\ wt a _ ‘az n v." v: ., *1 :,. >‘ V ‘ n r: ”‘ if ..» $1 2 " n ;.21 k V}; ‘1 v . A ‘1}! I 3..., « #5:.an . 1 [El] 5 , .2! 1!». In T 5 n» I .lrl~ 95 32.0.?! .5.“- fi. .| .0!!! . , n5)... ‘(T‘V ) I ‘ I s. ‘5). .91:\ 71)»: .5)lV\5§.1b7£.n. ~V. tu‘lum’ )3...“- g mfalflfl}. . .. n1!!! ‘ 1‘ In)... « I... i I: i! I . I , :.( .. .r .3 S??Irtqu-t}lfl)n.1ab& n)» 1 fun“ 75%.!) l. ‘1- ...J; h .. \ Y ‘ a» 3 .x. 7).!V\ ifiziggztfivzt7\)§n ) 1:711. 1 ...n _ 1 ‘ «.35 . .. a \nA: ‘. )nltxfliig)!‘ §\ssillan4ll I! lv VJ ff). ‘1. .ni‘xniSn‘clinhr . .112». A. \ \ I \ w§.I.7thfll!I-. nln.13‘nii‘¥ii\\fin‘fii ¢ . . .r n :3. . it‘nit, . 3...“; x 3 34.1.7. . a \v \11 3 1r1A I :3)(. i . .vl.\\l..l:f.u.\1 3,111.11 x ~ “83.21. infinity)!!! at. .i ’ltfils 7 , . 5.1. )n x x wizi‘1nzili51 “3.31.1: i A. . , 3. «31.1.:ryl}.t$ 31.»)...33‘ (It...) 3)). Hui!!! ‘ u .‘ ‘ltigsil links)! . 3.15. 3:37.. C... 2 :91? 11. 1}. an. .1. x0. \tga £31.91. .3:\\ ... \i 3111..., awn ELISE“: 2:2,. .. _. a . .4 I J\ L \ , 115.! u rt; nI\ \.. #1.}. .‘ ‘fiulnivli1’aaz \an \\ ‘\ .v nilhhkn.‘ k§\a\n\n.7~:1 «rights t. ~ . N. 1.» $315 3.3.3.153)!!! 11...!1r: §,....§ 11 4““. xiii!!- §\!5!\?rnl>.l{x)k (It \ on! \(‘nlar’nlis‘xul-I 3245\5’3‘ 713‘! 2“"? 1“; l. . ..$)¢M¢ti 1703155113»: 33353321.”! i a»)? . $1. $!\!¢\~§ 111 35.24 {1 35:51.4 .1 _ !. )(Jfi5unni‘fli ‘5‘V\|Ir)9nlv , 3!! 1’1 {(1 {’34.} inbpsiznh} 132.3 .51)!» G ! nils‘xlnnnsthSISIinn‘l; ‘ .1. 8‘ 21% I : n ,1, a .2. u 1.. in!) «.109. k! ‘ s. \fis i \I Sauna; 41:. 3 1.1 a x 1‘ n. 5:! _ . . 3’91 1... ..\ 1.35.711 z. (.31.... 3\n’$$$.133< . « .t. :2. x . ‘ $159.?)n ti .. . \.- inn... . a l! \ . 31 “all. ‘ 11 n .1. s 3.) . . . n 1 .n Jr’s-find! \. . .. T3 Influx; . s. .1.) ‘ 1.42. 6. . «13:1 1) x 5 \tcmmnkwcn : A. x I . ‘- 2 iii-‘93.! . ill n 5.50;; I!!! ’9. I HHWJHD'NHEK 1?...1»; v (Fill; n)... , inliill‘gfi’fni! I! v ‘5‘ ,. n 1 Pgtn!3.ii!in§.%§§§nl.£f. n 1114121113.... . ziyaigli‘!ll 1.1.! a} 1 ‘m . §I}¥‘§§£.£ . ‘ vii . .nl'll’li . 1131:1931! 1‘ 53.. x . .ll .1. .I (In! ,I . ‘ii‘iéYlig’iziivgfgflfiflM-ii in“ I!!! . i!¥\7‘13§31\§§v a I , CA 7A.v ‘ . V ‘ , . inn“. . $15!;qu 3g 1 I E isl‘Vlnninnvi. 31.3.1.3 .3715 1-33.11 S‘iilt!i§%. 3.13.!- !x \gig7l 3:933 . 0%., "full I‘M . i1 . h . . 71 n I. .~ . luflnrifitlnknrstilrffnslmrhhnll,‘tn‘. .Nn\!...§ . In krinnn‘nlt .v‘ \ i n in; King . .l flag 3!. unplug; . .xi. . , ..; .. . r.... .51 2.5.120..\\...,4..,.,.~.M.,.MM (will. in}; , ft!- Lllllv \ .rt. .(nnll‘rllil f2|£ iéfi if Egg!!! nnnnnzstsnlstl... innit-1.92%.!tlxxn ‘ git nlil, éh nullinl..!il!!tn}x§n 3.5!!!! iii: Quit. . 3).!!! ..... . it... . .!{n!¢ ;§.§E ; nnr , in)! 34.38%”le "lIHLmJTITMLTTITyLMMMinimum 1H Michigan State University This is to certify that the thesis entitled ANTICIPATED RELOCATION BY THE AGED: LIFE CHANGES AND URINARY FREE CORTISOL presented by Ann Maxwell Eward has been accepted towards fulfillment of the requirements for Ph.D. degeeinFamTIx Ecology aways U Major professor Date 5157/{7 / / 0-7 639 OVERDUE FINES; 25¢ per du per item RETURNING LIBRARY MTERIALS: M ‘ - ”-11 Place in book return to ream 'fi" ' _ "W charge from circulation racer y . \- “r3? 3., ANTICIPATED RELOCATION BY THE AGED: LIFE CHANGES AND URINARY FREE CORTISOL By Ann Maxwell Eward A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Human Ecology I980 © Copyright by ANN MAXWELL ENARD l980 ABSTRACT ANTICIPATED RELOCATION BY THE AGED: LIFE CHANGES AND URINARY FREE CORTISOL By Ann Maxwell Eward The purpose of this investigation was to study the stress of anticipated relocation as related to recent life changes and the impact on individuals as measured by urinary free cortisol. This study observed fifteen functionally healthy, active applicants with a mean age of 80 who were anticipating imminent relocation into a retirement village where, within the same building, a supportive care unit and health center could provide continuous care when and if needed. A control group included fifteen persons living independently in the community who were matched for age and other demographic characteristics. Each participant entered the study as part of a larger, longitudinal investigation in progress. The pre—move assessment included the inventory of life change events and the collection of a twenty-four hour urine specimen for radioimmunoassay. The number, frequency, and clustering of life change events within the past three years was compared by year for pre-movers and non-movers. The two groups were further divided as to category of low, medium, and high numbers of life change events (LCE). Cortisol excretion rates were observed by group and also between and within LCE group categories. The major findings of the study follow: l. The difference in the variance of urinary free cortisol (UFC) excretion rates for pre—movers reporting a high number of life change events over the past three years was statistically significant when compared with non-movers reporting high numbers of LCE. This finding remained significant when compared to pre-movers and non- movers reporting medium or low numbers of LCE. _The observed difference suggests an individual response to a stress threshold which activated the person‘s defense mechanisms and coping style. However, a larger sample size is advised in studies which attempt to replicate this finding. 2. Major life changes numbered two to three yearly for all persons studied. The loss of social support due to death of spouse, other relatives, and or close friends accounted for approximately one-third of those LCE reported. 3. An observation of physiological covariation of cortisol excretion rates for those couples who evidenced an affective relation— ship was interesting; however, the number of couples in the study was small with N=5. 4. A reported decline in physical activity was not observed to limit the general interenvironmental mobility of those aged who are capable of independent living. Recommendations for further study would include a longitudinal study of similar design to repeat observations of life changes, a clinical assessment of health status and coping styles, and the urinary free cortisol excretion rates during and following relocation. The final recommendation would encourage the development of housing models for the aged that provide accommodating environments to attract the active, vigorous aged, offer a choice of services, and provide alternative care areas within the same architectural structure. This would enable the aged to enhance their social support system While enjoying a degree of functional health, moderate activity and independence, and thereby avoid a relocation during a period of cognitive and physical decline. This study is presently within the context of a larger, longitudinal investigation. The investigator would caution against premature judgments related to the stress of relocation based solely on the pre—move data. DEDICATION To my grandmother Nealie Nettles Stallworth born in l882 a kind and gentle person ACKNOWLEDGEMENTS This effort was made possible through the love, support, encouragement, and professional guidance of many people. It is a reminder that no persons ultimately achieve in and of themselves, and that the ability to initiate and sustain meaningful relationships is a tribute to all of us who dare to venture. I, therefore, wish to pay tribute to the following people who contributed foremost as professionals and became good friends, and who are intimately responsible for this achievement with the author. In alphabetical order, and with a deep sense of gratitude, I thank you Dean Norma S. Bobbitt, Ph.D. Douglas Cook George C. Curtis, M.D., Major Consultant Steven M. Downing, Ph.D. Sister Mary Honora Kroger, Ph.D., Research Director Dean Lois A. Lund, Ph.D. Linda J. Nelson, Ph.D. Gertrude L. Nygren, Ph.D. Charles E. Morrill, M.D. E. Jane Oyer, Ph.D. Joseph A. Papsidero, Ph.D. John C. Peirce, M.D. Dean Robert K. Richards, Ph.D. George F. Sargent, Ph.D. Barbara S. Stowe, Ph.D., Major Advisor A. E. VanderMale Mary Ellen Yager As often in the past, I again thank my family, Don, Kenneth, and Michael, whose cooperate efforts made this family affair a family achievement. And to those, who, having successfully aged in our society, gave generously of themselves as voluntary participants in this research effort, I thank you. ii TABLE OF CONTENTS Page ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . ii LIST OF TABLES. . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . . vi Chapter I. STATEMENT OF PROBLEM . . . . . . . . . . . l Overview of Chapter I . . l Statement of the Problem . . . . . . . l Purpose of this Research . . . . . . . 3 Significance of this Research 7 Research Questions and Hypotheses . . . . . 8 Definition of Important Terms . . . . . . . 9 Summary. . . . lO Overview of Other Chapters in Study l0 II. REVIEW OF RELATED LITERATURE . . . . . . . . l2 Overview . . . l2 Relocation as Related to Physical Decline and Illness . . . . . . l2 Stress and Adrenocortical Response . . l8 Life Change Events as Related to Illness and Physical Decline . . . . . . . . . . 25 III. RESEARCH DESIGN AND PROCEDURES . . . . . . . . 29 Design . . . . . . . . . . . . 29 Design Over Time. . . . . . . . . . 29 Experimental Treatment. . . . . . . . 30 Subjects . . . . . . . . . . . . 3l Instrumentation . . . . . . . . . . 34 Validity Concerns . . . . . . . . . . 37 Design Over Variables . . . . . . . . . 4l Research Hypotheses. . . . . . . . . . 44 Procedures . . . . . . . . . . . . . 45 IV. ANALYSIS OF RESULTS . . . . . . . . . . . 53 Findings . . . . . . . . . . 53 Interpretation of Findings . . . . . . . . 70 111 Chapter Page V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS . , . . . 80 Summary and Conclusions . . . . . . . . . . 80 Recommendations . . . . . . . . . . . . 87 REFERENCES 9l APPENDICES 98 Appendix A: Life Change Events Inventory . . . . . 99 Appendix 8: Life Change Events As Factored . . . . l03 Appendix C: Hypothalamo- -Pituitary-Adrenocortical Axis Review. . . . . . . 104 Appendix D: Life Change and Illness Model . . . . lO6 Appendix E: Grant Proposal: Decline of the Aged Related to Stress of Relocation (Abstract) 109 iv Table 10 ll 12 l3 l4 l5 l6 l7 T8 19 20 2T LIST OF TABLES Demographic Characteristics Religious Preference. . . . . . . Frequency Categories of LCE Frequency Distribution of LCE Categories for Movers and Non-movers Number of LCE for Movers and Non—movers by Year. Number of LCE by Year for Movers and Non-movers, Christmas Excluded . Number of LCE for Movers and Non-movers by Factors Page 47 54 55 55 56 of Social Support, Life Style, Events, and Relocation 57 Comparison of Loss by Death Within the Past Three Years for Movers and Non-movers . . . Comparison of Social Support and Life Style Factors by Year for Movers and Non-movers . Significant Life Style Changes Personal Injury or Illness by Year by Group . Significant Events by Group Decision to Move by Year x2 of Decision Year by Frequency of LCE UFC Excretion Rate by Group Frequency Distribution of UFC for Total Sample . Comparison of UFC Values by Group and LCE Frequency Category . . . . ANOVA for UFC and LCE Variance of UFC by LCE Category . UFC Summary Data of LCE Frequency Category by Group Comparison of UFC Mean Difference in Married Couples by Relationship . 