.3: a.” ,

r .1:

R}...
an _

. ..l ‘

..;m.d4.. uh 1. cv.

. . .4. tau... 7 .s

.3. fr, 1.... 5.3.x”

no

. -.cI4.‘A
_ 2.. 1
u zil‘ql‘l - b ‘
a ‘4 o II III a: "I. V O‘cl.iut 9‘
. Lu] nu.) .. .11: ..
. .2227 r. fimmun . of ..
‘ -2. .fimwhflcifl-f‘! C1,...ewxn 1.41.0. d. . L
- .1: its},
.WAOrrYfic,‘ ..
.0... I

O

In}.
. v u“..-
2 )Ivn...fl. .o
.-......u.a...13:ox 3:4..- 1
. .. 1h... .
V . n . ‘tJoNN X Halffltflfifui
‘ 14.81.91! . ' inn 1?
It ) Club. 1
“5720 v 14 . .
A A
u... i A
1

7|.

. . .WA A. _ l. .

‘Ilt l C a n.
4 . A3,...uh.qflu.u....4fi.cr...g. . . ., . . .

‘ .Ile. .3...um.u»4...v._v. ‘. .. I . .

_ . i . ... .. .U. ‘. a . . u. . ‘ . . 6.... .
.. _. , .. . .. A .. _. ‘ , . ‘ i. 2.3»... in"?

. , . ... n. ; , ... .. ._ . A .. . . t orfiflmuruftv‘dal 0. ..

n l t n . Va: 5 I . V . ,p ‘ If ‘h; V "l " I ‘ D
. . ,. . 51.... 35315:... .42.. +hkmraf. .ha
. . . . . .u E. «.10; 5:0,..v
o t . I I lgfitp."l-O‘O I Ch‘
_ . . .1... ...1flo.m. . agritil/
qfllclcfi

l. I...- .11! n
.l ...rl4

)1! .
.X, '

. ;'§"E" .-
1"-
‘1.’ .

1
4...
all
..
1.1
.II.
. ‘v
, t.
.0
WW" I.
v Al ‘
.‘1 .1
. .\ “.1?
hail: A v
0. 9|:~.t1
5-1....11.
ll. .13.:551'93.‘
uuu 9|.le tn:
DID-I1.X.§. P. In;

.
. HUROOA‘I7
I r

' 3 1m :31,"

U13:

. rdtdtmmuu .
v Elaltiyvv
I ; . ..

. 4 .1!
I»...
v

‘0..- A4- I
. . - ill... )1 .. Il.1l}v...l]hurhflhn0w
II .. ‘ . . $.41) .flfivrbv‘rflflanuxuflb.h-tl
. . fbmfifiuwuul. 7) {HM annulfiwder ‘ Fl . u. MN»! .
‘u‘nl 4‘ 1x‘l-l-hil'l‘0x an.“

. I. |.:~I.v.n¢l .

,.y.v.\.'. . “1A».
In: 1104.“: v0, UOIHWNILI .|l 1.

. v .V- . . ... ‘
. . . if 1I.\, 3.3 ‘ On. p . 1
A.|\.O.1.1 [\lll. .l,\.‘¢ull\-n‘.0\.‘

 

THE81S STAr su

z 1111111111 IUIIIIIHIHIIHUIIHIIHHHIHIIIII

(1061331293 01710 2926

 

 

 

This is to certify that the

thesis entitled

Development and validation of short instruments
to assess dietary risk for low intake of
fiber and readiness to change fiber consumption

presented by

Jen-Hon Collin Shih

has been accepted towards fulfillment
of the requirements for

M.S. Human Nutrition

 

 

degree in

Major professor

Date 5'14'q :1”

0-7639 MS U is an Aflinnatt've Action/Equal Opportunity Institution

 

 

 

 

  

TuanAnv

Mlchl
Unmfifi.

     

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1M cJClm.p65-p.14

Development and validation of short instruments to assess
dietary risk for low intake of fiber and readiness to change
fiber consumption

By

Jen-Hon Collin Shih

A THESIS

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

Department of Food Science and Human Nutrition

1 997

 

ABSTRACT

Development and validation of short instruments to assess
dietary risk for low intake of fiber and readiness to change
fiber consumption

By

Jen-Hon Collin Shih

In community-based health promotion there is an emerging need for dietary assessment tools to
rapidly screen for nutritional risk. At the same time, the application of behavioral change models,
such as the Stages of Change Theory, to dietary intenrentions has yielded promising results to help
researchers understand a broad range of problematic behaviors. This study addressed the
development of instruments to screen for dietary risk of low fiber intake and to assess stage of
readiness to increase fiber intake in a sample of 99 middle-aged, predominantly female clerical
workers of Michigan State University. Results showed the test checklist had high sensitivity (0.80)
for distinguishing subjects with low fiber intake, and moderate correlation (r=0.52, p<.01) with Zday
food records. Subjects at Maintenance stage had higher average daily intake of fiber than those at
Precontemplation and Action stages and received more social and environmental support than
subjects at the other four stages. These findings: 1) support that the fiber checklist can feasibly
identify a high risk population with low fiber intake; and 2) provide preliminary evidence that people
at different stages of readiness for fiber consumption have somewhat different attitudes and beliefs.

 

ACKNOWLEDGEMENTS

I would like to give my hearty appreciation to Dr. Sharon Hoerr, for her patient and supportive
instruction that guided me to the shrine of advanced intellectualism. Her thoughtful advice outside
the classroom also broadened my perspectives on various issues that enriched my three-and-half-
year “American experience”. Words are not enough to express how much her tutoring meant to
me.

Many thanks to my committee members, Dr. Ralph Levine and Dr. Won Song, for their
heuristic assistance on thesis writing.

Special thanks to my computer consultants here in MSU. Their altruistic help transformed
me from a computer illiterate to an amateur troubleshooter, which enabled me to keep the
indispensable computers in good shape.

There will never be enough thanks to my parents. Their unconditional love and support were
second to none in this universe.

 

TABLE OF CONTENTS

 

 

 

 

 

 

 

Ust of Tables vi
Ust of Figures vii
introduction 1
Research Objectives 4
Hypotheses 5
Glossary 5
Review of Related Uterature 7
Health implications and food sources of dietary fiber 8
Health implications of dietary fiber 8
Status and source of dietary fiber intake 10
The development of dietary fiber analyses 10
Remarks about literature reviewed 14
Factors associated with intake of dietary fiber 15
Determination of potential factors 15
Demographic factors 15
Personal knowledge 16
V Attitudes and beliefs - 16
Remarks about the literature reviewed 18
Modification of dietary behaviors using Stage of Change theory 19
Components of Stages of Change theory 19
The applicability of SCT in dietary behavior change 21
Remarks about the literature reviewed 25
The validity and reliability of dietary assessment instruments 26
Reliability and validity, the concepts 26
Methods of dietary assessment in community-level 26

programs
Validation of new, short instruments in dietary assessment 29
Remarks about the literature reviewed 32
insight of the research from the reviewed literature 34
Methods 35
Subjects 35
Procedures 35
Instruments 37
Checklist 38
Stages of Change Questionnaire 40
Attitudes and Beliefs Questionnaire 42
Statistical analyses 43
Schedule 45
Results 46
Discussion 65
Checklist Validation 65

 

Stages of Change Questionnaire 67

 

 

 

Distribution of subjects across the stages of change 68
Attitudes and Beliefs Questionnaire 69
The rank of foods based on the quantitative contribution to total 70
fiber intake
Strengths 71
Limitations 71
Conclusion 73
Bibliography 74
Appendices 87
Appendix A Permission from the University Committee on 87
Research involving Human Subjects (UCRIHS)
Appendix B The brief explanatory letter sent to all 1,891 CTU 88
members
Appendix C The consent form sent to the selected 401 CTU 89
members
Appendix D The cover memo sent to subjects that didn't return 90
Appendix E The fiber checklist used in this study 91
Appendix F The Stage of Change questionnaire used in this study 92
Appendix G The Attitudes and Beliefs questionnaire used in this 93

study

 

Table 1

Table 2
Table 3

Table 4

Table 5
Table 6

Table 7
Table 8

Table 9
Table 10

Table 11

Table 12

Table 13

Table 14

Table 15

Table 16

Table 17

Table 18

Table 19

Table 20

List of Tables

Physiological effects and proposed mechanisms in vivo of
dietary fiber

Sources of five components of dietary fiber

Recent studies measuring individual's dietary attitudes and
behaviors

Summary of various lengths of time for maintenance used
for selected dietary behaviors in recent researches

Studies that applied SCT in dietary behavior change

Summary of studies developing new dietary instruments
using a behavioral approach

The sources and foods selected for the Checklist

Staging questions and algorithm for dietary fat reduction
used by Curry et al.,1992

Response rate and demographics of subjects

Mean (iSD) of age, height, weight, and BMI of subjects by
gender

Average daily fiber intakes form checklist and 2-day food
record and log transformed correlation

The distribution of subjects [n=64) in high, medium, and
low subgroups of fiber intake assessed by checklist
and 2-day food record

Fiber (g) assessed fiom the checklists and 2-day food
records at each stage of change

Spearman's rank order correlation between stage of
change and estimated grams fiber intake from two
instruments

Mode, mean, and standard deviation of subjects' [n=95)
responses to Attitudes and Beliefs questions

The percentage of variance of responses to AB
questionnaire for which each construct accounts

Correlations of each question with five constructs and or of
three constructs with more than one question

The mode of responses to Attitudes and Beliefs questions
and their distribution by five stages of change

Mean ($313) of the scores for three major constructs at the
subjects’ different stages of readiness to consume
fiber

Stepwise prediction of fiber intake from 2-day food record
using fiber intake from checklist, the subject’s stage of
change, and subjects scores of the three major
constructs in AB quesitonnaire

vi

11
17

21

24
30

39
41

46
46

50

52

54

54

56

57

58

59

60

61

 

 

Table 21

Table 22

Table 23

Table 24

Table 25

Ranking of foods from the checklist for their contribution to
total fiber intake based on the average weekly
number of portions consumed and grams of fiber
consumed from that food per week by each subject

Foods not listed on the checklist but were selected by
subjects more than once

Subjects' response to suggestions and barriers to
increasing fiber intake

Comparison of gram amount of dietary fiber intake
(meaniSDj by stage of change between Shih’s

checklist and data of Glanz et al., 1994; Spomy et ai..

1995; and Huang et al., 1997
Distribution of subjects in five stages of change for fiber
intake from several relevant studies

 

62

63

67

68

69

Figure 1
Figure 2

Figure 3

Figure 4
Figure 5

Figure 6
Figure 6a
Figure 6b
Figure 6c
Figure 6d

Figure 7

List of Figures

Diagram of Stages of Change Theory

The evolution of the classification of dietary fiber
components

Change processes most useful in particular Stages of
Change

A section of eriett-type food frequency questionnaire

A section of 13-item Block-type screening questionnaire
for fat intake

Distribution of fiber intake from checklist vs. 2-day food
records

Original distribution of fiber intake (9) from checklist

Logarithmic distribution of fiber intake (g) from checklist

Original distribution of fiber intake (9) from 2-day food
records

Logarithmic distribution of fiber intake (g) from 2-day
food records

Distribution of subjects in five stages of readiness to
change fiber intake (n=99] [number of subjects are
shown at the tip of bars)

viii

28
29

47
48
49

49

53

 

General Introduction

The importance of dietary fiber to health has emerged in the past two decades with
advanced understanding of its role in human physiology. It has been well documented that
low-fat, high-carbohydrate diets containing 20—35 g dietary fiber per day are beneficial for
health promotion and disease risk reduction (ADA reports, 1993). Nevertheless, the US.
public seems less concerned about fiber than about fat (Putnam, 1994) , even though aware
ofthe linkage between cancer and insufficient amount of fiber in diet (Glanz et al., 1994).
Americans’ dietary fiber intake has been reported as 11 g/day in NHANES II (Lanza et al.,
1987) and 14.8 g/day in NHANES III (Alanio et al., 1994), both about halt‘the goal for
dietary fiber for the Year 2000 (Healthy People 2000, 1990). Possible explanations for
people's low intake of fiber have included: 1) unclear expectations regarding the diet-
disease linkage (Smith et al., 1992; Smallwood et al., 1994); 2) difficulty in making
dietary change (Kristal et al., 1990); and 3) lack of knowledge for selecting high-fiber
foods (Levy et al., 1993; Smallwood et al., 1994).

Assessing the dietary intake of free-living individuals remains among the most
difficult measurement issues in epidemiological research and health promotion program
evaluation (Kristal et al., 1990). The three generally accepted methods of dietary
assessment—24-hour recalls, three-day food records, and food frequency questionnaires——
have been problematic for respondent burden and validity in community-level
interventions where measurement must be rapid, simple, and inexpensive (Kristal et al.,
[990). In this research, we sought to develop instruments which could satisfy the need of
community-level interventions with low respondent burden, while still maintaining

acceptable validity and reliability.

 

7

Studies have shown that the efficacy of behavioral changes will be enhanced, if the
intervention programs are based on appropriate psycho-theoretical models (Heimendinger

’

1993; Glanz et al., 1994 ). A new theory, the Stages of Change Theory—originally

Precontemplation I j
Maintenance |

I \‘ ontemplation '
Relapse I
A]

C
/
Action | \ j
E Preparation |

Figure 1. Diagram of Stages of Change Theory

developed for intervention programs on smoking cessation and substance abuse
management—is rapidly being applied to food consumption (Prochaska et al., 1985;
Sandoval et al., 1994). The Stages of Change Theory proposes that change is a measurable
process through which people progress to achieve a certain desirable behavior, such as
smoking cessation (Prochaska et al., 1988; DiClemente et al., 1991) or reduction of dietary
fat (Curry et al., 1992; Greene et al., 1994; Spomy et al., 1995). The process of change
consists of precontemplation, contemplation, preparation, action, and maintenance stages.
The dynamic, cyclic-like relationship of these stages is shown in Figure l (Prochaska et
al., 1994; Sandoval et al., 1994; Si gman-Grant, 1996). The Stages of Change model can be
beneficial for health professionals and program participants alike, because it allows
demonstration of progress toward desirable change, even lapses, prior to actual change
(Shelton, 1995), The early stages do not involve observable change of behaviors, but
changes and attitudes or beliefs which can be measured. Lapses and relapses are
anticipated as opportunities for improvement and not as absolute failure (Prochaska et al.,

1994).

Because 85% of the US. adult population is employed, the worksite setting offers
immense potential for health promotion efforts (Morbidity and Mortality Weekly Report,
1993). The fact that six of the ten leading causes of death in the US. are linked to diet has
persuaded the industries to provide worksite health promotion programs to their employees
(Healthy People 2000, 1990; Sciacca, 1987). Chronic diseases such as heart disease,
cancer, stroke, and liver cirrhosis have shown clear association with poor lifestyle factors,
especially dietary habits. Knowing that these diseases represent tremendous costs of
medical care and loss of working power, more and more business are aware that modifying
behaviors which contribute to the development of debilitating conditions is more cost
effective than treating established illnesses (Sciacca, 1987). The Surgeon General singled
out nutritional habits as one of five categories as critical to the health and well-being of
Americans and as targets for health promotion programs (US. Surgeon General, 1986).
Many health risk factors for chronic disease are lifestyle behaviors which are alterable, and
dozens of worksite programs have been exclusively designed to cope with them (Pritchard
et al., 1990', Chenoweth, 1987).

Programs to promote desirable eating habits are likely to be effective if based on an
understanding of factors influencing food choices and at the stage at which the individual
is (Sandoval et al., 1994; Laforge et al., 1994). Most interventions for dietary behavior
change are designed for individuals at the action stage, while most people in the general
population are at earlier stages (Glanz et al., 1993; Ockene et al., 1992). Theoretically,
specific strategies should be planned for people at a specific stage of dietary change
(Prochaska et al., 1994; Sandoval et al., 1994). So far few efforts have been made on
applying the theory to increase dietary fiber intake (Glanz et al., 1994; Laforge et al.,
1994).

To sum up, the purpose of this research is to classify a sample population into
different stages of change relative to their consumption of fiber and find out the
characteristics of attitudes and beliefs for each stage, which we believe will improve the
design of future nutrition intervention. Thus, in this research we will: 1) develop and test a
short checklist foods to estimate subjects’ dietary fiber intake; 2) develop and test a stages

of change questionnaire to categorize subjects into different stages for dietary fiber intake;

 

and 3) develop and test a questionnaire to examine individual beliefs and attitudes about
dietary fiber intake.

The reasons follow for selecting the target population, members of Clerical
Technical Worker’s Union (CT U). 1) The majority are mid-aged females who tend to be
the main food providers in family, thus their preference of foods and diets can influence
other household members, especially young dependents. 2) CTU’s are literate and have
experienced questionnaire administration, allowing them to answer the test instruments
without detailed interpretation by researchers. 3) An intramural health promotion
program, MSU Healthy U, provides nutrition and health information routinely to its
participants which are mainly female staff of MSU. The free dietary analysis offered is
considered beneficial to the target population, because it is a costly health service.
Another benefit for participants is that the importance of dietary fiber to health is
reinforced. The results and instruments of this research will be available to the MSU

Healthy U Committee to aid in future program design and evaluation.

Research Objectives

1. Develop a short checklist of foods to assess individual fiber intake and validate it with
concurrently collected 2-day food records from members of Clerical Technical
Worker’s Union (CTU) of Michigan State University.

2. Develop a stage of change (SC) questionnaire for categorizing subjects into different
stages by adapting the algorithm by Curry et al.

3. Develop a short questionnaire assessing individual attitudes and beliefs (AB
questionnaire) related to dietary fiber by using components from Health Belief
Model (HBM) that significantly correlate with dietary intake.

4. Determine the association offiber intake with the different stages related to dietary fiber
by analyzing the checklist and stages of change questionnaire.

5. Determine the association of subjects’ attitudes and beliefs relative to dietary fiber

intake and the stages at which they are.

 

6. Investigate the degree to which fiber intake can be predicted from fiber intake from

checklist, stages of change, and the major constructs in AB questionnaire.

Hypotheses

——

. There will be positive correlation between the gram amounts of dietary fiber intake

assessed from the checklist and from the 2-day food records.

N

. There will be no significant difference between the gram amounts of dietary fiber
assessed from the checklist and the weighted 2-day food records.

b)

. The sensitivity of the checklist to detect high and low intake of dietary fiber will be _>_.7O
when compared to weighted 2-day food records.

A

. There will be a significant difference between the gram amounts of dietary fiber intake
for different stages of change related to dietary fiber.

kit

. Subjects’ responses to the attitudes and beliefs questionnaire will differ according to

their different stages of readiness for fiber consumption.

O\

. Subjects’ fiber intake from the weighted 2-day records can be predicted from 3 short
instruments: the checklist, SC questionnaire, and the major constructs in the AB

questionnaire.

