This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Maintaining Self-Rated Health Through Social Comparison in Old Age

¹ Department of Applied Social Studies, City University of Hong Kong.
² Department of Psychology, Chinese University of Hong Kong.
³ Asia-Pacific Institute of Ageing Studies, Lingnan University, Hong Kong.

	Abstract

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
When self-related health (SRH) is under threat, people may use self-enhancement through social comparison to buffer against the threat. Because SRH is under greater threat among older than among younger adults, the benefit of social comparison on SRH may be greater for them. To test this, Study 1 examined ratings on physical attributes for self and "someone of the same age" among 592 Hong Kong Chinese, across young, middle, and old age. Findings suggested that perceiving one's physical self as better than others produced a larger gain in self-rated health in older than in younger people. Study 2 showed longitudinally that, among older adults, an increase in physical symptoms over time was associated with a worsening SRH, which was then positively associated with physical self-enhancement. This improved physical self in turn partially recovered the damage to SRH.

Self-rated health (SRH) predicts subsequent mortality and functional decline independent of objective health conditions (e.g., Idler & Benyamini, 1997; "Idler", Russell, & Davis, 2000; Kaplan, Strawbridge, Camacho, & Cohen, 1993; Spiers, Jagger, Clarke, & Arthur, 2003). Vulnerability to health problems, general physical functioning, activity restrictions, sleep quality, healthy behaviors, and social well-being are found to predict SRH cross-culturally (Cheng & Chan, 2006; Pinquart, 2001). With age, the correspondence between objective and subjective health weakens (Borchelt, Gilberg, Horgas, & Geiselmann, 1999; Idler, 1993; Pinquart). Longitudinal studies show that SRH remains relatively stable despite physical decline in old age (Leinonen, Heikkinen, & Jylhä, 1998; Maddox & Douglass, 1973), and when a significant decline was observed, the magnitude was generally small (Miller & Wolinsky, 2007; Svedberg, Gatz, Lichtenstein, Sandin, & Pedersen, 2005). For instance, over a 9-year interval, those individuals aged 70 years or older reported a decline of SRH (in standardized scores) of only –0.55 to –0.68 (Svedberg et al.).

Given that older adults seem to be able to maintain their SRH in the face of objective health declines, it is theoretically important to investigate the particular mechanisms that enable them to do so. A potential mechanism that contributes to the relative stability of SRH and one that has received little attention in the literature is social comparison. In the literature, researchers typically assess SRH by asking respondents to rate their overall health along a scale of excellent to poor. Sometimes, respondents are also asked to indicate if their health is better or worse than same-aged peers. Studies examining both questions have reported moderate correlations between the two (Baron-Epel & Kaplan, 2001; Deeg & Kriegsman, 2003), suggesting that SRH is partly based on social comparison. In this article we extend this literature by studying whether perceiving one's physical self as better than the physical self of age peers has a stronger beneficial effect on SRH in older than in younger adults (Study 1), and whether such a perception is a protective factor against the declining trajectory of SRH in later life (Study 2).

Festinger (1954) argues that social comparison is a human drive and is important for self-evaluation. When the self is compared with others, an individual may find the self as doing better (self-enhancement), the same, or worse (self-effacement) than others. More recent studies suggest that self-enhancement serves protective functions. According to Wills (1981), when one experiences threat to subjective well-being, then self-enhancement, done either passively (i.e., depending on available comparison targets) or actively (e.g., derogating another person), can be used to restore well-being and a positive self-image (e.g., subjective health). Numerous other studies have demonstrated the beneficial effects of self-enhancement in people whose physical and psychological well-being are under threat, such as cancer patients (see reviews by Taylor & Lobel, 1989; Wills, 1981). Events other than health threats can also benefit from social comparison. For example, in a longitudinal study of women who had to move their homes and adapt to a new living environment, seeing oneself as increasingly healthier than others was associated with enhanced psychological outcomes over time (Kling, Ryff, & Essex, 1997).

Social Comparison in the Context of Aging
Aging presents many challenges to physical well-being. Normative aging is associated with moderate increases in physical symptoms over time (see the Normative Aging Study; Aldwin, Spiro, Levenson, & Rosse, 1989). Worse, the culmination of chronic and life-threatening illnesses may leave one in pain or incapacitated on a day-to-day basis. It is therefore no wonder that although most people maintain a positive SRH regardless of their age, satisfaction with health inevitably declines with age (Cheng, 2004; Herzog & Rodgers, 1981). In the Berlin Aging Study, a poor functional health was found to be strongly associated with negative affect cross-sectionally, and functional decline further predicts declining positive affect over time (Kunzmann, Little, & Smith, 2000).