58 59 6O 6O 61 62 62 64 65 67 68 68 68 69 Figure 3.l 3.2 3.3 3.4 LIST OF FIGURES Design Over Time Life Change Events Variable Matrix Urinary Free Cortisol Variable Matrix Urinary Free Cortisol by Life Change Events Variable Matrix vi Page 30 42 42 43 CHAPTER I STATEMENT OF PROBLEM OVERVIEW OF CHAPTER I The first topic, STATEMENT OF THE PROBLEM, indicates that relocation of the aged is a major life change, a presumed stressor, and is often associated with physical decline and mortality. Next, the PURPOSE OF THIS RESEARCH is to study the stress of anticipated relocation as related to recent life changes and the impact on the limbic system as measured by urinary free cortisol excretion rates. The third topic, SIGNIFICANCE OF THIS RESEARCH, is to provide anticipatory, pre-movedata to demonstrate the relationship of life changes and cortisol secretion to the stress of imminent relocation. These data artepart of an initial observation in a longitudinal study of individual responses to the stress of relocation over time. The sections of RESEARCH QUESTIONS AND HYPOTHESES and DEFINI- TION OF IMPORTANT TERMS follow. A SUMMARY of chapter I concludes with an OVERVIEW OF OTHER CHAPTERS IN STUDY. STATEMENT OF THE PROBLEM Relocation stress is a real problem associated with the aged and for which there is good evidence of its phenomenal existence and its dire consequences to physical health. The aged appear to respond to the potential stress of relocation in a manner that is complex, heterogeneous,anuiunpredictable. Evidence indicates that the aged respond in a highly individual manner to changes in their environment and that a dramatic increase in death rate may occur in seemingly low- risk populations following relocation. Initially, interest in the study of relocation of elderly persons was spurred by reports of statistically significant increases in mortality and morbidity rates three months prior to three months following a relocation. Subsequent studies have suggested that relocation results in higher mortality and morbidity rates for some ' older persons but lower mortality and morbidity rates for others. During the past decade, studies have challenged purported predictors of high risk in areas concerning certain demographic variables, health status measures, voluntary versus involuntary relocation, and the mediation effects of relocation programs and social services. Disagree- ment exists as to which measures are the most replicable as indicators specific to healthful adjustment and which are significant indices of failure to thrive. Frequently concluded from such studies is the expressed, persistent need for an interdisciplinary, longitudinal approach in studying the problem. The decline of functionally healthy persons who have been relocated was initially reported by Aldrich and Mendkoff (l963), Blenkner (1967), and Lawton and Cohen (l975). Kiritz and Moos (l974) have reported considerable evidence showing that individuals experiencing change in their environment undergo physical changes and are more susceptible to disease. They further state that a feeling of relatedness and environmental support (,0 within one's environment tend to “homogenize” the values of cortisol and positively affect the person-environment fit. In addition, a large body of evidence demonstrates an associa- tion between stressful life events and both psychological and physical illness. The most striking findings according to Miller, Ingham, and Davidson (l976) appear to be the highly significant correlations between threatening events and psychological symptoms. It is postulated that when life events are linked to physical ailments, the exacerbation is initiated by psychological symptoms, i.e. anxiety, depression. For an event to cause ill effects, a stress threshold must be exceeded. Findings suggest that events which fall above this threshold are probably additive -- and the more events there are, the worse the psychological symptoms. PURPOSE OF THE RESEARCH The preceding section has illustrated the need for an objective indicator in assessing the individual differences in response to the stress of relocation. Psychological influences are among the most potent natural stimuli known to affect pituitary-adrenal cortical activity. Marked individual differences in pituitary-adrenal cortical response to any given situation have been a striking and consistent feature of psycho- endocrine studies. The emotional response to a situation both colors the experience and initiates attendant physiological responses within the person's autonomic, psychoendocrine, and neuromuscular apparatus. It is difficult to distinguish the operation of an individual's perceptions of a situation from the defenses or c0ping strategies. A number of studies have illustrated the interaction of personality and environ- mental variables without differentiating between perceptual and coping factors. The measurement of physiological responses which are mediated by difference in perception, coping and defense might enable us to make environments healthier in general or improve person—environment fit for specific groups of individuals. Reliance has been placed upon situational criteria of stress with a relative lack of systematic evaluation of the important and often marked individual differences between subjects and their emotional defensive reactions to a given situation. It has been repeatedly shown with human subjects that there are wide individual differences in psychoendocrine responses to the same situation or stimuli. No matter how seemingly threatening or drastic the life situation or stimulus, it cannot be assumed that all, or even most, subjects will experience substantial emotional arousal or distress. Psychoendocrine measurements such as the cortisol assays provide the behavioral scientist with a sensitive, objective index of physiological reflection of emotional state that represents a balance between forces promoting affective arousal or distress versus those of a protective or defensive nature which prevent, minimize, or counteract arousal or distress. While the dissemination of psychoendocrine data has remained largely within the psychiatric literature, the profound and practical implications for the study and understanding of physiological integration (functions) is a viable challenge to interdisciplinary scientists. As the endocrine glands are regulated in life by both physical and psycho- logical stimuli, the psychoendocrinologist is perhaps in as much danger of ignoring physical or constitutional factors as the physiologist is of ignoring psychological factors in the conceptualization of endocrine regulation in experimental study. In this study, a physiological measure is being recorded and analyzed for its psychological significance, with recognition that the physiological system being monitored is subserving a primary physio- logical function. A fundamental distinction between the physiological and psycho- logical correlates of a physiological measure is in terms of the relevancy or irrelevancy of the physiological change to the direct needs of the body. Because the system is subserving, the use of physio- logical measures in a psychological context is limited; however, with adequate precautions and the careful choice of measure, these physio- logical considerations can be minimized so that valid data with reliable psychological implications can be obtained. With the develOpment of accurate biochemical methods for the measurement of corticosteroid levels in blood and urine, it has been established that there exists an important relationship between psycho— logical function and the adrenocortical activity as evidenced in the earlier studies by Wolff et al. (l964), Mason (I968), and Friedman et al. (I963). Plasma and urinary corticosteroid levels as a function of psychological or physical stress have been of particular interest to numerous investigators currently; among these are George Curtis, M.D. (T972, I976, I977). Measuring the relationship between stress and the activation of the adrenal cortex has been greatly facilitated by the radioimmuno- assay approach which demands less broad biochemical training and thus merits consideration by the psycho-social scientist. Using the corti- sol assay in a recent study, Tennes, Downey and Vernadakis (l976) found that the ratings of separation anxiety and levels of urinary cortisol after the imposed stress were positively correlated. Urinary free cortisol has been used as an indication of adrenocortical function for many years and is used in the human factors research to relate corticosteroid excretion to human performance as reported by Miller et al. (l970). The specificity of determination of urinary free cortisol increased with the advent of the competitive protein binding assays and has increased even more with radioimmuno- assay. The radioimmunoassay method for urinary free c0rtisol provides a very sensitive index of adrenocortical function and is simple, rapid and reproducible as indicated by Sokoloff and Hilderbrand (l977) and by Borsi et al. (l976). The procedure has been applied to over l,OOO urine specimens as part of an ongoing program to correlated biochemical constituent values and occupation stress and performance levels among Naval personnel as reported by Miller et al. (l970). In summary, the primary purpose of this research is to study the stress of relocation as related to recent life changes and the impact on the person as mediated through the hypothalamo-pituitary- adrenocortical axis. (Appendix C). This study is presently within the context of a larger, longitudinal investigation in progress. (Appendix E). SIGNIFICANCE OF THIS RESEARCH The significance of this research is that the assay of the urinary free cortisol prior to relocation will provide a measure of adrenocortical activity as a dependent variable which may then be observed in repeated measures in a longitudinal study to include observations during and following relocation. The schedule of life change events will provide an added measurefOr clarifying and interpreting the findings as associated with the objectification of the cortisol measure. The longitudinal study is calculated to give good evidence of relocation stress by demonstrating the relationship between emotional stress and augmented cortisol secretion and production rates which may be predictive of physical decline. The expected rate of success is high given the objectification of a cortisol measure within the relationship dimension of the person-environment fit. This relationship dimension is conceptualized within a reloca- tion time frame which begins prior to the move. This study of the anticipatory phase, Anticipation of Imminent Relocation, observes the dynamics within the following triad: l. perceived imminent loss of a familiar environment and the anticipated demands for coping with a new environ- ment 2. accrued number of life change events over time 3. relationship to cortisol excretion. The following research questions and hypotheses are of primary interest during the anticipatory phase of relocation. RESEARCH QUESTIONS AND HYPOTHESES RESEARCH QUESTIONS I. Is there an association between an increased number of life change events and the anticipation of imminent relocation? Is there a relationship between the anticipation of imminent relocation and urinary free cortisol excretion? The first question addresses the concept of the "crescendo'I or clustering effect and is to determine if relocation is associated with other major life changes. The second question addresses the noted stability of the adrenocortical measure and is to determine if the anticipation of imminent relocation, apart from other major life change events, causes a change in adrenocortical activity. RESEARCH HYPOTHESES I. There is no difference in the number of Life Change Events between persons who Anticipate Imminent Relocation and those who do not. There is no difference in urinary free cortisol values between persons who Anticipate Imminent Relocation and those who do not. There is no difference in Urinary Free Cortisol Values between persons who have high, medium, and low numbers of Life Change Events. DEFINITION OF IMPORTANT TERMS Definitions for the key terms used in the study follow to provide a common basis for understanding. I. 2. 3. 