Glossary

* Dietary fiber: the indigestible residue in food, composed of the carbohydrates cellulose,
pectin, and hemicellulose; vegetable gums; and the noncarbohydrate lignin.
(Dreher, 1987)

* Precontemplation: the stage in which that the individuals have no intention to make a
dietary behavior change (e. g., increase dietary fiber intake or cut fat intake), at

least not within the next six months. (Glanz et al., 1990)

 

* Contemplation: the stage in which that the individuals seriously think about changing a
behavior within the next six months. (Glanz et al., 1990)

* Preparation: the time that serious commitment to change has been made. For example,
the individuals in this stage are those who tried to change their dietary behavior in
the past year and who plan to do it within next month. (Glanz et al., 1990)

"‘ Action: the span of time ranging from zero to six months after the initiation of overt
change toward a desired dietary behavior. (Glanz et al., 1990)

* Maintenance: the period beginning six months afier the action has been started and
continuing until the desired behavior change has been adopted. (Glanz et al., 1990)

* Criterion validity: the validity established on the agreement with a criterion measure
or ”gold standard” which might be found either concurrent with the instrument
being validated or in the future. (Patrick et al., 1991)

* Content validity: the validity marked by the extent to which a measurement tool or
instrument is consistent with a specific area of substantive content. (Patrick et al.,
1991)

* Face validity: the validity marked by the extent to which a measurement tool or
instrument is pertinent to a specific area. (Patrick et al., 1991)

* Predisposing factors: the factors affecting a specific dietary behavior utilization
including: 1)belief in the diet-disease connection; 2)perceived benefits of healthy
diet; and 3)knowledge about a healthy diet. (Glanz et al., 1993)

* Enabling factors: the factors affecting health services utilization including: 1)perceived
barriers to healthy diet; 2)socia1 support; and 3)perceived norms for healthy eating.
(Glanz et al., 1993)

 

 

Review of Related Literature

This chapter reviews the literature related to the key concepts of this research. The
literature review is divided into four sections: 1) health implications and food sources of
dietary fiber; 2) factors associated with intake of dietary fiber; 3) modification of dietary
behaviors using Stages of Change Theory; 4) the validity and reliability of dietary

assessment instruments.

 

Health implications and food sources of dietary fiber

Health imglications of dietary fiber

Three decades ago Burkitt and Trowell hypothesized that dietary fiber intake played a
role in the etiology of colon cancer and other diseases (Burkitt et al., 1975) and stimulated
subsequent research. Although interpretation is clouded in some case-control studies by
confounding dietary parameters and the absence of reliable information on fiber content
(Block et al., 1987), the role of dietary fiber has been promoted by nutrition and medical
researches from simply a laxation facilitator to a potential preventive of several chronic
diseases, such as diverticular disease (Brodribb, 1980), colorectal cancer (MacLennan et
al., 1978, Modan et al., 1975; Kaaks et al., 1995), Type II diabetes (Comi et al., 1995;
Anderson, 1985; Manhire et al., 1981; Nuttall et al., 1979), prostate cancer (Ross et al.,
1990; Weisbruger, 1987), hypertension (Margetts et al., 1988; Schlamowitz et a1, 1987;
Wright et al., 1979), and hypercholesterolemia (Jenkin et al., 1995 ; Jenkin et al.,1980;
Jenkin et al., 1979).

Despite the variety of health benefits associated with dietary fiber, however, many of
the mechanisms in vivo remain unclear (Eastwood, 1992). Table 1 summarizes the
physiological effects of dietary fiber and proposed mechanisms. Some of the health
benefits of dietary fiber components might be in part its ability to bind water and other
factors such as bile acids and estrogens (Chenoweth et al., 1976; Key et al., 1979;
Lindegarde et al., 1984). Other benefits might come from the increased transit time of
fecal bulk, colonic fermentation of soluble fiber to butyric acid and slowed absorption of
glucose. When fiber intake is increased, water and fluids must be increased as well to
avoid digestive problems (ADA Reports, 1988). Excess dietary fiber intake may result in a
decreased availability of bivalent metals (Fralich, 1995), however, consumption of a
balanced diet from a variety of foods should not lead to overt vitamin/mineral deficiencies

(ADA Reports, 1988).

 

  

38:85.5 5 <55 5.. 55.5 .52

 

 

 

 

 

 

 

3558.8. mg 5% 38:35.3 >555
5.5585885 555.5 55955 8:55. 55.555: 855885 8:55:85 5:5 .:5 5.55.55. 5:88 5.. 9.33555 5. 5... .8».
5255555 .:5 .555. 8.: ...5. 55555555 .:5 8:58.55: 5. 85585:... 5.55 555.... .55»
.:5 555:. 5:5 5: 5:25:55 55 5.55.. :35 o. 855. 8:5:5 58:5: .:5 88:
5:5 55.55.
28...... =5 35.58.55 5. 5.5 55... 5:5 =5 58555.5: 5. 5:535:55 5.... 5555 3. 88:5 1... 5. 5... .8.
53535.5: 5. 85:5 .55: 5:5: 55: .55.. 8 5 5555555.. :1 5 .:5 5.55.55.
.55: 5:5 55.. 5555555 =5 5:555 55.5.... o. 855:5 55. 55:58:: 5.5 5555
58 8.5:. 35.85:?
0555 85:5 53.85.59 8:55 .55. 55: 585.55 .525 55.... 5:5:_ 5 <5... :55 58.5. .53
5:58:55..." 58:53.55 5:5 55: .5955 =5 5.8.5 5.5 o. 8.5: 55:55: 55.5.
5.55:8 =5 5258:5555 5.55.5.5: .5 555.55 =5 55:88.5: 5. 5:58:55 5:5 3.8.5:. 5. 5... .58
55.8555 .85 =5 5.8.555. 5555555 5x885 o. 5855.5...585 8 .:5
5.55.55 55.. 3.588 :5 55:55 5.5.5 55:52 5 55:.
15885588555. 555555 5.5 555 .8555 9.5 8 5:55:55 5. 5.5 55... 5588.5: 3. 855.5 .551: =5 55:55.5 5. 5... .55.

58.55535 5:555

1.88.8555 5.55.5

5.5538555 5.55.5

52:55. .555. 8.5. 55. 55:. 555.5. .

>..5. 555.. 5.55.55. 5586.5: 5. 5. 5:: 5.5: 5:55.555 58885 88:55:. 8 5 m5::55.:5: 5. 5... .85.
55.55.. «5.5 5. 88:55.5 5588.5... 5:..55 5. 5... .85

525555 85:5 52:85.8: 5. :55: 5:: =5 88:55.... 55.55.. 5:35:55 5.... 5555 >358: 5. 5... .8...
58:55.5 5:: 555.58. 5:5: 35 5:5: :5 5.5 58.55.58. 5 555.5. Q... 55 5. 5... .58

m85 @5855 555.55. 888.5: 5:5 8955 85852.5. 58.5 .55.. 5:5: 55: 38.... I852. 5. 5....58
.:5 58.5 5555.. 5:5 @588 55:58.. 5 85:58. 55.5. .

95:95.: 5.... 55.55. 555.5. 585.85 5:5 55.55 5555 58.555: 3 .:5 52:85.8: <88. 5. 5... .85
5. 855.5 .55. 55: .5 8,5 588 8.55:8 5:: .:5 58.5 55:5..55. >358: 5. 5... .8.

$55.. 5: .:5 553.55 5. :5: 5.8.. 258:5 5585:5255 5.. :5: 552.: 5:55.588: .:5 5.5:. 5. 5... .me
38.5.5535 5.555 5. .55: 55: 8:85:55: 855. 5.85 25585.

255.55 58.5 58:55.5: 5:: 5:5:55 =5 855.5: 5. 5.5555: 5:5 5555555855. 1555. 5. 5... .53

5:5: 5 85 55 5.55. 5.5558 5 .53.

 
 

 

Status and source at dictum: trber intake

Nutritionists, food and supplement manufacturers, and health promoters have made
intensive efforts to spread the word about the virtues of dietary fiber, however, intakes
remain far below the levels recommended by US. Department of Agriculture (USDA), the
US. Department of Health and Human Service (DHHS), and other health authorities
(Smallwood et al., 1994). The third National Health And Nutrition Examination Survey
(NHANES III) revealed the mean and standard error of dietary fiber intake was 17.001027
grams for males and 12.755019. grams for females (Alaimo et al., 1994), while the
recommended daily intake is 20 to 30 grams by the National Cancer Institute (NCI) (Pilch,
1987) or 25 grams per 2,000 kcal diet by Foods and Drugs Administration (FDA) on food
labels (FDA, 1993). Surprisingly, data from USDA’s 1989-90 Diet and Health Knowledge
Survey (DHKS) and its associated Continuing Survey of Food Intakes by Individuals
(CSFII) revealed that 51% of 2,880 respondents thought their diets were “about right” in
fiber and those who thought they should lower their fiber intake consumed only 7.3 grams
per day (Smallwood et al., 1994). These findings imply an need for nutrition education to
educate people about their actual consumption of dietary fiber as well as the rich sources of

dietary fiber and its benefits to health.

The develo men! 0 dieta tber anal ses

Dietary fiber comes only from plant components and was first defined by Hipsley
(Hipsley, 1953) to describe the plant cell wall components of food including insoluble
lignin, cellulose, and hemicellulose. The definition of dietary fiber has been broadened by
some to include soluble substances such as pectins, gums, and mucilages (Trowell, 1976).
Table 2 summarizes the general sources of dietary fiber. This current broad definition
acknowledges the significance of fiber as a chemical and physiological component of the
diet.

As a result of findings from medical research on the health benefits of fiber, the public

policy on dietary fiber has been modified in the past three decades. Based on the definition

ll

that fiber is ‘the insoluble material remaining after severe acid and base hydrolysis’, the

US. Department of Agriculture (USDA) Handbook 8 provided the first comprehensive

data on the amount of crude fiber in foods in 1963. As fiber analysis methods advanced,
USDA released version 4 of the USDA nutrient database in November, 1990, which
included both soluble and insoluble dietary fiber values of foods (USDA, 1990). Because

this database is machine-readable, it is widely used for individual food intake surveys. In

January, 1993, Food and Drug Administration (FDA) and USDA issued final food-labeling

regulations including nutrition labeling and health claims for dietary fiber (Lee et al.,

1995). However, there is no level set for dietary fiber in the current Recommended Dietary

Allowances (National Research Council, l989).

 

 

 

 

 

 

Table 2

Sources of five components of dietary fiber

Cellulose Wholewheat flour Wax beans
Bran Broccoli
Cabbage Brussels sprouts
Young peas Cucumber Skins
Green beans Peppers
Apples Carrots

Hem/celluloses Bran Cereals
Whole grains Mustard greens
Brussels smouls Beet root

Gums Oat meals Rolledoat products
Dn‘ed beans

lignln Breakfast cereals Eggplant
Bran Pears
Older vegetables Green beans
Strawberries Radishes

Pectin Squash Cabbage
Apples Dried peas
Citrus fruits Carrots
Cauliflower Strawberries
Green beans Potatoes

 

The sophisticated structures and physico—chemical properties of dietary fiber

components have provoked much argument about how they should be classified. The

l2

advance of analytical methods for these components and better understanding of their

physiological effects on health has helped the development of an unanimous classification

system. In 1995, an international survey was conducted to get the views of 147

professionals in the field on the definition of dietary fiber (Lee et al., 1995). The survey

also solicited opinions on analytical methods for nutrition research. Figure 2 shows the

most updated classification system of dietary fiber components and its evolution.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1_l
Non-starch Non-cellulose Gum and Other 833:"
F°°d 9 m Polysaccharidee Polysaccharides 9 ”"53““ '
Polysaochaides Dietary Fiber Mes
Pectin
Henioellulose
Cellulose
Lignin
Enzyme-
Resistant
Starch
,Degradable.
l Starch l
l l

 

 

 

‘The solubility of some components depends upon the method of analysis used. Thus, some of the pectin fraction
may be treated as “insoluble”, or a pan of hemicellulose may be treated as “ soluble".

Figure 2. The evolution of the classification of dietary fiber components (ILSI, 1994)

13

Two types of analytical methods, gravimetric methods and the Southgate method, are
commonly used to determine dietary fiber content of foods (Asp et al., 1992). The
gravimetric methods, used over 130 years ago, weighed the insoluble residue (crude fiber)
after extraction with boiling dilute acid, alkali and enzymes. Gravimetric methods can
provide results varying from the amount of total fiber to its fractionated components,
depending on the degree of sophistication of extraction. Soluble fiber is not included in the
results of gravimetric methods, because the intensive extraction severely damages fiber
structure. Some proteins and minerals may be in the residue which can falsely increase the
amount of total fiber (Asp, 1995). To correct these analysis problems, a frequently
performed method uses alcohol or enzymes to precipitate soluble fiber before the intensive
extraction (Selvendran et al., 1984). The enzymatic-gravimetric methods were named the
most practical and simplest ways to determine major components identified as dietary
fiber (Prosky et al., 1984; Lee et al., 1995) and adopted by the Association of Official
Analytical Chemists (AOAC) as Official Methods 985.29, 991.43, 992.16 (AOAC, 1990).

The second type of methods to analyze dietary fiber content devised by Southgate
(Southgate, 1982) were originally used to provide data for the food composition tables in
the United Kingdom. These methods employed more sophisticated extraction than the
gravimetric methods, therefore, the results have more detailed chemical information about
the fractions of both insoluble and soluble fiber, especially low molecular weight
polysaccharides (Asp et al., 1992). The most updated Southgate methods use automated
gas-liquid or high performance liquid chromatography for separating sugars after
hydrolysis of the fractionated fiber components.

Each type of method—gravimetric, and Southgate, has strengths and limitations, so
there is no best choice for every analytical attempt. Gravimetric methods usually suffice
where information on dietary fiber is required for regulatory purposes. If the research
emphasizes on the nature and physiological properties of dietary fiber polysaccharides,
however, Southgate methods are more feasible and preferred (Asp, 1995; Asp et al., 1992).

AA

[4

Fiber is rich in foods such as whole grains, fruits, vegetables, and legumes (Anderson
et al., 1994). A recent report stated most Americans fail to consume even one whole-grain
food per day, much less than three, which might explain why the average daily intake of
fiber is less than half of the recommended value (Alberton et al., 1995). From the 24-hour
recalled dietary intakes of the NHANES 11 survey, Block and colleagues discovered the
major single contributors of dietary fiber in the US. diet were vegetables, which provided
28% of daily fiber, breads provided 19% and fruits 7% (Block et al., 1987). Similar
findings reported by Marlett and colleagues indicated that one third of the daily fiber intake
of 200 college students was from grain products and legumes (Marlett et al., 1981).

A secondary analysis on specific food sources of fiber using data from the 1985 CSFII
by Krebs-Smith and colleagues found that tomatoes, yeast breads, potatoes, green beans,
soy products, and French fries contributed about 40% of total dietary fiber in diets of 1,459
women 19 through 50 years old (Krebs-Smith et al., 1992). The same analysis also
revealed that subjects received an average of 42% of their dietary fiber from cereal and
bakery products and 28% from vegetables and potatoes (Smallwood et al., 1994). In the
Bogalusa Heart Study, conducted from 1976-1988 on three different age groups of 2,1 1 8
children and young adults, Nicklas and colleagues reported that vegetables, soups, breads,
and grains accounted for 53% and 70% of the total fiber in the diets of 10- and 13-years-
olds, respectively (Nicklas et al., 1995).

Remarks about literature reviewed

Over the last 30 years, the role of dietary fiber in nutrition has been changed from a
“non-nutritive ingredient” to a factor “representative of healthy diet” as analytical methods
and experimental dietary studies have progressed. Much has been written concerning the
advantages of a high-fiber diet on various chronic diseases, however, the average daily
intake of Americans still falls far behind the recommended value and people appear largely

unaware of what foods are good source of fiber.

61
AA

15

Factors associated with intake of dietary fiber

In the past two decades increasing concern and knowledge about the role of diet in the
etiology of chronic disease have led to the development of dietary guidelines by health
authorities in US. and other industrial countries. One major component these guidelines
share is the emphasis on reduced fat intake and increased fiber consumption (National
Research Council, 1989). To achieve the goals of these dietary guidelines, such as an
increase in average consumption of fiber to 20-30 grams per day suggested by National
Cancer Institute, an understanding of factors that strongly influence individual’s food

choice is indispensable for the development of intervention programs.

Determination of gotential (actors

A broad range of determinants contributing to dietary behavior has been recognized,
including intrapersonal, interpersonal, institutional, community, and public policy
influences (Glanz et al., 1990). Research examining dietary behavior has tended to
concentrate on personal factors such as nutrition knowledge (Johnson et al., 1985); health
and diet-related beliefs (Jensen et al., 1992); confidence of making dietary changes (Vega
et al., 1988); and feelings of personal control over dietary behavior (Saltzer et al., 1978).
Once potential determinants of specific dietary behavior are identified, these can provide a
foundation for developing effective strategies for dietary change (Glanz et al., 1993).
Additionally, measuring these variables might make it possible to evaluate change in
mediating factors, in addition to the major endpoints, of food consumption (Kristal et al.,

l990).

Demographic factors: Studies have found low socioeconomic status associated with
lower dietary fiber intake (Baghurst et al., 1990; Smith et al., 1992', Kushi et al., 1988).
Patterson and colleagues reported that people who had low-fat, high-fiber diets tended to
be older women with high incomes and college education (Patterson et al., 1994). Specific

dietary preference is another significant factor influencing dietary fiber intake. Vegetarians

 

l6

typically consume a high levels of fiber—perhaps 40 grams or more daily—compared to
non-vegetarians, while the historically carnivorous Eskimos eat virtually no fiber at all
(Eastwood et al., 1983).

Personal knowledge: Correlated with low socioeconomic status, lack of knowledge
may be another factor affecting fiber intake. A study using a telephone survey of
foodservice specialists in 242 military installations found only 24% of the respondents,
who nearly all perceived that they offered a whole-grain bread, actually did. Findings
indicated brand names, misleading labels, and lack of knowledge in identifying real
fiber-rich foods were associated with low fiber intake (Warber et al., 1996).

In a telephone survey on 407 adults residing in Providence, RI, Laforge and
colleagues used psychosocial approaches to probe for factors associated with low fruit and
vegetable consumption (Laforge et al., 1994). They found that education was directly
related to fruit and vegetable intake, but was not correlated with those who usually
consumed fewer than 5 servings and had no intention of adopting the national S-A-Day
program. Studies done by other researchers (Bunch, 1990; Subar et al., 1992) also

confirmed that low educational level is likely to be associated with low fiber intake.

Animdgs and beliefs: Based on data from 1985 CSFII and 1989-90 DHKS, another
psychosocial factor influencing dietary fiber consumption proposed by Frazao and
Cleveland was the individual’s assessment of the adequacy of his or her own intake
(Frazao et al., 1994). They found people who regarded themselves as consuming enough
dietary fiber in the diet, actually ate less dietary fiber than those who thought their fiber
intake was not enough, even though both groups fell well below the recommended daily
values for dietary fiber.

To acquire quantitative data for assessing abstract dietary attitudes and beliefs,
numerous researchers have generated questions using concepts from the Health Belief
Model (HBM) (Contento et al., 1990; Dittus et al., 1995; Patterson et al., 1995; Kristal et
al., 1995; Trenkner et al., 1990). The HBM is one ofthe most commonly applied models in

nutrition education (Contento, 1995) and a major organizing framework for explaining and

.-

l7

predicting acceptance of health recommendations (Glanz et al., 1990). For the concept of
improving disease-related dietary behavior, the HBM would suggest that behavior is
influenced by a readiness to take action and intervening beliefs about the behavior.
Readiness to take action related to health could result from the perceived susceptibility to a
disease state or concern about nutrition (Hayes et al., 1987; Dittus et al., 1995). The key
components of the HBM are: l) perceived threat; 2) outcome expectation; and 3) self-
efficacy (Glanz et al., 1990). A list of recent studies using IBM and other psychosocial

theories to measure individual’s dietary attitudes and beliefs is tabulated in Table 3.

