals Surveyed in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia,

in August, 2000 ........................................................................................ 154

xii

Table 4.8.

Table 4.9.

Table 4.10.

Table 4.11.

Table 4.12.

Univariate Analysis of Interviewed Individuals’ Understanding of the
Routes of Transmission and Ways to Prevent Brucellosis and the
Occurrence of Clinical Signs and Symptoms Typical of Acute and Chronic
Brucellosis in the 100 Individuals Surveyed in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia, in August, 2000.
N=100; 16% “Yes” for Typical Acute Brucellosis; 20% “Yes” for Typical
Chronic Brucellosis .................................................................................. 155

Association of Herder Knowledge of Routes of Brucellosis Transmission
and Brucellosis Prevention and the Practice of Behaviors or Risk Factors
Associated with Clinical Signs of Brucellosis Infection in Surveyed
Individuals in the Khangai Mountain Region of Bayankhongor, Mongolia,
Interviewed in August of 2000 ................................................................ 156

Final Reduced Logistic Regression Model of Risk Factors for Typical
Signs of General (Acute and/or Chronic) Brucellosis Reported in
Livestock Herders Interviewed in the Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia, in 2000 ............................. 157

Final Reduced Logistic Regression Model of Risk Factors for Typical
Signs of Acute Brucellosis Reported in Livestock Herders Interviewed in
the Khangai Mountain Region of Bayankhongor Aimag (Province),
Mongolia, in 2000 ......... ~ ........................................................................... 158

F inal Reduced Logistic Regression Model of Risk Factors for Typical
Signs of Chronic Brucellosis Reported in Livestock Herders Interviewed
in the Khangai Mountain Region of Bayankhongor Aimag (Province),
Mongolia, in 2000...........; ........................................................................ 159

xiii

Figure 1.0.

Figure 1.1.

Figure 2.1.

Figure 2.2.

LIST OF FIGURES

Appendix A. Map of 19 Aimags (Provinces) and 2 Cities (Ulaanbaatar
and Darkhan-Uul of Mongolia. Inset: Mongolia in Central Asia ........... 168

The Aimags (Provinces) of Mongolia for which Historic Brucellosis
Prevalence Data are Available for Livestock and Humans ....................... 16

The Study Region. Inset: The Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia. Below: The Five
Contiguous Sourns (Districts) of Gurvanbulag, Zag, J argalant, Galuut,
Erdentsogt and the Bagh (Sub-District) of Shargaljuut ............................. 48

Map of the Khangai Mountain Region of Bayankhongor Aimag
(Province), Mongolia. Labeled points indicate the Location (Global
Positioning System) and Brucellosis Status (squares = positive herds;
circles = negative herds) of Each Cattle Herd Selected for Testing .......... 62

xiv

INTRODUCTION

MONGOLIA: A BRIEF HISTORY

Mongolia is located in North Central Asia, bordered by the Russian Federation to
the north and the Republic of China to the south (See Figure 1). Mongolia is the 5th
largest country in Asia in land area with one of the lowest population densities in the
world. The nation’s human population of 2.4 million is distributed across an area of 1.56
million km2 (1.54 persons/kmz). Approximately 1/3 of Mongolians (750,000) live in
Ulaanbaatar, the national capital. Another 1/3 of Mongolians live in smaller cities and
towns, and 1/3 of Mongolians are nomadic, moving seasonally in search of pasture for
their livestock (Mongolia, 20010).

Present day Mongolia, formerly “Outer Mongolia”, gained its independence from
China in 1921. By 1924, after a series of power struggles involving Mongolia’s Buddhist
religious leaders, Chinese military commanders, the White Russians and the Russian
Communists, the Mongolian People’s Republic was founded as the world’s second
communist state. Mongolia continued to be closely aligned with Russia and the former
Soviet Union until the late 1980’s. In 1990, Mongolia began its latest economic and
political transition with the democratization of the political system and a change from a
centrally controlled economy to a market-driven system (Rossabi, 1999).

Extensive livestock agriculture has been the backbone of the region’s economy,
shaping its nomadic cultural tradition, since the 13th century when the Mongols
established the largest contiguous land empire the world has ever known under the

leadership of Chinggis Khan. Present day Mongols are predominantly (85%) of the

Khalkha ethnic group. Khalkha Mongolian is the language spoken by 90% of the
population. The adult literacy rate in Mongolia is over 90% and the predominant religion
(96%) is Tibetan Buddhist Lamaism (CIA, 2004).

The region’s climate, characterized by extreme temperatures with short growing
seasons, has forced Mongolian herders to migrate ﬁ‘equently (2-10 times/year) in search
of sufﬁcient pasture, water and minerals for their livestock for centuries (Rossabi, 1999).
The grassland steppe in central Mongolia continues to support the majority of the
region’s present day population of >30 million head of livestock. Although this system
of migratory, extensive, livestock agriculture has been in place for centuries, the
organizational structure of herds and the system of livestock ownership has experienced
changes that reﬂect the political and economic transitions of Mongolia’s contemporary
history.

Before 1924, secular or religious nobles controlled pasture use in Mongolia and
the seasonal migrations of herders and the livestock they tended. Between 1924 and
1990, pasture use and livestock management was controlled by Mongolia’s Soviet-
inﬂuenced socialist government. By 1960, all livestock herders in Mongolia had joined
livestock cooperatives and they received a salary for herding state-owned livestock. The
collective administration directed livestock migrations and subsidized transportation, the
building of wells and veterinary care. Democratic elections in 1990, and the
liberalization of the pastoral economy, led to the dismantling of herder cooperatives in
1992, and the privatization of the majority of the state owned livestock. Pastureland

(81% of Mongolia’s land area) remains state-owned until this day, and privately-owned

livestock are herded in common pastures deﬁned by the same districts (soums) and sub-
districts (baghs) used during the collective era (Femandez-Girnenez, 2000).

Today roughly 1/3 of Mongolians are nomadic pastoralists, moving seasonally
with their livestock. Half the nation’s population relies, either directly or indirectly, on
the pastoral economy, and the livestock sector accounts for 36% of Mongolia’s gross
domestic product. These numbers actually represent an increase in both the numbers of
individuals who identify “livestock herding” as their primary employment activity (186%
rise since 1990), and those living as nomadic pastoralists. Between 1990, and 2000, the
number of households herding livestock in Mongolia increased by 156% (74,710 to
191,526). This increase was seen not only in the absolute numbers but also in the
relative percentage of households involved in livestock herding, which rose from 16.6%
in 1990, to 34.6% in 2000. The increased reliance on livestock agriculture at the
household level, and the liberalization of the pastoral economy, led to an increase in the
number of livestock in Mongolia from 25.9 million in 1990, to a record high of 33.6
million in 1999, and then down to 30.2 million in 2000, as a result of losses during the
dzud (drought and severe snow storms) of 1999/2000 (Mongolia, 2001c).

The dramatic changes of the 1990’s have had a signiﬁcant impact on veterinary
care delivery in Mongolia. The level of State support for the livestock health sector
began to decline in the 1980’s as the large scale Soviet subsidies (once 1/3 of Mongolia’s
G.D.P.) dried up. Government sponsored livestock vaccination campaigns, including
efforts to control zoonotic disease such as brucellosis, were signiﬁcantly downscaled. In
addition, in 1999, soum (district) veterinary clinics and their staff were privatized. This

forced an overnight transition from government provided veterinary care to a fee for

service veterinary care delivery system. The impacts of these drastic changes in the
provision of veterinary care delivery on livestock health and productivity have not been
documented.

A number of authors have raised concerns regarding the impacts of the recent
political and economic transition in Mongolia on both human (F oggin et al., 1997;
Purevdawa et al., 1997) and livestock health (Roth et al., 2003) due to the changes in the
health care delivery system which have occurred over the past 10-15 years. The study
presented in this thesis examines the epidemiology of one disease, brucellosis, in a single
region of Mongolia, the Khangai mountain region of Bayankhongor Aimag (Province),
Mongolia. Brucellosis is a good example of a disease of economic and public health
importance that would be affected by a breakdown in national veterinary and human
health care provision. Additionally, Bayankhongor Aimag (Province) has experienced
the same trends in human population shifts, increasing livestock numbers, and the

transformation of the veterinary care delivery system seen across Mongolia.

BAYANKHONGOR AIMAG (PROVINCE)

Bayankhongor Aimag (Province) is one of 21 Provinces that make up Mongolia
(See Figure 1). Located in the southwestern portion of the country, Bayankhongor covers
an area of 116,000 kmz. It is home to 85,300 people and 2.4 million head of livestock.
The Khangai mountain range traverses the northern third of Bayankhongor where the
majority of the Province’s yaks are herded. Bayankhongor contains all three of the major
steppe grassland ecological zones found in Mongolia. The northern 1/3 of Bayankhongor

is in the “mountain steppe” ecological zone, the central 1/3 in the “steppe” ecological

zone, and the southern 1/3 of the Province, that stretches south to the Gobi desert and the
Chinese border, is comprised of “desert steppe” (F emandez-Gimenez, 2000).

Bayankhongor Aimag is divided into 19 administrative districts or soums. Each
soum contains approximately 4 sub-districts, called baghs. The soum and bagh system
was set up across the country during the collective era in Mongolia to direct grazing
patterns. Today, the majority of Bayankhongor’s nomadic population migrates within a
particular soum and generally rotates pastures within a single bagh. The study described
in this thesis includes the 5 northern Khangai mountain region soums of Gurvanbulag,
Zag, Jargalant, Galuut, and Erdentsogt, and the sub-district or bagh of Shargaljuut.

As in all Provinces outside the “city-Provinces” of Ulaanbaatar, Darkhan-

Uul, and Orkhon, livestock herding is the predominant occupation and means of
livelihood in Bayankhongor. The Province has experienced the same increase in
numbers of livestock (1.6 million to 2.4 million), individuals involved in herding (149%
increase), households herding livestock (176% increase) and overall percentage of
households involved in herding (33.6% to 68.2%), documented in Mongolia as a whole
from 1990, to 2000 (Mongolia, 2001b).

Veterinary service delivery, at the soum or district level, in Bayankhongor was
privatized in 1999, as it was across Mongolia. A state—sponsored veterinary laboratory
and service center still exists in the Province center where a small staff collects records of
livestock disease outbreaks in the Province. The facility is only semi-operational and
very few diagnostic tests are performed. The Province veterinary center does serve as a
distribution point for available livestock vaccines and veterinary products that are now

purchased and administered by private veterinarians. Since 1990, state-sponsored

preventive veterinary medicine has been signiﬁcantly scaled back in Bayankhongor, as it
has been across the country. As of 2002, the last year livestock were vaccinated for

brucellosis in Bayankhongor was 1988.

STUDY RATIONALE

Brucellosis is an economically important agricultural disease and a signiﬁcant
public health concern in many regions of the world (Halling and Boyle, 2002) (Nicoletti,
1980b). Measuring the burden of disease, and understanding the epidemiology of
brucellosis in a region, are essential components of informed livestock disease and public
health policy.

Brucellosis is reported to be an endemic disease in Mongolia, however, neither
the extent of brucellosis distribution, nor the economic and public health impact of the
disease is known (OIE, 2002b). A number of authors have suggested that the rapid
political changes of the early 1990’s, and the subsequent economic crisis, severely
disrupted the national health care system in Mongolia leading to a rise in the incidence of
multiple infectious diseases (Foggin et al., 1997; Purevdawa et al., 1997). Health
ofﬁcials in Mongolia today have highlighted brucellosis, along with hepatitis (B & C)
and tuberculosis, as one of their top three infectious disease public health priorities
(Ebright et al., 2003; Mikolon, 2000). The veterinary community has responded to these
concerns with the reinstitution of a livestock brucellosis vaccination program which was
began in 2001.

This study sets out to collect data on the epidemiology of livestock and human

brucellosis in the Khangai mountain region of Bayankhongor Aimag (Province),

Mongolia. Bayankhongor Aimag was selected due to the predominance of yak and cattle
herding in the region, and anecdotal reports of brucellosis-like disease in the area. The
type of information collected in this study will be essential in establishing an
understanding of the current state of brucellosis in the region, and will form a baseline for
future evaluation of local and national disease control programs.

Extensive livestock agriculture has proven to be the economic “safety net”
through multiple periods of economic and political transition in Mongolia. The
maintenance of a healthy national livestock herd is important not only to the ‘/4 of
Mongolian households who herd the country’s 30 million head of livestock for a living,
but also for the entire nation who rely heavily upon the grassland steppe and the livestock

it supports for their health and well being.

PROBLEM STATEMENT

The epidemiology of brucellosis in Mongolia today, both the distribution and
prevalence of disease in livestock, and the burden of disease in humans, is largely
unknown. Brucellosis control in Mongolia, in the form of comprehensive livestock
vaccination and systematic surveillance, ceased to exist in the late 1980’s as a result of
the abrupt loss of economic aid from the former Soviet Union. Brucellosis surveillance
in Mongolia from 1990, until 2001, has consisted of low-level and sporadic herd level
testing of bovine, caprine, and ovine herds. Data on the prevalence and incidence of
brucellosis in humans is based primarily on hospital case reporting. Published

information on brucellosis in Mongolia after 1990, consists primarily of summaries of

reports made by Mongolian government ofﬁcials to the Ofﬁce International des
Epizooties or the World Health Organization.

Well-designed population-based studies that assess current levels of brucellosis
infection in livestock and humans in Mongolia, and/or identify risk factors associated
with disease, do not exist. Recent publications have identiﬁed brucellosis as an important
public health concern in Mongolia, (Ebright et al., 2003) (Roth et al., 2003; Zinsstag et
al., submitted 2004), but little to no disease data is presented. The collection of
comprehensive and robust brucellosis data should be an essential component of
Mongolia’s new national brucellosis control program. Systematic disease surveillance
will enable the veterinary and public health community to evaluate and improve the

effectiveness of the recently implemented livestock vaccination program in Mongolia.

OBJECTIVES

This thesis, therefore, focuses on the epidemiology-of brucellosis in livestock and
humans in the Khangai mountain region of Bayankhongor Aimag (Province), Mongolia,
where cattle (primarily yaks) are the predominant livestock species. The speciﬁc
objectives of this study were to:

1) Determine the prevalence of brucellosis-seropositive cattle herds in the Khangai
mountain region of Bayankhongor Aimag (Province), Mongolia

2) Identify the herd management risk factors for brucellosis-seropositive cattle herds
in the region

3) Identify clinical signs or changes in production measures associated with the

brucellosis-seropositive status of cattle herds in the region

4) Determine the prevalence of clinical signs and symptoms suggestive of
brucellosis in the human (livestock herder) population in the region

5) Identify livestock product consumption and/or livestock management practice risk
factors for the occurrence of clinical signs and symptoms suggestive of

brucellosis in the herder population

OVERVIEW

Each chapter in this thesis has an abstract, introduction, hypotheses, methods,
results and discussion, with the exception of Chapter 1, which is a literature review and
does not have a hypotheses or methods. Chapter 1 is a review of the literature published
in internationally available journals on brucellosis in Mongolia dating back to the 1960’s.
The material covered includes studies and accounts published in the Russian and German
language as well as reports published in the English language. The literature review
provides a historical account of brucellosis epidemiology, research and control in
Mongolia, describes current brucellosis surveillance and control programs, and identiﬁes
areas for ﬁrrther research. Chapter 2 describes the prevalence of brucellosis-seropositive
cattle herds in the study area and provides a detailed description of the sampling and
testing methods used. Chapter 3 uses the brucellosis prevalence data presented in
Chapter 2 and identiﬁes livestock management practices that are risk factors for cattle
herd brucellosis-seropositive status in the study area. Chapter 3 also describes clinical
signs of brucellosis in the cattle population and identiﬁes changes in production
parameters related to disease. Chapter 4 describes the prevalence of clinical signs and

symptoms suggestive of brucellosis in the human herder population. Risk factors for the

occurrence of brucellosis-like disease in humans are identiﬁed. The human disease
occurrence data are also analyzed for relationships with cattle herd brucellosis
seroprevalence data presented in Chapter 2. The thesis includes an overall summary and

recommendations for ﬁrture research.

10

Chapter 1

BRUCELLOSIS IN MONGOLIA (1950-2002): A REVIEW

1.1 INTRODUCTION

Brucellosis is a zoonotic disease with a global distribution caused by bacteria of
the genus Brucella (Corbel, 1989). Major efforts have been undertaken around the world
to control brucellosis due the signiﬁcant economic losses associated with the occurrence
of the disease in domestic livestock and the oﬁen debilitating disease associated with
human infection (Halling and Boyle, 2002; Reddy, 2003). Clinical signs of brucellosis in
domestic livestock include late-term abortions, decreased milk production and lowered
fertility in both males and females (Drost and Thomas, 1996). The disease in humans is
associated with a number of nonspeciﬁc complaints including irregular or intermittent
fevers, night sweats and headaches. Untreated brucellosis in humans can lead to chronic
osteo-articular complications manifesting as joint pain, joint effusion and debilitating
arthritis (Benenson, 1995; Black, 2004a; Young, 1989). Particular biovars of Brucella
melitensis are associated with pathogenicity in different domestic animals. B. melitensis
abortus is the biovar most commonly associated with brucellosis diseasein cattl (Reddy,
2003). The vast majority of human infection is the result of direct or indirect
transmission of the disease from infected livestock to humans through the ingestion of
contaminated milk or dairy products or through direct contact with aborted material or

uterine ﬂuid from infected animals (Nicoletti, 1989; Shapiro, 2004; Slack, 2004a).

11

Brucellosis is classiﬁed as a “List B” disease by the Oﬂice International des
Epizooties (O.I.E.). List B diseases are “transmissible diseases that are considered to be
of socio-economic and/or public health importance within countries and that are
signiﬁcant in the international trade of animals and animal products” (OIE, 2003).
Livestock agriculture accounts for a signiﬁcant portion of Mongolia’s economy
producing 1/3 of the country’s gross domestic product (GDP) and employing over half
its working-age population (Mongolia, 2001b; Mongolia, 2001c). In addition, the
livelihoods of over half of Mongolia’s population are linked either directly or indirectly
to the livestock sector (Femandez-Gimenez, 2000). The econmic and public health
impacts of brucellosis on a country so heavily reliant on livestock agriculture are
signiﬁcant.

Data on the historical occurrence of brucellosis in Mongolia is limited to a series
of articles published in the international literature over the last 5 decades that describe a
number of population based epidemiological studies, vaccine ﬁeld trials and the
implementation of brucellosis control programs in the region (Beulig et al., 1969; J ezek
et al., 1972; Jezek et al., 1974; Otgon, 1968; Pinigin et al., 1968a; Pinigin et al., 1968b;
Tserendash, 1980, 1984; Tserendash and Shumilov, 1970; Zherikova et al., 1972).
Published data on the current status of brucellosis infection in the region is also limited
but a number of recent publications indicate that brucellosis occurrence and control is
still a major concern (Ebright et al., 2003; Erdenebaatar et al., 2002; Roth et al., 2003;
Zinsstag et al., submitted 2004). The publish studies addressing brucellosis in Mongolia

are described below and summarized in Table 1.1.

12

The objectives of this literature review were to:

1. Explore the historical data available on the epidemiology of brucellosis in
Mongolia, drawing primarily on material available in the non-English veterinary
literature published between 1960 and 1990.

2. Identify and discuss reports and sources of data on the current status of brucellosis
in Mongolia.

3. Highlight information gaps and areas for further research that would facilitate the
collection of meaningful data and broaden our understanding of the epidemiology

of brucellosis in Mongolia.

1.2 BRUCELLOSIS RESEARCH IN MONGOLIA

A signiﬁcant effort to study brucellosis in Mongolia has been in place since the
early 1960’s. Mongolian veterinarians, physicians, scientiﬁc researchers, and
government ofﬁcials have collected information on the occurrence of brucellosis in
domestic livestock and humans for decades. The availability of that information to the
international scientiﬁc community, however, has been heavily inﬂuenced by the political
climate of the times. It was, and still is, primarily through collaborative research efforts
with international scientists that Mongolians have the opportunity to submit their work
for publication in international journals. The availability and location of information on
brucellosis in Mongolia in the international literature, therefore, reﬂects patterns of
international aid, and economic and political agreements between Mongolia and her

political allies over the decades.

13

Publication of reports on studies of brucellosis in Mongolia conducted from the 1960’s
through the 1980’s were the results of collaborative efforts between Mongolian and
international scientists from the former Soviet Union and Eastern Europe. A few studies
were published by Mongolian scientists in the Russian veterinary literature (Otgon, 1968;
Tserendash, 1980, 1984). Other studies were published as co-authored manuscripts with
scientists from the former Soviet Union (Pinigin et al., 1968a; Pinigin et al., 1968b;
Tserendash and Shumilov, 1970; Zherikova et al., 1972) former Czechoslovakia (Jezek et
al., 1972; J ezek et al., 1974), Hungary (Denes, 1997) and the former German Democratic
Republic (Beulig et al., 1969).

More recent collaborative efforts in the ﬁeld of brucellosis research in Mongolia
have been between Mongolian veterinary specialists and scientists from Japan,
Switzerland, South Korea, Germany and the United States. Literature published on
brucellosis in Mongolia since 1990, includes an account of a veterinary assistance project
administered by the Christian Veterinary Mission (JAVMA, 1998), a research note on the
use of Agar Gel Immunodiffusion (AGID) testing to differentiate vaccinated from
infected animals in Mongolia (Erdenebaatar et al., 2002), a study funded by the World
Health Organization, Food and Agriculture Organization, Swiss Tropical Institute, Swiss
National Science Foundation and Swiss Agency for Development and Cooperation (Roth
et al., 2003) and an article on emerging infectious disease in Mongolia that highlights

brucellosis as a disease of importance (Ebright et al., 2003).

14

1.3 BRUCELLOSIS SURVEILLANCE IN MONGOLIA

The presence of brucellosis in Mongolia was ﬁrst reported in the international
scientiﬁc literature in the late 1960’s and early 1970’s. Initial reports described the
detection and identiﬁcation of Brucella infection in yaks. These reports were followed
by a series of epidemiologic studies carried out by the Mongolian National Veterinary
Organization with assistance ﬁ'om the German Democratic Republic, Hungary, and the
former Soviet Union. The early studies examined the distribution, prevalence, and
incidence of brucellosis in the livestock and human population in various regions of
Mongolia. Figure 1 indicates the speciﬁc aimags (provinces) and species for which

historic brucellosis prevalence data are available.

15

Figure 1.1. The Aimags (Provinces) of Mongolia for which Historic Brucellosis
Prevalence Data are Available for Livestock and Humans.

 

Key: List of Aimags (Provinces) and Cities (Ulaanbaatar and Darkhan-Uul)

1. Bayan-Olgi 8. Bayankhongor 15. Ulaanbaatar (Capital)
2. Uvs 9. Bulgan 16. Dundgovi

3. Khovd 10. Ovorkhangai 17. Khentii

4. Zavkhan 11.0mnogovi 18. Govisurnber

5. Govi-Altai 12. Selenge 19. Domogovi

6. Khovsgol 13.Darkhan-Uul 20. Domod

7. Arkhangai 14. Tuv 21. Sukhbaatar

A. Brucellosis Testing

A number of internationally accepted methods of testing were used in the
reviewed studies to diagnosis brucellosis in livestock and humans in Mongolia. The
methods described included the following:

1. Allergic Sl_dn Test: Also called the “Huddleson” reaction or the “Burnet” test.
An intradermal (0.1 ml) or subcutaneous (0.5 ml) injection of brucellin allergen
(an extract prepared from Brucella) is followed by injection site examination at
48-72 hours. A delayed hypersensitivity skin reaction develops speciﬁc to the
genus Brucella in animals or humans sensitized by infection or recent vaccination.
In livestock the injection site (usually the caudal tail fold) is observed and
palpated for detect swelling. A 6-level scale for classifying the reaction in
humans is used which ranges ﬁ'om a “++++ = red, indurated, grossly edematous
lesion with necrosis, or a tendency to necrosis, at the center” to “0=absence of any
lesion” (Alton and Jones, 1967).

2. Serum Tube Agglutin_ation Test (STAT): The STAT is used to test for the
presence of antibodies in the subject’s serum produced in response to brucellosis
infection. Brucella antigen is usually added to 5 narrow tubes containing 1:5,
1:10, 1:20, etc. of the serum under test and phenol saline. The degree of
agglutination in the tubes is read after a 24-hour incubation at 37°C. The highest
serum dilution showing 50% or more agglutination is recorded as the end-point
(Alton and Jones, 1967).

3. Complement Fixation Test (CF'D: The CFT is a serologic test designed to

determine the Brucella titer of the serum under test. The CFT utilizes Brucella

17

antigen, complement and sheep red blood cells to measure the degree of
hemolysis. The titer of the serum under test is read as the highest dilution
showing 50% or less hemolysis. In cattle a titer of 1:40 indicates an infected
animal and one of 1:20 is suspicious, indicating a need to retest in 2 weeks (Alton
and Jones, 1967).

. The Rose BengzﬂTestﬂlBT): A buffered Brucella antigen test used on serum.
Standardized plates are used to observe the reaction between the subject’s serum
and the added standardized Brucella antigen. After a short period of rocking the
plates are read and judged positive or negative based on the degree of
agglutination (0==no agglutination to 3=coarse clumping, deﬁnite clearing) (Alton

etaL,l988)

B. Livestock Brucellosis

The ﬁrst international report describing the epidemiology of brucellosis in

Mongolia was published in 1968 (Otgon, 1968). The author gives a brief history of

brucellosis detection in Mongolia including the ﬁrst laboratory conﬁrmed case of

brucellosis diagnosed in a cow from Selenge Aimag (a Mongolian Province bordering

Russia) by a State laboratory in Songino, Mongolia, in 1932. In 1942, veterinary workers

diagnosed brucellosis in the large “Orkhon” State sheep farm also located in Selenge

Aimag and by 1951, brucellosis infection was conﬁrmed in a number of State farms in

several Aimags (Provinces) of Mongolia. Livestock were imported from both China and

Russia during this time but the exact origin of brucellosis infection in Mongolia in

unknown.

18

A national survey of livestock conducted by Otgon and the national veterinary
organization in 1956, revealed that brucellosis infection was widely distributed across the
country. The highly sensitive but less speciﬁc, allergic skin testing methods were used to
identify infected livestock. The proportion of livestock with positive reactions was
higher in the northern region of Mongolia (prevalence in cattle =4.7%; sheep and
goats=3.7%) than in the south (cattle=2.l%; sheep and goats=0.02%). The author notes
that there was a higher density of livestock in the northern regions of Mongolia (65% of
the total population) and that livestock in the north where often clustered in and around
open grazing areas. State farms were also more prevalent in the north where livestock
were corralled during the winter months and provided harvested forage. Distribution of
breeding stock from State farms in the north to smaller operations around the country also
likely contributed to the spread of brucellosis in Mongolia (Otgon, 1968).

An article published in the German veterinary literature in 1969, describes
brucellosis testing that was performed in Selenge Aimag (Province) as a collaborative
effort between Mongolian veterinarians and veterinary specialists ﬁom the German
Democratic Republic. The article provides details of the types of animals tested and the
methods used for testing in Selenge. Cattle older than l-year and rams were tested for
brucellosis using the serum tube agglutination test (STAT) and complement ﬁxation
(CFT) test. Goats and camels were tested using STAT and sheep, other than rams, were
tested with allergic skin tests. Brucellosis prevalence data is not presented but it was
noted that positive animals were identiﬁed and extensive training on diagnostic
techniques for brucellosis was provided to Mongolian veterinary specialists (Beulig et al.,

1969).

19

More extensive data on livestock brucellosis in Mongolia are provided by two
reports by Pinigin et al., 1968 (Pinigin et al., 1968a; Pinigin et al., 1968b). The ﬁrst
report describes the typing of Brucella strains isolated from a number of yak raising
regions in Central Asia and Siberia, including Mongolia. The second report contains
seroprevalence data on yaks that originated in Mongolia and were imported into the
Autonomous Soviet Socialist Republic (ASSR) of Tuva in 1961, as well as data on yaks
from Mongolia delivered to two slaughterhouses in the Irkutsk region of Russian and the
ASSR of Buriatia in 1962.

The ﬁrst report by Pinigin, et a1. 1968, describes the examination of 20 Brucella
cultures, 17 isolated from yaks (aborted fetuses, internals organs and lymph nodes), 2
isolated from dogs and 1 isolated ﬁ'om a raven in the yak brucellosis “epicenters”. One
of the cultures described was obtained in Mongolia and the others were from the two
former ASSRs of Tuva and Buriatia that border Mongolia to the north. The methods
used to type the Brucella cultures were those approved by the Brucella taxonomy
subcommittee and united expert committee of the Food and Agriculture Organization
(FAO)/World Health Organization (WHO) on brucellosis. The morphological,
serological and biochemical properties of each of the isolates were largely identical.
They were all classiﬁed as Brucella abortus, biotype 3 (Pinigin et al., 1968b).

In a second article Pinigin et al., 1968, present slaughterhouse data from the
testing of yaks of Mongolian origin in 1961 and 1962. The Mongolian yaks tested in
1961 were imported into the ASSR of Tuva. Serum tube agglutination testing (1:100 =
positive) of 121 animals indicated that 44 or 37% were strongly positive. The same

serum was tested with the Huddleson (allergic skin test) reaction and produced 109

20

positives, a prevalence of 90%. The author dismisses the Huddleson reaction results due
to what was interpreted as an unacceptable number of false positive results. The STAT,
however, is less sensitive and more speciﬁc than the allergic skin testing methods so the
discrepancy between the results using the different tests (STAT=37% positive and
Huddleson=90% positive) would be expected.

A second group of yaks from Mongolia were tested at two slaughterhouses in the
Irkutsk region of Siberia in 1962. In this study the STAT and the complement ﬁxation
test (CFT) were used. The CFT is often used as a conﬁrmatory test because it is
generally more speciﬁc than the STAT. Of 1,404 animals tested with the STAT at one
slaughterhouse, 308 (22%) were positive. The CFT was run on 1,378 of the same group
of animals yielding 151 positives or a CFT seroprevalence of 11%. Testing preformed at
the second Irkutsk slaughterhouse involved 394 Mongolian yaks. Of this group 254 of
394 yaks (64%) were classiﬁed as positive according to the STAT and 67 of 370 yaks
(18%) were recorded as positive based on the CFT (Pinigin et al., 1968b). Although the
exact origin of the Mongolian yaks is unknown, these initial reports conﬁrm that
brucellosis was present in the yak population of Mongolia by the early 1960’s. The
serologic evidence of bmcellosis in this population was supported by the reported
isolation of Brucella abortus from aborted material collected from Mongolian yaks
delivered to the ASSR of Buriatia (Pinigin et al., 1968b).

Brucellosis in camels (Camelus bactrianus) appears to have been widespread in
Mongolia in the late 1960’s, with ser0positive animals detected in each of the 8 Aimags
(Provinces) surveyed (See Table 1.1). The relationship between the reported results of

the testing of camels with the STAT vs. the CFT is contrary to what would be expected.