4. Anticipation of Imminent Relocation - Approved applicants are notified that relocation may be expected within eight months. Observations were made when the applicant anticipated that the move would occur within six weeks. Relocation Time Frame ~ Relocation is conceptualized within a longitudinal time frame with four phases, e.g. (l) the anticipatory phase, (2) the actual move and “novelty” phase, (3) the short—term adjustment, and (4) long-term adjustment. This study examines the first phase, Anticipation of Imminent Relocation. Urinary Free Cortisol Excretion — This is a quantitative measurement Life of free cortisol within the urine and is indicative of adrenocortical activity. Quantification is determined by radioimmunoassay and may be referenced in the laboratory as cortisol production rates. Change Events - The Holmes and Rahe Scale contains forty-three items which include relocation as “change in residence.” (Appendix A). These events usually require some adaptive or coping behavior on the part of the involved individual and represent positive as well as negative items on the inventory. The Holmes and Rahe (l967) list of forty-three events was based on the ”social readjustment ratings” empirically found to be associated with illness onset. lO SUMMARY In summary, Chapter I presented the problem of the decline and disability which is often associated with relocation of the aged. Since relocation has been identified as a major life change event, the study sought to determine the number, frequency, and clustering of other life change events within the three-year period prior to the anticipation of a move. A change in cortisol production and excretion rates has been observed in studies where people are involved in various life change events; however, no study to date has compared cortisol values and other life change events within a relocation time frame. Since cortisol changes may be predictive of c0ping or defense mechanisms, a pre-move assessment may provide valuable information for predicting post-move adjustment as related to health. The significance of this study is to establish basal cortisol values for each participant as a comparison across time for repeated measures in another experimental context. The research questions and hypotheses were generated from studies by other investigators whose works are cited in Chapter II entitled, Review of Related Literature. OVERVIEW OF OTHER CHAPTERS IN THIS STUDY Chapter II provides a REVIEW OF THE RELATED LITERATURE pertinent to three major areas of concern, (l) Relocation of the Aged and Related Physical Decline and Illness, (2) Stress and Adrenocortical ll Response, and (3) Life Change Events as Related to Illness and Physical Decline. Chapter III details the RESEARCH DESIGN AND PROCEDURES, and Chapter IV states the ANALYSIS OF RESULTS. Finally, Chapter V contains the SUMMARY, CONCLUSIONS, DISCUSSION AND RECOMMENDATIONS. CHAPTER II REVIEW OF RELATED LITERATURE OVERVIEW OF CHAPTER II The review of related literature is divided into three main areas: (l) Relocation as Related to Physical Decline and Illness, ‘1... (2) Stress and Adrenocortical Response, and (3) Life Change Events F as Related to Illness and Physical Decline. The parameters of each Ai area presented in the review of literature are interrelated. RELOCATION AS RELATED TO PHYSICAL DECLINE AND ILLNESS Investigators have reported that elderly patients often become seriously ill, sometimes fatally, when they are moved from a familiar environment. Studies of how the external environment affects people undergoing relocation and those measures of health, personality, life style, emotional stability and perception of the move have as yet failed to yield baseline data needed to identify those people placed in special jeopardy by relocation. Consequently, there is little information about how older people should be prepared for a move and what the effects of preparation are in facilitating adjustment to a new setting. The anticipation of relocation and the disorganization during and immediately after a move are associated with physical decline in certain elderly people. l2 . .— T3 In a series of published reports, (Relocation Reports, l97l, l976), the Institute of Gerontology in Ann Arbor disclosed their findings concerning the decision of the Washtenaw County Board of Commissioners to close the old county home in the spring of l97l. The transfer of most of the population of elderly patients to the Whitmore Lake Convalescent Center, a new private facility about l5 miles away, was expected to cause sharp disruption in the patients' lives. The move would be from an old building and a familiar ;! environment to a very modern and unfamiliar milieu. A relocation preparation program was presented in three phases. Phase I permitted all patients who were medically able to be taken to the new convales- cent center and given a tour of the facility. Phase II involved dividing the group in half. The first group visited the new setting three more times. The second group remained at the county home and was acquainted with the convalescent center through slides and pictures. Discussion groups for both Group I and Group II were similar in content and structure. Phase III was the move itself. County facility staff escorted all patients to the new facility on moving day. The results of the program indicated that the multi-visit preparation program had a dramatic effect on the patient mortality rate. Within one year of the move, more than half of Group II (single visit) had died (52%). The mortality rate for the group making four visits was 27 percent. The groups were closely matched on age, sex, degree of illness, physical limitation, mental status, and level of behavior. Overall, the patients from the Washtenaw County Medical Care Facility (WCMCF) died at a much higher rate during the year of relocation than during the l4 previous five years at the same institution. A total of 34 percent of the 6l patients at WCMCF died during the six months before and six months following the move to the Whitmore Lake Convalescent Center. The average annual death rate during the five years before the move at WCMCF was 2l percent. During the year of the move, the relocated WCMCF patients died at a rate three times that of the matched control patients at the Sandusky County Home in Fremont, Ohio who were no: relocated. However, during the three months before the move, the death rate for the relocated WCMCF group was six times greater than the rate for the stable Sandusky County Home group. The rate was about eight times higher during the three months immediately after the move. For a comparison, this study cited statistics from the National Center for Health which showed a mortality rate of 27.5 percent for the l974 Nursing Home Population of the United States. The findings from the Pennsylvania Nursing Home Relocation Program (l976) suggest that the strongest predictor of mortality among the Relocation Program population is age, followed closely by prognosis and mental status at the time of relocation. Other factors associated with mortality are distance moved, and pre- and post-move attitude. Correlated with these three predictors, advanced age, poor prognosis, and confused mental status, are other variables such as incontinence and limited activity. These studies further indicate that patients with a neutral pre-move attitude died at a higher rate than those who were either accepting or rejecting. Or it may be a reflection of the hi I l5 very old sick and/or confused person's inability to express a clear reaction to an anticipated event. In contrast, post—move mortality was 4l percent of those rejecting and I4 percent among those who accepted relocation. Gutman and Herbert (l976) compare the mortality rate of ambulatory and non-ambulatory patients in an involuntary relocation of the entire population of the male extended care ward at Vancouver General Hospital. Of the 24 ambulatory patients, 50 percent died within the twelve months follow-up period. The mortality rate for non-ambulatory patients was only 26.3l percent. A similar trend was observed at 2I months with a 66.66 percent mortality among ambulatory and 43.63 percent among the non—ambulatory patients. In the five control groups, death rates were higher among non-ambulatory than among the ambulatory patients. This finding is in contrast to the results obtained by Killian (I970) who reported excessive mortality rates among non-ambulatory patients but diminished mortality among ambulatory patients. In a study by Rodstein, Savitsky, and Starkman (l976) of ICC elderly persons consecutively admitted to the Jewish Home and Hospital for the Aged, the majority of significant medical changes which occurred in the first month after admission were cardiovascular. Patients with pre-existent cardiac disease and low circulatory reserve are easily thrown into a state of acute coronary insufficiency, acute cardiac failure or acute cardiac arrhythmia by a variety of physical and emotional stresses. Previous studies have supported the concept that life stresses play a significant role in precipitating cardiac and other illnesses. Zohman (I973) related the devel0pment of overt l6 coronary artery disease in the aged to such stresses as change to more restricted living quarters, absence of privacy in living quarters, neglect of visiting by relatives, loneliness, and serious conflicts with housemates such as non-relatives or neighbors. Zweig and Csank (l976) have presented an exceedingly interesting study of the effects of mass transfer of disabled veterans in a chronic disease unit at Ste. Anne's Hospital, Canada,t0 a new building. A transfer-stress—prevention program was introduced one year prior to the actual move in an effort to counter the untoward effects of relocation. That year had the highest absolute mortality count with the lowest population at risk of all years studied. At worst, it should not have been much different from the two years which preceded it. An additional irony is that the move itself was followed by a decline in mortality. Anxiety is basically anticipatory and may account for the paradoxical results of the stress-prevention relocation program and subsequent move. Since anxiety can be inhibited and then displaced after a period of time, it may well be that one of the crucial periods is not in the months that follow relocation but some time later. Further studies by Zweig and Csank give support to the observation that a high-risk period may occur some time after relocation. In contrast to the lengthy preparation period cited above, Watson and Buerkle (l976) describe an involuntary relocation of geriatric Patients at the Veterans Administration Hospital, St. Cloud, Minnesota. One week prior to reorganization, patients were told briefly about the impending transfer. No formal orientation was given. l7 Approximately one-fifth of the patients on a ward were transferred one morning to wards whose architecture, furniture, staff attitudes, and daily routine varied little from those of their previous setting. Many continued the same Rehabilitative Medicine Program to which they had become accustomed. A post reorganization comparison between relocates and those in two comparison groups actually indicated a slight morbidity/mortality advantage but not one of statistical significance. These findings indicate that involuntary transfer can be accomplished without serious traumatic effects on physical health and should be viewed as an attempt to delineate the limits within which the relocation hypothesis operates. Aldrich and Mendkoff (I963) indicate that when disabled persons, especially the elderly, are transferred from one institution to another for reasons which have nothing to do with their health or their family's attitude, there is a marked increase in the death rate, and particularly within the first three months following relocation. For patients whose pre-relocation behavior pattern was neurotic, depressed, or compulsive, or who denied their physical disabilities, the death rate was three times higher than for patients with a satis- factory adjustment and more than twice as high as that for patients whose adjustment was characterized by hostility and demanding behavior. For those patients demonstrating a philosophical or overtly angry response to the news of relocation, the mortality rate was especially low and for those showing an anxiety response, the rate was lower than average. The mortality rates for depressed patients and particularly for the small group who denied the reality of the impending relocation, were much higher than average. 