 

 

Table 3
Recent studies measuring indiviglral's dietary attitudes and behaviors
Researchers Subjects Dietary behavior Psychosocial model(s) used in

investiga_ted designing Questionnaire
Cotugna et al., 22,043 adults cancer prevention diets Not specified
1992 nationwide
Dittus et al., 1995 1,069 Washington fruits and vegetable HBM
State residents intake

 

 

 

Contento et al., 117 adult comprehensive, positive HBM
1990 supermarket dietary changes Theory of Reasoned Action
sho rs Health Locus 'of Control
Smith et al., 1995 487 adult 12345+ Food and HBM
Australians Nutrition Plan diet Stages of Change
Health Locus of Control
Kristal et al., 16,287 healthful diets including HBM
1995 participants of increase fiber, fruits Stages of Change
Working Well and vegetable intake Social Learning Theory
Trial and decrease fat Diffusion of Innovation

intake Social Supgrt Theory

 

Remgrks about the literature reviewed

Aside from the traditional advocacy focusing solely on the health virtues of fiber, an
understanding of the psychological, socioeconomic, and other potential attitudinal factors
that influence intake of dietary fiber may be helpful for nutrition educators and health
promoters to screen for those with low intake and design effective interventions. The
concepts of the HBM offer comprehensive perspectives on individual’s perceived barriers
and motivations for desired dietary change. However, for dietary fiber, more efforts are
needed to clarify the contribution of these psychosocial factors to tangible change of
dietary behaviors such as attitudes, beliefs, and self-efficacy. The research does suggest
that some of these psychosocial factors might predict those at risk for low intakes of dietary
fiber.

 

19

Modification of dietary behaviors using Stage of Change
theory

Comgonems ot Stages 0t Change theory

Stages of Change theory (SCT) was developed by Prochaska, Diclemente and
colleagues in the early eighties (Prochaska et al., 1982; McConnaughy et al., 1983) in an
attempt to understand how and why people make health related changes. They proposed

that change occurs through a series of stages and processes shown in Figure 3.

 

Piece/item lat/on Contem lation Pre rat/“on Action Ma/htenance

Consciousness-Raising _

Social Liberation
Emotional Arousal
Self-Reevaluation
Commitment
Reward
Countering
Environment Control
Helping Relationships

Figure 3. Change processes most useful in particular Stages of Change (Prochaska et al.,
1992; Prochaska et al., 1994)

I A'

The definitions of these stages are: r ‘ r or not thinking about
making a change; contemplation—seriously thinking about changing; preparation——
making definite plans to change; action—actively modifying an undesired behavior; and

' ‘ ' ' U the new, favorable behavior for certain length of time which
might be from three months to one or two year. Figure 3 also shows the types of
intervention efforts thought to be appropriate at each stage to lead people to the next stage

of readiness to change.

 

A4

20

The SCT initially was designed for psychotherapy intervention programs such as
smoking cessation (Prochaska et al., 1983; Prochaska et al., 1985; Prochaska et al., 1988;
DiClemente etal., 1991); alcohol intake reduction (DiClemente et al., 1990; Norcross et
al., 1991 ); adoption of physical activity (Marcus et al., 1992); and prevention of coronary
heart disease (Ockene et al., 1990). Some claim that the SCT might be an integrative
theme for understanding and accelerating change in a wide range of problem behaviors,
including dietary ones (Prochaska et al., 1992). Several researches have applied SCT to
promote fat intake reduction (Curry et al., 1992; Greene et al., 1994; Spomy et al., 1995);
cholesterol reduction (Southard et al., 1992); fiber intake increase (Brinberg et al., 1990);
and adoption of healthy diets (Glanz et al., 1994). Broader applications have used SCT for
weight management (O'Connell et al., 1988; Prochaska et al., 1992), which is not really a
behavior but requires a group of behavioral changes in diet and exercise patterns and social
support, to achieve optimal weight.

In SCT, stages link to one another in a dynamic, linear-like pattern. Each stage,
however, does not inevitably lead to the next one, because it is possible for the individual
to move back to the previous stages or leave the system entirely (Prochaska et al., 1992).
For this reason, some researchers have questioned the use of the term “Stage” to refer to
levels of readiness to change wherein individuals can regress (personal communication).
In physiology, for example, a stage refers to an immutable progression of development
towards maturity (Rhoades and Pflanzer, 1992). Perhaps “state of change” would be more
appropriate term to use than “stage of change” for this model.

In the maintenance stage, there is no a priori reason for a particular time span for a
particular dietary behavior and this is an area which could benefit from research.
Researchers customarily have decided the length of time span for the maintenance stage
based on that in previously related studies (Spomy et al., 1995). The time lengths used for
maintaining desired dietary behaviors in some related studies are listed in Table 4 and vary
from “made the change” to “longer than 2 years". “Longer than 6 months” has been the
length of time used most frequently and by researchers funded by National Cancer Institute

to study intake of fat and fruits and vegetables (Glanz et al., 1994).

49

A4

 

21

 

 

 

 

Table 4

Summary of various lengths of time for maintenance used for selected dietary behaviors in
recent researches

Desiregdie_tarxbehavior Time length for maintenance Source

Low-fat and high-fruit and presently trying Campbell et al., 1994
vegetable diet

Low-fat diet longer than 6 months Cuny et al., 1992
Low-fat and high-fiber diet longer than 6 months Glanz et al., 1994
Low-fat diet longer than 6 months Greene et al., 1995
Seal/$353 frurts and longer than 6 months Laforge et al., 1994
Overall dietary change already made change Smith et al., 1995
Low-fat and high-fiber diet longer than 2 years Spomy et al., 1995

 

The agglicabilitv of SC T in dietary behavior change

Because most diet-related health problems develop gradually and do not present
immediate symptoms, to be effective, programs promoting healthy dietary behaviors must
reach large segments of the population and influence the diverse factors which determine
eating patterns (Glanz et al., 1988). The first use of SCT in programs promoting healthy
eating habits was regarded as a conceptual and methodological challenge (Glanz et al.,
1993). Glanz and Mullis explained that the attempt to clearly and correctly categorize a
person into a specific stage for dietary behavior change required attention to the unique
characteristics of eating patterns and practical issues about the eating environment which
were barely researched at that time (Glanz et al., 1988). Subsequently, some algorithms
and instruments were tested and validated for their effectiveness on categorizing
individuals into different stages of behavioral change (Curry et al., 1992; Glanz et al.,
1994).

A4

22

For healthy people of normal weight, most diet-related health problems do not present
immediate or dramatic symptoms, and are not the targets of social stigma compared to
alcohol or drug abuse or obesity. Without a provocative psychological reasons, most
healthy people are not motivated to make dietary behavior change, even if their diets are
inadequate (Glanz et al., 1994). Researchers have found that traditional health education
messages and action-oriented interventions may be counterproductive for people at the
Precontemplation stage, because people in this stage are highly resistant to change and do
not believe any change is necessary or that their current behavior threatens their well-being
(Ockene et al., 1992; Ershoff et al., 1989). Because the SCT underscores the need and
strategies to move from precontemplation to contemplation, its concepts offer individuals
promise of future change in dietary behaviors (Brownell et al., 1995) and are considered as
effective ways to decrease early dropouts (Prochaska et al., 1986).

A few studies have reported the applicability of the SCT used in promoting dietary
change in healthy populations. A summary of several reviewed studies is in Table 5.
Curry and colleagues assessed stage of dietary fat reduction and its association with fat
intake using telephone interviews with two predominantly white samples in Washington
state. They found consistent results across the two samples indicating more women than
men in action and maintenance stages, and significant associations between stage of
readiness for dietary fat reduction and percent of calories from fat (Curry et al., 1992).
Rossi and colleagues evaluated distributions for four different staging algorithms for
dietary fat reduction and compared the findings across three samples in New England
using in person, mail, and telephone methods. The distributions differed between the
algorithms and, with all four algorithms, dietary fat consumption levels decreased with
successive stages of change (Rossi et al., 1993).

Glanz and colleagues assessed the psychological stages of change in relation to both
dietary fat and fiber intake on two samples— predominantly middle-aged, white males—
using food frequency and questionnaires about psychosocial factors. Findings indicated
that a greater proportion of the population had actively tried to reduce fat intake than to
consume more fiber; and intake of fiber and vegetables was significantly lower in

Precontemplation, Contemplation, and Preparation stages than that in Action and

 

AA

23

Maintenance stages (Glanz et al., 1994). In a national 5-A-Day campaign, Laforge and
colleagues studied 405 adult respondents to investigate psychosocial factors related to fruit
and vegetable consumption using the SCT. Results show that education was positively
related to fruit and vegetable intake and inversely related to being in the Precontemplation
stage. Males were twice as likely as females to be in the Precontemplation stage and eat
fewer than 2 servings of fruits and vegetables a day. Such findings imply that stage of
readiness to change should be considered as well as other factors in planning interventions
for increasing fruit and vegetable consumption. Targeting precontemplators with
information about the health benefits of consuming more servings of fruits and vegetables
daily, will likely be more effective than giving these people ways to meet the 5-A-Day
objective (Laforge et al., 1994).

 

24

 

 

 

 

USSR was; 58338 3858

 

 

 

 

 

 

0:3. 8. m... 88 28 838.8 88 «828. 8882 8. 8:98: :88 8.8 =8: 838.8 8 n 8888388: 88
8:88 > 58 m 8283 man :88 88.8 =8: 8.8 5 m 3m§m=m=8
8:88 0: am 858 8888 $88.
8 m 8.8 38.8 888838 moo. 8:88? ma... m3 238: a 888 8.8528
38.88:. 8:88 m :8 88.38 8888: is .88 8. :8: 28
88888 a foam 858 £8 5038.
828.888 5 m 3283.86: . 88cm 9. 8882 a: 8:98: £8 33583:. 880888
88. 88288 2.22 a :85. £8 88% 0: 888m :03 8=
888: 5 E8288: Q8
8888.
0:28 a, m... 88 8.2: 838888 a 5838 8.83 88$ 8:: 8. 98.88 5 883858838 «88 8.. 8=. 8. ma
_ 58.. 88:8 22%. 08:8 0: =8: 8: A .80 :8: 8.8 8 8 0.8:. 888. ea 88: 8:88.
5.8288 .88 a 08:8 :8 «8:382: 8,808.8 5.8 8».
8.2 82.8w 0: 3:8 28 38:. 28 3:8 28 <8¢8¢8 58:8 5 m 838
@288“ BE.
8888 a 8... 8m 8:: 88.888 2 8 =28 mi 58828 832398: 8.8 $8 3.8 mm 28:. mm 88.8 8 8 8 =8
82 828386= 8.888 2.23. 888383238: .4988 m8 8, 8818: M
8 oa<a¢=8_ 2.. 8288 m 8*

8« 8:88 «5.88 7 G <88 0. 8:883 s88
:88 :8: 8 ea N o: 85: 825cm 9 :8 ma

 

 

6888.8 83‘. .
98.3 a m... 2m 853383 m39o<m8 8 88:3 8, 8:98: 9888 828: 8.88 5 =8 888283288 28
88 83:88 8383888: «88 .m 5 =8: 8888 89:

288.8 =8: 83.28:. .8. 5 =8: 88...: 8852.
8:583 888: o: 8828 8:88 <88
888888 :5. 8:8 8:8 28 8585.8 =8

888.. 28B?

25

Remarks about the literature reviewed

Nutrition interventions implemented using psychosocial theories can intensify the
credibility of program design, and improve effectiveness of interventions (Sigman-Grant,
1996). Instead of relying on traditional dichotomous outcomes of success or failure to
change diet, the alternative measures of success for nutrition interventions offered by the
SCT such as movement from precontemplation to contemplation stage are considered as
needed elements ofien missing from other theoretical models of health behavior (Spomy et
al., 1995). Additionally, the adoption of stage-appropriate interventions has the potential
to diminish the demoralization and frustration frequently expressed by nutrition and health
professionals as well as program participants (Shelton, 1995). Given the potential values
of using the SCT in designing nutrition intervention, further research, including an
exploration of dietary habits, processes of change and length of maintained behavior
associated with each stage, is clearly warranted (Curry et al., 1992). The finding that
dietary intake of fat, fiber, fruits and vegetables parallels stages of readiness to consume
these dietary components and foods suggests that SCT might be useful in predicting people
at risk for low intake of dietary fiber.

 

26

The validity and reliability of dietary assessment

Reliabiligy and val idigy, the concegts

A new test or measurement is generally accepted by the scientific community when
its reliability and validity have been established, otherwise results from the new
measurement are open to criticism (Patrick et at., 1991; Guyatt et al., 1987). Reliability is
often thought of as the extent to which concepts and their measures are defined precisely
and replicably, that is, the ability of the measurements to produce the same result
repeatedly. The definition of validity can be described as the accuracy of measurements,
or the extent of the closeness for the new tool to measure what it is supposed to measure
(Patrick et al., 1991; Block et al., 1989).

The assessment of reliability and validity of the measurement is crucial in the
evaluation of health promotion programs, although the process is regarded elusive by many
program designers, possibly because they are not familiar with the statistical methods
involved (Patrick et al., 1991; Patrick et al., 1979). Commonly used methods for reliability
analysis including: split-half reliability test; test-retest reliability test; and the calculation
of Cronbach's alpha (Patrick et al., 1991). For validity analysis, there are three types of
tests which are particularly important in the design of questionnaires for health promotion
including: content validity; criterion validity; and construct validity (Carmines et al., 1979;
Baranowski et al., 1991). Examples of how these analyses apply in development of new

instruments are given later in this Section under “Validation of new short instruments in

dietary assessment”.

 

Diet has an important impact not only on health but also on daily functioning,
cognitive performance, and probably psychological well-being. Data collected from
dietary assessment therefore can serve as an early indicator of individual nutrition risk as

well as be useful for implementing desirable dietary changes (Brownell et al., 1995).

,

27

There are four methods traditionally used by researchers to conduct dietary
assessment: 24-hour recall, multiple days food record, lengthy food frequency
questionnaire (FFQ) or checklist, and diet history (Block, 1988; Lee et al., 1993). Each
technique has particular strengths and limitations, but all place considerable demands on
participants for interview time, careful completion of lengthy questionnaire, and detailed
record keeping (Glasgow et al., 1996). As demands have emerged for health promotion
and public health researchers to conduct research and intervention in less constrained
locations and in original field settings—such as worksites or homes, the traditionally time-,
cost-, and labor-intensive dietary assessments such as 24-hour recalls and food records
have become less practical (Kristal et al., 1990; Glasgow et al., 1996).

Kristal and colleagues have argued that assessing patterns of dietary behavior instead
of, or in addition to, traditional measures of nutrient intake, may be useful for designing
and measuring the effects of community-level dietary interventions (Kristal et al., 1990).
Based on this premise, several studies have focused on the development and validation of
short food checklists as new dietary assessment techniques using behavioral approaches,
such as the psychosocial factors that influence specific food and nutrient intake (Glanz et
al., 1993; Kristal etal., 1990; Kristal et al., 1990; Kristal etal., 1990).

The F FQ, or increasingly a short checklist, has become the primary tool of dietary
assessment in epidemiological research (Willett, 1990; Block 1989; Block et al., 1989).
Additionally, the short F FQ or checklist is considered as an appropriate tool to collect
dietary information from all survey participants in large prospective cohort studies under
constraints of time and limited resources (Willett et al., 1995; Boeing et al., 1989).

The two most widely used FFQ seen in nutrition research are those developed by
Willett et al.(Willett et al., 1985) and Block et al. (Block et al., 1986). Both consist of lists
of about 100 or fewer individual foods or food groups that are important contributors to the
population’s intake of energy and selected nutrients with space for subjects to indicate the
frequency they usually consume of the foods (Lee et al., 1993). The Willett-type FFQ was
developed through stepwise regression analysis of data obtained by questionnaires from
nearly 100,000 women (Willett et al., 1985). Willett selected foods based on their ability

to distinguish between women with high, moderate or low consumption of fat and fiber,

28

and other nutrients selected for their suspected role in cancer. The Willett FFQ was
cross-validated in a small sample (n=27) comparing data from the FFQ to 1-year diet
records over the same period (Willett et al., 1987). The Willett-type FFQ specifies one
“standard” portion size for each food and does not provide subjects’ choice of portion size,
because Willett believed that frequency of food consumption, rather than difference in
portion size, could potentially influence the results of sorting individuals into groups of
high or low intake of nutrients (Willett, 1990). See Figure 4 for an example used in a

study conducted on 27 men and women living in Beltsville, MD, area (Willett et al., 1987).

 

I AveraggUse Last Year
Never 1-3 per 1 per 2—4 5-6 1per 2-3 4-5 6+ per
rless month week per per day per per day

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Foods and Amounts hnir; week week day day
per
month
airy Skim or low fat milk (8 oz.
oods Flass)
Whole milk (8 02. glass)
[Sour cream (1 D

 

 

Figure 4 A section of Willett-type food frequency questionnaire (Willett et al., 1987)

The Block-type FFQ allows subjects to indicate whether their usual portion sizes are
small, medium, or large with respect to a stated medium or “standard” portion for certain
age/sex groups. The food items in Block-type FFQ were selected on the basis of their
contribution to total population intake of energy and each of 17 nutrients in the NHANES
11 data, and represent over 90% of each of those nutrients. Block believed that difference
in portion size had significant influence on total intake of nutrients and need to be specified
in dietary assessments. Portion sizes on Block-type FFQ were derived from median
amounts reported in 24-hour dietary recalls obtained form nearly 12,000 adults during
NHANES 11 in 1976-1980 (Block et al., 1986). See Figure 5 for a section of l3-item

29

Block-type questionnaire developed by researchers at the National Cancer Institute to

identify groups whose mean percent fat intake is high or low (Block et al., 1989).

 

Your serving
Medium size

serving 8 M L Day Week Month Year Neve

How often ?

 

Foods

 

 

flamburger, meatloaf 1 medium

 

 

[Beef steaks, roasts 4 oz.
Eork, including chops, 2 chops/4 02.
est

 

 

 

 

 

 

 

 

 

 

 

 

Wigs 2 dogs
am, lunch meats 2 slices
Figure 5 A section of 13-item Block-type screening questionnaire for fat intake

(Block et al., 1989)

 

Val iciation oinew. short instryments in dietary assessment

Patrick and Beery suggested that the emphasis with development of new dietary
assessment tools for health promotion programs should be on the tools’ validity, rather
than reliability. Validity is thought to be more important than reliability because most
assessments and health promotion campaigns are cross-sectional and investigators are not
likely to administer the same instrument to the same subjects (Patrick et al., 1991). Also
respondents’ desire to provide information he/she feels the questioner wants to obtain,
instead of what the respondent really eats, can substantially bias the validity of testing
instruments (Patrick et al., 1991). These reasons explain why the importance of validity of
new dietary assessment instruments for health promotion outweighs the need for
reliability.

Some recent studies have documented how to establish validity and, when possible,
the reliability, of new instruments for dietary assessment. A summary of these studies is

shown in Table 6.

