ALEX C. MICHALOS, BRUNO D. ZUMBO and ANITA HUBLEY

HEALTH AND THE QUALITY OF LIFE ? (Accepted 22 July, 1999) ABSTRACT. The aim of this investigation was to explain the impact of people’s self-reported health on their levels of satisfaction with their health, and the impact of these things plus satisfaction with other specific domains of their lives on the perceived quality of their lives. The latter was operationalized as general happiness, satisfaction with life as a whole and overall satisfaction with the quality of life. Seven hundred and twenty-three (723) usable questionnaires returned from a mailout random sample of 2500 households of Prince George, British Columbia in November 1998 formed the working data-set for our analyses. Among other things, mean respondent scores on the SF-36 health profile were found to be lower than published norms from the UK, USA, Netherlands and Sweden, but higher than scores from Aberdeen, Scotland. Mean scores on the CES-D depression scale also indicated that our respondents tended to have more depressive symptoms than comparison groups in Winnipeg and the USA. A review of trends in mean scores on 17 quality of life items (e.g., satisfaction with family life, financial security, recreation, etc.) from 1994, 1997 and 1998 revealed that there were only 7 statistically significant changes across the four year period and they were all negative. Multivariate regression analysis showed that health status measured with a variety of indicators could explain 56% of the variation in respondents’ reported satisfaction with their health. A combination of health status plus domain satisfaction indicators could explain 53% of the variation in respondents’ reported happiness, 68% of reported life satisfaction and 63% of reported satisfaction with the overall quality of life. Sixty percent of the explained variation in happiness scores was attributable to self-reported health scores, while only 18% of the explained variation in satisfaction with life and with the overall quality of life scores was attributable self-reported health scores.

?

The authors would like to express their appreciation to Shelley Rennick and Shelley Webber for the help they gave us in the development, administration, analysis and reporting of the survey, and to George Deagle and J.C.J.M. de Haes for comments on the first draft of the paper. Social Indicators Research 51: 245–286, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.

246

ALEX C. MICHALOS ET AL.

INTRODUCTION

The aim of this investigation was to explain the impact of people’s self-reported health on their levels of satisfaction with their health, and the impact of satisfaction with health and satisfaction with other specific domains of their lives on the perceived quality of their lives. The latter was operationalized as general happiness, satisfaction with life as a whole and overall satisfaction with the quality of life. We begin with an extensive literature review that reveals a wide variety of ways in which researchers have connected subjective well-being broadly conceived to health or aspects of health. The review is a bit longer than usual because we want to give readers of this journal a good view of the work already done in this area, most of which has been published elsewhere. This is followed by a brief review of the 36-item Medical Outcomes Study Short Form (SF-36) profile and its use in explaining the quality of life. We then provide an overview of our sampling technique, questionnaire and sample characteristics. After characterizing the demographics of our sample, we summarize the information obtained about various aspects of respondents’ health and, wherever possible, provide some comparative statistics. We review results regarding smoking, drinking and socializing habits, and respondents’ feelings about their age. Then we reveal some assessments that some respondents made of health care services in the area, and we show some trend data on satisfaction levels with several domains of life and life as a whole since 1994. Following this descriptive material, we review the results of several regression analyses in which the dependent variables are satisfaction with health, satisfaction with life as a whole, overall quality of life, and general happiness whereas the independent variables are measures of various aspects of health and satisfaction with health as well as satisfaction with other domains of life.

HEALTH AND SUBJECTIVE WELL-BEING

Many studies have shown a positive association between subjective well-being (operationalized in a variety of ways) and objectively and subjectively measured health, and satisfaction with one’s own health (e.g., Campbell et al., 1976; Andrews and Withey, 1976). Okun et

HEALTH AND THE QUALITY OF LIFE

247

al. (1984) did a meta-analysis of 104 American studies published before 1980 dealing mostly with elderly subjects, and reported that objectively and subjectively measured health explained between 8% and 14% of the variance in subjective well-being. According to Aaronson et al. (1986), “a review of clinical trials undertaken between 1956 and 1976” revealed that at most five percent of the studies involved some relatively obvious measure of the impact of treatment on the quality of patients’ lives, with the latter measured in a variety of ways. Fayers and Jones (1983) identified over 200 papers with the term “quality of life” in their titles. Bergner (1989) reviewed papers abstracted in Index Medicus in order to compare usage of the phrases “quality of life” and “health status”. She found that most people using the former phrase conceptualized it as a discrepancy between desired and attained goals, and many people regarded it “as a risk factor or cause of illness as well as an outcome of medical care” (p. S149). On the other hand, health status is usually conceptualized as an extension of functional status. So, it usually includes performance measures for activities of daily living, measures of body functions, emotional status, symptoms and cognitive functions. Following the World Health Organization’s (WHO) definition of health as a “state of complete physical, mental and social wellbeing”, the promotion of good health might be indistinguishable from the promotion of a good life. The 1952 U.S. President’s Commission on Health Needs of the Nation seems to have taken this view when it wrote that “All of those measures aimed at improving the health aspects of the environment in which people live and at improving personal health practices constitute the promotion of health. Better housing, better nutrition, better working conditions, better education will enhance the health of our people just as certainly as will better physicians’ care” (Quoted by Breslow, 1989: pp. S211–S212). Breslow correctly remarked that “When health is seen as enabling people to lead a full (high-quality) life, health is clearly different from quality of life itself. Health professionals need to be concerned about the quality of life but their task is to enlarge the capacity for it, not to ‘take it over’ ” (p. S215). Excellent reviews of quality of life measures used in the medical literature may be found in Hollandsworth (1988), Goodinson and

248

ALEX C. MICHALOS ET AL.

Singleton (1989), and in several articles in Spilker (1996). Hollandsworth compared 69 studies published in the 1980–1984 period to 23 studies published in the 1975–1979 period and found, among other things, that (1) 60% of the recent studies included a subjective measure of quality of life compared to only 10% in the earlier period, (2) the recent studies tended to use larger (but still mainly convenience) samples and more control groups, and (3) 51% of the recent studies reported either negative or mixed results for treatment interventions compared to the typically favourable results reported in the earlier period. Several studies have shown that, among older adults, there is a positive association between defining oneself as younger than one’s chronological age and both physical and mental health (Busse et al., 1970; Milligan et al., 1985; Montepare and Lachman, 1989). Carp and Carp (1981) found that, in a sample of 352 older adults, both self-reported and observer-rated physical health were related to self-identifying as middle-aged rather than old, elderly, or aged, but were not related to chronological age. Logan et al. (1992) indicated that functional health affects the probability of reporting an old identity but not the probability of reporting a young identity. Andres et al. (1990) reported that better self-rated health and less illness was related to a younger subjective age in a sample of 326 elderly men. In two different studies of older adults, Hubley (1991) found that feeling younger than one’s age was related to better self-reported general health and health relative to others in one’s age group, fewer chronic diseases, and fewer restrictions on one’s instrumental activities due to health problems. Markides and Pappas (1982) reported that subjective age was a better predictor of mortality than age or either objective or self-reported health among 460 predominantly working- or lower-class Mexican American and Anglo elderly. There appears to be a gender difference in the relationship between subjective age and health. Mutran and Burke (1979) reported that subjective age was directly related to health for women. For men, however, subjective age was only indirectly related to health (through retirement). In a later study, Steitz and McClary (1988) found that, in a small group of 93 male executives, poor self-rated health was related to thinking of oneself as being older.

HEALTH AND THE QUALITY OF LIFE

249

A younger subjective age has also been linked to life satisfaction (Logan et al., 1992) and quality of life. Mutran and Burke (1979) reported that subjective age was directly linked to life satisfaction for women, but not for men. Steitz and McClary (1988) reported that neither job satisfaction nor life satisfaction was related to a perception of being younger or older, but financial satisfaction did show a positive relationship. Staats et al. (1993) examined the relationship between various aspects of subjective age and both present and future quality of life. They found that the age group that one’s activities are most like is a significant predictor of present perceived quality of life. That is, identifying one’s activities as belonging to an age group younger than one’s own age group is related to a higher self-reported current quality of life. However, the age one’s interests are most like is a significant predictor of future (i.e., 5 years from now) quality of life. Again, identifying one’s interests as belonging to an age group younger than one’s own age group is related to a higher self-reported future quality of life. Hayes and Ross (1986) cite several earlier studies indicating a high correlation between physical health and psychological wellbeing, with no clear causal order. In their probability sample of 401 Illinois residents, they found interesting and complex interactions among physical and social factors influencing physical health. In particular, contrary to several authors cited below, they found that the impact of social forces was not primarily through social comparison mechanisms. Social factors influence physical health. The determinants of physical characteristics are not purely biological or genetic. The well-educated exercise more, are less likely to be overweight, and are in better physical health than the poorly educated. Persons with high incomes exercise more and are in better physical health than those with low incomes. Married persons exercise less and are more likely to be overweight than the unmarried. And women are less likely to be overweight than men. Thus, social forces affect weight, exercise, and physical health. The latter two, in turn, affect psychological well-being, apparently through interactions with others . . . the link between body and mind is socially mediated, not because our bodies affect our minds by way of social evaluations of others, but because social position affects the shape of our bodies (pp. 397–398).

Scheier et al. (1989) studied the effects of dispositional optimism on 51 middle-aged men before and after coronary artery bypass surgery. They found optimism positively correlated with postsur-

250

ALEX C. MICHALOS ET AL.

gical quality of life at six months, with problem-focused coping, faster rate of physical recovery, faster rate of return to normal life activities, satisfaction with each of health care, medical staff and the emotional support received from family and friends. As well, optimists were less hostile, less depressed and used denial as a coping mechanism less frequently than pessimists. While pessimists might benefit from extra care, the authors note that “providing the extra care that is needed may not be the easiest thing to do” because “pessimists were more hostile, more depressed, and expressed less satisfaction with the treatment they had been receiving by the medical staff. They also asked less questions of the medical team and were generally less involved in the recovery process” (p. 1039). Mai et al. (1990) studied 24 survivors of heart transplants 12 months after surgery. Remarkably, Pre-operatively, 14 had a psychiatric diagnosis and this figure had dropped to five at 12 months follow-up. Those patients without a psychiatric diagnosis preoperatively had not developed one at follow-up” (p. 223). The authors concluded that “heart transplantation in selected subjects with terminal heart disease results in a substantial improvement in psychosocial adjustment and quality-of-life 12 months following surgery (p. 223).