21

The data indicate that the CFT identiﬁed 4 times as many brucellosis-seropositive camels
as the STAT when the opposite result is usually found when the two tests are used in
other species. An explanation for this unexpected result is not given but Tserendash, et
al., suggest the use of both tests in combination for seroprevalence studies in camel
populations. Clinical signs of brucellosis in the tested camels or attempts to isolate
Brucella were not reported. A record of the isolation of Brucella abortus from camels in
Mongolia was, however, found in other sources (Crawford et al., 1990).

C. Human Brucellosis

Mongolia is a country with a history of a signiﬁcant level of brucellosis in both its
livestock and human populations (Corbel, 1989; Joshi, 1991). Publications summarizing
the distribution of brucellosis in the region imply that the burden of disease was similar to
the pattern seen in the former Soviet Union where the incidence of brucellosis in humans
was widespread by the 1930’s and only began to decrease once livestock vaccination was
instituted in the 1950’s (Kolar, 1989). The collection of data on the epidemiology of
human brucellosis in Mongolia began in the 1950’s. Livestock vaccination did not begin
until the 1970’s as described in section 1.4. Although information speciﬁc to the
epidemiology of human brucellosis in Mongolia is limited, the articles described below
provide evidence of a signiﬁcant burden of disease.

Medical clinics in Ulaanbaatar, the capital of Mongolia, have records of patients
with brucellosis dating back to 1954 (Otgon, 1968). The ﬁrst published population-based
study of human brucellosis in Mongolia was conducted by Otgon in 1956, and involved
the screening of 1,516 people. The population included slaughterhouse workers from

Ulaanbaatar and livestock workers from 6 State farms in 6 Aimags (Provinces). The

22

individuals were screened for brucellosis, using allergic tests (Bumet and Huddleson).
Thirty-four percent of the individuals tested had positive reactions to the test, indicating
previous exposure to Brucella. Children as young as 6 showed positive reactions and
both males and females were equally affected. On State farms where the prevalence of
brucellosis infection in livestock was 3.7-4.7%, the prevalence of positive reactions in
farms workers was 30.3%. Six percent of the individuals with positive reactions had
clinical signs of brucellosis (Otgon, 1968). The observation that humans suffered from
both mild and severe clinical disease and the fact that both small ruminants and cattle
showed evidence of brucellosis infection, led the author to suggest that both Brucella
melitensis and Br. abortus were present in Mongolia; however, isolation and typing of the
agent was not performed. The study does establish the presence of signiﬁcant brucellosis
exposure in these two high-risk populations of farm and slaughterhouse workers.

Two studies published in the early 1970’s represent the most complete data
published on human brucellosis in Mongolia before disease control through livestock
vaccination was instituted in the 1970’s (J ezek et al., 1972; J ezek et al., 1974). The ﬁrst
study revealed that brucellosis was a serious public health concern in Mongolia,
documenting evidence of Brucella antibodies in all age groups, in both sexes and across
occupations in all regions of the country surveyed including the urban population. The
ﬁrst study consisted of a serologic survey of close to 5,000 individuals aged <4-60+ from
the ﬁve Aimags or Provinces of Uvs, Govi-Altai, Khovsgol, Domogovii, Domod, and the
national capital, Ulaanbaatar (See Figure 1). A multistage sampling method was used to
select the clusters of 60-70 individuals. Approximately 500 individuals were screened in

every age group (< 4, 5-9, 10-14, 15-19, 20-29, 30-39, 40-49, 50-59, 60+) with the

23

exception of the 60+ age group that only contained 226 individuals. An additional 758
individuals with potentially high occupational risks for brucellosis infection were targeted
and screened. The serum was tested with both the STAT and CFT. The prevalence of
brucellosis-ser0positive individuals of all age groups combined was higher in the rural
population (4.4%) than it was in the urban population (1.7%). In both populations the
prevalence increased with age with individuals in the 50—59 age group from the rural
population having the highest seroprevalence (8.1%). There is also evidence of early
childhood exposure in both the urban (1.1% prevalence in < 9 age group) and rural (2.4%
prevalence in < 9 age group) populations. Individuals who identiﬁed themselves as
shepherds (herding sheep and goats) had the highest prevalence of brucellosis
seropositivity (9.9%).

A second study, designed to determine the incidence of brucellosis in
Cenkhermandal soum (district) in Khentii Aimag (Province), supports the classiﬁcation
of brucellosis in Mongolia as a public health priority (J ezek et al., 1974). At the
beginning of the study, serum samples were collected from over 1,000 individuals (ages 3
to 60+) of the 2,000 nomadic and settled Mongolians who made up the population of the
Cenkhermandal soum (district). The serum was tested with both the STAT and the CFT.
The initial overall seroprevalence of the Cenkhermandal cohort was 11.7% with a level of
10.2% in the settled and 13.5% in the nomadic population; similar to the national
prevalence of brucellosis seropositivity reported in the ﬁrst study (J ezek et al., 1972).
However, prevalence levels higher than the national average (6.3% and 13.7%) were seen
in both the settled and nomadic children ages <4-9 years. There were no differences in

prevalence between the sexes and prevalence increased slightly with age. A higher

24

prevalence (15.4%) was seen in individuals who classiﬁed themselves as agricultural
workers vs. 10.7% prevalence in non-agricultural workers. The allergic skin test was also
performed 989 of the 1164 individuals tested serologically. The overall prevalence of a
positive skin reaction (raised edematous plaques of 10 mm or larger accompanied by
erythema) was 21.4%.

The second examination of the Cenkhermandal soum cohort occurred 6 months
later, which was 6-8 weeks after the end of lambing/kidding season. Of the 1,042
individuals who tested negative during the ﬁrst examination, 901 were re—bled and their
serum was tested with the STAT and CFT. The percent of converters in the 6 month
period was 2.4% overall with 1.3% and 3.8% in the settled and nomadic population,
respectively. The highest proportion of new positives was in the 20-29 year-old age
category. The incidence of sero-conversion in the 20-29 year-olds in the nomadic
population was 12.8%.

Recent studies of human brucellosis in Mongolia are not available. However, the
World Health Organization (WHO) has indicated through various reports and
publications that brucellosis remains a signiﬁcant public health concern in Mongolia
today (Joshi, 1991; WHO, 1998, 2003). Additionally, brucellosis has been highlighted
as a re-emerging disease of public health importance in Mongolia by the WHO (WHO,
2001) and the Mongolian medical community (Ebright et al., 2003). The Mongolian
Statistical Yearbook reported 850 cases of brucellosis 1995, 1,482 in 1999, and 992 in
2000 (Mongolia, 2001d). The method of reporting, case deﬁnition, and testing
techniques used to identify these brucellosis cases is unknown. However, if they are new

cases, the national incidence rate of human brucellosis in 2000, was 4.13 cases per

25

10,000. This indicates a return to the levels of brucellosis reportedly present in Mongolia

(4.8 per 10,000) before livestock vaccination was begun in the 1970’s (Mikolon, 2000).

1.4 BRUCELLOSIS CONTROL IN MONGOLIA

A. Historical Account

Vaccination of livestock against brucellosis in Mongolia began in the 1960’s and
1970’s with the use of the Brucellosis abortus Strain-19 vaccine in cattle and the Br.
melitensis Rev-1 vaccine in small ruminants (Corbel, 1989; Tserendash, 1980). The
vaccination program was deemed a success by Mongolian scientists who reported that
there was a reduction in the number of livestock abortions recorded on farms
(Tserendash, 1980) and by WHO ofﬁcials who reported that the annual incidence of the
disease in humans dropped to less than 1 case/ 10,000 according to Roth and Zinsdag et
al., 2003. There are three scientiﬁc publications that evaluate the initial livestock
vaccination program in Mongolia and explore the use of both the Br. abortus Strain-19
and Br. melitensis Rev-1 vaccine to control the disease (Denes, 1997; Tserendash, 1980,
1984)

The use of the Br. abortus Strain-19 vaccine in cattle in Mongolia was effective in
controlling brucellosis but its use was associated with a short period of protection (not
longer than 1 year) and the persistence of vaccine induced antibodies to Brucella that
complicated subsequent testing efforts (T serendash, 1980). The use of the Br. melitensz's
Rev-l vaccine in cattle was investigated as an alternative to Strain-19 in hopes of
showing equal protection but with a longer period of immunity and less vaccine-related

false positive tests. This work, presented by Tserendash in a series of two articles

26

(Tserendash, 1980, 1984), provides data on a vaccine trial of Rev-l in cattle, use of the
Rev-1 vaccine in cattle in the ﬁeld, and accounts of the relationship between brucellosis
infection prevalence on farms and the percent of abortions recorded in the herd.

Tserendash compared the safety and effectiveness of the Rev-l and Strain-l9
vaccine for preventing brucellosis in cattle in a herd of 199 animals. He concluded that
the two vaccines had similar characteristics in terms of safety since neither induced
abortion in cattle when administered at 2-3 months of gestation. He also concluded that
the Rev-1 vaccine was equal in efﬁcacy to the Strain-19 vaccine because cows
immunized with the two vaccines had similar abortion and calving rates when challenged
with Brucella. In addition Brucella organisms were isolated ﬁ'om a similar percentageof
challenged cows vaccinated with Rev-1 and Strain-19. It was unclear which strain of
Brucella was used for the infection challenge though the use of both Br. abortus and Br.
melitensis was mentioned. Further studies were performed in yaks and cattle of varying
ages. Similar conclusions were drawn regarding safety and efﬁcacy and it was also noted
that post-vaccine serological reactions waned more quickly in calves vaccinated with
Rev-1 than those vaccination with Strain-l9 (Tserendash, 1980).

Despite the somewhat limited scope of the Rev-1 vaccine trial described above,
the use of Rev-1 in cattle and yaks in Mongolia was extended to non-experimental herds
in the mid 1970’s through the early 1980’s. Tserendash refers to the vaccination of over
100,000 head of cattle in multiple regions of the country. The use of the vaccine was
associated with both a drop in the percent of reactors in the herds studied (14.3% to 5.3%
after 2 years, to 0.2% after 3 years and to 0% after 4 years) and a parallel drop in the

percentage of abortions in the herds (from 8.5% to less than 1%). There was also a drop

27

in the percent of abortions attributed to brucellosis based on bacterial culture results
(T serendash, 1984). The reports by Tserendash also provide pre-vaccination
seroprevalence data from multiple regions of Mongolia included in table 1.1.

A Hungarian sponsored livestock vaccination program implemented in
collaboration with the Mongolian veterinary agencies to eradicate brucellosis from the
cattle herds of agricultural cooperatives, State farms and small household stock, began in
Mongolia in 1987, (Denes, 1997). The program consisted of brucellosis surveillance, the
removal and slaughter of infected animals, and the systematic annual vaccination of
healthy young cattle. The Hungarian project covered 7 Aimags in Mongolia including
Bulgan, Domod, Khentii, Selenge, Tov and 2 unlisted Aimags (Denes, 1997).

The Hungarian project was also concerned with the limited period of protection
offered by the Strain-19 vaccine and the complications associated with vaccine-related
false positive serological tests so opted to explore the use of Rev-1 in cattle. Citing
encouraging ﬁndings reported by Tserendash (Tserendash, 1980, 1984), on the superior
immunogenicity of Rev-l, Denes supported the large-scale use of the Brucella melitensis
Rev-l vaccine in cattle as an alternative to the Br. abortus Strain-l9 product. The
Hungarian vaccination protocol included the serological testing of all cattle, 20-25 days
apart, before vaccination was begun. The cattle were tested with the Rose Bengal test
(RBT), STAT and CFT. Brucellosis-ser0positive cattle were marked for slaughter and
healthy cattle were vaccinated. In the ﬁrst year of the program all female cattle over 3
months of age (more than 1A: a million head) were immunized with the Rev.1 vaccine. In
the second year heifer calves older than 3 months, yearlings not vaccinated the ﬁrst year,

heifers older than 2 years and cows were vaccinated. The vaccine was administered

28

subcutaneously at a dose of 5X109 viable organisms in 2 m1 of vaccine re-suspended in
physiological saline. In herds formally vaccinated with Strain-19, the use of Strain-l9
was continued using the same schedule as that outlined for Rev-1 above (Denes, 1997).

The prevalence of ser0positive cattle observed by Denes ranged from 3.8% to
35%, and in the 20-25 days between the two initial testing periods an average of 10%
more cattle sero-converted per herd (Denes, 1997). The combination of the high
prevalence of brucellosis seropositivity and active disease transmission in these herds
supported the institution of regular vaccination, and not a test and slaughter program, as
the most effective and feasible method for controlling brucellosis in this population of
cattle (Denes, 1997).

Denes followed the cattle herds vaccinated with Rev-1 as well as those vaccinated
with the Strain-19 vaccine. The researchers found that cattle vaccinated with the Rev-1
had higher rates of post-vaccination seroconversion (92.7% in heifers and 90% in cows)
than the cattle vaccinated with Strain-19 (68.8% in the heifers and 89.7% in the cows)
though the authors note that this higher degree of antigenicity does not necessarily reﬂect
a higher degree of protection from infection (Denes, 1997). Conversely, the positive
serologic status of cattle 1 and 2 years after vaccination was greater in cattle vaccinated
with Strain-19 (13.2% after 2 years) than in cattle vaccinated with Rev-1 (7% after 1
year). However, the presence of detectable antibodies to Brucella in proportion of the
cattle tested by Denes 1 and 2 years after vaccination are likely a result of a recent natural
Brucella challenge and not the result of previous vaccination.

Unfortunately a comprehensive evaluation of the livestock vaccination program

begun in Mongolia in 1987, in cooperation with Hungarian veterinary experts, was never

29

completed. The political and social changes that took place in Mongolia in the early
1990’s effectively halted the brucellosis control program and prevented a ﬁnal
assessment of the Hungarian project (Denes, 1997).

B. Present and Planned Brucellosis Control Programs

Brucellosis control in Mongolia, in the form of comprehensive livestock
vaccination and systematic surveillance, ceased to exist in the late 1980’s as a result of
the abrupt loss of Soviet economic aid (Mikolon, 2000). The brucellosis control program
in Mongolia from 1990, until 2001, consisted of low-level surveillance and sporadic herd
level testing for brucellosis in bovine, caprine, and ovine herds (Roth et al., 2003). In the
early 1990’s Mongolia reported to the OIE that the occurrence of bovine, caprine, and
ovine brucellosis was “enzootic with a ubiquitous distribution” (OIE, 1990-1994),
seemingly indicating that the Hungarian project described above had very limited
success. Later in the decade the occurrence was reported as “low sporadic” and in 2001 ,
and 2002, the reports of occurrence of bovine brucellosis were “based only on serological
evidence of the disease” (OIE, 1996-2002). Data provided by the State Veterinary and
Animal Breeding Department, Mongolian Ministry of Food and Agriculture, on the
nationwide prevalence of brucellosis in cattle, sheep and goats between 1991 and 1999,
indicated that the prevalence in cattle ranged from 0.80% to 3.04% and in sheep and
goats between 0.50% and 1.11%. During this period (1990-2002) the Mongolian Ministry
of Food and Agriculture reported a number of outbreaks of brucellosis in livestock as
well as measures taken to control the disease in the form of slaughter, vaccination,

quarantine and a modiﬁed stamping out policy (OIE, 1990-1994, 1996-2002).

30

Efforts to recreate and fund the latest brucellosis control program in Mongolia
began in 1999. The World Health Organization (WHO) provided technical support for
the preparation of a National Brucellosis Control Plan that called for the mass vaccination
of ruminants and was due to begin in 2000 (Mikolon, 2000; WHO, 2000a; Zinsstag et al.,
submitted 2004). Mongolia’s current brucellosis control strategy consists of a 10-year
vaccination program designed to vaccinate all adult bovine, caprine, and ovine herds in
the country twice within 6 years. One third of the total adult population will be targeted
per year. All animals born during the 10-year program will be vaccinated once at less
than 1 year of age (Zinsstag et al., submitted 2004). The 10-year vaccination plan has
been set in motion by the Mongolian government despite signiﬁcant set backs posed by
the combination of drought and severe snowstorrns in the 1999-2002 period, and a lack of
signiﬁcant foreign ﬁmding for the vaccination program. Although a recently published
report indicated that “as of 2001, no uniform animal vaccination program existed in
Mongolia” (Ebright et al., 2003) no data are yet available on the degree of

implementation or impact of the current Mongolian brucellosis control program.

1.5 CONCLUSIONS AND AREAS FOR FUTURE RESEARCH

Human and livestock populations in Mongolia have been affected by brucellosis
from the 1950’s through the 1980’s and most likely continue to suffer from a signiﬁcant
burden of the disease today. Multiple studies performed in the late 1950’s through the
1980’s collected data on brucellosis in Mongolia that were published in the international
literature (Beulig et al., 1969; Denes, 1997; Jezek et al., 1972; J ezek et al., 1974; Otgon,

1968; Pinigin et al., 1968a; Pinigin et al., 1968b; Tserendash, 1980, 1984; Tserendash

31

and Shumilov, 1970). The authors also cite historical accounts of cases dating the

recognition of brucellosis in Mongolia back to the early 1930’s (Otgon, 1968). These

published reports provide information primarily on the prevalence of brucellosis in

humans and individual animals. Details of the testing methods employed are provided

but the methods used for selecting individuals for testing are less clear.

Conclusions that can be drawn from the studies performed in the 1950’s through

the 1980’s include the following:

1.

2.

Brucellosis was present in multiple livestock species in Mongolia.

Both rural and urban human populations were affected by brucellosis.

. Brucellosis was distributed throughout Mongolia with some regional variation in

the prevalence of the disease in livestock.
Brucella abortus was responsible for brucellosis infection in Mongolia. The

presence of Brucella melitensis was suspected but undocumented.

. Regions of the country with a higher density of livestock and State farms were

associated with a higher prevalence of disease in livestock.

Higher risks of brucellosis-seropositivity in humans were associated with contact
with livestock (veterinarians, slaughterhouse workers, and herders).

High rates of disease transmission were recorded in both the livestock and human
populations.

Methods implemented to control brucellosis in livestock (primarily vaccination)
led to a decrease in seroprevalence and clinical signs of brucellosis in livestock
herds, though control programs implemented in the late 1980’s were never

evaluated.

32

Scientiﬁc studies assessing the effects of brucellosis control in Mongolia on the
incidence of the disease in the human population are unavailable. Published information
on brucellosis in Mongolia post-1990, consists primarily of summaries of reports made
by Mongolian government ofﬁcials to the Oﬁ‘ice International des Epizooties or the
World Health Organization. The most recent publications explore the use of more
discriminatory diagnostic methods for brucellosis testing (Erdenebaatar et al., 2002),
highlight brucellosis as 1 of 4 most important infectious human diseases in Mongolia
(Ebright et al., 2003), and outline the economic and human health beneﬁts of a
comprehensive program to control brucellosis in livestock in Mongolia (Roth et al., 2003;
Zinsstag et al., submitted 2004).

The epidemiology of brucellosis in Mongolia today, both the distribution of
disease in livestock and the burden of disease in humans, is largely unknown. Well-
designed population-based studies are needed to assess the current level of infection in
livestock and humans in Mongolia. Strain typing, disease incidence data and an analysis
of risk factors associated with disease would greatly enhance the understanding of disease
transmission and provide needed information for designing effective disease prevention
programs. Ideally a nationwide brucellosis surveillance system should be put in place
with the use of standardized testing methods, uniform case deﬁnitions and reliable
reporting mechanisms. With limited funds and the logistical constraints associated with
an extensive livestock system, the surveillance system would need to be adapted
appropriately. Mongolia’s move to implement its national brucellosis control program
through the mass vaccination of its livestock is admirable. Coupling this disease

prevention program with a streamlined yet functional brucellosis surveillance system

33

would allow for an assessment of disease control progress, the identiﬁcation of areas for

more intensive measures, and a more successful program overall.

34

326815 5.582 833 Emu 88m aim—«MEAN: was 223.5: 2: wEg—oﬁ $08383 amen—E o van ggnneﬁb 89¢ 380:8 88 :«ESE .

 

 

 

 

 

3 Eu
A88: 2 3.5 52-38
3 Eu
5%: 2 SE 329
3 .Eo
£843 a; :3 95
3 Eu
3.38 E 5.5 a_o-§am 83: >o=§nm Ea sausage
mea_
85 3: to
can is 5.5
$2: 9: Eu
$9.: an" .25 £882 38: 8538» Ba eaaa
Nea_
cm: 35 83625
cm: at” Sam Eomeoz $8: 8330M Ea ewes
Hea_
8.: 3 .55 505:8
.63: 3.... cm 5. .65 .5582 $8: .530
ems“
Ewan—m 3:30 ua¢U\A—Qo__m u—uuaU 30H. £05521—
:vgmou m0—ﬂaam a: «a e\e 00:25»?- mmmc——ou=.~£ vegan—mumm— =¢muuo=onv 5am— uc use?
.ooBom 23
80> .3 «sowcoz E meocﬂsnom 58:3 98 x0053: E coco—«Bi Emozoop—m mo $83M 35:95 Mo 52:56 AA 03.3.

35

 

$2.0 3. :83: .299.

so: em ..
58 em ..
sot m... ..
83 5. ..
£8 3 59m
3 no: we: ESE ﬂag
5 s 33.. eons.» weésesm
and 3:2 23me 338%....
cm: 353 e383 can
98v 33.2 eons; vase banana Eu 8 eﬁmﬁﬁm
3 Eu
5%: 3 5.5.
.3 Eu
ace 9... 5.5
a: #6
A85 2 Sam
3 Eu
58: 3. Sam
Had—=5 3:30 gawk—00.5 2:50 amok.

 

 

e038 v.2 83 we a e\.. cone—«>95 $2585.:— ecu—«Ecum—

 

eoEoQ
_>oonoQ
~owm>o§

52-38
m>D

333535

Homeoﬁcdxmm
await...
_omm>o§
§§>mN

nauseou—

Qan as we onon
N53

88: seesaw ea aaeeocae
wea~

538.30 3.5 no 53>

 

eoaeueoo .3 2%...

6
3

 

:0: we we Nwaﬁ

 

 

3 .. .. 52

m... .. .. 82

E .. .. ES

3 .. .. $2

m.~ .. .. :2

ma .. .. £2

3. .. .. £2

3 .. .. E2

88an 9: Him 53:53 m5

Gougenc maﬁa Quamngm .Euﬂ :SNNEUH: so Eﬁwoa cowmﬁoog 2t 8cm 380:8 8% 028385 3me ammvcouomb

$3.32

88m; 2: 35 $5585 83: 58:25

. , . ~5a_
8% EN 5% 55.9

$5: v." 12.55
6va 3. session 2382 .. .. ..
63 3 8:238 8:3 Avocozoa 3% 3 59,25 €29
E: a h: 452‘

a3 02 25233 23802 .. ..

A28 2: 8:238 850m 5925 65.9 $3 : .3 a x33

ehm_
ﬂags: 3:50 uaow>uuo£w 2360 amok. ﬂamenco:—

aﬁoumuu moan—am a: *3 .X. ovum—ugvu mums—03:: Boqumumm Hagan—BU «nan— .«O .50?

 

8:588 .2 23h

37

o§~=on< can voom .«o EmE_E .Eogaoﬂ magnum 98 bactu~o> 88m Snowaoz moocﬂzogm mﬁmo=ooam ﬁaozaz “358 35320 ..

 

 

8:253 a... SEES 3 .. .. 32

823mm: 2. $325 3 .. .. M32

332v: a... @239 S .. .. 82

5: SM: 2 2825 3 .. .. 82

$232: a... @2682 N2 .. .. R2

@2122: m... 5239 3 .. .. $2

@382: a... @289 3. .. .. .82

@352 c 3. @223 a... .. .. 82

82%: c m... €385 a... £288 £025: 32

Laos .3 s @382

22-32

3 .. .. $2

a... .. .. $2

3 .. .. 32

S .. .. 22

3 .. .. $2

2 .. .. 2.2

a.” .. .. E2

3 .. .. £2

E .. .. £2

88an 3 3,5 Eggm 2.2

Avocm>0umv wage hammnnxdm .Ehﬂ .._N#Duumm.. co gang comamcmouw> 05 woﬁw 6302—00 Saw o>muoomm0um ANmoﬁv nmmﬁﬁohowh
58:: .255 39325 2:5 .5. 5:83 82-2.2
“—533 maﬁa—am a: #5 e\.. mono—Fro.— mmmO—EuF—a Gauss—Sana 5:36:00 Sun— a: .30?

 

 

Banana .3 2...?

38

Chapter 2

THE PREVALENCE OF BRUCELLOSIS-SEROPOSITIVE CATTLE/Y AK
HERDS IN THE KHANGAI MOUNTAIN REGION OF BAYANKHONGOR

AIMAG (PROVINCE), MONGOLIA

2.0 STRUCTURED ABSTRACT

Objective: To determine the prevalence of brucellosis-ser0positive cattle herds in the
Khangai mountain region of Bayankhongor Aimag (Province), Mongolia, and describe
livestock management and human/livestock interaction in the region.

Design: Cross-sectional study

Sample Population: A stratiﬁed random sample (n=105) of the 5,062 livestock herds in
5 contiguous soums (districts) and 1 sub-district in the Khangai mountain region of
Bayankhongor Aimag (Province), Mongolia.

Procedure: Cattle herd brucellosis status was determined by testing pooled milk
samples from each herd with the Brucellosis Milk Ring Test (BRT). Additionally, serum
from a subset of individual cattle in each herd was tested with the Brucellosis Card Test
(CARD). A structured questionnaire was administered to collect data on livestock
management, livestock/human interaction and human health.

Results: The prevalence of brucellosis-ser0positive cattle herds in the region was 41.3%
(95% C .I.; +/- 9.8%) based on the results of the BRT testing. Additional herds detected

with the CARD test increased the proportion of positive herds to 53%. The proportion of

39

ser0positive individual animals was 14.9%. Bovine abortions were reported in 21% of
the herds and 84% of herds experienced some degree of neonatal mortality. The majority
of herder households (75%) herded multiple species of livestock and 100% of the
households herded and milked cattle (predominantly yaks). Twelve percent of the
herders interviewed reported consuming raw milk and 33% reported preparing dairy
products from raw milk. Twenty-four percent of the herders interviewed reported clinical
signs suggestive of brucellosis.

Conclusions: The high prevalence of brucellosis-ser0positive cattle herds in the Khangai
mountain region of Bayankhongor Aimag is consistent with data from other regions of
the world where there is an absence of brucellosis control programs and established
records of economic losses and serious public health implications due to zoonotic
transmission of the disease. The proportion of ser0positive individual cattle (15%) found
in this study was higher than that ofﬁcially reported for Bayankhongor (5.4% overall,
8.8% in female cattle). Livestock management factors, livestock/human interaction
behavior, and the occurrence of human illness reported in this study can be used in a
future evaluation of risk factors associated with livestock and human brucellosis in this
region of Mongolia.

Relevance: This study provides data that support the anecdotal reports of brucellosis as a
re-emerging disease in Mongolia and a potential major public health concern. The data
will also serve as important baseline information on the cattle herd brucellosis-
seroprevalence in the region before the implementation of the national brucellosis control

program.

40

2.1 INTRODUCTION

Brucellosis is an infectious disease of livestock and one of the most widespread
zoonotic diseases in the world (Corbel, 1989). The worldwide distribution of endemic
brucellosis includes Mongolia, a sparsely populated country (1.5 persons/kmz) in Central
Asia where livestock herding is the backbone of the county’s economy and an important
part of its cultural heritage (Mongolia, 20010; OIE, 2002b). Major efforts have been
undertaken around the world to control brucellosis due the signiﬁcant economic losses
associated with the occurrence of the disease in domestic livestock and the often
debilitating disease associated with human infection (Halling and Boyle, 2002; Ragan,
2002; Reddy, 2003). Clinical signs of brucellosis in domestic livestock include late-term
abortions, decreased milk production and lowered fertility in both males and females
(Drost and Thomas, 1996). The disease in humans is associated with a number of
nonspeciﬁc complaints including irregular or intermittent fevers, night sweats and
headaches. Untreated brucellosis in humans can lead to chronic osteoarticular
complications manifesting as joint pain, joint effusion and debilitating arthritis
(Benenson, 1995; Young, 1989). .

Particular biovars of the species of Brucella melitensis are associated with
pathogenicity in different domestic animals. The disease in cattle is most commonly
caused by the biovar abortus (Reddy, 2003). The vast majority of human infection is the
result of direct or indirect transmission of the disease from infected livestock through the
ingestion of contaminated milk or dairy products or through direct contact with aborted

material or uterine ﬂuid from infected animals (Nicoletti, 1989).

41

Historically, Mongolia has had a high incidence of brucellosis in its human
population (Otgon 1968; J ezek Rusinko et a1. 1972; J ezek, Rusinko et a1. 1974) as a
consequence of the high prevalence of the disease in domestic livestock (Beulig et al.,
1969; Jezek et al., 1972; Jezek et al., 1974; Otgon, 1968; Pinigin et al., 1968a; Pinigin et
al., 1968b; Tserendash, 1980, 1984; Tserendash and Shumilov, 1970; Zherikova et al.,
1972). Brucellosis prevention, through livestock vaccination, began in Mongolia in the
1960’s (Corbel, 1989; Tserendash, 1980). A second prevention program, lead by the
World Health Organization (WHO), began in the mid-1970’s and focused on the
vaccination of small ruminants (Corbel, 1989). The former Soviet Union continued to
support a brucellosis livestock vaccination program in Mongolia (Brucellosis abortus
strain-19 vaccine in cattle and Br. melitensis Rev-1 vaccine primarily in small ruminants)
through the late 1980’s. Initially the control programs were deemed a success by
Mongolian scientists (Tserendash, 1980) and by WHO ofﬁcials who reported that the
annual incidence of the disease in humans dropped to less than 1 case/ 10,000 in the early
1980’s (Mikolon, 2000; Roth et al., 2003).

Despite evidence in the early 1980’s of a signiﬁcant reduction in the incidence of
brucellosis in Mongolia, concerns have been raised recently about the re-emergence of
the disease in the country (Ebright et al., 2003; Joshi, 1991; Roth et al., 2003; WHO,
2000b, 2001, 2003). Brucellosis surveillance in Mongolia from 1990, until 2001, has
consisted of unsystematic sampling of individual livestock and sporadic herd-level testing
(Roth et al., 2003). These data do not allow one to quantify the regional prevalence of
disease or conﬁrm the re-emergence of brucellosis, however, they do demonstrate that

livestock brucellosis continues to persist in Mongolia. In the early 1990’s Mongolia

42

reported that the occurrence of bovine, caprine and ovine brucellosis was enzootic with a
ubiquitous distribution (OIE, 2002b). Later in the decade the occurrence was reported as
“low sporadic” and in 2001, and 2002, the reports of occurrence of bovine brucellosis
were based only on serological evidence of the disease (OIE, 2002b). Data provided by
the State Veterinary and Animal Breeding Department, Mongolian Ministry of Food and
Agriculture, on the nationwide prevalence of brucellosis in cattle, sheep and goats
between 1991 and 1999, indicated that the prevalence in cattle ranged from 0.80% to
3.04% and in sheep and goats between 0.50% and 1.11%. During this period (1990-2002)
the Mongolian Ministry of Food and Agriculture reported a number of outbreaks of
brucellosis in livestock as well as measures taken to control the disease in the form of
slaughter, vaccination, quarantine and a modiﬁed stamping out policy (OIE, 2002b).