30

 

 

 

 

288.388 9.28... 83.3 33338... 5.838: 33.88 .83...»
5.38. 3 m... 8 38.88 5.8388: 8.83 88 88.3.3 88.3.3. =8 888.8: 82.8: 83.8 =8 3.3 3 8:88.83. 5m 98

 

 

 

.08. 53¢: 3...: 8 (.8 8:: 8.9.8:. 33888 3 33338. 83. .so 83. 88 .38.. 8: m «8.8 88.8
8:8 3 8. 38.8... 8882 8888 88.8 .9... 883... 88 08 38.38 25 8 88:88 3 88:8 .33
8. .8. =8 38.8. 88.383. 3 38.8 o. 8..
888... 8. :8 8.83.... 88. .3538. 88.389. 3
83.6 33335:. .28 can
.338. 3 m... 8 3.8-88 8.8.8.358 m 38:. 888.... 9.8.88.8. =8 8.38.8: 8.58 ._8 83.3. 88.38:... 8.58 8..-
.m8. 3 $838... 18.... 53...... :8 8... 8.. 33335:. 88 38: o. .5 3. 83.3338 38..
_ :8. 3835833888 8.. 88 88.8 883838.88 8.. .so ~.88..
588:. 8 883.. 88. 8. . =8 8.88.8: 8.18: 83.3. 88 88.8
888.8 8.. 8.8:. 8.88 33338. 8.. 38: o. 38 .58 away... :8 8: 88.
.83 8.. m3. 888:. .89. m8..-.<8 25 m3. .5 f8: 8.. .x. o. 8.88 .33 8..
. 88 888“. 888.8 8.. m3. 888...
.82.
.938. 3 m... 8 38-88 .538 38 m .68 8.8.8: 98.3.3 ammo. =8 8.838. =88 33.38. .. .83.. 9.. o. .o :8
.8... 38 12.0 353.88 838.33. 8 53:: . 8.28: $0 83. ~38... 88.. 8880.8 8.. m 3 =8 83
€8.88 38... =8 88.... 8: 88 39.8 :8 8:858
8:338 933. =8 39.88
83
088 3 8... 8» 830.8% 3 m 88:333. $.83 =8 888.88 82.8: 83.8 a 8:858 .85 38.38:. .
.8“. 28.3... $8.. 338. 9.8.8338 38338 33335:. 8.... 08 NP 0:5 mm. 8.88.88 5...: 88 .83
85:888. .888 3.3838 .83 388.50 15 .358 8:38 .83 Rob. 8
.3- 88 38.8.38 8882 =8 3838. 88.389. 3 8.888. no.8
88.8: 98388 3.3: 8:32.38 a 8:333: .85 3838.
88.382 388. =8:
88.8 98

 

 

 

31
In a research funded by the National Cancer Institute (NCI), Kristal and colleagues

developed an l8-item food checklist assessing dietary pattern of fat intake on 99 mid-aged
women (Kristal et al., 1990). Researchers tested the criterion validity of the new checklist
by comparing results from the checklist with those from two 4-day diet records and one
modified Block F FQ, which served as criterion measurements. For the estimation of
percentage of calories from fat, the correlation between the test instrument and criterion
measurements was 0.68 (p<.001). The new instrument also had adequate test-retest and
internal consistency reliabilities ranging from r=0.67-0.90 and r=0.54-0.76 for percentage
of calories from fat and for its five scales of dietary behavior related to fat. The five scales
were: I) avoid fat as seasoning; 2) avoid meat; 3) modify high-fat food; 4) substitute high-
fat foods with specially manufactured lower-fat foods; 5) replace high-fat foods with low-
fat alternatives.

Similar validation procedures were followed with another short checklist administered
to 97 disclaimers of the Women’s Health Trial to assess their intake of total fat, fiber, and
saturated fat (Kristal et al., 1990). These women were disclaimed from the randomized
clinical trial of low fat diet and breast cancer, because the screening showed they already
consumed low fat diets. The researchers found the correlations between the test
questionnaire and criterion measurement, which was a mean of two 4.day food records,
were r=0.52, 0.40, and 0.61 for total fat, dietary fiber, and saturated fat, respectively. These
correlations were similar to those observed between a modified Block FFQ administered
concurrently and 4-day food records (r=0.57, 0.48, 0.63, respectively), suggesting the
development of short dietary questionnaires with acceptable validity, as assessed by
moderate correlations, was a feasible approach.

In the development of the Food Behavior Checklist (FBC) , a simplification of a 24-
hour diet recall consisting of 19 simple yes/no questions about foods consumed during the
previous day (Kristal et al., 1990), researchers established validity by comparing the
responses of 96 women on the F BC to information collected during a professionally

administered 24-hour diet recall. For the agreement between the checklist and 24-hour

32
recall, ll items out of 19 had kappa values greater than 0.80 and five of the rest had kappa

values greater than 0.60. The authors concluded that data collected by F BC were similar to
those obtained from a 24-hour recall, and the FBC could be an inexpensive and technically
simple dietary assessment tool for community-level use. The authors assumed that the 24-
hour recall was appropriate for criterion validity, however, the validity of 24-hour recall for
typical food intake has been highly questioned by other researchers (Block, 1982).

A questionnaire developed by Glanz and colleagues used 24 Likert-scale items for 12
constructs to measure psychosocial factors influencing fat- and fiber— related dietary
behavior on 652 participants of the Working Well project - a National Cancer Institute
funded, multicenter controlled trial of worksite health promotion interventions (Glanz et al.,
1993). Researchers tried to establish criterion validity of their test questionnaire by
correlating the responses with the information collected from a 22-item and a 88-item
Block FFQ and then examining the partial correlations between psychosocial items and fat
and fiber intake. Among 12 constructs, the constructs of self-rated diet, past success at
change, and motivation to eat low-fat foods were found to have significant correlations with
fat and fiber data from the FFQs ranging from r=0.31 to r=0.35 (p<0.000l ). The constructs
of intentions to change, self-efficacy for change, and success at changing had internal
consistency greater than 0.50, which is desirable for scale construction in'behavioral
research (Glanz et al., 1993 ). Thus, a new dietary questionnaire with the questions designed
from those three constructs might have similar correlations with subjects’ actual nutrient

intake.

Remarks about the literatyre reviewed

An understanding of the factors that influence people to change their diets in a positive
direction may provide us with some insights useful for the design of effective nutrition
education and counseling programs (Contento et al., 1990). As the demand increases for

simple, rapid, and accurate dietary assessments in low-intensity community-based

33
intervention trials, measuring dietary behavior seems like a promising approach to meet

these needs (Kristal et al., 1990). Nevertheless, further research efforts in behavioral
science and psychology are indispensable to assure the validity and reliability of developing
instruments. Several short checklists for fat intake have been developed and validated
primarily by correlation coefficients exceeding >050. Food checklists do not exist
specifically for fiber. Also, one might question whether a correlation coefficient of0.50 is
adequate to establish validity, or whether it is even the appropriate technique for validation
of such tools. Perhaps the sensitivity, or ability ofa tool to detect who is at dietary risk
would be more meaningful for establishing the validity of a rapid community-level

screening tool.

34
Insight of the research from the revlewed literature

Knowing that increased dietary fiber intake helps reduce the risks of several chronic
diseases and peoples’ average daily intake is still far below the recommended amounts,
more efforts on promoting the public awareness and recognition of benefits of fiber seem
necessary. Nutrition interventions combining the information of what changes to make and
how to make changes might be useful to achieve this goal. The thesis research serves as a
prologue for the development of such intervention.

Validating a short checklist of foods for fiber intake examines the feasibility ofa rapid
method to collect dietary intake data, or a food component of interest. Developing a SCT
flowchart provides a simple way to stratify subjects into different stages of readiness for
different intervention strategies. Composing a questionnaire of attitudes and beliefs toward
dietary fiber intake reveals the characteristics of subjects at difierent stages, which might be
used as intervention points in program design as well as to validate the SCT paradigm for
dietary change. The aggregation of such instruments as these three might be an integral
piece for screening subjects at risk for low intake of dietary fiber before the onset of core

intervention, and providing baseline information for evaluating the efficacy of intervention.

35

Methods

Subjects

Subjects were the 99 respondents from a random sample of 401 adults selected from
the members of Clerical Technical Worker’s Union (CTU) of Michigan State University
(MSU), currently 1,891 members. The average age for this population was 42 years old and
nearly 92% of them are female. The name list and mailing labels for subjects were obtained
for $25.00 from MSU Human Resource Information Data Service in September, 1996.

The reasons follow for selecting this target population. 1) The majority of CTU’s are
mid-aged females who tend to be the main food providers in family, thus their preference of
foods and diets can influence other household members, especially young dependents. 2)
CTU’s are literate and have experienced questionnaire administration, allowing them to
answer the test instruments without detailed interpretation by researchers. 3) An intramural
health promotion program, MSU Healthy U, provides nutrition and health information
routinely to its participants which are mainly female staff of MSU. We expected the
research to be mutually beneficial to CTUs (increased awareness of dietary fiber) and
Healthy U program (provide information for future program design and screening those at

risk for low intake).

Procedures

Data collection began after permission from the University Committee on Research
Involving Human Subjects (UCRIHS) (Appendix A) and the MSU Healthy U Committee.
A pilot test was conducted on the secretaries (n=10) of Food Science and Human Nutrition
(FSHN) Department after the permission from chairperson of FSHN was obtained.
Preliminary support from the MSU Clerical Technical Worker’s Union (CTU) was sought,
and at first granted, by the President of the CTU. Unfortunately afier all questionnaires

36
were developed and prepared, the CTU President found, after extensive discussions with

her advisory board, that the CTU could not officially support the survey with a letter of
recommendation. This was because the Union was initiating formal health-benefit contract
negotiations with the university. Eventually, the test instruments were mailed without the
support letter from the CTU President. A brief explanatory letter was sent to all 1,891 CTU
members regarding the purpose of the research before the sample selection occurred
(Appendix B).

The sample of 401 people received the consent form. (Appendix C), the test
instruments, and an incentive. The test instruments, which include one food checklist, one
Stages of Change (SC) questionnaire, a one attitudes and beliefs (AB) questionnaire and
food record forms for one weekday and one weekend day, were mailed to selected subjects
in October, 1996. As an incentive, one dollar coupon to the campus Dairy Store and an offer
of free dietary analysis was provided. Interested subjects provided their age, height, weight,
and exercise pattern for a free dietary assessment. The test instruments were self-
administered with brief directions. It took about 10 minutes to complete the food checklist
and two questionnaires. More time, about 30 minutes, was needed for recording 2-day’s
food intake. Due to the poor response rate (16%) for this first mailing, a second mailing of
test instruments was mailed again with a cover memo (Appendix D) to subjects that did not
return. The final response rate reached 25%. All survey responses were collected by
researchers and kept in a file cabinet to ensure confidentiality of subjects.

The fiber content of foods on the checklist was extracted from the MSU NutriGuide
2.0 program (Song W.O., Nutritional Analysis Computer Program, Michigan State
University, 1990) and entered into an [BM-compatible computer using Microsofi Excel 7.0
program (Microsoft Inc., Seattle, WA, 1995). Results served as the database to calculate
subjects’ fiber intake assessed by the checklist.

To judge the validity of the test checklist, results were compared to the fiber intake
computed from the concurrently-collected 2-day food records analyzed by using MSU
NutriGuide 2.0. MSU NutriGuide analyzes diets for 27 nutrients in 953 different foods.

37
Ninety-nine percent of the foods in the software are complete for fiber composition. Its

nutrient composition data originally came from Michigan State University main-frame
database and USDA Handbook-8 (USDA, 1990). However, sources like food
manufacturer’s information, McCance and Widdowson (Holland et al., 1991), and
Pennington (Pennington, 1989) were also included in the database.

Before the primary data collection on CTUs, both the checklist and the questionnaires
were pilot-tested on all secretaries in the Food Science and Human Nutrition Department
(n=10). The draft test instruments were modified slightly based on response to the pilot test.
In the checklist, not-selected food items were omitted, such as cowpeas. Standard portion
size was added to each food for better accuracy in estimating actual consumption. In the
Stages of Change questionnaire, a slight change in wording was made to include people
already on a high-fiber diet in the maintenance stage. For the attitudes and beliefs
questionnaire, one question was eliminated because of similar wording to another question.
Two open-ended questions were added to acquire subjects’ detailed description of
motivation and barriers for increasing dietary fiber intake. The summarized responses will

be made available to Healthy U for improving future program design.

Instruments

One checklist of food sources of fiber and two questionnaires related to changing
dietary behavior related to fiber were developed for administration by mail. It was
hypothesized that these three short instruments, taking less than 10 minutes to complete
would be able to identify people at risk for low intake of fiber as well as the more lengthy
2-day food records.

38

(fhecklist

The purpose of developing the checklist was to establish a valid tool for ranking people
by their dietary fiber intake. Establishing criterion validity of this checklist was one of this
study’s goals.

The 2-day food records from Healthy U participants in three previous projects were
used to develop a list of food items contributing to 95% of the fiber intake in participants’
diets. The computer program written in Turbo Pascal (Borland International Inc., Scott
Valley, CA, 1987) language developed by Huang and colleagues (Huang et al., 1994) was
used to accomplish the sorting job. The “top 40” foods contributing fiber were assumed to
reflect the fiber consumption of our target population. The amount of dietary fiber intake
was determined by the following formula:

Fiber (g/day) = Quantity of serving X Frequency (daily) X Fiber content of the food (g)

Further selection of food items for the checklist were from two sources: 1) the food
sources of dietary fiber in diets of 1,032 women 19 through 50 years old in 1985 Continuing
Survey of Food Intakes of Individuals (CSFII) data, which lists the food items contributing
to 70% of dietary fiber resources (Krebs-Smith et al., 1992); 2) the dietary checklist formed
by Kristal and colleagues, because it has fair criterion validity (F 0.57) with two 4.day food
records and one Block-type FF Q and fair test-retest reliability (weighted Kappa= 0.48)
(Kristal et al., 1990; Laforge et al., 1994). See Table 7 for sources of foods in the checklist
developed for this study.

The final edition of the checklist used in this study (Appendix E) contained 47 foods
divided into 7 groups: Fruits, Vegetables, Beans/Legumes, Breads/Grains/Pasta, Nuts,
Snack foods, and Cereals. A few foods were kept on the checklist which did not come from
the sources in Table 7, because these foods might be “indicator” foods for people who
consume high fiber diets. A blank for one food item in each group was also included to
allow subjects to add foods not on the checklist. A standard portion size, closest to usually

consumed amounts, was selected for use on the checklist to keep the format concise. -The

39

small, medium, and large serving sizes in the Block-type FFQ were not included, because

studies suggested the variations in food intake are mainly determined by frequency of

consumption rather than by portion sizes (Samet et al., 1984). Also, subjects tend to pay

very little attention to the serving size information on FFQ without the presence of

instructor (Smith et al., 1991; Subar et al., 1995).

 

 

 

Table 7

The sources and foods selected for the Checklist

Diets of Healthy U CSFII data, 1985 Dietary checklist by Fiber-rich,
participants by Huang et al., Kristal et al.,1990 ‘lndicator'
1994 foods
apple apple baked beans berries
bagel broccoli broccoli celery
banana corn brown rice granola
bran flakes crackers Brussels sprouts grape
broccoli French fries cauliflower nectarine
carrots green beans kidney beans peach
French fries lettuce lima beans popcorn
green beans lima beans melon spinach
green peas noodles orange

noodles pasta/spaghetti peanut butter

oat bran peanut butter peanuts

oatmeal peanuts potato chips

orange potato pmnes

pasta/spaghetti potato chips raisins

peanut butter ready-to-eat cereals ready-to—eat cereals

pear tomato walnuts

pizza tortillas whole-wheat bread

potato white bread zucchini

potato chips whole-wheat bread

prunes

raisin bran

raisins

tomato

tortillas

tossed salad

white bread

whole-wheat bread

 

 

 

 

 

 

 

40

Stages 0t Change Questionnaire

One Stages of Change (SC) questionnaire was designed to categorize subjects into one
distinct stage of readiness to eat fiber-rich foods, based on the work of Curry and
colleagues. Curry and colleagues developed an algorithm and set of questions to determine
at which stage individuals were for dietary fat reduction (Curry et al., 1992)(Table 8).
Their results suggested that individuals could be clearly classified into one of five stages of
dietary fat reduction, which were inversely associated with self-reported intake of fat (-0.93
for men, 073 for women in Sample B, p<0.001). Therefore, their algorithm was modified
to readiness to consume adequate fiber, for the purpose of stratifying subjects in this study.
The subjects’ stage of change relative to fiber intake was determined by their responses
according to the algorithm. Although there was no a priori reason for a particular time span
for a particular dietary behavior (Spomy et al., 1995), and no study has addressed stage of
change for fiber intake (Glanz et al., 1994), we chose six months to be the time span for
maintaining high fiber diet. Six months was selected based on the remarks by Prochaska
and colleagues, “Being able to consistently engage in a new incompatible behavior for more
than six months are the criteria for considering someone to be in the maintenance stage”
(Prochaska et al., 1992).

The SC questionnaire was modified from multiple-choice questions into a flowchart
(Appendix F) for several reasons. 1) The flowchart appeared to be easier and quicker for
subjects to answer than a multiple choice format. 2) The format is shorter and clearer, thus
subjects’ interest of answering should be enhanced. 3) The original questions and algorithm
by Curry et al. left some people between stages; the flowchart format avoids this problem
entirely. By using the flowchart-style Stages of Change questionnaire adapted from Curry’s
algorithm (Curry et al., 1992), each subject can be categorized into one stage of change

related to his/her attitude and consumption of dietary fiber.

41

Table 8

Swing questions and algorithm for dietary fat reduction used by Curry et al.,1992
Questions

1. Have you ever changed your eating habits to decrease the amount of fat in your diet?
Yes 1; No 2 (Skip to #2)
1A. If Yes, are you currently limiting the amount of fat in your diet?
Yes 1; No 2 (Skip to #2)
181. If Yes, how long have you been limiting the amount of fat in your diet?
Less than 30 days 1; 1-6 months 2; 7-12 months 3; Over 1 year 4
182. If Yes, would you say your are now eating a low-fat diet?
Yes 1; No 2
2. In the past month, have you though about changes you could make to decrease the amount of fat in
your diet?
Yes 1; No 2
2A. How confident are you that you will make some of these change during the next month“?

Very confident 1; Somewhat confident 2; Mildly confident 3; Not at all confident 4

 

 

Algorithm
Stage Question(s) Answer(s)
Precontemplation l or 1A No

2 No
Contemplation 1 or lA No

2 Yes

2A Mildly or not at all confident
Decision/Preparation l or 1A No

2 Yes

2A Somewhat or very confident
Action I and 1A Yes

18 6 months or less
Maintenance 1 and 1A Yes

13 7 months or more

 

42
Attfiydes and Beliefs Questionnaire

The Attitudes and Beliefs (AB) questionnaire was designed for assessing subjects’
attitudes and beliefs related to dietary fiber intake. Carruth and Anderson proposed that
scaling criteria for developing and evaluating an attitude instrument should include
determination of content validity, test-retest reliability (stability), and internal consistency
reliability (unidimensionality) (Carruth et al., 1977). In this cross-sectional research, only
content validity and internal consistency reliability were examined.

Glanz and colleagues defined three domains of psychosocial factors as sensitive
indicators of dietary behavior improvement which are : predisposing, enabling, and
change-related factors (Glanz et al., 1993). A 12-construct, 24-item questionnaire based on
these three domains was developed and tested on the 652 participants of worksite health
promotion interventiOn. Results showed that self-rated diet, past success at change, and
motivation to eat low-fat foods were factors most strongly associated with dietary intake.
Adopting these three constructs and building on the work of Glanz and colleagues, we
designed our questions to assess subjects’ dietary attitudes and beliefs related to fiber
consumption. .

Other researchers have used the Health Belief Model (Rosenstock, 1974) for
constructs related to attitudes and beliefs important for dietary behaviors. Contento and
Murphy found dietary self-changers were differentiated from non-changers primarily on the
basis of perceived susceptibility to diet-related diseases, perceived benefits of changing
their diets, normative beliefs, overall health concern, cues to action, chance locus of
control, and self-efficacy (Contento et al., 1990). Likewise, Schwarzer proposed that self-
efficacy was a major determinant of behavioral intention in both the decision-making phase
and the action phase (Schwarzer, 1992). Therefore, questions related to these factors were
added to the AB questionnaire in order to increase its content validity and to relate
responses to appropriate Stages of Change.

Several questions on the AB questionnaire were adapted for fiber from the original

questionnaire developed by Glanz and colleagues (Glanz et al., 1993). Most questions were

43
designed according to appropriate constructs of the HBM (Rosenstock, l974) to reflect

stages of readiness to consume foods high in fiber. At first the questions were grouped into
four constructs for the pilot study: I) predisposing factors; 2) enabling factors; 3) change-
related factors; and 4) self-efficacy, but later the titles of these constructs were removed for
concrseness. Expert reviewers determined the soundness of the questions for assessrng the
fiber-related attitudes and beliefs to assure the content validity. For conciseness, items With
less face validity and complex wording were eliminated. The final edition of the AB
questionnaire (Appendix G) contained 13 questions, each question With a numeric Likert

response scale answers ranging from 1 = strongly disagree to 5 = strongly agree.