Despite all the challenges, older persons, except perhaps those in very old age and suffering from poor health, appear to maintain positive well-being at a level that is at least comparable with that of younger people, a phenomenon known as the paradox of aging (also known as the well-being paradox; see review by Cheng, 2004). It has been proposed that social comparison is a strategy that protects older people from the negative effects of aging.

In Heckhausen and Schulz's life-span theory of control (Heckhausen & Schulz, 1995; Schulz & Heckhausen, 1996), social comparison is conceptualized as a compensatory secondary control strategy so that a positive self-view can be restored when goal attainment is unsuccessful. The theory argues that because aging is associated with declining resources to manage the environment, motives such as self-improvement or self-assessment become gradually less salient with age. In contrast, motives for self-enhancement by way of social comparison are strengthened because of the increasing need to stabilize the self-concept despite realistic demands to have it modified.

A similar prediction is made by Brandtstädter and colleagues (Brandtstädter & Greve, 1994; Brandtstädter & Rothermund, 2002). They propose a dual-process framework to describe how aging individuals reduce discrepancies between their current and ideal situations. One process is known as the assimilative mode, which involves "intentional efforts to modify the actual situation in accordance with personal goals" (Brandtstädter & Rothermund, p. 117). The other process is known as the accommodative mode, which adjusts personal goals to make them consistent with situational constraints and limited resources. Age-related declines and the resulting losses in resources lead people to favor the accommodative process over the assimilative one as they age (Brandtstädter & Renner, 1990). Self-enhancement through social comparison is an effective accommodative strategy.

Indeed, an upward age trend in seeing oneself as possessing a more favorable developmental trajectory than most other people of the same age has been observed. Heckhausen and Krueger (1993) asked West Berliners aged 21–80 years to rate one's own development in adulthood, relative to that of other people. They found that, compared with younger people, older people rated themselves as experiencing fewer losses in desirable attributes and fewer gains in undesirable attributes over time. This finding demonstrated how perceived developmental norms could serve as a frame of reference for positive self-assessment.

Another study (Heckhausen & Brim, 1997) examined the difference between perceptions about the self and about "most people my age" with regard to 12 domains of life (e.g., health, marriage, and job) among a large, representative sample of U.S. adults aged 18 years or older. Findings of the study revealed that the presence of personal (including physical) problems was associated with a person's seeing more problems in most other people, thereby allowing the person to maintain a favorable self-view. Moreover, this tendency to downgrade others because of personal problems was most pronounced in older persons, but in domains that were relatively problem free (marriage, stress, job, and leisure) only.

Although these studies shed light on how perceptions of same-aged peers can be adapted to suit one's need for positive self-concept in the context of developmental decline, they did not examine whether self-enhancing comparison was actually used more by older persons. A diary study by Heckhausen (1999) was, to the best of our knowledge, the only one that examined this issue. In this study, Berlin adults aged 20–81 years responded to a questionnaire on a daily basis for 14 days by referring to notes on social comparison they made during the daytime. On the basis of whether the comparison target was rated as better (upward social comparison), same (lateral social comparison), or worse (downward social comparison) than self, the frequencies with which these three kinds of comparison activities were engaged in were tabulated. The results showed an upward age trend in the use of lateral comparisons and a downward age trend in upward comparisons, but a lack of age difference in the use of downward comparisons. Thus although older persons can take advantage of age stereotypes to maintain a positive self-view, they are not more likely than middle- and young-aged persons to actively engage in self-enhancement through social comparison in everyday life.

Although older people do not necessarily use social comparison more often than younger people do in a self-enhancing fashion (Heckhausen, 1999), a possibility that remains to be investigated is whether the effect of such social comparison on SRH is stronger with increasing age. A cross-sectional survey of a nationally representative sample of Israeli persons aged 45 years or older (Baron-Epel & Kaplan, 2001) suggested that this might indeed be the case. This study showed that those individuals aged 65 years or older were more likely to give a more favorable rating of their health when asked to compare themselves with people of the same age and sex than when the rating was made without a comparison instruction. According to Heckhausen, self-enhancement is a more important motive in later life because of the need to stabilize the self amidst increasing difficulties to control events in life (e.g., preventing health problems). As a result, the effect of self-enhancement on SRH should be greater for older than for younger people. Because of the nature of the study, we focus on social comparison in the physical health domain only, not social comparison in general. In line of this thinking, we hypothesize that SRH is a stronger function of social comparison in the physical domain for older than for younger adults, and that such comparisons serve as a buffer for older people against the threat to SRH that is due to increasing physical problems.