Angel and Thoits (1987) hypothesized that the perceived normality or abnormality of one’s physical and mental condition is a function of the similarity of one’s condition to that of most others in one’s reference group. According to these authors, the features of one’s condition that are typically noticed or overlooked at the pre-symptom stage are subtle cultural effects on health and illness identification that are largely neglected in the medical sociological literature. They also cite others who have shown that perceptions of symptom seriousness are influenced by one’s reference group. Suchman and Phillips (1958) found that elderly people tended to rate their health conditions more favourably than their physicians rated them. Mechanic and Angel (1987) cited several similar studies, and claimed that some researchers just think that elderly people are “health optimists”, while others appeal to an adaptation/expectation sort of hypothesis (e.g., Thomas, 1981). Results of their own survey of 2 431 Americans suggested that the “Subjective evaluations of health [of older people] . . . are not absolute but, on the contrary, are

HEALTH AND THE QUALITY OF LIFE

251

made in a context of self and other comparisons, and are always relative to some degree” (pp. 137–138). Schulz and Fritz (1987) compared a sample of 120 adults aged 65 years and older to a sample of 77 students from 17 to 32 years of age. Regarding eight problem areas, including poor health, both groups were more likely to attribute difficulties to people over the age of 65 than to themselves. When asked to “base their judgments on older persons they know”, both groups judged the lives of elderly persons more favourably than when basing judgments on generic older adults. Mechanic and Hansell (1987) studied the self-assessments of health of 1 057 American adolescents in 19 public schools, and found an interesting twist on the social comparison process. “Selfassessments of better health,” they wrote, “were influenced directly by higher levels of competence, as measured by grades in school and participation in sports and exercise, and by greater psychological well-being, as indicated by lower levels of depressed mood. In contrast, the effect of physical symptoms was indirect, and was mediated through initial levels of self-assessed health” (pp. 370–371). Although their analyses were unidirectional, they explicitly acknowledged their belief in “reciprocal causal relationships involving self-assessed health” and school achievement, participation in sports and exercise. Blalock et al. (1989) examined types of social comparisons made by a group of 75 women with rheumatoid arthritis. Earlier studies of breast cancer patients by Taylor (1983) and Wood et al. (1985) and of other arthritis patients by Affleck et al. (1987) and DeVellis et al. (1990) had found that most patients compared themselves to other patients who were worse off than themselves. That is, most of them behaved as predicted by downward comparison theory (Michalos, 1985). Blalock, McEvoy, DeVellis and DeVellis found that most patients compared themselves to people without the disease when they were setting standards of desired performance for themselves and they compared themselves to other patients when they were experiencing performance difficulties. In both cases satisfaction was enhanced by social comparisons, but downward comparisons were only made in the latter case.

252

ALEX C. MICHALOS ET AL.

In a later study, Blalock et al. (1990) studied the directionality of social comparisons by a group of 85 patients who had been suffering from rheumatoid arthritis for an average of eleven years each. The researchers found that nearly two-thirds of the comparisons made by their patients were either upward or lateral. “After controlling for differences in physical health status,” they “found that patients who emphasized their similarity to, as opposed to their differences from, individuals not affected by RA exhibited better psychological adjustment . . . by comparing with non-affected individuals, and focusing on one’s similarity to those individuals, patients may be able to cast their own limitations into a more positive light” (p. 368). The results of this study were therefore similar to those of Blalock et al. (1989), and to those of Schultz and Decker (1985), who worked with victims of spinal cord injuries. Buunk et al. (1990) constructed some interesting experiments showing that the directionality of a comparison is an ambiguous indicator of its impact on one’s feelings. The ambiguity arises briefly from the facts that, on the one hand, comparing oneself to someone better off (upward comparison) reminds one that one is worse off (producing negative affect) but that things may get better (producing positive affect), while on the other hand, comparing oneself to someone worse off (downward comparison) reminds one that one is better off (producing positive affect) but that things may get worse (producing negative affect). Faced with an upward or downward comparison to another person, the downside or the upside of the comparison may be more salient. Buunk et al.’s experiments with 55 cancer patients and 632 married people showed that downward positive affect comparisons were the most common (as predicted by downward comparison theory), but upward positive affect comparisons were almost as common. Furthermore, “Both studies demonstrated that comparison direction (upward or downward) is not intrinsically linked to affect, as the previous literature has often assumed” (pp. 1245–1246). Taking his lead from the World Health Organization, Brown (1989) distinguished disabled from handicapped people, with the latter defined as disabled people with a poor fit between their functioning and their social roles and environment. From this point of view, handicaps may be created or removed by altering the person-

HEALTH AND THE QUALITY OF LIFE

253

environmental fit regardless of disabilities. Among other things, Brown concluded that “Relatives and professionals tend to underestimate the ability of developmentally disabled persons to control their environment” (p. 557). According to Bryant and Marquez (1986: pp. 143–144), some studies have found that “women are more likely than men to report mental health problems such as depression, nervous breakdown and alienation, even when the effects of other demographic variables, including education, are controlled. Moreover, women are more strongly oriented toward help-seeking as a strategy for coping with mental health difficulties, whereas men are more strongly oriented toward self-help”. These authors’ own analysis of a 1976 national American sample of 960 males and 1304 females showed that men’s self-evaluations were a function of educational status but women’s were not. Several other examples of the social construction of knowledge and value were cited in Michalos 1991. All of these examples show that both sides of perceived discrepancies, what is versus something else, are thoroughly woven into and to some extent constructed by contingent historical circumstances. Burckhardt (1985) measured the quality of life of 94 adults with various forms of arthritis using an index consisting of life and domain satisfaction scores and a single item quality of life scale. She found that 46% of the variance in the index could be explained by positive self-esteem, perceived social support, internal control over health and a low negative attitude toward illness. Severity of impairment had no direct effect on the index, although it did have direct effects on self-esteem and internal control over health, which gave it significant indirect negative effects. Personal control over one’s health was also found to be an important predictor of psychosocial well-being in studies of patients with end-stage renal disease (Devins 1989), epilepsy (DeVellis, DeVellis, Wallston and Wallston 1980) and spinal cord injury (Schulz and Decker 1985). Devins and Seland (1987) speculated that it was a major factor relating to patients’ coping with multiple sclerosis, primarily through a self/wants mechanism. Gfellner (1989) studied 40 Canadians, 80 years of age and older, and found that their “subjective health perceptions were a

254

ALEX C. MICHALOS ET AL.

better index of life satisfaction than number of health conditions, instrumental limitations, or perceived functional abilities” (p. 208). Gray et al. (1985) found that the spouses of 50 patients with endstage renal disease had lower estimates of the patients’ feelings of satisfaction with themselves than the patients own perceived selfsatisfaction. Epstein et al. (1989) compared the general health status reports of 60 elderly people seeking health care service with reports about the former from 60 proxies who were family members or friends. Proxy reports were significantly positively correlated with subject reports at r = 0.64 for social activity, r = 0.62 for emotional status, r = 0.71 for overall health, r = 0.73 for functional status and r = 0.43 for satisfaction with health care service. Nevertheless, “the proxies described the subjects as less emotionally adjusted and as less satisfied [with health care] than the subjects described themselves” (p. S96). Spitzer et al. (1981) found that cancer patients rated the quality of their lives higher than their physicians rated it. Okun and George (1984) found that people’s self/rated health scores were more strongly related than physician ratings to people’s subjective well-being scores. Hurny et al. (1987) cite several studies showing that cancer patients have relatively higher scores than healthy controls on scales of social desirability and lying, which would explain some of the discrepancy between patients’ reported healthiness, satisfaction and so on and estimates by friends, family and physicians. Their own research with 41 colorectal cancer and 40 bladder cancer patients revealed that both groups had relatively higher Crowne Marlowe Scale scores of social desirability than samples in the general population. Sabourin et al. (1989) found that “self-deception and impression management accounted for only 9% of the variance in client satisfaction scores” (p. 128) for their sample of 81 former mental health outpatients in Quebec. In another study, Sabourin et al. (1989) found that both client “satisfaction reports and psychological distress scores were contaminated by socially desirable responding” (p. 352). Lomas et al. (1987) compared suggested quality of life measures of two groups (N = 9, N = 5) of language-disabled (i.e., aphasic) adults with nine clinicians, and found that the patient measures

HEALTH AND THE QUALITY OF LIFE

255

were more similar to each other than to the clinician mea- sures. The patient measures had relatively more specific items with greater focus on social needs than the clinician measures. Pearlman and Jonsen (1985) found that physicians’ assessments of the quality of life of patients “demonstrated marked variability”. In developing his quality adjusted life years scale, the health economist Alan Williams (1985) found that doctors“ “tended to give lower scores than did other people”, presumably because doctors had “a much greater aversion to disability and distress than the population at large” (quoted by Mulkay et al., 1987, p. 554). The latter authors make an excellent case for the claim that “quality of life is not a measurable phenomenon out there in the social world, but an interpretative by-product of the social interaction between ordinary people, doctors, NHS [National Health Service] managers and, most important of all in this context, health economists” (p. 559). The relative impact of one’s objective health status on one’s subjective well-being is more modest than one might expect even in the presence of the most catastrophic circumstances, e.g., following spinal cord injury (Mayer and Andrews 1981; Chwalisz et al., 1988); for multihandicapped adults (Wacker et al., 1983; Stensman 1985; Ramund and Stensman 1988); for cancer patients (Haes and Van Knippenberg 1985); for the visually impaired (Gillman et al., 1986); and for chronic mental patients (Lehman et al., 1982 and Lehman, 1983). Again, Haes and Van Knippenberg’s (1985) review of research on the quality of life (QL) of cancer patients found no differences with respect to most QL indicators: satisfaction with family, friends, work, income, values, activities, community, local government, health and the overall quality of life . . . , psychological functioning . . . , anxiety, depression, positive well-being, mental and general well-being . . . , daily activities . . . and work rehabilitation . . . . Mastectomized patients do not differ from benign controls with respect to quality of life . . . . The emotional well-being of melanoma patients being treated is equal to the well-being of “normals” . . . . No difference was found between chemotherapy patients and “normals” with respect to psychological complaints, satisfaction with life as a whole and care from partner and others . . . . It is remarkable that comparisons between cancer patients and others do not seem to support the assumption that the QL of cancer patients in general is poorer than the QL of other groups (p. 811).

256

ALEX C. MICHALOS ET AL.