Statistical analyses

The collected data were analyzed with the use of MSU NutriGuide 2.0 (Song W.O.,
Nutritional Analysis Computer Program, Michigan State University, 1990), SPSS 6.1 for
Windows (SPSS Inc, Chicago, IL, 1995) and Excel 7.0 (Microsoft lnc., Seattle, WA, 1995).
The descriptive statistics of subjects’ demographic characteristics including age, sex,
height, weight, and BM were analyzed. From the checklist, each subject’s gram amount of
fiber intake was determined from the fiber content of each food in a standard portion
multiplied the frequency of consumption, then summed for all foods selected. From the 2-
day food record, the average daily intake of fiber was the sum of 5/7’s of the weekday intake
plus 2/7’5 of the weekend intake. Different weights were applied to reflect subjects’
different eating pattern during the weekday and weekend. The size, mean, and standard
deviation of fiber intake assessed from checklists and weighted 2-day food records of each
subgroup were calculated.

Validity of the checklist was examined by three methods: 1) Pearson’s correlations
of gram amount of fiber intake between checklist and weighted 2-day food records; 2)

paired t-tests conducted on the gram amount of fiber intake assessed by the checklist and

44
weighted Z-day food records; and 3) the extent of sensitivity of the checklist to predict those

at risk for low intakes of fiber.

Low intake of fiber was defined two ways: 1) as _<_ 10 g/day based on findings by
Smallwood and Blaylock using data from CSFII (Smallwood et al., 1994) where people not
aware of health problems consumed less than this amount; and 2) as less than adequate or
20 g/day. High intake of fiber was 2 20 g/day because NCI recommends 20-30 g fiber
intake per day. The definition of sensitivity in this study is the percentage of subjects whose
actual fiber intake is defined as low (_<_ 20 g/day) confirmed by checklist (Greenberg et al.,
1993).

Subjects’ responses to SC questionnaire were coded 1 through 5, corresponding to
precontemplation through maintenance, for statistical analysis. In each stage, the
distribution of subjects and average daily intake of fiber assessed from the checklist and
weighted 2-day food record were calculated. ANOVA (Duncan’s multiple range test) was
used to test the significant difference of fiber intake among these subgroups. The
distribution of subjects in each stage combining with the results of exploratory factor
analysis from AB questionnaire were used to explain the characteristics of attitudes and
beliefs related to fiber intake.

The mode and mean (i SD) of each question on AB questionnaire for each stage was
determined to reveal particular characteristics of attitudes and beliefs related to fiber
consumption. Exploratory factor analysis was performed to determine how many
constructs really existed in the AB questionnaire and then appropriate construct titles were
given. The internal consistency of each construct was examined and the specific attitudes
and beliefs at each stage of change were determined by subject’s response to each construct.
Responses to AB questionnaire were cross-tabulated by distribution of Stages of Change to
see whether responses made sense conceptually according to stage of readiness to consume
fiber-rich foods.

In additional analyses, the rank of foods from the checklist was calculated based on

the contribution to the weekly number of portions and grams of fiber. The qualitative

 

45
responses to the barriers and suggestions to increase fiber intake were tabulated. Results

and implications of this research will be provided to the Healthy U Program to aid in future

evaluation and program design.

 

Schedule
Spring/95 *Started proposal writing
Summer/95 *Finished draft questionnaire and checklist
*Had 2nd committee meeting
*Acquired UCHRIS approval
Fall/95 *Examined 3-day food records from previous Healthy U
results
*Designed the format of checklist and questionnaires
Spring/96
*Finished preliminary data analysis of foods selection for
the checklist
*Finished questions wording for the AB questionnaire
*Had 3rd committee meeting
*Made the checklist and questionnaires ready
Summer/96
*Started literature review
*Completed pilot test on secretaries of FSHN Dept.
Fall/96
*Obtained the permission from MSU CTU
*Mailed out the test instruments and started data collection
from MSU CTUs.
*Had 4th committee meeting
*Started thesis and manuscript writing
Spring/97 *Submit manuscript to the Journal of American Dietetic
Association
*Finish all research works

 

 

 

 

46

Results

Of the 401 subjects contacted, 74 (18%) returned refusals to participate. In total, 99

 

(25%) subjects responded: 96 completed food frequency checklists; 67 completed 2—day
food records; and, most of the subjects were female (90%). The gender distribution was

similar between subjects who responded (n=99) and total subjects contacted (n=401).

 

There appeared to be a slightly higher percentage of men than women who refused to
participate. See Table 9 for the distribution of respondents by gender. The mean (i SD) age
for all respondents was 42.8: 17.1 years. See Table 10 for detailed demographic statistics.
The demographics for gender and age of the respondents was identical to those of the total
CTU pool (n=1891).

 

 

 

 

Table 9
Response rate and demggraphics of subjects
Male Female All
Total subjects contacted: 36 365 401
number of responders: 9 90 99
number of returned refusal: 10 64 74
number of non-respondents: 17 214 231
Subjects completing Checklists: 8 88 96
Subjects completing Zday Food Records: 7 60 67
Subjects completing both instruments 7 57 64

 

47

 

Table 10
Mean (: SD) of age, height, weight, and BMI of subjects gy gender

Male (n=7) Female (n=60) All (n=67)

 

Mean Age: (yr) 42.1 :11.8 41.3: 10.3 42.8: 17.1
Mean Height: (in) 71.3: 6.4 65.5:32 66.1 :40
Mean Weight: (lb) 184.1 :45.2 147.2: 27.6 151.0:316
Mean BMl: (kg/m2) 25.3:3.9 24.1 :37 24.2: 3.7

 

Figure 6 shows the scatterplot of fiber intake data from the two dietary assessment
tools. Figure 6a, 6c show the distribution of the original fiber intake data collected from
checklists and weighted 2-day food records, respectively. Distributions were skewed and
needed to be normalized for correlational analyses. Figures 6b, 6d show the logarithmic
distribution of the original data. All following statistical analyses were performed without
the subject who reported 95 g/day of fiber intake (Figure 6a), a physiologically impossible

intake, because this abnormal value would be a cause of significant bias.

 

 

 

 

8 50
(D
a. o
'O
8 40-
I..—
'C.’
N O
30‘ ° 0 .
E . ‘
g . . o . 00
. .
g 20‘ .0 .09. .0. : .
to: . ‘ o . . t 0.. o .
H— o o
g 10l o .0 :0 9‘. . t.
93
U) 0 t v v V
O 10 20 30 40 50

gram of fiber assessed from checklist
Figure 6: Distribution of daily fiber intake from checklist vs. weighted 2oay food records

48

 

30

 

 

 

 

 

 

 

Figure 6a. Original distribution of fiber intake (9) from checklist (n=96)

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6b. Logarithmic distribution of fiber intake (9) from checklist (n=96)

49

 

 

 

 

 

 

 

 

 

 

 

 

2%.:

 

 

 

1.5 .5 .5 .5 . ' 37'.5 42.5
10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0

 

Figure 6c. Original distribution of fiber intake (9) from weighted 2—day food records (n=67)

 

 

 

 

 

 

 

 

 

 

 

 

.88 ' 1.00 ' 1.13 ' 1.25 ' 1.3a ' 1.50 ' 1.63

Figure 6d. Logarithmic distribution of fiber intake (9) from weighted 2-day food records (n=67)

50
Although the fiber intake data appeared to be normally distributed after log

transformation, the log values were only used in the calculation of correlation coefficients
because it is not appropriate to describe if the subject was at the risk of low fiber intake or

not using log values.

Hypothesis 1. There will be positive correlation between the gram amounts of dietary fiber

intake assessed fiom the checklist andfrom the 2-day food records.

The daily fiber intake assessed from checklists was 16.9: 12.5 g/day, ranging from 3 to
an abnormally high intake of 95.5 g/day. After the subject with abnormal fiber intake was
excluded, the daily intake became 16.1 :9.6 g/day, ranging from 3 to 47.6 g/day. The daily
fiber intake assessed from weighted 2-day food records was 16.1 :7.3 g/day, ranging from
6.3 to 45.1 g/day.

Both original and logarithmic data showed significant correlation between the
checklist with the 1" day and the average of the weighted 2-day food records (Table 11),
suggesting that the eating pattern of subjects might have a larger fluctuation on weekend
days than that on weekdays. Although the fiber intake appeared higher on weekdays than
on the weekend, there was no significant difference (p=0.12) The range of Pearson’s
correlation coefficiences were lower (r=0.48 to 0.55, p<0.01) than those expected and
desired (r=0.60 to 0.80).

 

Table 11
Median and mean (: SD) of average daily fiber intakes form checklist and weighted 2-day food
record and log transformed correlation

 

 

 

 

Instrument N' Mediagg) Mean: SM Correlation (r)

Checklist 95 13.6 16.1 :96 —-

weighted 2-day food record: 66 Log
1" day (weekday) 15.0 17.0 : 8.5 545* 479*
2"" day (weekend) 14.0 15.4:8.6 .233 .198
averge 14.9 165:7.4 .519" .481‘

 

 

' number of paired t-test performed: 64
"' p<.01

 

5|

The range of “limits of agreement” , which indicated how the difference in grams of
fiber intake between two measurements distributes, was calculated by following formula
(Saba et al., 1995) : '

mean difference (0.33 9) : I. 96 X standard deviation of the differences (9.42)

The large range of “limits of agreement” (-18. 12, 18.80) also confirmed the moderate
correlation between these two measurements.

Based on these results, the Hypothesis 1 was weakly supported.

Hypothesis 2. There will be no significant diflerence between the gram amounts of dietary

fiber assessed from the checklist and the 2-day food records.

The paired t-value between the fiber intake from these two measurements was 0.29
(df=63, 95% Cl= -2.015 to 2.693). Because the 95% confidence interval includes 0 in this
range, no significant difference was detected in gram amounts of dietary fiber assessed from

both measurements. Thus, the Hypothesis 2 was supported.

Hypothesis 3. The sensitivity ofthe checklist to detect high and low intake ofdietaryfibcr

will be _. 70 when compared to 2—day/ood records.

The distribution of subgroups for high, medium, and low fiber intake is shown in

52
Table 12. The number of subjects being classified in the same subgroups by the two dietary

assessments is 32 (7+17+8), which divided by total valid subjects (n=66) is defined as the
percentage of agreement (0.48) between the checklist and weighted 2-day food record.

To calculate sensitivity, subjects in Table 12 with fiber intake < 20 g/day from both
dietary assessments were grouped (n=40) as well as the subjects with fiber intake < 20 g/day
from the weighted 2-day food record but 2 20 g/day from checklist (n=10). The sensitivity,
or the ability to detect subjects who actually had low fiber intake, of the checklist was 0.80,
is higher than the expected value (0.70). Therefore, the Hypothesis 3 was supported.

 

Table 12
The distribution of subjects (n=66) in high, medium, and low subgroups of fiber
intake assessed by checklist and wjghted 2-day food record

 

 

 

weighted 2-day food record
Low Medium High
Checklist (<10 g/d) (10 to <20 gfl) (_>_20 g/d)
Low (<10 g/d) 7 11 3
Medium (10 to <20 g/d) 5 17 5

High (320 gld) 1 9 8

 

Hypothesis 4. I here will be a significant difference between the gram amounts ofdietary

fiber intake for different stages ofchange related to dietary fiber.

To address this hypothesis, subjects were classified into stages of readiness to
consume adequate dietary fiber as assessed by the SC questionnaire. Figure 7 shows the
distribution of subjects by stage of change, with number of cases on the tips of the bars. The
largest group was people at the Maintenance stage, claimed by 35% of all subjects.

Eighteen subjects did not indicate their stages on the flowchart, a surprisingly large number.

 

53

 

4O

30‘

 

20‘

 

 

10'

      

    

 

 

 

 

 

 

 

 

Count
0

   

 

 

 

 

 

 

 

 

Missing Contemplation Action ’
Precontemplation Preparation Maintenance

Figure 7. Distribution of subjects in five stages of readiness to change fiber intake (n=99)
(number of subjects are shown at the tip of bars)

The results for mean (: SD) fiber intake at each stage assessed by both instruments are
shown in Table 13. Duncan’s multiple range t-test revealed that the fiber intake assessed by
the checklist at the Maintenance stage was significantly higher than that at the
Precontemplation and Action stages. No significant differences were found between stages
for the fiber intake assessed from weighted 2-day food records. Although people at the
Preparation stage appeared to have highest amount of fiber intake from both measurements,
the differences were not significant by either dietary assessment measure, perhaps due to

the small number in this stage.

54

 

 

 

 

 

Table 13

Fiber (9) assessed from the checklists and weighted 24iay food records at each stage of change
Stage Checklist“ wgghted 2-d Food Record‘I
Precontemplation 12.5:90 (17)” 14.9:96 (14)
Contemplation 14.7: 10.6 (11) 13.8: 3.9 (6)

Preparation 20.4:138 (6) 18.4:7.1 (5)

Action 13.1 :72 (13)* 15.8: 3.7 (7)
Maintenance 19.8:9.3 (34) 17.0: 7.9 (23)

All valid cases 165:9.9 (81) 16.1 :7.5 (55)

' Mean:SD (n)

b Duncan’s multiple range test, Precontemplation, Action vs Maintenance

* p<.05

Table 14 shows the Speannan’s rank order correlation between stage of change with

estimated grams of fiber intake from checklist and weighted 2-day food record. We found a

positive, significant correlation between the checklist and stage of change, not too

surprisingly because both are based on people’s self-perception of their diet and attitudes.

However, there was no correlation between stage of change and estimated fiber intake from

weighted food records which reported foods actually consumed for 2 days (a weekday plus

a weekend).

 

Table 14
Spearman’s rank order correlation between stage of change and
estimated grams fiber intake from two instruments

 

 

Instniment N Spearman's r
Checklist 81 367*
weighted Zday food record 55 .176

 

' p<.005

55

The Hypothesis 4 was supported in part for the checklist. The fiber intake at
Precontemplation stage is the lowest and was the highest at Maintenance stage. However, a
clear stepwise progression from low to high intake by stage of readiness to eat fiber-rich
foods was not observed, although there was a low but significant correlation for the

checklist. The low intake for the Action stage is puzzling.

Hypothesis 5. Subjects" responses to the attitudes and beliefs questionnaire will difler

according to their different stages of readiness for dietary fiber.

Table 15 shows subjects’ responses to the 13 questions on attitudes and beliefs related
to fiber intake and the original constructs. In general, subjects knew about the health
benefits of fiber (Q1 , Q2, Q9, Q13) and showed interest on fiber-rich foods (Q3, Q10, Q12).
Subjects’ responses were neutral to environmental factors (Q8, Q1 1) and childhood history
of high fiber intake (Q4). However, subjects reported overall a lack of social and family
support for eating fiber-rich foods (Q5, Q6, Q7).

56

 

Table 15

Mode“, mean, and standard deviation of subjects’ (n=95) responses to Attitudes and Beliefs
guestions

Mode Mean SD
Q1 Eating a lot of fiber-rich foods decreases my chances of

getting serious diseases such as cancer and heart disease. 4 4'0 0'94
Q2 What I eat is important for my health 5 4.6 0.74
Q3 Fiber-rich foods taste good 4 3.9 0.96
Q4 I ate a lot of fiber-rich foods when I was growing up. 3 3.1 1.08
QS I've gotten a lot of advice about how to eat fiber-rich foods. 1 2.9 1.18
06 My friends encourage me to eat fiber-rich foods 1 2.0 1.06
Q7 My family encourages me to eat fiber-rich foods. 1 2.3 1.18
Q8 It is easy to select fiber-rich foods when I eat out 3 2.8 1.09
Q9 Eating fiber-rich foods is important 5 4.4 0.85
Q10 I‘m willing to try a new food if it is fiber-rich 4 3.7 1.05
Q11 It is hard to increase my intake of fiber-rich foods 3 3.0 1.10
012 I can increase intake of fiber-rich foods in next 6 months 4 3.8 0.88
013 I can be healthy, if I eat fiber-rich foods 4 4.0 0.98

 

' Likert scale from 1=Strongly Disagree to 5=Strongly Agree

When exploratory factor analysis was performed, the five constructs emerging from
these questions were: Health benefits of fiber; Social and environmental support; Interest in
fiber-rich foods; Childhood intake; Difficulty of increasing intake. These five constructs
accounted for 73.1% of the total variance of responses. Table 16 shows the percentage of

variance for which each construct accounts.

57

 

 

 

Table 16

The percent—ang variance of responses to A8 questionnaire for which each construct accounts
Construct Pementagg Cumulative Percentage
Health benefits of fiber 28.5 28.5

Social and environmental support 17.2 45.6

Interest in fiber-rich foods 10.7 56.3

Childhood intake 9.0 65.4

Difficulty of increasLngintake 7.7 73.1

 

Questions were put together which had correlation coefficients 2 0.60 for a specific
construct. Table 17 shows the correlations between the 13 questions for the five constructs.
Confinnatory factor analyses were performed for the questions with correlations < 0.60 and
linked to more than one construct, e.g. Q7 and Q13, to decide to which constructs the
questions belong (Hunter, 1977).

As a result, the construct of “Health benefits of fiber” included Q1, Q2, Q3, Q9, and
the less dominant Q13. Construct of “Social and environmental support” included Q5, Q6,
Q7, Q8. Construct of “Interest in fiber-rich foods” included Q10 and Q12. Both the
constructs of “Childhood intake”and of “Difficult of increasing intake” had only one
question each which were Q4 and Q11, respectively. The a reliability of the three

constructs with more than one question ranged from 0.61 to 0.83 as shown in Table 17.

58

 

Table 17
Correlations ' of each question with five constructs and or of three constructs with more than one
gjiestion

Health benefits Social and Interest in fiber- Childhood Difficulty of

 

 

of fiber environmental rich foods intake increasing
support intake
Q1 .80
02 .86
Q3 .66
Q9 .84
013 .42 .52
05 .78
Q6 .81
Q7 .64 .42
08 .68
010 .72
Q12 .83
Q4 .93
01 1 .89
a .83 .73 .61 —- --

 

' only correlations > .40 are shown

Table 18 shows the responses to the 13 questions within the five new constructs by
subjects at different stages of readiness to consume fiber-rich foods. For the construct of
“Health benefits of fiber”, subjects across all five stages thought fiber in the diet was
important to health. People at Preparation stage, however, did not seem to like the taste of
fiber-rich foods (Q3). This finding was interesting given that subjects at the Preparation
stage reported high intakes of fiber. These people were probably at the transition of
adopting new foods and perhaps experiencing difficulty in the interim.

Responses to the construct of “Social and environmental support” was low for all
stages. People at Preparation and Maintenance stages appeared to be getting the most

advice. The highest scores overall for “Social and Environmental support” were at the

59
Maintenance stage, as expected in SCT. Unfortunately average support from friends (Q6)

and family (Q7) was still low or neutral, even at the Maintenance stage.

Subjects at all stages above Precontemplation showed moderate self-efficacy (Q10,
Q12) in the response to the construct of “Interest in fiber-rich foods”. There was no
apparent relation between stages of change for fiber and the two one-question
constructs—“Childhood intake of fiber rich foods” and “difficulty of increasing intake”.
Because a one-question construct is inadequate to contribute to a scaled score—these two

single question constructs were dropped from further analyses.