The persistence and possible re-emergence of brucellosis as a major public health
problem in Mongolia has been blamed on the collapse of the veterinary service delivery
system and public health infrastructure in Mongolia which began in the late 1980’s
(Ebright et al., 2003; Mikolon, 2000). The Mongolian veterinary system in the 1990’s
reﬂects the changes that were experienced by the country as a whole as Mongolia
transitioned from a centrally planned economy and socialist system of government to a
democratic system with a market economy (Ebright et al., 2003; Rossabi, 1999). The
combination of the abrupt loss of economic aid from the former Soviet Union, the
privatization of livestock in 1990, and the privatization of veterinary service delivery in
1999, meant the end of large scale, government subsidized, livestock health campaigns
including comprehensive livestock vaccination programs and systematic surveillance for

brucellosis (Mikolon, 2000).

43

Recently collected data on the prevalence, incidence and distribution of
brucellosis in Mongolia’s livestock and human population is largely unavailable. Ofﬁcial
government reports provide some information on disease occurrence but the sampling
methods employed are unknown and little insight into the level and distribution of
disease can be gained. The aim of this study was, therefore, to systematically collect data
on the prevalence of brucellosis in livestock in Mongolia. The decision was made to
focus on the Khangai mountain region of Bayankhongor Aimag (Province), where
brucellosis was reported to occur and to estimate the prevalence of brucellosis-
seropositive cattle herds in the region. The speciﬁc aims were the following:

1. To use the brucellosis ring test (BRT) on milk samples and the Brucellosis
Card (CARD) agglutination test on individual serum samples to determine
the brucellosis status of cattle herds in the study area.

2. To demonstrate the efﬁciency of using stratiﬁed random sampling to
estimate disease prevalence in an extensively managed (nomadic)
livestock population.

3. To conduct 2 questionnaires designed to characterize livestock
management practices, human/livestock interaction, and the occurrence of
clinical signs suggestive of brucellosis in the human population in the

region.

2.2 HYPOTHESES
1) The prevalence of brucellosis-positive cattle (yak) herds in the Khangai
mountain region of Bayankhongor Aimag (Province), Mongolia, can be estimated by

field testing randomly selected herds in the region.

2) Structured questionnaires can be used to collect information on livestock
management practices and the occurrence of clinical signs and symptoms suggestive of

brucellosis in the human herder population in the study region.

2.2 MATERIALS AND METHODS
A. Study Site: Bayankhongor Aimag (Province), Mongolia

This study took place in Bayankhongor Aimag (Province) in southwestern
Mongolia. Bayankhongor is the third largest of Mongolia’s 19 Aimags or Provinces.
The country is made up of the 19 Aimags and 2 cities, the national capital, Ulaanbaatar,
and an industrial center to the north, Darkhan—Uul (See Appendix A). Each Aimag is
divided into an average of 20 soums or districts. The soums are further divided into the
soum centers (small settled communities) and an average of 4 smaller administrative
units called baghs or sub-districts. This study area covered approximately 1/3 of the
116,000 km2 of Bayankhongor Aimag (Mongolia, 2001a).

Bayankhongor is one of the most ecologically diverse provinces in Mongolia.
The southern 1/3 of the province is part of the Gobi desert where it forms a section of
Mongolia’s border with China. The central 1/3 of the province is less arid and transitions

from the desert to the grassland steppe. The northern 1/3 of Bayankhongor is in the

45

mountain steppe zone and contains part of the Khangai mountain range where elevations
reach 4,000 meters and the primary livestock species is the yak. The ﬁve soums
(districts) and one sub-district (Shargaljuut) that make up the study area contain 41% of
the cattle (Bos taurus and Bos grunniens), 25% of the herder households, and 15% of the
total number of livestock in Bayankhongor (Mongolia, 2001b). Bayankhongor was the
leading province in number of livestock in 1999, with 2,475,100 head of camels, cattle,
horses, sheep and goats (Mongolia, 2001b).

B. Study Design and Sampling Frame

A cross-sectional approach was used to determine the prevalence of brucellosis-
seropositive cattle herds in the Khangai mountain region of Bayankhongor Aimag
(Province), Mongolia. The unit of interest was the herd and all herds located in the 5
contiguous Khangai soums (districts) of Gurvanbulag, Zag, Galuut, J argalant, Erdenstogt,
and the sub-district of Shargaljuut, were chosen for inclusion in the study (See Figure
2.1). This region was selected based on the predominance of cattle (mostly yaks) in the
composition of livestock herds in the region (Bayankhongor, 1999).

A livestock census is performed in Mongolia on a yearly basis. The herd, or
herder household, is considered the economic unit in rural areas and forms the basis for
measuring agricultural output in Mongolia’s extensive livestock management system
(Mongolia, 2001f). Livestock census data available at the Aimag (Province) level
include the total number of herder households in each soum (district) and the total
number of each livestock species per soum. These data from the most recent livestock

census (1999) were obtained and used to determine that the 6-strata (5 soums and 1 sub-

46

district) sampling frame contained 5,062 herder households, or sampling units, and close
to 200,000 head of cattle (See Table 2.1).

Livestock census data available at the soum (district) level include the number of
herder households per soum (district) as well as the number and species of livestock in
each herd, the bagh or sub-district in which the herd is grazed, the name of the herd
owner and the number of household members associated with the livestock herd. These
data were available in the form of lists obtained from the soum (district) government

ofﬁces.

47

Figure 2.1. The Study Region. Inset: The Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia. Below: The Five
Contiguous Soums (Districts) of Gurvanbulag, Zag, J argalant,
Galuut, Erdentsogt and the Bagh (Sub-District) of Shargaljuut.

      

 

100E

Russla
46 N
Chlna IV

 
      

48

C. Sample Size Calculation

Equation 1 (Thompson, 1992) was used to calculate the number of herds or
sample size required to estimate the proportion (prevalence) of brucellosis-ser0positive
cattle herds in the ﬁnite population of 5,062 herds in the study region. The goal was to
estimate the prevalence of brucellosis-positive herds in the 5 adjacent soums and 1
administrative unit of the Khangai mountain region in Bayankhongor Aimag (Province)
within 10% of the actual prevalence and a 5% probability of Type I error. For the
sample size calculation, the prevalence of brucellosis-ser0positive herds across the study
region was estimated to be 50%, allowing for the most conservative (maximum value)
calculation of required sample size and based on ﬁndings of other population-based
studies of livestock brucellosis (Al-Talaﬂrah et al., 2003; Kabagarnbe et al., 2001;
Nicoletti, 1980a). If we had reason to believe that the prevalence of ser0positive herds
was signiﬁcantly different from soum to soum we would have used a sample size
calculation for stratiﬁed random sampling that adjusts for both the relative weight of the
stratum and the expected prevalence of each stratum. When the expected prevalence is
equal across strata, the calculated sample size will be identical to that calculated for a
simple random sample using the formula below.
Equation 1:
Sample Size = n = ND “-9)

(N-1)d2/ZZ(1-o/2)+ P(1-p)

Where, “N” is the total number (5,062) of herds in the study; “p” is the estimated
prevalence of brucellosis-seropositive cattle herds in the region (0.5); “d” is the precision

or maximum acceptable error rate (0.1); and “a” is the probability of Type I error (0.05).

49

Using the formula in Equation 1, it was determined that brucellosis status data would
need to be collected from 95 herds to estimate the prevalence of brucellosis-ser0positive
herds in the 6-strata study area with the desired level of precision. The calculated
sample size of 95 was increased by 10, to 105, to account for the potential failure to
collect data due to the inability to locate herds, the absence of cattle in the herd, reﬁisal of
herders to participate or diagnostic test failure.

D. Herd Selection

The selection of the105 herds was performed using a stratiﬁed random sampling
technique. A simple random sample of herds was selected within each of the 6 stratum
(soum/district or bagh/sub-district). The number of herds selected per stratum was based
on the relative proportion of herder households in each stratum in the study area (See
Table 2.1). The stratiﬁcation and proportional allocation of sampling was done for
practical reasons. The census reports containing the individual herd identiﬁers were only
available at the soum (district) government ofﬁces. It was most efﬁcient to calculate the
overall sample size based on the aimag-level census data, which provided the number of
sampling units (herder households/soum), and then obtain the soum-level census data
with the herd identiﬁers upon arrival in the soum (district) centers.

The simple random sample of individual herds was selected from each soum list
by ﬁrst assigning a number to each herder household and then using a hand held
calculator to generate random numbers for herd selection (See Table 2.2). The location
and testing of the herds selected in each stratum began as soon as the herds were
identiﬁed. The herds were located with the assistance of the local soum (district)

veterinarians and the bagh (sub-district) governors.

50

E. Data Collection Team and Informed Consent

The data collection team consisted of two veterinarians, a physician, a driver who
also assisted with animal restraint, and a volunteer data recorder. In each soum the data
collection team met with local ofﬁcials and the soum (district) veterinarian. The soum
veterinarian accompanied the data collection team on all herd visits. The local soum
veterinarian made introductions to the herder household members. The Mongolian
bilingual (English and Mongolian) collection team member explained the project and
informed the herders that their participation in the study was voluntary. All team
members were available for questions. A record of informed consent was kept and data
collection proceeded only if consent was obtained.

F. Specimen Collection and Testing

Data collection took place in July and August of 2000. The summer months in
Mongolia are considered peak milk production periods. All of the livestock species are
seasonally bred and offspring are born in the spring and early summer.

Milk testing:

The Brucellosis Milk Ring Test (BRT) was used to test raw milk samples
collected from the selected herds in the study area. The BRT is an internationally
recognized serological method for detecting brucellosis in a herd of cattle and classifying
it as a positive or negative herd (OIE, 2002a). Blue haematoxylin-stained Brucella
abortus antigen is used to detect Brucella antibodies in the milk of affected cows (Alton
and Jones, 1967). The BRT was developed as a tool to use at the herd level to screen
dairy cattle for brucellosis. The sensitivity of the test was reported to be 72.2% (95%

CL; 58.l%-83.1%) with a speciﬁcity of 90.5% (95% CL; 76.5%-96.9%) when used to

51

test dairy cattle herds of varying size (1 1-399), (V anzini et al., 2001). The number of
cows contributing to the mixed bulk milk samples tested in this study was never exceeded
30 cows. Samples were stored at 4° C and tested using standard international protocols
for herds of less than 100 milking cows (OIE, 2002a).

Milk samples were collected soon after the morning or evening milking had been
completed. Individual cattle were routinely milked into buckets that were then emptied
into aluminum milk cans or an equivalent small bulk storage container. The raw milk in
the milk cans was mixed thoroughly and a single sample was withdrawn for testing. The
milk samples were stored in a portable refrigerator powered by a 12-volt car battery and
tested within 24 hoursz. The milk samples were tested for the presence of Brucella
antibodies using standard B. abortus strain 1119-3 BRT antigen provided by the United
States Department of Agriculture (USDA), Animal and Plant Health Inspection Service
(APHIS), National Veterinary Services Laboratories (NV SL), Ames, Iowa, USA. One
ml of whole milk from each sample was transferred to a narrow, sterile plastic test tube.
BRT antigen (0.03 ml) was added to each of the samples and mixed well. The milk
samples were then incubated for 1 hour at 37 0 C in a portable incubator powered by a 12-
volt car battery. Concentration of dyed antigen in the cream layer, forming a dark blue
ring, indicated a positive reaction (Alton’et al., 1988; CE, 2002a). Results were read at
the conclusion of incubation and recorded.

Serology:

Ten ml of blood were collected from approximately ‘/2 of the female cattle over

the age of 6 months in each of the selected herds. Sterile 20-gauge, 1-inch needles and a

 

2 The standard minimum storage time of 24 hours was reduced to 12-24 hours in this study. The protocol
was adjusted due to the difﬁculty in maintaining the samples at the required 4° C for a 24-hour period in
the ﬁeld. The samples were tested in less than 24 hours to avoid curdling of the milk.

52

10 ml Vacutainer tube or syringe were used to draw blood from the jugular or middle
coccygeal vein. The samples collected by syringe were transferred to clean test tubes and
the test tubes and Vacutainer tubes were centrifuged for 3-5 minutes to separate the red
blood cells from the serum, using a portable centrifuge powered by a 12-volt car battery.

Serum samples were tested for the presence of antibodies against Brucella using
the commercially prepared Brucellosis Card Test Kit with Rose-Bengal stained, buffered
Brucella abortus strain 1119-3 antigen produced by the USDA, APHIS, NVSL, Ames,
Iowa, USA. The brucellosis card test (CARD) is a modiﬁed serum agglutination test that
was developed in the United States to meet the need for a rapid ﬁeld test for the
serological diagnosis of brucellosis in the ranch or stockyard setting (Prior et al., 1975).
The test was well suited to ﬁeld conditions in Mongolia.

The CARD test is generally characterized as having a high sensitivity and low
speciﬁcity. Used in a population of non adult-vaccinated cattle, the sensitivity and
speciﬁcity the CARD test was determined to be 99.0% and 30.6% respectively, when
evaluated based on bacteriologic culture and isolation of B. abortus as the deﬁnitive test
(Huber and Nicoletti, 1986). Agreement between the CARD test and the serum tube
agglutination test (STAT)3 under ﬁeld conditions is reportedly high (O'Reilly and
Cunningham, 1971; Prior et al., 1975). The CARD test was used in this study in addition
to the BRT, allowing for the identiﬁcation of brucellosis-positive animals in the non-

milking portion of the cattle herd.

 

3 The serum tube agglutination test (STAT) is used to test for the presence of antibodies in serial dilutions
of the subject’s serum produced in response to brucellosis infection. The highest serum dilution showing
50% or more agglutination is recorded as the end-point Alton, G., Jones, J ., 1967, Laboratory Techniques
in Brucellosis. World Health Organization, Geneva.

53

Serum samples were tested using the protocol outlined in the Brucellosis Card
Test Kit instructions. One drop (0.03ml) of serum and one drop (0.03ml) of Brucella
antigen (USDA Rose-Bengal stained, B. abortus strain 1119-3) were placed in an
individual well on the standard card, mixed with an individual wooden stick and then
rocked for 4 minutes. The contents of the wells were examined in good light for signs of
agglutination and results were recorded in the ﬁeld notebook as “positive agglutination”,
“weak agglutination”, “very weak agglutination” or “absent agglutination“.

Brucellosis card test (CARD) positive serum samples were stored and submitted
to the Mongolian State Central Veterinary Diagnostic and Sanitary Laboratory for further
testing. Serum from CARD-negative animals was discarded. The serum from the
CARD-positive animals, including samples with a “very weak” positive agglutination
reaction, was retested in the laboratory using the Mongolian Complement Fixation Test
(CFT). The CFT is accepted and often used as a conﬁrmatory test for bovine brucellosis
(OIE, 2002a). The CFT is a serologic test designed to determine the Brucella titer of the
serum under test. The CFT utilizes Brucella antigen, complement and sheep red blood
cells to measure the degree of hemolysis. The titer of the serum under test is read as the
highest dilution showing 50% or less hemolysis. In cattle a titer of l :40 indicates an
infected animal and one of 1:20 is suspicious (Alton and Jones, 1967).

The sensitivity and speciﬁcity of the CFT, at a test titer > 40, in a population of
non adult-vaccinated cattle was 64.8% and 81.4%, respectively (Huber and Nicoletti,
1986). The best relationship is seen between the CARD test and CF T when the level of

infection in the population under test is high (Tirnbs et al., 1978).

 

’ Positive agglutination = complete clunrping of antigen; Weak agglutination = ﬁne particles observed with
some clumping; Very weak agglutination = very ﬁne particles present with no clumping; Absent
agglutination = No visible agglutination

54

Bacteriology:

Milk can samples from positive herds were saved after testing, refrigerated and
eventually frozen. The samples were submitted to the Central Research Laboratory at the
National Medical University of Mongolia in Ulaanbaatar for bacteriologic exam. The
equipment and facilities necessary for Brucella culture were, however, unavailable and
no culture results were obtained.

G. Case Definitions

Sampled herds were classiﬁed as brucellosis-ser0positive if either a BRT-positive
result was detected in a mixed raw bulk milk sample or at least one individual animal
was classiﬁed as a reactor to the CARD test. An individual was classiﬁed as a reactor if
their serum had a “positive” or “weak” agglutination reaction on the CARD test. An
animal with a serum sample exhibiting “very weak” or “absent” agglutination on the
CARD test was considered negative.

H. Herd Management Survey

A 3-page, structured livestock management questionnaire was conducted as an in-
person interview with an adult member of the herder household involved in livestock
rearing and self-identiﬁed as the “head of household” (See translation of “Livestock
Management” questionnaire in Appendix B). The interview was conducted in the
Mongolian language by a bilingual (Mongolian and English) native-Mongolian speaker.
Questions included requests for information on livestock numbers, herding practices, and
a history of vaccination and veterinary interventions. The herders’ awareness of the
characteristics of livestock brucellosis was assessed using two open-ended questions.

Interview responses were recorded in Mongolian and later translated into English.

55

1. Head of Household Survey

A 2-page questionnaire was conducted as an in-person interview with the
available male or female individual self-identiﬁed as the “head of the herder household”.
The interview was conducted in Mongolian by a bilingual (Mongolian and English)
native Mongolian speaker. The questions were designed to characterize herder contact
with livestock and the practices surrounding the preparation and use of livestock
products. Additionally the interviewees were asked to indicate the presence or absence of
the occurrence of a list of 17 non-speciﬁc clinical signs and symptoms of disease within
the past 12 months. Speciﬁc information on the subjects’ previous brucellosis diagnosis,
testing and treatment was also collected. The ﬁnal questions assessed the degree of
herder awareness of the zoonotic nature of livestock brucellosis and the modes of disease
transmission (See translation of “Human Health/Livestock Contact/Product

Consumption” questionnaire in Appendix C).

2.3 DATA ANALYSIS

A. Herd-Level Data

Data from the BRT and CARD test results were transferred from ﬁeld notebooks
to a computer database. The test results were entered in a Microsoft Excel spreadsheet
database (Microsoft, 2000) along with the herd identiﬁer, the number of cows
contributing to the mixed milk samples and information on the age and breed of
individual animals. The prevalence of brucellosis-ser0positive herds in the 6-soum study
area was calculated as the proportion of test positive herds among those herds tested. The

percent prevalence and 95% conﬁdence intervals were adjusted to account for the

S6

stratiﬁed sample using the ﬁnal sampling weights. Equation 2 (Lohr, 1999) was used to
determine the overall proportion of brucellosis-positive herds in the region and the

variance of this stratiﬁed sample was calculated using Equation 3 (Lohr, 1999).

Equation 2
H
Estimated Prevalence = Z Nh/N (Ph)
h=1
Equation 3
H
Estimated Variance = z: (l-nh/N) (Nb/N)2 [Ph (1- Ph)/ nh -1]

h=l

H=strata (6); h=stratum; N=total herder households in sampling frame; Nh=total herder
households in stratum; Ph=proportion of positive herds per stratum; nh=number of herder
households sampled per stratum.

B. Individual Animal-Level Data

The number of cattle with a positive CARD test in the herd divided by the total
number of cattle tested in the herd was used to approximate the prevalence of brucellosis-
seropositive cattle in each herd. The overall individual cattle brucellosis-seroprevalence
level in the study area was calculated as the total number of cattle with a positive CARD

test divided by the total number of cattle tested.

57

C. Concordance Measurement

The “interrater” reliability of the BRT and CARD test as methods for determining
cattle herd brucellosis status was measured by calculating the Kappa statistic, a measure
of concordance between the BRT and CARD classiﬁcation of the cattle herds.

D. Questionnaire Data

Data from the “livestock management” and “human health/livestock
interaction/product consumption” questionnaires were entered into Microsoft Excel
spreadsheets (Microsoﬁ Ofﬁce 2000). Imported Excel spreadsheet data were analyzed
using the SAS software program (SAS version 8.2, Cary NC: SAS Institute, Inc. 1999-
2001). Cattle abortion rates, calving rates and neonatal mortality rates were computed
and entered as new variables. The individual animal data was used to calculate the within
herd prevalence of brucellosis CARD reactors and the average age of cows in the herd.
Descriptive statistics were computed for herd characteristics and management factors,
human /livestock interaction, livestock product consumption habits and human health

status reporting.

2.4 RESULTS

A. General Characteristics of Livestock Herds

Herd Composition:

Yaks (Bos grunniens) were the predominant milk producing species in the study
region. All of the herder households raised cattle and milked at least one cow. The mean
size of cattle herds in the study was 20 with a mean of 7.7 milking cows per herd. F iﬁy-

ﬁve percent of the 100 heads of herder households interviewed reported that they herded

58

exclusively the yak species of cattle. The majority of the remaining cattle herds were
composed of crossbred (Bos grunniens and Bos taurus) animals or both species. Ninety
percent of the individual cattle selected for testing were Bos grunniens or Bos grunniens
crosses. The term “cattle” in this thesis refers to the bovine livestock in the region that
were predominantly the yak or Bos grunniens species.

The majority (70%) of livestock herds included in this study were mixed herds of
cattle, sheep, goats and horses. The total number of livestock could only be calculated
for approximately 60% of the herds due to errors in data recording. The mean number of
livestock per herd was 95, with a median of 67, and a range of 3-390. The mean number
of cattle per herd was 20, with a median of 17, and a range of 1-70. The mean number of
sheep was 65 and the mean number of goats was 26. Data on the total number of horses
was not collected. Five percent of the herds included a few camels in their composition.
Of the 61 herds for which we had total livestock number data, 80% were composed of
fewer than 150 head of livestock, the Mongolian govemment’s ofﬁcial deﬁnition of
poverty.

LivestockLManagement:

All of the herder households included in this study were nomadic, moving their
livestock and felt dwellings, or gers, an average of 4 times per year. Groups of herder
households often migrate together and generally graze their livestock on shared pasture.
In addition, 75% of herders reported the mixing or their livestock with other herds at
water sources, 75% mixed with other herds in sheltered winter areas and 73% of herds

mixed with other livestock overnight.

59

All of the herder households included in the study milked cattle. Only 6% of the
herders also milked sheep and 13% also milked goats. Sixty-eight percent of the herders
reported that their livestock had received veterinary services within the past 12 months.
When asked speciﬁcally about brucellosis vaccination, 4 herders reported that their cattle
had been vaccinated against brucellosis. Two herders reported that the vaccination had
occurred in the last year.

Livestock Health and Production Parameters:

Abortions were observed in cattle in 21% of the herds, in sheep in 12% of the
herds and in goats in 25% of the herds. Neonatal calf mortality was reported in 84% of
herds and the average calf mortality per herd was 10%. Horses with draining sores in the
vicinity of the withers were only reported in I herd.

Herders reported that the average age of a ﬁrst calf heifer was 3 or 4 years. The
mean age of lactating cattle per herd was 7 years. The calculated mean calving rate per
herd in the sampled population was approximately 80%.

Livestoﬂg Brucellosis Knowledge:

Typical symptoms of livestock brucellosis (abortions and decreased milk
production) could be described by 15% of the herders interviewed. Only 10% indicated
that they knew that livestock brucellosis could be prevented with vaccination. The routes
of disease transmission were identiﬁed by less than 10% of the individuals interviewed.

B. Prevalence and Distribution of Brucellosis-Seropositive Cattle Herds

One hundred and four of the 105 herder households selected for sampling were
located. The exact location of the herds ranged from N 46°33 ’13.2”, E 101°26’46.7” to

N 47°23’37.7”, E 098°25’29.9”, a point to point distance of approximately 250 km. Only

60

one head of household refused to enroll their livestock herd in the study or respond to the
questionnaires. Bulk milk samples were collected from 102 of the 103 herds
participating in the study and results were obtained from 98 milk samples tested with the
brucellosis milk ring test (BRT).5 Forty of the 98 milk samples tested were positive. The
proportion of BRT-positive herds, adjusted for stratiﬁcation, was 41.2% (95% CI. +/-
9.8%). The soum (district) with the highest proportion of BRT-positive herds was
Gurvanbulag with a prevalence of 53.8% (7/ 12). The soums (districts) with the lowest
proportion of BRT-positive herds were lag and Erdentsogt, both with a prevalence of
30% (3/ 10 and 6/20), (See Table 2.3).

Blood samples were collected from 513 cattle representing 101 different herds.
Serum from a sub-sample of cattle from each herd was tested with the CARD serum
agglutination test and an additional 14 herds were identiﬁed as brucellosis-positive with
this method. Forty herds had at least one animal classiﬁed as CARD-test positive. The
proportion of cattle herds with at least one CARD-positive bovine, adjusted for
stratiﬁcation, was 39.2% (95% CI. +/— 9.3%). The proportion of cattle herds classiﬁed as
brucellosis-ser0positive after testing in series with the CARD followed by the Mongolian
CFT was 22.7% (95% CI. +/- 8.1%). The proportion of cattle herds classiﬁed as
brucellosis-positive by both the BRT and CARD test was 24.3 (95% CI. +/- 8.1%).
When either a positive CARD test or a positive BRT sample was used to classify a herd
as positive, the adjusted estimate of brucellosis-ser0positive herds in the study region was

53.0% (95% CI. +/- 9.5%).

 

5 Four of the milk samples appeared grossly abnormal at the time of testing. The milk samples had curdled
during the storage period and test results were not considered valid.

61

A summary of the test results and the proportion of brucellosis-ser0positive herds
are presented in Table 2.4. The brucellosis-ser0positive herds were distributed across the

study area (See Figure 2. 2) with no obvious geographical clustering of positive herds.

Figure 2.2. Map of the Khangai Mountain Region of Bayankhongor Aimag
(Province), Mongolia. Labeled Points Indicate the Location (Global
Positioning System) and Brucellosis Status (squares = positive herds;
circles = negative herds) of Each Cattle Herd Selected for Testing.

100 E _
Russra
ﬂ Mongolia
_ 46 N
.'

 

     
 

China "

   
 

° Brucellosis - Positive Herds
I Brucellosis - Negative Herds

62

C. Concordance of Herd-Level Brucellosis Testing

The use of the BRT to test bulk milk samples and the CARD test to detect
individual reactor cattle resulted in a similar estimation of the prevalence of brucellosis-
seropositive herds, 41 .2% and 39.2% respectively, in the study area. The kappa statistic,
which quantiﬁes the degree of concordance between the BRT and CARD test, was 0.45
(See Table 2.5). The individual cattle tested with the CARD method and those that
contributed to the mixed milk sample were not always the same animals. The CARD
test, performed on serum, had the potential to detect brucellosis-positive non-milking
cattle in the herd.

D. Individual-Animal Brucellosis-Seroprevalence

Serum samples ﬁ'om 513 individual cattle from 101 of 103 herds in the study
were tested with the CARD test for brucellosis. Seventy-one of the serum samples
exhibited positive agglutination reactions. The overall prevalence of CARD reactors in
this population of cattle was 13.8%. The soum with the highest proportion of CARD-
positive individual cattle was Galuut with 19.2% reactors and the soum with the lowest
proportion of CARD-positive cattle was Erdenstogt with 10.8% (See Table 2.6).

Over 90% of the individual cattle tested were yaks (Bos grunniens),
approximately 3% were “Mongol” cattle (Bos Taurus) and 2% were hainaks (F -1 cross of
Bos grunniens and Bos taurus). Yaks, “Mongol” cattle and hainaks were all represented
in the CARD-positive population of cattle. The mean age of the cattle tested was 6.67
years and the median age was 6.00 years.

The CARD test results on individual cattle were used to estimate the ‘ﬁvithin

herd” prevalence of brucellosis-ser0positive cattle in the herds. The gross “within herd”

63

seroprevalence levels ranged ﬁom 0 to 100% with a mean of 13% and 95% CI. (9.0%,
17.0%). The mean within herd prevalence among the herds with at least one CARD-
positive animal was 33.3%.

E. Human-Livestock Interaction

Data was collected on the nature of contact between livestock and hmnans in each

herder household by interviewing the male or female head of the household and
administering a questionnaire (See Appendix C). Information was gathered from 101 (all
but 3) herder households. The average age of the interviewed herders (heads of herder
households) was 42, median 41, with a range of 17-81. The interviewed herders had
received an average of 5.7 years of formal education, median 8 years, with a range of O to

12 years. A slightly higher proportion of the herders interviewed were female (52.75%).
Herder Occupation:

The majority (92%) of interviewees identiﬁed “livestock herding” as their
primary occupation or profession. Six percent of the individuals identiﬁed other
occupations including 3 veterinary technicians and a human health worker. None of the

interviewed individuals were involved in commercial livestock slaughter or trade.
Livestock Contact:

As expected, all members of the herder household participate in livestock rearing.
As mentioned before, 75 % of these herds were composed of multiple species of
livestock. The presence of cattle was recorded in the largest number of herds (100%)
followed by sheep, goats and camels. Home slaughter of livestock for personal

consumption and hand milking of individual animals is common practice among

64

Mongolian herders. Of the 100 individuals interviewed, 84% reported participation in
cattle slaughter, 87% in sheep, 80% in goat and 3% in camel slaughter. Ninety-six
percent of the individuals interviewed reported hand-milking cattle, 24% reported hand-
milking goats and 16% reported hand-milking sheep. Speciﬁc questions addressing
herder assistance of livestock during parturition or dystocia were not asked, however,
39% of herders reported that weak neonates (primarily lambs and kids) were brought into

the family’s living quarters for observation and extra care.
Meat/MilkﬁConsmnption:

Mongolian livestock herders generally consume the meat and milk products of the
livestock they raise. One hundred percent (100/ 100) of the heads of herder households
interviewed reported consuming beef, 93% (93/ 100) reported consuming mutton, 75%
(75/ 100) reported consuming goat meat and 3% (3/ 100) reported consuming camel meat.
Traditionally meat is thoroughly cooked, either by boiling or roasting. One respondent
reported consumption of raw meat of camel, cattle, sheep and goat origin. Tire
consumption of raw milk was reported by 15% (l 5/ 100) of respondents. Twelve percent
(12/ 100) reported consuming raw cow or yak milk, 3% (3/ 100) raw goat milk and 3%

(3/ 100) raw sheep milk. Thirty-three percent (33/ 100) of respondents reported making

dairy products with raw milk.

