Self-compassionate Responses to Aging

Abstract

Purpose: Evidence suggests that self-compassion may be beneficial to older adults who are struggling to cope with the aging process. The purpose of this study was to assess the thoughts of self-compassionate older adults and to determine whether self-compassionate thoughts relate to positive responses to aging. Design and Methods: Participants (n = 121, M = 76.2 years, approximately 65% female) completed measures of self-compassion and self-esteem; were randomly assigned to write about a positive, negative, or neutral age-related event; and completed questions about the event and their reactions. Responses were coded for self-compassionate themes and emotional tone. Results: Analyses indicated that self-compassion predicted positive responses to aging and that self-compassionate thoughts explained the relationship between trait self-compassion and emotional tone as well as the belief that one’s attitude helped them cope with age-related events. Implications: Although older adults who were low versus high in self-compassion experienced similar age-related events, participants high in self-compassion thought about these events in ways that predicted positive outcomes. Encouraging older adults to be more self-compassionate may improve well-being in old age.

As people get older, the likelihood of experiencing stressors such as the death of a partner or friend, health problems, and lowered ability to engage in enjoyable and fulfilling activities increases (Baltes & Baltes, 1990; Lee & Markides, 1990). Yet, despite the physical, mental, and social challenges that often accompany aging, people show great variability in how they cope with these changes, and many older adults report high subjective well-being (Blazer, 2008; George, Okun, & Landerman, 1985; Kunzmann, Little, & Smith, 2000; Strawbridge, Wallhagen, & Cohen, 2002). Thus, researchers have been interested in factors that are associated with taking a proactive approach to staying healthy and maintaining well-being (Kahana & Kahana, 2003). What emotional, cognitive, and behavioral mechanisms differentiate older individuals who deal effectively with aging from those who do not? Some such mechanisms—such as perceived control, social support, and self-esteem—have been shown to serve as protective factors in the aging process (Heckhausen & Schulz, 1995; Krause, 2007), but a great deal of variance in psychological well-being remains unexplained.

The purpose of this research was to examine the role of self-compassion in well-being as people age. Self-compassion involves treating oneself compassionately—with caring, concern, and kindness—when problems arise in life, much like people treat loved ones who experience difficulties. Neff (2003a) defined self-compassion as “being open to and moved by one’s own suffering, experiencing feelings of caring and kindness toward oneself, taking an understanding, nonjudgmental attitude toward one’s inadequacies and failures, and recognizing that one’s experience is part of the common human experience” (p. 224). As a trait, self-compassion is associated with patterns of motives, cognitions, and behaviors that are associated with treating oneself with caring and concern when things go wrong. People who are self-compassionate treat themselves kindly, do not engage in undue self-criticism or self-pity, resist personalizing their difficulties, and maintain an objective perspective on the events that they experience. Individual differences in self-compassion are measured using the Self-compassion Scale (Neff, 2003a), which has been shown to be both reliable and valid in numerous studies (for a review, see Neff, 2009). Self-compassion is associated with adaptive cognitive, emotional, and behavioral processes in young adults (Neff, 2003b), and evidence suggests that it may be associated with well-being in older adults as well (Allen, Goldwasser, & Leary, 2012).

Studies of young adults show that self-compassion strongly correlates with indicators of mental health and adjustment, including lower depression, lower anxiety, and higher life satisfaction (Neff, 2003b). It also has medium to strong positive associations with happiness, optimism, wisdom, initiative, and agreeableness, and a strong negative relationship with neuroticism (Neff, Rude, & Kirkpatrick, 2007). Furthermore, studies in which people who score low versus high in self-compassion are led to experience negative events (such as failure, rejection, awkward interactions, or reminders of previous stressful events) show that self-compassion moderates reactions to unpleasant and difficult situations (Leary, Tate, Adams, Allen, & Hancock, 2007).