18 STRESS AND ADRENOCORTICAL RESPONSE In I936, Hans Selye demonstrated that the pituitary-adreno- cortical axis which influences many metabolic functions responded to “mere emotional stimuli” in rats. Studies in neuroendocrinology and the behavioral sciences (Harris, l956) demonstrated the functional significance of the hypothalamo-hypophyseal portal capillary system as a neuro-hormonal link between the hypothalamus and the anterior pituitary gland. These hormonal responses are viewed as reflecting intrapsychic processes for a relatively objective approach to quanti- tative and qualitative study of psychological mechanisms operating in response to stress. It is of some import to note that stress is a collective term for an area of study. Lazarus (I966) indicates that ”stress” as a generic term for a complex phenomenon includes the stimuli producing stress reactions, the reactions themselves, and the various intervening processes. Therefore, the parameters of the field of stress include the systemic or physiological which is concerned primarily with the disturbances of tissue systems, (e.g. Cannon, I929; Seyle, l956); the physiological stress with cognitive factors leading to the evaluation of threat (Lazarus, l966); and social stress with the disruption of a social unit or system (Smelser, I963). It is within this collective context that we have operationalized a study of stress and relocation. (Appendix D). The study of stress was facilitated by a major breakthrough in hormone assay methodology in I952 with the chromotographic Nelson- Samuel's method for determining l7—OHCS levels in plasma, and more 19 recently, radioimmunoassay (RIA) procedure. Dr. Rosalyn Yalow and Dr. Solomon Berson have been credited with the development of the sensitive radioimmunoassay procedure which increased the specificity of determination of urinary free cortisol. Further study of interassay and intra-assay variabilities by Sokoloff and Hilderbrand (l977) has demonstrated a high specificity and reliability in the RIA method for determining urinary free cortisol. Since the pituitary-adrenocortical system is known to be altered by psychosocial stimulation, it is the focus of considerable attention because of evidence primarily derived from pharmacological studies. Studies which seek to correlate stress reactions associated with relocation and concomitant illness and physical decline are of interest to researchers from diverse disciplines. In a series of investigations begun some years ago, Kral, Grad, and Berenson (I968) did a follow-up study on the stress reactions resulting from the relocation of an aged population from the Gerontol- ogic Unit of the Allan Memorial Institute of Psychiatry, McGill University and the Maimonides Hospital and Home for the Aged, Montreal. The subjects were categorized as normal male (NM), normal female (NF), psychotic male (PM), and psychotic female (PF). The study concludes that more psychotic subjects died than did those termed as normal; however, some interesting morbidity statistics seem to correlate with findings in the plasma corticoids which were measured before and after relocation. Following relocation the normal men responded with an increase in plasma cortisol (PC) from l4.l to l7.4 mcg percent, while the normal females showed a decrease at the same time from I5.5 to 20 l4.3 mcg percent. The difference in response was statistically significant. There was no significant difference in the PC level between men and women before relocation. Twenty-five percent of the men died within six months of relocation and none of the females died. When subjects were compared according to symptomatology, l8 out of 54 subjects (33%) had no complaints, 23 (43%) had symptoms without organic signs and I3 (24%) had organic signs. The complaints of the 23 subjects without organic signs involved vertigo with falls, insomnia, backache, complaints relating to the anal region and to the passage of urine; however, 4l percent related to the abdominal area. Of the l3 subjects with organic symptoms, six had respiratory tract related infections. Four presented cardiovascular signs, one developed a stroke, another had intraven— tricular conduction defect, and two had myocardial infarcts. Three had symptoms referable to the eye and two of these belonged with the respiratory group. Two subjects developed dermatitis. Eleven of the thirteen subjects developed organic signs within four months of re- location and of these, six developed them between two to four weeks after relocation. Examination of the data indicates that the normal men who pre- sented organic signs showed a greater increase in the plasma cortisol levels than thosw eho presented complaints without organic signs. The mean PC levels of the latter were in turn higher than for those NM who made no complaints regarding their health. Analysis of NC (no complaint) subjects was statistically significant in men only. The four deaths with- in the first six months occurred only in men (NM), and the death rate within 23 months of relocation was 42 percent for male and 20 percent 2T for female subjects. Interestingly, the appearance of organic symptoms and complaints without symptoms occurred at least two weeks after the increase in plasma cortisol. Inasmuch as the greatest plasma cortisol change was observed in those subjects with the severest symptoms, this suggests that thepflasnmlcortisol response to a stress or change may be a prognostic indicator of who will fall ill subsequently, as well as to what degree of illness may develop. Kiritz and Moos (l974) have concurred that there is consider- able evidence that individuals experiencing change in their environment undergo physiological changes and are more susceptible to disease. In his review paper in I968, Mason implicates change and novelty as influences leading to elevation in the l7-hydroxycorticosteroid level. Mason further suggests that while it was generally assumed l5 years ago that adrenal cortical responses in any given stressful situation were caused by ”physical” factors unless proved otherwise, it now appears increasingly important to take the position that concomitant emotional stimuli must be ruled out before it can be concluded that a “physical” stimulus is capable, in itself, of eliciting increased adrenal cortical activity, as Selye's “stress” concepts suggest. It has been repeatedly shown with human subjects that there are wide individual differences in psychoendocrine response to the same situation or stimulus. No matter, suggests Mason, how seemingly threatening or drastic the life situation or stimulus, it cannot be assumed that all, or even most, subjects will experience substantial emotional arousal or distress. 22 There is interesting data concerning the physiological correlates of one individual's involvement with another person, or his involvement in a task or situation (Insel and Moos, l974). Kaplan, Burch, and Bloom (I964) suggest that positive or negative involvement with others in one's psychosocial environment promotes physiological covariation. Other findings (Insel and Moos, l974) support the conclusion that affiliation and not simply exposure to common environmental stimuli is responsible for physiological covariation. Furthermore, Kiritz and Moos state that closely related to affiliation and cohesion is the notion of involvement in one's social environment. Involvement implies a strong affective relationship towards the members and goals of the environment in which one is participating. Clearly some environments are more involving for most of their members than other environments. Mason and Brady (I964) believe that pituitary-adrenal cortical activity can be seen, in part, as a general index of interaction or involvement of the animal with the physical and social environment. They present evidence that the mere presence of others with whom to interact leads to increases in l7-OHCS excretion. In one experiment, rhesus monkeys were housed in individual cages in the same room. When the cages were removed to separate rooms, l7-OHCS excretion dropped. In another study, a hospitalized patient experienced a drop in l7-OHCS excretion every weekend, when he was transferred from a bustling surgical ward to a quiet hospital across the street. In his extensive review of psychoendocrine research, Mason concluded, in part, that elevation of l7-OHCS level is not related to 23 a highly specific affective state such as stress or anxiety, but ”appears to reflect a relatively undifferentiated state of emotional arousal or involvement, perhaps in anticipation of activity or coping.” It is particularly necessary, in the case of the dimension of involvement, to consider the interaction of personal environmental variables. This is exemplified in a study by Friedman, Mason and Hamburg (I963) who measured the l7—OHCS secretion of parents of children hospitalized with leukemia and found that the steroid levels of the 43 subjects tended to remain within a relatively restricted range, even when the parents were exposed to severe acute stresses associated with the illnesses of their children. However, the subjects' rankings of mean daily l7-OHCS excretions were related to how emotionally vulnerable they were to their children's plight, as judged by the experimenters. The most vulnerable parents exhibited the most extreme elevations of l7-OHCS. Those parents who were judged to use denial defenses had the lowest l7-OHCS levels. These parents were, in a sense insulated by their psychological defenses from involvement in the social environment which included their fatally ill children. Thus, even extreme environmental press for involvement can be countermanded, at least temporarily, by the individual's defenses and other stable personality mechanisms. It is often difficult or impossible to distinguish the operation of an individual's perceptions of a situation from the use of defenses or coping strategies. However, a number of studies illustrate the interaction of personality and environmental variables without attempting to differentiate perceptual from coping factors. Katz, 24 Weiner, Gallagher, and Hellman (l970) assayed l7-OHCS in 30 women hos- pitalized with breast tumors, several days before biopsy was to be performed, to determine if the patients had breast cancer. The subjects showed a broad range of values of l7-OHCS secretion, none of them remarkably high. In addition, the subjects were extensively interviewed to determine their patterns of coping and adequacy of ego defenses. The three measures of adequacy of defenses were significantly correlated with l7-OHCS secretion; those with least effective defenses showed the greatest l7-OHCS. Patients who used one of the three defensive patterns labeled ”stoicism-fatalism, prayer and faith," and "denial with rationalization" experienced considerably less disruption, as judged by the steroid hormone secretion rates and the psychiatric rating scores. Summarizing much of this literature, Mason points out that too much reliance has been placed upon situational criteria of stress and upon mean group values, with a relatiVe lack of systematic evaluation of the important and often marked individual differences between subjects in their emotional defensive reactions to a given situation. We have noted empirical studies of cortisol levels as an indicator of perceived stress associated with a change in the person's physical and/or social environment. A study by Volicer (I978) which commented on Mason's observation that "newly admitted subjects exhibited increased urinary corticosteroid levels, decreasing to normal levels during the first week of hospitalization," has found that actual physical changes believed to be indicative of stress can be observed shortly before or at the time of hospitalization. This study further 25 observed that increased life stress prior to hospitalization is associated with high pain scores for medical patients and low physical status scores for surgical patients and less improvement after discharge with a lower degree of return to usual activities than for those who scored low in hospital stress. LIFE CHANGE EVENTS AS RELATED TO ILLNESS AND PHYSICAL DECLINE Life change events are generally temporal in their occurrence and take place over a relatively brief span of an individual's total lifetime; however, the duration of psychological and physiological effects of a person's recent life changes appearSto vary from a few weeks to a year or more. Various studies have associated life change events with the onset of such diverse medical entities as tuberculosis, diabetes, mental illnesses, coronary heart disease, various neoplastic diseases, and the gamut of minor complaints. These life change events have included personal, family, marital, occupational, recreational, economic, social, interpersonal, and religious changes in the psycho- social environment. In studying the effects of specific life changes on illness susceptibility, Syme, Borhani, and Buechley (I965) found that subjects with a history of residential and/or job mobility have significantly higher prevalence rates for coronary heart disease than do persons who are less mobile. The significance of psychosocial stress factors in the etiology of cardiovascular disease has also been reported by Theorell and Rahe (l97l) in an analysis of data collected by Stewart Wolf which provides evidence of the importance of stressful recent life changes as related to coronary death. 