F. Human Health Status and Brucellosis Awareness

Human Health Status:

Twenty-seven percent (26/97) of interviewed heads of herder households reported
a history of brucellosis diagnosis. Twelve percent of the interviewed herders

(approximately ‘/2 with a history of brucellosis) had been diagnosed within the previous

65

12 months. Eight percent of the individuals reported that they had been tested for
brucellosis with a blood test and 5% reported skin testing. None of the interviewed
individuals reported having been treated for brucellosis either as outpatients or in
hospitals or clinics. None of the individuals reported that they had been rechecked for
brucellosis after initial diagnosis and 1 person reported experiencing a relapse of disease
(See Table 2.7).

In addition to the series of questions on brucellosis diagnosis and testing, the
heads of herder households were asked to indicate which of the 17 clinical signs and
symptoms on the list provided, they had experienced within the previous 12 months. (See
questionnaire in Appendix C). The most common symptom experienced was joint pain at
18%. The other symptoms experienced were headache (16%), lower back pain (12%),
stiff joints (11%), swollen joints (10%), muscle pain (9%), extreme weakness (9%),
dizziness (9%), bone pain (9%), abdominal distension (6%), difﬁculty sleeping (4%),
night sweats (2%), fever (2%), enlarged lymph nodes (2%), abdominal pain (1%) and 0%
experiencing chills and depression. When a number of these non-speciﬁc signs were
combined into a group characterizing acute brucellosis (chills, headache, fever or night
sweats) and a group characterizing chronic brucellosis (joint pain, stiff joint, swollen joint
or lower back pain), the prevalence of acute brucellosis-like symptoms was calculated as
16% (16/ 100) and the prevalence of chronic brucellosis-like symptoms in this
interviewed population was calculated as 20% (20/ 100), (See Table 2.8).

Knowledge of Prevention and Transmission:

Twenty-ﬁve percent of the heads of herder households interviewed understood

that brucellosis was a zoonotic disease and that infected livestock were a risk to human

66

health. Seven percent of the herders interviewed identiﬁed both livestock product
consumption and contact with aborted material as a means of brucellosis transmission to
humans. An additional 18% identiﬁed either consumption of contaminated livestock
products (meat and milk) or direct contact with aborted material as the means of
transmission of disease from livestock to humans.

Eighteen percent of the heads of herder households correctly identiﬁed ways to
prevent brucellosis occurrence in humans. Overall 12% identiﬁed either livestock
vaccination/prevention or sanitary measures (cooking/boiling meat and milk and use of
protective clothing). Six percent of interviewees identiﬁed both livestock brucellosis

prevention and the institution of sanitary measures.

2.6 DISCUSSION

A. Brucellosis Prevalence and Distribution

The adjusted estimated prevalence of Brucellosis Ring Test (BRT)-positive cattle
herds in the Khangai region of Bayankhongor, Mongolia, was 41.3% (95% CL; +/-
9.8%). The addition of herds detected as brucellosis-positive with the CARD test brings
the herd prevalence in the region to over 50%, demonstrating a signiﬁcant burden of
disease in this population. Recent herd level data from Mongolia is unavailable for
comparison, however, the estimated prevalence of 41 .3% found in the study area exceeds
or was similar to the herd prevalence reported in countries with no systemic brucellosis
control program and similar systems of nomadic or extensive livestock management
(Salman et al., 1984). Recent reports include a study of cattle brucellosis in Eritrea

reporting a 36.6% prevalence of ser0positive cattle herds (Omer et al., 2001), a study in

67

Syria reporting a prevalence of 17.48% (Darwish and Benkirane, 2001), and a study of
cattle herds belonging to nomadic pastoralists in Chad reporting a brucellosis-
seropositive herd prevalence of 64% (Schelling et al., 2003).

The gross individual animal sero-prevalence of 13.8% was higher than expected.
The 1999, nationwide livestock brucellosis surveillance report produced by the State
Veterinary and Animal Breeding Department of the Mongolian Ministry of Food and
Agriculture, reported the sero-prevalence in cattle in Bayankhongor Aimag (Province) as
5.4% (114 positive/2,129 tested), and 1.3% nationwide. When the ofﬁcial data are
evaluated for adult male and female cattle separately, the reported prevalence in male
cattle in Bayankhongor is 2.1% (23 positive/ 1,098 tested) and 8.8% (91 positive/ 1,031
tested) for female cattle. The individual animal brucellosis-seroprevalence calculated in
this study was based on the testing of female cattle only. Although a higher (13.8% vs.
8.8%) estimate of brucellosis-seroprevalence was found in this study than that reported
for Bayankhongor as a whole, together the study data and the ofﬁcial reports indicate that
signiﬁcant levels of brucellosis infection are present in Bayankhongor’s cattle population.

Another explanation for the discrepancy between the ofﬁcial report of brucellosis
prevalence in Bayankhongor Aimag (Province) and that found in the study area is the
possible non-uniform distribution of brucellosis-ser0positive livestock. The 6 strata (5
soums and 1 bagh) selected for inclusion in this study represent a region for which there
was anecdotal evidence of signiﬁcant levels of bovine brucellosis. It is possible that the
ofﬁcially reported Bayankhongor Aimag-level prevalence is lower due to the sampling
from both the study region and the potentially lower prevalence southern 2/3 of the

province. Little information is available on the sampling or testing methods used in the

68

national brucellosis survey. Most Mongolian laboratories use a version of the highly
sensitive serum agglutination test on bovine serum. Sampling methods are generally
unsystematic and do not allow one to describe or evaluate data on the geographical
distribution of disease.

Cattle herds in this study were classiﬁed as brucellosis-ser0positive based either
on the brucellosis milk ring test (BRT) or the USDA serum agglutination CARD test.
When the herds are classiﬁed with the BRT and CARD test independently, the adjusted
estimated herd prevalence is very similar (BRT = 41.2 % +/- 9.8% and CARD = 39.2 %
+/- 9.3%). In both cases the herd is classiﬁed based on the testing of a subset of
individuals in the herd. Milking adult female cattle contributed to the mixed sample
tested with the BRT and approximately V2 the female cattle > 6 months of age were
included in the group of individuals tested with the CARD test. The discordance between
the BRT and CARD test results was due in part to the fact that the individual cattle tested
with the CARD method (not necessarily milking) and those that contributed to the mixed
milk sample (only milking cows) were different.

Classiﬁcation of a herd as ser0positive if either the BRT or an individual CARD
test were positive involved the use of the, greatest amount of information (number of
cattle tested) to make the decision on herd status. The adjusted seroprevalence estimate
of herds classiﬁed as positive based on the BRT or CARD test was 53.0%.

Misclassiﬁcation of cattle herds as brucellosis-ser0positive is also a concern in
this study due to the relatively low speciﬁcity of both the BRT and CARD tests (Huber
and Nicoletti, 1986; Nielsen et al., 1996; Vanzini et al., 2001). To increase the

speciﬁcity of our testing system we used a parallel testing strategy whereby both BRT

69

and CARD test results had to be positive to establish a more “conservative” adjusted
estimated herd prevalence of 24.3%.

Serial testing systems are also often used in brucellosis monitoring programs to
conﬁrm screening tests and improve the speciﬁcity of the testing system (Emmerzaal et
al., 2002). The complement ﬁxation test (CFT) is assumed to be 100% speciﬁc in non-
vaccination populations (Dohoo et al., 1986) and was used to conﬁrm the positive results
of the CARD test on individual serum samples in our study. The brucellosis
classiﬁcation of cattle herds based on serial (CARD + CFT) testing results in the adjusted
estimated herd prevalence of 22.8%.

None of the serologic tests used in this study can distinguish between brucellosis
antibodies produced due to an infection and those generated after Br. abortus strain-l9
vaccination (Omer et al., 2001). Although 4 of the herders surveyed reported that their
cattle had been vaccinated for brucellosis in the past, we received assurance from the
Bayankhongor Aimag (Province) veterinarians that no brucellosis vaccines (Br. abortus
Strain-19 or Br. melitensis Rev-1) had been administered or distributed in Bayankhongor
since 1988. Due to the centralized system of veterinary service delivery, the state
monopoly on vaccine production and strictly enforced vaccine importation bans, it is
most likely that the practice reported by the Aimag veterinarians is correct. Local soum
(district) veterinarians were also interviewed about brucellosis vaccination practices and
they conﬁrmed that brucellosis vaccination had not been performed in Bayankhongor
since at least 1990.

With the potential for false positive reactions due to the presence of vaccine-

induced antibodies eliminated, conﬁdence in the speciﬁcity of our screening tests is

70

increased. False-positive results in the CFT and CARD test can also occur as a result of
cross-reactive antibodies induced by a variety of Gram-negative bacteria (Emmerzaal et
al., 2002), however, the cattle screened in this study appeared healthy and there was no
reason to believe that widespread infection with pathogenic Gram-negative bacteria was
present in the study population. The speciﬁcity of the BRT can be reduced if samples are
grossly abnormal (curdled) or if the sample contains colostrum (OIE, 2002a). None of
the cattle tested in this study were in early lactation (no colostrum) and abnormal milk
samples were discarded and not subjected to testing. Calls have been made to review the
use of the BRT as a screening test for brucellosis due to the increasing size of dairy herds
in some regions of the world (Vanzini et al., 2001). Concerns surrounding the reduced
sensitivity of the BRT applied to bulk tank milk samples to which large (>100) numbers
of cows have contributed, are not applicable in this study. The number of cows
contributing to a milk sample in this study was always less than 30 and the test was used
on a “milk can” scale as it was originally intended (Alton and Jones, 1967).

B. Livestock Management

The general system of livestock management is similar across the Khangai
mountain region of Bayankhongor Aimag (Province). People extensively manage mixed
herds of livestock as mobile pastoralists. Although very little mixing of livestock occurs
at sale yards or in markets, herders do graze their livestock in communally owned
pastures. Veterinary service delivery, in terms of coverage and interventions, is relatively
uniform and livestock are raised primarily for family consumption. The occurrence of
the yak as the primary milking animal in the region and the relative similarity of livestock

management systems across the study area allowed us to estimate the brucellosis sero-

71

prevalence for the region as a whole. Differences do, however, exist from herd to herd
in terms of production parameters and abortion reporting that will allow us to compare
characteristics of herds with BRT-positiVe milk or CARD-positive animals to those herds
classiﬁed as brucellosis-seronegative. These differences will be further analyzed with
the use of univariate and multivariable statistical techniques reported in chapter 3.

C. Human/Livestock Interaction and Human Health

The heads of herder households interviewed represented a normal distribution of
individuals in terms of age and educational level with an equal number of males and
females surveyed. As livestock herders, all of the household members have regular and
repeated contact with livestock. Potentially hi gh-risk interactions with livestock include
home slaughter of multiple livestock species and hand milking of cattle, sheep and goats.
Generally adults are involved in livestock slaughter with men involved in the killing,
skinning and deboning of the animals and women in the cleaning and preparing the
gastrointestinal tract and internal organs. Both adults (primarily women) and children are
observed milking cattle, sheep and goats. The questionnaire did not include speciﬁc
information on human exposure to livestock placentas, uterine ﬂuid or aborted material
but weak neonates, especially kids and lambs, were brought into the dwellings of 39% of
the herders interviewed.

Of greater concern in this population was the high percent of interviewees who
reported consuming raw cow’s milk (12%) and preparing dairy products with raw milk
(33%). It is probable that the habits of the entire household would reﬂect the actions of
the head of household in terms of raw milk consumption and that all members of the

household would consume dairy products produced from the raw milk. Although milk is

72

routinely boiled immediately after collection these responses indicate that there is the
potential for a signiﬁcant level of human brucellosis exposure through the consumption
of raw milk and dairy products in this population.

Of further concern was the high number, 26/97 or 27%, of interviewees who
reported that they had a history of brucellosis diagnosis. Fourteen of 20 individuals with
a history of brucellosis diagnosis (54%) who responded to the follow-up question about
the timing of their brucellosis diagnosis were diagnosed within the previous 12 months.
The fact that laboratory testing was only reported to have been performed on a subset of
the individuals who had been diagnosed (7 of 26 or 27%) calls into question the accuracy
of the diagnosis. In addition, none of the individuals reported receiving any treatment for
brucellosis.

As an alternative to “brucellosis history”, questioning on the occurrence of non-
speciﬁc signs of illness was employed and results indicated that a portion of the
interviewees had experienced the clinical signs often associated with acute (16%) and
chronic (20%) brucellosis within the previous 12 months. The variables characterizing
livestock/human interaction, education level, brucellosis knowledge, and livestock herd
brucellosis status are compared in individuals who did and did not report a history of
brucellosis diagnosis or clinical signs and symptoms associated with brucellosis in
chapter 4 and analyzed with univariate and multivariable statistical techniques.

D. A Herder Knowledge of Brucellosis

Less than 25% of the herders interviewed could identify the clinical signs and
symptoms associated with livestock brucellosis or means of livestock brucellosis

prevention and transmission. A smaller percent were aware of the risk of brucellosis to

73

human health and ways to prevent zoonotic transmission. Due to this demonstration of a
lack of brucellosis knowledge, educational pamphlets describing brucellosis as a zoonotic
disease and means of disease prevention, were created and distributed to all of the herders
in the study area (both those selected and those not selected for inclusion in the study)

through their local veterinarians.

2.5 CONCLUSION

The prevalence of brucellosis-ser0positive cattle herds in the Khangai mountain

region of Bayankhongor, Mongolia, is undoubtedly associated with livestock production

losses and probably associated with the occurrence of clinical signs and symptoms
suggestive of brucellosis reported by the humans in the region. Whether the more
conservative estimate of herd prevalence based on serial serum testing (22.8%) and BRT
and CARD agreement (24.3%), or the less conservative estimates of prevalence based on
the BRT (41.3%), CARD (39.2%) and the combination of BRT and CARD test
information (53%), the data support concerns about the re-emergence of brucellosis in
Mongolia (Ebright et al., 2003; Roth et al., 2003; WHO, 2001) and its classiﬁcation as a
major public health issue.

Although this study represents data from only a small portion of Mongolia’s
livestock population there is no reason to believe that the occurrence of brucellosis in
cattle is restricted to this study region. Historical publications (J ezek et al., 1972; J ezek
et al., 1974; Otgon, 1968; Pinigin et al., 1968b; Tserendash, 1980, 1984; Tserendash and
Shumilov, 1970) and more recent reports (OIE, 2002b) indicate that brucellosis is found

throughout Mongolia. Additionally, the individual animal data collected in this study

74

supports ofﬁcial reports that highlight Bayankhongor as a high prevalence Aimag
(Province), but also indicate that the ofﬁcial reports may be underestimating the
prevalence of brucellosis in Bayankhongor and possibly the nation as a whole.

Since the elimination of Mongolia’s systematic brucellosis control and
surveillance program in the late 1980’s, efforts have been made to recreate and fund a
new brucellosis control program in Mongolia. The World Health Organization (WHO)
provided technical support for the preparation of a National Brucellosis Control Plan in
1999, that called for the mass vaccination of ruminants and was due to begin in 2000
(Mikolon, 2000; WHO, 20003; Zinsstag et al., submitted 2004). Despite the lack of
international funding for the program and signiﬁcant set backs posed by the combination
of drought and severe snowstorms in 1999 through 2002 (UNDP, 2001), the Mongolian
government has pledged to follow a brucellosis control strategy that consists of a lO-year
vaccination program designed to vaccinate of all adult bovine, caprine and ovine herds in
the country twice within 6 years (Zinsstag et al., submitted 2004). The 10-year
vaccination plan was due to be launched in Bayankhongor Aimag (Province) 1-2 years
alter the completion of this cross-sectional study. The data collected in this study will
provide important baseline information for the evaluation of the success of the
vaccination program in the Khangai mountain region of Bayankhongor. The
identiﬁcation of risk factors for both livestock and human brucellosis, discussed in
chapters 3 and 4, will contribute to the understanding of the epidemiology of brucellosis
in Mongolia today and enhance the effectiveness of current and future brucellosis control

and public health education programs.

75

Table 2.1. The 5,062 Herder Household Sampling Frame in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia, and the Relative
Proportion of Herder Households from Each Stratum.

 

Stratum (Soum/Bagh) No. Herder Households Relative Proportion (%)

 

Gurvanbulag 755 14.9
Zag 562 . 11.1
Jargalant 1183 23.4
Galuut 1375 27.2
Erdentsogt 989 19.5
Shargaljuut 198 3.9

TOTAL 5,062

 

76

Table 2.2. The Number and Percent of Herder Households Sampled in Each Stratum
and the Overall Proportion of Participation.

 

 

Stratum No. Herder No. Proportion (Herds/Stratum)
(Soum/Bagh) Households Sampled Sampled (%)

Gurvanbulag 75 5 15 2.0

Zag 562 12 2.1

Jargalant 1 183 22 ' 1.9

Galuut 1375 27 2.0

Erdentsogt 989 20 2.0

Shargaljuut 198 7 3.5

TOTAL 5062 1037105 = 98% Participation Proportion

 

77

Table 2.3. The Prevalence of Brucellosis Ring Test (BRT)-Positive Cattle Herds in
East Stratum and the Adjusted Overall Estimated Prevalence and 95%
Conﬁdence Intervals of Brucellosis-Seropositive Cattle Herds in the
Khangai Mountain Region of Bayankhongor Aimag (Province), Mongolia.

 

 

Stratum No. Herds No. Positive Prevalence 95% C.I.
(Soum/Bagh) Tested Herds (%)

Gurvanbulag 12 7 58.3

Zag 10 3 30.0

J argalant 22 9 40.9

Galuut 27 12 44.4

Erdentsogt 20 6 30.0

Shargaljuut 7 3 42.9

TOTAL 98 40 41.2% 19.8%

 

78

Table 2.4. A Comparison of the Adjusted Estimates of the Prevalence of Brucellosis-
Seropositive Cattle Herds in the Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia, Based on the Results of the
USDA Brucellosis Ring Test (BRT), the USDA Brucellosis Card Test
(CARD), the CARD and BRT (Both Positive), the CARD and the
Mongolian Complement Fixation Test (CFT) in Series and the CARD or

 

 

BRT (Either Positive).
Test No. +/Total Herds Prevalence of Positive Herds 95% C. I.
BRT+ 40/98 _ 41.2 % (31.4, 51.0)
CARD+ 39/101 39.2 % (29.9, 48.4)
CARD+ and BRT+ 25/96 26.5 % (17.8, 35.1)
CARD + and CFT+ 23/101 22.7 % (14.6, 30.8)
CARD+ or BRT+ 54/103 . 53.0 % (43.5, 62.6)

 

79

Table 2.5. The Degree of Concordance (Kappa Statistic) Between the Brucellosis
Ring Test (BRT) and the Brucellosis Card Test (CARD) as a Herd-Level
Brucellosis Testing Method.

 

CARD Positive CARD Negative Total

BRT Positive 25 14 39
BRT Negative ll 46 57
Total 36 6O 96

Kappa = 0.45; 95% CI. (0.27, 0.64)

80

Table 2.6. Individual-Animal Brucellosis-Seroprevalence in Cattle Tested in the
Khangai Mountain Region of Bayankhongor Aimag (Province), Mongolia,
with the USDA Brucellosis Card Test (CARD) and conﬁrmed with the
Mongolian Complement Fixation Test (CFT).

 

 

Stratum Number Number Positive Proportion Positive
(Soum/Bagh) Tested CARD (CFI‘) CARD (CFT)
Gurvanbulag 85 12 (8) 14.1% (9.4%)
Zag 49 6 (2) 12.2% (4.1%)
J argalant 118 13 (7) 11.0% (5.9%)
Galuut 130 25 (14) 19.2% (10.8%)
Erdentsogt 93 10 (2) 10.8% (2.2%)
Shargaljuut 38 5 (3) 13.2% (7.9%)
TOTAL 513 71 , (36) 13.8% (6.2%)

 

81

Table 2.7. The Reported History of Brucellosis Diagnosis, Testing and Treatment in
the Surveyed Heads of Herder Households in the Khangai Mountain
Region of Bayankhongor Aimag (Province), Mongolia.

 

Reported History Reported History Reported History Reported History

of Brucellosis of Brucellosis of Brucellosis of Brucellosis

Diagnosis Diagnosis Within Testing Treatment
Previous 12 months

 

 

Overall Overall Diagnosed Overall Diagosed Overall
26/97 14/97 14/26 11/97 7/26 0/0
(26.8%) (14.4%) (53.8%) (11.3%) (26.9%) (0.0%)

 

82

 

 

Table 2.8. Reported Occurrence of Clinical Signs and Symptoms Suggestive of
Acute, Chronic, and General Brucellosis in the 100 Surveyed Heads of
Herder Households in the Khangai Mountain Region of Bayankhongor,
Mongolia.
Acute Chronic General
(Acute + Chronic)
Chills, Fever, Joint Pain, Stiff Joint(s), Chills, Fever,
Headache, or Swollen J oint(s),‘ or Headache, Night
Night Sweats Lower Back Pain Sweats, Joint Pain,
Stiff Joint(s), Swollen
J oint(s), or Lower
Back Pain
Prevalence 16% 20% 24%

 

83

Chapter 3

MAN AGEMENT—RELATED RISK FACTORS FOR BRUCELLOSIS IN
CATTLE/Y AK HERDS IN THE KHANGAI MOUNTAIN REGION OF

BAYANKHONGOR AIMAG (PROVINCE), MONGOLIA

3.0 STRUCTURED ABSTRACT

Objective: This component of the study was undertaken to determine the speciﬁc
management practices associated with increased risk of brucellosis-ser0positive status of
cattle (predominantly yak) herds in the Khangai mountain region of Bayankhongor
Aimag (Province), Mongolia. The primary objective was to examine livestock
management factors within the broad categories of grazing and watering, migration
ﬁequency, herd species composition, contact with outside herds, veterinary care, age
structure, and size of herd for associations with cattle herd brucellosis status. The
secondary objective was to determine whether herd brucellosis seropositivity affected
calving or neonatal mortality rates on the herd level or was associated with the prevalence
or occurrence of livestock abortions. .

Design: Cross-sectional study

Sample Population: A stratiﬁed random sample (n=105) of the 5,062 livestock herds in
5 contiguous soums (districts) and 1 bagh (sub-district) of the Khangai mountain region

of Bayankhongor Aimag (Province), Mongolia.

84

Procedure: A livestock management questionnaire was administered as an in-person
interview to the head of each selected herder household. Cattle herd brucellosis-
seroprevalence data was based on testing with the Brucellosis Milk Ring Test (BRT) and
the Brucellosis Card Test (CARD). Livestock management data and reports of disease
occurrence (abortions and neonatal deaths) were collected from the questionnaire.
Descriptive statistics were generated and the data were analyzed with univariate and
multi-variable methods. Attempts were made to identify herd-level and individual-level
risk factors for brucellosis—seropositivity and assess the association between brucellosis-
seropositivity and cattle health and production parameters.

Results: The stratiﬁcation-adjusted prevalence of brucellosis-ser0positive cattle herds in
the study region was 53% (95% CI. +/- 9.5%) based on having either/or a positive BRT
or CARD test. The adjusted prevalence based on using the BRT and CARD test in
parallel was 24.3% (95% CI. +/- 8.1%). A signiﬁcant association was found between
larger cattle herd size and an increased risk of brucellosis seropositivity. Brucellosis herd
ser0positive status was associated with an increased number and prevalence of cattle
abortions.

Conclusions: Bovine brucellosis is an important livestock disease in the Khangai
mountain region of Bayankhongor Aimag (Province), Mongolia, and the ﬁndings of this
study support the concerns surrounding the re-emergence of brucellosis in Mongolia. The
reproductive losses (bovine abortion and infertility) affect both the supply of replacement
stock and milk production, which in turn signiﬁcantly impact the food security and
economic viability of herder households. The human consequences of the circulation of

brucellosis in livestock are a concern given the zoonotic nature of this disease.

85

3.1 INTRODUCTION

Brucellosis is an infectious disease of livestock and is one of the most widespread
zoonotic diseases in the world (Corbel, 1989). The worldwide distribution of endemic
brucellosis includes Mongolia, a sparsely populated country (1.5 persons/kmz) in Central
Asia where livestock herding is the backbone of the county’s economy and an important
part of its cultural heritage (Mongolia, 2001c; OIE, 2002b). Major efforts have been
undertaken around the world to control brucellosis due the signiﬁcant economic losses
associated with the occurrence of the disease in domestic livestock and the often
debilitating disease associated with human infection (Halling and Boyle, 2002; Ragan,
2002; Reddy, 2003). Clinical signs of brucellosis in domestic livestock include late term
abortions, decreased milk production and lowered fertility in both males and females
(Drost and Thomas, 1996).

Particular biovars of the species of Brucella melitensis are associated with
pathogenicity in different domestic animals. The disease in cattle is most commonly
caused by the biovar abortus (Reddy, 2003). The vast majority of human infection is the
result of direct or indirect transmission of the disease from infected livestock through the
ingestion of contaminated milk or dairy products or through direct contact with aborted
material or uterine ﬂuid from infected animals (Nicoletti, 1989).

Historically, Mongolia has had a high incidence of brucellosis in its human
population (Otgon 1968; J ezek Rusinko et a1. 1972; J ezek, Rusinko et al. 1974) as a
consequence of the high prevalence of the disease in domestic livestock (Beulig et al.,
1969; Jezek et al., 1972; Jezek et al., 1974; Otgon, 1968; Pinigin et al., 1968a; Pinigin et

al., 1968b; Tserendash, 1980, 1984; Tserendash and Shumilov, 1970; Zherikova et al.,

86

1972). Brucellosis prevention, through livestock vaccination, began in Mongolia in the
1960’s (Corbel, 1989; Tserendash, 1980). A second prevention program, lead by the
World Health Organization (WHO), began in the mid-1970’s and focused on the
vaccination of small ruminants (Corbel, 1989). The former Soviet Union continued to
support a brucellosis livestock vaccination program in Mongolia (Brucellosis abortus
strain-19 vaccine in cattle and Br. melitensis Rev-l vaccine primarily in small ruminants)
through the late 1980’s.

In the later half of the 20th century, the risk factors associated with livestock
brucellosis in Mongolia included the establishment of state dairy farms and the more
intensive management of cattle and small ruminants (Otgon, 1968). In addition the
movement of cattle across the country was associated with the regional spread of
livestock brucellosis (Denes, 1997; Otgon, 1968).

Initially the brucellosis control programs were deemed a success by Mongolian
scientists (Tserendash, 1980) and by WHO ofﬁcials who reported that the annual
incidence of the disease in humans dropped to less than 1 case/ 10,000 in the early 1980’s
(Mikolon, 2000; Roth et al., 2003). Despite evidence in the early 1980’s of a signiﬁcant
reduction in the incidence of brucellosis in Mongolia, concerns have been raised recently
about the re-emergence of the disease in the country (Ebright et al., 2003; J oshi, 1991;
Roth et al., 2003; WHO, 2000b, 2001, 2003). Brucellosis surveillance in Mongolia from
1990, until 2001, has consisted of unsystematic sampling of individual livestock and
sporadic herd level testing (Roth et al., 2003). These data do not allow one to quantify
the regional prevalence of disease or conﬁrm the re-emergence of brucellosis, however,

they do demonstrate that livestock brucellosis continues to persist in Mongolia.

87

The persistence and possible re-emergence of brucellosis as a major public health
problem in Mongolia has been blamed on the collapse of the veterinary service delivery
system and public health infrastructure in Mongolia which began in the late 1980’s
(Ebright et al., 2003; Mikolon, 2000). The Mongolian veterinary system in the 1990’s
reﬂects the changes that were experienced by the country as a whole as Mongolia
transitioned from a centrally planned economy and socialist system of government to a
democratic system with a market economy (Ebright et al., 2003; Rossabi, 1999). The
combination of the abrupt loss of economic aid from the former Soviet Union, the
privatization of livestock in 1990, and the privatization of veterinary service delivery in
1999, meant the end of large scale, government subsidized, livestock health campaigns
including comprehensive livestock vaccination programs and systematic surveillance for
brucellosis (Mikolon, 2000).

Recently collected data on the prevalence, incidence and distribution of
brucellosis in Mongolia’s livestock and human population is largely unavailable and no
published reports of the risk factors associated with disease exist. Ofﬁcial government
reports provide some information on disease occurrence but the sampling methods
employed are unknown and little insight into the level and distribution of disease can be
gained. The aim of this study was, therefore, to systematically collect data on the
prevalence of brucellosis in livestock in Mongolia and examine livestock management
practices used in the region. In addition an effort was made to quantify the production
level of the herd and examine the data for associations between cattle herd brucellosis

status and the clinical signs and symptoms often associated with bovine brucellosis. The

88

decision was made to focus on the Khangai mountain region of Bayankhongor Aimag

(Province), where brucellosis was reported to occur.

3.2 HYPOTHESES

1) There are speciﬁc management practices associated with increased risk of
brucellosis-ser0positive status of cattle herds in the Khangai mountain region of
Bayankhongor Aimag (Province), Mongolia, within the broad categories of: grazing and
watering, migration frequency, herd species composition, contact with outside herds,
level of veterinary care, age structure and size of herd.

2) Herd brucellosis-seropositivity is a risk factor for the occurrence of abortions, a

reduction in calving rate, and increased rates of neonatal mortality.

3.3 MATERIALS AND METHODS

A. Study Design and Herd Selection

A stratiﬁed random sample (n=104) of the 5 ,062 herds in the 5 contiguous
Khangai soums (districts) of Gurvanbulag, Zag, Galuut, J argalant, Erdenstogt, and the
bagh (sub-district) of Shargaljuut were selected for sampling. The study area, sample
size calculation and randomized herd selection methods are described in chapter 2
(Section 2.3).

A cross-sectional survey was used to determine the association between the
prevalence of brucellosis-ser0positive herds in the Khangai mountain region of
Bayankhongor Aimag (Province) and livestock management practices as described

below.

89

B. Data Collection Team and Herd Location

The data collection team consisted of two veterinarians, a physician, a driver, and
a volunteer data recorder. The data collection team met with local ofﬁcials and the soum
(county) veterinarian in each soum center. The selected herder households were located
with the assistance of the soum veterinarian and local ofﬁcials. The soum veterinarian
accompanied the data collection team on all herd visits and introduced the team to the
herder household members. The Mongolian bilingual (English and Mongolian)
collection team member explained the project and informed the herders that their
participation in the study was voluntary. All team members were available for questions.
Data collection proceeded only if consent was obtained.

C. Brucellosis Testing of Cattle Herds

The brucellosis status of the selected cattle herds was determined by using
serological tests on milk and serum samples obtained from each herd. The milk samples
were tested for the presence of Brucella antibodies using standard B. abortus strain 1119-
3 BRT antigen provided by the United States Department of Agriculture (USDA),
Animal and Plant Health Inspection Service (APHIS), National Veterinary Services
Laboratories (NV SL), Ames, Iowa, USA.6 Serum samples were collected ﬁ'om
approximately ‘/2 of the individual female cattle > 6 months of age in each herd and tested

with the USDA brucellosis card test kit (CARD)7. The milk and serum samples were

 

6 The Brucellosis Milk Ring Test (BRT) is an internationally recognized method for detecting brucellosis in
a herd of cattle. Blue haematoxylin-stained Brucella abortus antigen is used to detect Brucella antibodies
in the milk of affected cows Alton, G., Jones, J ., 1967, Laboratory Techniques in Brucellosis. World Health
Organization, Geneva.

