| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
RESEARCH ARTICLE |
Department of Sociology, University of Chicago, Illinois.
Address correspondence to Yanni Hao, University of Chicago, Department of Sociology, 1126 East 59th Street, Chicago, IL 60637. E-mail: yanni{at}uchicago.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Using four waves of Health and Retirement Study data collected from a sample of 7,830 individuals aged 55 to 66, I estimated growth curve models to assess the effects of productive activities on mental health trajectories. The analytical strategy took into account selection processes when examining the beneficial effects of activities. The analyses also formally attended to the sample attrition problem inherent in longitudinal studies.
Results. The results indicated that activity participants generally had better mental health at the beginning of the study. Full-time employment and low-level volunteering had independent protective effects against decline in psychological well-being. Joint participants of both productive activities enjoyed a slower rate of mental health decline than single-activity participants.
Discussion. The results are consistent with activity theory and further confirm the role accumulation perspective. The finding that full-time work combined with low-level volunteering is protective of mental health reveals the complementary effect of volunteering to formal employment. Methodological and theoretical implications are discussed.
Key Words: Productive activities • Psychological well-being
IN light of the increasing life expectancy and improving health conditions of older adults, the productive social engagement of the aging population has received increasing attention. Older adults are involved in productive activities in many forms. Some continue to engage in paid work, and many others participate in unpaid activities that contribute to the public good. Both paid job and unpaid volunteer work represent two major forms of productive activity in formal contexts (Bass, Caro, & Chen, 1993
; Luoh & Herzog, 2002). Studies have shown that 30% to 50% of Americans aged 55 and older engage in volunteer work (Zedlewski & Schaner, 2005) and that partially retired people with more education and better health are more likely to participate (Chambre, 1987; Luoh & Herzog, 2002; Warmurton, Le Brocque, & Rosenman, 1998). Considering paid employment, less than 70% of 60-year-old men engage in paid work; among women the corresponding proportion is lower (Luoh & Herzog, 2002). Although there have been decreases in the levels of labor force participation among older adults for decades, some scholars have pointed out that this trend may reverse itself (Burkhauser & Quinn, 1997; Woodbury, 1999). If so, many people would continue to work until old age.
The literature investigating productive activities has found positive support for the psychological well-being of older participants (Li & Ferraro, 2005, 2006; Luoh & Herzog, 2002; Morrow-Howell, Hinterlong, Rozario, & Tang, 2003; Thoits & Hewitt, 2001; Umberson, Chen, House, & Hopkins, 1996; Van Willigen, 2000). Volunteers report fewer symptoms of anxiety, a higher level of life satisfaction, and better personal control (Fengler, 1984; Greenfield & Marks, 2004; Hunter & Linn, 1981; Jirovec & Hyduk, 1998). Research about paid work among older adults has focused on the impact of retirement and involuntary job loss (Gallo, Bradley, Siegel, & Kasl, 2000; Kim & Moen, 2002). Only a few studies have directly addressed the effects of late-life work, and some of them have documented that working to an older age has a number of psychological benefits (Bosse, Aldwin, Levenson, Workman-Daniels, & Ekerdt, 1990; Calvo, 2006; Duncan & Whitney, 1990; Erikson, Erikson, & Kivnick, 1986). Social gerontologists widely employ activity theory (Herzog & House, 1991; Lemon, Bengtson, & Peterson, 1972) to address the association between activities and mental health among older participants. The theory suggests that activities in general, and interpersonal activities in particular, are beneficial for psychological well-being because they offer channels for acquiring role supports that sustain one's self-concept (Lemon et al., 1972). According to the theory, older persons who engage in productive work have a higher level of personal control and mastery (Hayward, Friedman, & Chen, 1998). The activity can be a fulfilling experience that bolsters meaning in later life (Su & Ferraro, 1997; Wethington, Moen, Glasgow, & Pillemer, 2000). The altruistic nature of volunteering offers a way of gaining social approval in addition to improving self-esteem (Siegrist, Knesebeck, & Pollack, 2004; Thoits & Hewitt, 2001). Frequent social interactions also increase the chances of finding social support and social contacts (Lin, Ye, & Ensel, 1999). Activity theory received renewed interest in the late 1990s in light of the model of successful aging that postulates social activity as a crucial component of healthy aging (Rowe & Kahn, 1998).
Studies on beneficial effects of volunteering have focused primarily on the amount of service performed. The results in general document a nonlinear effect of volunteer hours on mental and physical well-being (Luoh & Herzog, 2002; Morrow-Howell et al., 2003; Musick, Herzog, & House, 1999; Van Willigen, 2000). As for paid work, scholars have investigated employment status rather than the level of involvement. One study by Ross and Mirowsky (1995) did differentiate between full-time and part-time employment, with the authors finding beneficial effects only for full-time work status. Apart from the amount of service, another aspect of activity that merits attention is concurrent engagement in different activities. As indicated from prior research, people who are partially retired are more likely to volunteer (Warmurton et al., 1998). As a result, many older adults participate in both paid work and volunteer work simultaneously. According to Zedlewski and Schaner (2005), only 6% of older adults aged 55 and older report paid work as their only form of social engagement, and about 29% of people combine work with formal or informal volunteering. The literature shows that occupying multiple roles is associated with better mental health (Adelmann, 1994; Pietromonaco, Manis, & Frohardt-Lane, 1986; Thoits, 1986). Adelmann found that the number of social roles older people occupy is negatively related to their level of depression. Researchers have employed role accumulation theory (Moen, Dempster-McClain, & Williams, 1992; Sieber, 1974) to explain how occupying multiple roles has more positive mental health outcomes. According to this theory, a large number of role identities promotes human development partly because interactions with a wide range of others augment power, resources, and socioemotional support (Bronfenbrenner, 1979). Specifically, people derive four types of rewards from role accumulation: role privileges, overall status security, resources for status enhancement and role performance, and enrichment of the personality and ego gratification (Sieber, 1974). Multiple-role occupancy may be especially important for older adults, because a broadened role repertoire allows older people to maintain a certain level of social involvement at a time when various commitments are being reduced (Morgan, 1988). With regard to combining volunteering with other social roles, some researchers have called the enhanced benefits from multiple roles the "complementary" effect of volunteering (Oman, Thoresen, & McMahon, 1999; Van Willigen, 2000). For example, Van Willigen found that adults older than 60 who were employed reported higher life satisfaction when they volunteered than did those who did not volunteer. She also found that younger adults did not experience the same benefits from combining roles.
Although previous studies have provided rich evidence on the mental health benefits of activities, most of these studies emphasized the outcome measures at one point in time. This leaves open the question of the change trajectories of health conditions. It is therefore unclear whether the beneficial effects of productive activities extend to changes in health over time. In addition, rarely has previous research investigated paid work and formal volunteering in one study, and none of these investigations simultaneously considered the effect of the amount of time commitment together with that of the number of engagements.
Examination of the health benefits of later life activity is also complicated by selection processes. Studies have found that volunteers tend to be healthier and more socially integrated individuals (Li & Ferraro, 2005, 2006; Thoits & Hewitt, 2001). By the same token, older adults with prolonged labor force participation represent a rather select group. They are more likely to have higher occupational status and educational attainment (Hayward et al., 1998). Still, some researchers (Li & Ferraro, 2005, 2006) have pointed out that compensation may act as another self-selection process. For example, depression may promote voluntary participation in later life for overcoming negative effects (Li & Ferraro, 2005). As a result, failure to account for selection mechanisms may bias the effects of activity participation. To rule out such selection effects, conventional studies control for health conditions that precede activity participation. This strategy is suboptimal, though, given that in observational studies the possibility always remains that some uncontrolled prior health heterogeneity may exist. To my knowledge, only a few studies (Li & Ferraro, 2005, 2006; Thoits & Hewitt, 2001) have systematically investigated both beneficial and selection mechanisms. In the present study, I gave explicit attention to beneficial mechanisms while taking into account the selection processes.
Thus, in this article, I investigate whether the beneficial effects of productive activities extend to the change trajectories of mental health for older Americans, a group that is experiencing important life transitions such as retirement and health deterioration. Guided by activity theory and the role accumulation perspective as well as the current empirical research, I developed two general hypotheses. I hypothesized, first, that both paid work and volunteering would predict slower rates of mental health decline among older adults, and second, that concurrent engagement in both productive activities would have a stronger positive effect on mental health than single-activity participation. Moreover, I assessed how the level of time involvement committed to each activity impacts the change trajectory of psychological well-being.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
As is true for most panel studies, the HRS sample has experienced significant attrition over time. At baseline of the present study there were a total of 7,830 age-eligible respondents. By Wave 2, 148 people had died, and the number of deaths increased to 336 in total by Wave 3 and 585 by Wave 4. Some participants were unable to be reached for follow-up (487 at Wave 2, 393 at Wave 3, and 313 at Wave 4). Relatively few respondents have item-missing values (126 at wave 2, 109 at wave 3 and 115 at wave 4), and most are present in depression items. Consequently, the sample sizes for the three follow-up waves were 7,069, 6,505, and 5,937. There were a total of 27,341 observations in the time-person data set. I weighted all analyses to adjust for unequal sampling probabilities and response rates and estimated logistic regression models to predict the odds of sample attrition. Compared to the study sample at large, those who died after the first wave were more likely to be male, single, and older. These people also had less education and poorer physical health and tended to be unemployed. Among survivors, people who did not complete the study or had missing data on major measures tended to be Black, to be male, to have low education and poor health at baseline, and to participate less in volunteer work (p <.05). The attrition pattern was consistent with existing studies (Li & Ferraro, 2005; Thoits & Hewitt, 2001) that have reported that people who are underrepresented in panel studies are more likely to be of lower socioeconomic status and less socially integrated. This sample bias may have led to the underestimation of the true relationship between activities and health (Berk, 1983). In a recent work, Li and Ferraro (2005) attended to this problem by using multigroup analyses for complete and incomplete data (Allison, 1987; Li & Ferraro, 2005). In the present study, by including a time-varying indicator for attrition risk, I provided an alternative approach to accounting for the attrition problem. I address the operationalization of the indicator later.
Dependent Variables
In the HRS, depressive symptomatology consists of eight items from the Center for Epidemiologic Studies–Depression scale (CES-D). The eight items in the modified CES-D scale in the HRS were selected based on factor analysis results reported in Radloff (1977). Respondents were asked whether they had experienced eight specific symptoms in the past week, including (a) was depressed, (b) everything was an effort, (c) sleep was restless, (d) was (not) happy, (e) felt lonely, (f) (did not) enjoy life, (g) felt sad, and (h) could not get going. Dichotomous response categories were consistent from Wave 1 to Wave 4 and signified the presence (1) or absence (0) of a symptom. In the present analysis, I summed the responses for the eight items, with higher scores indicating more depressive symptoms.
Time-Varying Indicators
The present study controlled for major time-varying covariates. I created two binary variables indicating status for paid work (yes or no) and for volunteering (yes or no), respectively. The analysis also included time-varying marital status (yes or no). I measured general health status with a participant self-rated score ranging from 1 to 5, with a higher score indicating better health. As for the indicator of functional disability, eight items probed difficulties with activities of daily living. I summed the responses for these items and coded them into one indictor denoting number of limitations (ranging from 0–8). Respondents were also asked about eight chronic medical conditions, and I created an index for the number of chronic diseases.
I created a time-varying attrition risk indicator based on survival models. Inspired by Heckman's probit-based methodology to adjust for sample attrition bias (Berk, 1983; Heckman, 1979), I used Cox proportional hazards models to estimate a time-varying hazard ratio for attrition, which I employed in turn as a time-varying control in the main analyses. The model was specified as follows:
|
|
i(t)/0(t) is the hazard ratio of attrition for respondent i at time t; and Xpi represents predictors for attrition, including age, gender, race, education, income, marital status, employment status, volunteer status, self-assessed health status, and chronic disease index (a summed index of number of chronic diseases ranging from 0–8). The attrition risk indicator created from this method represented the changing likelihood of attrition across waves for each respondent. Specifically, it denoted the relative risk of attrition compared to that of a hypothetical individual—a White married man with average levels of age, education, income, and health status who was employed but did not volunteer. The hazard ratio captured the expected values of the disturbances due to nonrandom attrition, creating an adequate control and hence removing them as a source of biased estimates.
Time-Invariant Indicators
Baseline productive activities
The survey questions on paid work contained in the 1996 sample asked each respondent whether he or she was doing any work for pay at the time of the study. Respondents who answered affirmatively for paid employment were then asked the number of hours per week and the number of weeks per year usually worked on that job. Annual hours were obtained by multiplying the number of hours per average week by the number of weeks worked. According to the U.S. Department of Labor regarding full-time versus part-time employment, I created two binary variables denoting work level: I classified less than 1,680 hr annually as part time, and 1,680 hr or more as full time. The questions on volunteer work contained in the 1996 survey asked each respondent about hours spent in the past 12 months doing volunteer work for religious, educational, health-related, or other charitable organizations. Previous research has shown that the effect of volunteer work on health resides in the difference between performing 100 hr or more annually versus performing less or no volunteer work (Luoh & Herzog, 2002; Morrow-Howell et al., 2003; Musick et al., 1999; Van Willigen, 2000). Based on the information, I created two binary variables denoting the group volunteering less than 100 hr annually and the group volunteering a minimum of 100 hr.
Concurrent engagement in two activities
Using the 1996 baseline survey, I categorized respondents' activities into four mutually exclusive groups: no activity, paid work only, volunteering only, and both paid work and volunteering. According to this categorization, I generated three dichotomous indicators representing three forms of activity engagement, with the reference group being no activity.
Demographic characteristics
Several demographic characteristics were assessed in the analysis. Age was measured in years. Gender was coded as (1) for women and (0) for men. Race was represented by binary variables: Black (=1), Hispanic (=1), and White or other (the reference). Years of education ranged from 0 to 17, and income level was indicated by the total household income in 1996 (in $10,000). I used imputed values provided on HRS public use files for the cases with missing data on income.
Analytical Strategy
I used growth curve modeling (Raudenbush & Bryk, 2002) to estimate individual health trajectories. As one type of hierarchical modeling, growth curve models represent a dual-level analysis of change process. The first level models an individual's status on some trait as a function of an individual growth trajectory. At the second level, the parameters of the individual growth trajectories vary as a function of individual background characteristics (e.g., volunteer activities). The analyses estimated linear growth curve models.
Model specifications are presented in the following equations. Equation 2 represents the Level 1 within-subjects model. In this model, each respondent's health trajectory was a function of time, controlling for time-varying covariates:
|
|
0i is the expected value for mental health of person i at baseline and when all other covariates are held constant; 1i is the per-year change rate of mental health for person i.
The Level 1 model defined nine coefficients. The coefficients 0i and 1i became outcome variables at Level 2. The other seven coefficients in Level 1 were constrained to have fixed effects.
The Level 2 model incorporated a respondent's demographic controls and time-invariant indicators for paid work and volunteering activities:
|
|
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
100 hr) volunteering. As for concurrent participation in both paid and volunteer work, 21% of older adults fell into this category. Over the course of the study, respondents reported a steady decrease in involvement in paid work, dropping from 56% to 37%. Volunteer participation was relatively constant over time, ranging from 32% to 34%. The hazard ratios for attrition were generally small, with mean ratios and standard errors decreasing over waves. This indicated that attrition over time led the study sample to be more homogeneous in terms of attrition traits (less likely to drop out).
|
|
In a similar vein, Model 2 explored the effects of volunteer work. Estimated coefficients from the model suggested that volunteers at both levels of volunteering had lower levels of psychological distress at the beginning of the study in contrast to nonvolunteers. Moreover, low-level volunteering at baseline also predicted a slower rate of decline for psychological well-being.
In Model 3 and Model 4, I assessed the effects of concurrent engagement in two activities compared to single-activity participation. I assessed three binary variables representing volunteered only, worked only, and both volunteered and worked at baseline. The reference category was the group that did not engage in either of the two activities. The results in Model 3 showed that both single-activity and dual-activity participants enjoyed better psychological well-being at baseline. The magnitude of the coefficient for dual-activity participants was stronger than for single-activity participants, suggesting that people who engaged in both activities enjoyed better mental health than single-activity participants at baseline. The results for the change trajectory revealed interesting findings. Respondents who only worked or volunteered at baseline shared the same change rate with those who performed no activities. Only respondents who were concurrently involved in both activities had a slower decline rate in psychological well-being in subsequent waves. Because Models 1 and 2 revealed that full-time employment and low-level volunteering had an independent promoting effect on mental health over time, I further explored the effect of amount of service within the group of dual-activity participants. I divided the dual-activity group into four subgroups: part-time work and low-level volunteering, part-time work and high-level volunteering, full-time work and low-level volunteering, and full-time work and high-level volunteering. Results of Model 4 showed that, consistent with results from Models 1 and 2, it was the combined activity of full-time work and low-level volunteering that yielded benefits on the change rate of psychological well-being.
To summarize, the results indicated that productive activity participants generally began with better psychological well-being. Dual-activity participants also displayed better health than single-activity participants. This may have been the outcome of previous activities or selection processes, or most likely a combined result of the two mechanisms simultaneously. As for the rate of change of mental health, I found that both full-time work and low-level volunteering had protective effects against decline in psychological well-being. Although single-activity participants shared the same change rate with those who had no activity, dual-activity participants did derive a slower rate of mental health decline. These were net effects after controlling for time-invariant demographic indicators and major time-varying covariates.
In addition to the main analyses, I conducted ancillary analyses to test whether the effect of working and volunteering on the mental health change rates was the same for men and women and for married and unmarried individuals by adding interaction terms in the analysis. The results did not yield significant differences in terms of the change rate in mental health across gender or marital status categories.
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The findings reveal that concurrent participation in paid work and volunteering has protective effects against mental health decline for older adults. This supports the role accumulation perspective. However, the effect of multiple-role occupancy may not always be beneficial. Some prior studies have shown detrimental effects of multiple-role attachment (Moen et al., 1992; Pruchno, 1999). These studies have focused on activities such as caregiving and informal helping that are primarily obligatory responsibilities from family and personal networks. Older participants in these activities frequently experience role strain from multiple-role attachment (Cantor, 1983; Goode, 1960; Pruchno, 1999; Wilcox & King, 1999). I should note that because these activities are usually less discretionary, multiple engagements under these circumstances often operate as burdens for older adults versus outlets for personal growth (Cantor, 1983; Pruchno, 1999; Wilcox & King, 1999). Absence of strain from concurrent engagement in paid work and volunteering in this study might suggest that the effects from role attachment are, to some extent, dependent on the congruity between the nature of the role and individual preferences (Herzog, House, & Morgan, 1990; Van Willigen, 2000). The levels of satisfaction and autonomy in role engagement may be important in mediating the multiple-role occupancy and health outcomes (Coser, 1975).
The finding that only full-time work combined with low-level volunteering is protective of mental health reveals the complementary effect of volunteering to formal employment. The results are consistent with Van Willigen's (2000) study on adults aged 60 and older. Moreover, this study also reveals a more nuanced understanding of this complementary effect. The fact that other combinations of activities are not protective implies differential significances of paid work and volunteering for older adults. Formal employment, as a crucial channel for social participation, represents resources, power, and prestige. Apart from tangible economic benefits, participants also reap noneconomic rewards such as public recognition, opportunities for self-expression, and chances for ascendance in the workplace (Mannheim, 1975; Phelan, Bromet, Schwartz, Dew, & Curtis, 1993; Repetti, Matthews, & Waldron, 1989). These benefits, however, may accrue differently to those involved in full-time and part-time employment. Full-time employees reportedly receive better compensation, better health and retirement insurance coverage, and more social support compared to their part-time counterparts (Bosse et al., 1990; Gable & Hollon, 1982). They also have relatively high control over their work and high job satisfaction (Hall & Gordon, 1973; Miller & Terborg, 1979). This in turn may explain the advantageous position of older full-time employees with regard to mental health preservation compared to part-time workers. As for volunteer activity, the existing literature suggests that volunteering in old age is oftentimes a welcome complement to formal employment. Under many circumstances, it is an extension of formal employment for older adults. A modest amount of service is usually sufficient for health benefits, with higher levels of participation not necessarily leading to further gains (Luoh & Herzog, 2002; Morrow-Howell et al., 2003). As a consequence, participants benefit from the very fact of occupying the activity role, rather than from the extent of involvement. However, as the current study limited the sample to persons between ages 55 and 66, it cannot extrapolate this complementary effect to other age groups. Patterns and processes might differ for older adults because the importance of paid work is often attenuated and the significance of other roles is elevated in later life.
One recurring question in the social roles and health literature is whether one type of role identity can substitute for the absence of another. Some research studies have found that volunteering serves as a protective factor against the disadvantage to psychological well-being that accompanies role losses in other life domains (Fengler, 1984; Greenfield & Marks, 2004; Musick et al., 1999). A study conducted by Fengler documented a stronger relationship between being a volunteer and life satisfaction for those whom he called "disadvantaged" older adults, that is, those with personal and social resource deficits. Greenfield and Marks's study suggested that volunteering may protect adults with relatively fewer social roles from psychological distress. In contrast, the findings of the present study imply that those with role absence of employment derive no benefits from the substitute of volunteering. Conversely, older adults who are already "advantaged" with formal employment enjoy enhanced mental health from the extra role of volunteering. There may be two explanations for the inconsistency between the findings from this study and the prior literature. First, most previous studies supporting the substitution effect have been cross-sectional in research design. Although recent empirical research has employed longitudinal data, none of these studies examined change in health over time. Therefore, it may be possible that volunteering as a substitution does not accrue benefits that are strong enough to impact health change processes. Second, studies of which I am aware that support substitution effects predominantly focused on adults aged 65 and older. The literature suggests that effects of volunteering are elevated to greater significance in later life (Li & Ferraro, 2006; Van Willigen, 2000). Given that the present study examined people younger than 66, I suspect that the substitution effects of volunteering may be contingent on the period of life course and hence are most important for older individuals when major role losses are not revertible and volunteering is more valuable for maintaining equilibrium in role sets. To summarize, the negative findings on substitution effects in this study raise some intriguing questions for future analysis. A further investigation in these areas would benefit the understandings of the substitution role of formal volunteering.
Like previous research on the effects of activities, this study faced the task of disentangling the effects of social selection (Li & Ferraro, 2005; Thoits & Hewitt, 2001). In the current analyses, the dual-level model specification considered mental health heterogeneity in both initial status and later change trajectory in relation to activity participation. This made possible an effective delineation of the beneficial mechanisms of participation when selection processes were considered simultaneously. The results verify the beneficial mechanism of activities, as the difference in change rate is a net result of initial status. Moreover, this study shows that dual-activity participants generally begin with better psychological well-being, and they also enjoy a slower rate of mental health decline. Therefore, as long as older adults continue to engage in productive activities, the mental health discrepancy between dual-activity participants and single-activity participants or nonparticipants will become bigger over time. This may suggest evidence of a positive cycle of selection and beneficial processes. Based on this fact, what is required in future studies is a dynamic approach to social integration and well-being. The present study also formally attended to the sample attrition problem inherent in longitudinal studies. Because of the incorporation of a time-varying indicator for attrition hazard, respondents in this study were under the same risk of attrition during the process of the study. This may provide an alternative approach for attending to the attrition problem in longitudinal studies.
Nevertheless, there are limitations to this study. Plausible mechanisms that might help account for the positive effects of productive activities were left unconsidered. However, in order to understand the association between social activities and mental health, it is important to pinpoint precisely the underlying processes that are responsible for these salutary effects. Additional concerns can be raised regarding the types of paid work and volunteering people do. Researchers have suggested that the type and condition of work are influential on mental health (Calvo, 2006; Musick & Wilson, 2003). For example, older employees who work in low-stress jobs with the schedule they desire tend to experience better mental health (Calvo, 2006; Herzog, House, & Morgan, 1991). Similarly, volunteer work covers an enormous range of activities. Studies have shown that religious-based volunteering is more beneficial for older volunteers' mental health than is secular activity (Musick & Wilson, 2003). Due to space limitations, I did not cover these other dimensions of activities in this study. Thus, further efforts are required to address these issues.
Finally, the findings on the salubrious effect of productive activities are also linked to development of social programs to facilitate the activities of older individuals. Public health programs that bring these individuals into productive activities will improve the general well-being of older generations. Moreover, this study suggests that older Americans represent a tremendous resource to society. According to the U.S. Census Bureau (2004), 30% Americans will be 55 years or older by 2030. Policies and programs are needed to focus on ways to increase productive contributions for older adults. In conclusion, continuing to investigate the productive social participation of older people will provide researchers and practitioners with a more comprehensive picture of what the optimal activity pattern is and how this can be incorporated into successful aging for older adults.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received for publication April 19, 2007. Accepted for publication November 30, 2007.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Schwingel, M. M. Niti, C. Tang, and T. P. Ng Continued work employment and volunteerism and mental well-being of older adults: Singapore longitudinal ageing studies Age Ageing, May 27, 2009; (2009) afp089v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
