id‘s. 11v. .1 1! .1... .1 2...». _. 1...”. 111...”: .........”m.... 1.1.1.1.... 111W 1w1/11.1.1u?£:..11 .1 .. 11.. ........,......;1...:.1....1.....1”..1..1..1.1...~1.._1.1....:/”1.wm...1 .115... . 1..., .Ia; 1 11'»: o 11'1aJ1 1 1; . 1:61 . 1.1 J. 1.;«111. 1r... 1. 1.11 1.1.3? .1 11/111171. 1.11. , 1 1.. c .1 1 1.11.1.5 1 111.: 1/1 94. @13- ST. an AL m w ——ru--“-‘-w-v“‘ . . 1 . . 1 . .11 . . . 1 . . . r . . . r ..1 . . . . . . . . .v 1 1.. 1' 0 <...,..1....-.11.n . , .I . ., 1 1 .1. . , 1 1 1 . 1 . v .1 5v..~.:1.l.. It .12, 1 cl... 11 . I v1.1.1.1: -n.....11l.111.v)1§...v v1”. ..,, 1....“ .121, 11.11:: ' l i {p.yl -‘1 I \ll"\, I ' *’I'|||‘ v e 7 LIBRA R 12 Michigan State WIT ‘ST‘AT N‘IVE ll LTfl ET ‘ mvcmty MN.“ ‘I “ml “Ml lllllllll ‘ 3 1293 01085 9308 This is to certify that the thesis entitled Body Weight Extremes in Adult Females: An Analysis of Research Procedures to Study Health Related Correlates and Precursors presented by Ruth Evelyn Dennis has been accepted towards fulfillment of the requirements for _PhL___ degree in Mience fidlml' Lw'u Major professor Date August 10, 1973 0-7639 DE 8‘ aazo 093 J M AGIC 2 gii“ .N§%fig _U1 99 BC All ARALY ABSTRACT BODY WEIGHT EXTREMES IN ADULT FEMALES: AN ANALYSIS OF RESEARCH PROCEDURES TO STUDY HEALTH RELATED CORRELATES AND PRECURSORS By Ruth Evelyn Dennis Important to the understanding and prevention and/ or the physical condition of extremes in human body-weight, are the variables precursing and correlating with this condi- tion. The purpose of this study was to explore and explain alternative models and protocols of research which would pro- vide empirical evidence in resolution of cause/effect questions arising from observed relationships between obesity and chronic disease. Directionally the study points to limita- tions and potentials of previous studies, demonstrates use of the sample survey, and suggests models of broader potential for continued study of such physiological conditions developing over a period of time. I Traditional studies of bodily weight have been mainly concerned with obesity, implicating the latter as a cause of various chronic diseases. The validity of these findings and conclusions is here challenged for two reasons: is unrepre and the (ii: in bodily l fissible m- mterpreta afthe tra I. S This izediate 3fpossib1 .3’Cint-in-t Between Va A derr Sented. I body Weigh health and the adVant wider use PCPUIatiOr II. The I identify 1 and Chroni Health inc wereases The C Ruth Evelyn Dennis the unrepresentative nature of the populations considered, and the disregard of the developmental process of extremity in bodily weight. Three research designs are presented as possible means of overcoming these limitations in design and interpretation. As conceptual models, they are adaptations of the traditional modes of research. I. Sample-Survey of Cross-Sectional Groups This model was posed as one of the most accurate and immediate means of selecting a population broadly indicative of possible variables. The method is amenable to testing one- point-in-time hypotheses questioning existing relationships between variables. A demonstration of the use of the sample survey is pre- sented. The research of the design investigated levels of body weight as related to a variety of personal habits, health and chronic disease in adult females. It indicates the advantages of this design over previous designs including wider use of variables applicable to only generalized total population. II. Longitudinal Study of Communities or Neighborhoods The protocol of research in this design proposed to identify limited areas of high-risk in body weight extremes and chronic disease by observing frequencies of negative health indicators (e.g. squalor, congestion, crime) and also increases in physical deterioration and social disorganization, The design further proposed to provide opportunity for abservati. cmrelate neighbor‘r. III. 'rlS :hronic c .) r‘f‘ U the de he indix Ruth Evelyn Dennis observation of the development of common precursors and correlate of body weight extremes and chronic disease at the neighborhood level. III. Longitudinal Study of Individuals This study similar to those undertaken to investigate chronic disease development was suggested for consideration of the developmental processes of body weight extremes in the individual. It was noted that the advantage of this design is the study of a population before observable mani- festation of pathology, host and environmental factors early in life (precursors) then relatable to subsequent development. The method is notable as the only one that improves the accuracy of cause/effect prognosis over a period of time. In addition to the observation of precursors and corre- lates at the neighborhood level, the research protocol here proposes observation of precursors and correlates at the personal level, especially the effects of stress and depri— vation on individual health over a period of time. The emphasis then is on measurement of change. BODY HEIGHT EXTREMES IN ADULT FEMALES: AN ANALYSIS OF RESEARCH PROCEDURES TO STUDY HEALTH RELATED CORRELATES AND PRECURSORS By Ruth Evelyn Dennis A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Social Science 1973 Q? DEDICATION This thesis is dedicated to my beloved parents, John and Pennie Gaines, whose continued support of love and understanding has helped to sustain me. I am assistance To th firna Barba contribute To Dr ETateful f graEEful t its COuUSe I WOu iOctm’al c (I) 3: OIOEY; ‘ Harry valuable s I am if public also to tl‘. passible t LIEUCa t iOn 13m Eel ACKNOWLEDGEMENTS I am profoundly grateful to many people for their assistance and contributions during the course of this work. To three very significant persons, Mrs. Rowena Grey, Dr. Anna Barbara Grey and Miss Marry Thomas, whose love and support contributed greatly to my reaching this point. Thank you. To Dr. Peter Manning, my former major professor, I am grateful for his counseling and assistance. I am especially grateful to Dr. Robert Lewis, my present major professor for his counseling, patience, encouragement and guidance. I would also like to thank the other members of my doctoral committee, Dr. J. Allan Beegle, Professor of Sociology; Dr. John M. Hunter, Professor of Geography; and Dr. Harry M. Raulet, Professor of Anthropology for their valuable suggestions and criticisms of this work. I am indebted to many persons at the Michigan Department of Public Health for their technical and financial support. Also to the Michigan Department of Public Health who made possible the grant assistance from the Department of Health, Education and Welfare (HEW). I am particularly grateful to Ms. Jean Burl and Ms. Dawn Thelen for the typing, and Ms. Marilyn Jessers and Ms. Mary Davis for the editing, as well as many other friends for their moral and material support. ii My greatest thanks goes to my children for their love, patience, and understanding during the years of my doctoral study. TABLE OF CONTENTS LIST OF TABLES ............................. LIST OF FIGURES ............................ Chapter I. II. III. IV. INTRODUCTION ......................... The Problem ......................... Significance ........................ Purpose ........................... Objectives ......................... TRADITIONAL STUDIES OF BODY - WEIGHT ............. Introduction ........................ The Review of the Literature ................ Health Influences ...................... Association with Diabetes .................. Arthritis .......................... Hypertension ........................ Environment ......................... Spatial ASpects of Obesity ................. Socio-Cultural Influences .................. Socio-Economic Differences ................. Limitations ......................... THEORY AND HYPOTHESES ..................... Theory ........................... Hypothesis ......................... Hypothetical Models ..................... Interpretation of Models and Hypothesis ........... ECOLOGICAL APPROACH TO STUDY OF BODY WEIGHT EXTREMES ..... Introduction ........................ The Survey ......................... Hypothesis ......................... A Community-Based Longitudinal Study (Model 11) ....... A Brief Suggestive Protocol for a Community Based Longitudinal Study .................... iv Page vi Page Selecting the Population for Study ............ 37 Prospective Individual Study ............... 42 Suggestive Protocol for a Longitudinal Study of Body Weight Extremes .................. 43 Selecting the Population for Study ............ 44 Relevant Variables .................... 45 V. A DEMONSTRATION OF THE SURVEY APPROACH TO STUDYING BODY HEIGHT EXTREMES ................... 48 Source of Data ..................... . 48 Method .......................... 49 Limitations of Design and Interpretation ......... 53 Results ......................... 55 Discussion ........................ 91 Summary ......................... 94 VI. CONCLUSIONS AND RECOMMENDATIONS .............. 99 Conclusions ....................... 99 Recommendations ..................... 102 BIBLIOGRAPHY ............................ lO4 APPENDICES Appendix A - Tables of Responses ................ lll Appendix B - The Metropolitan Life Insurance Company's Height-Height Table ................ 148 Appendix C - Michigan Health Survey .............. I49 EH9 l. Descriptic 2. QdCldl C0? Race wi 3- Marital S 4' OCCUpatio 5- EOUCdtlor 5' Females l \‘ Females FEITlal e g Femal ES Table 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. LIST OF TABLES Description of Sample by Age and Height ............ Racial Composition of Sample by Age and Height, Ratio by Race within Age ...................... Marital Status by Age and Height Occupational Status Educational Status by Age and Weight Females Females Females Females Females Females Females Females Females Females Females Females Females Females Females Females Females Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting Reporting 000000000000000 by Age and Weight Seen Doctor Due to Illness 000000000 Seen Doctor Due to Illness 0-12 Grade . . . . Seen a Doctor Due to Illness l3+ Grade Overnight Hospital Stay ........... Overnight Hospital Stay 0-12 Grade Overnight Hospital Stay l3+ Grade . . Trouble with High Blood Pressure ...... Trouble with High Blood Pressure 0-12 Grade . Trouble with High Blood Pressure 13+ Grade Trouble with Emphysema ........... Trouble with Emphysema O-12 Grade ...... Trouble with Emphysema l3+ Grade Trouble with Diabetes ............ Trouble with Diabetes 0-12 Grade ...... Trouble with Diabetes 13+ Grade ....... Trouble with Kidney or Bladder Disease Trouble with Kidney or Bladder Disease 0-12 Grade ............................ vi Page 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 . zl r\, 4-- t...- ()1 Females P l3+ Bra Females P Females R . Fenlesl Females F Females F Females F Famlesl Females l Femalegl Females Table 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Females Reporting Trouble with Kidney or Bladder Disease 13+ Grade ......................... Females Reporting Trouble with Heart Disease . . . ..... Females Reporting Trouble with Heart Disease O-l2 Grade . . . Females Reporting Trouble with Heart Disease 13+ Grade Females Reporting Regular Cigarette Smoking ..... Females Reporting Regular Cigarette Smoking O-l2 Grade Females Reporting Regular Cigarette Smoking l3+ Grade . . . . Females Reporting General Medical or Physical Exam ..... Females Reporting General Medical or Physical Exam O-l2 Grade Females Reporting General Medical or Physical Exam l3+ Grade Females Reporting Taking Vitamins Regularly ......... Females Reporting Taking Vitamins Regularly O-l2 Grade Females Reporting Taking Vitamins Regularly l3+ Grade . . . . Females Reporting the Regular Use of Tranquilizers ..... The Ranking of Obese Females with Specified Negative Health Indicators, Grand Rapids, Michigan ............ vii Page 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 "HIE ()1 o 0‘ o \‘J a Cu 0 . Hypothet Height L Accumula Pyaothet Height L Constant Nested a Common 1 Disease Alternal and 015! Figure 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. LIST OF FIGURES Hypothetical Models Indicating Body Weight by Age and Weight Level in Case of Unspecified Diseases - Accumulative Process .................... Hypothetical Models Indicating Body Weight by Age and Weight Level in Cases of Unspecified Disease - Constant Process .................... Nested Hypotheses Leading to the Identification of Common Precursors to Body Weight Extremes and Chronic Disease .......................... Alternative Models for Studying Body Weight Extremes and Disease Conditions ................... Sample Description - Weight by Age ............. Marital Status by Age ................... Percent of Females by Education Level for Selected Michigan Cities ...................... Percent “Seen a Doctor Because of Illness or Medical Problems" ......................... Percent "Overnight Hospitalization During the Past Year" . . Percent "Hypertension During the Past Year" ........ Percent " Emphysema During the Past Year" ...... Percent "Diabetes During the Past Year" .......... Percent "Kidney or Bladder Trouble During the Past Year" Percent "Heart Trouble During the Past Year" ....... Percent "Smoking Cigarettes Daily" ............ Percent "Medical or Physical Examination" ......... Percent "Vitamins Daily" .................. Percent "Drugs for Medical and Non-Medical Reasons Percent of Underweight Adult Females . . . .1 ........ viii Page 25 26 23 31 56 58 60 61 64 66 68 70 72 74 76 78 79 81 82 Figure 20. 21. 22. 23. 24. 25. 26. Page Percent of Normal Weight Adult Females ....... . . . . 83 Percent of Overweight Adult Females ............ 84 Percent of Obese Adult Females ............... 85 Percent of Families with Income Below Poverty Level . . . . 87 Illegitimate Births by Place of Residence per 100,000 Persons as Reported by the l970 Federal Census ....... 88 Attempted Suicide Rate/1000 ................ 89 Percent Deteriorated Structures .............. 9O ix Rese :ainzent Prevalenc Of Preve: 3: reseai Stldies ‘ design a: 920gfeSS advanCeS H 3391118 s CHAPTER I INTRODUCTION Research in community health aims at systematic ascer- tainment of the extent and distribution of disease and related problems and investigating the causes of fluctuation in the prevalence of such conditions (Shattuck, 1850) for purposes of prevention and cure. The success of such study is dependent on research design which reflects integration of prior relevant studies with the study-at-hand. By avoiding past problems in design and analysis, a study becomes part of constructive progress in its field and adds to the viable base of future advances. Thus it is the intention of this study of Body-Weight Extremes in Adult Females to emphasize evaluation of past re- search designs, and to suggest alternative models useful in determining precursors and developmental correlates of body- weight extremes and disease. Traditionally, studies of body-weight have been concerned mainly with obesity and have been of limited usefulness to subsequent investigation of development and variations in body- weight extremes. Their research-design and data-interpreta- tion indicate a narrowed focus on consequence rather than direction to the broader reaches of etiology. Dealing with special groups and the effects of obesity on chronic disease within these groups, their underlying assumption seems to be that some 1 52311 or n: suggest th. Body-Weigh This sective, reactions or gradual 5113’. 38311 longituain fC‘ir appar in terms C Alter 2 that some chronic diseases are the result of obesity. The small or non-existent gains resulting from their conclusions suggest that present knowledge of prevention and cure of Body-Weight Extremes is incomplete and perhaps distorted. This study, geared to etiology and longitudinal per- spective, is an exercise designed to reveal the rhythm of reactions such as chronic disease, which although delayed or gradual in emergence, has definite precursors. Conceptu- ally, assuming a chain of events, circumstances and conditions longitudinally operative and within discernible risk fields, four apparent etiological alternatives may be differentiated in terms of body-weight extremes and disease conditions: Alternative I: Body-weight extremes and disease conditions are both related to similar precursors, but not directly related to each other. Alternative II: Body-weight extremes and disease condi- tions related only after coexistence. AlternativeIII: Body weight extremes, following a set of precursors, become themselves precursors of a new pathogenic condition. Alternative IV: Conversely, disease conditions following a set of precursors, become precursors of body-weight extremes. The Problems The standard bivariate analysis applied to a set of selected individuals can show and has shown statistical relationships between body-weight extremes and disease conditions. However, because of the nature of the method, and because of the restriction of data availability, these analyses '51 :he follow f\) m (I) I. IfA FY’ t“'rezes 3 analyses have not developed evidence answerable to any of the following: 1. Extreme body weight is a precursor of disease condition. 2. Extreme body weight is not a precursor but an element interactive with disease conditions only after both coexist. 3. Both disease and body weight depend upon the same precursors, no interaction occurring between them. Significance If Alternative I (A-I) pertains, i.e. both body weight extremes (BWE) and disease conditions are related to similar precursors but not to each other, prevention of both is a function of efforts directed toward precursors. If A-II pertains, i.e. coexistence determines their relationship, as is suggested in the literature, then treatment of either BWE, or disease conditions, or both, may influence the course of one, both, or neither. If A-III pertains, prevention of BWE would effect pre— vention of disease conditions; also, successful treatment of BWE would modify the degree of disease conditions. If A-IV pertains, then attention is demanded by the etiological complex resulting in disease, i.e. the host's resistance to susceptibility, the disease-producing agent, and the nature of the environment in which host and agent are joined, and the cumulative effects of these three elements ‘When the dimension of time is added. The ' weight am extreme f confined iisease m If t 05 obesit disease d iescribec latecl to the addii -. ”llr'xa ‘5‘ l; The literature's suggestion that coexistence of body weight and disease conditions increases the risk of more extreme forms of the disease or weight implies that treatment confined to changing the weight level offers only temporary disease modification, not preventive consideration. If the literature's assumption is not true, treatment of obesity may enhance a disease condition. For example, a disease developing because of conditions of stress as described by Dodge and Martin (1969) and also causally re- lated to extremes in body weight, may be a further risk by the additional stress of altering the body weight. Purpose It is observed that current views of obesity and chronic disease call obesity causative of chronic disease (according to a number of studies, an implied definitive "relationship"). The focus of the methods and the restriction on data availa- bility have characteristically stopped at "showing a relation- ship" but not indicating the nature of the relationship. The purpose of this study is to explore and explain alternative modes and protocols of research which would pro- vide empirical evidence of cause/effect relationships arising out of observed relationships between obesity and chronic disease. These models will encourage insights into the developmental processes between chronic diseases (CD) and (BWE), indicating methods of prevention, and also clarifying the effecti treatmen E . Driectives The 0‘ l. T previOUS, to weigh t sations in SilOl’chmi 2. I rFlathIis‘: 3. '1 Cation Of 5 the effectiveness and appropriateness of present approaches to treatment . Objectives The objectives are: l. The review of positive and negative aspects of previous, present and alternative research designs in order to weigh the values of available methods, so that investi- gations in this area may proceed with full awareness of short-comings and potential in a particular design. 2. The exploration of models of the developmental relationship between BWE, CD and related factors. 3. The development of models to facilitate identifi- cation of precursors to BWE and CD. CHAPTER II TRADITIONAL STUDIES OF BODY-WEIGHT Introduction Primarily because of the convenient accessibility of data, special-group studies have provided practically all of the information currently available on obesity. Subjects already grouped as hospital patients, insurance-policy holders, etc., are usually found convenient to work with, or their generally available records have provided requisite data. In addition to aiding in the identification of factors associated with obesity and the variability of these factors, such groups offer research a relatively inexpensive mainte- nance of large populations and expanded dimension. The Review of the Literature A review of the literature reveals that a vast majority of past research concerning BWE and its correlates is retro- spective but focused on one point in time. The multiplex of variables already indicated as influ- ential make it possible to classify the literature under three headings: influences pertaining to human health, physical environment and socio-cultural environment. 6 ‘1, ‘~ Health Influe Althoug? health of ar. mdeteminec weight - eii certain typ his reason rates among 31. 1961) Much . 1531‘);ng b0 StatistiCS IES‘jranCe for aPPrm “Eight eXl SO SeleCt bias Stud: One Health Influences Although the influence of body weight on the physical health of an individual or a group of individuals is largely undetermined, it is generally agreed that extremes in body weight - either obesity or underweight - increase the risk of certain types of diseases or undesirable health conditions. This reasoning is based on the higher mortality and morbidity rates among the groups with body weight extremes (Kannel, et a1, 1967) (Mark, 1960) (Mayer, 1967). Much of the data concerning the medical significance of varying body weights has come from insurance studies based on statistics of such companies as the New York Metropolitan Life Insurance Company. Although considered a significant source for approximating the mortality and morbidity risk of body weight extremes (Metropolitan Life Insurance Company, 1960), so select a population is considered by some critics to bias study results. One such long-term study, the Body Build and Blood Pressure study of the Society of Actuaries (the body weight standards are used in the present study) showed that among five million insured persons, the mortality rate in persons 15 to 69 years of age was one-third greater in those 20 percent or more overweight than in those who weighted less. Overweight women showed a lower mortality rate than over- weight men, although mortality in women also increased with weight and age. Despit seem more 1 :ory disea: weight cla: The l associatin advanced E that a nu: Specific i pected, s: Z'Edical If that in El etiologic and hYPer Bery notes the aSSOciate and Logan oxygen de f 1 8 Despite the paucity of data on underweight persons, they seem.more vulnerable to death by tuberculosis, acute respira- tory disease, and hypertension than do people not in their weight class. (U. S. Public Health Service, 1966). The literature reveals that clinical observations associating obesity with heart attacks and strokes have been advanced for more than 2,000 years. Wilson and Wilson noted that a number of classical Greek and Roman sources record specific instances of such association as if they were ex- pected, such grouped references were commonplace in early medical literature (Wilson, 1969). The Wilsons further stated that in the 1940's obesity was considered the most common etiologic agent in arteriosclerotic cardiovascular disease and hypertension. Berylne supports Wilson in the latter contention, and notes that the syndrome of cardio-respiratory failure associated with extreme obesity was first recognized by Kerr and Logan in 1936. He lists obesity, cyanosis, arterial oxygen desaturation and polycythemia as essential features of the cardio-respiratory syndrome (Berylne, 1958). The Framingham study, conducted in a small American community of 5,127 men and women who had been followed for a period of more than 12 years for signs of intial develop- ment of coronary heart disease (CHD), is widely considered to be the most extensive epidemiological study as yet under- taken. In the 12 years, 252 