7 The Brucellosis Card Test Kit (CARD) with Rose-Bengal stained, buffered Brucella abortus strain 1119-
3 antigen is produced by the USDA, APHIS, NVSL in Ames, Iowa, USA. The CARD test is a modiﬁed
serum agglutination test that detects the presence of antibodies against Brucella and was originally
developed in the United States to meet the need for a rapid ﬁeld test for the serological diagnosis of
brucellosis in the ranch or stockyard setting Prior, M.G., Niilo, L., Reeker, W.H., 1975, Use of the

90

collected and tested using standard international protocols (OIE, 2002a) as described in
chapter 2 (section 2.3). Serum ﬁom the CARD-positive individuals was submitted to the
Mongolian State Central Veterinary Diagnostic and Sanitary Laboratory for conﬁrmatory
testing with the Complement Fixation Test (CFT)8.

D. Case Definitions

Two primary case deﬁnitions of a brucellosis-positive herd were used in this
component of the study. A cattle herd was classiﬁed as “Positive” if either a BRT-
positive result was recorded for the mixed raw milk sample or at least one individual
animal was classiﬁed as a reactor to the CARD test. A herd was classiﬁed as “Both
Positive” if the BRT and CARD tests, were positive. These two case deﬁnitions of a
brucellosis-ser0positive herd were used in the univariate and multi-variable data analyses
described below (section 3.4). The individual animal sero-prevalence data were used to
create two alternative case deﬁnitions. A herd was classiﬁed as “> 1 CARD-positive”, if
more than 1 individual in the herd was identiﬁed as a CARD-test reactor. A herd was
classiﬁed as “CFT-conﬁrmed CARD-positive” if an individual in a herd tested with the
CARD and CFT in series was positive on both tests. These two secondary case
deﬁnitions were used as alternative outcome variables in the multi-variable logistic

regression analyses described below (section 3.4).

 

brucellosis card test for screening cattle in Saskatchewan. Canadian Journal of Comparative Medicine 39,
107-109.

8 The complement ﬁxation test is a serologic test designed to determine the Brucella titer of the serum
under test. The CFT utilizes Brucella antigen, conrplement and sheep red blood cells to measure the degree
of hemolysis. The titer of the serum under test is read as the highest dilution showing 50% hemolysis. In
cattle a titer of 1:40 indicates an infected animal and one of 1:20 is suspicious, indicating a need to retest in
2 weeks Alton, G., Jones, J ., 1967, Laboratory Techniques in Brucellosis. World Health Organization,
Geneva.

91

E. Collection of Data Relating to Possible Risk Factors for Brucellosis-
Seropositive Cattle Herds

A 3-page questionnaire was conducted as an in-person interview with an adult
member of the herder household involved in livestock rearing. The interview was
conducted by a bilingual (Mongolian and English) native Mongolian speaker. (See
translation of questionnaire in Appendix B). The in-person interview and collection of
bovine milk and serum samples for brucellosis testing were performed during the same
herd visit. The livestock management data collected included: herd composition
(species/numbers of livestock and young/year), herding practices (number of
migrations/year, mixing with other livestock at water points, winter and night shelters), a
history of veterinary care (brucellosis vaccination and other interventions) and livestock
milking practices. In addition data were collected on the occurrence and timing of
livestock abortions, male infertility and neonatal mortality. The ﬁnal section of the
questionnaire assessed herder awareness of characteristics of livestock brucellosis.

Interview responses were recorded in Mongolian and later translated into English.

3.4 DATA ANALYSIS

A. Univariate Analysis of Herd and Individual-Level Data

Data from the livestock management questionnaires and BRT, CARD and CFT
test results were entered into Microsoft Excel spreadsheets (Microsoft Ofﬁce 2000).
Data were subsequently analyzed using the SAS software program (SAS version 8.2,
Cary, NC: SAS Institute, Inc. 1999-2001). Descriptive statistics were computed for the
risk factors and dependent variables of interest. Cattle abortion rates, calving rates and

neonatal mortality rates were computed for each herd. Multiple herd mixing variables

92

(water, shelter, night) were collapsed into 1 variable, “mix”. A poverty variable was
created based on the number of livestock per herder household. If the herder household

owned less than 150 head of livestock they were classiﬁed as “poor”.

Individual animal data (age, sex, breed and CARD test results) were transferred
from ﬁeld data sheets and entered into Microsoft Excel spreadsheets (Microsoft Ofﬁce
2000). Descriptive statistics were computed for the individual animal level risk factors.
The individual animal data were used to calculate the within herd prevalence of
brucellosis reactors and the average age of cows in the herd. These two calculated herd-
level characteristics were incorporated into the primary data set as new herd-level

variables.

The t-test (PROC TTEST) was used to test for signiﬁcant associations between
herd brucellosis status and continuous risk factors such as herd size, number of individual
species and numbers of migrations per year. The Wilcoxon nonparametric test (PROC
NPARIWAY) was used in addition to the t-test for those continuous risk factors with a
skewed distribution. Pearson’s X2 test (PROC FREQ; “TABLES/CHISQ”) was used to
test for signiﬁcant associations between herd brucellosis status and categorical risk
factors such as mixing with other herds, the occurrence of livestock abortions, and herd
location. The X2 test for trend (PROC FREQ; “TABLES/CHISQ”; MH) was computed
when the categorical risk factors represented ordinal levels. The t-test and Pearson’s X2
test were also used to test for signiﬁcant associations between the individual animal

brucellosis status continuous and categorical risk factors such as age and breed.

93

B. Multi-Variable Analysis of Herd-Level Data

The original complete herd-level data set contained 44 variables. A subset of the
original variables were considered for inclusion in the full logistic regression model.
Variables excluded from the multi-variable logistic regression analysis included those
variables with a large proportion of missing responses (>20% missing); redundant
variables; and those deemed not to be biologically relevant to the occurrence of
brucellosis. Eleven risk factors or disease indicators were examined in the full logistic
regression model. Soum (county) was entered as a class variable. No variables were
forced into the model and no interactions were included. Backward elimination was used
to model the relationship between cattle herd brucellosis status (classiﬁed by “BRT and
CARD-positive”, by “BRT or CARD-positive”, by “>1 CARD-positive” and by “CFT-
conﬁrrned CARD-positive”) and livestock management practices. Observations with

missing data were removed during the modeling procedure.

3.4 RESULTS

A. Description of Sample Population

The head of the herder household volunteered to be interviewed for the livestock
management questionnaire portion of this study in 103, of the 104 herds (99%),
contacted. Information on cattle herd brucellosis status was available for the same 103
herds. The total number of cattle herds found to be seropositive for brucellosis based on
either the Brucellosis Milk Ring Test (BRT) or the Brucellosis Card Test (CARD) was 54

of 103, an adjusted prevalence of 53% (95% CI. +/- 9.5%). The number of cattle herds

94

found to be ser0positive on both the BRT and CARD test was 25 of 96, an adjusted cattle
herd prevalence of 26.5% (95% Cl. +/- 8.7%).

The completeness of data collected in each of the interview sessions varied. The
variables with the greatest amount of missing data included the following: reported
abortion in ewes (44% missing), reported abortion in does (3 7% missing), number of
reproductive-age female cattle (35% missing), number of goats (27% missing), number
of sheep (26% missing), number of cattle (22% missing), and total number of calves
(21%). The data for the remaining variables covered in the livestock management survey
were 80% to 100% complete. The primary explanations for the missing data were
incomplete interviews due to time constraints and recording errors. The degree of
missing data on livestock numbers also affected the completeness of some of the
calculated variables including calving rate, calf survival, and cattle abortion rate.

B. Descriptive and Univariate Analyses of Herd Management Risk
Factors

The data on total number of cattle (mean = 20.26), total cows or reproductive-

age females (mean = 10.93), and total milking cows (mean = 7.71) indicate the presence
of an association between herd size and brucellosis seropositivity in this population. The
mean number of both milking cows and adult females was statistically signiﬁcantly
higher at the p<0.05 level in the cattle herds classiﬁed as brucellosis-ser0positive by both
or either the BRT and CARD tests (See Table 3.1). The same trend was observed in the
total number of cattle but the difference in means was not signiﬁcant at the p=0.05 level,
(See Table 3.1). The data on cattle herd size were also categorized and analyzed using the

X2 test for trend. A signiﬁcant (p=0.0013) trend of increasing risk associated with

95

increasing herd size was found (See Table 3.2). An association was not found between
the total number of livestock (cattle, sheep and goats) and cattle herd brucellosis status.

All of the herders interviewed reported that they milked cattle. Neither the
presence or the number of sheep, nor the presence or the number of goats in the herd was
signiﬁcantly associated with cattle brucellosis seropositivity. Milking goats appeared to
be a protective factor for brucellosis in cattle herds (O/ 12 herds that milked goats were
sero-negative, compared to 22/54 that did not milk goats), however, the association was
not found when cattle herds were classiﬁed as brucellosis-ser0positive based on either a
positive BRT or CARD test (See Table 3.2).

No statistically signiﬁcant (p-value <= 0.05) associations were found between
cattle herd brucellosis seropositivity and any other of the livestock management or herd
characteristics examined, including the degree of mixing with other livestock, number of
migrations per year, level of veterinary service, reported use of brucellosis vaccine, breed
of cattle, the socio-economic status of the herder household or the level of knowledge of
livestock brucellosis symptoms and prevention (See Table 3.1 and 3.2).

No association between individual animal brucellosis sero-status and cattle breed,
pure yak (Bos Grunniens) or yak crossed with “Mongol” cattle (Bos Taurus), was found.
(See Table 3.3). The mean age of CARD-positive individual cattle was signiﬁcantly
higher (p=0.0253) than the mean age of CARD-negative individuals (See Table 3.3). The
X2 test for trend, however, did not show a signiﬁcant increasing risk of CARD-positive
brucellosis status with increasing age class (See Table 3.3). In addition “average cow
age” as a herd-level risk factor was not associated with cattle herd brucellosis

seropositivity.

96

C. Descriptive and Univariate Analyses of Disease Indicators

A statistically signiﬁcant association was found between cattle herd brucellosis-
seropositivity (positive BRT and CARD) and reported cattle abortion (p=0.0017),
number of cattle abortions (p=0.0429) and percent of cattle abortions in a herd
(p=0.0077), (See Table 3.4 and 3.5). The same trend was seen in those herds classiﬁed as
brucellosis-ser0positive by either or the BRT or CARD test but the associations were not
signiﬁcant at the p=0.05 level, (See Table 3.4 and 3.5). No association was found
between cattle herd brucellosis seropositivity and reported abortions in goats but a
signiﬁcant association was found between CARD and BRT-positive cattle herds and
reported abortions in sheep (p=0.0398). The only herd with a horse with draining sores
over the withers was from a brucellosis-ser0positive cattle herd.

No signiﬁcant associations were found between cattle herd brucellosis
seropositivity and the reporting of bovine neonatal deaths. No signiﬁcant differences in
the mean calving rates, calf mortality or survival was seen between brucellosis-
seropositive and negative herds, although the calculated calving rate was lower in
brucellosis-ser0positive herds, (See Table 3.5).

D. Multi-Variable Analysis of Management Risk Factors and Disease
Indicators Associated with Herd Brucellosis-Seropositive Status

The full multi-variable logistic regression model was based on 68 observations
(66% of cattle herds) and included 11 variables, (See Table 3.6). The number of
observations was reduced ﬁ'om the original 103 herds due to missing data on herd size.
The overall prevalence of brucellosis-serbpositive herds in the 68 herds included in the
logistic regression model was 22% for herds classiﬁed as positive if both the “BRT and

CARD” test were positive and 54% for herds classiﬁed as positive if either the “BRT or

97

CARD” test were positive. The brucellosis herd sero-prevalence calculated for the 68
herds included in the logistic regression model was the proportional to that calculated for
the 103 herds in the original sample; 22% vs. 23% for the “BRT and CARD”, and 54%
vs. 53% for the “BRT or CARD”, in the 68 and 103-herd sample, respectively.

The ﬁnal reduced model (Table 3.7) included 2 variables; “number of milking
cows” and “observed bovine abortion”. An increasing number of milking cows was
associated with an increased risk of cattle herd brucellosis-ser0positive status. The odds
of a herd being classiﬁed as brucellosis-ser0positive according to the “BRT and CARD”
and “BRT or CARD” deﬁnition was increased by 0.24 and 0.13 times, respectively, for
every 1 cow increase in herd size. The observation of bovine abortion in the herd in the
previous year was associated with an increased risk of cattle herd brucellosis-ser0positive
status in herds classiﬁed by the “BRT and CARD” and the “BRT or CARD” testing
methods. The increase in odds of classiﬁcation as brucellosis-ser0positive by the “BRT
and CARD” in herds in which bovine abortion had occurred was 25 (95% CL; 4.5 —
139.5) compared with herds reporting no bovine abortion. The increase in odds of
brucellosis-ser0positive classiﬁcation by “BRT or CARD” was 4.4 (95% CL; 1.2 — 16.3)
compared to herds that reported observing no bovine abortions in the previous year.

The full logistic regression model was also run with the two secondary case
deﬁnitions, “>1 CARD-positive” and “CFT—conﬁrmed CARD-positive” as the outcome
variable. The ﬁnal reduced logistic regression model with the two secondary case
deﬁnitions as the outcome variable also identiﬁed “the occurrence of bovine abortion”
and an “increasing number of milking cows” as risk factors for brucellosis-ser0positive

herd status. The odds ratios and 95% conﬁdence intervals are presented in Table 3.7.

98

3.5 DISCUSSION

A. Distribution of Brucellosis-Seropositive Cattle Herds

Serologic testing of cattle herds in the Khangai mountain region of Bayankhongor
Aimag (Province), Mongolia, revealed a high prevalence of brucellosis-ser0positive
herds (53.0% based on having either/or a positive BRT or CARD test and 26.5% based
on using the BRT and CARD test in series) in the region as well as a high prevalence of
individual reactors to the brucellosis card test (13.8%). This level of disease is similar to
the prevalence described in countries with no systemic brucellosis control programs and
similar systems of nomadic or extensive livestock management, including Mexico
(Sahnan et al., 1984), Eritrea (Omer et al., 2001), Syria (Darwish and Benkirane, 2001),
Jordan (Al-Talaﬂrah et al., 2003) and Chad (Schelling et al., 2003).

Brucellosis-ser0positive cattle herds were detected throughout the study region,
in each of the soums (districts) and bagh (sub-district) tested. A visual inspection of the
mapped points indicating the location of positive and negatives herds (See Figure 2.2) did
not reveal any degree of clustering, and herd location (soum) was not determined to be a
signiﬁcant risk factor for either cattle herd or individual animal brucellosis seropositivity.

B. Livestock Management Risk Factors for Cattle Herd Brucellosis
Seropositivity

Livestock management practices were largely consistent across the study region.
Yaks (Bos grunniens) were the primary livestock species in the Khangai mountain region
of Bayankhongor Aimag (Province,) yet almost all of the herds were composed of mixed
species. Herders migrated seasonally (approximately 4 times per year) in search of

pasture for their livestock, and all of the land in the region was commonly owned. The

99

uniformly extensive management system, and the largely unrestricted contact between
herds, resulted in the identiﬁcation of very few differences from herd to herd.

The univariate analyses of livestock management practices examined revealed
signiﬁcant associations between herd disease status and variables relating to cattle herd
size, especially the total number of adult female cattle and the total number of milking
cows. The mean number of total adult female cattle and total milking cows was
signiﬁcantly larger in the ser0positive herds. Increasing cattle herd size has often been
found to be associated with an increased risk of brucellosis infection (Nicoletti, 1980a)
which can be explained by the increase in cattle density that is often associated with
larger herds and leads to an increase in cattle to cattle transmission, the inability to
identify and remove infected cattle and the greater likelihood of the addition of
replacement cattle from outside sources.

The multi-variable logistic regression model conﬁrmed the trends seen in the
univariate analyses. With the exception of increasing herd size, none of the other risk
factors examined (presence of sheep or goats, mixing of herds, brucellosis vaccination,
veterinary care, herder knowledge of livestock brucellosis, herd location or migration
frequency) emerged as signiﬁcant risk factors in the modeling of disease in this
population.

The association between increasing cow herd size and the risk of brucellosis-
seropositive status was stronger (larger odds ratios and smaller p-values) when the herd
was classiﬁed as positive using the more conservative, and arguably more speciﬁc, case
deﬁnition of “both BRT and CARD positive” vs. the less conservative case deﬁnition of

“either BRT or CARD positive”. The association is likely weakened when the less

100

conservative and less speciﬁc case deﬁnition is used, due to misclassiﬁcation bias
resulting from false positive tests.

C. Cattle Brucellosis Disease Indicators

Bacteriologic culture and Brucella isolation and identiﬁcation was unavailable in
Mongolia at the time this study was conducted. To compensate for the inability to
culture-conﬁrm Brucella infection in the ser0positive herds, data were collected on the
prevalence of clinical signs (abortion, decreased milk production, orchitis, male
infertility, and neonatal weakness or death) associated with brucellosis in cattle (Reddy,
2003). None of the herders interviewed had observed or suspected any degree of male
infertility within their livestock herds, however, no systematic qualitative or quantitative
assessment of male fertility was practiced.

Signs of female infertility in the form of observed late term abortions were
reported frequently for cattle, sheep and goats. All of the cattle abortions were reported
to have occurred in January, February or March, the time period that corresponds to the
ﬁnal 100 days of gestation in this seasonally and naturally bred population. Univariate
analyses revealed a strong association between the occurrence, number and prevalence of
abortions and brucellosis-ser0positive cattle herds. The calving rate was lower in the
ser0positive cattle herds though the difference in the mean calving rate in the brucellosis
positive and negative herds was not signiﬁcant.

The multi-variable logistic regression model revealed that the occurrence of cattle
abortion is highly associated with brucellosis positive herd status with an adjusted odds
ratio of25.0 (95% CL, 4.5 — 139.5) and 4.4 (95% CI, 1.2 — 16.3) in the “BRT and

CARD” and “BRT or CARD” classiﬁed herds, respectively. A similarly strong

101

association was seen when the more speciﬁc case deﬁnitions for a positive herd were
used. Odds ratios of 14.8 (95% Cl, 2.7-81.8) and 10.2 (95% CL, 2.5 — 42.0) for the “>1
CARD-positive” and “CFT-conﬁrmed CARD-positive”, respectively, were strongly
supportive of a causal relationship between bovine abortion and brucellosis in these
herds.

Although other causes of abortion in this population could not be ruled out,
brucellosis is thought to be the main infectious cause of abortions in cattle in this region
of Mongolia (Mondongadis, 2000). It is possible that some of the late term abortions
noted were a consequence of nutritional deﬁcits related to a severe winter, or trauma
associated with migration from winter to spring pastures, however, the strong association
of reported abortion and herd brucellosis-ser0positive status points to Brucella infection
as an important cause of bovine abortion in this population. A strong association
between reported cattle abortions and herd brucellosis-ser0positive status has been
reported in other population based studies of cattle brucellosis (McDermott and Arimi,
2002) (Schelling et al., 2003). Although brucellosis infection could not be conﬁrmed
with bacteriologic culture in this study, it is clear that the seroprevalence data is
associated with the clinical signs and morbidity we would expect with Brucella infection
in cattle.

We were unable to conﬁrm the strain of Brucella circulating in cattle of the
Khangai mountain region of Bayankhongor Aimag (Province), Mongolia, within the
scope of this study, however, data on the occurrence of abortion in sheep and goats in the
sampled herds points toward the abortus biotype. A signiﬁcant association was seen

between reported abortions in sheep and the ser0positive status of the cattle herd, but not

102

between the reported occurrence of abortion in goats and the ser0positive status of the
cattle in the herds sampled. B. abortus is reported to have a low pathogenicity for sheep
and goats although it has been associated with a number of outbreaks in sheep (Nicoletti,
1980b). B. abortus infection in goats has only rarely been recorded (Corbel, 1989). The
fact that the same herds experienced abortions in their sheep and cattle but not their goats
supports the assumption that the observed abortions in cattle were due to B. abortus
infection.

D. Sew-Prevalence and Risk Factors Associated with Disease in
Individual Cattle

This study provided the opportunity to collect data on the prevalence of
brucellosis-ser0positive individuals in the herds selected for testing. The overall
prevalence was CARD test-positive individuals was 13.8% (71/513). The prevalence of
CARD-positive individuals that were also complement ﬁxation test (CFT)-positive was
6.2%. Although there were very few non-yak bovines in the sample, cattle breed did not
appear to be a risk factor for disease, as has been shown in other studies of brucellosis in
cattle (Crawford et al., 1990). The prevalence of ser0positive animals was greatest in the
2 to 4-year age range (27.3%) and the lowest in those less than 2 (4.5%). This
corresponds to previous ﬁndings that suggest that young cattle are less susceptible to
brucellosis infection than older sexually mature cattle (Crawford et al., 1990).

E. Limitations

The lack of bacteriological conﬁrmation of the serological ﬁndings in this study is
unfortunate, however, this population of cattle did not have a history of brucellosis
vaccination thereby eliminating the primary cause of false-positive results associated with

the testing methods used.

103

The stratiﬁed random sampling technique used to select cattle herds for testing in
this study allows us to extrapolate our ﬁndings on the prevalence of brucellosis-
seropositive herds, and the risk factors and clinical signs associated with the disease, to
the entire Khangai mountain region of Bayankhongor Aimag (Province), Mongolia. The
external validity of our ﬁndings beyond this region, however, is limited. Other
mountainous regions of the country where the yak is the primary livestock species, and
where the cattle density is similar, would most likely experience a similar level of bovine
brucellosis. The data collected in this study support the presence of a high prevalence of
brucellosis infection in the region. The past implementation of national brucellosis
control programs in Mongolia lead us to believe that livestock brucellosis is an important
disease across the country. It is also recognized that the prevalence of cattle brucellosis
and the risk factors associated with disease transmission and persistence, will vary from .
one ecological zone to another where the composition of livestock herds, cattle density
and general livestock management practices differ.

The cross-sectional nature of this study allows us only to describe the disease
burden in a snap shot in time. The patterns described reﬂect associations only, and do not
provide the data necessary to make ﬁrm inferences on temporal associations or cause and
effect relationships. The study was, however, conducted in the late summer in an attempt
to reduce the limitations of data obtained in a cross-sectional study. In this seasonally
bred population of cattle, the late summer represents the post-calving period when one
would expect the clinical signs and symptoms associated with bovine brucellosis to be
most apparent and the livestock management practices that would limit or promote the

spread of disease to have the most impact.

104

The limited herd-to—herd variability made the identiﬁcation of herd-level risk
factors for brucellosis seropositivity difﬁcult. It is possible that the livestock
management risk factors and the indicators of disease that were identiﬁed in this study
were associated not with bovine brucellosis-ser0positive status but another unidentiﬁed
factor. This possibility is unlikely given that the associations identiﬁed are biologically
plausible and consistent with the results of others. The data collected indicate the need
for a broader survey of livestock brucellosis in Mongolia and provide an interesting basis

for further research.

3.6 CONCLUSION

Bovine brucellosis is an important livestock disease in the Khangai mountain
region of Bayankhongor Aimag (Province), Mongolia. The direct losses due to bovine
abortion and impaired fertility are clear. These reproductive losses affect both the supply
of replacement stock and milk production, which in turn signiﬁcantly impact the food
security and economic viability of herder households. The human consequences of the
circulation of brucellosis in livestock are a concern and the ﬁndings of a further
investigation in this area will be presented in the following chapter (chapter 4). On a
national level, the presence of brucellosis in livestock impacts access to markets for both
live animals and livestock products. The national health program is undoubtedly
burdened, and human productivity is likely hampered by the zoonotic transmission and
incidence of human brucellosis infection in Mongolia.

The ﬁndings of this study indicate that concerns surrounding the re-emergence of

brucellosis in Mongolia (Ebright et al., 2003; Roth et al., 2003; WHO, 2001; Zinsstag et

105

al., submitted 2004) are well founded. The national brucellosis control program,
currently underway in Mongolia, will need to be evaluated as to its effectiveness in
reducing and eventually limiting the spread of brucellosis in Mongolia. Greater attention
should be paid to areas with higher herd prevalence, larger herds or an increased density
of livestock, as indicated both in this study of a population of cattle in Mongolia, and

historically in regions across the globe.

106

3:3 He: Ba 8285.» 5:3 2:95 05 830.com 8 3508 Damage—eta 05 mo 3: 05 33065...

 

 

25 a?“ a. ozawoz 93 3.4 3 ozamoz c8» an ac
$8.0 Lao 3N 2.4 8 ozznoa 33¢ 33 o: m; S 228a Eczema:
N32 No.8 2.. oénwoz own: 3% am oéowoz can:
Std ammo 3.8 3.8 mm nausea 85¢ .26 9.8 and M: Eamon 8285
8; cos we ozawoz can 33 : ozawoz 382$
«mad and and 0: mm 25.25 3.25 :3 Rd NE 3 358a ow... 38
2.... So we ozamoz NS. 3.0 on oénwoz 3:39:
886 O3 a; o; mm 338 $85 93 £6 32 am 3:38 ago Foe
own new mm oénmoz mam mg a. o>€moz 283
88.0 imam was :.2 mm ozzaoa 88¢ Nam mg 2.2 2 258a $835
on: 3: mm ozamoz 9.: a3: 3 ozamoz
9.25 :2 8.2 8.8 a. 328m ~83 am? 32 8.8 mm 228a 2:833
.— S dd :82 a 55.0 a 3 dd :82 a A520 3:5. in.
he 5:: 253m ES .55 2.23m
mums—02:” the: mung—035m— the:

 

.2920 can Emu Steam

238a x976 8 Se Runes: 328a x920 can .26 cauz x920 8 EB STE as 920

was .55 2.: main doom S @8on .m:om:o§ Aooﬁ>emv maﬁa Howcoﬁcwzam .«o :2on 58558 HwSﬁM 05 E
modem 90 958m ammo—Enacm Bum 8m BEBE xma 9320M “soEomewE x8633 305E300 mo mowing Boar—«>33 Ad «Bah.

107

 

 

$3 3.2 mm ozawoz 32 «4.3 on oéawoz
dado So on? 3.2 Na 8:58 £42 86 42: 3.: om 228a anoooz
2.8 when mm ozamoz 2cm Eda d ozamoz
Edd Ed Sim on? a. ozznom N83 dd «26 8.3 .8 gamed goodoz
a E dd :82 a 55.0 a E dd :82 a A920 .88... and
.8 5:: dead 23 .28 need

mmmO=OU=hm «who:

dado—.35m Eu:

 

 

32:80 dd 2%...

108

.m 55 $2 $550 26: £3 05 .3 $3.. dog? “do“ Homxo 355$ 05 89a banged voEdtm 05 oo ed: 05 82863 a.

 

 

amono mm; m d *mwomo 2o N d nannies; duo—303m
«Sumo nmo mm vm mm _ m .o Nvo 3 2 83.8w Scroooxw
E. d No m d .o mm om abovmo moo S 2 x8635 6593 52
$33 woo o o nommoo mod .2 o 380 M52
nooood moo m m roommo m mm o o 325. v52
.. -- mv om .. .. co mm 0:30 SEE
E. Ko m d .o F. on moomo $5 3 S 8085 $30
. mmwmo om.o mm mm vmomo moo to C 230.5 doosm
ooood go o d -- -- o o deem 88$
$92 dam .3 Z dam
.5:— .8 any .5:— 65 go
madam d_de=8=._m dﬁﬁm 38:93.5
2:57: L" 1.3: .3 zoned—5.....— o=_a>tn "X 6.5! .3 5:25lo ~33.”— v7.2

 

£920 2a Ammo Ramona

258a x920 s 56 «dunno: 2225 A920 Ea ammo dauz ”85.0 8 Se meta as 920

can Ham 2: $55 .oooN E 350,—. .mzomdoE 68383 maﬁa dowdoidwzmm mo dammed £8882 ﬁmSﬁM 05 E
doom oo dBﬁm ﬂdozooEm to: 0:30 So duodenum SEE Rotowoumo coon—om “condemned—Z Eu: oo dmdbﬁe onE>ED .N.n 035—.

109

.m 56 82 $550 23: £3 05 .«0 $3. 5:3 «m8 888 mLonmE 2: 80¢ 33530.5 qum-~ 05 mo 8: 05 8:863 ..

 

 

 

 

ﬂ a v w 9‘60 NYE
m w a v 958 m7:
3 w E N $58 36
wmood mm.m 2 m ﬂood a2: 2 _ 958 m;
2:9»-.— NX Engaging—32 .woz .2:— o=_u>.a «X .oumaoa=-_3=a2 .woz .mom «Em Eu:
65:. .8.— amoh NX
*ooooA Ed v m .5266 de o m cos—555 £82335
6089595 "”5903
*oooo._ cod ‘ o n {tweed mmé . o o mmmozooam mo 3838mm .

35:83 ﬂop—om
wovmd 86 _~ mm “.580; $6 2 m M 38395 Cm VV
mEonomsom “com
wooﬁd $.m m S .mnmmd $6 3 m 33.33 8m 98
armwazzxm .83 352.60on

.9 Z .2:— .N. Z .8.—

.Hmm .8 any .5:— 15 go

2.3% ﬁne—BEG 253m 38:85.5
03.9.; "N the: a: 5:250..."— oEgd "x Eel %a zone—:5; 38.5 “E”—

 

8:528 .3” 2...;

110

Table 3.3. Univariate Analysis (Xz-Test) of Cattle Breed, Cattle Age and Cattle Age
Class (X2 Test for Trend) as Risk Factors for Individual Bovine
brucellosis Sero-Positivity in a Population of Cattle in the Khangai
Mountain Region of Bayankhongor Aimag (Province), Mongolia, Tested
in 2000, Using the CARD Test (N= 516; Positive Cattle=71)

 

 

 

 

 

 

 

 

Breed Pos. Neg. Mantel-Haenszel X2 p-value
Pure (Bos grunniens) 69 424 0.5202 0.7559*
Cross (Bos taurus) 2 21

[1 Mean Age S.D. [t] p-value
Positive 71 6.15 2.76 2.28 0.0253
Negative 438 6.97 3.01
Age Class Pos. Neg. Mantel-Haenszel X2 p-value
(1-2 years) 1 22 2.95 0.0858

(>2-4 years) 21 77
(>4-6 years) 23 112
(> 6 years) 26 227

 

111

.m 55 m8— $§8 023 £30 0513 $3 5:? $2 888 Manama 2: Bob 3:338.» 32¢.-N 05 mo 3.: 05 3:385 ..