Assessing Social Comparison
The target against which comparison is made may have profound effects on the findings obtained. In survey research, the typical instruction is to compare oneself against "most other people of the same age" (e.g., Heckhausen & Krueger, 1993; Kling et al., 1997). This being the case, the respondent is expected to call upon stereotypic representations of others in his or her assessment of the self. This creates a problem when such stereotypic views exist for some attributes but not others, especially attributes that are not readily observable (e.g., sleep quality). More importantly, when people take diary notes of their social comparison activities, they seldom, if ever, compare themselves against an undefined group such as most other people; rather, the typical comparison is engaged with someone known or met (Heckhausen, 1999; Wheeler & Miyake, 1992). Hence "most other people" is simply an irrelevant reference.

Another problem with the typical instruction is that an explicit comparison with others (e.g., by requiring respondents to rate whether they are better or worse than others) tends to provoke a contrast effect (i.e., seeing oneself as different) and hence bias the study toward finding self-enhancing effects (McFarland, Buehler, & MacKay, 2001). Research shows that this contrast effect is diminished when respondents are not prompted to make explicit comparisons with the target (McFarland et al.). Moreover, making the comparisons mentally requires the participants to have a certain level of cognitive resources, which may not be available to older adults with limited education. Hence we adopted an approach that tapped ratings of self and others independently, and then we took their differences as our measure of social comparison. This approach has been widely adopted in the study of discrepancy ratings between self and others (e.g., Bradshaw & Hazan, 2006; Klein, Monin, Steers-Wentzell, & Buckingham, 2006; Mussweiler & Ruter, 2003), and it has been shown to be responsive to experimental conditions designed to manipulate the self-enhancement bias (Klein, 2001).

In addition, research on social comparison in everyday life has shown rather different findings from those obtained in experimental situations (Wheeler & Miyake, 1992). In general, the literature suggests that social comparison is typically made against routine targets, thus conserving cognitive resources in the exercise (Mussweiler & Ruter, 2003). Although such routine targets are usually significant others (Mussweiler & Ruter), they are not the preferred targets for achieving positive self-evaluations (Wheeler & Miyake) because it can be painful to see one's significant other not doing well (Beach & Tesser, 1995; McFarland et al., 2001), especially in the area of health. For self-enhancing comparison, the preferred targets are those who are sufficiently distant from the self (e.g., ordinary friends, neighbors, colleagues) but whose information in specific domains is readily available (Buunk, Collins, Taylor, Van Yperen, & Dakof, 1990; Wheeler & Miyake). This implies that one often engages different targets in different comparison domains if self-enhancement is the goal. In a survey research such as the present one, it is important that we do not fixate the comparison target across all attributes. As a result, we adopted an approach that allowed participants to shift targets across comparison domains. Our focus was not on how and with whom participants conducted social comparison, but on what happened to the participants' SRH as a result of the comparison. As shown in previous research (Mussweiler & Ruter), our approach (see the subsequent discussion) should elicit images of the most salient and accessible social other with whom the participants compare in specific domains in daily life.

To investigate age differences in social comparison and its effect on SRH, Study 1 examined whether SRH was related to perceiving one's physical self as better than that of others, and whether such relationship differed by age. Study 2 focused on older adults. Over a 12-month period, it investigated (a) if increases in physical symptoms were associated with greater self-enhancement through social comparison in the physical domain, which (b) in turn contributed to better SRH. Collectively, the two studies aimed at shedding light on social comparison as a mechanism that may help older adults to maintain SRH in the face of physical declines.

	Study 1

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
METHODS
Participants and Procedure
Because of profound differences in educational attainment and literacy between younger and older cohorts of the population (Census and Statistics Department, 2002), we adopted a mixed strategy for data collection for different age groups. Following Cheng (2004), we recruited recent graduates of a psychology program to distribute a questionnaire to unrelated persons they have known from formal or informal contacts. We endeavoured to achieve a roughly sex-balanced sample within each of the following age groups: 20–24, 25–29, 30–34, 35–39, 40–44, 45–49, 50–54, and 55–59 years old. A total of 395 questionnaires were distributed and 387 (98.0%) completed questionnaires were returned.

Additionally, we recruited 208 persons aged 60 years or older on a convenience basis from social centers for a longitudinal study on aging and well-being. We used the Wave 1 data in this study for comparison with younger age groups. We excluded 3 individuals with scores of <20 on the Mini-Mental State Examination (Folstein, Folstein, & McHugh, 1975; also see Chiu, Lee, Chung, & Kwong, 1994), leaving 205 for the analysis. The entire sample of 592 individuals had a mean age of 47.1 (SD = 21.30).

Participation was voluntary. Younger participants filled out the questionnaire individually and returned the questionnaire in a sealed envelope. Older participants were interviewed, also individually, for approximately 1 hour by use of the unfolding approach (i.e., they were first asked the direction of their response before they chose the actual degree of response on a frequency or Likert scale). Cue cards with enlarged fonts displaying the response options of each question were provided. Ethics approval was obtained from the Ethics Subcommittee of the Research Committee of the City University of Hong Kong.