men and 128 women developed CHD. There was a and women at wdght and 2 Not al logical fac hiy1incide fleet how I the use of Of coronar‘ 9 There was a marked increase in mortality rates in both men and women more than 20 percent overweight over those of normal weight and a lesser degree of overweight (Kannel, et a1, 1967). Not all researchers agree that obesity is a major etio- logical factor in heart disease. Hinkle suggests that the high incidence of obesity in coronary heart disease may re- flect how the study population is selected. He notes that the use of the interview to conduct an epidemiological study of coronary heart disease is not an efficient way to deter- mine the incidence of the disease, since the examination of those who are detected during surveys is based on electro- cardiographic abnormalities. In many cases, the subject had not been aware that these were present, just as a proportion of all myocardial infarctions occurs in people without pre- vious evidence detectable by any means as coronary heart disease. It is actually likely that a central group selected on the basis of an interview will contain many people who have coronary heart disease. Also, persons of higher socio-economic status are more likely to be under medical care and to know if they have a coronary heart condition (Hinkle, 1968). Wardwell and others analyzed 87 white males who had coronary heart disease, and concluded that "men of middle class protestant background have the highest artios of ob- served to expected cases of coronary heart disease even when other sociological and selected physiological variables (hypertension, obesity, smoking, and diet) are controlled (Wardwell, 1964). Association 1 The ass cussed in ma inprolongec :derance o< obese sub je. carbohydrat Craddc 1 It ac 2. :4. w‘ o TN'OO 10 Association with Diabetes The association of diabetes with obesity has been dis— cussed in many studies, there being general agreement that in prolonged obesity a gradual deterioration in carbohydrate tolerance occurs. The results of Ogilvie's study of 11 obese subjects for 18 years showed that all had a diminished carbohydrate tolerance, three of them becoming acutely dia- betic (Ogilvie, 1934). Craddock (1970) distinguishes two types of diabetes: 1. Juvenile diabetes present in children and young adults, is due to a complete lack of insulin. 2. Maturity-onset diabetes present in older adults who are usually obese, and due to a relative lack of insulin or a graduating insensitivity to insulin. Two other trends of thought on the association of obesity and diabetes are that obesity appears to lead to insulin re- sistance which, after many years, brings about the diabetic state (Karan, 1965); and, the diabetic tendency manifests itself first as obesity (Medley, 1965). Arthritis Leavell and Clark (1953) say of the term: Arthritis is applied to the heterogeneous group of conditions which have commonly only the fact that they represent the end products of the actions of various combinations of factors involving agent, host, and environment acting upon the same organ system - the joints of the body. The term rheumatism is also applied to arthritis, Smillie and Kilbourne dividing the condition into four classes: 1. Rheumatical Arthritis (atrophic) 3 Gou 4 In! Osteoa in older pe on the weig posture, i in the dis H‘Perterts: x 11 2. Osteoarthritis (hypertrophic) 3. Gout (metabolic arthritic) 4. Infectious (arthritic) Osteoarthritis, a degenerative disease occurring usually in older persons, is known as a disease of wear and tear” on the weight-bearing joints. Obesity, with resultant faulty posture, is thought to be an important contributing factor in the disease (Smillie and Kilbourne, 1963) (Radin, 1972). Hypertension As in some of the above pathological conditions, the association of body weight to hypertension is still under debate. It has been held that obesity is the most common etiologic agent in hypertension (Wilson, 1960). Hypertension (a sustained increase in blood pressure) and obesity, diabetes and gall bladder diseases, are important coexisting conditions which may have a common denominator in abnormal lipid metabo- lism, notably cholesterol (Waterman, 1955). The "Nutritions Review" notes that obesity as a risk factor in coronary arterial disease has generally been rele- gated to a level of minor importance. The Framingham Study suggests that, apart from associated Hypertension or hyper- cholesteremia, overweight per se has weak, if any, influence on the development of new atherosclerotic manisfestations. They further report that isolated obesity has a significant effect upon the development of atherosclerosis in men, but not in women (Nutrition Reviews, 1967). Chains a between 0V9” excess weight and morbidit‘) apertensim socio-econox‘. in this grou general COEI factors cont interrelat'u Swironrrent Extrem are thought obesity is attributed are of com '39 engende: and w The p CODSISts aCEIVity’ Obesi 12 Chaing and Perlman (1971) reviewed the relationship between overweight and hypertension, their study revealing excess weight as substantially contributive to the mortality and morbidity of the hypertensive diseased. Among blacks hypertension is higher than among whites, regardless of socio-economic status. It is not clear whether or not BWE in this group is similarly high, but it is clear that the general confounding of hypertension and obesity with other factors contributes to the present confusion concerning their interrelationship. Environment Extreme overweight and its opposite, extreme underweight, are thought to be symptoms of individual human stress. While obesity is attributed to overeating and extreme underweight is attributed to undereating, it is the reasons for both that are of concern to scientists. These reasons are believed to be engendered by the multiple factors of heredity, constitution and environment. The physical environment, as it relates to body weight, consists of the availability of food, the amount of physical activity, and the amount of stress in the immediate surroundings. Obesity is commonly thought to be a problem of affluent societies whose intake of sweets (carbohydrates) is high, and whose amount of exercise is generally reduced, while under- weight is characteristic of the less-developed areas where seasonal hunger is evident. However, in the United States, obesity is most prevalent among the poor. 13 It has been observed that extremes in body weight and associated disease disorders have had high positive corre- lations with stress-producing environments. Dodge and Martin show in their study that while body weight is to a large ex- tent genetically determined, it is more susceptible to environmental influences than are other measurements (Dodge, 1970). This suggests thatthese external variables are pre- disposing, precipitating, or causally related to weight extremes. Much of this influence is attributable to the psychological and socio-cultural environment to be discussed later in this paper. Lack of physical activity has been found to be an important factor in increasing body weight, a particular pro- blem in the United States where there is an increasing use of the automobile and home appliances. Lincoln (1960) found that decreased physical activity was more important to increasing body weight than caloric intake. Using controls, he studied 28 obese girls and 14 obese adolescent boys and discovered that the obese group ate less than the non-obese group. This type of study was repeated in 1961 by Rose, and in 1966 by three separate studies conducted by Heumann, McCarthy, and Maxfield. Further studies regarding physical activity conclude that: l. Obese girls were significantly less active than controls (Johnson, et a1, 1956). l4 2. Obese boys were apt to expend less energy than non-obese boys during participation in exercise activities (Stefanik, et a1, 1959). 3. Individuals with thelighest percentage of body fat spent most of their time in less strenuous activities than those with lower percentages of body fat (Hutson, et al, 1965). Spatial Aspects of Obesity The relationship between social class and obesity also suggests a distinct spatial pattern, since lower socio- economic populations form geographic patterns that are identifiable. The idea that overweight women form a distinct geo- graphic pattern, especially within an urban area, is sup- ported by a study of 1,503 females 18 years and older in Flint, Michigan (Dennis, 1970). Of these women, 1,165 (78 percent)were white, and 338 (22 percent) were non-white. Thirty-six percent of the females were overweight. Fifty- five percent of the black females and 30 percent of the white females were overweight; about 3 percent of black and white females were underweight. The study showed that the higher percentage of overweights (both non-white and white females) were located in virtually the same neighbor— hood, areas rated low on the socio-economic scale. Some underweight females were reported in nearly all neighborhoods, but no spatial or socio-economic patterns of distribution emerged. The data for this study were taken from the Michigan Health Survey. The: aspects 0 Essever, aatic bou and avail anvil-0mm Pmspers Whenever It c 9- «ea LUre 1 . 946 an: 1. 2. 15 There appesrs to be little, if any, data on the spatial aspects of body weights within a limited geographical area. However, there exist differences in body build across cli- matic boundaries, because of physical adaptation to weather and available food (Hunter, 1969). Socio-Cultural Influences Social scientists generally agree that the psychological aspect of an individual's inner environment is the sum total of his life experiences and upbringing, while the outer environment is the culture and the society in which he prospers or fails, and that psychological stress results whenever the inner environment is in conflict with the surrounding world in terms of tasks and expectations. It has been noted by some researchers that obesity is a feature of several psychological disorders. Richardson in 1946 and Mendelson in 1964 note that obese subjects manifested: l. A high level of anxiety; 'over-compensation' (retreat behind a wall of obesity). 2. A lack of emotional satisfaction ('regression to the infantile state'). 3. An increased drive for oral satisfaction; a defense against depression; a true addiction to food; a sexual conflict situation. Silvertone maintains that if these psychological disturbances were of primary etiological importance in obesity, one would expect obese patients to show much more general psychological disturbance than do non—obese subjects. Moore, et a1. investi- gated the relationship between obesity and mental health in a random 55 110,000 i ".hese res intervie'. The health we ) p . h ononic r response did thos (imatur StatiSti lacing o 16 random sample of 1,660 persons selected as representative of 110,000 inhabitants in a residential area of New York City. These residents were subjected to standard psychological interviews. The relationship of obesity to nine measures of mental health was investigated, holding constant age and socio- economic variables. The obese persons made more pathological responses (or scored lower on mental health measures) than did those of normal weight; and for three of the measures (immaturity, suspiciousness, and rigidity), the results were statistically significant. The author notes that in corre- lating obesity with a variety of other factors, correlations alone cannot tell us which, if any, of the factors are primary, or indicate whether mental health factors cause obesity or are the results of being obese in a society that devalues obesity (Moore, et a1. 