 

 

 

owmmd mm." 2 amommd cod 3 c mason— 3382
*wmvud K; m *wmmod . vmd m m 28922 noonm
336 mod n $990 93 2 m 30922 :80
336 mod 2 L506 no. 2 n m meow—en? 0:30
.5:— .S 20 .5:— ES 20
25:5 33:35: 2555 38:83:
2:5»...— NX Eu: .3 %uaoaaouh 2:919 «X 6.5: .3 xenon—Bum 33am awe

 

azgo Ea Em: mmumeom 328m x920 s Ea «museum 328m x920 Ba .28 eeuz A920
s 58 EYE use 920 e5 Em 2: was: .88 a ease 3882 .68383 $82 Homaoﬁamm .8
commom 58:32 EwSEM 05 5 35m («0 9.58m Emo=ooam Bum 0:30 Ho.“ £8885 0.80me we mmmbmé Saginaw .vd «Bah.

112

:33 Ho: 03 moo—85> 3:3 02a?“ 05 cameos—om 9 H552: 3835388 05 .«o um: 05 8306:?

 

 

 

NS 23 om 26802 So 8.0 t. 03532 and
$30 :8 35 Rd mm 238m $33 a; 85 m3 2 268m 3.26
w; 85 R ozsmoz 26 8.0 9. 3:802 :35
~22 86 mod :5 8 968m 203 $6 86 3o : 228m 26.x.
:5 8o 3 3:802 go 8.0 mm Eamoz 33%
ago .8; Rd wwd wm 328m 8mg. moo NS 3o 2. 328m . 26.x.
o; 86 R 3:802 mod 86 a. 3:802 825$.
33 m3 m3 cod 2 .2688 EEO Fm 3o 26 2 338m 235:.
a: one 9. 96802 m3 :3 S 3:802 225?.
$25 «2 SN 3o 5. 02:8.” 335 8a 2; NI 2 328m 2:82
a E d.m :82 a A520 a E dd :32 a 55.0 58m :5.

3 5:5 253m 23 .26 seam

ammo—Euah— Eu: m_me__ou=..m Eu:

 

.2920 Ea Em: Suﬁsm 328m x920 s .28 312...: 228m x920 us... .26 canz x920 8

56 8W7: as 920 as Em 05 was: doom 5 889 £882 $838.: was? howeoﬁﬁsm Co Soum
58552 Ewan; 2: 5 35m .«o 33m mmmozooEm who: no.“ $8865 0385 3033300 mo mombmﬁx oEmthdD .m.m 2...;

113

.m :05 $2 $5.00 0.6: 0:00 05 00 $3. 00:? 0000 80x0 mLBEm 05 80b 359303 0020-N 05 mo 00: 0% 0000065 ...

 

 

.525; Ed 0 m .5306 and c m :0::0>0E 000600ch
.583 80 0 5 :83 m3. 0 0 meoassm 30052
£8de mm; m ﬂ *mwomd m _ .o N 3 02005023 032—0095
33.0 3.0 mm vM mm _ md $6 9‘ 2 00E0m E§0~0>
E. Kd m 3 .o mm 0m $300 No.0 3 2 0.005034 0050? 0&2
mind 26 S cm 836 0~.o 3 2 E0005 $000
mmwmd omd mm mm 33.0 mod to .2 E0005 000.5
035.0 mod n 2 £506 50.: n m 003022
0:30 00t0a0m

.wo Z .m0m .w0z .00.—

EE 8 528 can Ea 528
0530 «60:03.5 0530 $02—0an ﬂotsam 01.5—
0251“ «X 9.0—.— 3 500.50.:— 0Ea>un La 9.0: .3 ionosvohm 33.8330

 

.2920 05 5:0 21000 3280 x920 a .500 enumeom 3E8.”
x920 Ea Em: emuz x920 s Em: .2:qu as E0 05 920 .30 2: 05...: doom a Bag .2308:
6003003 maﬁa Howcoidmxam m0 :0E0m £95.02 898an 05 5 038m 0>Emom000m 062—0095 0003 0900
00m Eouomm 03M 0000—0M0802Q 0:0 E080w0§2 030.0024 «0 6002 :0_000&0m oummwoq =an 05 E ﬂoaomm 03% dd 030,—.

114

:33 “on 2m 80593 can? can?“ 05 883% 8 3508 common/steam 05 we um: 05 83085.."

 

 

 

 

 

o3 omo 3 9:332 a? moo on oémwoz 3=ng
ovooo oo.~ o; o; No "Eamon ovooo a2 33 one oo 3:58 £60 335
o; mod Q oémmoz o3 moo E 02332 Eaton E
omooo Loo mod m: on ”summon mmmoo 3o 9: m3 8 «Ewan 823%:
o E do 532 a 556 o E do :82 a 556 388m in.
.8 5.8 2.35 E: .28 53m 32.5280
mumE—ouem 6.3m.— mmmc——ou=.—m— the:
J o snowmaom
ozoo o m> m Soho o m> m m @38on
ago o 2, v 833 o m> o o. 323
oEo o 2, m Soho o 9 m m oﬁamﬁ
33o o 2, N was o e, N m mﬁ
oomoo o m: ooooo o m: M 33525
2:5»..— Nx 25$ amok 2:31— NN 25$ «ooh wamucU 8.5m
A520 3 one 2.23m A520 25 one 2:5 23...; $20

m_mo__8=._m P5:

ammo—Enem— FEE

 

 

3:528 .5 033

115

 

 

om; I No.“ v: momoo um; I no; om; good 2 u 8385 use

0:30

3%: .oz

Nmom I EN. mod ﬂood In; I cod 9.3 Eooo newton?

2:8 823%

33.5 .552: cuam 350 2:5»..— AEaE .928:— ctaﬂ 3:5 2:51— .385— «Ed
acumen—SD .xho «N 355 852.280 $3 «N 25$

A9518.—

-.Em BaE..8-EB 258m €3.83: .5:

again—1mm 5 2.35 €2.82: .5:

 

 

 

om; I mo; m: Smoo mm; I oo._ om; omooo C H $8.85 “EB
0330
mag: .02
No.: I oo._ omd vomoo omdm I mom 6.3 moooo newton?
0:30 ootoqom
Ago—«E :3..an 8.3— 360 2:979 GEE .9»..er .53— 2:5 2:91: .885— ii
ooaoccuoo $3 "X 25>? 3:02.250 gmo "N 25$
A920 .o 58 mag $2.82: E»: 556 E: 58 25% again .5:
.831 Emu o5 mag “.52 mmmozooEm <DmD 2: 5E doom 5 women.
.ozowaooé 62.38% maﬁa Howaoﬁamxom mo zoomed £8552 1&5an 05 E £63m o>£mom65m mmmozooEm Rom
0:30 So 828m xma BEoMbmooma Ea EoEommamE x0883: me $602 coﬁmoowom ocﬂwoq 8960M 35m Ann 035.

116

Chapter 4

THE OCCURRENCE OF CLINICAL SIGNS SUGGESTIVE OF HUMAN
BRUCELLOSIS AND ASSOCIATED RISK FACTORS IN LIVESTOCK
HERDERS IN THE KHANGAI MOUNTAIN REGION OF BAYANKHONGOR

AIMAG (PROVINCE), MONGOLIA

4.0 STRUCTURED ABSTRACT

Objective: This component of the study was undertaken to 1) record the occurrence of
clinical signs and symptoms suggestive of human brucellosis in livestock herders in the
Khangai mountain region of Bayankhongor Aimag (Province), Mongolia, and 2) identify
the risk factors associated with brucellosis-like disease occurrence in the Bayankhongor
herder population. In addition, attempts were made to compare the prevalence of
brucellosis in livestock (cattle) and the occurrence of brucellosis-like disease in humans.
Design: Cross-sectional study

Sample Population: The male or female head of household of a stratiﬁed random
sample of livestock herder households in 5 contiguous soums (district) and 1 bagh (sub-
district) of the Khangai mountain region of Bayankhongor Aimag (Province), Mongolia.
Procedure: A questionnaire was administered as an in—person interview to the head of
each herder household. The questionnaire collected information on the history of
diagnosis and treatment of brucellosis in the herders as well as the occurrence of non-

speciﬁc clinical signs and symptoms experienced within the previous 12-month period.

117

Information on the consumption of meat and dairy products and contact with livestock
was collected. Descriptive statistics were generated and the data were analyzed with
univariate and multi-variable methods.

Results: Twenty-six 100 herders interviewed reported a history of brucellosis diagnosis,
and 14 reported that they had received a diagnosis of brucellosis within the previous 12
months. Similarly, 20 of the 100 individuals interviewed reported experiencing clinical
signs suggestive of chronic brucellosis, and 16 reported experiencing clinical signs
suggestive of acute brucellosis within the previous 12 months. Thirty-three percent of
herders reported preparing dairy products from raw milk, 12% reported drinking raw
cow’s milk and 30% reported bringing weak neonates (lambs, calves and kids) into their
living quarters for extra care. All three of these behaviors were determined to be
signiﬁcant risk factors for the occurrence of clinical signs suggestive of brucellosis.
Conclusions: Data on the occurrence of brucellosis-like symptoms in the herder
population indicates that the disease has a signiﬁcant impact on the health and well being
of the human herder population. Consumption of raw milk, and close contact with weak,
newborn livestock, were identiﬁed as risk factors for the occurrence of clinical signs
suggestive of brucellosis in this herder population.

Data from this study support the Mongolian government’s classiﬁcation of
brucellosis as a priority public health concern. Because brucellosis is a zoonotic disease,
the focus of Mongolia’s planned national brucellosis control program is livestock
vaccination. The program is designed to control the disease in livestock and thereby

reduce and eventually eliminate the source of human Brucella infections.

118

4.1 INTRODUCTION

Brucellosis is an infectious disease of livestock and one of the most widespread
zoonotic diseases in the world (Corbel, 1989). The worldwide distribution of endemic
brucellosis includes Mongolia, a sparsely populated country (1.5 persons/kmz) in Central
Asia where livestock herding is the backbone of the county’s economy and an important
part of its cultural heritage (Mongolia, 2001c; OIE, 2002b). Major efforts have been
undertaken around the world to control brucellosis due to the signiﬁcant economic losses
associated with the occurrence of the disease in domestic livestock and the often
debilitating disease associated with human infection (Halling and Boyle, 2002; Ragan,
2002; Reddy, 2003). Health ofﬁcials in Mongolia today have highlighted brucellosis,
along with hepatitis (B & C) and tuberculosis, as one of their top three infectious disease
public health priorities (Ebright et al., 2003; Mikolon, 2000).

Brucella infections in humans are almost exclusively of animal origin and human
to human transmission of brucellosis is extremely rare (Nicoletti, 1989). Humans are
infected either by direct or indirect contact with infected animals or animal products or by
the ingestion of infected milk or dairy products (Slack, 2004b). In endemic areas of the
world the majority of cases occur in people who have occupational contact with livestock
such as livestock herders, dairy farmers, butchers, abattoire workers or veterinarians or in
children if they live in close proximity with domestic animals (Slack, 2004b). The
biovars of Brucella melitensis that are most commonly associated with brucellosis in
humans in decreasing order of occurrence are B. melitensis (primary hosts sheep, goats,
and camels), B. abortus (primary host cattle and camels), B. suis (primary host swine),

and B. canis (primary host dogs) (Black, 2004b; Reddy, 2003; Slack, 2004b).

119

Brucellosis in humans is often associated with a number of nonspeciﬁc symptoms
including fever (“undulent fever”), night sweats, anorexia, fatigue, headache, myalgia,
malaise and depression (Benenson, 1995; Black, 2004b; Young, 1989). Brucellosis is a
systemic infection and can involve any organ system and potentially cause osteoarticular,
gastrointestinal, pulmonary, neurologic and cardiovascular complications (Black, 2004b).
The incubation period is variable (weeks to months) and the onset of disease can be acute
or insideous. Untreated brucellosis in humans can lead to chronic localized infections
often manifesting as joint pain, joint efﬁrsion and debilitating arthritis (Benenson, 1995;
Young, 1989).

Historically, Mongolia has had a high incidence of brucellosis in its human
population (Otgon 1968; J ezek Rusinko et al. 1972; J ezek, Rusinko et a1. 1974) as a
consequence of the high prevalence of the disease in domestic livestock (Beulig et al.,
1969; Jezek et al., 1972; Jezek et al., 1974; Otgon, 1968; Pinigin et al., 1968a; Pinigin et
al., 1968b; Tserendash, 1980, 1984; Tserendash and Shumilov, 1970; Zherikova et al.,
1972). Brucellosis prevention, through livestock vaccination, began in Mongolia in the
1960’s (Corbel, 1989; Tserendash, 1980). A second prevention program, lead by the
World Health Organization (WHO), began in the mid-1970’s and focused on the
vaccination of small ruminants (Corbel, 1989). The former Soviet Union continued to
support a brucellosis livestock vaccination program in Mongolia (Brucellosis abortus
strain-19 vaccine in cattle and Br. melitensis Rev-1 vaccine primarily in small ruminants)
through the late 1980’s.

Initially the brucellosis control programs were deemed a success by Mongolian

scientists (Tserendash, 1980) and by WHO ofﬁcials who reported that the annual

120

incidence of the disease in humans dropped to less than 1 case/ 10,000 in the early 1980’s
(Mikolon, 2000; Roth et al., 2003). Despite evidence in the early 1980’s of a signiﬁcant
reduction in the incidence of brucellosis in Mongolia, concerns have been raised recently
about the re-emergence of the disease in the country (Ebright et al., 2003; J oshi, 1991;
Roth et al., 2003; WHO, 2000b, 2001, 2003). Brucellosis surveillance in the human
population of Mongolia since 1990, has consisted primarily of the unsystematic sampling
and screening of high risk populations (herders and slaughterhouse workers) and reports
on the number of hospitalized cases (Mikolon, 2000; Mongolia, 2001d; Roth et al., 2003).
Both occupational hazards, through direct contact with infected livestock, (Roth et al.,
2003) and the consumption of unpasturized milk and dairy products have been identiﬁed
as modes of disease transmission ﬁ'om livestock to humans in Mongolia (Mikolon, 2000).
The persistence and possible re-emergence of brucellosis as a major public health
problem in Mongolia has been blamed on the collapse of the veterinary service delivery
system and public health inﬁ'astructure in Mongolia which began in the late 1980’s
(Ebright et al., 2003; Mikolon, 2000). The Mongolian veterinary system in the 1990’s
reﬂects the changes that were experienced by the country as a whole as Mongolia
transitioned from a centrally planned economy and socialist system of government to a
democratic system with a market economy (Ebright et al., 2003; Rossabi, 1999). The
combination of the abrupt loss of economic aid ﬁ'om the former Soviet Union, the
privatization of livestock in 1990, and the privatization of veterinary service delivery in
1999, meant the end of large scale, government subsidized, livestock health campaigns
including comprehensive livestock vaccination programs and systematic surveillance for

brucellosis (Mikolon, 2000).

121

Recently collected data on the prevalence, incidence and distribution of
brucellosis in Mongolia’s livestock and human population is largely unavailable and no
published reports of the risk factors associated with disease exist. In this study data on
the possible burden of brucellosis disease in livestock herders were assessed alongside a
study of the prevalence of brucellosis in cattle in the Khangai mountain region of
Bayankhongor Aimag (Province), Mongolia. By collecting information on the
occurrence of clinical signs and symptoms suggestive of brucellosis disease in the human
population, and information on the history of brucellosis diagnosis, attempts were made
to identify risk factors associated with brucellosis-like disease in the livestock herders of

the region.

4.2 HY POTHESES

1) Clinical signs and symptoms suggestive of brucellosis disease in humans
occur in livestock herders in the Khangai mountain region of Bayankhongor Aimag
(Province), Mongolia, where livestock brucellosis is endemic.

2) Consumption of unpasturized dairy products and direct contact with
livestock are associated with an increased risk of the occurrence of brucellosis-like

clinical signs and symptoms in livestock herders.

4.3 MATERIALS AND METHODS
A. Study Design
A cross-sectional survey was used to determine the prevalence of clinical signs

and symptoms suggestive of brucellosis, the prevalence of a history of brucellosis

122

diagnosis, and the patterns of livestock product consumption and livestock contact, in a
population of nomadic herders in the Khangai mountain region of Bayankhongor Aimag
(Province), Mongolia. The unit of interest in this portion of the study was an individual
herder in each herder household. One individual, identiﬁed as the head of the selected
herder household, was interviewed to collect the information on their individual
experience of symptoms suggestive of brucellosis and their practice of potential risk
factors associated with the occurrence of brucellosis-like disease.

B. Herder Household Selection

A sample (n=104) of the 5,062 herder households located in the 5 contiguous
soums (districts) of Gurvanbulag, Zag, Galuut, J argalant, Erdenstogt, and the sub-district
of Shargalj nut, in the Khangai mountain region of Bayankhongor Aimag (Province),
Mongolia, were selected using a stratiﬁed random sampling technique. The sampling
frame, sample size calculation and herder household identiﬁcation methods were
identical to those used for the selection of herder households for cattle/yak brucellosis
testing. The methods were summarized previously in Section 2.3.

C. Data Collection Team and Herd Location

The same data collection team who conducted the livestock brucellosis testing
and livestock management questionnaire was responsible for this portion of the study.
Livestock herder households enrolled in this portion of the study were the same as those
enrolled in the livestock brucellosis portion of the study. The same methods for locating
herder households and obtaining and recording informed consent were employed. See

Section 3.3.

123

D. Head of Herder Household Interviews: Health Status and Livestock
Contact/Product Consumption Data

A 2-page questionnaire was conducted as an in-person interview with the male or
female head of the selected herder households. The questionnaire (Appendix C) was
adapted from a one used previously for a. brucellosis survey in eastern Mongolia. The
interview was conducted in Mongolian by a bilingual (Mongolian and English), native
Mongolian speaker. The in-person interview was conducted during the herd visit as
described in chapters 2 and 3, often with the same individual who participated in the
livestock management survey.

Basic demographic information (age, sex, education level) was collected at the
beginning of the interview and the questionnaire was coded to correspond with the
livestock herd identiﬁcation number. The ﬁrst part of the questionnaire focused on
recording the nature of herder contact with livestock through both speciﬁc occupational
and routine activities in the herd. Patterns of livestock product consumption were also
recorded, including practices related to the preparation and consumption of both meat and
milk products (See Appendix C).

The second part of the questionnaire was designed to collect information on the
occurrence of non-speciﬁc clinical signs and symptoms of disease in the herder
population that could then be further categorized as typical of “acute” or “chronic”
brucellosis. Data on the history of brucellosis diagnosis was also collected and herder
knowledge of the routes of brucellosis transmission from livestock to humans and the
means of brucellosis prevention were assessed. Interview responses were recorded in

Mongolian and later translated into English.

124

E. Case Definitions

The responses recorded in the “non-speciﬁc clinical signs and symptoms” section
of the questionnaire were used to identify herders who, in the previous 12 months, had
experienced clinical signs and symptoms suggestive of brucellosis disease. A case
deﬁnition of “acute” brucellosis was created based on the reporting of one or more of the
signs usually experienced at the onset of brucellosis infection, including chills, headache,
fever or night sweats (Benenson, 1995; Black, 2004b). A case deﬁnition of “chronic”
brucellosis was created based on the reporting of one or more of the signs usually
experienced with the osteoarticular complications associated with chronic localized
brucellosis infections, including joint pain, stiff joints, swollen joints or lower back pain
(Benenson, 1995; Black, 2004b). A third case deﬁnition of “general brucellosis” was
also identiﬁed that included individuals who reported clinical signs and symptoms typical
of both, or either, “acute” or “chronic” brucellosis.

Information on the history of brucellosis diagnosis in this population of herders
was also collected and used as an alternative “case deﬁnition” in the analyses, described

below (Section 4.4).

4.4 DATA ANALYSIS

A. Univariate Analysis of Risk Factors Associated with Clinical Signs
Suggestive of Brucellosis and Reported Brucellosis Diagnosis in
Livestock Herders

Data from the human health status/livestock contact questionnaire were entered

into Microsoft Excel spreadsheets (Microsoft Ofﬁce 2000). Data ﬁles containing

information on the cattle herd brucellosis status, collected as described in chapter 2, were

125

merged with the Excel ﬁle containing the human questionnaire data.9 Imported Excel
spreadsheet data were analyzed using the SAS software program (SAS version 8.2, Cary
NC: SAS Institute, Inc. 1999-2001). Descriptive statistics were computed for the risk
factors and dependent variables of interest. Datasets were created within SAS that

corresponded to the disease classiﬁcation and case deﬁnitions described in Section 4.3, E.

The t-test (PROC TTEST) was used to test for signiﬁcant associations between
the occurrence of brucellosis-like clinical signs and symptoms and the history of
brucellosis diagnosis in humans and the continuous variables of age and years of
education. Pearson’s X2 test (PROC FREQ; “TABLES/CHISQ”) was used to test for
signiﬁcant associations between the occurrence of brucellosis-like clinical signs and
symptoms, history of brucellosis diagnosis in humans and potential categorical risk
factors for disease. An initial investigation of the presence of confounding by gender was
investigated by performing a gender-stratiﬁed analysis of the categorical data and
computing the Mantel-Haenszel X2 test for stratiﬁed tables (PROC FREQ;
“TABLES/CHISQ”; MH).

B. Multi-Variable Analysis of Risk Factors for the Occurrence of
Clinical Signs Suggestive of Brucellosis in Livestock Herders

The complete human disease data set contained 76 variables. A subset of the
original variables were considered for inclusion in the ﬁll] logistic regression model.
Variables excluded from the multi-variable logistic regression analysis included those

variables with a large proportion of missing responses (>20% missing); redundant

 

9Serologic tests (BRT, CARD, and CFT) were used to test the milk and serum of cattle in the selected
herds. Six classiﬁcations of a “positive” herd were created based on the positive results of l or a series of
serologic tests. Generally the “BRT or CARD-Positive” classiﬁcation was the most sensitive but least
speciﬁc classiﬁcation and the “>1 CARD-Positive” and “CFT-Conﬁrmed CARD—Positive” the less
sensitive but the most speciﬁc classiﬁcation.

126

variables; and those deemed not to be biologically relevant to the occurrence of
brucellosis.

Sixteen independent variables were included in the full logistic regression model.
Soum (district) and gender were entered as class variables. Biologically relevant
interaction variables were included in the full model. Backward elimination was used to
model the relationship between potentialrisk factors and confounders (livestock product
consumption practices, livestock handling practices, age, and gender) and the occurrence
of brucellosis-like clinical signs and symptoms (general, acute, and chronic) in the human
population. A signiﬁcance level of 0.05 was used for removal of effects ﬁ'om the model.
An assessment of confounding was preformed using non-statistical means. If the
removal of a covariate resulted in more than a 10% change in the point estimate of the
remaining covariates the removed covariate was added back to the model. Observations
with missing values for the response or explanatory variables were removed during the

modeling procedure.

4.5 RESULTS

A. Description of Sample Population

Participation and Demographics:

The human health and risk factor questionnaire was administered to an individual
identiﬁed as the head of the herder household in 101, of the 104 herds (97%), targeted for
inclusion in the study. The three herds for which we were not able to collect
questionnaire information were not signiﬁcantly different from other herds in the study.

The reasons for our inability to collect questionnaire information from the three herds

127

included one herder who refused to enroll the herd in any aspect of the study, one herder
who refused to participate in the questionnaire due to lack of time, and a data recording
error that prevented us from matching the data from the completed questionnaire to the
appropriate cattle herd.

The mean age of the interviewed herders was 42, (median 41 , range 17-
81). The interviewed herders had received an average (mean) of 5.7 years of formal
education (median 8, range 0 to 12). Just over half of the herders interviewed were
female (52.75%).

Herder Occupation:
The majority (92%) of interviewees identiﬁed “livestock herding” as their

primary occupation or “profession”. Six percent of the individuals identiﬁed other
occupations, including 3 veterinary technicians and a human health worker. None of the

interviewed individuals were involved in commercial livestock slaughter or trade.
LivestocLContact:
Traditionally all members of the herder household participate in livestock rearing.

As detailed in chapter 2, 75% of these herds were composed of multiple species of
livestock. The presence of cattle was recorded in the largest number of herds (100%)
followed by sheep, goats and camels. Horne slaughter of livestock for personal
consumption and hand milking of individual animals is common practice among
Mongolian herders. Of the 100 individuals interviewed, 84% reported participation in
cattle slaughter, 87% in sheep, 80% in goat and 3% in camel slaughter. Ninety-six

percent of the individuals interviewed reported hand milking cattle, 24% reported hand-

128

milking goats and 21% reported hand-milking sheep. Speciﬁc questions addressing
herder assistance of livestock during parturition or dystocia were not asked, however,
30% of herders reported that weak neonates (primarily lambs and kids) were brought into

the family’s living quarters for observation and extra care.

Meat/Milk Consumption in Herder Households:

Mongolian livestock herders generally consume the meat and milk products of the

livestock they raise. One hundred percent of the heads of herder households interviewed
reported consuming beef, 93% reported consuming mutton, 75% reported consuming
goat meat and 3% reported consuming camel meat. Traditionally meat is thoroughly
cooked, either by boiling or roasting. One respondent reported consumption of raw meat
of camel, cattle, sheep and goat origin. The consumption of raw milk was reported by
15% of respondents. Twelve percent reported consuming raw cow or yak milk, 3% raw
goat milk and 3% raw sheep milk. Thirty-three percent of respondents reported preparing

dairy products with raw milk.
Knowledge of Brucellosis Prevention and Transmission:

Twenty-ﬁve percent of the heads of herder households interviewed understood
that brucellosis was a zoonotic disease and that infected livestock were a risk to human
health. Seven percent of the herders interviewed identiﬁed both livestock product
consumption and contact with aborted material as a means of brucellosis transmission to
humans. An additional 18% identiﬁed either consumption of contaminated livestock
products (meat and milk) or direct contact with aborted material as the means of

transmission of disease from livestock to humans.

129

Eighteen percent of the heads of herder households correctly identiﬁed ways to
prevent brucellosis occurrence in humans. Overall 12% identiﬁed either livestock
vaccination/prevention or sanitary measures (cooking/boiling meat and milk and use of
protective clothing). Six percent of interviewees identiﬁed both livestock brucellosis
prevention and the institution of sanitary measures.

Prevalence of Brucellosis Diagnosis, Testing_and Treamnent History:

Twenty-seven percent (26/97) of interviewed heads of herder households reported
a history of brucellosis diagnosis. Twelve percent of the interviewed herders
(approximately ‘/2 with a history of brucellosis) had been diagnosed within the previous
12 months. Eight percent of all of the individuals survey reported that they had ever been
tested for brucellosis with a blood test and 5% reported skin testing. None of the
interviewed individuals, including those with a history of brucellosis diagnosis, reported
having been treated for brucellosis either as outpatients or in hospitals or clinics. None of
the individuals reported that they had been rechecked for brucellosis after initial
diagnosis and 1 person reported experiencing a relapse of disease.

Prevalence of Reported Non-Speciﬁc Clinical Sigr_rs and Symptoms:

The surveyed individuals indicated which of a list of non-speciﬁc clinical signs
and symptoms they had experienced within the previous 12 months. (See questionnaire in
appendix C). The most common symptom experienced was joint pain at 18%. The other
symptoms experienced were headache (1.6%), lower back pain (12%), stiff joints (11%),
swollen joints (10%), muscle pain (9%), extreme weakness (9%), dizziness (9%), bone

pain (9%), abdominal distension (6%), difﬁculty sleeping (4%), night sweats (2%), fever

130

(2%), enlarged lymph nodes (2%), abdominal pain ( 1%) and 0% experiencing chills and
depression (See Table 4.1).

The prevalence of individuals experiencing one or more symptoms typical of
brucellosis (chills, fever, headache, night sweats, joint pain, stiff joints, swollen joints or
lower back pain) was 24% (24/ 100). When the list of typical brucellosis-like symptoms
was further divided into those more characteristic of acute disease brucellosis (chills,
fever, headache, or night sweats) and those more typical of chronic infection (joint pain,
stiff joints, swollen joints or lower back pain) the prevalence of acute brucellosis-like
symptoms was 16% and the prevalence of chronic brucellosis-like symptoms was 20%.

B. Univariate Analyses: History of Brucellosis Diagnosis

Association with Brucellosis-Like Symptom;

No signiﬁcant associations were found between the history (“recent” or “ever”) of
brucellosis diagnosis and the reporting of clinical signs and symptoms typical of either
acute or chronic brucellosis (See Table 4.2). Additionally no signiﬁcant associations
were found between the history of brucellosis diagnosis and any of the individual clinical
signs and symptoms reported by the surveyed individuals.

Association with Brucellosis Status of Cattle Herd:

No signiﬁcant associations were found between the reported diagnosis of
brucellosis in livestock herders and the brucellosis status of their cattle herd by any of the

testing methods employed (See Table 4.3).

131

Association with LivestockProduct Consumption and Herding Practices:

No signiﬁcant associations were found between the history of brucellosis
diagnosis and any livestock product consumption or herding practices with the exception
of “bringing neonates into living quarters” (X2 = 6.71; Fisher’s p-value = 0.0108).

Age and Gender:

There were no associations between either age or gender and the reported
diagnosis of brucellosis (“ever” or “recent”) in the herder population.

Association with Knowledge of Brucellosis Prevention and Transmission:

There was no association found between knowledge of brucellosis prevention or
routes of transmission and the reported history of brucellosis diagnosis (“ever” or
“recent”) in the herder population.

C. Univariate Analysis: Reporting of Clinical Signs and Symptoms
Suggestive of Acute and Chronic Brucellosis

Association with Brucellosis Status of Qttle Herd:

Signiﬁcant associations were found between the reporting of typical signs of
“chronic” brucellosis and the “BRT and CARD-Positive” classiﬁcation of cattle (X2 =
4.91; p-value = 0.0267) and the “CARD-Positive” classiﬁcation of cattle in the same herd
(X2 = 4.77; p-value = 0.0290). Though not signiﬁcant at the .05 level, a strong
association was also found between the reporting of typical signs of chronic brucellosis
and the “>1 CARD-Positive” classiﬁcation of cattle in the same herd.

No signiﬁcant associations were found at the 0.05 level between the reporting of
typical signs of “acute” brucellosis and the brucellosis positive status of cattle in the same
herd, however, associations nearing signiﬁcance were found between typical signs of

“acute” brucellosis in humans and the “BRT and CARD-Positive” (X2 = 3.27; p-value =

132

0.0707) and the “>1 CARD-Positive” (x2 = 3.12; p-value = 0.0775) classiﬁcation of their
cattle herds (See Table 4.4).