26 The relationship of certain life events on man's well being was noted by Farr (I837) who observed that imprisonment resulted in higher mortality rates of inmates. Early in this century, Sigmund Freud, along with colleagues such as Alfred Adler, outlined concepts of early life stress and illness susceptibility, including the notion of illness specificity. Walter B. Cannon (I929) and later Hans Selye (I956) clearly demonstrated physiological phenomenon associated with emotional stress in animals. Franz Alexander (I966), Adolf Meyer(l95l), Stanley Cobb (l976), Harold Wolff (I953), Steward Wolf (I970), and George Engle (I955) contributed to major studies and life stress and illness in human subjects. A shift in methodological emphasis from subjective testimonies of the effects of life stress to more objective analyses of the life stresses themselves was apparent in the early I950's. Meyer (I95I) generated a biographical method for demonstrating the relationship of biological, psychological, and sociological phenomena of processes of health and disease in man. Subsequent research efforts by Wolff, and Wolf, and Hare (I958) suggested that stressful life events evoked psychophysiological reactions and were casually implicated in many diseases. Rahe (I968) used a weighted scale of life change units based on changes in health, work, home, family, personal and social activity as a predictor of illness. Evidence strengthens the inference that change rather than undesirability is the characteristic of life events that should be measured for their effects such as serious psychological and organic diseases as well as for relatively mild psychological symptoms. Many of the life events denoted by Meyers (I95I) are found in the 43 life events inventory by Holmes and Rahe (I967). 27 Nuckolls, Cassell, and Kaplan (I972) have studied life change events for expectant mothers and found that more complications of pregnancy and parturition occurred in the expectant mothers who experienced higher levels of life changes. Paykel et al. (I975) published a study which found that life change events occurring within the past six months were nearly twice the level for patients admitted to the hospital for depression compared to a healthy matched control group. For those patients who attempted suicide, the level of recent life change events was four times that of the controls. Theorell (l97l), Rahe (I97l, I973, I974), and colleagues in Sweden have demonstrated a significant life change buildup six months prior to myocardial infarction, and a three times greater level of recent life change events for patients who died suddenly from heart disease. In other studies by Madison and Viola (I968), as well as Parkes (l97l), it was found that widows and widowers deveIOped significantly more illness during the six months to one year immediately following the death of their spouse than did matched controls. Thus, the findings of a number of different researchers using diverse recent life change and near-future illness methodologies have shown similar findings. However, most studies provide only mean life changes and illness data, applicable to the group as a whole; individual variability in life change buildup and disease onset was often not examined. Studies by Miller, Ingham, and Davidson (I976) found a high correlation between threatening events and psychological symptoms, II 28 especially anxiety. They postulated that when life events are linked to physical ailments the exacerbation is initiated by psychological symptoms such as anxiety or depression. These investigators failed to observe an association between symptomatology and non—threatening events. However, Dohrenwend and Dohrenwend (l974) found that gharge_ rather than undesirability is the characteristic of life events that more accurately assesses their stressfulness. (Horowitz et al. I979) observed that cumulative stress from the impact of life events, looking at measures of frequency, amount, and clustering of life changes, was the variable of greatest import. Rahe (I980) has identified a certain imprecision of thought that has become apparent since the popularization of the life change and illness concept. It is simplistic to view this concept as a close and immediate relationship such as is apparent between the staphylococci endotoxin and acute dysentary. A model which embraces a series of intervening variables is required to identify the subtle and complex process of how certain life changes result in physical decline and even death for certain persons. In conclusion, the numbers and diversity of life changes and illness studies to date have firmly established a significant relation— ship between subjects' recent life change experience and their subse— quent illness and even death. Retrospective investigation has demonstrated a powerful relationship between recent life changes and near future illness; however, while demonstrating that recent life changes remain a significant predictor of near-future illness, prospec- tive studies indicate that the relationship is a relatively weak one. CHAPTER III RESEARCH DESIGN AND PROCEDURES This chapter has two major divisions, the DESIGN of the study, and the PROCEDURES that were followed to implement that design. DESIGN There are two separate design sections in this chapter. First, the Design Over Time section illustrates the overview of the procedures, and builds the foundation for the discussion on the validity of the research, i.e., Validity Concerns. Second, the Design Over Variables section details the interrelation of the variables used in the study and their analysis. It provides the foundation for the Research Hypotheses section. DESIGN OVER TIME This study takes the form of a post-test-only, control group design, similar to Campbell and Stanley's (I963, p. 25) design number 6. While the pretest is a concept deeply embedded in the thinking of researchers in psychology, it is not actually essential to eXperimentaI designs according to Campbell and Stanley. A post-test—only design was chosen because it was not possible to pretest subjects prior to their 29 3O decision to apply for admission to the retirement village. Figure 3.l shows the relationship between the treatment (X), assignment of subjects (R), and measurements (0's). R X 0] 02 R O] 02 Figure 3.l DESIGN OVER TIME Legend: X: Treatment R: Group assignment 0: Observations O] - Life Change Events 02 — Urinary Free Cortisol Excretion In considering the DESIGN OVER TIME, the following components of Figure 3.l will be explained: EXPERIMENTAL TREATMENT (X), SUBJECTS within the experimental and control groups, and INSTRUMENTATION (0's). EXPERIMENTAL TREATMENT The experimental treatment is the anticipation of imminent relocation into a retirement village whose facility includes a progressive care option within the same building. This facility is presumed to be the final habitat and consequently, the last major move for the subjects. Discussion regarding the admission procedure and selection appears in the major section of this chapter entitled, PROCEDURES. 3l SUBJECTS The population under investigation represents persons over 65 years of age who are living independently in Western Michigan. This population is healthy in terms of functional activity but will exhibit chronic diseases characteristic of the aged population at large. The experimental sample of fifteen subjects was drawn sequentially from applicants accepted for residence at an estab— lished retirement community extended care facility serving a healthy middle to upper socio-economic status p0pulation in the Grand Rapids area. Over 45 percent of those applicants contacted agreed to participate in the study, a higher rate than expected due to the arduous nature of the overall design. The control sample of fifteen subjects was drawn from a list of participants in the Voluntary Action Program (R.S.V P.) in Grand Rapids and from additional lists of members in area church ”circles.” Only 35 percent of those contacted actually participated in the study. Those who refused stated that they were too busy for such commitment. In addition to being matched for age, functional health, educational and socio-economic background, the control group must have lived in their present residence for two years or longer with no current plans for relocation. The nature of the study resulted in utilizing a purposive sample that met specific criteria. The study preferenced one 32 retirement facility to avoid interinstitutional variability. Although this particular sampling procedure may not yield a representative sample of the universe which could result in biased estimators of the variables studied and thereby restricting generalizations, the advantages and nature of the study appear to warrant its use. Table l describes certain demographic characteristics in the experimental and control samples. Anticipated Use of Facility by Applicants All of the applicants to the retirement village were function- ally healthy and capable of independent living in spite of the advanced mean age of eighty. Each applicant expressed the desire to chooSe with some feeling of reassurance both where and how they would be cared for in the event of a decline in health. The facility offers independent living units with meals and housekeeping services available for sub- scription. A supportive care unit is housed within the same archi- tectural structure which permits contact with the other residents in the independent living unit while providing additional supportive care. The health center, also located within the building, provides 24-hour skilling nursing care. Admission to this unit may also be either temporary or permanent. Interestingly, several movers anticipated using the facility during the winter months and returning to homes in Northern Michigan in late spring. A few planned to live in Florida during the winter and return to Grand Rapids for the remaining nine months as had been their 33 TABLE l DEMOGRAPHIC CHARACTERISTICS SAMPLE TOTAL EXPERIMENTAL CONTROL VARIABLE N=3O N=l5 N=l5 Age mean 78.9 80.7 77 standard deviation 4.l 6 range 75-87 69-86 Sex female 25 l3 l2 male 5 2 Marital Status married IO 4 6 widowed I8 IO 8 divorced l l 0 single I O I Education (grade completed) 8 - II 4 2 2 l2 - I5 ll 6 5 I6 and over l5 7 8 Socio-economic Status I) Executives, professionals and business managers for large concerns - l5 7 8 2) Administrative personnel, proprietors of small businesses - 9 5 4 3) Clerical, sales workers, technicians, skilled employees - 5 3 2 4) Machine operators and semi-skilled employees - l O 34 life style. Most movers, however, planned to live in residence for the twelve months and would see their family and friends as often as before the move. However, there was generally expressed concern that the families would feel less obligated to visit once the mover was in a more protective environment. The vacancy rates within the independent living units are calculated from previous records at 2.5 per month. During the study, an unusual situation developed where no vacancies were available for a six month period. Most of the subjects had anticipated relocation prior to the time of this study and expressed concern for moving during the winter months. Indeed, several subjects had sold their homes and most of their possessions, and were living in modest surroundings. The movers expressed the feeling that this period of waiting was more stressful than the move would be. Indeed, some of those who had refused to participate in the study had indicated that they felt they I'couldn't take any more stress at this time." Others had indicated a willingness to participate in the study “after moving." Clearly, this period of waiting was perplexing, and expressions of anxiety were related to the ”not knowing” when the actual move would occur. If anticipatory anxiety alone can effect cortisol production rates, then the reported stress in the experimental sample was expected to be positively correlated with cortisol excretion. 