1962). The most outstanding result of Moore’s study was the striking relationship between social class and obesity. Obesity was found to be seven times more frequent among women of the lowest socio-economic level than among those of the highest level. Also in keeping with other studies, the prevalence of obesity was found to increase with increasing age. Socio-Economic Differences Maddox, et a1. (1970) suggested that one of the socio- economic reasons for differences in body weight might be differing gatients a center, t‘r deviance 1 In 0' the ‘ self perc quen Phil lacks 5e: (even by the heal with 10m which he emphaSiz attitude Or SU‘DC‘ StudieS l. l7 differing attitudes toward obesity. In a study of 100 new patients at a public out-patient medical clinic of a medical center, they tested the assumption of overweight as social deviance and disability. The authors note: In our society, overweight figures significantly in the perception of others and in the conception of self; there are many indications that overweight is perceived as social deviancy and, partly, in conse- quence of this imputation, is a social deviancy. While an overweight person is looked upon as one who lacks self-control and is responsible for his unsightly state (even by physicians and middle class individuals) despite the health implications (Keys, 1955), leanness is associated with longevity, good personal appearance, and self-denial which has historically been suffused with a Protestant ethic emphasizing appropriate rewards, including good health. This attitude toward obesity is not shared by all ethnic groups or subcultures even within our own society. Maddox, et a1 studies the difference in attitudes of four groups: 1. White males, 60 years of age and over, with at least an eighth grade education, living in a small town or rural area in the west appeared to be most satisfied with or least interested in their weight. 2. White females 40-60 years of age, with a college education, living in cities of a million or over in the east appeared to be most unhappy with their size, wanting to be smaller, even when they were not likely to report much success in reducing. 3. Negro males 20-40 years of age, with college edu- cation, living in cities of a million or over in the west were more likely than any other group to want more weight. \lit' economic essentie to be cl the fin Child 1' likable de3101151 expect; valLles 'I to Ref 18 4. Negro females, with at least a high school edu- cation or more, living in cities of a million or over in the mid-west appear unconcerned about weight reduction. With this cross-section of geographic areas and socio- economic levels, their data overwhelmingly supported the essentially negative evaluation of overweight hypothesized to be characteristic of our society. This study supports the findings of Richardson and his associates that the obese child ranked consistently (with few exceptions) the least likable. Also the negative evaluation of obesity that was demonstrated by Maddox, et al. was consistent with the expectations based on impressionistic assessments of cultural values (Maddox, 1970). The study along with others suggest that women of higher socio-economic status are under greater social pressure to maintain normal weight than others. In these studies, age, sex and socio-economdc status were found to be factors affecting body weight. I In addition to the above factors or interacting with them, is that of individual eating behavior. Studies have shown that the eating behavior of obese persons is distinctly different from that of the non-obese, regardless of socio- economic status. Bruch did a study that suggested that obese people may not be able to distinguish physiological hunger from such states as fear, anger, and anxiety (H. Bruch, 1961). S. Schochter (1968) showed that unlike normal subjects, obese sub He also 5 intake we ‘-4 jects. examined, normal 01 good by a good, fa subjects 19 obese subjects' urge to eat was unrelated to hunger signals. He also showed in another study that external cues for food intake were more important to the obese than to normal sub— jects. When the effect of taste on eating behavior was examined, it was shown that obese subjects ate more than normal or underweight subjects when the taste was rated fairly good by all groups. (The ratings were terrible, bad, not very good, fairly good, very good, and excellent) Underweight subjects tended to eat more than other subjects when the taste was rated as bad or terrible. They concluded that the external, or non-visceral, taste cue had different effects on the eating behavior of underweight, normal, and obese subjects. Other studies were made to test the effect of external cues on persons of different body weight. Schochter con- cluded from these studies that obese people are rather insen- sitive to food deprivation signals of a physiological nature, but are very sensitive and responsive to visual, taste, or food-related environmental cues. Since it is unlikely that only persons of lower socio- economic status will have a tendency to respond in an obese or non-obese way to food, and yet high socio-economic females tend to keep their weight lower on the average, these studies support the contention that women of higher socio-economic status are probably more influenced by social pressure to keep their weight near normal than others. Further, as in studies of socio-economic class, there is the question of whether members of higher socio-economic class are in this class because they or if the ma_i lower socio-e rentally inc. It has ' theory) is n access to up for white fe Finallj Weights in : t0 the Cute seen to ex‘n likely to t 1 m . The l: 20 because they are mentally more capable of upward mobility, or if the majority of the obese drifted or remained in the lower socio-economic class because they were obese and/or mentally incapable of upward mobility (drift theory). It has been suggested that this contention (drift theory) is not applicable to non-white females, since the access to upward mobility is different for them than it is for white females. Finally, it is believed by some that the varying body weights in females is a good index for measuring responses to the outer environment (external stimuli),since the latter seem to exhibit a greater sensitivity to change and are more likely to turn to food as a way of coping. Limitations The limitations of these classes of studies stem.mostly from their resistance to generalization beyond the population being studied. In some cases, the population at risk is not represented, e.g. insurance policy holders may not include a representative proportion of persons from the low socio-econo- mic population where BWE's are found to be more prevalent. The same holds true for weight reducing salons and classes where studies on obesity have been carried out. Hence, the validity of the findings and the conclusions reached are questionable since the population is not representative and major factors relating to the condition may not have been in- cluded in the analysis. One such factor is the longitudinally developmental process involved in BWE and its correlates. CHAPTER III THEORY AND HYPOTHESES Theory Extremes in body weight among adults have customarily been attributed to faulty metabolism or faulty diet. 0f the two, it is noted that metabolic/genetic causes account for a statistically small percentage of extremes, leaving diet and the implied causes of its regulation accountable for most cases. Five conditions predispose faulty dietary patterns: 1. Home environment--culturally influenced habits, individually exercised choices. 2. Economic poverty--cheap starchy foods, ignorance of nutrition. 3. Occupation--housewives, persons working in food industry or service. 4. Emotional factors--eating as compensation for monotony, domestic, or financial, social, business problems, illness. 5. Aging and disease. When obesity is specifically associated with heart conditions, hypertension, diabetes, arthritis, stroke, kidney malfunction, and when tuberculosis and acute respiratory disease are specifically associated with underweight, the above condi- tions are agents of increased morbidity and early mortality. Hypothesis The testing procedure for Alternative I (A—I) and (A-II) involves a series of hypotheses. These hypotheses fOIIOW’the 21 nested methc which in tn 'aewish to weight ext- (A-II) be Ely]. (j = rESpectiv It i SUbSEQUer , \ Ba rd are factors i'hile t' in th a Very invite the the true 22 nested method, each hypothesis being composed of sub-hypotheses which in turn are divided into further sub-hypotheses. Suppose we wish to test (A-I), i.e. that disease conditions and body weight extremes are independent. Let Hxi e (A-I) and Hyj 5 (A-II) be a series of hypotheses-)Hxi (i = l,2,3,--k) and Hy. J respectively (see chart of nested hypotheses). Figure 3. (j = 1,2,3,--m) are sub-hypotheses of (A-I) and (A-II) It is generally agreed that the variations in causes and subsequent manifestations indicate that the three major factors to consider in explaining the development of disease and/or BWE are genetic, traumatic and environmental. The latter two factors are concerned with culture, diet, activity, etc., while the genetic is concerned with metabolic disturbances in the form of inborn errors in metabolism. Mayer noted that a very small percentage of BWE are genetic (Mayer, 1968). Environmental and traumatic conditions seem to account for the majority of the cases. This same line of reasoning is true for most disease conditions. It seems logical to follow the above sequence of hypotheses through a number of sub- hypotheses to discriminate between concurrent development, and causative relationship in either direction for both disease conditions and BWE. Too, as an alternative to this approach, separate dietary habits from environmental influences in consideration of BWE and disease, is untenable in the light of the volume of evidence supporting general environmental impact (Social, cultural and physical) as a significant factor in the 233 «acumen uvcocgu ecu uzm a» mtomtauuga coeeou mo copuuuvmwucoeu me» o» nepumun mmmozuon»: uuumozu-ru.m weaned .meuo h + comm co cowuuotwucw .mtomtauuca + L.oza x5 vomaou min cu coesoo evocu an camamu + cc. mzm coaxuun coon can no new mzm cmmxuoa + mauuwmo acvmomtucw one u<¢>z muummco mnemoocoev oz» Ho¢>x + » u s canon—v upcoggu tea was uauuuu .ou use use on meantauosa 3 $0232.. PS . .Bo «to mass» can» .xutu>oa .oc.u~o a...» 5.: 2328 933 amoumpv o'coceu can wxm .ue.u no.2 momuwtuc. unaware u_=oL;u vco use 5233 258522 2: "5% 5.953 358523 2: “on,“ .350 ”SEES ”5x: 53... .323. “on: a .mco 91‘ Ii» .uwucou + .ms.» mo u=302o anon» mo ummco vacuum 233:3 o L8 3.23-3 883 .8 no :33 has» Laue» apco mcvoon umaompu uwcoceu «Luminous; cossou mgomcsuogq umoompv upeotgu use mam <~ new mam amorous ow o>o= uo: on < coesou u>az o couxpon avgmcovuo_oc ugh " >= nvzmcovuopuc on» “ >1 wzm use oncompo u Nx: mzm new «acumen " Nxz all .|\» e, » + uamuuo pounce unmouo acoucoqou=_ use mzm ecu meowuyucou «acumen qux: » » J .uouunot use mzm was meowuwucou omoomwo ”op>: 24 development of disease and body weight extremes (Dodge, et a1, 1970). Hypothetical Models Extremes in body weight are often observed as coincident with particular diseases and poor health, certain environmental, social and personal conditions precipitating such a chain reaction. Or, when BWE antecede disease, they may appear to be cause of disease. Or, they provide occasion for the appearance of disease symptoms and so affect the condition although not necessarily causing it. Finally, BWE may act as a determinant of disease outcome, rather than an initiating force. In the latter case, BWE is precursor of disease. The hypothetical considerations emerging from.longi- tudinal study of alleged effects of BWE range from demonstrable absence of visible effect to strong positive or negative effect. Figures 1 and 2 show relational analyses implying conceivable statistical results which may be observed between BWE and disease. Figure 1 may be interpreted as indication of positive accumulative relationship between body weight extremes and unspecified diseases. This hypothetical observation is suggested in the literature; however, the contention that the longer the duration the higher the frequency of disease has been inconclusive. Figure 2 shows a hypothetical situation of no relation- ship between BWE and a disease condition. The parallel lines 'BY A: 3ERCENT IN ME GROUP ERCENT IN AGE SROUP 25 FIGURE 1. HYPOTHETICAL MODELS INDICATING PERCENT OF BODY WEIGHT BY AGE AND WEIGHT LEVEL FOR UNSPECIFIED DISEASES ACCUMULATIVE PROCESS UW = Ufiderweight '3 NW = Norma] weight OH = Overweight 3 08 = Obese * /1/ 65+ years // / 60 v 50 ~ 40-64 years 40 0. 25-39 years 30 6- 20 u 18—24 years 1 4r 0 - , On-Set ; {I/ 0 ';fi ‘f; A ¢ UN NW ow OB WEIGHT LEVEL FIGURE 2. CONSTANT PROCESS 60 fir 504» 40._.~.:= — «33:- 9:: .1 >92... as: a... tzl v 19531.3. he 9:: wt <=\O= “was! C» r< #2351923? 3: 5. .L,.< ._> _ :‘c ‘2 k: s... .aw.‘ u> _ g :vCQ rzc_h_=zcw» ..+.( :._= 21" ‘31:.Ll — s~9¢~b3 5:3: Uz~>::—fi I:‘ .. ~ :2: ~>~h<21~kg< II .‘ OL‘a‘u‘ \ 31 «I. «5.33:8 co v33 uéuaflh 60C -553 £023 -93 ouECUO . .._....