Association with Livestock Product Con_sumption and Herding Practices:

The practice of drinking raw milk and producing dairy products from raw milk
were highly associated (X2 = 7.56-20.06; p-values = 0.0060 - <0.0001) with the
occurrence of both “acute” and “chronic” brucellosis-like symptoms in the herder
population (See Table 4.5). The practice of milking, slaughtering or eating meat from
camels, cattle, sheep or goats was not associated with either acute or chronic brucellosis-
like symptoms in the herder population (See Table 4.5). Keeping neonates (primarily
kids and lambs) in the family’s living quarters was associated with the reported
occurrence of both “acute” brucellosis-like symptoms (X2 = 4.62; p-value = 0.0317) and
“chronic” brucellosis-like symptoms (X2 = 11.03; p-value = 0.0014) in the herder
population. All of the herders in this population but 1 reported that they herded yaks
(Bos grunniens), cattle (Bos taurus) or hybrid crosses of the two species. Herders’
reports of both “acute” and “chronic” brucellosis-like symptoms were positively
associated with the practice of herding strictly Bos taurus cattle whereas the practice of
herding strictly Bos grunm’ens cattle was positively associated with the absence of
brucellosis-like symptoms in the interviewed herders (See Table 4.5).

Age and Gender:

The mean age of herders reporting the occurrence of signs and symptoms
suggestive of “acute” and/or “chronic” brucellosis was slightly older but not signiﬁcantly
different from the mean age of those who experienced no clinical signs and symptoms

(See Table 4.6). Being female was associated with the reporting of the occurrence of

133

“acute” brucellosis-like symptoms (X2 = 3.89; p-value = 0.0495). There was no
difference in the odds of men and women having a history of brucellosis diagnosis, or
experiencing chronic or general brucellosis-like symptoms (See Table 4.7). Males were
more likely to report their participation in cattle slaughter (X2 = 6.33; p-value = 0.0119).

Occupation:

Individuals who identiﬁed themselves as animal health workers were more likely
to report the occurrence of clinical signs and symptoms of “acute” and “chronic”
brucellosis, though the association did not reach statistical signiﬁcance; “acute”
brucellosis (x2 = 5.91; p-value = 0.0658) and “chronic” brucellosis (x2 = 4.21; p-value =
0.1011), (See Table 4.5).

Knowledge of Brucellosis Prevention and Transmission:

No associations were found between herders’ knowledge of brucellosis
transmission (either or both direct and indirect routes) and their reporting of “acute” or
“chronic” brucellosis-like clinical signs and symptoms. Similarly, no associations were
found between herders’ knowledge of brucellosis prevention (either or both livestock
disease control and persona/food safety) and their reporting of acute or chronic
brucellosis-like clinical signs and symptoms (See Table 4.8).

An additional analysis was performed to determine whether or not there was an
association with herder understanding of brucellosis transmission and prevention and the
livestock product consumption and herding practices found to be associated with the
occurrence of brucellosis-like symptoms in this population. The consumption of raw

milk (all species), the consumption of raw cow’s milk, preparing dairy products with raw

134

milk, bringing neonates into the living quarters and gender (being female) were all
examined. No signiﬁcant associations were found (See Table 4.9).
B. Multi-Variable Analysis of Livestock Product Consumption and
Livestock Herding Practices and the Occurrence of Brucellosis-Like
Symptoms in the Human Herder Population

General Brucellosis-Like Symptoms:

The full multi-variable logistic regression model of the occurrence of brucellosis-
like symptoms, the occurrence of “acute” brucellosis-like symptoms and the occurrence
of “chronic” brucellosis-like symptoms in the human population was based on 60
observations (61% of interviewed heads of herder households, and 59% of all herds) and
included 16 variables. A relatively high prevalence of missing data on age (24%) led to
observations being removed ﬁ‘om the model. There were 20/60 (33%) individuals
reporting “general” brucellosis-like symptoms, 14/60 (23%) reporting “acute”
brucellosis-like symptoms, and 16/60 (27%) reporting “chronic” brucellosis-like
symptoms in this sample.

The ﬁnal reduced model for “general” brucellosis-like symptoms included 2 risk
factors; “preparation of dairy products from raw milk” (OR=16.25; 95% C.I.=2.83-93.38)
and “neonates in living quarters” (OR=10.60; 95% C.I.=1.92-58.40). Age was a
signiﬁcant variable but not a signiﬁcant confounder. A 1-year increase in age
represented an 8.8% increase in risk of the occurrence of “genera ” brucellosis-like
symptoms (See Table 4.10).

The ﬁnal reduced model for “acute” brucellosis-like symptoms (Table 4.11)

included 2 risk factors; “preparation of dairy products from raw milk” (OR=9.33; 95%

C.I.=2.22-39.31) and “drinking raw cow’s milk” (OR=6.52; 95% C.I.=1.24—34.40).

135

Gender was a signiﬁcant confounder. Being female was associated with a 5 fold increase
in the odds of the occurrence of clinical signs suggestive of “acute” brucellosis (OR=5.3;
95% C.I.=1.03-27.32), (See Table 4.11).

The ﬁnal reduced model for “chronic” brucellosis-like symptoms included 3 risk
factors; “drinking raw cow’s milk” (OR=6.52; 95% C.I.=1 .24-34.40), “neonates in living
quarters” (OR=10.60; 95% C.I.=1.92-58.40), and “herding strictly Bos taurus cattle”
(OR=5.64; 95% C.I.=1 .00-31.80). Age was a signiﬁcant variable but not a confounder.
A l-year increase in age represented a 5.4% increase in risk of the occurrence of

“chronic” brucellosis-like symptoms (See Table 4.12).

4.5 DISCUSSION

A. Hypotheses

Hypothesis 1, that clinical signs and symptoms suggestive of brucellosis occur in
the population of livestock herders in the Khangai mountain region of Bayankhongor
Aimag (Province), Mongolia, where livestock brucellosis is endemic; and hypothesis 2,
that human behaviors related to the consumption of unpasturized dairy products, and
direct contact with livestock are associated with an increased risk of the occurrence of
brucellosis-like clinical signs and symptoms in livestock herders in the Khangai region of
Bayankhongor Aimag (Province), Mongolia; were both found to be true and conﬁrmed in

this cross-sectional study.

136

B. Prevalence of Reported Symptoms Suggestive of Brucellosis and
History of Brucellosis Diagnosis in Herder Population

The prevalence of brucellosis-like symptoms (24/ 100; 24%) and the reported
history of brucellosis diagnosis (26/ 100; 26%) in this population of livestock herders
from the Khangai mountain region of Bayankhongor Aimag (Province), Mongolia, is
high. Studies of brucellosis sero-prevalence among livestock herders in other regions of
the world with endemic livestock brucellosis report sero-prevalence of 3.8% (Schelling et
al., 2003). Although the brucellosis disease prevalence data collected in this study are
not laboratory-conﬁrmed, the prevalence recorded in this Bayankhongor Aimag
(Province) herder population is similar to the ﬁgure of 25% brucellosis sew-prevalence in
Mongolia’s herder population reported by A. Mikolon (Mikolon, 2000) and in the
Mongolian Health Sector Review cited by Roth, et al, in which 16% of herders and
livestock workers were classiﬁed as infected with brucellosis (Roth et al., 2003).

Although the ﬁgure of brucellosis prevalence in this population of herders that is
based on reported diagnosis and the reported occurrence of brucellosis-like clinical signs
and symptoms is similar, the correlation one would expect between these two means of
classiﬁcation does not exist. There is no association between individuals who reported
that they had a history of either “ever” or “recent” brucellosis diagnosis, and their report
of the occurrence of clinical signs and symptoms associated with human brucellosis
disease. Further investigation revealed that none of the individuals who reported a
history of brucellosis diagnosis had ever been treated for the disease. The individuals
were questioned about this in more than one way as they were asked to identify what they
were treated with, whether or not they had been hospitalized, treated as an outpatient or

in a community clinic. The medical background of the individual making the diagnosis

137

of brucellosis in this instance is unknown. Additionally only a fraction of the herders
surveyed reported that they had been tested with a blood or skin test for brucellosis. The
uncertainty of the source of brucellosis diagnosis in these individuals and the lack of
correlation with a history of clinical signs of disease, led us to suspect that this case
deﬁnition (self-reported “ever diagnosed” with brucellosis) was prone to serious
misclassiﬁcation error.

Interestingly, when the “history of brucellosis” disease classiﬁcation was used to
test for associations (univariate and multi-variable) with potential risk factors for disease
(livestock product consumption, livestock herding practices, gender and age) no
signiﬁcant associations were found with the exception of the practice of bringing weak
neonates (primarily lambs and kids) into the living quarters for extra care and attention.
This risk factor also proved to be signiﬁcant when alternative case deﬁnitions for
brucellosis were used, and will be discussed below in terms of its potential role as a
signiﬁcant “high-risk” human behavior for the acquisition of brucellosis.

C. Cattle Herd Brucellosis Status and Human Brucellosis-Like Disease

Although the presence of brucellosis in livestock in itself does not guarantee the
transmission of disease to the humans who tend the livestock and consume its products,
one would suspect that the chance of zoonotic transmission would be higher in those
herder households in which there was evidence of the presence of brucellosis in livestock,
and especially those herder households in which there was evidence of more than a single
case of livestock brucellosis. This was found to be true in this population of herders.

Cattle herds meeting the more stringent deﬁnition of a brucellosis ser0positive

herd (more than 1 positive test, or more than 1 individual with a positive brucellosis test)

138

were found to be signiﬁcantly associated with the reporting of the occurrence of
“chronic” brucellosis-like symptoms in the interviewed head of herder household.
Relationships nearing signiﬁcance were found between brucellosis-positive classiﬁed
cattle herds and the report of “acute” brucellosis-like clinical signs in the head of the
herder households.

No signiﬁcant associations were found between cattle herd brucellosis status and
the reported history of brucellosis diagnosis in the head of the herder household.

Livestockﬂerding Practices as Risk Factors for the Occurrence of Brucellosis-
Like Symptoms in Herders:

The livestock herding practice most consistently associated with the reported
occurrence of brucellosis-like symptoms in interviewed herders was the practice of
bringing weak neonates (primarily lambs and kids) into the family’s living quarters for
extra care and attention (OR. of 15-20). Interviewed herders were not asked questions
about their contact with aborted livestock fetuses and placentas, or practices associated
with assisting livestock species during the birthing process. Direct human contact
(through breaks in the skin or conjunctiva) with uterine ﬂuid, aborted fetuses and
especially placentas of brucellosis infected livestock, have been identiﬁed as signiﬁcant
modes of zoonotic transmission of brucellosis (Benenson, 1995; Reddy, 2003). The
practice of brining weak neonates into the living quarters characterizes the close
involvement of the livestock herders in the birthing process and the close contact they
and their family have with neonates potentially coated with brucellosis containing uterine
and amniotic ﬂuid. The subset of herder households that follow this practice are
therefore at increased risk of contracting brucellosis through direct contact with their

infected livestock.

139

The practice of hand-milking livestock or involvement in the slaughter and
butchering process does not appear to be associated with an increased risk of
experiencing clinical signs and symptoms characteristic of acute or chronic brucellosis in
this population. The practice of herding strictly yaks (Bos grunnz'ens) appears to be
somewhat protective while herders who reported a strictly Bos taurus composition of
their cattle herd were more likely to experience brucellosis-like symptoms. This
relationship held true especially in the multivariable model of chronic brucellosis-like
symptoms. The history of cattle production in Mongolia may explain this pattern. As
previous authors have mentioned, the regional spread of brucellosis in Mongolia in the
1950’s - 1970’s was linked to the establishment of intensive state owned dairy
cooperatives, often with imported cattle (Denes, 1997; Otgon, 1968). When livestock
were privatized in Mongolia in 1990, the state farms were dismantled and the cattle
distributed to Mongolian herders. Herders who received and maintained Bos taurus
cattle from the state farms may have a more persistent problem with brucellosis in their
herds today than do herders whose herds are composed of indigenous cattle and yaks that
have been raised extensively and at lower densities for decades.

Livestock Product Consumption Practices ai Risk Factors for the Occurrence of
Brucellosis-Like Symptoms in Herders:

The consumption of raw cow’s milk and the use of raw milk in the preparation of
dairy products were both identiﬁed as signiﬁcant risk factors for the occurrence of
brucellosis-like symptoms in this populations of livestock herders (O.R. range = 6-30).
Slight variations in the estimation of the odds ratios were found in the multi-variable
models when the deﬁnition of disease was changed from “general”, to “acute”, to

“chronic” brucellosis-like symptoms. The strongest association with brucellosis-like

140

symptoms was seen with the practice of consurrring of raw cow’s milk followed by the
practice of preparing dairy products from raw milk. The identiﬁcation of these two
livestock product consumption risk factors is consistent with the knowledge that
unpasturized milk and diary products are potential sources of infection for humans
(Benenson, 1995; Reddy, 2003).

Occupation:

Livestock herders in general represent a high-risk population for brucellosis
infection in countries where the disease is endemic in livestock (McDermott and Arimi,
2002). The majority of individuals interviewed in this study identiﬁed livestock herding
as their primary occupation. Within this population of herders there were no individuals
who identiﬁed an alternative occupation considered a hazard for brucellosis infection
with the exception of the three individuals who identiﬁed themselves as livestock health
workers. Two of these three individuals reported the occurrence of clinical signs and
symptoms consistent with both “acute” and “chronic” brucellosis infection. Although the
numbers are small, we can speculate that livestock health workers are subjected to an
increased risk of brucellosis infection within this herder population due to their increased
opportunities for contact with infected livestock.

Gender and Age:

Age is a signiﬁcant variable but it does not confound the relationship between the

identiﬁed behavioral risk factors for brucellosis (drinking raw milk, preparing dairy
products from raw milk and bringing neonates into the living quarters) and the
occurrence of “general” and “chronic” brucellosis-like symptoms in this population of

livestock herders. Brucellosis is an infectious disease with chronic sequelae so the slight

141

increase in the odds of symptom occurrence with age is expected. Older individuals have
had a longer period of time in which to become exposed to the infectious agent. It is also
likely that there is a general increase in the prevalence of many of the non-speciﬁc
symptoms that are consistent with both brucellosis and many other conditions in older
individuals.

Being female was associated with a signiﬁcantly higher odds (OR = 5.31; 95%
CI. = 103-27.32) of reporting clinical signs as symptoms consistent with “acute”
brucellosis infection. One would not expect to see any association between gender and
brucellosis infection since males and females are equally susceptible to the disease.
However, in some cases there is a division of labor that falls along gender lines in
livestock herding practices that may lead to differences in exposure to infection. In this
population of herders the only identiﬁed herding or livestock consumption product
consumption practice that was associated with gender was cattle slaughter. Signiﬁcantly
more men (X2 = 6.33; p-value = 0.0119) reported that they slaughtered cattle, though this
did not prove to be a signiﬁcant risk factor for disease in the reduced logistic regression
model.

The higher probability of reporting signs typical of “acute” brucellosis among
women in this herder population may be due to a gender speciﬁc role in an aspect of
livestock herding not captured in the study survey. As mentioned previously, speciﬁc
questions surrounding the herders’ involvement in obstetric procedures and the handling
of aborted material and livestock placentas were not asked. Personal observations
indicate that women and girls generally take care of female stock nearing parturition and

newborn stock. Men and boys spend time grazing adult livestock in the surrounding

142

pastures. This study was performed in July and August, approximately 1-2 months after
peak calving, lambing and kidding season. The higher prevalence of women reporting
clinical signs associated with “acute” brucellosis may be related to their more frequent
exposure to infected (aborting) livestock.

D. Limitations

A signiﬁcant amount of data in this portion of this study was collected using a
questionnaire that was conducted as an in-person interview. Individuals were asked to
recall the occurrence of a list of clinical signs and symptoms within the past 12 months
and also asked to report on their typical practices and habits surrounding livestock
herding and livestock product consumption. Although we can expect that individuals
will remember the occurrence of signs of illness over a 12-month period, it is still a
signiﬁcant time and the data may be susceptible to recall bias. We have no reason to
believe that the questions concerning livestock herding practices or livestock product
consumption would have led individuals to hesitate to give us accurate answers but the
normal variation in practice from day to day and season to season may have led to some
misclassiﬁcation bias with regards to those who did and did not participate in the
potentially risky practices. These points represent the limitations of a cross-sectional
study using self-reported data. Since the data is collected at one point in time, making
seasonal trends or variation over time hard to capture.

One of the major limitations of this study was that conﬁrmatory laboratory testing
of individuals for brucellosis infection was not available. The information collected on
the occurrence of clinical signs and symptoms was subject to misclassiﬁcation bias due to

the non-speciﬁc nature of many of the clinical signs associated with both “acute” and

143

“chronic” brucellosis. Despite this drawback to our method, the case deﬁnitions based on
clinical signs and symptoms seemed much more consistent with disease than did the case
deﬁnition based on the report of a previous or recent diagnosis of brucellosis.

Only the heads of households were interviewed and asked to report the
occurrence of their clinical signs and symptoms over the lZ-month period. We were able
to interview roughly equal numbers of men and women but the majority of the
individuals were middle-aged adults and no data on the occurrence of brucellosis-like
symptoms in infants, children or very young adults were collected. The burden of
disease in children, found to be signiﬁcant in previous studies (J ezek et al., 1972; J ezek et

al., 1974), was not assessed.

4.6 CONCLUSION

Bovine brucellosis is an important livestock disease in the Khangai mountain
region of Bayankhongor aimag, Mongolia, and the data on the occurrence of brucellosis-
like symptoms in the herder population indicates that the disease also has a signiﬁcant
impact on the health and well being of the human herder population. The practices
identiﬁed as risk factors for disease in this region of the world are not different from
those that have been identiﬁed in other areas where livestock brucellosis is endemic.
Raw milk, consumed directly or used in the preparation of dairy products, will continue
to be a source of infection for humans as long as brucellosis is circulating in the livestock
population. Educational campaigns will have some effect on herder understanding of the

risks associated with their livestock production consumption habits but, as was seen when

144

examining these data, knowledge of brucellosis transmission routes has little effect on the
practices that enhance its transmission.

The second major risk factor for human disease identiﬁed in this study was a
practice closely associated with the routine care of newborn livestock. In a population
where livestock herding represents both cultural and economic survival, individuals will
be resistant to giving up a practice, like bringing weak livestock into warm and sheltered
human living quarters, which are perceived to increase the productivity of their ﬂocks.
The only way to reduce the burden of human disease associated with brucellosis infection
is to reduce, and eventually eliminate, the disease in livestock.

Inﬂuenced by both national and international concerns surrounding the re-
emergence of brucellosis in Mongolia (Ebright et al., 2003; Roth et al., 2003; WHO,
2001; Zinsstag et al., submitted 2004), the Mongolian government has identiﬁed
brucellosis as a priority economic and public health concern. A national brucellosis
control program is currently underway in Mongolia. Data from this study contributes to
the evidence that brucellosis control in livestock is an important step in not only
improving the productivity of the livestock sector, but also in enhancing the health and

well being of all Mongolians.

145

Table 4.1. Prevalence of Clinical Signs and Symptoms of General Illness
Experienced Within the Previous 12 Months and Reported by the 100
Individuals Surveyed in the Khangai Mountain Region of Bayankhongor
Aimag (Province), Mongolia, in August 2000.

 

 

Symptom Yes No Period (12 months) Prevalence
Joint Pain 18 82 18%
Headache 16 84 16%
Lower Back Pain 12 88 12%
Stiff Joint(s) ll 89 , 11%
Swollen J oint(s) 10 90 10%
Extreme Weakness 9 91 9%
Muscle Pain 9 91 9%
Bone Pain 9 91 9%
Dizziness 7 93 7%
Abdominal Distention 6 94 6%
Difﬁculty Sleeping 4 96 4%

Enlarged Lymph Nodes 2 98 2%

Fever 2 98 2%
Night Sweats 2 98 2%
Abdominal Pain 1 99 1%
Chills 0 100 0%
Depression 0 100 0%

 

146

.m 55 $2 3550 023 £30 05 .00 $3 5:? $2 838 Manama 05 89a tam—£303 Boga 2: 00 um: 05 83065 *

 

 

 

*momno omo Q N *3; mo 3N 3 o @6288 No 55ka
£8:me £95395
*NmRo ovo mm v _ *NGmo $4 . VN N €30:me 82m:
£8:me ammozooam
oz v.8» e Z 8?
m_me=oo=..m «wedge—Em
3.3.50 .3 339:. 33¢. no 329:. mmmeuwaa
253.: "X ESQ—Pmm . 2:51. "X aeeaﬁmm ammo—.02.:—
.m_mo=ooEm 2:020 3393. .5.“ 28>: $o~
ammo—Guam 88¢. 3293. poo rang: $2 mooﬁuz .ooom Jmswz< E .azowaoz Gunﬁght maﬁa Homnoﬁcgam oo
commom £8582 RwSEM 05 E 9&0de 3.26385 oo_ 05 3 98:85.5 2:020 one oSo< Mo 3293. mesa—Pam
ER 2me 33:20 00 85580 venomom 05 one 2883* E £8:me mac—30:5 32.83% 00 mix—«5‘ BethD .Né 035—.

147

.m 82: 0.0.2 8580 023 0:00 2: oo $mN :08.» :00: 00000 0.0055 2: So: DSEBEQ :20me 05 mo 00: 2: 00:00:05 ...

 

 

 

*oowvo omo 2 .0 «echo oo.o 3 o 03:00.“ a 0
0080::00IHm0
*omooo oo.o m o N *vommo mm; m: N 0>£mom
my? 0 “A
onwoo oo.o N o mmNNo 5: N N. 0>Emom
Hmm
*oNono “No R v ooao oo.o N o 0>:_mom
g 0
wand No 3 w oomoo oo.o on no 0>:mmom
% 0 .8 .55
*wVNno Ed 2 N *ooSo m _ .N m: m 03:09“
my? 0 0:0 Hmm
02 00> 0 Z 00>

£02.35 ammo—003m 33:03.5 5:» 2505 ammo—.00....m
020?; ”N A68 N~ 5:5 aE0000? 030.»-9 ”X .0850:— au0>me :00: 0500

062—0095 2:39 $002035

.0688 N_ 53-30093. 8.: 200?. $9.3 $620005 5:3 oomoqwﬁa $0.0m: 8.: ..00>: oo.oN mooonz

.oooN Jmswn< E dmowcoz 0203050 weﬁdq. Swaoﬁagmm mo GOE0M £8502 398an 0.: E :0>0>.Sm
20:03:05 oo: 2: E £8:me ammo—003m Etomom 0:0 038m 0000:0005 0000 0:000 00 07.302 80.590025 .m.v 030B

148

.m 55 $2 $580 26: £3 05 mo $3 5:? “m3 888 955E oﬁ 88m b23305 BE?N 2.: mo 8: 06 8:365 *

 

 

 

.823 Ed 5 m *vmﬂmd mo; om N o>£mom go
cogmcoU-EU
nmcod 3d 2 o 38.0 aﬁm S m 9568
20 _A
vmmmd 31 an m @306 vmd mm c 2580A
Ham
oomod 5: cm N" we“ ﬁo 3N mm a gamma
go
Nmmﬁo ac.“ mm 2 voKd Ed Q a 338m
20 no Ham
38.0 3.? S m 88.0 5N4” M: c 2560A
20 was Ham

52 3% :2 no?

mama—.ouaum mac—.33:
2.5.30 .3 .333. 8.54 no 339:. 323m ﬁns—35.5
2:5»..— NX 2:39.53 03573 "X magma—mm Ea: 2.30

.m_mo=ooam 0:850

32.5 5 :8? $8 mm50:02.5 3:2 323. § awe? :2 moenz .88 .33 a anomcoz 4855.6
map—2 Homaogaam mo somwom 58502 ngg 05 E vohotsm 835m 2: 05 E mug—002m 030.30 cam BEE
no 329$. mﬁoﬁﬁzm Ea mammm 13:50 no 85580 2: 28 335 mmmosooam 9on 0330 mo mmmbmé 03535; .96 oEah

149

hmomd mmd 5 3 coco; oo.o mo 2 802 300 “mm

 

5- E. ow om ---- ---- em 3 :62 0:30 Em
*mavd 36 m ~ $390 56 ~ N :62 6:80 Em
383 85 K M: %33 3o 2 2 3625 862$ 85

£08834 Hounwaﬂm
*wmmnd mod mo 2 $23 omd E 2 80am “8.33%
ommmd and no 3 camd 34 ac ~ ~ Snow Simsﬂm
”323 So 8 2 *8de m2 Q 2 2:8 5:335
38o; Euo m o *oooo._ and m 0 £250 52%sz
$286 8.: M: .3 . :86 mod . R m 50 E 83:82.
2 8% oz mo>
ammo—.35.:— £258....“—
oEEAU .3 Eumﬁmh 83¢. .3 389:.
2.3.1— Nx maﬁa—Pa 2:97.“ "X 2:33:58 33am :ma

 

.m_mo__ooEm 2:030 33% H 8m amo>s ASou ammo—303m 83<

329$. .8.“ :8? $3 mooﬁnz .oooN Jmswz< E 433952 68323 maﬁa Homcoﬁthwm mo commom £8552

RwSﬁM 05 E coxogm EEGEBS oo— 05 E ammo=ooam 2:020 93 23" mo 329$; mEoEEzm was mama
Row—=0 no 3:05.80 05 EB ”Sauna vim BaﬁoqaoumasooO Ea 8223mm BEBE Eucouom .3 2min: 03.5383 .mé oEah

150

Soov

wooed

*E‘ood

* Kcod

wooed

*oooé

*memd

*vnwmd

camd

*ovmod

3.2

3.2

unﬁ—

50$

N03

mod

36

:4

$4

mmd

E

E.

w_

Z”

2

mm

2

E

500V

ooood

*mmood

*mmood

33.0
336
“*8de

.525. _

wodm

ems

3‘12

£9

34
mmd
24

mod

ON

mm

3

a

Nﬁ

wk.

m # v22 3mm 88m
326on EEO 0.89on

o Auﬁooam 88on 87c
3:2 and M55

m é: 9825 %m V35
M £5 £80 Em aim

Anzmmaazgm 8m 98 32:8 85
0 V52 9300 3am x55

o Va: m._o§u 5a v35
3 . 332:3 8S 83.» é:
m googm é:
m 380 55
V 3:33 83 2:8 é:
o Esau £5

2 so: 32a 3m

 

82.850 .3 2...;

151

.m 55 83 3:58 023 33 05 .8 $3 can? 62 Hoax“. $855 2: Bob ban—Boa BETN 05 («o 08 05 83085 ..

 

Nmmwd

*:d_d

*owmod

m—mdd

anmd

waded

Eugd

*mdomd

vod

ﬁnd

3d

mmd

and

cmd

voé

mm;

Vb

S

on

5»

mm

3

dd

3

2

2

.525. M

*wmcdd

addddH ﬂ

wwddd

owomd

mwmhd

m Sod

“50: d

cod

36

mod

hwd

Ed

:d

2.3

OWN

R

we

mm

_m

mm

N

2

[\

2

2

noummsooo :850:

9582 £me 38:3
ﬁgatouoxr

88H
835.5 89834

33%—3m
#08834 3808800

850on 88on 829
88824 Bum

A8mm=§gm 83 83> Eon
325 P53

380 Bum

A888 83 2sz Hum

£250 Bum

 

8:368 .m... 2...;

152

 

 

 

mamd co; 8.3 at; no Hoz
3.2 5.3 ~ ~ “mo> 3.3.50
mom ﬁo $1 2.3 3.3 3 H02
wmdﬁ wwév 2 ”8% 83¢
02.3.9 E dd ow< :32 :
doom .Hm:w=< E azowcoz Avast/8nd was“? Howcoﬁaaxmm .«o
commom c5562 «$53M 2: E coxogm mﬁsﬂtcﬂ 02 05 E Emo=ooam 088:0 can 83< mo 339C. mEoEgnm
Ea 29m 30:50 mo 85580 05 can ow< 830: 5250mm 5:3083‘ 05 mo 338-5 mix—«5‘ BathD .wé 035—.

153

 

 

 

mﬁmd mod 3“ a 336 ow.m am v 3333
R : 0m .2 A225”:
”houseﬂy
cl no? :2 8%
38:93.5 mac—Enem—
o_=e.:_U .3 183:. 83< .8 ﬁnish
«EST.— NX 3.3.—ﬂea 2:51— NX mEcua—Emm .58.:— gm?—
doom Jmsw=< E .mzowcoE Aooﬁ>okc maﬁa
homcoﬁgzmm mo somwom E8552 RwEﬁM 05 E voxogm wﬁsﬂzcﬁ o3 05 E mmmozooem 3:020 9a 8:3“ mo
E29? mEoEgnm can ,3me 33:20 (Ho $55.80 05 93 5980 cooéom seaﬂoommaw 05 mo max—mg oatgab .56 035.

154

.m 55 $2 $580 at“: 230 05 mo Xmm 5:3 63 SSE 985$ 05 Bob onmnoE ooEm-N 05 mo 8: 05 83065 ...

 

 

*oooo; moo m _ m #553 mod 3 v 93.3w ooogmcoﬁom
S 3280 9.335
x8333 Bough:
N 55:96.5
:33 Ed v N *cﬁmd a: V N 3an 895883
Ba 1580 88me
x0285 Raucous
u ecu—5.6.5
*vamo mm." 3 n Pomo on; 2 o AmoSom 6285 .8
825 wouucoo:
N sigma—.28
*ono 3:0. . v m *woomo oo.o m N Gounod 8865 23
625 35:53
g acmmmmﬁmauuh
oz 8> cz mo>
38:03.:— «$2.33.:—
3530 .«c .333. 8.54 .3 .333. ”sausage—ED
2:57.— NX maﬁa—8.5m 2:57.— NX maﬁa—8mm mac—Egan

 

$602835 quEU 329$. So :mo?. Oéom ammo—$03.5 83< Roi»... 8o 28>: $2 Moo_uZ .oooN
Jmswaq. E .msomaoE Aooﬁ>€$ was“? Howcoﬁdgmm .«o commom 58552 RwSﬁM 05 E @3356 £56385
oo~ 05 E mmmo=ooEm 3:030 98 83< mo 18E» 8 mEBEPAm can mama 325.0 mo 082.380 05 van m_mo=33m

E965 8 93>? o5 sommﬂﬁmcﬂb mo 880M 05 mo mama—«$523 Manor/65 332285 mo max—«Ex BathD

.m.v «Bah

155

.m 55 82 £550 02“: £3 oﬁ mo $3 :23 $3 888 985E 05 Bow 325305 32m-m 05 .50 um: 05 8:365 *

 

 

$53 8.5 mm a 583 9: R : 38553805 6m

”85.5 2.5 mm m 52.5 $2 mm : 5:2 335 Sam

$260.5 ED 2.305

.583 8.5 2 N .583 8.5 : v 65°on 89% 57¢

52 Em inn

5555.5 55.5 . _ 5 _ .523 $5 2 N Q§§EM 8m 55. 355 83

5:5 938 Em aim

85.5 8.5 mm 5 82.5 2% M: S 50 a $5282
a Z 8? a Z 8?