 

Table 18
The mode“ of responses to Attitudes and Beliefs questions and their distribution by five stages of

change

 

 

 

Health benefits of Social and Interth in Childhood Difficulty of
fiber environmental fiber-rich intake increasing intake
support foods
Stagefln) 01 02 03 09 013 05 Q6 07 08 010 012 04 011
P (17) 4 5 4 4 3 2 1 1 3 3 3 3 4
C(10) 4 5 4 5 4 1 1 1 3 4 4 2 2
D (6) 4 5 2 5 3 4 1 1 3 4 4 3 3
A (13) 4 5 4 5 5 3 1 2 3 4 4 3 4
M (35) 5 5 4 5 5 4 2 3 4 5 4 3 3

 

 

 

 

 

 

' Likert scale from l= Strongly Disagree to 5 = Strongly Agree
b P= Precontemplation, C= Contemplation, D= Preparation, A= Action, M= Maintenance

To examine if subjects’ responses to first three constructs (with two to five questions
each) would demonstrate significant difference by stage, the “score” of each construct was
calculated. The score of a construct was defined as the sum of subjects’ response, based on
the Likert scale, to all questions within the construct. Table 19 shows the mean (:SD) of
the scores for each construct at different stages. Subjects at the Maintenance stage had the

highest scores (p<.05) for the construct of “Health benefits of fiber” compared to those at

60
other four stages. The social and environmental support that precontemplators and

contemplators perceived for consumption of fiber-rich foods was significantly less than
what people at the Maintenance stage perceived. Precontemplators also had significantly
less interest in fiber-rich foods than people at other stages, except for Preparation,
confirming the concepts of SCT. These three constructs only accounted for 56.3% of total
variance of responses, indicating they did not completely reveal all attitudes and beliefs

related to fiber consumption. Thus, Hypothesis 5 was particially supported.

 

Table 19
Mean (: SD) of the scores for three major constructs at the subjects’different stages of readiness to
consume fiber

Health benefits of Social and Interest in fiber-rich foods

 

fiber (5) ‘ environmental support (2)
(4)
Precontemplation (17)b 19.9: 2.4” 8.3: 2.6"c 6.4: 2.0
Contemplation (10) 20.9: 2.6‘ 8.4:3.1* 77:11"d
Preparation (6) 18.8: 2.8“ 92:29 7.8: 1.7
Action (13) 21 .2:2.6* 10.2:39 8.1 :1 .5"
Maintenance (35) 22.8:2.1 11.3: 2.9 78:15"

 

' number of questions in the constnict

b number of subjects at the stage

.- Duncan’s multiple range test, stages vs Maintenance

" Duncan‘s multiple range test, stages vs Precontemplation
"' p<.05

Hypothesis 6. Subjects 'jiber intake from the 2-day food records can be predicted from 3
short instruments: the checklist, SC questionnaire, and the major constructs

from the AB questionnaire

Multiple regression (enter) was used to determine if fiber intake from the weighted 2-

day food records (dependent variable) could be predicted from the fiber intake from

61

checklist, the subject’s stage of change, and scores of the three major constructs in AB

quesitonnaire (independent variables). Table 20 shows the hierarchical and stepwise

prediction of dependent variable from five independent variables.

 

Table 20
Stepwise prediction of fiber intake from weighted 2-day food record using fiber intake from
checklist, the subject’s stage of change, and subject’s scores of the three major constructs in AB

 

 

guesitonnaire

Step Variable r R2 p

1 Gram amounts of fiber intake from checklist .52 .27 <.001
2 Stages of Change .13 .02 .35
3 Construct of “Health benefits of fiber“ .17 .03 .20
4 Construct of “Social and environmental support” .25 .06 .06
5 Construct of “Interest in fiber-rich foods” .19 .03 .17

 

The regression equation was:

Y = 0.46 X1 + 8.54

X I : gram amounts of fiber assessed from the checklist

(n = 64; F = 29.43; df= l; p< .0001)

Y .' gram amounts of fiber assessed from the weighted 2-day food record

Other independent variables (step 2-5) are not shown in the equation because their

correlations with the dependent variable are not significant, although the construct of

“Soical and environmental support” showed a low correlation that approaches significance

(p=0.06). The five independent variables predicted a total of 44% of the variance in fiber

intake—suggesting that the relationship is not likely linear. Therefore, hypothesis 6 was not

supported for this population of predominatly women.

62
Additional Analysis

From the checklist, the rank of foods for their contribution to total fiber intake was
calculated based on two factors: the weekly number of portions consumed; and the grams
of fiber consumed per week by an individual. Table 21 shows the top 20 foods for each of
these two factors. Wheat bread was the food with the largest weekly consumption and apple

was the food providing largest amount of fiber in the subjects’ diet.

 

Table 21

Ranking of foods from the checklist for their contribution to total fiber intake based on the average
weekly number of portions consumed and grams of fiber consumed from that food per week by
each subject

 

 

Food no. of portion/wklperson Food grains of fiber/wklperson
wheat bread 4.50 apple 8.32
apple 3.18 whole grain cereal 7.95
white bread 2.85 wheat bread 7.16
lettuce 2.74 bran cereals 7.05
tossed salad 2.57 potato 5.38
potato 2.56 banana 5.38
banana 2.48 prunes 5.01
tomato 2.28 refried beans 5.00
carrots 2.25 tomato 4.88
bagel 1 .96 oatmeal 4.37
whole grain cereal 1.91 peanut butter 4.36
pasta 1 .45 carrots 3.96
crackers 1.42 baked beans 3.65
broccoli 1.40 green peas 3.16
green beans 1.34 green beans 2.92
peanut butter 1.07 broccoli 2.83
com 1 .05 pasta 2.72
tortillas 0.99 bagel 2.45
oatmeal 0.96 orange 2.30
bran cereals 0.94 pear 2.08

popcorn 0.91 tossed salad 2.01

63

Shown in Table 22 are foods that subjects in this study ate frequently, but were not
listed on the checklist. The foods selected more than three times (white rice, cabbage,
pepper, grapefruit) might be necessary to be included in the checklist in the future to

decrease the inaccuracy between the data from the checklist and actual intake.

 

Table 22
Foods not listed on the checklist but were selected by subjects more than
once

Food Times being selected
White rice
Cabbage
Pepper (green, red)
Grapefruit
Feud
Lentil
Pinto bean
Kiwi fruit
Black bean
Sunflower
Cucumber
Almond
Onion
Gabanza bean

. Sesame seed
Bulgar
Pecan
Radish

mmmmmmmmwwwwwwekam

64
Subjects’ responses were tabulated in Table 23 to the two open-ended questions

“What could you do to increase your intake of dietary fiber ?” and “What keeps you from
eating more fiber-rich foods ?” in the AB questionnaire. Generally, subjects thought the
best way to increase intake of dietary fiber was to consume more fiber-rich cereal, fruits,
and vegetables. It is noticable that two subjects reported use of fiber supplements was a
way they would increase their fiber intake. The major factors for avoidance of fiber-rich
foods were lack of convenience, not enough time to cook, difficulty in having fiber-rich
foods on the go or when eating out. Concerns about the gas production and unpaltable taste

of fiber-rich diet were also significant reasons for not eating fiber-rich foods.

 

TabI023

Subjects’ response to suggestions for and barriers to increasing fiber intake
(n=95)

What could you do to increase your intake of dietary fiber ?

Counts .
Eat more fiber-rich cereals, fniits, and vegetables 40
Improve meal planning 7
Avoid empty calorie snacks 7
Increase availability of fiber-rich foods 5
Learn how to select fiber-rich foods 3
Use fiber supplements 2
What keep you fiom eating fiber-rich foods .7
Lack of convenience (time to prepare, availability) 30
Health concerns (on special diet, avoid gas production) 13
Lack of interest to change 11

Unpalatable taste of fiber-rich foods 8
Feelings of enough consumption 6
Influence from family or spouse 4
Cost 4
Strict dietary preference (small appetite) 3
Lack of knowledge for selecting fiber-rich foods 3
Lack of recipes 1

‘ based on number of opinions generated from open-ended questions, not number of
subjects

65

Discussion

The high sensitivity (80%) of the checklist demonstrated its ability to distinguish
subjects who actually ate low fiber diets from those who reportedly ate low fiber diet. High
sensitivity of a dietary assessment instrument is required for intervention programs to
ensure the strategies are targeted to those who actually are at dietary risk, even though they
might regard their diets as adequate. The good sensitivity of the checklist supports its
feasibility and effectiveness of use as a screening tool for a population similar to this sample

at high risk for insufficient intake of dietary fiber.

Checklist Validation

The correlation between the test checklist and the criterion measure of weighted 2-day
food records was 0.52, which was identical to the result of a similar study by Saba et a1
(1995). Low to moderate correlation (r =0.30-0.50) in measuring dietary fiber intake
between the test instrument and criterion measurement seemed prevalent in the studies
reviewed. The short checklist developed by Kristal et al. had a correlation of 0.40 with
mean of two 4-day food records (Kristal et al., 1990). The FFQ developed by Willett et al.
had a correlation of 0.46 with mean of four 7-day food records (Willett et al., 1985). The
FFQ developed by Jain et al. had correlations of 0.39 for men and 0.52 for women with one
7-day food record (Jain et al., 1996).

Moderate correlations between two dietary assessment techniques might be caused by
random errors incurred in measurement of the diet by both methods. Sources of random
errors include: 1) subjects forgetting to record what they ate; and 2) over- or underestimate
of actual consumption. Other sources of error include: 1) checklist or FFQ design does not

match subject’s ordinary food choices; 2) the foods not included in the database of the

66
analytical software were mistakenly substituted by researcher’s subjective interpretation;

and 3) incorrect portion size transformation by researchers.

All methods of dietary intake measurement require either a direct probe of portion size
or referral to a standard size in order to estimate energy or nutrient content. If portion sizes
are estimated incorrectly, this source of error itself might be sufficient to explain the
discrepancy between self-reported caloric intakes and rates of obesity in the US. (Young et
al., 1995) Moreover, studies have shown that in fact subjects pay little attention to portion
size information on FFQ (Subar et al., 1995). The moderate correlation between the
checklist and the weighted 2—day food record in this study might largely result from this
reason.

Another reason for the moderate correlation might be subjects’ confusion on filling out
the checklist. It was quite common to find subjects entering self-contradictory responses
(> 80%), e. g., filling in 1 time daily and 3 times a week for the same food. This problem
was addressed by uniformly choosing the answer with largest frequency, which might have
led the checklist to overestimate actual fiber intake. A third reason for the moderate
correlation might be that foods were not on the checklist, but were selected by the subjects,
forming a missing source of fiber intake. There were 36 subjects (36%) who indicated at
least one food they frequently ate which was not shown on the checklist. A few of these
foods such as green pepper and cabbage might be significant fiber sources. The fiber
content of these foods was not included, because it would have contaminated the validation
check of the checklist in this study.

The ability of the checklist to classify subjects into low, medium and high intakes, or
percentage of agreement with the weighted 2-day food record, was a modest 47%, but
similar to that found in the self-administered FFQ (53%) where Saba and colleagues used a
7-day weighed food record for validation (Saba et al., 1995). In the present study, 4 subjects
(6%) were placed into opposite subgroups of fiber intake. That is, their fiber intakes were

determined as high by one measurement but low by the other one. This misplacement

67
percentage for the checklist was lower in the present study than the 12% misplacement error

in the study by Saba et al.

Stages of Change Questionnaire

Surprisingly 18 (18%) subjects returned their SC questionnaires blank, while they
completed the other instruments. The problem might be caused by subjects’ lack of
attention to the self-administered questionnaire, but it is more likely a sign of the subjects’
unfamiliarity with the questionnaire in flowchart format. Although the SC questionnaire
did achieve the goal of stratifying subjects effectively, because no subject chose two stages
at the same time, this might also be part of the problem, if some people did not fall neatly
into one of the five stages. The large number of subjects in the Maintenance stage
confirmed that a voluntary, health-related survey is more likely to recruit respondents who
are self-conscious about diet and health.

Table 24 shows a close comparison of fiber intakes at different stages. from the present
study with those from studies done by Spomy (Spomy et al., 1995), Glanz (Glanz et al.,
1994). and Huang (Manuscript in process, 1997). In this study, fiber intake did not increase
significantly, perhaps due to the small sample size, compared to the significant increase of
fiber consumption found earlier at the Action stage in both Spomy’s and Glanz’s studies. In
Huang’s study, the significant increase was not found until the Maintenance stage.

As a way to examine the theoretical relevance of the SC questionnaire to intake of
dietary fiber, the mean and standard deviation of grams fiber consume at each stage among
the cited studies were compared. Except for the Maintenance stage in Spomy’s study and
the Contemplation and Preparation stage in Glanz’s study, there was no difference in the
fiber intake at each stage between this study and the other two cited. These findings

indicate fair compatibility between the checklist and the SC questionnaire in demonstrating

68
the relationship between the actual fiber intake and different stages of readiness to increase

fiber consumption.

 

Table 24
Comparison of gram amount of dietary fiber intake (mean:SD) by stage of change between Shih’s
checklist and data of Glanz et al., 1994; Spomy et al., 1995; and Huanget al., 1997

n Precont. Contemplation Preparation Action Maintenance
Glanz etal.,1994I 17,039 12.6:0.1’ 13.0.0.1b 13.8:02“ 15.4:01d 19.6:02"

 

Spomyetal.,1995 618 118:2.8a 118:2.9a N/A 13.7:36" 14.7:3.7°
Huangeta|.,1997 128 N/A 15.4.1.4: 165:1.1" 16.2:09” 19.6:1.2°

Shihetal.,1997' 81 12.5:90‘ 14.7:106 20.4:13.8 13.1:7.2‘ 19.8:9.3

I Based on 2,000 kcal diet

.4.- Means not sharing the same superscript are significantly different at p<.05 or beyond, using post-hoe
Schefi’e constract.

f Duncan’s multiple range test, Precontemplation, Action vs Maintenance

* p<.05

Distribution of subjects across the stages of change

Table 25 shows the percentage of distribution of subjects at each stage in this study
compared with results from other studies. The highest percentage was at the Maintenance
stage, indicating almost half of the subjects in this study were aware of benefits of dietary
fiber and willing to maintain a fiber-rich diet. In Huang’s study, subjects were the
participants of MSU Healthy U program who already seemed motivated to learn about a
healthy lifestyle. The distribution by stages of change confirmed that assumption and
revealed that the largest group was people at the Action stage. It is believed that a

69
voluntary, health-related mail survey is more likely to recruit respondents who are self-

conscious about diet and health. Both the distributions of this study and Spomy’s
confirmed that assumption, because the methods used for data collection were similar. It
appears that when the survey is conducted on participants of intervention programs,

however, the percentage of people at the Action stage tends to be higher than that at the

 

 

 

 

other stages.
Table 25
Distribution of subjects in five stages of change for fiber intakefrom several relevant studies

n Precont. Contemplation Preparation Action Maintenance
Shih et al., 1997 99 17.1% 10.1% 6% 13.1% 35.3%
Huang et al., 1997 128 0% 7.3% 22.6% 54.7% 15.3%
Spomy et al., 1995 618 7.7% 9.2% N/A 37.8% 45.1%
Glanz et al., 1994 16,980 12.2% 28.1% 8.7% 33.4% 17.6%

 

Attitudes and Beliefs Questionnaire

The AB questionnaire showed a fair content validity with the three major constructs
within in it accounting for 53% of the variance of subjects’ response to attitudes toward
fiber consumption. The AB questionnaire demonstrated ability to reveal characteristics
related to fiber intake at some stages of change. The major three constructs in
questionnaire—health benefits of fiber, social and environmental support, and interest in
fiber-rich foods— resemble three of 10 social psychological variables in Spomy’s study—
perceived benefits, social modeling, and overall health concern (Spomy et al., 1995).

70
Subjects at the Maintenance stage in both studies reportedly perceived more health benefits

of fiber and more social and environmental support to consume fiber-rich foods than their
cohorts at the Precontemplation and Contemplation stage. Subjects at the
Precontemplation stage were also found to have the lowest score in the construct of

“interested in fiber-rich foods” in both studies.

The rank of foods based on the quantitative contribution to
total fiber intake

Compared to Huang’s similar analysis of dietary fiber from 3-day food record of MSU
Healthy U participants (n=128)(Huang, manuscript in process), subjects in this research
showed similar dietary patterns for fiber consumption. Thirteen out of 20 foods appear in
Huang’s list with slight difference in the order. Some less-frequently consumed foods, such
as prune and refiied beans, were listed in the top 20 foods because of their high fiber
content.

Compared to Block and Lanza’s list of dietary fiber sources in U.S.(Block et al., 1987),
analyzed from NHANES II data, a different dietary pattern of fiber consumption was found
in the present study. The top 1 and 2 foods on Block and Lanza’s list were white bread and
pinto and other dried beans, respectively, which were not in ranking of top 20 foods
consumed by this population. White bread provided 11.4% of total fiber intake in
NHANES II (n=1 1,658), but it only accounted for 1.2% in the present study. The top 1 food
in the present study was apple, which accounted for 7% of total fiber intake but only 3.6% in
NHANES II.

71
Strengths

This research has the following strengths.

1) A high sensitivity (80%) for the checklist distinguished subjects who actually had
low fiber intake from those who reportedly had high intakes.

2) The short administration time of the three test instruments was good. It took less
than 10 minutes to complete the checklist and both SC and AB questionnaires,
based on the responses from the pilot test.

3) The study prombted awareness among respondents of dietary fiber intake, a topic
ofien neglected in health promotion interventions.

4) Good compatibility was found among three instruments to form a validated
screening package for fiber intake and readiness to increase fiber consumption in

this sample.

Limitations

This research also has the following limitations.

1) There was only a moderate correlation (0.52) between the checklist and the
weighted 2-day food records for fiber intake.

2) There was a small sample size, due to the poor response rate of 25%. The large
proportion of respondents not answering the SC questionnaire lowered the
statistical power of analysis.

3) The SC questionnaire did not go through a separate validation process.

4) The 18% of non-respondents to the SC flowchart was worrisome and must be

explored further before this tool can be recommended for use.

72

5) The two constructs of “Childhood intake of fiber” and “ Difficulty of increasing in

take of fiber” had only one question each in the factor analysis, yet together
accounted for nearly 20% of the variance in response to attitudes and beliefs.
Future research might expand the number of questions for these two constructs so

that they could be added to the scale of “Attitudes and Beliefs about Fiber”.

The following reasons might be responsible for the low response rate in this study.

1)
2)

3)
4)

5)

We were not able to send an official endorsement letter from the President of the
CTU with the survey.

Incentives were not attractive enough to the target population.

The time-consuming 2-day food recording task reduced subject’s interest.
Non-respondents might have had a negative influence on those who were hesitant
at first and who worked within the same department, because subjects from the
same department tended to be respondents or non-respondents collectively.

Subjects had low interest in the health benefits of dietary fiber.

73

Conclusion

This study demonstrated the development and validation of a new dietary assessment
tool using behavioral approach. Results demonstrated that the combination of the
theoretical model and brief dietary assessment was a feasible way to identify a high risk
population before the intervention begins. Due to limited time, money, and resources,
findings from this study must be considered preliminary and should be repeated with larger,
more diverse populations than the 99 CTU’s at MSU. Future research efforts might focus
on:

1) The validation of the checklist with larger, more diverse populations.

2) Expanding the number of questions for the constructs—“Childhood intake of
fiber” and “Difficulty increasing fiber intake”——perhaps by using responses from
the open-ended questions. ‘

3) Exploring the reasons for non-response to the flowchart SC questionnaire and
redesign an usable version.

4) Developing appropriate intervention strategies for people before the Action stage.

Bibliography

74

ADA. Position of The American Dietetic Association: health implications of dietary fiber.
J Am Diet Assoc. 1988;88:216-221.

ADA reports. Position of the American Dietetic Association: Health implications of dietary
fiber. J Am Diet Assoc. 1993;93:1446-1447.

Alaimo K, McDowell MA, Briefel RR, Bischof AM, Caughman CR, Loria CM, Johnson
CL. Dietary intake of vitamins, minerals,and fiber of persons ages 2 months and
over in the United States: Third National Health and Nutrition Examination
Survey, Phase 1, 1988-91. Advance Data, National Center for Health Statistics.
No.258,Nov,1994.