-.. -.. 2, 5:: 3 >2? .eovmov avauu was» we mecca.» cvmotaoomvu on» coon u>og auou .eo 33:26 can .38 65333 £233.33.” 23 ea 55 -S on 3:23 12» 23.8 28. 8 2.. ~30 yo «3:2»: 3 mos—u acute?“ on A2. 3 .33» .3 .335 .5 .88 9:32.? o'cl. -VDOIDIIIIIIII“ .. .431 . $3332 a. eggs—aezcou 38: £358 «- 203 3 33¢ .325 co.ua_anoq ecu neaum 5.;qu co 323°: .2 dog .6 H do? 3 >333 9 3:32 3.1.0... 8 6:8 as: :5 V- -VII .ca.moo seas“ as» a. 2033 2283.235 use . pacer: £32.! 83.“. -c. cu o—n_mmoa a. a“ .3: 35.: 23522.8 :32 3 9.3.7.: «Luau: 3.3“.— .283... 5.35 $82.3 .822". :83 can. v._uauu ca.»._=aoa 7 I'Ei' V'DIH' El .352.- fistulfiidt .303 8.3338 an!» 3.78. 2.. £77.32 9.38 : .39. a... a .52! ea 85:53.8 . .395» «5.5 2339: mo .25 7?...» S5528 as! o 2.- up! .85.. 3:533. .8052. «>vo .3 egg» m_~:u_>_vc. 28. 53.. as. 3»: Bots?! 2.» a. 2953:. .2353: 3 «83 «>3 5!. not: 5:558 an 95383 9.502 .3323»; 9.. .5383.- .5-3 3 a»... vast-.5935 5 338.2 60:35.: on 3: .26 Tatar-q 7:038 :5 on 9.50: 32:2: 05 EEoucou 33.585 .36! .032. flotaufln 3.3 mg 53 $02.33 3 AI 35038 3.82.2.8 use “guano—v m:o_t~> we. 62:37.38 p335... 55532.2... :88 60C; .53 to «28% 3:25 83 23 man go «88.5.5 05 L8 moo..- xut 5:. 3533322”. 3.: we E3 5 35298 2 353236 28 .25 5 358 2.3383. as use... 55:. 5322.2. we £2320» 53:8. .233 4...: 52.2 93.3.. .8 can... «9...; ea 9:995 goofing: -II 6'39... 52!. .550 2.. m3 3 Centaur... a... an... 3 9: $37.28 £88... .2339:— 522 he 13.. 3.. 3388. 8.5 .33 $3333 33% v5 2. Bo. 3 3:88; patiu uni-k Vt . I :~ 439. 02838. :2. 3 :2: swan» ....usa..¢.a _~ .vaum soaoa... J43» 2,2 333.. we. use .32.. mean 05 E :33... a... 5332.886 :33... veg 333.. .28 ea 33:850.... 5:. :5 53 332.55 53 -Zafuoou a... man on .30.... S 489. mama—22 3a to» 5 :1 363m 82.85 3. .3 a» .538 a as” 3.82.. .263 05 $5 233...; 0.5.. 23 .5293 on 3:2» «cl $3.5 9:33! a... «83388 .9: -28.:3 .232. a: E 536:8 05 333m 55 soil 23 I: 5 :28 82. on An: .rotauta a.» -53 33268. 3 {2380... 3389:. 832,330 95 L8 to:- uoc on 838 33” of mo «:32. .eeovuuom-umotu as» .005. so .380 he. »:3- 531.3.— -5 332-93 3926.... .2338 298.5“ 9322 :05 E; $38.. :33 33:38 .293 .6 a3: 3033—33 £05232. 665. 3 tau—.288 3:53. .538 33» 3.353389 .uco—u -->sonao uuuvto>_u_:: .co.uu_ .saoa o>.uauconotoos a on eovuuo—on seven: .33... 5:2 2.. 3.328 3 82:2 as. moiflsuxo 232. as E: .332 5:2 2.. 33.9.8. yo 28:88:. a... $33 use; 232, as 3: .43? co 32:38 one T3» 3 :1 89.5.». 293 533 «53.5 . 2652. .3 58:2: 8 a... SSS—o8“. co vow-n pennant» «32.3.5 .388 mu 8:82.28 a? ouaoavv use» 8—338... 3538 $323383 32. an: 3 8.. «32.8.8 -5 323.3. £5 32:533.: .533 .325 $32.32. ” [PC «:20: z. 3 SI gut. an 83:23.. cvvcvw 0N*p~L , 2.3 8 3:3. at... 60:53 on»! 53138 canon _ _. . _ _ . 699353 E rogue» 82: to 3:32...) 9: 2.. man 5.3 “.3283. 393 9.32203 5 an; .33 93:09.05 333:! an 55238.. at: £352. 3 3:5 .33 ea 323.383 .35 .33» I...» 5 22x. 08 _ .muco.u.a 353 June—2 3&8 85.52: on: . .a . Jae—98a maou—au—to... .3... 3% f2 05 L33; «5 .3308 we use: 2: L335 9: ns. 3.328 .3 285385 a. 57.08 3... $995 pa 58$ T5232 use—om 853m 3032; guy—50 o» m< Era—a. andmngh act»! 52% no .5an u>~h¢¢uz 82...! SP5 .3 guard uttmcn 35292 no 92¢! <=\on 339‘ £838.89 mag:— 92 urge: has: 58 wise..." «a... 38?. u>:. l3+ GRADE 50" 50 4—e 0 10 Figure 8. -- Percent of Females Answering "Yes" to Seen By Doctor Because of Illness or Medical Problems (1970-7l) ‘ in?“ 6? The indication that the percent of illness did not in- crease continually with age (time) in any of the age classes may be explained in part by the fact that most of the females in the 18-24 year age group are in the peak childbearing stage. The pattern of morbidity and weight is not as clear for those with education above the 12th grade (women in the upper socio-economic class). Illness in this group seem to be influenced more by age (time). However in the older ages, those with weight extremes (both underweight and obese) re- ported a greater percent of illness (significant at a = .05). Except for the underweight category, the pattern of weight-extremes follows the hypothetical model in Figure 2 where there is an initial interrelation between general morbidity and weight, but the increase is not cumulative. However, the null hypotheses of no difference between the proportion of persons who are ill with normal weight and obese as well as underweight is rejected at the .05 level of significance for the former and .01 level for the latter. Females responding "yes" to this question may have one or more acute and/or chronic conditions, so the sum total of morbidity is accounted for here. The first level hypotheses tested here confirm the contention of increased morbidity with increased weight. So the question of whether the weight extremes had a causal re- lation to morbidity, remains, i.e. whether BWE occurred first and was followed by the illness, or whether the illness was already present in some dominant form, causing obesity or 63 underweight which manifested itself at a later date. Hospitalization Overnight: Figure 8 shows that the differences from normal~weight in morbidity for obese and underweight females were significant at older ages and lower socio-economic groups. The examination of the question of hospitalization overnight is an attempt to determine the extent or severity of the illness. The underweight female showed a higher percentage of overnight hospital stays than any other weight class at all ages except for the 18-24 age group. The overweight group is identical with the normal weight group at all ages except the 18-24 age group (Figure 9). The underweight females reporting overnight hospital stay were mostly in the lower socio-ecohomic group (0-12 grade). The pattern for hospital use by weight was not as distinct for the upper socio-economic group as it was the lower socio- economic group. Hospital stay overnight for young females may have been related to pregnancy, since it seemed to decrease with age until after 50 years of age where it started to level off. The difference between overweight or obese females' and normal weight females' frequency of hospital use is not significant, while the underweight females show a higher frequency of hospital use than the females in other weight categories. This difference is significant ( a = .05) only at the older ages and for the lower socio-economic group. This pattern for the underweight female is similar to 64 m -.. _ Underweight Normal weight - + _ + Overweight Obese 40, / PERCENT 30 r -\ / . , YES 204_ TOTAL lo. 0 - - . r . i 4 lo 20 30 40 50 60 70 80 AGE 40‘ PERCENT 30 * ’/.:’ "YES“ 2 .. O-lZ 0 GRADE 10 a 0 e 4r 4. # t 4 ¢ 5 lo 20 30 40 50 60 70 80 AGE 40‘ , PERCENT 3o . "YES" 13+ 20 i GRADE lO«» 0 e e— Lfi e e e 5 4 lo 20 30 40 50 60 70 80 AGE Figure 9. -- Percent of Females Reporting "Yes" to H0§pitalization During the Past Year (l972) by Age, Height and Education 65 their morbidity pattern. The overweight female seems to fare better than the other weight-extreme categories both on morbidity and hospital use. Chronic Diseases: Obesity has been associated with four different types of hazards to health: 1. Changes in various normal body functions. 2. Increased risk of developing certain diseases. 3. Detrimental effects of established diseases. 4. Adverse psychological reactions. In this section of the analysis, the emphasis is on item 2 above, i.e. increased weight increases the risk of developing certain diseases. The five chronic diseases used to test the contention in item 2 and to verify previous findings are hypertension, heart trouble, kidney or bladder trouble, diabetes and emphysema. Hypertension: The data were obtained in response to the question, "In the past year, has . . . had trouble with high blood pressure?" There was a higher percentage of females in the sample reporting hypertension (15.1%) than any of the other chronic diseases presented in this study. About 28% of the obese ‘women reported hypertension compared to about 9% of normal weight women. Approximately half (49.7%) of the obese women 65 and over reported high blood pressure. Hypertension is positively related to age and weight (Figure 10). The underweight female showed a frequency similar to that of the normal weight female. While the PERCENT ll YES ll TOTAL PERCENT "YES" 0- l2 GRADE PERCENT "YES" 1 3+ GRADE Figure l0. -- 30p 20“ lo" sol. 4o» 200 'IO" 50« 20: 10* f 66 ._ - .. -Underweight Normal weight _. + - + Overweight Obese lO 10 Percent of Females Reporting "Yes" to Hypertension During the Past Year (l972) by Weight, Age and Education JG». 67 overweight females reported a higher frequency of hypertension, this frequency is significant only in higher socio—economic groups at older ages. The difference between the normal weight female and obese female is significant ((1= .05) after 40 for both low and high socio-economic groups. The frequency pattern in Figure 10 is similar to the hypothetical model in Figure l, where it is interpreted as being an increased cumulative process. However, the present data do not show what other variables may be responsible for the development of this cumulative process, 95 if the disease itself is responsible for developmental obesity. Epphysema: There was a higher percentage of obese females reporting emphysema for all age groups except the 39-64 age group, where the underweight females show a higher percent (Figure 11). Overweight females reported less emphysema than the other weight groups for all ages except the 24-39 age group. The pattern of reporting emphysema for the underweight females is quite different for the low and high socio-economic groups, i.e. females age 40-64 in the low socio-economic group show a pattern reverse to the same age group of high socio- economic status. While the low socio-economic obese female reports a consistently higher frequency of emphysema in all age groups, this difference from normal weight is not significant except after'age 70--the underweight females in the low socio- economic group in the 40-64 age group showed a significant difference (<:= .05) from all other categories. This was PERCENT "YES" TOTAL PERCENT "YES". 0-12 GRADE PERCENT "YES" 13+ GRADE 15‘- iolL 68 —.~w—-Underweight Normal weight -+-+Overweiqht —— Obese lS‘t 10:» 15.4. 10.» Figure ll. -- Percent of Females Reporting "Yes" to Having Emphysema by Weight, Age and Educational Level 69 also true for the 65+ age group in the high socio-economic group. Emphysema did not increase significantly with age except in the high socio-economic group. About 5% of the females reported having emphysema in the following descending order: obese females reported 6.5%, underweight females reported 4.6%, normal weight females reported 4.4%, and overweight females (showing a consistently lower rate at older ages) reported 3.7%. Diabetes Mellitus: The association between obesity and diabetes has been well documented (Karon, 1965) (Medley, 1965). However, neither the mechanism for this coexistence nor the role that genetic determinants play is clearly understood. It is thought that obesity increases the risk of diabetes among those with a predisposition for insulin resistance. (Karon, 1965). Approximately 5% (4.7%) of the females in the sample reported having diabetes. This is the reported rate esti- mated to be present in the United States population (Craddock, 1970). Females between the ages 18-24 reported very little diabetes regardless of their weight level (Figure 12). Obese females showed a sharp increase in frequency thereafter. The reported frequency increased with age in all weight categories. Here again the frequency for the obese did not appear to be cumulative, since the slopes of the lines between weight categories are nearly parallel. Hypotheses Hx10 would be PERCENT "YES ll TOTAL PERCENT "YES" 0-12 GRADE PERCENT "YES" 13+ GRADE Figure l2. 20» 15R 10«» 15+ 10« 51 201. 204 15" 10.. 51. T 70 —.