Sociodemographic characteristics of the sample are shown in Table 1. For the sake of presentation, we present the data separately for three age categories (20–34, 35–59, 60+). In the actual analysis shown in the paragraphs that follow, we treated age as a continuous variable. The sample as a whole was similar to the population in terms of these characteristics, except that the younger adults were more highly educated (Census and Statistics Department, 2002). Nevertheless, sociodemographic differences across the age cohorts resembled those observed in the general population (see Table 1). On the whole, older persons were more likely to be women, widowed and living alone, and to have less education and income. Hence, in the main regression analysis, we included age, sex, marital status, education, and living situation as control variables (we did not include income because it was highly correlated with educational level, =.67, thus creating the problem of statistical redundancy).

Measures
Social comparison
We measured social comparison in terms of the difference between perception of self and perception of others. To generate a list of attributes relevant for this population, we asked 405 young and midlife people (M_age = 38.23, SD = 7.06, range = 21–53), as well as 125 older people (M_age = 73.73, SD = 6.72, range = 60–89), to provide up to three possible selves, three hoped-for selves, and three feared selves for the coming year (Cross & Markus, 1991). We analyzed these statements for content and found them to cover four broad domains: the social self, the physical self, the material self, and the work self (including voluntary work). Statements with similar contents were grouped together; this resulted in 16 items for physical self, 18 for social self, 15 for material self, and 10 for work self. We included both desirable and undesirable (reverse-scored) attributes for each domain. However, many older participants had difficulty responding to the work items, and we dropped this domain in the first place. Furthermore, because a preliminary analysis showed that the effects of comparison in the social and material domains were redundant with that of comparison in physical aspects, and because this article is primarily concerned with health, we limited this report to the physical domain only. (Results concerning social and material selves can be obtained from S.-T. Cheng. Physical self items are shown in the Appendix.)

Physical status
We included two measures for physical status. We assessed self-rated health by asking respondents to rate their overall health on a 5-point scale from 1 (very poor) to 5 (excellent). We assessed physical symptoms by using a 12-item physical symptoms checklist. The original scale has 20 items (Cheng & Hamid, 1996); we selected the 12 items with the highest factor loadings for this study. Items were rated on a frequency scale of 0 (never) to 4 (often) against the past week, with an alpha coefficient of = 0.87.

	RESULTS AND DISCUSSION

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
Because of the disparity in educational levels across the age categories, and because the younger adults were more educated than their age peers in the general population, we had initially weighted the sample to represent the population distribution in education. However, because weighting produced essentially the same results, and because education had no effect on SRH (see the subsequent discussion), we report the findings without the weights applied.

On the whole, participants saw themselves in relatively good health (M = 3.46, SD = 0.84) and reported few symptoms (M = 12.49, SD = 9.83). One-sample t tests showed that the mean social comparison score at 0.13 (SD = 0.64) was significantly different from zero, t(591) = 5.06, p <.001. This score was not correlated with age (r =.02, ns). Hence, in comparison with others, these participants tended to see their physical selves as slightly better, and this was the same regardless of age.

Predicting Self-Rated Health
We conducted hierarchical regressions to see if social comparison explained SRH beyond the effects of sociodemographics and symptoms, and if this effect increased with age. We regressed SRH on the following variables in order of entry: (a) sociodemographic variables, including, age, gender, marital status, education, and living alone (except age, all were coded as dummy variables); (b) symptoms; (c) social comparison; and (d) Age x Social Comparison and Age x Symptoms interactions. We predicted that Age x Social Comparison would bear a significant positive sign, meaning that the effect of social comparison would be greater for older than for younger people. Because studies have shown that subjective health is increasingly unrelated to objective conditions as people age, we included the interaction term Age x Symptoms and we expected a positive sign (as symptoms and SRH were negatively correlated) accordingly. We formed interaction terms from standardized scores of the constituent variables to minimize collinearities between them (Cohen, Cohen, West, & Aiken, 2003). The results are shown in Table 2.

View larger version (11K): [in this window] [in a new window]   Figure 1. Study 1: Self-rated health as a function of (a) social comparison of physical self (controlling for symptoms) and (b) physical symptoms (controlling for social comparison of physical self), with age at –1 SD, mean, and +1 SD; gender, marital status, educational level, and living status are controlled for

	Study 2

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
In addition, to directly test whether the same pattern of relationship between social comparison and SRH observed in Study 1 could be obtained without the potential sampling confound, we examined the relationship longitudinally among older adults in Study 2. The main purpose of this study was to examine if the use of social comparison changes over time in a way that serves to maintain SRH, after the effect of symptom changes is controlled for. Specifically, we predicted that an increase in physical symptoms would be associated with lowered SRH over time, which would then trigger more self-enhancing evaluation as a secondary control strategy to stabilize SRH (i.e., a nonrecursive model).