Chwalisz et al. (1988) compared the emotional experiences of 23 nonhandicapped students to 18 spinal-cord-injured people and 14 other people who were wheelchair-bound for other reasons. On average, the student group was six years younger than the other two groups. The six primary emotions measured were joy, love, sadness, anger, sentimentality and fear. Although satisfaction with their own health is not measured by any of these emotions, one would have thought that the great divergence in the health status of the student group versus the others would make a difference to their emotional status. However, “There were no differences in subjective well-being between the handicapped and nonhandicapped groups in this study . . . . Thus, there appears to be no direct relation between severity of one’s handicap and subjective well-being” (p. 826). Remarkably, The spinal-cord-injured persons with the most limited bodily feelings often reported feeling several emotions more intensely than before their injury. A majority of the spinal-cord-injured subjects in the low-feedback group reported increases in the intensity of feelings of love, joy, sentimentality, and sadness in their lives. Furthermore, spinal-cord-injured persons significantly more often reported increases in the intensity of fear than those in the other groups (p. 825).

Johnson et al. (1982) compared four groups of patients with endstage renal disease and found that those on hemodialysis or with successful transplants reported levels of life quality similar to those of the general population, while those with unsuccessful transplants reported relatively lower levels of life quality than all other groups. Binik and Devins (1986–87) studied 150 patients with end-stage renal disease and could not replicate the Johnson, McCauley and Coply results. They concluded that their “results indicated no significant differences in levels of pain, depression, normal mood states, perceived intrusiveness, or perceived control among dialysis and transplant patients associated with a history of transplant failure. In fact, these and other analyses we have performed . . . consistently have failed to identify systematic differences in quality of life data among any dialysis as compared to transplant patient groups” (p. 288). Chuang et al. (1989) assessed the psychosocial distress of 24 gay or bisexual men with asymptomatic human immunodeficiency virus (HIV) infection, 22 with acquired immune deficiency syndrome

HEALTH AND THE QUALITY OF LIFE

257

(AIDS)-related complex and 19 with AIDS. Among other things, they found that the men with AIDS were less distressed than those in the other two groups, but there were no significant differences in levels of reported happiness. Callan and Hennessey (1989) compared the lives of 52 women with explained infertility, 25 women with unexplained infertility and 54 mothers, and found no significant differences among the groups in reported levels of self-esteem and happiness. In another study, Callan and Hennessey (1988) found that infertile women reported lower levels of life satisfaction than mothers who had subsequently become infertile. Granting that the overall impact of objectively measurable health problems on one’s subjective well-being may be modest, it is nevertheless negative and can be very debilitating for some people. For example, Gillman et al. (1986) reported that for their sample of 486 public housing residents over 60 years of age, “Thirty percent of the visually impaired said their lives were unpleasant, boring and worse than their friends; they stated that they felt useless and trapped and generally were less satisfied with their lives than their sighted counterparts” ( p. 590). Kibele (1989) mentioned several studies showing that “adults with significantly limiting disabilities including cerebral palsy cited a low degree of life satisfaction, exemplified by difficulty obtaining regular medical care, coping with sexuality, and achieving independence” (p. 371). Klonoff et al. (1986) worked with 71 patients with closed-head injury, two to four years postinjury, and reported that “quality of life is adversely affected by increased severity of head injury and greater residual motor deficits” (p. 469). Similar findings were reported by Lezak (1987) based on his study of 42 white male volunteers with traumatic head injuries. Magilvy’s (1985) research with 66 hearing-impaired older women found that those “women experiencing a high degree of hearing handicap had a lower perception of quality of life” (p. 143). Several studies have found that hearing impairment has a negative impact on interpersonal interaction, self-concept and morale, and is positively associated with depression, bitterness, hypochondrial complaints, resentment, hostility and feelings of inferiority (Cooper 1976; Blackwell and Levy 1987).

258

ALEX C. MICHALOS ET AL.

Shadish et al. (1985) reported that their sample of 74 nursing home patients were significantly less happy than a national random sample of Americans (N = 2 225), a random sample from Chicago (N = 252), a sample of unemployed males from a small town in Iowa (N = 390) and a sample of schizophrenics (N = 178). However, the mean reported happiness level of the nursing home patients was not significantly different from that of a sample of urban renewal black people (N = 614). Blodgett (1981–82) reviewed 19 studies of hemodialysis patients and noted that the “median incidence of emotional maladjustment was slightly less than 50%” (as quoted by Nadel and Clark, 1986: p. 47). Nadel and Clark’s own research with 24 patients who had undergone more than one kidney transplant indicated that about 50% of them had “life satisfaction ratings of ‘good’ and ‘very good’ ” (p. 41). Query et al. (1990) interviewed 36 patients who had shunts for hydrocephalus between 14 and 36 years earlier, their parents and siblings, and reported that while “most judged the quality of their family relationships and lives to be good”, “the families of patients with continuing or ultimately fatal problems reported the most negative effects” (p. 126). One of the first models used to explain overall quality of life operationalized as life satisfaction was a simple linear model. In this model the dependent variable is satisfaction with life as a whole and the independent variables are some measures of satisfaction in particular domains. For example, in Michalos and Zumbo (1999) we summarized results of seven studies in which life satisfaction and happiness scores were regressed on scores for satisfaction with one’s own health, financial security, family relations, job, friendships, housing, area lived in, recreation activities, religion, self-esteem, transportation, government services, living partner and formal education. In the context of these predictors, satisfaction with one’s own health had average standardized regression coefficients (beta values) of 0.10, with a range of 0 to 0.18, when the dependent variable was life satisfaction and when it was happiness. This simple linear model was also used, among other things, in Michalos (1991, 1991a, 1993, 1993a). In this project samples of

HEALTH AND THE QUALITY OF LIFE

259

undergraduates were drawn from 58 universities in 39 countries, including about 250 students in each sample. (Taiwan was a special case in which all 26 universities on the island were sampled.) Applying this model to a combined sample of about 5000 students, it was found that satisfaction with one’s health had a beta value of 0.11 when the dependent variable was life satisfaction and 0.08 when the dependent variable was happiness.

SF-36 AND PREVIOUS RESEARCH EXPLAINING QUALITY OF LIFE

The review in the previous section clearly reveals that a wide variety of alternative instruments have been used to measure health as an independent variable to explain and predict the quality of people’s lives operationalized primarily as overall life satisfaction or happiness. Thanks to the pioneering work of John Ware and his colleagues, a multi-item scale of overall healthiness has been developed and has been shown to have good reliability and validity (Ware and Sherbourne, 1992; McHorney et al., 1993; McHorney et al., 1994; Ware et al., 1993; Ware et al., 1994). Indeed, the 36-item Medical Outcomes Study Short Form (SF-36) profile has achieved such a high status in the research community that it is sometimes referred to as the ‘gold standard’ for health status measurement and its norms for several populations (e.g., American, Australian, British, German) provide useful benchmarks for other developed countries. That is why it was included in our questionnaire. Exhibit I summarizes its basic elements. McHorney et al. (1994) reported results of a panel study of 3 445 patients with chronic medical and psychiatric conditions which, among other things, allowed researchers to assess the relative amount of difficulty different kinds of patients would have completing the questionnaire. They found that Younger patients, those better educated, and MI [myocardial infarction] patients had the highest rates of item completeness, while the worst groups were the elderly, the less educated, and the poor (p. 54).

With the exception of MI patients, the groups having more or less difficulty with the SF-36 questionnaire were precisely the groups

260

Exhibit I Content of the SF-36 Questionnaire Number Meaning of low scores of items

Physical Functioning

10

Role Physical

4

Bodily Pain General Health

2 5

Vitality 4 Social Function- 2 ing Role Emotional 3 Mental Health 5 Reported Health 1 Transition

Meaning of high scores

Limited a lot in performing all physical activities Performs all types of physical activities includincluding bathing or dressing due to health. ing the most vigorous without limitations due to health. Problems with work or other daily activities as a result No problems with work or other daily activities of physical health. as a result of physical health. Very severe and extremely limiting pain. No pain or limitations due to pain. Evaluates personal health as poor and believes it is Evaluates personal health as excellent. likely to get worse. Feels tired and worn out all of the time. Feels full of pep and energy all of the time. Extreme and frequent interference with normal social Performs normal social activities without interactivities due to physical or emotional problems. ference due to physical or emotional problems. Problems with work or other daily activities as a result No problems with work or other daily activities of emotional problems. as a result of emotional problems. Feelings of nervousness and depression all of the time. Feels peaceful, happy, and calm all of the time. Believes general health is much better now than one Believes general health is much worse now than year ago. one year ago.

Source: Ware et al., 1993: p. 3:5.

ALEX C. MICHALOS ET AL.

Concepts

HEALTH AND THE QUALITY OF LIFE

261

that have more or less difficulty with most printed questionnaires. The internal-consistency reliability (Cronbach’s alpha) for each of the eight concepts measured in the SF-36 with this panel ranged from 0.78 to 0.93, with a mean of 0.85. Using a random sample of 2 474 Americans, Ware et al. (1993: p. 9:24) correlated scores from the eight scales of SF-36 with scores on the General Psychological Well-Being (GPWB) measure taken from Dupuy (1984) and they found statistically significant positive zero-order correlations running from 0.19 to 0.60. The highest correlation, 0.60, was with the Mental Health scale, which one might have expected given the content of the two measures. The GPWB is a 22-item scale covering six dimensions: anxiety, depression, general health, positive well-being, self-control and vitality. Respondents are asked to rate their lives “during the last month” on six-point scales ranging from “extremely happy, could not have been more satisfied or pleased” to “very dissatisfied or unhappy most of the time”. Naughton et al. (1996: pp. 122–123) indicated that the GPWB typically has Cronbach’s alpha values in the 0.90 to 0.94 range. In our view, the GPWB scrambles so many different things together that it is difficult to know exactly what it is measuring, although it does have reasonable levels of concurrent validity with well-known measures like the Beck Depression Inventory. For our purposes, the SF-36 scale works best as a measure of a health profile with eight dimensions, rather than as a single summative measure. Ware et al. (1993: p. 12:4) explained that they had tried several combinations of the eight scales and that the best combination contained only the two summary indexes for Physical Functioning and Mental Health. Furthermore, they reported that Others have experimented with a summary measure based on a simple summation of the eight SF-36 scales (Katz et al., 1992), which we do not recommend. MOS [Medical Outcomes Study] studies to date indicate that this summary index is: (1) less sensitive to differences in health status (compared with SF-36 scales and Summary Indexes), and (2) not specific (i.e., not interpretable) in relation to an underlying health concept. . . . [Katz et al.] . . . observed that their summary score for the SF-36 scales was no worse than other summary measures studied, including those for the SIP [Sickness Impact Profile, Bergner et al., 1981], sAIMS [shortened Arthritis Impact and Measurement Scales, Katz et al., 1992] and FSQ

262

ALEX C. MICHALOS ET AL.