—--Underweight Normal weight .. + - +Overwei ght Obese 10 -- Percent of Females Answering "Yes" to Having Diabetes During the Past Year (l972) by Weight, Age and Educational Level 7l rejected as this data support claims in the literature, i.e. diabetes shows a positive relationship to being obese. The normal weight, overweight and underweight females were effected by this condition in a similar way. The difference in reported frequency of diabetes between the normal weight female and underweight female, as well as the overweight female, was not significant. However, the difference between the proportion of normal weight reporting diabetes and obese reporting diabetes was significant at the .05 level. Figure 12 also shows that the frequency of diabetes increases with age and weight. The underweight female dropped out of the reporting entirely for the high socio-economic group while the obese female shows a difference from normal weight females that is even more pronounced than in the low socio-economic group. Kidney or Bladder Disease: Very little research has been conducted emphasizing kidney disease and weight. However, in the present study, 8.9% of the females reported having kidney or bladder trouble. This condition did not increase significantly with age except after age 50 when the underweight female in the low socio-economic group took a sharp upward turn that is highly significant (a== .01) after age 65. The obese female showed a significant (a = .05) difference from the normal weight female in the low socio-economic group (Figure 13). Heart Disease: Studies relating to heart disease and obesity have been inconclusive as to the impact of these two 72 _.—-——-Underweight Normal weight .. +_ + Overweight Obese / 20‘. ./ / / 15 ., ’ PERCENT / ,/ \(x "Y ,, a * “"‘7;&—'v——v ES mi \\ x a“. / *\ / TOTAL ./" \\.‘ #24. .. + .. +. .i‘: ... 5 ~ ‘ ‘ - .4 o A A A; A A J 1'0 2'0 30 4b 50 so 70 so AGE / 20 .. ./ / / PERCENT 15 0 / "YES" 0-12 10 ‘* GRADE 0 A J - g A A A. l u. 0 To 20 36 40 5o 66 70 so AGE 20 .. PERCENT l5 .. "YES" .\ ./ 10.. \ X'~+‘+‘.4/ 13+ \/ o \‘F \ * \ GRADE ,/ ' + \‘\ 0 . O \ \ \ o 10 20 30 4d fin‘ 60 7'0 80 AGE Figure 13. -- Percent of Females Answering "Yes" to Kidney or Bladder Trouble in the Past Year (1972) 73 conditions on each other. (Kannel, 1967 (Wardwell, 1970) However, it is inferred that extensive obesity puts an unusual ’workload on the heart, causing acute heart failure. Two percent of the females reported "having any kind of heart trouble during the past twelve months." Twice as many obese women as normal weight women reported heart trouble. Three times as many obese as normal weight women in the high socio-economic group reported heart trouble. The contrasting difference shown in Figure 14 is that underweight women in the low socio-economic group reported a higher frequency of heart trouble than any other weight group, while obese females reported a higher frequency (significant at a = .01 level for age 65+) in the high socio-economic group. The overweight female again showed a lower frequency than all other weight categories. This was more notable as age increased. Heart trouble was shown to increase with age as well as weight; however, reported differences were slight at younger ages. The increase after age 50 followed the pattern shown in the hypothetical model in Figure l. Health-Related Habits: Studies have shown that certain health-related habits are also precursors to specific diseases (Paffenburger, 1969); could this also be true for body weight extremes? Do persons with normal weight have health habits different from others? Could a probe into the health habits of persons with different weights help identify unique problems of persons with body weight extremes so that the solution to these problems would also aid in the prevention PERCENT II YES ll TOTAL PERCENT "YES" 0-12 GRADE PERCENT "YES" 13+ GRADE Figure 14. 74 -——--Underweight Normal weight -+-+Overweight ——Obese 30.. 20+ 101 O —l 00 N O (A) O A O 1! U1 O 05 O \l O 00 O1 301 200 To. 1 301[ 20.. 10" 0 . 0 1'0 20 30 4o 56 so 70 so AGE ' Percent of Females Answering "Yes" to Having Trouble With Any Kind of Heart Trouble 75 of the causes of body weight extremes? Six health-related habits are used as the respondent answered the question, "Does . . . on most days or every day?" Cigarette Smoking: About 30% of the females reported habitual smoking. The number of cigarettes smoked per day was not determined. Most of the smoking was reported by the underweight females (40.9%). Two-thirds (66.2%) of the under- weight females in the 40-64 age group reported smoking; this compares with 26.0% for the obese females in the same age group. Reported cigarette smoking increased with age for the underweight until 64 years and decreased with age for the overweight females. There was significant difference ( health professionals, the magnitude of the problem in 11r1derweight females is far less than it is for the obese females because the number of underweight females is much smaller. As previously mentioned, education is a multiplex sweariable itself, involving numerous sub-variables. As one ()1: ‘more of the sub-variables confounded in the multiplex ‘IEIriable, education seems to be related to the prevention of I><>dyweight extremes and chronic diseases; the higher the Gatiucation level, the lower the frequency of chronic disease 1111 all weight levels. The lower frequency of obesity noted 3111 the higher educated may be due to the commonly held con- tention that most middle class individuals consider extreme (Iverweight unsightly, immoral, and indicate of gluttony (Keys, 1955; Bruch, 1957). It seems reasonable to conclude, lfhen, that some of the variables confounded in lower educa- 1:ion may be precursory to BWE. Some of these variables may 15e nutrition, education, income, and/or the individual's ‘response to stress. Both illness and hospital use were higher among those of lower education levels. Illness is shown to be higher at both weight extremes. Overnight hospital use was higher among the underweight, a pattern for the underweight observed for all ages. Poverty and illness perception may be factors that have some bearing on this observation. 93 The overall pattern of normal weight females receiving medical examinations was reported to be higher than that of females in the other weight classes, for both education levels. The females with higher education reported a higher frequency of medical examination than those in all weight classes with less education. The geographic clustering of obese females was seen in areas of high negative-health indicators. J. May noted that the geographical pattern of disease is a result of environ- mental stimuli that provoke adjustment response to populations. The pattern of disease will change as the environmental stimuli and the host's characteristics change over a period of time (May, 1961). The high frequency of obesity in the low socio- economic area may be part of the variables confounded in the multiplex low socio-economic variable. Kruse points out three concepts of the pathogenesis of disease or degenerative condi- tions that may apply to BWE: 1. Etiological factors include noxious agents, stress, excess, and deprivation; - 2. They operate through their presence, excess, defeat, or absence; 3. There is an interrelation between them (Kruse, 1954). The identification of these stresses and/or agents along with the degree of excess or deprivation may be the identifi- able precursors. Economic, social, hygienic conditions, erroneous nutritional habits, and a difficult struggle for existence, especially during the growing years, can reshape the whole individual and his disposition (Gladstone, 1954). 94 The disadvantage of this design is the same as that of the previous designs as far as the time factor is con- cerned, because the data are treated as cross-sectional studies and the developmental process between BWE and negative health are not observed. In the present study, as well as others of this type, the time factor is taken into consideration by using analyses where individuals of later ages are compared to those of earlier ages. Also individuals of the same age are compared with each other. The limitations of this design have already been mentioned earlier in this paper. Epidemiological studies of this nature are based upon the central axiom that disease is not randomly distributed in the human population, but shows aggregation according to time, space and measurable human traits. (Stallones, 1963) Although this design seldom, if ever, provides positive proof of the "cause" of a disease or condition, it comes close to providing evidence of causal relationships by pro- viding some of the modifications of disease or conditions in populations which produce predicted modifications of disease or conditions in those populations. The hypothesis Hle tested in this study attempted to show "a relationship" be- tween BWE and specified variables in the population, which only begins the process of showing the cause of a condition. Summary This study has investigated the body weight levels of adult females as they relate to a variety of chronic diseases, 95 11ea1th and personal habits. The data showed that about 5% of the females were underweight, 21% were overweight, and 19% were obese . Education was used as a socio-economic factor or indica— tor in which 77% of the females had an education up to the 12th grade and 23% had college and above education. This study suggests that educational status has an impact on body weight extremes; the higher the educational status, the less likely is overweight or obesity. Hence, for some conditions, higher education acted as a deterrent to body weight extremes (especially obesity). As a result of this observation, all subsequent charts are presented in three parts: up to high- school education, college and above education, and the total of the two. Table 1 (Appendix A) shows the distribution of weight by age. The frequency of underweight females decreased con- tinuously with age. The frequency of normal weight females increased slightly until about the age of 50, and then sharply decreased. Overweight and obese females show a similar pattern revealing a low frequency at younger ages, but increasing sharply until age group 40-64. Thereafter, there is a rather sharp decrease in frequency. The frequency of females in the four different weight categories were observed for those that reported illness and overnight stays in hospitals for illnesses rather than pregnancy. The data supported the contention of increased 96 morbidity with increasing body weight extremes (both obesity and underweight). However, the data showed no difference between normal, overweight, and obese females in their reports of overnight hospital stays. The underweight females reported a higher frequency of overnight stays in hospitals. Five chronic diseases were used to measure the relationship between body weight extremes and the chronic disease conditions. Hypertension showed the greatest positive relationship for obesity. The normal weight female and the underweight female had practically the same frequencies. Diabetes was another chronic disease showing a significant (a = .05) relationship for the obese female. Diabetes and hypertension. increased with age and weight. Heart-trouble frequency as reported did not differ significantly between weight classes in the young female. However, after about age 50, the underweight female in the low socio-economic group showed a higher frequency, while in the higher socio-economic group the obese female reported the highest frequency. The reported frequency of emphysema created a pattern difficult to analyze, e.g. the underweight female in the low socio-economic group shows a higher frequency of emphysema than the other weight classes in the age group between 45-64. Obese females reported a higher frequency of emphysema at all age levels. This frequency is significant for only the high socio-economic group at the older ages. 97 Kidney or bladder disease for the low socio—economic group showed a similar frequency at ages under 50. There- after, both weight extremes showed notable increase signifi- cantly different from the normal weight category. The pattern of kidney or bladder disease was not clearly defined for the higher socio-economic group. Females with extremes in body weight generally showed a E1 difference in the selected health habits reported in this study. Normal weight females reported having more physical examina- 3 tions than other females. Obese and underweight females in tj the low socio-economic group reported fewer physical exams. The pattern of cigarette smoking was about the same for all weight classes except the underweight female. This was especially true in the low socio-economic group. The sharp difference for the underweight female was even more evident at the high socio-economic level. About 30% of all females reported taking vitamins. The underweight female reported taking the most, while the obese female reported taking