	METHODS

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
We interviewed the older participants in Study 1 (Time 1, or T1) again in 12 months (Time 2, or T2) by using the same measures. We excluded 6 individuals who could not be contacted, declined participation, or did not complete the follow-up measures, leaving 199 individuals for the longitudinal analysis. Because the results from Study 1 showed that the sociodemographic variables did not independently predict SRH, this study focused only on the interrelationships among social comparison, physical symptoms, and SRH.

We analyzed data by using structural equation modeling. We randomized the 12 items for measuring symptoms into three parcels, each being the average score of the 4 items allocated to the particular parcel. We conducted the same procedure for the items measuring social comparison. Item parcels not only help to correct distribution problems of the individual items and improve the reliability of the indicators (and hence the latent variables under estimation), but they also improve the reliability of the path coefficients because less information has to be estimated from the current, relatively small sample (Little, Cunningham, Shahar, & Widaman, 2002).

As for SRH, we specified the observed variable to be equal to the latent variable with zero residual. In the model to be tested, all T2 variables were predicted by their T1 counterparts. We assigned the amount of variance unexplained by the stability of each variable to a phantom variable, representing the residualized change of the variable. Thus we created three phantom variables for SRH, social comparison, and symptoms (Little, 1997).

At T1, SRH was predicted by both symptoms and social comparison. As in Study 1, this allowed us to assess the effect of social comparison on SRH, controlling for the effect of symptoms in this older sample. For the residualized change variables, SRH was predicted by symptoms to see how symptom change produced a change in SRH. At the same time, we allowed SRH and social comparison to influence each other mutually.

We then subjected the covariance matrix of the indicators to maximum likelihood estimation by use of LISREL version 8.52. We follow Hu and Bentler's (1999) combinational rule for maximum likelihood models. Besides the chi-square (²) statistic, we report the Comparative Fit Index (CFI) and the standardized root mean square residual (SRMR). A well-fitting model is expected to have a CFI 0.95 and an SRMR 0.08 (Hu & Bentler).

	RESULTS AND DISCUSSION

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
Participants reported worse SRH in T2 (M = 3.39, SD = 0.87) than in T1 (M = 3.54, SD = 0.87); t(198) = –2.45, p <.01. There was a tendency to report a more favorable physical self in comparison with same-aged peers in T2 (M = 0.20, SD = 0.69) than in T1 (M = 0.13, SD = 0.71), but it was only marginally significant, t(198) = 1.47, p =.07. There was also no change in reported symptoms 12 months apart; T2, M = 7.20; T1, M = 7.28; both SDs = 7.49; t(198) = –0.21, ns. It should be noted that even though there were no significant changes in social comparison and symptoms across time for the whole sample, the degree and direction of these changes could still vary across individuals. In other words, we could still test whether those with a greater increase in symptoms over time were more likely to perceive their own physical self as better than the physical self of others. We proceed to do so in the next section.

Testing the Structural Equation Model
Prior to analyzing the theoretical model, we performed confirmatory factor analyses to see if the same latent constructs were measured 12 months apart. Results showed that the items were loaded on the same factors at both points in time (i.e., factor-pattern invariance): ²(24, N = 199) = 32.09 (ns), CFI = 0.99, and SRMR = 0.05. Constraining the factor loadings to be equal across time resulted in ²(28, N = 199) = 35.57 (ns), CFI = 0.99, and SRMR = 0.05. As the chi-square change was nonsignificant, we deemed factor loadings to be equivalent, suggesting that the same constructs of symptoms and social comparison were measured at both time points.

Results of the full structural equation model are presented in Figure 2. Although the chi-square, which is sensitive to sample size, was significant, ²(71, N = 199) = 166.01, p <.05, the CFI and the SRMR, being 0.95 and 0.08 respectively, were both within acceptable limits. Thus the model fit the data rather well.

View larger version (27K): [in this window] [in a new window]   Figure 2. Study 2: Structural relationship between social comparison and self-rated health from baseline to 12-month follow-up, with physical symptoms controlled for. All paths were significant at.05

	GENERAL DISCUSSION

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 
We conducted the studies reported herein to examine the effect of perceiving one's physical self as better than that of others on SRH. Study 1 examined this effect among Chinese adults ranging in age across the entire span of adulthood. Findings from the study revealed that perceiving one's own physical self as superior to that of others was positively associated with SRH across age, but more so for older adults. This was the case even after we statistically controlled for the main effect of physical symptoms and the differential effect of physical symptoms on older relative to younger adults (i.e. Symptoms x Age interaction). Study 2 replicated this effect of social comparison on SRH longitudinally. We examined changes in social comparison, SRH, and physical symptoms across a 12-month period among older adults only. In this older sample, the association between social comparison and SRH was even stronger than that between physical symptoms and SRH. More importantly, the longitudinal findings suggest that a decrease in SRH gives rise to a greater use of social comparison to enhance the physical self, which partially buffers against the negative effect of physical symptoms on SRH.