[Functional Status Questionnaire, Katz et al., 1992]. The latter results may reflect problems with summary indexes in general (p. 12:4).

Ware et al. (1994: p. 2:1) claimed that further research showed that the two indexes for Physical Functioning and Mental Health accounted for “from 80 to 85 percent of the reliable variance in the eight SF-36 scales” and that “this result suggests that psychometrically-based summary measures have the potential to reduce the number of statistical comparisons required in analyzing SF-36 data from eight to two without substantial loss of information”. This literature suggests that if one wishes to use summary indexes then one should use two (PF and MH). However, the eight dimensions were consistently found to be better than the two summary indexes. Fryback et al. (1997) used scores on the eight SF-36 scales to predict Quality of Well-Being (QWB) scores, with a sample of 1 430 Americans in the Beaver Dam Health Outcomes Study. They found that “A six-variable regression equation drawing on five of the SF36 components predicted 56.9% of the observed QWB variance” (p. 1). The QWB scale measures levels of mobility, physical and social activity, and requires a trained interviewer to administer it. It was developed by Balaban et al. (1986) following work on the Index of Well-Being developed by Kaplan et al. (1976). Shmueli (1998) reported results of a 1993–1994 survey of 2 030 urban Jewish Israelis aged 45 to 75 years in which, among other things, scores from the eight scales and a summary score for SF-36 were used to explain overall health-related quality of life (HRQoL). The HRQoL scale was based on responses to the question “Imagine that health related quality of life could be graded on a numeric scale. The number 100 represents the quality of life of a person in perfect health and the number 0 represents death. How would you rate your health related quality of life during the last month?” Forty-five percent of the variance in HRQoL scores was explainable from the summary SF-36 scores, and 49% was explainable from the scores on the eight scales used as separate predictors. Shmueli (1998: p. 194) provided the following instructive overview of his findings. The scores on the eight SF-36 scales are not equally important to the HRQoL score. It is the more holistic scales – general health and vitality – that are the main predictors of the HRQoL scale. The scales measuring ‘roles performance’ have no

HEALTH AND THE QUALITY OF LIFE

263

net effects on quality of life. Not considering the general health and vitality scales, the PH [physical health = the mean of the physical functioning, role-physical and bodily pain scales] and MH [mental health = the mean of the mental health, roleemotional and social functioning scales] scales (and the six individual scales on which these summary scales are defined) are equally important in the determination of HRQoL. These results reiterate the importance of the distinction between measures of health which are validated by clinical considerations and those validated by more subjective measures of health related quality of life. Items which are powerful in classifying individuals into different clinical diagnosis groups might have no relation to different levels of self-reported quality of life (such as the roleperformance items), while items that are highly correlated with subjective quality of life might be weak in their clinical classification of persons (such as vitality or general health reports).

SAMPLING TECHNIQUE AND QUESTIONNAIRE

Twenty-five hundred 10-page questionnaires were mailed to a simple random sample of Prince George’s 34 000 households in November 1998. The first three and a half pages contained the SF-36 items. These were followed by a check list of 18 chronic conditions (heart disease, diabetes, etc.) and a single item about frequency of worries about one’s own health. The twenty-item Center for Epidemiologic Studies Depression (CES-D) scale (Radloff 1977) came next. It was designed to measure frequency of depressive symptoms in broad-based survey research populations. Seven social support items, followed by 12 items measuring life-stress in general and 11 items measuring stress at work came next. There were 5 items on smoking and drinking and 10 dealing with assessments of health care services in Prince George. The latter included open-ended questions asking about the best and worst things about health care services here, and about changes that could be made to improve these services. There was a page of questions about respondents’ satisfaction with particular domains of their lives (e.g., their family relations, jobs and housing), their overall happiness and satisfaction with the overall quality of their lives. The questionnaire ended with an item measuring subjective age (i.e., whether people felt older or younger than they actually were) and two pages of demographic questions.

264

ALEX C. MICHALOS ET AL.

SAMPLE CHARACTERISTICS

By the end of December 723 (29%) useable questionnaires were returned, which formed the working data-set for the survey. Of the 719 respondents who identified their gender, 49% (354) were females. Sixty-five percent (465) of those who answered the marital status question were now married. The average age was 46, and the range ran from 17 to 92. Two hundred and twenty-two (30%) had some university education, with 154 (21%) holding a university degree. Four hundred and sixteen (58%) were employed full-time and 87 (12%) part-time. Two hundred and one (29%) had total family incomes of from $40 000 to $69 999, with the median in the $50 000 to $59 999 category. According to the 1996 census, which is the last official full population count available from Statistics Canada, the average age of Prince George residents over 18 years of age was 41, 49% of the city’s adult residents were female, 50% were married and living with their spouse, 17% had some university education, 9% held a university degree and 66% were employed. The 1995 median family income for Prince George was $56 125. Broadly speaking, then, our working sample is a bit older, and has a few more married and university educated people than the population from which it was drawn. Since health tends to be positively correlated with education and negatively correlated with age, there may be some averaging effect from the tendency of our health-related figures to be inflated resulting from the former and deflated resulting from the latter.

ASPECTS OF HEALTH

In the next three exhibits we summarize the eight SF-36 profile scores for our respondents taken all together and for males and females, and compare their scores with normative scores from the United States, United Kingdom, Netherlands, Aberdeen (Scotland) and Sweden. SF-36 profile scores are designed such that 0 represents the worst and 100 represents the best state of health. When Prince George scores are compared to those of the UK and the Netherlands, the sample is reduced to include only people in the same age range as those surveyed in those countries.

265

HEALTH AND THE QUALITY OF LIFE

Exhibit II SF-36 Comparisons of Prince George Scores with Country and City Normsa Place

P.G.

USA

UK

Netherlands

Aberdeen

Sweden

PF RP BP GH VT SF RE MH

87.2 81.8 60.9 73.7 58.4 83.5 80.7 74.5

84.2 81 75.2 72 60.9 83.3 81.3 74.7

91.8 (89.1)b 88.8 (83.5) 83.7 (61.8) 75.3 (74.4) 61.7 (58.4) 88.5 (83.8) 82.1 (80.5) 73.1 (73.8)

88.9 (89.3) 82.0 (83.9) 81.5 (61.9) 70.6 (74.2) 65.6 (58.0) 85.1 (83.8) 81.6 (80.3) 74.1 (73.7)

79.2 76.5 76.9 68.7 61.2 78.6 75 73.7

87.9 83.2 74.8 75.8 68.8 88.6 85.7 80.9

a Prince

George N = 709, ages 18–92; USA N = 2474, ages 18 and older; UK N = 9332, ages 16–64; Netherlands N = 4041, ages 20–59; Sweden N = 8930, ages 15 and older; Aberdeen N = 542, ages 18 and older. b Numbers in parentheses are Prince George values using same age-range as other countries. Sources: Ware et al. (1993); Jenkinson et al. (1993); Han et al. (1998); Sullivan et al. (1995); Garratt et al. (1993).

Exhibit II lists the scores for our respondent group as a whole. The scores range from 87.2 for Physical Functioning to 58.4 for Vitality, with a mean of 75.1. This mean score is lower than those of four of the other five groups. Only Aberdeen has a lower mean score, at 73.7. If we take these means of the eight scores as broadly indicating the health of the six groups listed in Exhibit II, then it is fair to say that our respondents are relatively less healthy than those in four of the other five groups. Comparing the eight component scores for Prince George and the USA, we find that the American scores are superior on four components (Bodily Pain, Vitality, Role Emotional and Mental Health) and the Prince George scores are superior on the other four (Physical Functioning, Role Physical, General Health and Social Functioning). All of the UK and Swedish component scores are superior to the Prince George scores (except for MH for the UK), while five of the Netherlands scores and only two of the Aberdeen scores are superior to those of Prince George. Exhibit III lists the scores for males in the four groups for which we had data. The scores for Prince George male respondents range

266

ALEX C. MICHALOS ET AL.

Exhibit III SF-36 Comparisons of Prince George Male Scores with Country Normsa Place

P.G.

USA

UK

Netherlands

PF RP BP GH VT SF RE MH

87.1 83.6 61.9 74 60.2 85.3 81.5 75.5

87.2 86.6 76.9 73.5 63.6 85.2 83.3 76.4

91.8 (88.9)b 88.8 (85.7) 83.7 (62.9) 75.3 (74.8) 61.7 (60.1) 88.5 (86.0) 82.1 (81.8) 73.1 (74.9)

90.3 (89.0) 84.7 (85.7) 83.5 (63.1) 70.8 (74.6) 67.9 (59.7) 86.9 (86.0) 85.0 (81.6) 76.2 (74.8)

a Prince

George N = 361, ages 18–92; USA N = 1055, ages 18 and older; UK N = 9332, ages 16–64; Netherlands N = 4041, ages 20–59. b Numbers in parentheses are Prince George values using same age-range as other countries. Sources: Ware et al. (1993); Jenkinson et al. (1993); Han et al. (1998).

from 87.1 for Physical Functioning to 60.2 for Vitality, with a mean of 76.1. Again, this mean is lower than that of the three country means for males, although the range is wider than that of the USA and the Netherlands and narrower than that of the UK. Comparing the eight component scores for Prince George male respondents and the American male scores, we find that the former scores are superior on only General Health. All of the UK male scores are superior to the Prince George male scores (except MH) and only two of the latter are superior to the Dutch male scores (Role Physical and General Health). Exhibit IV lists the scores for females in the four groups. The scores for Prince George female respondents range from 87.3 for Physical Functioning to 56.5 for Vitality, with a mean of 74. So, on average our female respondents are less healthy than our males. The mean female score is also lower than that of the other three groups, while the range is wider than all the others. Comparing the eight component scores for Prince George female respondents and the American female scores, we find that only two of the latter scores are superior to those of the former (Bodily Pain and Vitality). In fact, the very low level of the Prince George scores

267

HEALTH AND THE QUALITY OF LIFE

Exhibit IV SF-36 Comparisons of Prince George Female Scores with Country Normsa Place

P.G.