35 INSTRUMENTATION This section first discusses the inventory of Life Change Events (0]), and second, discusses the 24-hour urine collection for radioimmunoassay of urinary free cortisol (02). Life Change Events Inventory The Life Change Events Inventory is a modification of the Holmes and Rahe Social Readjustment Rating Questionnaire. Originally designed as a self-administered questionnaire, the SRRS scale of forty-three items documents significant changes in the subject's life in those areas of personal, family, community, social, religious, economic, occupational, residential, and health experience over the past few years. The military version entitled, the Schedule of Recent Experiences (SRE), restricted inquiry of those changes to the past four years, rather than the past ten year period previously assessed in retrospective analyses. Also, each of the SRE yearly assessments is subdivided into two month intervals. Wording was changed for more meaningful communication with service men. Two thousand nine hundred United States Navy officers and enlisted men completed the SRE question- naire. The Holmes and Rahe scaling method has a high correlation on mean judgments of item intensity between culturally different populations with ”r” values usually over .90; however, the important variable in psychosomatic and psychological research is the cumulative stress from the impact of life events looking at measures of frequency, amount, and clustering of life changes. Thirty-seven of the items used in the Life Change Events Inventory for this study were obtained from the Social 36 Readjustment Rating Questionnaire (SRRQ). Two items on the SRRQ pertaining to loans and mortgages were combined into a single category, ”mortgage over SI0,000.” The five other items excluded were pregnancy, and a repetition of events related to formal schooling and current employment. Of the six items added to the inventory, a change in life style, fi.e., change in physical activity), accounted for one, and the other five related to changes in the social support of family and close friends. These were (I) marital problems, (2) trouble with close friend(s), (3) change in family situation of children, (4) change in health or behavior of spouse, and (5) change in health or behavior of close friend(s). Radioimmunoassay of Urinary Free Cortisol Tissue cortisol levels have been determined by fluorometry gas chromatography and gas chromatography-mass spectrOmetry, competitive protein binding assay, and most recently, radioimmunoassay. Rosalyn Sussman Yalow, an American-educated physicist, was awarded the I977 Nobel Prize for Medicine for her part in developing the sensitive radioimmunoassay (RIA) procedure. Although the specificity of deter- mination of urinary free cortisol increased with the advent of the competitive protein binding assays, the specificity has increased even more with RIA. (Sokoloff and Hilderbrand, I977) Explanation and Usage of the Test Cortisol is the major glucocorticoid which is secreted by the adrenal cortex. Among its physiological effects are anti-inflammatory activity, blood pressure maintenance, and the mobilization of lipid and carbohydrate stores. 37 The glucocorticoids transform protein to carbohydrate in response to adrenocorticotropic hormone (ACTH) which is secreted by the anterior lobe of the pituitary gland. In the normal individual, cortisol participates in a negative feedback loop with ACTH through the’ hypothalamus-pituitary—adrenal cortex axis (HPA). Pituitary ACTH is, in turn, regulated by corticotropin-releasing factor (CRF) which is secreted by the hypothalamus. The hypothalamic CRF is reponsive to cortisol levels and to stress. Physical and psychological stress, diurnal variation, and low blood sugars will affect the rate of cortisol' secretion. A malfunction of any of the organs in the HPA axis will result in alteration of the cortisol levels. The assay of circulating cortisol, therefore, assists in the clinical evaluation of individuals whose symp- tomatology suggests chronic adrenal insufficiency (Addison's Disease), Cushing's syndrome (adrenal overproduction), and hypopituitarism. Tissue cortisol levels as a function of psychological or physi« cal stress have been of particular interest in numerous investigations by Curtis (I972), Mason (I973), Miller (I970), Rose (I975), and Rubin (I969). Principles of the Method The GammaCoat 1251 Cortisol Radioimmunoassay was the commer- cially available kit selected for the quantitative determination of the urine cortisol level. This procedure is based on the competitive binding principles of radioimmunoassay and is used for the measurement of urinary free (unconjugated) cortisol. The cortisol present in the urine is first extracted into redistilled methylene chloride. An 38 aliquot of the extract is evaporated directly into the GammaCoat antibody-coated tube. To insure that the samples contain proper levels of protein, IO microliters of Cortisol Serum Blank is added to each GammaCoat tube. The remainder of the assay procedure including prep- aration of a standard curve and monitoring for extraction efficiency is referenced under the proprietary name, GammaCoat, and is also part of normal laboratory procedure for clinical assays. VALIDITY CONCERNS This section discusses possible concerns for first, the internal validity of the study, and then the external validity. Internal validity asks the question: was the experimental treatment the real cause of the observed change? (Campbell and Stanley, I963) External validity asks the question: to what populations, settings, treatment variables, and measurement variables can this effect be generalized? Violations of external validity, therefore, reduce the generalizability of the findings. INTERNAL VALIDITY The design over time is a form of the post-test-only control group design. This is a design which Campbell and Stanley show to have excellent internal validity and controls for HISTORY, MATURATION, STATISTICAL REGRESSION, and EXPERIMENTAL MORTALITY. No unusual circum- stances arose to warrant further discussion of the above concerns. However, the design's control over sources of invalidity arising from SELECTION,INSTRUMENTATIONand TESTING is less clear and does warrant further discussion. 39 Because randomization is the most adequate all-purpose assurance of lack of initial biases between groups, a purposive sample may result in differential selection of respondents for the comparison groups. Therefore, additional observations concerning the selection process for those admitted to the retirement village will be reported in the last section of this chapter entitled, PROCEDURES. The concern for the internal validity of an instrument requiring memory recall resulted in an instrument design which narrowed the time interval selections and repeated the responses for each item on the inventory. These intervals were collapsed into equal time segments for analysis. Therefore, events which the respondent indicated as having occurred within one week, one month, six months, or less than a year appear as ”less than a year" in the data analysis. The additional concern for internal validity relates to the efficiency of urine collection and preservation. Overcollection often occurs as a result of failure to discard the specimen at the beginning of the collection. A visual collection chart was designed as part of the collection procedure and careful instruction for each subject preceded each collection. The use of a commode pan with spout and a 3 inch cap on the gallon collection container facilitated the collection of accurate specimen. Appropriate storage of specimen in a 700 C freezer eliminated concerns for specimen deterioration. Furthermore, diurnal variation has been avoided by the collec- tion of a 24-hour urine sample rather than a single sample of urine, plasma or serum. The secretion of cortisol varies in a non-random, circadian rhythm (24 hours) with the greatest urinary excretion of 4O cortisol between 5:00 A.M. and 9:00 A.M. and the lowest excretion at approximately 9:00 P.M. (Lee, I977). Therefore, invalidity arising from single sample collection has been avoided. Questions relating to the effects of taking a test upon the scores of a second test might appear as a source of concern in a study designed to recall stressful life events and also measure a biochemical stress response, adrenocortical activity. The cortisol measure was chosen for its noted stability. Increased production of adrenocortico— steroids is associated with major life changes which are psychologically and/or physically stressful. Exceptions to this finding have been attributed to certain individuals' abilities to maintain effective psycho- logical defenses against stress, including a suppression of cortisol secretion. An undue concern for testing effect in this study is prob— ably unwarranted. EXTERNAL VALIDITY The threats to the external validity, or generalizability, are reasonably well controlled for in this design, although several possible threats to external validity identified by Campbell and Stanley warrant explanation. The INTERACTION of TESTING and TREATMENT is not a significant concern since the treatment, anticipation of imminent relocation, is not manipulated by the study. However, the INTERACTION and SELECTION of TREATMENT and possibly REACTIVE ARRANGEMENTS identified by Campbell and Stanley (p.26) are a possible concern for external validity. 