Our studies, perhaps the first examining social comparison in old age in a non-Western culture, contribute to the larger literature on social comparison. Although self-enhancement may be a less salient motive for Asians (Kitayama, Markus, Matsumoto, & Norasakkunkit, 1997; Ross, Heine, Wilson, & Sugimori, 2005), actual self-enhancement or self-effacement depends on domain (Yik et al., 1998). When it comes to attributes that are considered important to the self, self-enhancement would be more likely even in collectivistic cultures (Sedikides, Gaertner, & Toguchi, 2003). To the extent that physical well-being is a fundamental concern universally, the present studies reveal a general tendency, regardless of age, to see one's physical self as better than that of age peers among the Chinese population.

What is not clear from the literature, however, is whether the effect of self-enhancement changes with age. Moreover, the potential role of social comparison in enhancing SRH has not been systematically examined in the context of aging. Aging provides an excellent context for studying social comparison and SRH, because physical decline is both noticeable and inevitable with age. Although there are things that one can do to delay the process, such as by regular exercise and proper dieting (strategies referred to as primary control by Heckhausen and Schulz, 1995), the degree of control over one's health eventually diminishes with age. Under such circumstances, the primary motive would be to stabilize the image of one's health through modifying the way health is evaluated, such as by comparing oneself to less fortunate others. However, as Study 1 shows, age is not associated with portraying a wider gap in physical functioning between self and others. Rather, with age, social comparison takes on a more salient role in determining SRH (Study 1), and indeed, it is effective in partially compensating for the negative effect of increased physical symptoms over time in later life (Study 2). An unexpected finding in Study 1, one that echoes the literature on the increasing disagreement between subjective and objective health (Borchelt et al., 1999; Idler, 1993; Pinquart, 2001), is that, compared with younger persons, older persons do not rely so much on the extent of physical symptoms to evaluate their SRH. Together, these mechanisms help to regulate SRH vis-à-vis the challenges of declining health in later life (Brandtstädter & Greve, 1994; Brandtstädter & Rothermund, 2002) and may contribute to the paradox of aging and well-being—the maintenance of subjective health and well-being by older people at a level that is at least comparable with that of younger people despite undeniable declines and losses. It is therefore no wonder that SRH hardly declines in old age (Leinonen et al., 1998; Maddox & Douglass, 1973), despite moderate increase in physical symptoms (Aldwin et al., 1989).

We acknowledge three limitations in the present study. First, the sample might have suffered from some selection biases. The younger and middle-aged participants were recruited from formal and informal contacts of recent university graduates and the older participants were recruited from senior centers. Both methods of recruitment were likely to result in participants who were healthier than the general population. Although the fact that even this healthier sample tended to engage in self-enhancement in the physical domain made this finding particularly noteworthy, we cannot rule out the possibility that the findings may not be generalizable to the general population. Second, the procedure for data collection was not entirely the same across the age groups (i.e., self-administration for young and midlife adults, face-to-face interview for older adults). This might bias the findings by discouraging self-enhancement in the interview situation, thus leading to a lack of correlation between social comparison and age. This bias was minimized because the purpose of assessing social comparison was not apparent to the participants who provided independent ratings to self and other. Nonetheless, future research should try to use the same data-collection procedure across age groups. Finally, although the cross-sectional data of Study 1 did not allow more definitive inferences about causal relationships to be drawn, the longitudinal data of Study 2 were limited to older adults only. Longitudinal studies that include participants of a wider age range are needed in the future to examine the directions of the relationships among social comparison, SRH, and physical symptoms across age.

Despite these limitations, our findings contribute to the literature on SRH in two ways. First, the cross-sectional findings suggest that seeing one's physical self as superior to that of others has an enhancing effect on SRH, particularly for older adults. Second, longitudinal findings demonstrate the buffering role of such a self-enhancing evaluation on the SRH of older adults, against the damaging effect of increased physical symptoms over time. Both of these mechanisms may help to explain the puzzling observation that SRH remains relatively stable in old age despite moderate declines in physical health.

	Footnotes

 
Decision Editor: Karen Hooker, PhD

Received for publication October 27, 2006. Accepted for publication April 16, 2007.