USA

UK

Netherlands

PF RP BP GH VT SF RE MH

87.3 79.8 59.7 73.4 56.5 81.7 79.8 73.5

81.5 77.8 73.6 70.6 58.4 81.5 79.5 73.3

91.8 (89.3)b 88.8 (81.3) 83.7 (60.7) 75.3 (74.0) 61.7 (56.7) 88.5 (81.7) 82.1 (79.2) 73.1 (72.6)

87.6 (89.6) 79.7 (82.1) 79.7 (60.6) 70.4 (73.9) 63.5 (56.3) 83.5 (81.7) 80.4 (79.0) 72.2 (72.5)

a

Prince George N = 348, ages 18–92; USA N = 1412, ages 18 and older; UK N = 9332, ages 16–64; Netherlands N = 4041, ages 20–59. b Numbers in parentheses are Prince George values using same age-range as other countries. Sources: Ware et al. (1993); Jenkinson et al. (1993); Han et al. (1998).

on these two components is largely responsible for the low overall mean score. All of the UK female scores and four of the Dutch female scores (Bodily Pain, Vitality, Social Functioning and Role Emotional) are superior to Prince George female scores. Scores on the twenty-item CES-D scale provide a narrower view of the health of our respondents, a view focused on depressive symptoms. Items measure “depressed mood, feelings of guilt and worthlessness, feelings of helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance” (Radloff, 1977). Scores range from 0 representing no symptoms to 60 representing significant depressive symptoms. Barnes et al. (1988) reviewed fourteen studies in which the CES-D scale was used in broad surveys and they found that the most frequently used cut-point for an indication of depression was a score greater than or equal to 16. For our respondent group as a whole, scores ranged from 0 to 54, with a mean of 10.3. For males, scores ranged from 0 to 50, with a mean of 9.5, while for females, scores ranged from 0 to 54, with a mean of 11.1. About 84% of males had scores lower than 16, compared to 73% of females. Apparently, then, our female respondents experienced more depressive symptoms than our male

268

ALEX C. MICHALOS ET AL.

respondents. This gender difference is consistent with results from the National Population Health Survey 1994–1995, for the Northern Interior Health Region, British Columbia and Canada (Healthy Communities Committee 1997: p. 36). Barnes et al. (1988: p. 390) reported mean scores of 8.6 for the 231 males and 9.8 for the 290 females in their Winnipeg Area Study of 1983. Lowinsohn et al. (1997: p. 281) reported mean scores of 7.9 for the 419 men and 8.7 for the 586 women in their American sample of people fifty years of age and older in the 1982–1984 period. Compared to these two surveys, then, our results indicate that our respondents have higher mean scores or report more depressive symptoms. Another narrow view of health is provided by our eleven-item Stress Index. Respondents simply indicate ‘yes’ or ‘no’ to a battery of stress-related items (e.g., ‘Too much is expected of you by others.’). Scores ranged from 0 to 10 for the whole group, with a mean of 2.1. The Cronbach reliability coefficient alpha for the index was 0.74. The Stress Index items were followed by a single item asking people how stressful their lives were, all things considered, on a nine-point scale from ‘very unstressful’ to ‘very stressful’. The group mean score on this item was 4.8, which indicated their lives were ‘a little stressful’. Body Mass Index scores provided an additional aspect of health. These were obtained by dividing individuals’ weight in kilograms by height in meters squared. The calculations include people aged 20–64 and exclude pregnant women. Exhibit V summarizes results from our survey and provides comparisons with the National Population Health Survey 1994–1995 for Canada and a special sample of Prince George taken at the same time. The latter two columns indicate practically no change in the Prince George scores over the period from the 1994–1995 survey to ours, and both sets of figures show that we have relatively more excessively heavy people here than in the country as a whole. Exhibit VI gives the numbers and percentages of respondents reporting chronic health conditions. Slightly more than 67% reported some chronic health condition with over 37% reporting more than one chronic condition. The relatively high percentage

269

HEALTH AND THE QUALITY OF LIFE

Exhibit V Body Mass Index Scores Weight type

Canada 1994–95

P. G. 1994–95

P.G. 1998

Insufficient Acceptable Some Excess Overweight Total N =

7.6 43.3 23.9 25.2 12318

4.7 37.5 29.9 27.9 706

4.9 37.8 29.9 27.7 617

Exhibit VI Numbers and Percentages of Respondents with Chronic Health Conditions Number of conditions

Respondents

Percentages

No chronic problems One problem Two problems Three problems Four problems Five problems Six problems Seven problems

238 215 123 76 38 24 7 3 724

32.9 29.7 17 10.5 5.2 3.3 1 0.4 100

of people reporting chronic health conditions is to some extent a function of the number of conditions in our checklist. These included heart disease, hypertension, diabetes, epilepsy, back problems, bronchitis/emphysema, stroke, arthritis, asthma, headache, thyroid problems, cirrhosis of the liver, cancer, HIV/AIDS, sinusitis, dementia/Alzheimer’s, bowel problems, hearing loss/tinnitus, other. We obtained another indirect measure of health from our question about how frequently respondents worried about their health, on a five point scale from ‘almost never’ to ‘almost always’. Sixtysix percent reported that they ‘almost never’ or only ‘occasionally’

270

ALEX C. MICHALOS ET AL.

worried about their health, and another 24% said they worried about it ‘sometimes’. The mean score on this item was 2.1.

SMOKING, DRINKING, SOCIALIZING AND FEELING YOUR AGE

Seventy-seven percent of our respondents indicated that they ‘never’ smoke, compared to 9% who smoke ‘hourly’ and another 8% who smoke ‘daily’. The percentage of non-smokers in our survey of June 1998 was 73%. So, given our typical error margins, there was probably no change in the few months from the summer to the fall survey. About 69% of our respondents said they were ‘never’ exposed to second-hand smoke in their homes, compared to only 9% who said they were never exposed to second-hand smoke outside their homes. Seventy-four percent said they were at least ‘occasionally’ exposed to second-hand smoke outside their homes. Forty-two percent drank alcoholic beverages only once a month or less, with 12% saying they ‘never’ drank such beverages. Of the 88% who drank such beverages at all, 71% said that on average they had one or two drinks per sitting. Just about half of our respondents (49%) said they belonged to voluntary organizations or associations, such as school groups or church social groups. Of those who belonged to such organizations, 46% said they participated in activities sponsored by the groups ‘at least once a week’ and another 32% said they participated ‘at least once a month’. Apart from special occasions such as weddings and funerals, 65% of all our respondents said they ‘never’ attended religious services or meetings in the past 12 months. We combined responses to our four questions indicating that respondents had someone who they ‘could confide in’, ‘count on in a crisis’, ‘count on for advice’ and who ‘made them feel loved and cared for’ into an Index of Social Support. Scores ranged from 0 to 4, with a mean of 3.6. The Cronbach alpha for the Index was 0.75. Evidently, most of our respondents had good social support. In fact, nearly 80% of them answered ‘yes’ to each of the four questions in the Index. Regarding respondents’ age identity, while 11% said that ‘physically’ they felt ‘much younger’ than their age, only 2% said they felt ‘much older’ than their age. Nineteen percent felt ‘mentally’ ‘much

HEALTH AND THE QUALITY OF LIFE

271

younger’ than their age, compared to 3% who felt ‘much older’. Eight percent believed they ‘look’ ‘much younger’ than their age, but only 0.1% believed they look ‘much older’. Thirteen percent ‘would choose to be much younger’ than they are, while only 0.4% ‘would choose to be much older’.

HEALTH CARE ASSESSMENTS

As usual, most respondents did not offer any suggestions about the worst things, best things and things to change to improve the health care in Prince George. Regarding the worst things, 128 people mentioned the lack of available specialists and 70 mentioned the excessively long waiting times for surgery. Ninety-two people mentioned the general availability or accessibility of health care as the best thing about it and 40 mentioned walk-in clinics. By far some feature of the Prince George Regional Hospital was most often mentioned as needing improvement in order to improve health care services here. For examples, 131 people recommended things like building a new hospital or renovating the current one, updating and increasing equipment and services, allowing more operating and emergency room time, and increasing the number of available beds. Eighty-one people suggested bringing in more specialists and another 56 just wanted more doctors.

QUALITY OF LIFE ASSESSMENTS

Exhibit VII summarizes the results of three Prince George surveys in which the same battery of questions was asked about the quality of life in general and in specific domains of life. Generally speaking, for random samples of our sizes (about 700 respondents), score differences of 0.3 percentage points or greater are statistically significant at a 95% level of confidence. From June 1994 to November 1998 there were 7 significant changes in the 17 indicators, and all 7 were negative. Satisfaction with jobs and financial security deteriorated, perhaps as one might have expected given the economic downturn in this period. However, satisfaction with health, religion, recreation activities, self-esteem and the provincial government also

272

ALEX C. MICHALOS ET AL.

Exhibit VII Respondent Mean Quality of Life Scores: 1994, 1997, 1998

Domains of life

1998 Nov.

1997 Nov.

1994 June

Your house, apartment, mobile home Your neighbourhood as a place to live Your family relations, generally Your living partner Your job Your life as a whole Your friendships Your health Your religion or spiritual fulfilment Your financial security Your recreation activities Your self-esteem Federal government officials Provincial government officials Local government officials Your overall quality of life Your overall happiness Mean Score

5.8 5.9 5.7 5.9 5.1 5.5 5.6 5.3 5 4.3 4.7 5.2 2.8 2.2 3.4 5.5 5.6 4.9

5.5 5.3 6 6.3 5.4 5.8 5.8 5.4 5.3 4.6 4.9 5.5 2.8 2.5 3.6 5.6 5.5 5.1

5.8 5.7 5.9 6.1 5.4 5.6 5.8 5.6 5.4 4.7 5 5.5 3 3 3.5 5.7 5.6 5.1

deteriorated. From November 1997 to November 1998, there were 10 significant changes in the 17 indicators, 8 of which indicated some deterioration. The only two positive changes were for levels of satisfaction with people’s homes and neighbourhoods. While there were no significant changes in the mean scores for the global indicators of happiness and satisfaction with the overall quality of life in any of the surveys, there was a significant negative change in reported satisfaction with life as whole from the 1997 to the 1998 survey. The greatest deterioration occurred for satisfaction with the provincial government, which dropped from 3.0 in the summer of 1994 to 2.2 in November 1998.