4l The INTERACTION EFFECTS of SELECTION BIASES and the EXPERI- MENTAL VARIABLES are those associated with a purposive sample. Addition- ally, it should be noted that subjects will ordinarily vary in their willingness to participate, as was observed when approximately 60 percent of those contacted refused to participate. It should be further noted that those subjects participating demonstrated considerable effort towards the accurate recording of the requested data and the accurate collection of the urine specimen. The INTERACTION EFFECTS are the basis for this study and the untoward effect upon the external validity would be to delimit general- izability to a population of similar characteristics. The last concern for the external validity is the REACTIVE EFFECTS of EXPERIMENTAL ARRANGEMENTS. Under design 6 as selected for this study, REACTIVE ARRANGEMENTS are more controlled than in designs 4 and 5. The differential becomes even stronger in a research design for the social sciences than for research on educational instruction. Circumstances did not allow this experiment to be done so that subjects were unaware that they were participating in a study. Consequently, all subjects were equally informed that they were participating in a study and how they were being measured. All of the subjects believed they were in the experimental group. One group was told that the study concerned the stress effects associated with relocation. The other group was informed that the investigation was to study the stress effects associated with aging. The data collection from all subjects occurred at their places of residence. The experimental subjects were approached for observation as they appeared sequentially on a waiting liSt for admissions to the retirement village. Data collection was scheduled 42 when the experimental subjects anticipated imminent relocation within six weeks. The willingness of the experimental subjects to participate in a somewhat rigorous study while anticipating a major relocation may have resulted from an incorrect belief that their participation would posi- tively influence their position on the waiting list of the facility. The consent form stated that the decision to participate was completely voluntary and would in no way effect their relationship with the facility. In conclusion, the design of this study has been shown to be reasonably sound with respect to violations of either internal or exter— nal validity, thus the data collected is reasonably free from contamina- tion. DESIGN OVER VARIABLES Whereas the design over time gave the overall picture of the procedures of the design, the design over variables gives the inter- relationship of the variables used in the study. It leads to the research hypotheses and together they lead to Chapter IV, the Analysis of the Results. VARIABLE MATRICES There are two dependent variables being tested, first is the number of life change events, and the second is the amount of urinary free cortisol excreted during a 24 hour period. Life Change Events Variable. The independent variable in the Life Change Events variable matrix is the anticipation of imminent 43 relocation, from hence referred to as MOVERS. Each cell of the variable matrix contains the number of persons experiencing low, medium, or high numbers of Life Change Events for movers and non—movers. Life Change Events Group Low Med High Movers Non-movers Figure 3.2. Life Change Events Variable Matrix. Urinary Free Cortisol Variables. Two variable matrices will be used to examine the urinary free cortisol hypotheses. The independent variable in the first matrix is the anticipation of imminent relocation (MOVERS). The dependent variable is the mean value of cortisol measured in micrograms per 24-hour urine excretion. This is a discreet variable with expected values within 20 pg to 90 pg per 24-hour sample with repeated measures expected to show little variance in a non-experimental context. Urinary Free Cortisol Group Mean Variance Movers Non-movers Figure 3.3. Urinary Free Cortisol Variable Matrix 44 In the second UFC variable matrix, the independent variable is the frequency of life change events categorized by low, medium, and high number of LCE. In each cell the independent variable is the mean value of cortisol observed for each frequency category of LCE. LCE Frequency Group Low Med High Combined Movers Non-movers Figure 3.4. Urinary Free Cortisol by Life Change Events Variable Matrix. STATISTICAL MODEL OF ANALYSIS All hypotheses were tested at the .05 level of significance. The first hypothesis was tested using a X2 test of independence as illustrated in Figure 3.2. The usual assumptions of independence between and within groupsentzreasonably well met since subjects Dyself assignment were either anticipating an imminent relocation or were not. All subjects were equally eligible for application to the retirement home. The groups are assumed to represent a normal population. Five married couples participated in the study-~two in the experimental group-~which limited the independence within the groups to some extent. Analysis of the second hypothesis is a tgtest for differences between group means as shown in Figure 3.3. In addition to the assump- tions previously stated, population variances are assumed to be equal. 45 The third hypothesis uses a one-way analysis of variance to compare the variability within and between the groups as illustrated in Figure 3.4. This follows an £_distribution. Research has shown that even large violations of the assumptions of equality of variances and normal distribution have little effect on the ANOVA fiftest when the design is balanced and cell sizes are equal. Since the design is balanced with nearly equal cell sizes, there is no reason to expect large violations of assumptions, and the ANOVA fftest is considered to be both normal in power and true level of significance. RESEARCH HYPOTHESES The testable hypotheses, stated in the research form, are given below. They are divided into the Life Change Events and the Urinary Free Cortisol hypotheses to be analyzed via their corresponding variable matrices given in the Design Over Variables section. I. Life Change Events Hypotheses Ho: There is no difference in the number of life change events between persons who anticipate imminent relocation and those who do not. Hi: There is a difference in the number of life change events between persons who anticipate imminent relocation and those who do not. II. Urinary Free Cortisol Hypotheses A. Urinary Free Cortisol By Group Ho: There is no difference in urinary free cortisol 46 values between persons who anticipate imminent relocation and those who do not. Hi: There is a difference in urinary free cortisol values between persons who anticipate imminent relocation and those who do not. 8. Urinary Free Cortisol and Life Change Events Ho: There is no difference in urinary free cortisol values between persons who have low, medium and high numbers of life change events. Hi: There is a difference in urinary free cortisol values between persons who have low, medium and high numbers of life change events. PROCEDURES The procedures of the study are the actions that have been taken to implement the design. Many of the procedures of the experiment have already been outlined in the Design Over Time section of this chapter. This procedures section gives additional detail on the Admis- sions Procedure for Facility Applicants, the Selection Procedures for Experimental and Control Groups, the Interview Procedures, and additional modifications and instruction related to the Instrumentation and Data Collection. Admissions Procedure for Facility Applicants l. Personal Interview at the Retirement Village by the Administrative Director. Admission to the independent 47 living area depends on the physical functioning and mental Status of the applicant as observed by the Director during the interview. Persons who primarily rely upon a wheel— chair for mobility or who demonstrate mental confusion related to orientation are not given an application form. The Formal Written Application. The application includes the usual demographic descriptors and a request for a financial statement and current medical report from the applicant‘s physician. Home Visit. Two members of the Women‘s Board interview the applicantsin their home. Further observations were made by the study to determine possible discriminatory practices related to religious preferences. Table 2 shows the religious mix of the groups. Current residents of the retirement village also include persons of the Jewish faith and Christian Reformed Dutch which are well represented in the community at large. Review Process. The Admissions Committee reviews the recommendation from the Women's Board, the written applica— tion, and the Director's letter. Applicant names are presented at the following board meeting and a letter of acceptance with an estimated date of admission is sent to the approved applicants. 48 TABLE 2 RELIGIOUS PREFERENCE Affiliation Experimental Control Congregational Episcopal Presbyterian Protestant Methodist Christian Catholic Baha'i' None Baptist AOONO’AN—‘OW Data Collector. The experienced interviewer is a registered nurse with an extensive background in community and mental health, admini- stration, and housing and environmental design. Selection Procedures for Experimental and Control Groups. Experimentals. The Administrative Director of the Retirement Village sent a letter to each applicant on the current waiting list stating the purpose and sponsorship of the study and assuring the voluntary nature of the participation. The content of the letter followed the format of the consent form which described the types of information requested and the urine collection, an assurance of confi- dentiality, and further acknowledgment that participation in the study would in no way influence the relationship of the applicant with the retirement village. The letter further stated that an experienced 49 interviewer would telephone the applicant to answer any questions about the study and to set a time for the interview (home visit). No commit- ment to the study was requested at this time. Controls. For those contacts whose names appeared on area church circles, an introductory letter was sent from the Department of Community Health Science and was followed by a telephone call from the interviewer. A The contacts from R.S.V.P. were selected from the volunteer list of approximately 950 names. The selection criteria have been described in the Design Over Time section of this chapter. The letter of introduction was sent to these contacts from the Program Director of the Voluntary Action Center. This was followed by a telephone call from the interviewer. During the initial home visit, a sample interview packet including the larger study was presented to the potential participants. All instruments and instructions could be examined prior to the agree- ment to participate. The data obtained specifically for this study included demographic information on the admissions form, a signed consent form, and the inventory of Life Change Events. During this initial interview all contacts agreed to participate and were interviewed at that time. Following the initial data collection a date was confirmed for the collection of the 24-hour urine specimen. The collection most often occurred the following day; however, if a 24-hour period was to elapse prior to the onset of the urine collection, the interviewer agreed to call the participant on the evening prior to the collection. All 50 collections began upon the awakening of the subject on the morning designated for the collection. Construction of Data Collection Instruments. Life Changg_Events Inventory. The item inclusion on the Life Change Events Inventory have been described under Instrumentation in the Design Over Time sections of this chapter. Because this question- naire was originally designed to be self administered, the design was modified to facilitate aged populations. This included the choice of a more appropriate type size, double spacing between items, and repeated responses. Additionally, related items were grouped in such a way to facilitate memory recall. The first grouping included items relating to family, friends, and support relationships which were measured as a loss. These were followed by items relating to the work situations or job changes. Next were changes in life style which could be perceived as adaptive in some instances to other events. Last, a category of events was listed which could be viewed as either directional, i.e. Christmas, vacations, financial. 