the fewest. The frequency of vitamin consumption increased with age in the low socio-economic group; the pattern for the high socio-economic is not as clear. A definite spatial pattern emerged when respondents were mapped according to residence. This pattern is especial- ly observed in the normal weight and the obese female. The frequency of normal weight females ranges from about 41% in 98 the central part of the city to 62% on the periphery, while the frequency of obese females ranges from about 8% on the periphery to more than triple (29%) that in the urban center. These types of trends have been noted for chronic diseases; they were noted by Centerline's (1956) studies of hypertension and coronary heart disease mortality in Boston. His data showed that relatively high mortality from these causes tended to occur in highly urbanized areas. He suggested that some of the many possible causes for these trends may be diet, exercise and stress. The data presented in this section has demonstrated some of the beginnings of variable identification possible in studying the causes and consequences of BWE. The design has used some frequently used variables to verify or refute findings and also some variables that are seldom if ever used to extend our knowledge concerning this disorder. CHAPTER VI CONCLUSIONS AND RECOMMENDATIONS Conclusions This set of studies has been concerned with methods of inquiry into two anomolies of human growth and develogment: body weight extremes a d chronic disease. The aim has been to point out limitations and potentials of previous studies, to demonstrate use of the sample survey, and to suggest models potentially broader than the traditional explanations of body weight extremes or any other physical condition of cumulative progress. ' A comprehensive review of the literature showed most previous studies of BWE concerned with special groups; while these studies have made some contribution to the observed patterns of correlation between BWE and chronic disease, the contention here is that the interpretations of results based on special groups may be distorted or faulty. This conclusion is based on the observation that present treatment and prevention seem ineffective and may even be harmful. It was also observed that alleged precursors (e.g. education, age, low income, etc.) and consequences (e.g. 99 100 .increased morbidity) of BWE seem to be remarkably similar .in identity and levels of correlation to those variables ‘nhich have been shown to be related to certain diseases (e.g. hypertension, diabetes, arthritis, etc.). The sample survey (Model I) method was used: 1. To demonstrate its use for examining the relation- ships between BWE, and chronic disease and their correlates. 2. To refute or verify findings of previous studies. 3. To demonstrate the methodology, flexibility, and potential in overcoming some of the limitations of previous studies. 4. To broaden the basis for predicting high—risk groups for BWE. The data compiled and analyzed using Model I design sup- ;ported the findings in previous research that selected (disease showed a higher percentage at both under - and over- *weight extremes than at normal weight. The data showed that 'there was little difference in the frequency of disease llevels in persons with normal weight and persons less than £30% overweight, but a significant difference between the frequency of disease in the normal weight, and 913}; 30% 0f the overweight females. The major limitation of this design lies in the sto- chastic nature of the problem. Since the method of research does not allow for the observation of the developmental prC>¢=ess of BWE, it limits the interpretation of when and under what circumstances the disorder deve10ped. 101 These findings are supporting evidence to begin to test further hypotheses that some of the variables specified in Chapter II, and others causing personal anxiety may be precursors to BWE and to chronic disease. However, the one- shot Sample Survey approach is a static study process while BWE, its correlates and precursors are dynamic processes. The concept of area - or community-study design suggested that there are sufficient homOgeneous areas of high - and low- risk occurrence of a condition or disease that by studying the area (or the individual), the findings may expose the importance of environmental variables to BWE and shed more light on previous findings which have led to some of the current methods of prevention and treatment. The emphasis in the design is on the area or community. It has the advantage of being able to observe the area prior to the occurrence of high incidence of negative health conditions, so that some determination can be made about circumstances under which the disorder (BWE) developed. The community-based longitudinal approach is concerned vvith a fixed area that is observed over time. The population as well as the characteristics of the areas are subject to change. This design may be helpful in establishing appro- Ilrfiiate variables and statistical approaches leading to the c=<>hc>rt analysis suggested in a longitudinal study for BWE. The former is designed to observe the area while the latter is designed to follow the cohort. .29. 9.4:. [Hair I., x 102 It is even more efficient to observe the individual over a period of time than to observe the area when a de- velopmental condition is under study. The longitudinal design (Model III) has been used very successfully to study chronic disease as it developed. It has been a consideration in the present study that this method is the most promising for study of BWE, especially since both BWE and chronic disease have been frequently shown to develop under similar conditions. This method of analysis allows for studying the individual long before clinical manifestation of a disease or condition is evident. When design is used to study chronic disease, it usually uncovers associated factors and identifying pre- cursors. It may provide opportunities for the detection of environmental and personal factors in the development of BWE which may be susceptible to change or removal. Recommendations Although it is generally accepted that BWE creates an additional hazard to otherwise healthy people, the variation in the causes and subsequent consequences have been largely undetermined. ~ The magnitude of this problem seems to demand that more needs to be known about the development of BWE in order to provide adequate solutions to this problem. The material presented here leaves little doubt that BWE is confounded with genetic and environmental factors based on generally _ we... 103 accepted conclusions that hereditary and environmental factors include all conditional factors that determine completely the different characteristics of man and that the dynamic physiologic and phychologic changes taking place from conception to maturity affects the growth and development of the individual. While many studies have been concerned with the growth and develOpment of children, a continued surveillance of the individual into adult life and to senescence would certainly give more insight into the relationship of growth and deveIOpment of earlier life, to subsequent body weight levels in later life. If a condition that affects almost half the females in the country is to be understood and controlled, longitudinal studies such as the ones posed in this study should be rigorously undertaken as soon as possible in order that a foundation of fact and theory may be built. This would permit development of guidelines of selective therapy based on a better understanding of the etiological factors of the disorder. BIBLIOGRAPHY Abraham, 5., Collins, G. and Hordsieck, M. Relationship of Childhood Weight Status to Morbidity in Adults. HSMHA Health Repgrts, Vol. 86, Number 3, March, 1971, pp. 273-283. Armstrong, D.B., Dublin, L.I., Wheatley, G.M. and Marks, H.H. Obesity and It's Relationship to Health and Disease. Journal of American Medical Association, March 10, 1951. Baird, I., McLean and Howard, A.N. Education, Obesity; Medical and Scientific Aspects. E.& S. LivingstoneL.T}D.,‘Edinburgh8 LonHOn, 1969} Berkman, P.L. Life Stress and Psychological Well-Being: A Replication of Largner's Analysis in the Midtown-Manhattan Study. Journal of Health and Social Behavior, March 12, 1971. Berlyne, G.M. and Manc, M.B. The Cardio-Respiratory Syndrome of Extreme Obesity. The Lancet, November, 1958. Bertalanffy, Ludwig Von. Problems of Life. John Wiley and Sons, Inc. Watts & Company, London, 1962. Bloom, 8.5. Stability and Change in Human Characteristics: John Wiley and Sons, Inc., New'York, 1964. Boyle, J.F. Bioloqical Patterns in Hypertension by Race, Sex, Body Weight and Skin Color. JAMA, Sept. 7, 1970, Vol. 213, Number 10. Bray, G.A., Bruch, H., Ducon, T.G., Mayer, J., Penich, S.B., Salons, L.B., Seltzer, C.C., Van Itallie, T.B., Nestlake, R.J. Obesity: America's Endemic Problem. Brock, R.D. and Haggard, E.A. The Use of Multivariate Analyses of Var- iance in Behavorial Research. Handbook of Measurement and Assessment in Behaviorial Sciences (Nhitla, 0.k., Edl) Reading, Massachusetts. Asslson-Nealey,l968. Brook, C.G.D. Evidence for a Sensitive Period in Adipose-Peplocation in Men. The Lancet, Vol. 2, Sat. 23, Sept., 1972. Brozek, J. and Keys, A. Relative Body Weight, Age and Fatness. Geria- trics, 1959. Bragg, H.P. Psychological Aspects of Obesitv. Borden Rev. Nutr. Res., 58. 104 105 . Transformation of Oral Impulses in Eating Disorders: A Conceptual Approach. Psychiatric Quarterly, Vol. 35, pp. 458-481, 1961. Cassell, E.G., Lebowitz, M.D., Walter, D.W. and McCarroll, J.R. Health and the Urban Environment IX: The Concept of the Multiplex Independent Variable. American Journal of Public Health, Vol. 61, Number 12, Dec., 1971. Chaing, B.N., Perlman, L.V. and Epstein, F.H. Overweight and Hypertension: A Review. Circulation, 39:403-421, 1971. Coe, R.M. Sociology of Medicine. McGraw-Hill Book Co., New York, 1970. Craddock, D. Obesity and Its Management. E.& S. Livingstone, L.T.D., London, 19691 Dawber, T.R., Kannel, W.B. and Lyell, L.P. An Approach to Longitudinal Studies in A Community (The Framingham Study). Annual of New York Academy of Sciences, 1963. Dennis, R. A Studyof the Heights and Weights of Females in Flint, Mi. Unpublishedpaper, 1970. Dodge, 0.0. and Martin, T.W. Social Stress and Chronic Illnesses. Uni- versity of Notre Dame Press, Notre Dame, 1970. Dubos, Rene. Man,_Medicine and Environment. F.A. Praeger Publisher, New York, 1968. Enterline, P., et al. Geographic Pattern of Deaths From Coronary Heart Disease. Public Health Reports 9, 1955. Evans, J.G. and Rose, G. Hypertension: British Medical Bulletin, Vol. 27, Number 1, pp. 37-42, 1971. » Florey, C.D.V. The Use of Interpretation of Ponderal Index and Other Weight/Height Ratios in Epidemiological Studies. Journal of Chronic Disease, Vol. 23, pp. 93-103, 1970. Fruchter, B. The Scope of Multivariate Research. Multivariate Behavi- orial Research, p. 7, Jan., 1969. Galle, D.R., Grove, W.R., and McPherson, Miller. Population Density and Pathology: What are the Relations for Man? (Evidence from one city suggests that high population density may be linked to pathological behavior). Science, Vol. 176, pp. 23-30, April 7, 1972. Glass, D.C. (Ed.) Environmentgl Influences Biology and Behavior Series. The Rockefeller University Press & Russell-Sage Foundation, New YOFk, 196B. 106 Gladstone, I. Beyond the Germ Theory: The Roles of Deprivation in Health and Disease. NewJYork Academyof Medicine Book, published by the Health Education Council, New York, V01. 