Alberton AM, Tobelman RC. Consumption of grain and whole-grain foods by an American
population during the years 1990 to 1992. J Am Diet Assoc. 1995;95:703-704.

Anderson JW, Akanji AO. Dietary fiber -— an overview. Diabetes Care. 1991;14:1126-
1131.

Anderson J W, Smith BM, Gustafson NJ. Health benefits and practical aspects of high-fiber
diets. Am J Clin Nutr. 1994;59(suppl):124ZS-7S.

Anderson J W, Story L, Sieling B, Chen W-JL, Petro MS, Story J. Hypocholesterolemic
effects of oat-bran or bean intake for hypercholesterolemic men. Am J Nutr.
1984;40:1146-1155.

Anderson JW. Dietary fiber in the nutrition management of diabetes. In: Basic and Clinical
Aspects of Dietary Fiber. G. Vahouny and D. Kritchevsky, Eds. 343-350. New
York, NY: Plenum Press, 1985.

AOAC. Official methods ofanalysis. 15th ed. Arlington, VA, 1990.

Asp NG, Schweizer TF, Southgate DAT, Theander 0. Dietary fibre analysis. In Dietary
Fibre — a component of food. Nutritional function in health and disease, eds TF
Schweizer & CA Edwards, pp 57-101. London: Springer Verlag., 1992.

Asp NG. Dietary fibre analysis — an overview. Eur J Clin Nutr. 1995;49:842-S47.

Atwood JR, Aickin M, Giordano L, Benedict J, Bell M, Ritenbaugh C, Rees-McGee S,
Sheehan E, Buller M, Ho EE, Meysken FL, Albert D. The effectiveness of

adherence intervention in a colon cancer prevention field trial. Prev Med.
1992;21:617-53.

75

Baghurst KI, Record SJ, Baghurst PA, Syrette JA, Crawford D, Worsley A. Socio-
demographic determinants in Australia of the intake of food and nutrients
implicated in cancer epidemiology. Med J Aust. 1990;153:444-452.

Baranowski T, Simons-Morton BG. Dietary and physical activity assessment in school-
aged children: Measurement issues. J School Health. 1991;61:195-196.

Block G, Clifford C, Naughton D, Henderson M, McAdams M. A brief dietary screen for
high fat intake. J Nutr Educ. 1989;21:199-207.

Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based
approach to diet questionnaire design and testing. Am J Epidemiol.
1986;124:453-469.

Block G, Hartman AM. Issues in reproducibility and validity of dietary studies. Am J Clin
Nutr. 1989;59:1133-1138.

Block G, Lanza E Dietary fiber sources in the United States by demographic group. J
Natl Cancer Inst. 1987;79(1 ):83-91 .

Block G. Health habits and history questionnaire: Diet history and other risk factors.
Bethesda, MD: National Cancer Institute, 1988.

Block G. Human dietary assessment: Methods and issues. Prev Med. 1989;18:653-660.
Block G. Validations of dietary assessment methods. Am J Epidemiol. 1982;115:492-505.

Boeing H, Wahrendorf J, Heinemann L, Kulesza W, Rywik SL, Sznajd J, Thiel C. Results
from a comparative dietary assessment in Europe.- -I. Comparison of dietary
information derived from concurrently applied frequency questionnaires and
quantitative measurement instruments. Eur J Clin Nutr. 1989;43:367-377.

Bridges SR, Anderson JW, Deakins DA, Dillon DW, Wood CL. Oat bran increases serum
acetate of hypercholesterolemic men. Am J Clin Nutr. 1992;56:455-459.

Brinberg D, Axelson ML. Increasing the consumption of dietary fiber: A decision theory
analysis. Health Educ Res. 1990;5z409-420.

Brodribb AJM. Dietary fiber in diverticular disease of the colon. 1n Spiller GA. and Kay
R.M., eds: Medical Aspects of Dietary Fiber. New York, NY: Plenum Press, 1980.

Brownell KD, Cohen LR. Adherence to dietary regimens 2: Components of effective
interventions. Behav Med. 1995;20:155-164.

76
Bunch KL. 1989 California dietary practices survey, focus on fruits and vegetables
highlights, nutrition and cancer prevention program, California Dept. of Health
Services, California Public Health Foundation. Nutrition and cancer prevention
program, Sacramento, CA., 1990.

Burkitt DP, Trowell HC. Refined carbohydrate foods and disease. Some implications
of dietary fibre. London, England: Academic Press, 1975.

. - Campbell MK, DeVellis BM, Strecher VJ, Amerrnan AS, DeVellis RF, Sandler RS.

Improving dietary behavior: The effectiveness of tailored messages in primary care
settings. Am J Public Health. 1994;84:783-7.

Carmines G, Zeller A. Reliability and validity assessment. Beverly Hills, CA.:Sage,1979.

Carruth BR, Anderson HL. Scaling criteria in developing and evaluation an attitude
instrument. J Am Diet Assoc. 1977;70:42-50.

Chenoweth WL, Bennick HR. Hypocholesterolaemic effect of oat fibre. Fed Proc.
1976;35:495.

Comi D, Brugnani M, Gianino A. Metabolic effects of hypocaloric high-
carbohydrate/high-fibre diet in non-insulin dependent diabetic patients. Eur J Clin
Nutr. 1995;49:8242-S244.

Contento 1, Murphy BM. Psycho-social factors differentiating people who reported making
desirable change in their diets from those who did not. J Nutr Educ. 1990;22:6-14.

Contento IR, Maksymowicz-Murphy B. Psycho-social factors differentiating pe0ple who
reported making desirable changes in their diets from those who did not. J Nutr
Educ. 1990;22:6-14.

Contento IR. Theoretical frameworks or models for nutrition education. J Nutr Educ.
1995;27:287-292.

Cotugna N, Subar AF, Heimendinger J, Kahle L. Nutrition and cancer prevention
knowledge,beliefs, attitudes, and practices: The 1987 National Health Interview
Survey. J Am Diet Assoc. 1992;92:963-968.

Cummings JH, Branch WJ. Postulated mechanisms whereby fibre may protect againtst
large bowel cancer. In Dietary fibre in health and disease, eds G Vahouny &D
Kritchevsky, pp 313-325. Plenum: New York, NY, 1982.

Curry SJ, Kristal AR, Bowen DJ. An application of the stage model of behavior change to
dietary fat reduction. Health Educ Res. 1992;7297-105.

77

DiClemente CC, Hughes SO. Stages of change profiles in alcoholism treatment. J Subst
Abuse Treat. 1990;2z217-235.

DiClemente CC, Prochaska JO, Fairhurst SK, Velicer WP, Velasquez MM, Rossi J S. The
process of smoking cessation: An analysis of precontemplation, contemplation and

preparation stages of change. J Consult Clin Psych 1991;59:295-304.

Dittus KL, Hiller VN, Beerman KA. Benefits and barriers to fruit and vegetable intake:
Relationship between attitudes and consumption J Nutr Educ. 1995;27:120-126.

Dreher ML. Handbook of dietary fiber. New York, NY.: Marcel Dekker, 1987.
Eastwood MA, Passmore R. Dietary Fibre. Lancet. 1983;July:202-206.

Eastwood MA. The physiological effect of dietary fiber. An update. Annu Rev Nutr.
1992;12:19-35.

Ershoff, DH, Mullen PD, Quinn VP. A randomized trial of serialized self-help smoking
cessation program for pregnant women in an HMO. Am J Public Health.
1989;79:182-187.

FDA. Fed Register. 1993;58:2302-2964.

Frelich W. Bioavailability of micronutrients in a fibre-rich diet, especially related to
minerals. Eur J Clin Nutr. 1995;49:S1 16-S122.

Frazao B, Cleveland L. Diet-health awareness about fat and cholesterol—only a start.
Foodreview. 1994;Jan-Apr115-22.

Glanz K, Eriksen M. Individual and community models for dietary behavior change. J Nutr
Educ. 1993;25:80-6.

Glanz K, Kristal AR, Sorensen G, Palombo R, Heimendinger J, Probart C. Development
and validation of measures of psychosocial factors influencing fat- and fiber-
related dietary behavior. Prev Med 1993;22:373-387.

Glanz K, Lewis FM, Rimer BK. Health Behavior and Health Education: theory research and
practice. San Francisco, CA: Jossey-Bass Publishers, 1990.

Glanz K, Mullis RM. Environmental interventions to promote healthy eating: A review of
models, programs, and evidence. Health Educ Q. 1988;15:395-415.

Glanz K, Patterson RE, Kristal AR, DiClemente CC, Heimendinger J, Linnan L, McLerran
DF. Stages of change in adopting healthy diets: Fat, fiber, and correlates of nutrient
intake. Health Educ Q. 1994;21:499-519.

78
Glasgow RE, Peny JD, Toobert DJ, Hollis JF. Brief assessments of dietary behavior in field
settings. Addict Behav. 1996;21:239-247.

Greene GW, Rossi SR, Reed GR, Willey C, Prochaska JO. Stages of change for reducing
dietary fat to 30% of energy or less. J Am Diet Assoc. 1994;94:1105-1110.

Greenberg RS, Daniels SR, Flanders D, Eley JW, Boring JR. Medical epidemiology. 1" ed.
Appleton & LangezNorwalk, Conn, 1993.

Guyatt G, Walter S, Norman G. Measuring change over time: Assessing the usefulness of
evaluative instruments. J Chron Dis. 1987;40:171-8.

Haber GB, Heaton KW, Murphy D. Depletion and disruption of dietary fibre: Effects on
satiety, plasma-glucose, and serurn-insulin. Lancet. 1977;ii:679-682.

Havas S, Heimendinger J ,Damron D, Nicklas TA, Cowan A, Beresford SAA, Sorensen G,
Buller D, Bishop D, Baranowski T, Reynolds K. 5 A Day for Better Health— Nine
community research projects to increase fruit and vegetable consumption. Public
Health Reports. 1995;110:68-79.

Healthy People 2000: National health promotion and disease prevention objectives.
Washington, DC: US. Dept. of Health and Human Services, Public Health
Services; 1990.

Heimendinger J. Community nutrition intervention strategies for cancer risk reduction.
Cancer. 1993;72:1019-23.

Hill MJ, Crowther J S, Drasar BS, Hawksworth G, Aries V, Williams REO. Bacteria and the
aetioloty of cancer of the large bowel. Lancet. 1971;95-100.

Hipsley EH. Dietary “fibre” and pregnancy toxaemia. Br Med J. 1953;22420-422.

Hjollund E, Pederson O, Richelsen B, Beck-Nielson H, Sorenson NS. Increased insulin
binding to adipocytes and monocytes and increased insulin sensitivity of glucose
transport and metabolism in adipocytes from non-insulin dependent diabetics after
low-fat/high-starch/high-fiber diet. Metabolism. 1983;32: 1067-1075.

Holland D, Welch A, Unwin I, Duss D, Paul A, Southgate D. McCance and Widdowson’s
The Composition of Foods. 5‘h ed. Cambridge, UK: Royal Society of Chemistry,
1991.

Huang YL, Song WO, Schemmel RA, Hoerr SM. What do college students eat? Food
selection and meal pattern. Nutr Res. 1994;14:1143-1153.

79
Huang YL. Dietary change of women after the worksite nutrition programs predicted by
blood lipids, BM], dietary intakes, stages of change and psychosocial factors.
Manuscript in process, 1997.

Hunter JE. Cluster analysis: Reliability, construct validity, and the multiple indicators
approaches to measurement. Presented at a workshop titled “Advanced Statistics”
given at the US. Civil Services commission, 1977.

ILSI (International Life Science Institute). Dietary fibre: Nutritional function in health and
disease. Washington, DC, 1995.

Jacobs LR. Effect of dietary fiber on colon cell proliferation and its relationship to colon
carcinogenesis. Prev Med. 1987;16:566-571.

Jain M, Howe GR, Rohan T. Dietary assessment in epidemiology: Comparison of a food
frequency and a diet history questionnaire with a 7-day food record. Am J
Epidemiol. 1996;143:953-960.

Jenkins DJA, et al. Dietary fiber and bloods: treatment of hypercholesterolemia with guar
crisbread. Am J Clin Nutr. 1979;33:575-581.

Jenkins DJA, et al. Dietary fibre and blood lipids: reduction of serum cholesterol in Type II
hyperlipidemia by guar gum. Am J Clin Nutr. 1980;32:16-18.

Jenkins DJA, Wolever TMS, Vuksan V, Brighenti F, Cunnane SC, Rao AV, Jenkins AL,
Buckley G, Patten R. Singer W, Corey P, Josse RG. Nibbling versus gorging:
metabolic advantages of increased meal frequency. N Engl J Med.

1989;321 :929-934.

Jensen J. Counte MA, Glandon GL. Elderly health beliefs, attitudes and maintenance. Prev
Med. 1992;2 1: 483-97. - - .

Johnson DW, Johnson RT. The process of nutrition education: a model for effectiveness, a
synthesis of research. J Nutr Educ. 1985;17:81-44.

Johnson SB, Freund A, Silverstein J, Hansen CA, Malone J. Adherence-health status
relationships in childhood diabetes. Health Psych. 1990;92606-63 l.

Kaaks R, Riboli E. Colorectal cancer and intake of dietary fibre. A summary of the
epidemiological evidence. Eur J Clin Nutr. 1995;49:810-S17.

Kay RM, Strasberg SM, Peetruka CN, Wayman M. Differential absorption of bile acids by
lignins. In Inglett GE and Falkenhag SI eds: Dietary fibres, Chemistry and
Nutrition. New York, NYzAcademic Press, 1979.

80

Klurfeld DM. Dietary fiber-mediated mechanisms in carcinogenesis. Cancer Res.
1992;52:20558-2059s.

Krebs-Smith SM, Cronin FJ, Haytowitz DB, Cook DA. Food sources of energy
macronutrients, cholesterol, and fiber in diets of women. J Am Diet Assoc.
1992;92:168-174.

Kristal AK, Bowen DJ, Curry SJ, Shattuck AL, Henry HJ. Nutrition knowledge, attitudes
and perceived norms as correlates of selecting low-fat diets. Health Educ Res.
1990;5:467-477.

Kristal AR, Beresford SA, Lazovich D. Assessing change in diet-intervention research. Am
J Clin Nutr. 1994;59(supp1):1855-9s.

Kristal AR, Patterson RE, Glanz K, Heimendinger J, Hebert JR, Fcng Z. Probart C.
Psychosocial correlats of healthful diets: Baseline results from the Working Well
Study. Prev Med 1995;24:221-228.

Kristal AR, Shattuck AL, Henry HJ, Fowler AS. Rapid assessment of dietary intake of fat,
fiber, and saturated fat: Validity of an instrument suitable for community

intervention research and nutritional surveillance. Am J Health Prom.
1990;42288-295.

Kristal AR, Shattuck AL, Henry HJ. Patterns of dietary behavior associate with selecting
diets low in fat: Reliability and validity of a behavioral approach to dietary
assessment. J Am Diet Assoc. 1990;90:214-20.

Kritchevsky D, Story J A. Binding of bile salts in vitro by non nutritive fibre. J Nutr.
1974;104:458-464. ‘ ‘

Kushi LH, Folsom AR, Jacobs DR, Luepker RV, Elmer PJ, Blackburn H. Educational
attainment and nutrient consumption patterns: The Minnesota Heart Survey. J Am
Diet Assoc. 1988;88:1230-1236.

Laforge RG, Greene GW, Prochaska JO. Psychosocial factors influencing low fruit and
vegetable consumption. J Behav Med 1994;17:361-374.

Lanza E, Jones DY, Block G, Kessler L. Dietary fiber intake in the US. population. Am J
Clin Nutr. 1987;46:790-797.

Lee SC, Prosky L. International survey on dietary fiber: Definition, analysis, and reference
materials. J AOAC Int. 1995;78:22-36.

81

Levy AS, Fein SB, Stephenson M. Nutrition knowledge levels about dietary fats and
cholesterol: 1983-1988. J Nutr Educ. 1993;25:60-6.

Li BW. Determination of total dietary fiber in foods and food products by using a single-
enzyme, enzymatic-gravimetric method: lnterlaboratory study. J AOAC Int.
1995;78:1440-1444.

Lindegarde F, Larsson L. Effects of a concentrated bran fibre separation on HDL-
cholesterol in hypocholesterolaemic men. Hum Nutr Clin Nutr. 1984;38C:39.

Lwanga SK, Lemeshow S. Sample size determination in health studies: A practical manual.
World Health Organization; Geneva, 1991.

MacLennan R, Jensen OM, Mosbech J, Vuori H. Diet, transit time, stool weight, and colon
cancer in two Scandinavian populations. Am J Clin Nutr. 1978;31:5239-5242.

Manhire A, Henry CL, Hartog M, Heaton KW. Unrefined carbohydrates and dietary
fibre in treatment of diabetes mellitus. J Hum Nutr. 1981;35:99-101.

Marcus BH, Banspach SW, Lefebvre RC, Rossi J S, Carleton RA, Abrams DB. Using the
stages of change model to increase the adoption of physical activity among
community participants. Am J Health Promot. 1992;62424-429.

Margetts BM, Beilin LJ,Armstrong BK, Vandongen R. Vegetarian diet in mild
hypertension: effects of fat and fiber. Am J Clin Nutr. 1988;48:801-5.

Marlett J A. Content and composition of dietary fiber in 117 frequently consumed foods. J
Am Diet Assoc. 1992;92:175-186.

McConnaughy EA, Prochaska JO, Velicer WF. Stages of change in psychotherapy:
Measurement and sample profiles. Psychotherapy: Theory, Research, and
Practice. 1983;20:368-375.

Modan B, Barell V, Lubin F, Modan M, Greenberg RA, Graham S. Low fiber intake as an
etiologic factor in cancer of the colon. J Nat Cancer Inst. 1975;55:15-18.

National Cancer Institute. Cancer control objectives for the nation: 1985-2000.
Washington, DC: US Dept of Health and Human Services; 1986. NIH publication
86-2880, no.2.

National Research Council. Committee on Diet and Health, Food and Nutrition Board.
Diet and health. Implications for reducing chronic disease risk. Washington DC:
National Academy Press, 1989.

82

National Research Council. Recommended Dietary Allowances. 10Lb edition. Washington,

DC: National Academy Press, 1989.

Nicklas TA, Farris RP, Myers Leann, Berenson GS. Dietary fiber intake of children and
young adults: The Bogalusa Heart Study. J Am Diet Assoc. 1995;95:209-214.

Norcross JC, Prochaska JO, Hambrecht M. Treating ourselves vs. Treating our clients: A
replication with alcohol abuse. I Subst Abuse Treat. 1991;32123-129.

Nuttall FQ, Brunzell JD Principles of nutrition and dietary recommendations for
individuals with diabetes mellitus. J Am Diet Assoc. 1979;75:527-530.

Ockene JK, Kristeller JL, Goldberg R, Ockene I, Merriam P, Barrett S, Pekow P, Hosmer D,
Gianelly R. Smoking cessation and severity of disease: The Coronary Artery
Smoking Intervention Study. Health Psychol. 1992;11:119-126.

Ockene JK, Ockene IS. Helping patients to reduce their risk for coronary heart
disease: an overview. In Ockene IS, Ockene JK eds. Prevention of Coronary Heart
Disease. Boston, Mass: Little, Brown; 1990.

O'Connell DO, Velicer WF. A decisional balance measure of the stages of change model
for weight loss. Int J Addict. 1988;23:729-740.

Patrick D, Elinson J. Methods of sociomedical research. Handbook of medical sociology.
In Freeman H, Levin S, Reeder L eds. Prentice-Hall, Inc., Englewood Cliffs,
NJ .: 1979.

Patrick DL, Beery WL. Measurement issues: Reliability and validity. Am J Health Promov.
1991;52305-310.