Some researchers have wondered whether self-compassionate responses involve indifference, indulgence, or a refusal to accept responsibility. However, research shows that people who are self-compassionate are more likely to accept responsibility for their mistakes and failures than those who are less self-compassionate (Leary et al., 2007). Importantly, although self-compassion correlates strongly with trait self-esteem (presumably because people who treat themselves kindly feel better about themselves), the relationship between self-compassion and psychological well-being is independent of self-esteem (Leary et al., 2007; Neff, 2003b). Indeed, self-compassion may uniquely predict well-being more strongly than self-esteem does (Leary et al., 2007).

Self-compassion seems a likely predictor of well-being among older people and, in fact, self-compassion may become increasingly important as people age. As noted, aging involves undesired life changes, such as declines in mental and physical health, many of which involve lapses or failures, as when people display memory failures or cannot perform everyday tasks. Many older people become self-critical and angry, castigating themselves, and bemoaning their inability to function as they once did (Mirowsky & Ross, 1992). Because self-compassion attenuates reactions to negative events—both those for which the person has personal responsibility and those that are beyond one’s control (Leary et al., 2007, Study 1)—self-compassion may help older people deal with an array of negative events.

Allen and colleagues (2012) assessed the relationship between self-compassion, life conditions, and subjective well-being among older adults. High self-compassion was related to general well-being, successful aging, fewer emotional problems, and life satisfaction, and self-compassion was particularly beneficial for participants who reported more negative life circumstances. Participants high in self-compassion maintained higher well-being and life satisfaction when reporting greater physical pain, poorer hearing, more physical problems, or poorer health than those low in self-compassion. Given that self-compassion moderated the relationship between physical health and subjective well-being, it may explain some of the variance in how people deal with the negative aspects of aging.

The effects obtained by Allen and colleagues (2012) are probably due both to the psychological coping strategies that are inherent in the self-compassionate mindset (Allen & Leary, 2010) and to the fact that people who are high in self-compassion engage in behaviors that promote their well-being under difficult circumstances. Just as people show others compassion both by encouraging them to think about negative events in adaptive ways and by treating them kindly in ways that promote their well-being, people high in self-compassion may respond to their own problems in similarly caring ways. For example, Allen and colleagues (2012) found that older people who were higher in self-compassion were more willing to ask people to repeat themselves when they could not hear what they said, and among people who had difficulty walking, those who were higher in self-compassion were more willing to use a walker. These results suggest that people who are high in self-compassion are more accepting of their physical limitations and more willing to take steps to maintain their well-being (Leary et al., 2007; Neff, 2003a; Neff et al., 2007).

The present study was designed to examine the nature of the cognitions that underlie self-compassion vis-à-vis aging. What are people high in self-compassion thinking that leads them to treat themselves more kindly and to respond in a more adaptive manner to adversity? Understanding the nature of the self-compassionate mindset may be beneficial for helping older people think about themselves and about getting older in ways that maintain well-being.

Research on young adults shows that people high in self-compassion think about negative events differently than people low in self-compassion. Leary and colleagues (2007, Study 1) asked participants to report four negative events over a 3-week period and to rate how likely they would be to think certain thoughts. Self-compassionate participants were less likely to endorse thoughts such as “I seem to have bigger problems than most people do” and “Why do these things always happen to me?” and more likely to endorse the thought “This isn’t any worse than what lots of other people go through.” In another study, self-compassionate participants were less likely to have catastrophizing and personalizing thoughts, and more likely to have thoughts that promoted equanimity (Leary et al., 2007, Study 2). These studies provide a glimpse into the cognitions of people who are highly self-compassionate, which the present study extended by assessing the relationship between self-compassion and self-related thoughts in older adults.

This study addressed three questions: (a) How do the cognitions of more self-compassionate older adults differ from those who are less self-compassionate? (b) Do these cognitions differ because self-compassionate older adults downplay or ignore the negative factors associated with aging or because they fully recognize those factors but think about them in a compassionate manner? (c) Do these self-compassionate cognitions mediate the relationship between trait self-compassion and people’s emotional reactions? If people who score high in self-compassion think differently about their life situations as well as respond more adaptively to negative events, then presumably their thoughts mediate the relationship between trait self-compassion and their responses to aging.

Methods

Participants

One hundred and thirty-one participants were recruited from the local community via pubic fliers at senior citizen centers, word of mouth, previous participation, and the university’s aging registry. Of these 131 individuals, 121 (38 men, 79 women, 4 unknown) provided sufficient information to be usable by completing the Self-compassion Scale and writing about the age-related event (to be described). Participants ranged from 63 to 93 years in age (M = 76.2, standard deviation [SD] = 6.74). Ninety-one (75.6%) were Caucasian, 28 (23.5%) were African American, 1 (0.8%) was Asian American, and 1 participant did not indicate his or her race.

Procedure

Participants who indicated that they would like to participate received a packet by mail that contained an information sheet, informed consent form, questionnaire, and payment form. After participants returned their materials in the prepaid envelope, they received $10. The questionnaire was designed to take 45–60min, but we do not know how much time participants spent or the conditions under which it was completed.

Individual Differences.

Participants first completed the Brief Self-compassion Scale (BSCS; Raes, Neff, & Van Gucht, 2011), which consists of 12 items from Neff’s (2003a) original measure (the four items that correlated most highly with each of the three subscales). The BSCS is an economic alternative to the original, 26-item Self-compassion Scale and correlates .78 to .92 with the longer scale. In addition, the BSCS has demonstrated adequate reliability and validity with both college students and middle-aged adults (Raes et al., 2011). They also completed Rosenberg’s (1965) Self-esteem Scale, the most widely used measure of trait self-esteem (Blascovich & Tomaka, 1991).

Experimental Manipulation.

The participant was instructed to recall an age-related change or event that he or she had experienced. Participants were randomly assigned to one of three conditions that differed in how the question was framed. Participants in the positive event condition were asked to think about a positive age-related event, participants in the negative event condition were asked to think about a negative age-related event, and participants in the control condition were merely asked to think about an age-related event they had experienced. Examples of events provided by participants in the positive event condition included “retirement,” “yoga class,” and “seeing my granddaughters get married.” Participants in the negative event condition frequently reported physical and mental limitations such as “arthritis,” “loss of hearing,” and “memory loss.” Participants in the control condition were asked to think about any age-related event, but these events tended to be negative, such as “knee replacement,” “heart attack,” and “back pain.”

Outcome Variables.

Participants were then asked questions regarding their thoughts and feelings about the event. They rated how good or bad the event was (1 = extremely bad; 7 = extremely good), how good or bad the event made them feel about themselves (1 = extremely bad; 7 = extremely good), how much the event affected their life (1 = no impact at all; 7 = huge impact), and how much they thought about the change or event (1 = never; 5 = a great deal). Participants also rated the degree to which four factors helped them cope with the event, including their own attitude, realizing that the change or event was normal, trying to be nice to themselves, and deliberately trying to control their emotions (1 = not at all; 5 = extremely).

Open-ended questions asked participants to write about (a) their thoughts and feelings regarding the event, (b) three general thoughts about getting older, (c) three ways in which their thoughts about aging had become more positive or negative since they were 40 years old, and (d) three pieces of advice that they would give to people in their 30s, 40s, and 50s. Coders independently assessed the emotional tone of each statement (1 = very negative; 5 = very positive) as well as the presence of six basic features of self-compassion identified by Neff (2003a)—common humanity, isolation, self-kindness, self-judgment, mindfulness, and overidentification (0 = clearly not present; 1 = may be present; 2 = clearly present; kappa = .72). Common humanity versus isolation reflects people’s recognition that a negative experience is part of the human experience rather than being unique to oneself. Self-kindness versus self-judgment reflects being understanding and caring toward oneself rather than harsh and self-critical. Mindfulness versus overidentification reflects experiencing emotions without becoming carried away with them. Coders’ ratings of the six facets of self-compassion were used to create four quantitative values for each participant that reflected the participants’ self-compassionate cognitions regarding (a) the event, (b) thoughts about aging, (c) changes associated with aging, and (d) advice to younger adults.

Results

To confirm that participants wrote about positive versus negative events as instructed, two coders independently rated how objectively good or bad the event was (1 = extremely bad; 7 = extremely good). A one-way analysis of variance showed that, as intended, participants in the positive event condition (M = 4.9, SD = 1.25) reported events that were more positive than participants in the control (M = 2.8, SD = 1.08) and negative (M = 2.8, SD = 0.83) conditions, F(2,118) = 52.30, p < .001. Ratings of events in the control and negative conditions did not differ, suggesting that when asked simply to discuss an event associated with aging, participants tended to choose negative events.

Self-compassionate Cognitions

Table 1 provides means, SDs, and correlations between trait self-compassion, trait self-esteem, positive and negative self-compassionate cognitions for each of the four writing prompts, and the emotional tone of each of the four questions.

Table 1.

Means, Standard Deviations, and Correlations

	M	SD	SC	SE	PE	NE	PT	NT	PC	NC	PA	NA	TE	TT	TC
SC	3.67	0.55	—
SE	4.21	0.54	.58*	—
PE	1.08	1.29	.31*	.23**	—
NE	.70	1.15	−.26**	−.32*	−.48*	—
PT	1.82	2.05	.22**	.22**	.14	−.02	—
NT	1.29	1.73	−.16***	−.23**	−.09	.22***	−.50*	—
PC	2.22	1.96	.36*	.31**	.16	−.11	.37*	−.24***	—
NC	.62	1.04	−.27**	−.39*	−.16	.20***	−.30**	.33**	−.47*	—
PA	3.15	1.46	.13	.29**	.28**	−.26**	.22***	−.03	.36*	−.18***	—
NA	.13	0.72	−.14	−.25**	−.15	.06	−.10	.02	−.19***	.39*		—
TE	3.21	1.37	.32*	.25**	.64*	−.71*	.09	−.15	.07	−.11	.25**	−.15	—
TT	3.15	0.91	.25**	.33*	.15	−.12	.73*	−.77*	.39*	−.31*	.18***	−.16***	.13	—
TC	3.52	0.78	.27**	.37*	.27**	−.33*	.35*	−.35*	.69*	−.75*	.37*	−.43*	.28**	.45*	—
TA	3.91	0.47	.13	.18***	.10	−.16	.09	−.05	.26**	−.35*	.49*	−.61*	.20***	.14	.38*

Notes: NA = negative self-compassionate cognitions regarding advice to younger adults; NC = negative self-compassionate cognitions about changes associated with aging; NE = negative self-compassionate cognitions regarding the event; NT = negative self-compassionate cognitions regarding thoughts about aging; PA = positive self-compassionate cognitions regarding advice to younger adults; PC = positive self-compassionate cognitions about changes associated with aging; PE = positive self-compassionate cognitions regarding the event; PT = positive self-compassionate cognitions regarding thoughts about aging; SC = self-compassion; SE = self-esteem; TA = tone of advice to younger adults; TC = tone of changes associated with aging; TE = tone of the event; TT = tone of thoughts about aging.

*p < .001. **p < .01. ***p < .05.

Self-compassion was significantly associated with the positive emotional tone of participants’ responses for the event (r = .32), thoughts about aging (r = .25), and changes associated with aging (r = .27), but not for advice given to younger adults. Self-compassion was associated with more positive self-compassionate thoughts regarding the event (r = .31), general thoughts about aging (r = .22), and changes associated with aging (r = .36). Self-compassionate participants also generated fewer negative thoughts regarding the event (r = −.26), aging (r = −.16), and changes associated with aging (r = −.27). Self-compassion was not related to positive self-compassionate cognitions or negative self-compassionate cognitions regarding advice to younger adults, although both effects approached significance, all p’s > .10.

Consistent with previous research (Leary et al., 2007), self-compassion was not related to how negatively participants rated the event that they described, how bad the event made participants feel about themselves, how much of an impact the event had on their life, or how much they thought about the event, all p’s > .10. Thus, participants who were low versus high in self-compassion did not differ in the nature of the events they reported, even though they thought about them in quite different ways.

Regression Analyses

Four hierarchical multiple regression analyses assessed whether self-compassion, writing condition (positive, negative, control), and their interaction predicted self-compassionate cognitions regarding (a) the event, (b) thoughts about aging, (c) changes associated with aging, and (d) advice to younger adults. Trait self-esteem was entered in Step 1 as a covariate, self-compassion was entered in Step 2, dummy coded condition variables were entered in Step 3, and the self-compassion by condition interaction terms were entered into Step 4.

As shown in Table 2, with variance due to self-esteem removed, self-compassion significantly predicted self-compassionate cognitions regarding the event and changes associated with aging. Writing condition significantly predicted self-compassionate cognitions about the event, showing that participants in the positive writing condition expressed more self-compassionate cognitions than participants in the neutral and negative conditions. The self-compassion by condition interaction was not significant for any of the outcome variables suggesting that the valence of the prompt did not influence the relationships.

Table 2.

Regression Analyses Examining the Relationship Between Self-compassion, Condition, and Their Interaction on Self-compassionate Cognitions

	Self-compassionate cognitions
	Event		General thoughts		Changes		Advice
Predictor	∆R 2	β	∆R 2	β	∆R 2	β	∆R 2	β
Step 1	.10*		.07*		.12*		.07**
SE		.31*		.26***		.35*		.27**
Step 2	.03***		.01		.04***		.001
SC		.22***		.11		.26***		−.03
Step 3	.18***		.001		.04		.01
Condition D1		−.37*		.001		.001		−.08
Condition D2		−.47*		.04		.21		−.12
Step 4	.02		.01		.001		.02
SC × Cond D1		.22		.04		−.04		−.07
SC × Cond D2		.08		.13		.02		−.12
Total R 2	.33*		.09**		.21*		.10**
n	119		111		96		108

Notes: Cond D1 = dummy coded condition variable comparing the neutral and positive conditions; Cond D2 = dummy coded condition variable comparing the negative and positive conditions; SC = self-compassion; SE = self-esteem.

*p < .001. **p < .01. ***p < .05.

Mediation Analyses

To examine whether self-compassionate cogni tions mediate the relationship between self-compassion and participants’ responses, analyses were conducted with self-compassion as the predictor, positive and negative self-compassionate cognitions regarding the event as the mediators, and self-esteem as a covariate. Both positive and negative self-compassionate cognitions were included as potential mediators because debate exists regarding whether the positive or negative self-compassion subscales are more relevant to emotional well-being (Gilbert & Procter, 2006). Following the recommendations by Preacher and Hayes (2008), the significance of the indirect effects was estimated using a bootstrapping technique. Unlike the Sobel test, bootstrap estimation is unaffected by nonnormality, thus producing more accurate parameter estimates. Bootstrapping involves taking n cases with replacement from the sample and reestimating the relationships k number of times. For our purposes, k = 5,000, thus resulting in 5,000 estimates of the total and specific indirect effects of X on Y. These estimates of the indirect effect are then sorted from low to high with 100(α/2) defining the lower and upper confidence intervals (CIs). Evidence for mediation is obtained if the CI does not include zero.

Emotional Tone.

Self-compassionate cognitions were examined as potential mediators of the relationship between self-compassion and the emotional tone of participants’ responses. Self-compassion predicted positive self-compassionate cognitions (a₁ path = .63, p = .014) but not negative ones (a₂ path = −.23, p = .31). Both positive (b₁ path = .39, p < .001) and negative (b₂ path = −.62, p < .001) self-compassionate cognitions predicted the emotional tone of participants’ responses. The total effect of self-compassion on emotional tone was significant (c path = .66, p = .015), and when the mediators were entered into the analysis, the direct effect of self-compassion on emotional tone was no longer significant, suggesting full mediation (c′ path = .27, p = .136). Self-esteem was not a significant covariate in this model (B = −.18, p = .328). Point estimates and CIs are provided in Table 3.

Table 3.

Mediation Effects of Self-compassion on Emotional Tone

			Bootstrapping
			Percentile 95% CI		BC 95% CI		BCa 95% CI
	Point estimate (a*b)	SE (a*b)	Lower	Upper	Lower	Upper	Lower	Upper
Indirect effects
Positive SC cognitions	.250	.102	.051	.455	.069	.475	.068	.474
Negative SC cognitions	.142	.186	−.195	.526	−.219	.506	−.297	.462
Total	.392	.241	−.066	.872	−.087	.858	−.154	.820

Notes: n = 120. SC = self-compassion; SE = self-esteem; BC = Bias-corrected; BCa = Bias-corrected and accelerated.

Self-perceived Influences on Coping.—

Participants indicated the degree to which four factors helped them deal with the event they described—their attitude, recognizing that the change or event was normal, trying to be nice to oneself, and efforts to control one’s emotions. As before, mediation analyses assessed the degree to which self-compassionate cognitions mediated the relationship between self-compassion and the perceived influences on coping.

On the analysis of participant’s attitude as a coping factor, self-compassion predicted positive self-compassionate cognitions (a₁ path = .70, p = .007) but not negative ones (a₂ path = −.28, p = .22). Positive (b₁ path = .22, p = .01) but not negative self-compassionate cognitions (b₂ path = .01, p = .903) predicted the degree to which participants thought their attitude helped them cope with the event. The total effect of self-compassion was significant (c path = .47, p = .04), but the direct effect was not significant after accounting for the mediating variables (c′ path = .32, p = .16). Self-esteem was not a significant covariate (B = .05, p = .81). The point estimates and CIs for the coefficients are presented in Table 4. The CI for positive self-compassionate cognitions does not include zero suggesting that they mediate the relationship between self-compassion and the extent to which participants believed that their attitude helped them cope with the age-related event.

Table 4.

Mediation of the Effect of Self-compassion on the Extent to Which Participants Said Their Attitude Helped Them Adjust to the Age-Related Event Through Positive Self-compassionate Cognitions and Negative Self-compassionate Cognitions

			Bootstrapping
			Percentile 95% CI		BC 95% CI		BCa 95% CI
	Point estimate (a*b)	SE (a*b)	Lower	Upper	Lower	Upper	Lower	Upper
Indirect effects
Positive SC cognitions	.154	.081	.015	.328	.029	.363	.026	.352
Negative SC cognitions	−.003	.042	−.112	.068	−.114	.066	−.101	.084
Total	.151	.079	.003	.312	.029	.027	.027	.364

Notes: n = 119. SC = self-compassion; SE = self-esteem; BC = Bias-corrected; BCa = Bias-corrected and accelerated.

Consistent with the conceptualization of self-compassion (Neff, 2003b), self-compassion predicted the extent to which participants coped with the event by trying to be nice to themselves (c path = .66, p < .001), but the positive and negative self-compassionate cognition mediators did not predict the outcome variable, showing no evidence of mediation (b₁ path = .06, p = .41; b₂ path = −.05, p = .53, respectively).

Self-compassion did not predict the degree to which participants coped with the event by thinking of it as normal (c path = .31, p = .14) or the extent to which participants coped with the event by controlling their emotions (c path = .32, p = .17), and there were no significant mediator-to-outcome relationships.

Discussion

The primary goals of this study were to examine how older adults who differ in self-compassion think differently about age-related events and to test whether the nature of their thoughts mediate the relationship between trait self-compassion and how people cope emotionally with such events. Overall, the findings support the hypothesis that older adults who are high in self-compassion think differently about events that are associated with aging than those who are low in self-compassion. Compared with participants low in self-compassion, self-compassionate participants reported more thoughts that involved positive self-compassionate cognitions when discussing an age-related event, their general thoughts about getting older, and how their thoughts about aging have changed since turning 40. Conversely, self-compassionate participants expressed fewer negative thoughts about these events. Consequently, self-compassionate participants expressed greater overall positivity in their responses. In these ways, the cognitive concomitants of self-compassion among older adults are similar to those found among younger people (Leary et al., 2007). These effects were not moderated by experimental condition, showing that self-compassion predicted self-compassionate thoughts when writing about both positive and negative age-related events. In general, self-compassionate individuals generated more positive and more self-compassionate thoughts, findings that are consistent with the idea that people high in self-compassion engage in positive cognitive restructuring in which they identify and challenge maladaptive thoughts (Allen & Leary, 2010).

Importantly, low and high self-compassionate people reported equally negative events (i.e., they rated the events they described as equally bad) and indicated that those events had an equally strong impact on their lives. Thus, what distinguishes people who are low versus high in self-compassion is not their propensity to experience negative events or even their evaluations of those events but rather the ways in which they think about them. This pattern is consistent with research on young adults (Leary et al., 2007).

One variable for which self-compassion was not related to cognitions involved the advice that participants said they would give about aging to younger adults. Yet, the average number of positive self-compassionate thoughts was notably higher on this measure than those expressed in answers to the other prompts. Perhaps participants encouraged other people in self-compassionate ways irrespective of their own propensity to be self-compassionate. Consistent with other research, this finding suggests that people who are low in self-compassion may respond as positively toward other people as high self-compassion people do (Leary et al., 2007).

The mediation analyses showed that positive self-compassionate cognitions related to the event mediated the relationship between self-compassion and the emotional tone of participants’ responses, supporting the hypothesis that self-compassionate participants respond more positively because of the self-compassionate themes present in their thoughts. Furthermore, positive self-compassionate cognitions mediated the relationship between self-compassion and the extent to which people reported that their attitude helped them adjust to the event that they described. Self-compassionate participants think more self-compassionately, and, in turn those thoughts lead to a more positive attitude toward their experiences.

Notably, our hypotheses were supported while controlling for trait self-esteem. Self-compassion and self-esteem are highly correlated (Leary et al., 2007; Neff, 2003a; Neff, Pisitsungkagarn, & Hseih, 2008), but they are distinct at both a theoretical and an empirical level. In particular, self-compassion predicts well-being over and above self-esteem, showing that the benefits of self-compassion exceed those of self-esteem (Leary et al., 2007; Neff, 2003a; Neff & Vonk, 2009). At a conceptual level, negative events threaten self-esteem by challenging one’s sense of acceptance or superiority (Gilbert & Irons, 2005; Leary, 2006). In contrast, Gilbert and Irons (2005) proposed that in the face of negative events, self-compassion deactivates the threat system and helps people engage in thoughts and feelings characterized by self-warmth. In addition, research suggests that older adults who have higher self-esteem may ignore aspects of themselves that are in decline (Sneed & Krauss Whitbourne, 2001). Although this strategy promotes self-esteem, turning a blind eye to one’s problems and weaknesses is often not adaptive. Unlike self-esteem, self-compassion allows people to acknowledge and accept their negative characteristics, and this acknowledgement often leads to proactive behaviors (Allen et al., 2012).

Self-compassionate cognitions—cognitions charac terized by self-kindness, common humanity, and mind fulness—may be particularly important for maintaining well-being in older age. The results showed that positive self-compassionate cognitions predicted the positivity of participants’ emotional tone when reflecting on an age-related event. Additionally, participants’ self-compassionate cognitions were related to the extent that participants’ attitudes helped them cope with the aging event. Self-compassion is related to a variety of positive outcomes (Neff, 2003a; Neff et al., 2007), and this research suggests that self-compassionate thoughts may be the mechanism that underlies this relationship.

This study provides an initial look at self-compassionate thinking in older adults, but two limitations may affect the study’s generalizability. First, by posting fliers at senior citizen centers and using the university’s aging registry, we may have recruited a sample of older adults that was unusually engaged and high functioning. Additionally, people who agreed to participate were probably more open to research studies and eager to learn about themselves than those who did not participate. Even so, our findings might have been even stronger had the sample included a higher proportion of participants with the kinds of problems that precluded participation (see Allen et al., 2012). Second, the sample size (n = 131) was modest, raising questions about the power of the analyses and the generalizability of the findings. However, our findings replicate research showing that self-compassion is related to emotional well-being, and our assertion that self-compassionate cognitions are key to this relationship has already been suggested by previous research.

Although people differ in their tendency to be self-compassionate, evidence suggests that self-compassion can be increased and that people may become more self-compassionate with practice. Leary and colleagues (2007, Study 5) found that having participants write about a negative personal event using self-compassion prompts led them to think about the event in a self-compassionate way, take greater responsibility for the event, experience less negative affect, and report feeling more similar to other people who have experienced similar things compared with participants in two different control groups. In a study that dealt with guilt-induced eating (Adams & Leary, 2007), a self-compassion induction led highly restrictive eaters who had overeaten both to feel less distressed and to eat less in a subsequent taste test than restrictive eaters who did not receive the self-compassion induction. Finally, evidence suggests that training people to be self-compassionate reduces depression and anxiety and increases well-being (Neff & Germer, 2013); Gilbert & Procter, 2006). (Neff & Germer, 2013) are currently validating an 8-week Mindful Self-compassion Program. Preliminary evidence shows that participants experienced an increase in self-compassion and life satisfaction and decrease in anxiety, depression, and stress when compared with a wait-list control group. However, research is needed to address the possible long-term benefits of self-compassion training.

At present, research supports the claim that self-compassion is malleable (more so than self-esteem); however, research needs to address the extent of self-compassion’s malleability and identify the best techniques for increasing self- compassion. Nonetheless, practitioners who assist older adults in adjusting to age-related changes should strive to identify individuals who are particularly low in self-compassion. By teaching the basic principles of self-compassion, practitioners can help older people redefine how they treat themselves when experiencing negative changes or events. Several self-compassion activities are available as part of Neff & Germer (2013) Mindful Self-compassion Program. These activities help people identify their self-critical voice and offer suggestions for how to develop a self-compassionate mindset. As with any intervention, we expect that long-term effects will take time and practice. However, practitioners can be instrumental in introducing patients to a new way of thinking and providing them with the resources to practice self-compassion.

Regarding established psychoeducational programs, Neff & Germer (2013) offer self-compassion training workshops for individuals who are interested in becoming more self-compassionate as well as for practitioners who want to learn about self-compassion. Once trained, practitioners could facilitate self-compassion support groups where older adults share their noncompassionate self-thoughts and learn practical techniques for being more self-compassionate in everyday situations. The present research suggests that helping older adults think more self-compassionately may play an important role in increasing their emotional and subjective well-being. As a first step, practitioners should focus on helping patients identify ways in which they do not respond with self-directed compassion and kindness to their problems. Acknowledging these thoughts and recognizing their maladaptive nature are essential to becoming more self-compassionate. In addition, practitioners should focus on teaching self-compassionate principles and implementing self-compassion exercises that help people process negative age-related events in a more compassionate manner. As is evidenced by the preliminary findings from the Mindful Self-compassion Program (Neff & Germer, 2013), teaching self-compassionate principles results in lower depression, anxiety, and stress. We expect that other yet-to-be-developed approaches to promoting self-compassion will also be effective at increasing well-being.

In sum, self-compassion training programs could help older adults cope more effectively with the changes associated with aging (see Rejeski, 2008, for a comparable recommendation). Self-compassionate cognitions are central to the relationship between self-compassion and well-being, but more research is needed to understand the specific mechanisms underlying self-compassion’s positive benefits. In any case, the fact that older people who are high in self-compassion fare better than those who are low in self-compassion suggests that the field of gerontology would benefit from devoting additional attention to self-compassion. In fact, if psychoeducational programs are effective in increasing self-compassion, they may someday be offered more widely to the general public to promote the benefits of treating oneself compassionately.

Funding

This work was supported by the National Institute on Aging (AG034716).