	References

TOP Abstract Study 1 Results and Discussion Study 2 Methods Results and Discussion General Discussion References

 

• Aldwin, C. M., Spiro, A., Levenson, M. R., Rosse, R. (1989). Longitudinal findings from the normative aging study: I. Does mental health change with age? Psychology and Aging, 4,295-306.[Medline]
• Baron-Epel, O., Kaplan, G. (2001). General subjective health status or age-related subjective health status: Does it make a difference? Social Science & Medicine, 53,1373-1381.
• Beach, S. R. H., Tesser, A. (1995). Self-esteem and the extended self-evaluation maintenance model: The self in social context. In M. H. Kernis (Ed.), Efficacy, agency, & self-esteem (pp. 145–170). New York: Plenum Press.
• Borchelt, M., Gilberg, R., Horgas, A. L., Geiselmann, B. (1999). On the significance of morbidity and disability in old age. In P. B. Baltes & K. U. Mayer (Eds.), The Berlin aging study: Aging from 70 to 100 (pp. 403–429). Cambridge, England: Cambridge University Press.
• Bradshaw, C. P., Hazan, C. (2006). Examining views of self in relation to views of others: Implications for research on aggression and self-esteem. Journal of Research in Personality, 40,1209-1218.
• Brandtstädter, J., Greve, W. (1994). The aging self: Stabilizing and protective processes. Developmental Review, 14,52-80.
• Brandtstädter, J., Rothermund, K. (2002). The life-course dynamics of goal pursuit and goal adjustment: A two-process framework. Developmental Review, 22,117-150.
• Brandtstädter, J., Renner, G. (1990). Tenacious goal pursuit and flexible goal adjustment: Explication and age-related analysis of assimilative and accommodation strategies of coping. Psychology and Aging, 5,58-67.[Medline]
• Buunk, B. P., Collins, R. L., Taylor, S. E., Van Yperen, N., Dakof, G. (1990). The affective consequences of social comparison: Either direction has its ups and downs. Journal of Personality and Social Psychology, 59,1238-1249.[Medline]
• Census and Statistics Department. (2002). Hong Kong 2001 population census main report, Vols. I & II. Hong Kong: Author.
• Cheng, S.-T. (2004). Age and subjective well-being revisited: A discrepancy perspective. Psychology and Aging, 19,409-415.[Medline]
• Cheng, S.-T., Chan, A. C. M. (2006). Social support and self-rated health revisited: Is there a gender difference in later life? Social Science & Medicine, 63,118-122.
• Cheng, S.-T., Hamid, P. N. (1996). A Chinese symptom checklist: Preliminary data concerning reliability and validity. Journal of Social Behavior and Personality, 11,241-252.
• Chiu, H. F. K., Lee, H. B., Chung, W. S., Kwong, P. K. (1994). Reliability and validity of the Cantonese version of Mini-Mental State Examination: A preliminary report. Journal of the Hong Kong College of Psychiatrists, 4,(Suppl. 2), 25-28.
• Cohen, J., Cohen, P., West, S., Aiken, L. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Erlbaum.
• Cross, S., Markus, H. (1991). Possible selves across the life span. Human Development, 34,230-255.
• Deeg, D. J. H., Kriegsman, D. M. W. (2003). Concepts of self-rated health: Specifying the gender difference in mortality risk. Social Science & Medicine, 43,376-386.
• Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7,117-140.
• Folstein, M. F., Folstein, S. E., McHugh, P. R. (1975). "Mini-Mental State": A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12,189-198.[Medline]
• Heckhausen, J. (1999). Developmental regulation in adulthood: Age-normative and sociostructural constraints as adaptive challenges. Cambridge, England: Cambridge University Press.
• Heckhausen, J., Brim, O. G. (1997). Perceived problems for self and others: Self-protection by social downgrading throughout adulthood. Psychology and Aging, 12,610-619.[Medline]
• Heckhausen, J., Krueger, J. (1993). Developmental expectations for the self and most other people: Age grading in three functions of social comparison. Developmental Psychology, 29,539-548.
• Heckhausen, J., Schulz, R. (1995). A life-span theory of control. Psychological Review, 102,284-304.[Medline]
• Herzog, A. R., Rodgers, W. L. (1981). Age and satisfaction: Data from several large surveys. Research on Aging, 3,142-165.[Abstract]
• Hu, L., Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis' conventional criteria versus new alternatives. Structural Equation Modeling, 6,1-55.
• Idler, E. L. (1993). Age differences in self-assessments of health: Age changes, cohort differences, or survivorship? Journal of Gerontology: Social Sciences, 48,S289-S300.
• Idler, E. L., Benyamini, Y. (1997). Self-rated health and mortality: A review of twenty-seven community studies. Journal of Health and Social Behavior, 38,21-37.[Medline]