HEALTH AND THE QUALITY OF LIFE

273

Exhibit VIII Regression of Health Satisfaction Scores on Health Scores Explanatory Variables

Beta Values

General Health Vitality Body Mass Index Health Worries Stress Index Explained Variance

0.51 0.17 –0.06 –0.14 –0.09 56%

EXPLAINING RESPONDENTS’ SATISFACTION WITH THEIR HEALTH

As the previous exhibit reveals, mean scores for health satisfaction ranged from 5.6 in the summer of 1994 to 5.3 in the current survey. This is the first time we have measured health itself in order to explain its impact on satisfaction with health. We applied a stepwise regression procedure using as explanatory variables scores from the eight components of SF-36, CES-D, Stress Index, life stress item, Body Mass Index, numbers of chronic conditions and worries about health. Results of the regression are summarized in Exhibit VIII. Our health scores were able to explain 56% of the variation in health satisfaction scores. The General Health items from the SF-36 clearly dominated the set of significant explanatory variables, with a standardized regression coefficient (Beta value) of β = 0.51. The Vitality items had a relatively small impact, β = 0.17, and scores for Body Mass, Health Worries and Stress were appropriately negative at β = –0.06, β = –0.14 and β = –0.09, respectively. Using the Pratt measure (Thomas et al., 1998), we found that the General Health items accounted for 65% of all the variance explained by our variables.

274

ALEX C. MICHALOS ET AL.

EXPLAINING THE IMPACT OF HEALTH AND DOMAIN SATISFACTION ON HAPPINESS

In the next three sections we make similar analyses explaining in turn our global indicators of happiness with life as a whole, satisfaction with life as a whole and satisfaction with the overall quality of life. In this section, first we regress happiness scores on our battery of health scores. Second, we regress happiness scores on our battery of domain satisfaction scores. Third, we apply a hierarchical regression procedure entering the block of health scores first and then the block of domain satisfaction scores. Fourth, we reverse the entry of the two blocks, and finally, we enter all the explanatory variables at once. While the beta values of distinct variables do not change as a result of being entered in a block regardless of the order in which the block is entered, the percent of variance explained by the block may change considerably. Accordingly, we will review the impact on the explanatory power of the two blocks relative to their order of entry, but only the results of the first two and fifth regression analyses are summarized in Exhibit IX. The first column of Exhibit IX shows that our battery of health scores accounted for 45% of the variation in happiness scores, with CES-D scores having the greatest influence(Pratt = 45%), followed by the SF-36 Mental Health items (Pratt = 33%). Only two other scores found their way into the equation, the Stress Index score (Pratt = 16%) and the SF-36 General Health items (Pratt = 6%). The second column shows that our battery of domain satisfaction scores accounted for 40% of the variation in happiness scores, with self-esteem satisfaction dominating the set of predictors (Pratt = 45%), followed at some distance by satisfaction with family relations (Pratt = 18%), financial security (Pratt = 18%), and living partner (Pratt = 16%). Given the number of apparently more personal domains in our battery that failed to have any influence on happiness, we found it remarkable that satisfaction with the provincial government should have had a positive impact. The satisfaction with self-esteem and with the provincial government had Pratt indices of about 4% each. The third column shows that the two batteries together accounted for 53% of the variation in happiness scores, with CES-D scores having the greatest influence (Pratt = 31%), followed by self-esteem

275

HEALTH AND THE QUALITY OF LIFE

Exhibit IX Regressions of Happiness Scores on Health and Domain Satisfaction Scores Explanatory variables

Beta values Beta values Beta values

General Health Mental Health CES-D Stress Index Role Physical Vitality Role Emotional Chronic Cond. Family Rel. Sat. Living Partner Sat. Financial Sec. Sat. Self-esteem Sat. Prov. Gov. Sat. Fed. Gov. Sat Present Age Sat. Variance Explained a = variable not in equation

0.1 0.24 –0.32 –0.16 a a a a a a a a a a a 45%

a a a a a a a a 0.17 0.16 0.15 0.33 0.13 a a 40%

a 0.14 –0.26 –0.10 0.06 –0.04 0.07 0.04 0.10 a a 0.17 a 0.07 0.14 53%

satisfaction (Pratt = 18%). The seven health scores accounted for 60% of the total variance explained and the four domain satisfaction scores accounted for the remaining 40%. Results of the two hierarchical regressions indicated that if the block of health scores was entered first, it accounted for 46% of the variance in happiness scores, leaving only 7 percentage points to be added by the block of domain satisfaction scores. However, if the block of domain satisfaction scores was entered first, it accounted for 40% of the variance in happiness scores, leaving 13 percentage points to be added by the block of health scores. In fact, both procedures provided a misleading impression of the relative importance of health and domain satisfaction to happiness by exaggerating the importance of the first block entered into the equation. That

276

ALEX C. MICHALOS ET AL.

was revealed admirably by their Pratt measures in the previous paragraph.

EXPLAINING THE IMPACT OF HEALTH AND DOMAIN SATISFACTION ON SATISFACTION WITH LIFE AS A WHOLE

The first column of Exhibit X shows that our battery of health scores accounted for 47% of the variation in life satisfaction scores, with CES-D scores having the greatest influence(Pratt = 46%), followed at some distance by the Stress Index scores and the SF-36 Mental Health items (Pratt = 22% each). Only three other scores found their way into the equation, SF-36 General Health (Pratt = 6%), the life stress item (Pratt = 4%) and Physical Functioning (Pratt less than 1%). The second column shows that our battery of domain satisfaction scores accounted for 65% of the variation in life satisfaction scores, with self-esteem satisfaction dominating the set of predictors (Pratt = 24%), followed fairly closely by satisfaction with friendships (Pratt = 16%) and living partner (Pratt = 12%). Six other domains also had some influence, namely, job satisfaction, satisfaction with present age and health, financial security and religion. The third column shows that the two batteries together accounted for 68% of the variation in life satisfaction scores, with CES-D scores having the greatest influence (Pratt = 21%), followed by satisfaction with self-esteem (Pratt = 18%), health satisfaction (Pratt = 15%) and friendship satisfaction (Pratt = 12%). Results of the two hierarchical regressions indicated that if the block of health scores was entered first, it accounted for 38% of the variance in life satisfaction scores, leaving 30 percentage points to be added by the block of domain satisfaction scores. However, if the block of domain satisfaction scores was entered first, it accounted for 64% of the variance in life satisfaction scores, leaving only 4 percentage points to be added by the block of health scores. Again, both procedures provided a misleading impression of the relative importance of health and domain satisfaction to life satisfaction. The Pratt measures showed that the eight domain satisfaction scores actually explained about 82% of the explained variance in life satisfaction.

277

HEALTH AND THE QUALITY OF LIFE

Exhibit X Regressions of Life Satisfaction Scores on Health and Domain Satisfaction Scores Explanatory variables

Beta values

Beta values

Beta values

General Health Mental Health CES-D Stress Index Life Stress Item Physical Functioning Family Relations Sat. Living Partner Sat. Financial Security Sat. Self-esteem Sat. Job Sat. Friendship Sat. Present Age Sat. Health Sat. Religion Sat. Variance Explained a = variable not in equation

0.09 0.17 –0.34 –0.21 –0.06 –0.09 a a a a a a a a a 47%

a a a a a a 0.08 0.15 0.08 0.24 0.12 0.17 0.12 0.12 0.09 65%

–0.07 a –0.25 a a –0.07 0.07 0.13 a 0.19 0.10 0.14 0.15 0.19 0.06 68%

EXPLAINING THE IMPACT OF HEALTH AND DOMAIN SATISFACTION ON SATISFACTION WITH THE OVERALL QUALITY OF LIFE

The first column of Exhibit XI shows that our battery of health scores accounted for 42% of the variation in quality of life satisfaction scores, with CES-D scores having the greatest influence(Pratt = 52%), followed at some distance by the the SF-36 General Health items (Pratt = 21%), and Stress Index scores (Pratt = 19%). Only two other scores found their way into the equation, the life stress item (Pratt = 8%) and the Body Mass Index (Pratt is less than 1%). The second column shows that our battery of domain satisfaction scores accounted for 63% of the variation in quality of life satisfaction scores, with present age satisfaction dominating the set

278

ALEX C. MICHALOS ET AL.

Exhibit XI Regressions of Satisfaction with the Overall Quality of Life Scores on Health and Domain Satisfaction Scores Explanatory Variables

Beta values

Beta values

Beta values

General Health CES-D Stress Index Life Stress Item Body Mass Index Family Rel. Sat. Living Partner Sat. Financial Sec. Sat. Self-esteem Sat. Neighbourhood Sat. Friendship Sat. Present Age Sat. Health Sat. Recreation Sat. Local Gov. Sat. Variance Explained a = variable not in equation

0.21 –0.37 –0.18 –0.11 0.07 a a a a a a a a a a 42%

a a a a a 0.08 0.10 0.17 0.14 0.11 0.13 0.19 0.16 0.06 0.12 63%

a –0.16 a –0.08 0.06 0.12 a 0.10 0.14 0.13 a 0.21 0.18 0.08 0.07 63%

of predictors (Pratt = 16%), followed by financial security (Pratt = 15%), satisfaction with self-esteem and health (Pratt = 13% each), living partner (Pratt = 8%), and five other domains with Pratt values of 7% or less each. The third column shows that the two batteries together accounted for 63% of the variation in quality of life satisfaction scores, with present age satisfaction scores having the greatest influence (Pratt = 18%), followed by health satisfaction (Pratt = 15%), and CES-D scores and self-esteem satisfaction with Pratt values of 13% each. Results of the two hierarchical regressions indicated that if the block of health scores was entered first, it accounted for 33% of the variance in quality of life satisfaction scores, leaving 30 percentage points to be added by the block of domain satisfaction scores. How-

HEALTH AND THE QUALITY OF LIFE

279

ever, if the block of domain satisfaction scores was entered first, it accounted for 60% of the variance in quality of life satisfaction scores, leaving only 3 percentage points to be added by the block of health scores. Again, both procedures provided a misleading impression of the relative importance of health and domain satisfaction to life satisfaction. The Pratt measures showed that the eight domain satisfaction scores actually explained about 82% of the explained variance in quality of life satisfaction.