Patterson RE, Haines PS, Popkin BM. Diet Quality Index: Capturing a multidimensional
behavior. J Am Diet Assoc. 1994; 94: 57-64.

Pennington J A. Bowes and Church’s food values of portions commonly used. 15th ed. New
York, NY: Harper & Row, Publishers, 1989.

Pilch SM. Physiologic effects and health consequences of dietary fiber. Life Science
Research Office, Bethesda, MD: Federation of American Societies for
Experimental Biology, 1987.

Prochaska JO, DiClemente CC. Stages and processes of self-change in smoking: Toward an
integrative model of change. J Consult Clin Psychol. 1983;52390-395.

83
Prochaska JO, DiClemente CC. Towards a comprehensive model of change. In Miller W
and Heather N eds. Treating addictive behaviors, Plenum Press, New York,
NY.,1986. .

Prochaska JO, DiClemente CC. Common processes of self-change in smoking, weight
control, and psychological distress. In S. Shiffman & TA Wills (Eds), Coping and
Substance Use. San Diego, CA: Academic Press, 1985.

Prochaska J O, DiClemente CC. Stages of change in the modification of problem behaviors.
In Hersen M, Eisler RM, Miller PM eds.: Progress in Behavior Modification.
Sycamore, 111.: Sycamore Press, 1992.

Prochaska JO, DiClemente CC. The transtheoretical approach: Crossing traditional
boundaries of therapy. Melbourne, FL.:Krieger., 1991.

Prochaska JO, DiClemente CC, Norcross JC. In search of how people change: application
to addictive behaviors. Am Psycho. 1992;47:1102-1114.

Prochaska JO, DiClemente CC. Transtheoretical therapy: Toward a more integrative model
of change. Psychotherapy. 1982;20:161-173.

Prochaska JO, Norcross JC, Fowler JL, Follick MJ, Abrams DB. Attendance and outcome
in a work site weight control program: Processes and stages of change as process
and predictor variables. Addict Behav. 1982;7:133-142.

Prochaska JO, Velicer WF, DiClemente CC, Fava JL. Measuring processes of change:
Applications to the cessation of smoking. J Consult Clin Psych. 1988;56:520-528.

Prochaska JO. Self-change processes, self-efficacy and self-concept in relapse and
maintenance of cessation of smoking. Psych Reports. 1982;51:983-990.

Prochaska, JO, Norcross JC, DiClemente CC. Changing for Good. New York, NY. William
Morrow and Co. Inc. 1994.

Prosky L, Asp NG, Furda I, Devries JW, Schwezer TF, Harland BF. Determination of total
dietary fiber in foods, food products, and total diets: lnterlaboratory study. J Assoc
Off Anal Chem. 1984;67: 1044-1051.

Pusateii DJ, Roth WT, Ross JK, Shultz TD. Dietary and hormonal evaluation of men at
different risks for prostate cancer: plasma and fecal hormone-nutrient
interrelationships. Am J Clin Nutr. 1990;51:371-7.

Putnam JJ. American eating habits changing: Part 2, Grains, vegetables, fruit, and sugars.
, FoodRev. 1994;17:36-48.

84
Rhoades R, Pflanzer R. Human physiology. 2"d ed. New York, NY: Saunders College
Publishing, 1992.

Rosenstock IM. The Health Belief Model and preventive health behavior. Health Educ
Monographs. 1974;2z354-386.

Ross JK, Pusateri DJ, Shultz TD. Dietary and hormonal evaluation of men at different risks
for prostate cancer: fiber intake, excretion, and composition, with in vitro evidence
for an association between steroid hormones and specific fiber components. Am J
Clin Nutr. 1990;51:365-70.

Rossi SR, Greene GW, Reed G, Prochaska JO, Velicer WF, Rossi J S. A comparison of four
Stage of Change algorithms for dietary fat reduction. Presented. at the Society for
Behavioral Medicine Annual Meetings, San Francisco, CA., March, 1993.

Saba A, D’Amicis A, Turrini A. Evaluation of a fibre questionnaire using weighed records.
Eur J Clin Nutr. 1995;49:8135-8138.

Saltzer EB. Locus of control and the intention to lose weight. Health Educ Monographs.
1978;611 18-28.

Samet JM, humble CG, Skipper BE. Alternatives in the collection and analysis of food
frequency interview data. Am J Epidemiol. 1984;120:572-581.

Sandoval WM, Heller KE, Wiese WH, Childs DA. Stage of Change: A model for nutrition
counseling. Top Clin Nutr. 1994;9:64-69.

Schlamowitz P, Halberg T, Wamoe O, Wilstrup F, Ryttig K. Treatment of mild to moderate
hypertension with dietary fibre. Lancet. 1987;ii:622-3.

Schneeman BD. Dietary fiber: physical and chemical properties, methods of analysis and
physiological effects. Foods Tech. 1986;40:104-110.

Schneeman BO, Lefevre M. Effectsof fibre on plasma lipoprotein composition. In Dietary
fibre: basic and clinical aspects, eds GV Vahony & D Kritchevsky, pp 309-331.
New York, NY: Plenum Press, 1985.

Schwarzer R. Self-efficacy in the adoption and maintenance of health behaviors: theoretical
approaches and a new model. In: Schwarzer R, ed. Self-efficacy: thought control of
action. Washington, DC: Hemisphere Publishing, 1992.

Selvendran RR, DuPont MS. Analysis of dietary fibre. In Developments in food analysis
techniques, ed RD King, pp 1-68. London: Applied Science, 1984.

85
Shelton DL. How people change. Am Med News. 1995;Jan:9-10.

Shinnick FL, Hess RD, Fischer MH, Marlett J. Apparent nutrient absorption and upper
gastrointestinal transit with fiber-containing internal feedings. Am J Clin Nutr.
1989;49:471-475.

Sigrnan-Grant M. Stages of change: A framework for nutrition interventions. Nutrition
Today 1996;31:162-170.

Smallwood DM, Blaylock JR. Fiber: Not enough of a good thing? Foodreview.
1994;Jan-Apr:23-29.

Smith AF, J obe JB, Mingay DJ. Question-induced cognitive biases in reports of dietary
intake by college men and women. Health Psychol. 1991;10:244-51.

Smith AM, Baghurst K, Owen N. Socioeconomic status and personal characteristics as
predictors of dietary change. J Nutr Educ. 1995;27:173-181.

Smith AM, Baghurst KI. Public health implications of dietary differences between social
status and occupational category groups. J Epidemiol Comm Health.
1992;46:409-4 16.

Smith AM, Owen N. Associations of social status and health-related beliefs with dietary fat
and fiber densities. Prev Med. 1992;21:735-45.

Southard DR, Winett RA, Walberg-Rankin JL et al. Increasing the effectiveness of the
National Cholesterol Education Program: Dietary and behavioral interventions for
clinical settings. Ann Behav Med. 1992;14:21-30.

Southgate DAT. Definitions and terminology of dietary fiber. In Vahouny GV, and
Kritchevsky D eds: Dietary fiber in health and disease. New York, NY.: Plenum
Press, 1982.

Spomy LA, Contento IR. Stages of change in dietary fat reduction: Social psychological
correlates. J Nutr Educ. 1995;27:191-199.

Subar AF, Thompson FE, Smith AF, Jobe JB, Ziegler RG, Potischman N, Schatzkin A,
Hartman A, Swanson C, Kruse L, Hayes RB, Riedel-Lewis D, Harlan LC.
Improving food frequency questionnaires: A qualitative approach using cognitive
interviewing. Am J Diet Assoc. 1995;95:781-788.

Subar AS, Heimendinger J, Krebs-Smith SM, Patterson BH, Kessler R, Pivonka E. S-A-
Day for better health. A baseline study of American's fruit and vegetable

86

consumption, National Cancer Institute, National Institutes of Health, Rockville,
MD., 1992.

Trenkner LL, Rooney B, Viswanath K, Baxter J, Elmer P, Finnegan JR. Development of a
scale using nutrition attitudes for audience segmentation. Health Educ Res.
1990;5:479-487.

Trowell H, Southgate DA, Wolever TM. Dietary fibre redefined. Lancet. 1976;12967.

US. Department of Agriculture. Nutrient Database for Individual Food Intake Surveys.
Release 4. Washington, DC, 1990.

US. Department of Agriculture. Composition of Food: Raw, Processed, Prepared Agric.
Handbook No. 8. Washington, DC, 1990.

Vega WA, Sallis JF, Patterson TL, Rupp JW, Morris JA, Nader PR. Predictors of dietary
change in Mexican American families participating in a health behavior change
program. Am J Prev Med. 1988;42194-9.

Venter CS, Vorster HH. Possible metabolic consequences of fermentation in the colon for
humans. Med Hypoth. 1989;29:161-166.

Weisburger JH. On the mechanisms relevant to nutritional carcinogenesis. Prev Med.
1987;16:586-591.

Willett WC, Reynolds RD, Cottrell-Hoehner S, Sampson L, Browne ML. Validation of a
semi-quantitative food frequency questionnaire: Comparison with a 1-year diet
record. J Am Diet Assoc. 1987;87:43-47.

Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschie J, Hennekens CH,
Speizer FE. Reproducibility and validity of a semiquantitative food frequency
questionnaire. Am J Epidemiol. 1985;122:51

Willett WC. Nutritional Epidemiology. New York, NY: Oxford University Press, 1990.

Wright A, Burstyn PG, gibney MI. Dietary fibre and blood pressure. Br Med J
1979;ii:1541-43.

Young LR, Nestle M. Portion sizes in dietary assessment: Issues and policy implications.
Nutr Rev. 1995;53:149-158.

Appendices

Appendix A

87

MICHIGAN STATE
u N1 v E a s r r it

September 10. 199‘

ro: Sharon Hoerr
204 on. Trout Building
Michigan scaca University

RE: rear: 95-411
rrrca: rue APPLICATION or srAcas or CHANGE rxaoar ON
turnovanasr or socuaana AND INSOLUABLE orarAnr
crass rsrAxa
aavrsrou aaouasrso: 08/21/96
CAracoav: -c
APPROVAL DATE: 09/18/96

The university Commiceaa on Raaaarch Involving Human Subjects'IUCRIKSl
raviaw of this project is complete. I am ploaaad co advxsa than the
rights and waltara of rho humnn gubjacca appear to be adequately
pgocaccad and methods co obtain informad consanc are appropriate.
b:ratora. cha UCRIHS approved this project and any ravxsions llscad
a va.

-.

RIflIflhL: UCRIHS approval is valid tor one calendar year. beginning with
the approval data shown above. Investigators planning co
continua a projac: beyond one year must use the groan renewal
torn (enclosed with c a original a proval lancer or when a
projecc is renawadl to soak u dac certification. There is a
mnxxmum of tour such axpadl ranawals possible. Investigators
wishing to continua a reject beyond than time naad co submae I:
again or complete rav aw;

 

RIVISIONS: UCRIHS muse ravlaw any changaa in rocaduraa involving human
subjects. prior to lnaciacion o! c a change. If this is done a:
cha.clma o ranawal. plaasa uaa cba green renewal torn. To
ravasa an approvad protocol a: an ochar time during the year,
send your written request to the IRS Chair, requesting ravrsed
approval and ratarancang cha project's IRS 3 and cicla. Include
in your request a daacrapcron o: cha.change and any revised
Instruments. consent forms or advarcrsamancs ehac ara applxcabla.

PROBLIKSI .

cnaxcrs: Should aichar of tho followin ariaa during the course of the
work, investigators must not; DCRIHS promptly: (ll roblams
(unexpected aida attaccs, comp azncs. ccc.) involving uman
subjeccs or 12) changes 1n the research environmen: or new
intermacron rndicaclng graacar rlsk co the human sub aces than
exrscad when the prococol was pravzoualy reviewed an approved.

I: we can be of any fucura help. lease do not hesitate to contact us
an (517)355-2l80 or rAx (Sl7ls 2- I71.

kl“

Sincerely.

   

 
  

vid 3. Wright.
CRIHS Chair

DEH:b¢d

Appendix B

September 28, 1996

Dear MSU CTUs:

The importance of dietary fiber to prevent several chronic diseases—including
colonic cancer, diverticular disease, diabetes, etc—has been mentioned in various
publications. However, the US. public seems less concerned about fiber than about fat
in their diets. To promote the awareness of its health benefits by increasing daily fiber
intake, we are going to conduct a survey about your readiness to change fiber
consumption. Some of you will receive a package of short questionnaires via campus
mail next week. We also will enclose a $1 coupon for MSU dairy store in the package
and ofi‘er a $25 value of dietary assessment to express our appreciation for your
participation. This survey has been granted by MSU Healthy U Committee and
received the permission from the President of MSU CTU.

Best Regards

Sharon Hoerr, RD., PhD. ' Won Song, RD., PhD.
Associate Professor Associate Professor
Dept. of Food Science and Human Nutrition(FSPH\l) Dept. of FSHN

Ralph Levine, PhD.
PrOfessor
Dept. of Psychology

Appendix C

89

The purpose of this study is to develop aid validate: l) a short checklist to assess intake of
dietary fiber. 2) a flowchart to determine where you at mlative to dietary fiber intake; 3)a questionnaire
to assess your attitudes and behaviors related to dietary fiber. and 4) your ideas about how to
increase the intake of fiber-rich foods. it will take you about 10-15 minutes to complete all three
instruments. You are or course tree to refuse to answer any question that makes you uncomfortable.

We are also asking you to provide a 2<lay record of the foods you eat for our cross-
reference. All of the data you provide for the Questionnaires and dietary records will be analyzed
confidentially. Only the research team will have access to the original data and no individuals will be
identified by name.

To express our appreciation for your consideration and . we hope your cooperation. a $2.00
coupon for the MSU Dairy Store is enclosed for your use, whetheror not you complete and return the
loans. In addition. a tree analysis of your diet (a $25.00 vaue) is available for you if you check the
boxonthesheetof owtofilloutthefoodreoordlorm'.

If you have any questions. please contact Collin Shih at 355-1010 or via e-mail at
shihjenh@pilotmsu.edu or Sharon Hoerr RD. PhD. at 355-7701 or 20533sgh@ibm.msu.edu. this
study is part of the MSU Healthy U Progmm. Your participation is appreciated as we work to improve
the program. _ . '

if you agree to participate in this research. please sign your name on the line

 

 

(Name) ' , (Dare)

Note: Please return this form with your completed instruments. Thanks!

Appendix D

November 18, 1996

Dear MSU CTU:

Two weeks ago, we mailed you a set of questionnaires about your dietary fiber intake. We
are still looking forward to your response. Enclosed is another set of the questionnaires in case
you cannot find the originals. This will be our last contact with you, so please take a few minutes
to help us acquire valuable perspectives from you to design health promotion programs. Thanks
for your cooperation ! if you do not plan to participate, please return the blank questionnaires to us
and mark 'Regrets' on the top.

Sincerely

Sharon Hoerr, RD., PhD. ~
Collin Shih

Department of Food Science and Human Nutrition
Michigan State University

Appendix E

 

91

Check your dietary fiber intake

Please writein the number of servings you had of following foods over last month.
If you ate less than once a week, leave blank.

 

 

 

 

 

 

Dari} Tintes‘uk Daily Tirnes/uk

Fruits: (l serv=l/'2 cup/ 1 piece) Breads, min and pasta:

apple - ( ) ( ) (l serv-l/Z cup/ 1 slice)

banana ( ) l ) brown rice ( ) ( )

berries r ) r ) noodles ( ) ( )

grapes ( l ( ) pasta or Spaghetti ( ) ( )

melon ( i r ) white bread ( ) ( )

orange ( ) ( ) whole-wheat bread ( ) ( )

peach/nectarine ( ) l ) ether ( ) ( )

pear ( l ( )

prunes/plum ( ) l. ) Nuts: (1 serv=l/4 cup or 4 Tbsp)

raisins ( ) l ) PCMU! butter ( ) ( )

ether ( ) ( ) PC3005 ( ) ( )
walnuts ( l ( )

Vegetables, cooked or raw: ether ( ) ( )

(l serv=ll2 cup)

broccoli l ) ( ) Snack foods:

Brussels sprouts ( ) ( ) bagel (1 whole) ( ) ( )

carrOts ( l ( ) chips l 10 chips or 1 oz) ( ) ( )

cauliflower ( l l l crackers (4—6 pc) ( ) ( )

celery ( l t ) pizzal 1 slice of 16") ( ) ( )

corn r ) l ) pepcomll cupS) ( ) l )

French fries (' ) ( l tortillas“ whole) ( ) ( )

green beans ( ) ( ) cther ( ) ( )

lettuce ( ) l )

pctato ( .l ( ) Cereals: ( l serv=l cup)

spinach l l l ) bran cereals ( ) ( )

summer squashlzucchinil l ) ( ) granola ( ) ( )

tomato l l l l oatmeal ( ) ( )

tossed salad r l l l \\ hole grain cereals ( l ( )

ether l ) l ) cther ( ) ( )

 

 

Beans. legumes: l l sen =1 2 cupl
baked beans

l l l l
green peas I l l ' l
kidney beans r l ( l
lima beans l l l )
retried beans I l l l
ether l l ( )

 

Dent Ftlfld Sucrrtc .2 Human Nutrrtn-n ‘.l.\'i.' I‘Nh

 

 

Appendix F

92

Find the direction of your dietary fiber intake

F iber-riclr foods mean fruits, vegetables, beans, nuts, and hi in-fiber cereals.
Tomato sauce, refried beans, and salsa are also included. Follow the direction of

the arrows with your response, and W.

 

Have you ever tried to increase the amount N 0
‘C I of fiber-rich foods in your diet? OR ".
YES O Are you already on high-fiber diet? .‘
1

O
i V
In the past month. have you thought about. .

Ge you currently doing it? > IIIIII’ doing it?

 

 

 

 

 

I No I
l ' '
I I
I l :
yrs I YES : a
I I
I I
V V '
I
E
. How confident are you that you I
GOV” long have you done "7 > will start this change during the I No

 

next month?

 

 

 

 

I
I
I
I
I
I
I
I
I

Very or . I

somewhat M‘ldly 0‘ not at V

More than 6 m°mh5 < Less than 6 months) confide,“ all confident

Dept Food Scrence a: Human Nutrition
Michigan State Universrty. Iwn

 

Appendix G

93

What do you think about fiber-rich foods ?

Aga‘n, fiber-rich foods mean Iruits. vegetables. nuts. beans. whole grain breads.
and high-fiber cereah
Stroneg Disagree Stroneg Agree

I. Eating a let of fiber-rich foods decreases my chances of l 2 3 4 5

getting serious diseases such as cancer and heart disease.

2.‘What I eat is important for my health. I 2 3 4 S
3. Fiber-rich foods taste good. I 2 3 4 S
4. I ate a lot of fiber-rich foods when I was growing up. I .. 3 4 5
5. I've gotten a lot of advice about how to eat fiber-rich foods. I 2 3 4 5
6. My friends encourage me to eat fiber-rich foods. 1 2 3 4 S
7. My family encourages me to eat fiber-rich foods. I .. 3 4 5
8. It is easy to select fiber-rich foods when I eat out. I 2 3 4 S
9. Eating fiber-rich foods is important. I 2 3 4 5

10. I‘m willing to try a new food if it is fiber-rich. I 7 3 4 5
l I. It is hard to increase my intake of fiber-rich foods. I 2 3 4 5
I2. I can increase intake of fiber-rich foods in next 6 months. I 2 3 4 5
13. I can be healthy, if I eat fiber-rich foods. I 2 3 4 5

Your feedback to following questions will help Healthy U improve its program.

I What could you do to increase your intake of dietary fiber?

2. What keeps you from eating more fiber-rich foods?

Wires you are done with this pate. please told. staple and put It ln campus mall.
tiles please complete the two day: toorl records.

Dept I-u-rd Science k Human Nutrition
Michigan State Uruversm. I'M.

"illllllllllllllllllls