1, 1954. Gordon, T. and Kannel, W.B. The Framingham, Mass. Study 20 Years Later. (Kessler, I. and Morton, L.L. Ed.) The Community as an Epidemiological Laboratogy: A Casebook of Community Studies. The JOhns-Hopkins Press, Baltimore and London, 1970. Hadden, D.R., Montgomery, D.A.D. and Weaver, J.A. Myocardial Infarc- tion in Maturity Onset Diabetes. The Lancet, Sat. 12, Feb., 1972. Hinkle, Lawrence E., Jr., M.D. The New York Hospital - Coronel Medical Center in New York. Social Factors and Coronary Heart Disease: Some General Introductory Remarks. Social Science Medicine, 1968. Hollingshead, A.B. and Redlich, F.C. Social Class and Mental Illness. John Wiley and Sons, London and New York, 1958. Hunter, J. and Hughes, C.C. Disease and~Development in Africa. Hans Peter Dreitzel (Ed.) The Social Organization of Health. New York, MacMillan, pp. 150-214, 1971 (Recent Sociology Series, Vol. 3.) Huemann, R.L., Hampton, M.C., Shapiro, L.R. and Behnke, A.R. Adolescent Food Practices Associated with Obesity. Federation Proc., Vol. 25, p. 4, 1966. Hutson, E.M., Cohen, N.L., Kunkel, N.D., Stienkamp, R.C., Rourke, M.H. and Walsh, H.E. Measures of Body Fat and Related Factors in Normal Adults. J.M. Dietic Association, Vol. 47, p. 179, 1965. Johnson, N.L., Burke and Mayer, J. Relative Importance of Inactivity ~ and Overeating Energy Balance of Obese High School Girls. American Journal Clinical Nutrition, Vol. 4, P. 37, 1956. Kannel, et al. Serum Lipid Precursors of Coronary Heart Disease. Human Pathology, Vol. 2, pp. 129-151, March, 1971. . LeBouc, E.J., Dauber, T.R. and McNamara, P.M. Relation of Body Weight to Development of Coronary Heart Disease. Circulation, Vol. 35, p. 734, 1967. Karen, J.H., Grodsky, G.M. and Parelatoes, F.C. Critical Factors in Excessive Seru-insulin Responses to Glucose in Obesity and Maturity Onset Diabetes. The Lancet, Vol. 1, p. 286, 1965. Kessler, I. and Morton, L.L. The Community as an Epidemiologiggl Labor- atory: A Casebook of CommunityStudies. The Johns-Hopkins Press, Baltimore andlfondon, 1970. 107 Keys, A. The Diet and the Development of Coronary Heart Disease. Journal of Chronic Disease, Vol. 4, Number 4, October, 1956. Knutson, A. L. Patterns of Individual Variation. The Individual Society and Health Behavior. The Russell-Sage Foundation, New Yo?k,1965. Kruse, H.D. The Ratios of Health and Disease: How the Presence, Excess, Deficit or Absence of Conditions Envokes Disease. Be ond the Germ Theory. Gladstone, I. (Ed.) Health Education Council New Yfirk, p. 52, 1954. Lahav, E. Methodological Problems in Behavorial Research on Disease. Journal of Chronic Disease, Vol. 20, p. 333, 1967. Lawrence, P.S. and Tebbitts, C. Recent Long-Term Mprbidity Studies in Hagertown, Maryland. Methods in Public Health Research II. American Journal of Public Health, Vol. 41, pp. 101-107, Jan., 1951. Leavell, H.R. and Clark, E.G. Textbook of Preventive Medicine. McGraw- Hill Book Company, New York, p. 163, 1953. Lincoln, J.E. Calorie Intake, Obesity and Physical Activity. American Journal of Clinical Nutrition, Vol. 25, pp. 303-394, April, Linton, P.H., Conley, M. Kuechenmeister, C. and McClusky, H. Satiety and Obesity. American Journal of Clinical Nutrition, Vol. 25, April, 1972. Maddox, George L., Back, Kurt W. and Liederman, Veronica R. Overweight as Social Deviance and Disability. Journal of Health and Social Be- havior, p. 287, 1970. . Anderson, C. F. and Boghonoff. Overweight as a Problem of MEHical Management in a Public Outpatient Clinic. Journal of‘Medical Sciences, Vol. 252, pp. 394-402, October, 1970. Maldofdky, N.D., F.R.C.P. & Rathman, Arthur I., Ed. Personality, Disease Parameters and Medication in Rheumatoid Arthritis, Rheumatic Diseases Unit, Wellesley Hospital, Toronto, Canada. Journal of Chronic Disease, Vol. 24, pp. 363-372, 1971. Pergamon Press, Printediin Great Brita1n. Mark, H. H. Influence of Obesity on Morbidity and Mortality. Bull, New York Academy Medicine, Vol. 36, 1960. Maxfield, E. and Konish, F. Patterns of Food Intake and Physical Activ- ity in Obesity. Journal of American Dietic Association, Vol. 69, p. 406, 1966. ‘ i.li|l’j.‘. 1'1 J .1 .Izllllial luau. ’ \r II I 108 May, Jacques Studies in Disease Ecology. Hafmer Publishing Co., New York, 1961. Mayer. J. Overweight. Prentice-Hall, Inc., Englewood Cliffs, N.J., 1968. McCarthy, M.C. Dietary and Activity Patterns of Obese Women in Trin- idad. Journal of American Dietic Association, Vol. 49, p. 406, 1966. McQuodde, W. What Stress Can Do To You. Fortune, pp. 102-141, Jan., 1972. "“"" Mechanic, 0. Medical Sociology: A Selective View. The Free Press, New York, 1968. Medley, D.R.K. The Relationship Between Diabetes and Obesity. Ouart Journal of Medicine, Vol. 34, p. 111, 1965. Mendelson, M. Obesity as a Factor in Psychological Disorders. Mgg, Clins. N. Am., Vol. 48, p. 1373, 1964. Metropolitan Life Insurance Company. Statistical Bulletin, No. 41, 1960. Moore, M.E., Stunkard, A., Srole, L. Social Class and Mental Illness. JAMA, Vol. 181, Number 11, Sept. 15, 1962, pp. 962-966. Myers, I.K. and Bean, L.L. A Decade Later: A Follow-Up of Social Class and Mental Illness. Jfihn Wiley andCSons,Ihc., New York, 1968. Napier, J.A., Johnson, B.C. and Epstein, F.H. The Tecumseh, Michigan Community Health Study: Kessler, I. and Morton, L.L. (Ed.). Ihg_ Community as an Epidemiologigal Laboratory: A Casebook of Community Studies. "The JEhns-HOpkins Press, Baltimore and’London, 1970. Nutrition Reviews. Body Weight and Coronary Heart Disease, Vol. 25, p. 270,§ept., 1967. Overweight and Hypertension, Vol. 27, Number 6, June, 1969. Assessment, Etiology and Effects of Obesity. Feb., 1955. Ogilvie, R.F. Sugar Tolerance in Obese Subjects: A Review of 65 Cases. Quarter Journal Medicine, Vol. 28, p. 345, 1935. Paffenbarger, R.S., Jr., Nothin, J., Krueger, D.E., Wolf, F.A., Thorme, M.C., LeBauer, E.J. and Williams, J.L. Precursors of Coronary Mor- tality Chronic Diseases in Former College Students. American Journal of Public Health, Vol. 56, pp. 962—971, June, 1966. 109 Page, I.H. and Alvarec, W.C. (Ed.) A Personal View on Diet and Arteri- osclerosis. Modern Medicine, Dec. 27, 1971. Parks, R.E. Human Communities: The City and Human Ecology. The Free Press, New York. CollierJMacMillanLimited, [ondon, pp. 73-87, 1952. Parsons, W. and Crespell, K.R. Obesity, Disease and Health. Pierce, R.M. An Environmental Approach to Urban and Physical Stress: A Geographical Frameonk. Unpublished paper, May. 1972. Predicting Coronary Heart Disease from Body Measurements of Framingham Males. Damon, A. et al, J. Chr. Dist., Vol. 21, pp. 781-802, April, 1969. Radin, E.L., Paul, I.C., and Rose, R. Role of Mechancial Factors in Pathogenesis of Primary Osteoarthritis. The Lancet, p. 520, March 4, 1972. Rahe, R.M., Gunderson, E., Pugh, W.M., Rubin, R.T. and Arthur, R.J. Illness Prediction Studies: Use of Psychosocial and Occupational Characteristics as Predictor . Arch. Environmental Health, Vol. 25, pp. 192-197, Sept., 1972. Richardson, H.B. Psychological Factors in Obesity. Med. Clins. N. Am. Vol. 30, 1946. Rimm, A.A., Warner, L.H., Van Yserloo, B.S. and Bernstein, R. Summar of the Relationship Between Disease Conditions and Obesity. Paper Presented—at American Public Health‘AssociationlCOn erence, Nov.,1972. Rose, G.A. and Williams, R.T. Metabolic Studies on Large and Small Eaters. British Journal of Nutrition, Vol. 15, p. 1, 1961. Schuman, L.M. (Ed.) Research Methodology and Potential in Community Health and Preventive Medicine. Annuals of the New York Academy of Science, Vol. 107, Article 2, 1963. Seyle, H. Stress, Acta, Inc. Medical Publishers, Montreal, Canada, 1959. Shattuck, L. Massachusetts, Sanitggy Commission of Massachusetts 1850. (Introduction 5y C.E.A. Winslow ambridge,‘Howard’Ufiiversity ress, 1948). Silverstone, H.B. Psychological Factors in‘Obesity. Proc. R. Soc. Med. Vol. 61, p. 371, 1970. Smille, W.G. and Kilbourne, E.D. Preventive Medicine and Public Health. The MacMillan Co., New York, p. 460, 1963. 110 Shepard, R. et al. Factors Affecting Body Density and Thickness of Subcutaneous Fat. Data on 518 Canadian City Dwellers. American Journal of Clinical Nutrition, Vol. 22, pp. 75-89, Sept., I969. Society of Arteraries, Body Build and Blood Pressure Study. Chicago, 1959, Vbl.71. Stefanik, P.A., Heald, F.P., Jr., and Mayer, J. Caloric Intake in Relation to Energy Output of Obese and Non-Obese Adolescent Boys. American Journal of Clinical Nutrition, Vol. 7, p. 55, 1959. Stunkard, A., d'Aquil, E., Fox, S. and Filior, R.D.L. Influence of Social Class on Obesity and Thinness in Children. The Journal of American Medicine Association, Vol. 221, pp. 579-584, Number 6, August 7, 1972. '7' . Thomas, H.E., Jr., Kannel, W.B., McNamara, P.M. Obesity: A Hazard to Health. Medical Times, Vol. 95, pp. 1097-1106, October, 1967. Travis, G. Chronic Disease and Disability. University of CalifOrnia Press, Berkéley,Califbrnia, 1966. U.S. Public Health Service, 1966. Wardwell, W.I., Mentor, H. and Claus, B.B. Socio-Environmental Ante- cedents to Coronary Heart Disease in 87 White Males. Social Science and Medicine, Vol. 2, pp. 165-183, 1968. Waterman, T.L. and Lang, V.F. Chronic Illnesses. The C.V. Mosely Co., St. Louis, Missouri, 1955. Willard, W.R. The Present Status and Future Development of Community Health Research: A Critique From the Viewpoint of the Educational Institute in New York. New York Academy of Sciences, 1963. Wilson, H.L. and Wilson, L.W. Obesity and Cardiorespiratory Stress. Obesit , ngson, N.L. (Ed.) and Davis, F.A. Co., Philadelphia, pp. - , 9. Winslow, C.E. The Cost of Sickness and the Price of Health. World Health Organization Monogrgph, Series No. 7, 1951. APPENDIX A TABLES 0F RESPONSES 111 TABLE 1. -- Description of Sample by Age and Weight 18-24 25-39 40-64 65+ *" W. No. 0. #110. 2 TOTAL Underweight 111 36.5 80 26.3 68 22.4 45 14.8 304 9.9 5.6 3.1 5.1 5.4 Normal 758 24.7 881 28.7 1053 34.2 383 12.4 3,075 Weight 67.7 61.4 48.4 43.3 54.8 Up to 30% 146 12.4 237 20.2 541 47.1 250 21.3 1,174 Overweight 13.1 16.5 24.8 28.2 20.9 Over 30% 104 9.81 236 22.2 515 48.5 207 19.5 1,062 Overweight 9.3 16.5 23.7 23.4 18.9 GRAND TOTAL 1,119 1,434 2,177 885 5,615 100.0 100.0 100.0 100.0 100.0 112 TABLE 2.--Racial Composition of Sample by Age and Weight, Ratio by Race Within Age , .._—._-_._.-.... .. ..-— .v.-.——. .- -.-m-.————.—. .. —...---...--.— v.— ..“ -~——--..--.. -. —— -.-——.— 7““ —_-._-..—-. -—.—- _....--._ -wr- ._._.____,,-.-__‘~._._-,._ -—-.—..--.->.-_ -—. _._.... ~_-1 WH ITE NON -WHI T E Age No. % No. % Total Underweight 18-24 96 86.5 15 13.5 111 25-39 74 92.5 6 7.5 80 40-64 64 94.1 4 5.9 68 65 + 43 95.6 2 4.4 45 TOTAL 277 91.1 27 8.9 304 Normal 18-24 687 90.6 71 9.4 758 Weight 25-39 805 91.4 76 8.6 881 40-64 1,007 95.6 46 4.4 1,053 65 + 375 97.9 8. 2.1 383 TOTAL 2,874 93.5 201 6.5 3,075 Up to 30% 18-24 122 83.6 24 16.4 146 Overweight 25-39 187 78.9 50 21.1 237 40-64 492 90.9 49 9.1 541 65 + 238 95.2 12 4.8 250 TOTAL 1,039 88.5 135 11.5 1,174 Over 30% 18-24 85 81.7 19 18.3 146 Overweight 25-39 191 80.9 45 19.1 236 40-64 400 77.7 115 22.3 515 65 + 186 89.9 21 10.1 207 TOTAL 862 81.2 200 1118.8 1,062 GRAND TOTAL 5,052 90.0 563 10.0 5,615 113 TABLE 3. -- Marital Status by Age and Weight SEPARATED SINGLE MARRIED DIVORCED WIDOWED AGE No. No. i No. 1% No. % TOTAL Underweight 18-24 57 51.4 45 40.5 9 8.1 - - 111 25-39 5 6.3 61 76.3 14 17.5 - - 80 40-64 6 8.8 50 73.5 8 11.8 4 5.9 68 65+ 5 11.1 10 22.2 1 2.2 29 64.5 45 Total 73 24.0 166 54.6 32 10.5 33 10.9 304 Normal 18-24 344 45.4 375 49.5 38 5.0 1 .1 758 Weight 25-39 43 4.9 735 83.4 96 10.9 6 .7 881 40-64 47 4.5 844 80.1 81 7.7 81 7.7 . 1,053 65+ 20 5.2 135 35.2 14 3.7 214 55.9 383 Total 354 14.8 2,089 67.9 229 7.5 302 .8 3,075 Up to 30% 18-24 53 36.3 84 57.5 9 6.2 - - 146 Overweight 25-39 10 4.2 191 80.6 33 13.9 3 1.3 237 40-64 22 4.1 409 75.6 57 10.5 53 9.8 541 65+ 16 6.4 111 44.4 12 4.8 111 44.4 250 Total 101 8.6 795 67.7 111 9.5 167 14.2 1,174 Over 30% . 18-24 42 40.4 46 44.2 15 14.4 1 1.0 104 Overweight L 25-39 18 7.6 183 77.5 32 13.6 3 1.3 236 . 40-64 16 3.1 390 75.7 52 10.1 57 11.1 515 65+ 4 1.9 90 43.5 7 3.4 106 51.2 207 Total 80 7.5 709 66.8 106 10.0 167 15.7 1,062 GRAND TOTAL 708 12.6 3,759 67.9 478 8.5 669 11.9 5,615 ___._ _._ __.--, _ 1.. ____.___. -J II 1.1.1 . 1311‘ 111.11 1.1311 .1 lafllill semis: nee seq .3 are: 18126238814 GEE 114 N.N NNF F.m Nme F.N wFF m.m mmF m.m¢ o¢¢.N o.mm mFN.N u . N.m N 1 1 m.F N «.mF NF N.Ne «a N.Nm em «N1NF mom cm>Q a.N mN e.oF FNF N.N NN m.m Ne N.Ne mFm o.mm wee Fewer a. F m.o¢ NoF m. N N.N N o.o¢ ooF N.mF mm + me o. m N.N NF N.N NF m.m NF F.me mmN ¢.m¢ FmN «o1oe N.F a m. F N.F a m.¢ FF m.m¢ NFF N.Fe mm mm1mN azoFm3Lm> m.FF NF 1 1 F.N m F.e m m.Ne Nm N.mm mm 4N1NF mom cu a N.N 11 oFF m.o mmF R.FF mm m.N mm N.Ne mmN.F m.N¢ mFm.F FmpoF F. F ¢.¢¢ oNF N.m N ¢.m mF m.o¢ mmF o.FF Na + mm m. m N.N mN N.N mm N.N «N m.¢¢ Noe m.o¢ owe eo1oe N. F FF 1 1 F. F F: m.N ON m. Fm owe 4.? 8m «NAN £3qu m.NF mm 1 1 m. o F.¢ Fm o.mN NFN o.¢m mow eN1mF Fee...o o.m NN m.w oN N.N m m.¢ mF m.mm NFF m.om oFF Fence 1 1 N.Nm 8N N.N F N.N F o.o¢ NF N. F + mo 1 1 o.m N o.m e m.F F m.m¢ mm m.o¢ NN ¢m1o¢ N.N 1 1 m.F F N.N m m.m¢ om o.me on mm-mN N.FN «N 1 1 m.F N N.N m N.NN om N.F¢ we ¢N1NF ungmztoo=:_ m .oz m .02 a .oz m .oz m .02 m .02 ma< mucmuaum umngma mmmcFFF um oF we: mermmzo: chto: NemFoz use om< x: msunum choFumasooo11.¢ mg i J m