35:52 8592 8.5.8.: 8.5. 5852:.—

oEgﬁ NX 5:55...— 2:9»..— NX scammﬁwnauh .885— 57.2

 

.oooN mo “Swag E 332283 .wzowEE cowcoﬁcgmm mo coﬁom £8582 EmSEM

05 E £36385 ooxotzm E :ouoooﬁ mac—303m mo mama Bog—U £3, oBEQOmm< ﬂouomm 57.3 .8 ﬁomgnom mo
368.5 05 98 53535 Emo=ooam can :oEEEmStH mmmozooam «o 830% no 035230qu 6303 .50 5.235089% dd. oEuH

156

Table 4.10.

Final Reduced Logistic Regression Model of Risk Factors for Typical
Signs of General (Acute and/or Chronic) Brucellosis Reported in
Livestock Herders Interviewed in the Khangai Mountain Region of
Bayankhongor Aimag (Province), Mongolia, in 2000.

 

 

Wald X2 95% Conﬁdence
Risk Factor p-value Odds Ratio Interval (Wald)
Age 0.0114 1.09 1.20— 1.16
(Unit=1 year)
Age
(Unit=5 years) 0.0114 1.52 1.10 — 2.11
Neonates Housed
In Ger 0.0018 16.25 2.83 — 93.38
Consuming Raw
Cow’s Milk 0.0067 10.60 1.92 — 58.40

 

157

 

 

Table 4.11. Final Reduced Logistic Regression Model of Risk Factors for Typical
Signs of Acute Brucellosis Reported in Livestock Herders Interviewed in
the Khangai Mountain Region of Bayankhongor Aimag (Province),
Mongolia, in 2000.

Wald X2 95% Confidence

Risk Factor p-value Odds Ratio Interval (Wald)

Sex (Female) 0.0456 5.31 1.03 — 27.32

Consuming Raw

Cow’s Milk 0.0271 6.52 1.24 — 34.40

Preparing Dairy

Products with Raw

Milk 0.0023 9.33 2.22 — 39.31

 

158

Table 4.12. Final Reduced Logistic Regression Model of Risk Factors for Typical
Signs of Chronic Brucellosis Reported in Livestock Herders Interviewed
in the Khangai Mountain Region of Bayankhongor Aimag (Province),

 

 

Mongolia, in 2000.
Wald X2 95% Confidence

Risk Factor p-value Odds Ratio Interval (Wald)
Age 0.0646 1.054 1.00 — 1.12
(Unit=1 year)
Age
(Unit=5 years) 0.0646 1.30 0.98 — 1.72
Neonates Housed
In Ger 0.0012 20.06 3.29 — 122.44
Drink Raw
Cow’s Milk 0.0071 31.17 2.55 — 381.05
Herd Strictly
Bos taurus Cattle 0.0500 5.64 1.00 — 31.80

 

159

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

Summary

These studies have described the prevalence of brucellosis sero-positive cattle
herds in the Khangai mountain region of Bayankhongor aimag (province) Mongolia.
Stratiﬁed random sampling methods were used to select herder households for inclusion
in the studies. Livestock management risk factors for brucellosis sero-positive herd
status were identiﬁed, and clinical signs associated with the disease in livestock were
described. The prevalence of brucellosis-like clinical signs and symptoms in the humans
from the same herder households in the region was described. Livestock contact and
livestock production consumption practices associated with human brucellosis-like
infection were identiﬁed. Univariate and multi-variable logistic regression methods were
used to identify the risk factors for both livestock and human infection, and to model
brucellosis disease in this population.

The high prevalence of brucellosis sero-positive cattle herds recorded in this study
suggests that brucellosis is an important re-emerging disease in Mongolia. The
prevalence of brucellosis sero-positive cattle herds in the Khangai mountain region of
Bayankhongor aimag (province), Mongolia, was similar to historical accounts of
brucellosis in Mongolian livestock before any control measures were instituted in the
1960’s. Cattle in over 50% (54/103) of herds were recorded as brucellosis sero-positive
based on the results of either the Brucellosis Ring Test (BRT) performed on milk, or the
Brucellosis Card Test (CARD), an agglutination test performed on serum. Twenty-four

percent (23/101) of the cattle herds were classiﬁed as positive by both tests.

160

Evidence of brucellosis infection (antibody presence) in the cattle herds of the
Bayankhongor region was associated with signiﬁcant reductions in herd productivity.
The proportion and number of bovine abortions was higher in brucellosis sero-positive
herds and the report of bovine abortion was signiﬁcantly associated with brucellosis sero-
positive status (OR = 25.00; 95% CI. 4.49 — 139.55). None of the speciﬁc livestock
management factors examined in the study were associated with an increased risk of
brucellosis infection, however, increasing cattle herd size was associated with an
increased risk of brucellosis sero-positive status (5 unit increase; OR = 2.90; 95% CI. =
1.41 ~5.94). Larger cattle herd size is likely associated with a more mobile cattle herd and
a higher probability of contact with brucellosis infected cattle in the population.

The investigation of brucellosis prevalence in the human herder population
mirrored the levels of infection detected in the cattle. Although laboratory testing was
not available for conﬁrmation, we identiﬁed approximately 1/4 of the herders interviewed
who had experienced clinical signs and symptoms associated with brucellosis in the
previous 12 months. The occurrence of brucellosis—like disease in humans was
associated with the detection of the disease in their livestock, as would be expected with
this zoonotic disease in a system where livestock herders are primarily consuming milk
and meat from their own livestock. Additionally, those individuals who consumed raw
dairy products, or dairy products prepared with raw milk, were signiﬁcantly more likely
to experience brucellosis-like symptoms.

The practice of housing weak newborn livestock in the family’s living quarters
was associated with an increased risk of experiencing brucellosis-like symptoms. This

practice most likely characterizes the close contact that often occurs between herders and

161

the livestock they care for. The weak neonates, vs. health ones, may themselves be more
likely to be contaminated with infected uterine ﬂuids from their dams. This livestock
management practice, that is perceived to improve the productivity of the herd by
increasing lamb/kid/calf survival, may be a difﬁcult practice to change.

Interestingly, being female was associated with a signiﬁcantly higher probability
(OR = 5.31; 95% CI. = 1.03-27.32) of reporting clinical signs and symptoms consistent
with “acute” brucellosis infection. One would not expect to see any gender bias in
brucellosis infection since males and females are equally susceptible to the disease.
However, in some cases there is a division of labor that falls along gender lines in
livestock herding practices, or different habits of livestock product consumption between
men and women, that leads to differences in exposure to infection. A speciﬁc behavior,
with signiﬁcant gender bias that would explain the increased odds of women
experiencing clinical signs suggestive of brucellosis was not identiﬁed in this study,
however, it is the author’s observation that women and girls are more likely than men to

take care of livestock nearing parturition and the offspring they produce.

Conclusions
0 The prevalence of brucellosis sero-positive cattle herds in the Khangai mountain
region of Mongolia was high. The recorded prevalence of 25% -50% indicates
active circulation of infection in this population and evidence of the re-emergence
of brucellosis in this region of Mongolia.
0 Signiﬁcant production losses in the form of bovine abortion and decreased

calving rates were associated with brucellosis infection in this population.

162

Only increasing cattle herd size was identiﬁed as a risk factor for brucellosis sero-
positive cattle herd status in this region of Mongolia where livestock management
practices are relatively homogeneous.

The prevalence of clinical signs suggestive of brucellosis in the human herder
population in this region of Mongolia mirrored the brucellosis sew-prevalence
data collected from their cattle. Approximately 25% of the individuals surveyed
had experienced clinical signs consistent with brucellosis disease in the previous
12 months.

Risk factors for the occurrence of clinical signs suggestive of brucellosis in this
population included the consumption of raw milk, preparing dairy products from
raw milk, and the practice of housing weak neonates, primarily kids and lambs, in
the family’s living quarters.

Increasing age in humans was associated with a higher probability of reporting
clinical signs consistent with “chronic” brucellosis. Gender (being female) was
associated with an increased odds of the occurrence of “acute” brucellosis-like
symptoms. This gender difference may be related to a gender-based division of

labor in the care of cattle nearing parturition and their newborn offspring.

Recommendations
1. Increase Brucellosis Surveillance in Mongolia

This study revealed evidence of a higher than expected prevalence of brucellosis in
the Khangai mountain region of Bayankhongor aimag (province), Mongolia. Both the

prevalence of sero-positive livestock herds, and the prevalence of humans reporting the

163

occurrence of clinical signs and symptoms consistent with brucellosis, were higher than
would be expected based on the nationally reported levels of brucellosis in Mongolia.

The stratiﬁed random sampling methods for herd selection used in this study could be
applied to the testing of herds in other regions of Mongolia. The random technique
allows for the extrapolation of results across the sampled region with a calculated
precision. The technique poses signiﬁcant advantages over the quota sampling methods
currently employed in Mongolia.

The hospital case-based reporting of human brucellosis disease should be
expanded to include population-based studies of both the “high risk” herder population
and the semi-settled and urban population in Mongolia.

It is also essential that systematic surveillance for brucellosis in livestock and humans
be incorporated into any control programs so that the effectiveness of brucellosis control
measures can be evaluated.

2. Implement a Livestock Brucellosis Vaccination Program

The zoonotic nature of brucellosis dictates that the efforts of eliminating this
economically devastating, and public health impacting disease be focused on controlling
and eliminating the disease in livestock. The herder population in Mongolia has very few
means of protecting themselves from brucellosis. Pre-sale testing of livestock is not
available, and food products produced by their livestock are rarely subject to any
regulatory inspections. Although educational campaigns may be effective in changing
livestock product consumption behaviors (boiling milk and cooking meat) livestock
herders can do little to protect themselves from the risk of direct contact with infected

livestock.

164

As other authors have pointed out (Roth et al., 2003; Zinsstag et al., submitted
2004) brucellosis control in Mongolia should be a priority of multiple sectors including
human health, agriculture and the economy. The focus should be the elimination of
brucellosis in livestock with recognition of the beneﬁts this will have on livestock
production, human health and the overall agricultural economy. This study and others
indicate that the current prevalence of livestock brucellosis in Mongolia is too high to
institute a “test and slaughter” program, but re-instituting livestock immunization for
brucellosis should be seriously considered.

Subsequent to the completion of this study, the Veterinary Services Department of i
the Mongolian Ministry of Food and Agriculture, with some assistance from the World
Health Organization (WHO), designed a-lO-year plan for the control of livestock
brucellosis in Mongolia. The plan involves 2 years of mass immunization of livestock
(cattle with B. abortus Strain-l9, and small ruminants with Rev-1) followed by an 8-year
period of immunization of all young female stock. It is recommended that this plan be
implemented in its entirety, with frequent assessment of its effectiveness in reducing
brucellosis infection rates in Mongolia. The presence of vaccine related antibodies in the
livestock population once this control program is implemented will make disease
surveillance a challenge, but methods are available in Mongolia to distinguish between
vaccine and wild-type Brucella reactions. Funding from both the National Government
of Mongolia and international donors should be made available for the implementation of

a comprehensive brucellosis control program in Mongolia.

165

3. Further Research: Brucella Identification and Longitudinal Data

Culturing and identifying the Brucella species/strain circulating in this population
of livestock and humans would enhance our understanding of the epidemiology of
brucellosis in the Khangai mountain region of Bayankhongor aimag (province) and
provide updated information on the strains of Brucella that occur in Mongolia. A
longitudinal study involving livestock herders and the animals (sheep, goats, horses,
camels, dogs, and cattle) that compose their herds, would also allow us to assess
interspecies transmission of Brucella, calculate disease incidence rates, assess seasonal
trends in disease transmission, and possibly identify risk factors for disease not apparent
in the cross sectional study performed in July and August of 2000, described in this

thesis.

166

APPENDIX A

167

Figure 1.0. Map of the 19 Aimags (Provinces) and 2 Cities (Ulaanbaatar and
Darkhan-Uul) of Mongolia. Inset: Mongolia in Central Asia.

 

Mongolia

 

 

Key: List of Aimags (Provinces) and Cities (Ulaanbaatar and Darkhan-Uul)

8. Bayan-Olgi 8. Bayankhongor 15. Ulaanbaatar (Capital)
9. Uvs 9. Bulgan 16. Dundgovi

10. Khovd 10. Ovorkhangai 17. Khentii

l 1. Zavkhan 1 l. Omnogovi 18. Govisumber

12. Govi-Altai 12. Selenge 19. Domogovi

l3. Khovsgol l3. Darkhan-Uul 20. Domod

l4. Arkhangai 14. Tuv 21. Sukhbaatar

168

APPENDIX B

169

LIVESTOCK MANAGEMENT QUESTIONNAIRE
Khangai Mountain Region of Bayankhongor Aimag (Province), Mongolia

Date of interview:
Soum (district):
Bag (sub-district):

Please list how many and what type of livestock you have:

SHEEP
GOATS
CATTLE
CAMELS
HORSES
YAKS

List the number of adult female animals that gave birth to live babies this year:

SHEEP
GOATS
CATTLE
CAMELS

Where do your animals give birth?

This year, how many baby animals born alive died during the ﬁrst week of life?
LAMBS

KIDS

CALVES

CAMEL CALVES

FOALS

List the number of adult females that aborted or gave birth to dead babies this year:

SHEEP

GOATS

CATTLE

CAMELS

What months did the abortion(s) occur?

How did you dispose of the aborted fetus and placenta? In open space, etc.

170

Did you consult a veterinarian about the abortions?
If yes, how much money did they charge?

When was the last time your animals were vaccinated?
For what disease(s) were they vaccinated?

Have your animals ever been vaccinated for brucellosis?
If yes, when was the last time?

How much money did the veterinarian charge to vaccinate for brucellosis?
Do any male animals in your herd have orchitis or lumpy testicles?
Are any of your breeding males infertile?
Which animals do you milk (circle):
SHEEP GOATS CATTLE CAMELS HORSES YAKS
On average, how many weeks or months are your animals in lactation per year?
SHEEP
GOATS
CATTLE
CAMELS
HORSES
YAKS
How many times do you move your herd per year?
Do neighboring herds mix with your animals while grazing?
In the overnight area?
In the watering area?

In the winter shelter?

What do you think, what symptoms do animals with brucellosis have?

 

 

What can be done to prevent brucellosis in animals?

 

 

171

APPENDIX C

172

 

HUMAN HEALTH AND LIVESTOCK CONTACT/PRODUCT CONSUMPTION
QUESTIONNAIRE
Khangai Mountain Region of Bayankhongor Aimag (Province), Mongolia

Date of interview:

Soum (district): Sex: MALE FEMALE
Bag (sub-district): Age:
Occupation:

Highest level of education completed:

Have you worked with livestock? What kind of work?
veterinary service

commercial livestock slaughter

hide and wool trade

herding livestock

other

3"?pr

Do you herd the following livestock? (circle)
SHEEP GOATS CATTLE CAMELS YAKS
Do you keep baby animals inside the ger? 'YES NO

Do you personally milk the following livestock? (circle)

SHEEP GOATS CATTLE CAMLES YAKS
Do you drink raw (unheated or unpasteurized) milk? YES NO
If yes, milk from (circle) SHEEP GOATS CATTLE CAMLES

Do you prepare dairy products from raw milk? YES NO
Do you personally slaughter the following livestock? (circle)
SHEEP GOATS CATTLE CAMELS
Which animals do you use for meat? (circle)
SHEEP GOATS CATTLE CAMELS
Have you consumed raw or undercooked meat of the following types? (circle)

SHEEP GOAT CATTLE CAMEL

173

Do you handle cattle dung with your hands? YES NO

Do you have any horses with open draining sores on their withers or top of their head?
YES NO

How do you sever the umbilical cord of animals?

Have you ever been diagnosed with the disease brucellosis?
If yes, when?

How have you been diagnosed?
a. Have you ever had a skin test for brucellosis?
b. Have you ever had a serological (blood) test for brucellosis?

Which of the following symptoms, if any, have you had this year? (circle)

FEVER CHILLS HEADACHES NIGHT SWEATS JOINT PAIN
SWOLLEN JOINT(S) STIFF JOINT(S) BONE PAIN MUSCLE PAIN
LOWER BACK PAIN EXTREME WEAKNESS DIZZINESS
DEPRESSION ABDOMINAL PAIN ABDOMINAL DISTENTION

ENLARGED LYMPH NODES DIFFICULTY SLEEPING

What kind of treatment did you receive for-the disease? If possible, list medications.

 

How long were you treated?

a. Were you treated in a hospital/clinic? YES NO
b. Were you treated outside a hospital/clinic? YES NO

Did you experience any relapse of illness after treatment? YES NO
Did you see a doctor (medical ofﬁcer) at the end of your treatment? YES NO

How do you think people get brucellosis?

 

 

 

What do you think can be done to prevent brucellosis infection?

 

 

174

LITERATURE CITED

175

 

LITERATURE CITED

Al-Talafhah, A.H., Laﬁ, S.Q., Al-Tarazi, Y., 2003, Epidemiology of ovine brucellosis in
Awassi sheep in Northern Jordan. Preventive Veterinary Medicine 60, 297-306.

Alton, G., Jones, J ., 1967, Laboratory Techniques in Brucellosis. World Health
Organization, Geneva.

Alton, G., Jones, L., Angus, R., Verger, J ., 1988, Techniques for the Brucellosis
Laboratory. INRA, Paris, 190 p.

Bayankhongor 1999. Bayankhongor Aimag Animal Inventory Data (Bayankhongor,
Bayankhongor Aimag Veterinary and Animal Breeding Department).

Benenson, A. 1995. Control of Communicable Diseases Manual (Washington, DC,
American Public Health Association), pp. 71-74.

Beulig, W., Kopp, H., Vogel, M., 1969, [Veterinary medicine experts of the German
Democratic Republic in the Mongolian People's Republic--an example of socialist
help]. Monatsh Veterinarmed 24, 706-709. '

Black, F., 2004a, Brucellosis, In: Cohen J, P.W.G. (Ed.) Infectious Diseases. Mosby,
New York, pp. 1665-1667.

Black, F.T., 2004b, Brucellosis, In: Cohen J, P.W.G. (Ed.) Infectious Diseases. Mosby,
New York, pp. 1665-1667.

CIA 2004. The World Fact Book-Mongolia (United States Central Intelligence Agency).

Corbel, M., 1989, Brucellosis: epidemiology and prevalence worldwide, In: Young, B.,
Corbel, M. (Eds) Brucellosis: Clinical and Laboratory Aspects. CRC Press, Inc.,
Boca Raton, pp. 53—72.

Crawford, R, Huber, J ., Adams, B., 1990, Epidemiology and surveillance, In: Nielsen,
K., Duncan, J. (Eds.) Animal Brucellosis. CRC, Boca Raton, pp. 131-151.

Darwish, M., Benkirane, A., 2001, Field investigations of brucellosis in cattle and small
ruminants in Syria, 1990-1996. Rev Sci Tech 20, 769-775.

Denes, B., 1997, Serological ﬁndings obtained in cattle herds immunized with the

Brucella melitensis Rev.l and the B. abortus B19 vaccine in Mongolia. Acta Vet
Hung 45, 33-43.

176

Dohoo, I.R., Wright, RE, Ruckerbauer, G.M., Samagh, B.S., Robertson, F.J., Forbes,
LB, 1986, A comparison of ﬁve serological tests for bovine brucellosis. Can J
Vet Res 50, 485-493.

Drost, M., Thomas, P., 1996, Diseases of the reproductive system: female reproductive
disorders, In: Smith, B. (Ed.) Large Animal Internal Medicine. Mosby, St. Louis,
pp. 1555-1556.

Ebright, J .R., Altantsetseg, T., Oyungerel, R., 2003, Emerging infectious diseases in
Mongolia. Emerg Infect Dis 9, 1509-1515.

Emmerzaal, A., de Wit, J .J ., Dijkstra, T., Bakker, D., van Zijderveld, F.G., 2002, The
Dutch Brucella abortus monitoring programme for cattle: the impact of false-
positive serological reactions and comparison of serological tests. Veterinary
Quarterly 24, 40-46.

Erdenebaatar, J ., Sugar, 8., Yondondorj, A., Nagabayashi, T., Syuto, B., Watarai, M.,
Makino, S., Shirahata, T., 2002, Serological differentiation of Brucella-vaccinated
and -infected domesticated animals by the agar gel immunodiffusion test using
Brucella polysaccharide in Mongolia. J Vet Med Sci 64, 839-841.

Femandez—Gimenez, ME, 2000, The role of Mongolian nomadic pastoralists' ecological
knowledge in rangeland management. Ecological Applications 10, 1318-1326.

F oggin, P.M., Farkas, 0., Shiirev-Adiya, S., Chinbat, B., 1997, Health status and risk
factors of seminomadic pastoralists in Mongolia: a geographical approach. Soc
Sci Med 44, 1623-1647.

Halling, S.M., Boyle, S.M., 2002, Foreword. Veterinary Microbiology 90, 1-3.

Huber, J .D., Nicoletti, P., 1986, Comparison of the results of card, rivanol, complement-
ﬁxation, and milk ring tests with the isolation rate of Brucella abortus from cattle.
Am J Vet Res 47, 1529-1531.

J ezek, Z., Rusinko, M., Baldandordj, C., Mingir, G., Ochirvan, S., Hejdova, B., 1972,
Brucellosis serological survey in the Mongolian People's Republic. J Hyg
Epidemiol Microbiol Immunol 16, 426-439.

J ezek, Z., Rusinko, M., Ochirvan, S., Baldandordj, C., Mingir, G., 1974, Epidemiological
study on brucellosis of the Mongolian People's Republic. J Hyg Epidemiol
Microbiol Immunol 18, 160-171.

J oshi, DD, 1991, Organization of veterinary public health in the south Asia region. Rev
Sci Tech 10, 1101-1129.

177

 

Kabagarnbe, E.K., Elzer, P.H., Geaghan, J .P., Opuda-Asibo, J ., Scholl, D.T., Miller, J .E.,
2001, Risk factors for Brucella seropositivity in goat herds in eastern and western
Uganda. Preventive Veterinary Medicine 52, 91-108.

Kolar, J ., 1989, Brucellosis in Eastern European countries, In: Young, B., Corbel, M.
(Eds.) Brucellosis: Clinical and Laboratory Aspects. CRC, Boca Raton, pp. 163-
172.

Lohr, S. 1999, Sampling: Design and Analysis. Duxbury Press, New York, 494p.

McDermott, J .J ., Arimi, S.M., 2002, Brucellosis in sub-Saharan Africa: epidemiology,
control and impact. Vet Microbiol 90, 1 1 1-134.

Microsoft 2000. Microsoft Excel (Microsoft Office).

Mikolon, A. 2000. Report of the WHO mission to Mongolian from 14 August to 3
November 1999 (Geneva, World Health Organization).

Mondongadis 2000. Chief, Bayankhongor State Veterinary Laboratory.

Mongolia 2001a. Mongolian Statistical Yearbook 2000: Administrative Units, Territory
and Climate (Ulaanbaatar, Administration and Data Dissemination Department,
National Statistics Ofﬁce of Mongolia), pp. 21-26.

Mongolia 2001b. Mongolian Statistical Yearbook 2000: Agriculture (Ulaanbaatar,
Administration and Data Dissemination Department, National Statistics Ofﬁce of
Mongolia), pp. 115-150.

Mongolia 2001c. Mongolian Statistical Yearbook 2000: Gross Domestic Product
(U laanbaatar, Administration and Data Dissemination Department, National
Statistics Ofﬁce of Mongolia), pp. 76-83.

Mongolia 2001d. Mongolian Statistical Yearbook 2000: Health (Ulaanbaatar,
Administration and Data Dissemination Department, National Statistics Ofﬁce of
Mongolia), pp. 217-231.

Nicoletti, P., 1980a, The epidemiology of bovine brucellosis. Advances in Veterinary
Science & Comparative Medicine 24, 69-98.

Nicoletti, P., 1980b, The Epidemiology of Bovine Brucellosis, In: Advances in
Veterinary Science and Comparative Medicine. Academic Press, pp. 69-98.

Nicoletti, P., 1989, Relationship between animal and human disease, In: Young, B.,

Corbel, M. (Eds) Brucellosis: Clinical and Laboratory Aspects. CRC, Boca
Raton, pp. 41 -5 1.

178

Nielsen, K., Smith, P., Gall, D., Perez, B., Cosma, C., Mueller, P., Trottier, J ., Cote, G.,
Boag, L., Bosse, J ., 1996, Development and validation of an indirect enzyme
immunoassay for detection of antibody to Brucella abortus in milk. Vet Microbiol
52, 165-173.

OIE 1990-1994. HANDISTSTUS 11 Annual Animal Disease Status (Ofﬁce International
des Epizooties).

OIE 1996-2002. HANDISTATUS H Annual Animal Health Disease Status (Geneva,
Ofﬁce International des Epizooties).

OIE, 2002a, Bovine Brucellosis, In: Manual of Standards: Diagnostic Tests and
Vaccines. Ofﬁce International des Epizooties, Geneva, p. Chapter 2.3.1.

 

OIE 2002b. HANDISTSTUS II Annual Animal Disease Status: Brucellosis 1990-2002
(Geneva, Oﬂice International des Epizooties). E

OIE 2003. Terrestrial Animal Health Code (Oﬁ‘ice International des Epizooties).

Omer, M.K., Skjerve, B., MacMillan, A.P., Woldehiwet, Z., 2001, Comparison of three
serological tests in the diagnosis of Brucella infection in unvaccinated cattle in
Eritrea. Prev Vet Med 48, 215-222.

O'Reilly, D.J., Cunningham, B., 1971, An assessment of the brucellosis card test.
Veterinary Record 88, 590-594.

Otgon, T., 1968, [On the problem of the epidemiology and epizootology of brucellosis in
Mongolia]. Zh Mikrobiol Epidenriol Irnmunobiol 45, 120-125.

Pinigin, A.F., Kokourov, A.P., Petukhova, O.S., Merinov, S.P., 1968a, [Properties of
strains of Brucella isolated ﬁ'om Yaks]. Veterinariia 45, 18-20.

Pinigin, A.P., Kokourov, A.P., Petukhova, O.S., Merinov, S.P., Taranov, V.A., 1968b,
[Brucellosis in yaks]. Veterinariia 45, 46-49.

Prior, M.G., Niilo, L., Reeker, W.H., 1975, Use of the brucellosis card test for screening
cattle in Saskatchewan. Canadian Journal of Comparative Medicine 39, 107-109.

Purevdawa, B., Moon, T.D., Baigalmaa, C., Davaajav, K., Smith, M.L., Vermund, SH,
1997, Rise in sexually transmitted diseases during democratization and economic
crisis in Mongolia. Int J STD AIDS 8, 398-401.

Ragan, V.E., 2002, The Animal and Plant Health Inspection Service (APHIS) brucellosis
eradication program in the United States. Vet Microbiol 90, 11-18.

179

 

Reddy, C., 2003, Brucellosis, In: Veterinary Pathogenic Bacteriology. Michigan State
University Printing, East Lansing, pp. 153-161.

Rossabi, M. 1999. Mongolia in the 19908: From Comrrrissars to Capitalists (The Soros
Open Society). .

Roth, F., Zinsstag, J ., Orkhon, D., Chimed-Ochir, G., Hutton, G., Cosivi, O., Canin, G.,
Otte, J ., 2003, Human health beneﬁts from livestock vaccination for brucellosis:
case study. Bull World Health Organ 81, 867-876.

Sahnan, M.D., Meyer, M.E., Hird, D.W., 1984, Epidemiology of bovine brucellosis in
the Mexicali Valley, Mexico: data gathering and survey results. Am J Vet Res 45,
1 561-1566.

Schelling, E., Diguimbaye, C., Daoud, S., Nicolet, J ., Boerlin, P., Tanner, M., Zinsstag,
J ., 2003, Brucellosis and Q-fever seroprevalences of nomadic pastoralists and
their livestock in Chad. Prev Vet Med 61 , 279-293.

Shapiro, D., 2004, Infections acquired from animals other than pets, In: Cohen J, P.W.G.
(Ed.) Infectious Diseases. Mosby, New York, pp. 977-986.

Slack, M., 2004a, Gram-negative coccobacilli: Brucella species, In: Cohen J, P.W.G.
(Ed) Infectious Diseases. Mosby, New York, pp. 2245-2248.

Slack, M.P.E., 2004b, Gram-negative coccobacilli: Brucella species, In: Cohen J, P.W.G.
(Ed.) Infectious Diseases. Mosby, New York, pp. 2245-2248.

Thompson, S., 1992, Sampling. John Wiley & Sons, Inc., New York, 343 p.
Timbs, D.V., Digby, J .G., Doe, I., 1978, The relationship between the brucellosis card
test and the complement ﬁxation test used in the Brucellosis Eradication Scheme.

New Zealand Veterinary Journal 26, 67-70.

Tserendash, C., 1980, [Use of B. melitensis strain Rev. 1 as a vaccine for cattle in
Mongolia]. Trudy VIEV 52, 120-125.

Tserendash, C., 1984, [Experimental use of B. melitensis strain Rev. 1 as a vaccine for
cattle in the Mongolian area]. Trudy VIEV 61, 21-26.

Tserendash, C., Shumilov, K.V., 1970, [Diagnosis of brucellosis in camels]. Veterinariia
1, 116-117.

UNDP 2001. Dzud in Mongolia (United Nations Development Program and Mongolian
National Defense Board State Emergency Commission).

180

Vanzini, V.R., Aguirre, N.P., Valentini, B.S., Torioni de Echaide, S., Lugaresi, C.I.,
Marchesino, M.D., Nielsen, K., 2001, Comparison of an indirect ELISA with the
Brucella milk ring test for detection of antibodies to Brucella abortus in bulk milk
samples. Vet Microbiol 82, 55-60.

WHO 1998. Proposed Programme Budget 2000-2001: Mongolia (World Health
Organization Regional Ofﬁce of the Western Paciﬁc).

WHO 2000a. Combating Communicable Diseases: Brucellosis Control (Regional
Committee for the Western Paciﬁc Fifty-First Session).

WHO 2000b. Focus: Communicable disease surveillance and response (World Health
Organization Regional Ofﬁce of the Western Paciﬁc).

WHO 2001. Communicable disease surveillance and response: Brucellosis (Regional
Committee for the Western Paciﬁc F ifty-Second Session).

WHO 2003. Food Safety: Issues-Public Health Signiﬁcance of Foodbome Illnesses
(World Health Organization Western Paciﬁc Region).

Young, B., 1989, Clinical manifestations of human brucellosis, In: EJ, Y., MJ, C. (Eds)
Brucellosis: Clinical and Laboratory Aspects. CRC, Boca Raton, pp. 97-117.

Zherikova, A.D., Kupul, Z., Tserenchimid, O., 1972, [Achievements in the control of
infectious diseases in the Mongolian People's Republic]. Zh Mikrobiol Epidemiol
hnmunobiol 49, 16-19.

Zinsstag, J ., Roth, F., Orkhon, D., Chimed-Ochir, G., Nansalmaa, M., Vounatsou, P.,
submitted 2004, Dynamic evaluation of zoonoses control programmes for animal
and public health: The example of livestock Brucellosis vaccination in Mongolia.
Preventive Veterinary Medicine.

181

   

I[initiiggii][in]