'The three year time frame used in the analysis was actually broken down into five time responses. These were as follows: I. "less than a month“ 2. ”a month ago but less than six months” 3. ”six months to about a year ago“ 4. ”about one year ago but less than two years” 5. ”two years but less than three years ago“ The first three responses were collapsed into one interval for analysis as ”less than one year.” The fourth response was indicated 5l as ”less than two years, more than one“ and the last respOnse was noted as ”less than three years, more than two” in the data analysis. When the respondent seemed vague about the time, the inter- viewer would ask questions as to the season of the year or relationship of the event to a major holiday. In addition, when the death of a close friend or relative was noted by the respondent, the interviewer would ask if the person had been a close friend or merely an acquaintance. To avoid duplication of the same event within the responses the interviewer would ask for the specific change noted. For instance, a change in residence does not necessarily indicate a change in life style. On the other hand, if a person noted a change in recreation as not playing golf weekly, this was not duplicated as a ”change in social activities” or a ”change in physical activity“ but as a ”change in recreation." When specific dates were remembered, such as in the instance of ”death of a spouse,” these were noted even though they were not within the three year time frame and were, therefore, not part of the data analysis. Twenty-four Hour Urine Collection. The 24-hour collection of the urine sample was facilitated by the use of a commode pan, a gallon collection jug, and a written schedule of instruction. The written instructions were as follows: l. Good Morning! EMPTY YOUR BLADDER. DO NOT SAVE. Write the time here 2. Place the commode pan in place as demonstrated on the toilet rim under the seat. 52 3. Now save all urine specimens forthe rest of the day and night. Carefully pour each urine specimen into the brown jug. Save, save, save, save, save ..... and have a good night’s sleep ..... save ..... save. 4. Good Morning! Your last specimen to be saved in the brown jug will be at about the same time you wrote down yesterday. Write the time you give your last specimen now Thank you. All subjects were asked these questions during the explanation of the procedure: what time do you usually awaken in the morning? Following that response the subject was asked to repeat the collection procedure. Next, the question was: how often do you get up at night to empty your bladder? And: are you awake enough to remember to save the specimen? The interviewer demonstrated the appropriate use of the commode pan and the collection jug in the subject's bathroom and then requested that the subject repeat the collection procedure. Before leaving, the interviewer again noted the appropriate time when the first specimen was to be discarded and the last specimen was to be saved. At an appointed time, the interviewer collected the specimen and equipment. Next, the total volume was measured in a laboratory and appropriate samples were prepared for freezer storage. The mini-vials were labeled with the study number, name, date, time and volume of the collection. The polypropylene mini—vials (8 mm) were then stored in a -700 C freezer. 53 The collection procedure was facilitated by the use of a special commode pan and a collection jug. The commode pan was designed to fit under the rim of the toilet seat and had a collection capacity of 800 cc which was especially useful during the night-time collections. A small dip in the rim of the commode pan facilitated ease in pouring the specimen into the collection jug. The collection jug was a gallon jug with a 3 inch diameter cap for ease in both pouring and Opening the jug. The amber color was selected for light sensitive material to preserve the integrity of the collection and the polyethylene construction was both light weight and durable. The 3 inch opening facilitated the pouring from the commode pan and the direct collection by male subjects. The procedures have described the selective process for the purposive sample and further described measures taken to insure accuracy and efficiency of the data collection and process. CHAPTER IV ANALYSIS OF RESULTS This chapter is divided into two major parts. The first part reports the findings of the study, and the second part is a discussion of the interpretation of those findings. FINDINGS There are findings in three major areas: first are the findings associated with the life change events for movers and non-movers; second are those findings of urinary free cortisol (UFC) mean distribu- tion in movers and non-movers; and finally are those findings of UFC values associated with life change event (LCE) frequency distribution for the past three years. In each case, the hypothesis is restated, the summary table and associated tables are given, followed by a discussion related to the hypothesis as a summary statement. LIFE CHANGE EVENTS INTERVAL DETERMINATION FOR RATING AMOUNT OF LIFE CHANGE EVENTS The frequency and distribution of the numbers of life change events for each subject provided the interval selection data used for determining low, medium (med), and high (hi) categories of life change events (LCE). 54 55 Table 3 shows the frequency of the number of LCE reported by all subjects within the three year period and the interval selection for low, medium, and high categories with a mean of 8.7 events per subject. TABLE 3 FREQUENCY CATEGORIES OF LCE NUMBER LCE CATEGORY WITHIN 3 YEARS N FREQUENCY LOW 8 4 4 5 2 6 2 MED II 7 6 8 l 9 4 HI ll IO 3 ll 3 l2 2 I3 I I7 I 20 l Hypothesis: There is no difference in the number of life change events between persons who anticipate imminent re- location and those who do not. The distribution of life change events for movers and non- movers was analyzed by a x2 test of independence (x2 = I.409, d.f. = 2, NS). 56 In the x2 test of independence, we cannot reject the null hypothesis of no difference in the numbers of movers and non-movers who experience low, medium, and high numbers of life change events. Table 4 shows frequency in LCE categories by group. TABLE 4 FREQUENCY DISTRIBUTION OF LCE CATEGORIES FOR MOVERS AND NON-MOVERS LCE CATEGORY GROUP LOW MED HI MOVERS 3 7 5 NON-MOVERS 5 4 6 Since the decision to relocate was made during the preceding three year period, additional analyses were performed to determine if the number of life change events within each of these three years was significantly different between and within groups. Table 5 reports the summary of the data for number of LCE for movers and non-movers by year. TABLE 5 NUMBER OF LCE FOR MOVERS AND NON-MOVERS BY YEAR Life Change Events Movers Non-movers x2 d.f. a Less than ONE Year 34 62 39.57 28 .0722 Less than TWO Years 54 39 27.75 23 .2254 Less than THREE Years 39 33 22.68 24 .5386 Total l27 I34 30.58 32 .5382 57 While the data by year continues to support the findings of no significant difference in numbers of life change events for movers and non—movers, the raw count appears to show a greater number of LCE for movers in the interval of less than TWO YEARS, and a greater number of LCE approaching a level of statistical significance for non-movers in the interval of less than ONE YEAR. A further look at the data indicate that the item, Christmas, is perhaps a contaminating factor in these intervals. Since the LCE inventory contains both positive and negative items, Christmas was indicated by all respondents. Thirteen of the experimental subjects indicated that Christmas had occurred ”about a year ago but less than two years.” These subjects were inter— viewed in late fall and early winter prior to the holidays. Fourteen of the control sample were interviewed in mid-winter to late spring and therefore indicated that Christmas had occurred in ”less than a year.” Table 6 represents a revision of the total number of LCE per year for movers and non-movers with the exclusion of the item, Christmas. A more even distribution becomes apparent between years and within groups. TABLE 6 NUMBER OF LCE BY YEAR FOR MOVERS AND NON-MOVERS, CHRISTMAS EXCLUDED YEAR MOVERS NON-MOVERS FREQUENCY Less than ONE Year 32 48 80 Less than TWO Years 4l 38 79 Less than THREE Years 39 33 72 Total Il2 ll9 23l 58 ASSOCIATED ITEMS WITHIN THE LCE INVENTORY The LCE Inventory contains items related to changes in social support, changes in life style, and other events. Some studies have demonstrated an inverse relationship between social support and the effects of stress; therefore, the changes in social support were com- pared for movers and non-movers within this study. In addition, changes in life style, which may accompany changes in social support, were also analyzed. Table 7 summarizes the life change events by factoring the events into the four categories of (I) social support, (2) life style, (3) events, and (4) relocation. The item, relocation, was isolated as a contaminating variable since the inclusion criteria for the control sample restricted participation to those who had lived in their present place of residence for two years or longer. TABLE 7 NUMBER OF LCE FOR MOVERS AND NON-MOVERS BY FACTORS OF SOCIAL SUPPORT, LIFE STYLE, EVENTS, AND RELOCATION Factors Item N MOVERS NON-MOVERS Frequency SOCIAL SUPPORT ll 40 42 82 LIFE STYLE I4 37 37 74 EVENTS 7 42 53 95 RELOCATION I 8 2 . l0 TOTAL ' l27 I34 26l A x2 value equal to 4.74 with three degrees of freedom indicates nosignificant difference (p:s.l920) in the amount of change with factor factoreditems for movers and non-movers. 59 The legend for factored items is as follows: (Appendix) Social Support — l, 2, 3, IO, I2, l3, l4, l6, l7, l8, 3l Life Style - I9, 20, 2l, 22, 23, 25, 27, 28, 29, 30, 32, 33, 34, 35 Events - IS, 24, 36, 37, 38, 39, 4O Relocation - 26 Ten items on the 43—item Life Change Events Inventory received no response from the subject and were, therefore, excluded from the legend. Social Support The social support factor contains three items related to loss by death. These are (I) death of spouse, (2) death of close family member(s), and (3) death of close friend(s). Table 8 shows the distri- bution of loss for movers and non-movers. TABLE 8 COMPARISON OF LOSS BY DEATH WITHIN THE PAST THREE YEARS FOR MOVERS AND NON-MOVERS Loss By Death Items Movers Non-Movers Frequency Total 25 3I 56 Death of Spouse 4 O 4 Death of Close Family Member(s) 9 I8 27 Death of Close Friend(s) l2 I3 25 The overall loss by death appears comparably distributed for movers and non—movers with the noted exception of death of spouse. A 6O 2 x 2 contingency table was constructed for computing a frequency distribution for death of a spouse. The X2 = 4.6l5 with one degree of freedom is statistically significant at p S .05. This would suggest that the death of a spouse is associated with the decision to relocate. Table9 is a summary of the two factors, social support and life style, by year for movers and non-movers. TABLE 9 COMPARISON OF SOCIAL SUPPORT AND LIFE STYLE FACTORS BY YEAR FOR MOVERS AND NON-MOVERS Factors Group >l I>2 2>3 Frequency Social Support Movers 8 l3 I9 40 Non-Movers IO l5 I7 42 Llfe Style Movers 11 15 11 37 Non-Movers 24 8 5 37 The data reveal a comparable distribution of life change events for both movers and non-movers over the past three years. Changes in social support and life style are evenly distributed both between groups and factors. The changes in social support may be greater than is apparent since these changes are representative of II items in the social support scale; whereas, there are l4 items in the life style factor. Life Style The changes in life style were not statistically different between groups; however, the distribution between items indicates a 6T selectivity which may be characteristic of this age sample. Those changes in life style which occurred to more than 25 percent of the total sample within the three year period are reported in Table I0. TABLE lO SIGNIFICANT LIFE STYLE CHANGES Item Movers Non-movers Frequency Personal injury or illness ll IO 2l Change in eating habits 5 3 8 Change in physical activity 6 5 ll These reported life style changes are evenly distributed between movers and non-movers with x2 = .2407, two degrees of freedom at a p f .05 level. Table II shows the distribution of personal injury or illness by year for movers and non-movers. This table is of interest in observing whether personal injury or illness precedes the decision to move or may follow such a decision. TABLE ll PERSONAL INJURY OR ILLNESS BY YEAR BY GROUP Group By Year