CONCLUSION

The aim of this investigation was to explain the impact of people’s health on their levels of satisfaction with their health, and the impact of these things plus satisfaction with other specific domains of their lives on the perceived quality of their lives operationalized as general happiness, satisfaction with life as a whole and overall satisfaction with the quality of life. Seven hundred and twenty-three usable questionnaires returned from a mailout random sample of 2500 households of Prince George, British Columbia in November 1998 formed the working data-set for our analyses. Among other things, mean respondent scores on the SF-36 health profile were lower than published norms from the UK, USA, Netherlands and Sweden, but higher than scores from Aberdeen, Scotland. Mean scores on the CES-D depression scale also indicated that our respondents tended to have more depressive symptoms than comparison groups in Winnipeg and the USA. A review of trends in mean scores on 17 quality of life items (e.g., satisfaction with family life, financial security, recreation, etc.) from 1994, 1997 and 1998 revealed that there were only 7 statistically significant changes across the four year period and they were all negative. Multivariate regression analysis showed that health status measured with a variety of indicators could explain 56% of the variation in respondents’ reported satisfaction with their health. A combination of health status plus domain satisfaction indicators could explain 53% of the variation in respondents’ reported happiness, 68% of reported life satisfaction and 63% of reported satisfaction with the overall quality of life. Sixty percent of the explained variation in happiness scores was attributable to self-reported health scores, while only 18% of

280

ALEX C. MICHALOS ET AL.

the explained variation in satisfaction with life and with the overall quality of life scores was attributable self-reported health scores.

REFERENCES Aaronson, N. K., F. S. A. M. van Dam, C. E. Polak and R. Zittoun: 1986, ‘Prospects and problems in European psychosocial oncology: a survey of the EORTC study group on quality of life’, Journal of Psychosocial Oncology 4, pp. 43–53. Affleck, G.: 1987, ‘Downward comparison and coping with serious medical problems’, American Journal of Orthopsychiatry 57, pp. 570–578. Andrews, F. M. and S. B. Withey: 1976, Social Indicators of Well-Being (Plenum Press, New York). Andres, D., D. P. Gold, T. Arbuckle, A. Schwartzman and J. Chaikelson: 1990, Psychosocial determinants of self-reported health in elderly men, Paper presented at the Canadian Association on Gerontology conference, Victoria, BC. Angel, R. and P. Thoits: 1987, ‘The impact of culture on the cognitive structure of illness’, Culture, Medicine and Psychiatry 11, pp. 465–494. Balaban, D. J., P. C. Sagi, N. I. Goldfarb and S. Nettler: 1986, ‘Weights for scoring the quality of well-being (QWB) instrument among rheumatoid arthritis: a comparison to general population weights’, Medical Care 24, pp. 973–980. Barnes, G. E., R. F. Currie and A. Segall: 1988, ‘Symptoms of depression in a Canadian urban sample’, Canadian Journal of Psychiatry 33, pp. 386–392. Bergner, M.: 1989, ‘Quality of life, health status, and clinical research’, Medical Care 27, pp. S148–S156. Bergner, M., R. A. Bobbitt, W. B. Carter and B. S. Gilson: 1981, ‘The Sickness Impact Profile: development and final revision of a health status measure’, Medical Care 19, pp. 787-805. Binik, Y. M. and G. M. Devins: 1986–1987, ‘Transplant failure does not compromise quality of life in end-stage renal disease’, International Journal of Psychiatry in Medicine 16, pp. 281–292. Blackwell, D. and L. Levy: 1987, ‘Hearing impairment, self-concept, and morale among elderly’, Free Inquiry in Creative Sociology 15, pp. 21–26. Blalock, S. J., B. M. DeVellis and R. F. DeVellis: 1989, ‘Social comparison among individuals with rheumatoid arthritis’, Journal of Applied Social Psychology 19, pp. 665–680. Blalock, S. J., et al.: 1990, ‘Adjustment to rheumatoid arthritis: the role of social comparison processes’, Health Education Research 5, pp. 361–370. Blodgett, C.: 1981–1982, ‘A selected review of the literature of adjustment to hemodialysis’, International Journal of Psychiatry Medicine 11, pp. 97–123. Breslow, L.: 1989, ‘Health status measurement in the evaluation of health promotion’, Medical Care 27, pp. S205–S216.

HEALTH AND THE QUALITY OF LIFE

281

Brown, R. I.: 1989, ‘Aging, disability and quality of life: a challenge for society’, Canadian Psychology/ Psychologie Canadienne 30, pp. 551–559. Bryant, F. B. and J. T. Marquez: 1986, ‘Educational status and the structure of subjective well-being in men and women’, Social Psychology Quarterly 49, pp. 142–153. Burckhardt, C. S.: 1985, ‘The impact of arthritis on quality of life’, Nursing Research 34, pp. 11–16. Busse, E. W., F. C. Jeffers and W. D. Obrist: 1970, ‘Factors in age awareness’, in E. Palmore (ed.), Normal aging: Reports from the Duke Longitudinal Study, 1955–1969. (Duke University Press: Durham, NC). Buunk, B. P., et al.: 1990, ‘The affective consequences of social comparison: either direction has its ups and downs’, Journal of Personality and Social Psychology 59, pp. 1238–1249. Callan, V. J. and J. F. Hennessey: 1988, ‘The psychological adjustment of women experiencing infertility’, British Journal of Medical Psychology 61, pp. 137– 140. Callan, V. J. and J. F. Hennessey: 1989, ‘Psychological adjustment to infertility: a unique comparison of two groups of infertile women, mothers and women childless by choice’, Journal of Reproductive and Infant Psychology 7, pp. 105– 112. Campbell, A. P., P. E. Converse and W. L. Rodgers: 1976, The Quality of American Life (Russell Sage Foundation, New York). Carp, F. M. and A. Carp: 1981, ‘Mental health characteristics and acceptancerejection of old age’, American Journal of Orthopsychiatry 51, pp. 230–241. Chuang, H. T., G. M. Devins, J. Hunsley and M. J. Gill: 1989, ‘Psychosocial distress and well-being among gay and bisexual men with human immunodeficiency virus infection’, American Journal of Psychiatry 146, pp. 876–880. Chwalisz, K., E. Diener and D. Gallagher: 1988,‘Autonomic arousal feedback and emotional experience: evidence from the spinal cord injured’, Journal of Personality and Social Psychology 54, pp. 820–828. Cooper, A. F.: 1976, ‘Deafness and psychiatric illness’, British Journal of Psychiatry 129, pp. 216–226. Devellis, R. F., B. M Devellis, B. S. Wallston and K. A. Wallston: 1980, ‘Epilepsy and learned helplessness’, Basic and Applied Social Psychology 1, pp. 241– 253. Devellis, R. F., et al.: 1990, ‘The relationship of social comparison to rheumatoid arthritis symptoms and affect’, Basic and Applied Social Psychology 11, pp. 1– 18. Devins, G. M.:1989, ‘Enhancing personal control and minimizing illness intrusiveness’, in N. G. Kutner, D. D. Cardenas and J. D. Bower (eds.), Maximizing Rehabilitation in Chronic Renal Disease (PMA Pub. Group, New York), pp. 109–136. Devins, G. M. and T. P. Seland: 1987, ‘Emotional impact of multiple sclerosis: recent findings and suggestions for future research’, Psychological Bulletin 101, pp. 363–375.

282

ALEX C. MICHALOS ET AL.

Dupuy, H. J.: 1984, ‘The psychological general well-being (PGWB) index’, in N. K. Wenger, M. E. Mattson, C. D. Furberg and J. Elinson (eds.), Assessment of Quality of Life in Clinical Trials of Cardiovascular Therapies (Le Jacq Pub., New York), pp. 170–183. Fayers, P. M. and D. R. Jones: 1983, ‘Measuring and analysing quality of life in cancer clinical trials: a review’, Statistics in Medicine 2, pp. 429–446. Garratt, A. M., D. A. Ruta, M. I. Abdalla, J. K. Buckingham and I. T. Russell: 1993, ‘the SF-36 health survey questionnaire: an outcome measure suitable for routine use within the NHS’, British Medical Journal 306, pp. 1440–1444. Gfellner, B. M.: 1989, ‘Perceptions of health, abilities, and life satisfaction among very old adults’, Perceptual and Motor Skills 68, pp. 203–209. Gillman, A. E., A. Simmel and E. P. Simon: 1986, ‘Visual handicap in the aged: self-reported visual disability and the quality of life of residents of public housing for the elderly’, Journal of Visual Impairment and Blindness 79, pp.588–590. Goodinson, S. M. and J. Singleton: 1989, ‘Quality of life: a critical review of current concepts, measures and their clinical implications’, International Journal of Nursing Studies 26, pp. 327–341. Gray, H., D. Brogan and N.G. Kutner: 1985, ‘Status of life-areas: congruence/ noncongruence in ESRD patient and spouse perceptions’, Social Science and Medicine 20, pp. 341–346. Haes, J. C. J. M. de and F. C. E. Van Knippenberg: 1985, ‘The quality of life of cancer patients: a review of the literature’, Social Science and Medicine 20, pp. 809–817. Han, T. S., M. A. R. Tijhuis, M. E. J. Lean and J. C. Seidell: 1998, ‘Quality of life in relation to overweight and body fat distribution’, American Journal of Public Health 88(12), pp. 1814–1821. Hays, R. D. and C. E. Ross: 1986, ‘Body and mind: the effect of exercise, overweight, and physical health on psychological well-being’, Journal of Health and Social Behavior 27, pp. 387–400. Healthy Communities Committee: 1997, Report on the Quality of Life in Prince George (City of Prince George, BC). Hollandsworth, J. G.: 1988, ‘Evaluating the impact of medical treatment on the quality of life: a 5-year update, Social Science and Medicine 26, pp. 425–434. Hubley, A. M.: 1991, ‘The relationship of personality trait variables to two measures of subjective age identity: A symbolic interactionist perspective’, Unpublished Master’s thesis (University of Victoria, Victoria, BC). Hurny, C., E. Piasetsky, R. Bagin and J. Holland: 1987, ‘High social desirability in patientsbeing treated for advanced colorectal or bladder cancer: eventual impact on the assessment of quality of life’, Journal of Psychosocial Oncology 5, pp. 19–29. Jenkinson, C., A. Coulter and L. Wright: 1993 Short form 36 (SF-36) health survey questionnaire: normative data for adults of working age’, British Medical Journal 306, pp. 1437–1440.

HEALTH AND THE QUALITY OF LIFE

283

Johnson, J. P., C. R. McCauley and J. B. Copley: 1982, ‘The quality of life of hemodialysis and transplant patients’, Kidney International 22, pp. 286–291. Kaplan, R. M., J. W. Bush and C. C. Berry: 1976, ‘Health status: types of validity for an index of well-being’, Health Services Research 11, pp. 478–507. Katz, J. N., M. G. Larson, C. B. Phillips, A. H. Fossel and M. H. Liang: 1992, ‘Comparative measurement sensitivity of short and longer health status instruments’, Medical Care 30, pp. 917–925. Kibele, A.: 1989, ‘Occupational therapy’s role in improving the quality of life for persons with cerebral palsy’, American Journal of Occupational Therapy 43, pp. 371–377. Klonoff, P. S., L. D. Costa and W. G. Snow: 1986, ‘Predictors and indicators of quality of life in patients with closed-head injury’, Journal of Clinical and Experimental Neuropsychology 8, pp. 469–485. Lehman, A. F.: 1983, ‘The well-being of chronic mental patients’, Archives of General Psychiatry 40, pp. 369–373. Lehman, A. F., N. C. Ward and L. S. Linn: 1982, ‘Chronic mental patients: the quality of life issue’, American Journal of Psychiatry 139, pp. 1271–1276. Lewinsohn, P. M., J. R. Seeley, R. E. Roberts and N. B. Allen: 1997, ‘Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-residing older adults’, Psychology and Aging 12, pp. 277–287. Lezak, M. D.: 1987, ‘Relationships between personality disorders, social disturbances, and physical disability following traumatic brain injury’, Journal of Head Trauma Rehabilitation 2, pp. 57–69. Logan, J. R., R. Ward and G. Spitze: 1992, ‘As old as you feel: Age identity in middle and later life’, Social Forces 71, pp. 451–467. Lomas, J., L. Pickard and A. Mohide: 1987, ‘Patient versus clinician item generation for quality-of-life measures’, Medical Care 25, pp. 764–769. Magilvy, J. K.: 1985, ‘Quality of life of hearing-impaired older women’, Nursing Research 34, pp. 140–144. Mai, F. M., F. N. McKenzie and W. J. Kostuk: 1990, ‘Psychosocial adjustment and quality of life following heart transplantation’, Canadian Journal of Psychiatry 35, pp. 223–227. Markides, K. S. and C. Pappas: 1982, ‘Subjective age, health, and survivorship in old age’, Research on Aging 4, pp. 87–96. Mayer, T. and H. B. Andrews: 1981, ‘Changes in self-concept following a spinal cord injury’, Journal of Applied Rehabilitation Counseling 12, pp. 135–137. McHorney, C. A., J. E. Ware and A. E. Raczek: 1993, ‘The MOS 36-Item ShortForm Health Survey (SF-36): II. Psychometric and Clinical Tests of Validity in Measuring Physical and Mental Health Constructs’, Medical Care 31(3), pp. 247–263. McHorney, C. A., J. E. Ware, J. F. R. Lu and C. Donald Sherbourne: 1994, ‘The MOS 36-Item Short-Form Health Survey (SF-36): III Tests of Data Quality, Scaling Assumptions, and Reliability Across Diverse Patient Groups’, Medical Care 32(1), pp. 40–66.

284

ALEX C. MICHALOS ET AL.

Mechanic, D. and R. J. Angel: 1987, ‘Some factors associated with the report and evaluation of back pain’, Journal of Health and Social Behavior 28, pp. 131–139. Mechanic, D. and S. Hansell: 1987, ‘Adolescent competence, psychological wellbeing, and self-assessed physical health’, Journal of Health and Social Behavior 28, pp. 364–374. Michalos, A. C.: 1985, ‘Multiple discrepancies theory (MDT), Social Indicators Research 16, pp. 347–413. Michalos, A. C.: 1991, Global Report on Student Well-Being, Vol. 1, Life Satisfaction and Happiness (Springer-Verlag, New York). Michalos, A. C.: 1991a, Global Report on Student Well-Being, Vol. 2, Family, Friends, Living Partner and Self-Esteem (Springer-Verlag, New York). Michalos, A. C.: 1993, Global Report on Student Well-Being, Vol. 3, Employment, Finances, Housing and Transportation (Springer-Verlag, New York). Michalos, A. C.: 1993a, Global Report on Student Well-Being, Vol. 4, Religion, Education, Recreation and Health (Springer-Verlag, New York). Michalos, A. C. and B. D. Zumbo: 1999, ‘Public services and the quality of life’, Social Indicators Research, 48, pp. 125–156. Milligan, W. L., D. A. Powell, C. Harley and E. Furchtgott: 1985, ‘Physical health correlates of attitudes toward aging in the elderly’, Experimental Aging Research 11, pp. 75–80. Montepare, J. M. and M. E. Lachman: 1989, ‘’You’re only as old as you feel’: self-perceptions of age, fears of aging, and life satisfaction from adolescence to old age’, Psychology and Aging 4, pp. 73–77. Mulkay, M., M. Ashmore and T. Pinch: 1987, ‘Measuring the quality of life: a sociological invention concerning the application of economics to health care’, Sociology 21, pp. 541–564. Mutran, E. and P. J. Burke: 1979, ‘Personalism as a component of old age identity’, Research on Aging 1, pp. 37–64. Nadel, C. and J. J. Clark: 1986, ‘Psychosocial adjustment after renal retransplants, General Hospital Psychiatry 8, pp. 41–48. Naughton, M. J., S. A. Shumaker, R. T. Anderson and S. M. Czajkowski: 1996, ‘Psychological aspects of health-related quality of life measurement: tests and scales’, in B. Spilker (ed.), Quality of Life and Pharmacoeconomics in Clinical Trials (Lippincott-Raven, Philadelphia) pp. 117–131. Okun, M. A., W. A. Stock, M. J. Haring and R. A. Witter: 1984, ‘Health and subjective well-being: a meta-analysis’, International Journal of Aging and Human Development 19, pp. 111–132. Pearlman, R. A. and A. Jonsen: 1985, ‘The use of quality-of-life considerations in medical decision making, Journal of the American Geriatrics Society 33, pp. 344–352. Query, J. M., C. Reichelt and L. A. Christoferson: 1990,’Living with chronic illness: a retrospective study of patients shunted for hydrocephalus and their families’, Development Medicine and Child Neurology 32, pp. 119–128.

HEALTH AND THE QUALITY OF LIFE

285

Radloff, L. S.: 1977, ‘The CES-D scale: A self-report depression scale for research in the general population’, Applied Psychological Measurement 1, pp. 385–401. Ramund, B. and R. Stensman: 1988, ‘Quality of life and evaluation of functions among people with severely impaired mobility and non-disabled controls’, Scandinavian Journal of Psychology 29, pp. 137–144. Sabourin, S., L. Bourgeois, P. Gendreau and M. Morval: 1989, ‘Self-deception, impression management, and consumer satisfaction with mental health treatment, Psychological Assessment: A Journal of Consulting and Clinical Psychology 1, pp. 126–129. Sabourin, S., et al.: 1989, ‘Social desirability, psychological distress, and consumer satisfaction with mental health treatment’, Journal of Counseling Psychology 36, pp. 352–356. Scheier, M. F., et al.: 1989, ‘Dispositional optimism and recovery from coronary artery bypass surgery: the beneficial effects on physical and psychological wellbeing’, Journal of Personality and Social Psychology 57, pp. 1024–1040. Schulz, R. and S. Decker: 1985, ‘Long-term adjustment to physical disability: the role of social support, perceived control and self-blame’, Journal of Personality and Social Psychology 48, pp. 1162–1172. Schulz, R. and S. Fritz: 1987, ‘Origins of stereotypes of the elderly: an experimental study of the self-other discrepancy’, Experimental Aging Research 13, pp. 189–195. Shadish, W. R., R. G. Orwin, B. G. Silber and R. R. Bootzin: 1985, ‘The subjective well-being of mental patients in nursing homes’, Evaluation and Program Planning 8, pp. 239–250. Shmueli, A.: 1998, ‘The SF-36 profile and health-related quality of life: an interpretive analysis’, Quality of Life Research 7(3), pp. 187–195. Spilker, B.: 1996, Quality of Life and Pharmacoeconomics in Clinical Trials (Lippincott-Raven, Philadelphia). Spitzer, W. O. et al.: 1981, ‘Measuring the quality of life of cancer patients’, Journal of Chronic Disease 34, pp. 585–597. Staats, S., K. Heaphey, D. Miller, C. Partlo, N. Romine and K. Stubbs: 1993, ‘Subjective age and health perceptions of older persons: Maintaining the youthful bias in sickness and in health’, International Journal of Aging and Human Development 37, pp. 191–203. Steitz, J. A. and A. M. McClary: 1988, ‘Subjective age, age identity and middleage adults’, Experimental Aging Research 14, pp. 83–87. Stensman, R.: 1985, ‘Severely mobility-disabled people assess the quality of their lives’, Scandinavian Journal of Rehabilitation Medicine 17, pp. 87–99. Suchman, E. A. and B. S. Phillips: 1958, ‘An analysis of the validity of health questionnaires, Social Forces 36, pp. 223–232. Sullivan, M., J. Karlsson and J. E. Ware: 1995, ‘The Swedish SF-36 health survey – 1. Evaluation of data quality, scaling assumptions, reliability and construct validity across general populations in Sweden’, Social Science and Medicine 41(10), pp. 1349–1358.

286

ALEX C. MICHALOS ET AL.

Taylor, S. E.: 1983, ‘Adjustment to threatening events: a theory of cognitive adaptation’, American Psychologist 38, pp. 1161–1173. Thomas, D. R., E. Hughes and B. D. Zumbo: 1998, ‘On variable importance in linear regression, Social Indicators Research 45(1–3), pp. 253–275. Thomas, W. C.: 1981, ‘The expectation gap and the stereotype of the stereotype: images of old people’, Gerontologist 21, pp. 402–407. Wacker, D. P., D. C. Harper, W. J. Powell and A. Healy: 1983, ‘Life outcomes and satisfaction ratings of multihandicapped adults’, Developmental Medicine and Child Neurology 25, pp. 625–631. Ware, J. E., M. Kosinsky and S. D. Keller: 1994, SF-36 Physical and Mental Health Summary Scales: A User’s Manual (The Health Institute, New England Medical Center, Boston, Massachusetts). Ware, J. E. and C. Donald Sherbourne: 1992, ‘The MOS 36-Item Short-Form Health Survey (SF-36): I. Conceptual Framework and Item Selection, Medical Care 30(6), pp. 473–483. Ware, J. E., K. K. Snow, M. Kosinski and B. Gandek: 1993, SF-36 Health Survey: Manual and Interpretation Guide (The Health Institute, New England Medical Center, Boston, Massachusetts). Williams, A.: 1985, ‘Economics of coronary artery bypass grafting, British Medical Journal 291, pp. 326–329. Wood, J. V., S. E. Taylor and R. R. Lichtman: 1985, ‘Social comparison in adjustment to breast cancer’, Journal of Personality and Social Psychology 49, pp. 1169–1183.

Institute for Social Research and Evaluation University of Northern British Columbia Prince George, B.C. E-mail: [email protected]