• Idler, E. L., Russell, I. B., Davis, D. (2000). Survival, functional limitations, and self-rated health in the NHANES I epidemiologic follow-up study, 1992. American Journal of Epidemiology, 152,874-883.[Abstract/Free Full Text]
• Kaplan, G. A., Strawbridge, W. J., Camacho, T., Cohen, R. D. (1993). Factors associated with change in physical functioning in the elderly. Journal of Aging and Health, 5,140-153.[Abstract/Free Full Text]
• Kitayama, S., Markus, H. R., Matsumoto, H., Norasakkunkit, V. (1997). Individual and collective processes in the construction of the self: Self-enhancement in the United States and self-criticism in Japan. Journal of Personality and Social Psychology, 72,1245-1267.[Medline]
• Klein, W. M. P. (2001). Post hoc construction of self-performance and other performance in self-serving social comparison. Personality and Social Psychology Bulletin, 27,744-754.[Abstract/Free Full Text]
• Klein, W. M. P., Monin, M. M., Steers-Wentzell, K. L., Buckingham, J. T. (2006). Effects of standards on self-enhancing interpretations of ambiguous social comparison information. Basic & Applied Social Psychology, 28,65-79.
• Kling, K. C., Ryff, C. D., Essex, M. J. (1997). Adaptive changes in the self-concept during a life transition. Personality and Social Psychology Bulletin, 23,981-990.[Abstract]
• Kunzmann, U., Little, T. D., Smith, J. (2000). Is age-related stability of subjective well-being a paradox? Cross-sectional and longitudinal evidence from the Berlin Aging Study. Psychology and Aging, 15,511-526.[Medline]
• Leinonen, R., Heikkinen, E., Jylhä, M. (1998). Self-rated health and self-assessed change in health in elderly men and women—A five-year longitudinal study. Social Science & Medicine, 46,591-597.
• Little, T. D. (1997). Mean and covariance structures (MACS) analyses of cross-cultural data: Practical and theoretical issues. Multivariate Behavioral Research, 32,53-76.[Medline]
• Little, T. D., Cunningham, W. A., Shahar, G., Widaman, K. F. (2002). To parcel or not to parcel: Exploring the question, weighting the merits. Structural Equation Modeling, 9,151-173.
• Maddox, G. L., Douglass, E. B. (1973). Self-assessment of health: A longitudinal study of elderly subjects. Journal of Health and Social Behavior, 14,87-93.[Medline]
• McFarland, C., Buehler, R., MacKay, L. (2001). Affective responses to social comparisons with extremely close others. Social Cognition, 19,547-586.
• Miller, T. R., Wolinsky, F. D. (2007). Self-rated health trajectories and mortality among older adults. Journal of Gerontology: Social Sciences, 62B,S22-S27.[Abstract/Free Full Text]
• Mussweiler, T., Ruter, K. (2003). What friends are for! The use of routine standards in social comparison. Journal of Personality and Social Psychology, 85,467-481.[Medline]
• Pinquart, M. (2001). Correlates of subjective health in older adults: A meta-analysis. Psychology and Aging, 16,414-426.[Medline]
• Ross, M., Heine, S. J., Wilson, A. E., Sugimori, S. (2005). Cross-cultural discrepancies in self-appraisals. Personality and Social Psychology Bulletin, 31,1175-1188.[Abstract/Free Full Text]
• Schulz, R., Heckhausen, J. (1996). A life-span model of successful aging. American Psychologist, 51,702-714.[Medline]
• Sedikides, C., Gaertner, L., Toguchi, Y. (2003). Pancultural self-enhancement. Journal of Personality and Social Psychology, 84,60-79.[Medline]
• Spiers, N., Jagger, C., Clarke, M., Arthur, A. (2003). Are gender differences in the relationship between self-rated health and mortality enduring? Results from three birth cohorts in Melton Mowbray, United Kingdom. The Gerontologist, 43,406-411.[Abstract/Free Full Text]
• Svedberg, P., Gatz, M., Lichtenstein, P., Sandin, S., Pedersen, N. L. (2005). Self-rated health in a longitudinal perspective: A 9-year follow-up twin study. Journal of Gerontology: Social Sciences, 60B,S331-S340.[Abstract/Free Full Text]
• Taylor, S. E., Lobel, M. (1989). Social comparison activity under threat: Downward evaluation and upward contacts. Psychological Review, 96,569-575.[Medline]
• Wheeler, L., Miyake, K. (1992). Social comparisons in everyday life. Journal of Personality and Social Psychology, 62,760-773.
• Wills, T. A. (1981). Downward comparison principles in social psychology. Psychological Bulletin, 90,245-271.
• Yik, M. S. M., Bond, M. H., Paulhus, D. L. (1998). Do Chinese self-enhance or self-efface? It's a matter of domain. Personality and Social Psychology Bulletin, 24,399-406.[Abstract]

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation