Longitudinal Implications of Social Integration for Age and Gender Differences in Late-Life Physical Functioning

Abstract

Social integration has documented benefits for late-life health; yet, little is known about its impacts on trajectories of physical functioning. This study examines age and gender differences in the longitudinal associations between social integration and activities of daily living (ADLs) using a hierarchical linear model with three waves of survey data collected over 4 years from the Social Integration and Aging Study (N = 400; baseline mean age = 80.3). Findings indicated some interaction effects of age, gender, and/or social integration on ADL trajectories. Among those of more advanced age, women showed greater increases in ADL limitations than men, and individuals with lower social integration experienced greater increases in ADL limitations than those with higher social integration. Neither of these patterns were found among younger older adults. This study highlights the benefits of longitudinal research on social integration and the need to explore practical interventions for promoting social integration particularly among the oldest older adults.

Keywords

social integration physical functioning activities of daily living gender older adults

The benefits of various components of social integration for physical health and functioning across the lifespan, particularly in late life, are well documented (e.g., Berkman et al., 2000; Stephens et al., 2014; Thoits, 2011; Umberson & Montez, 2010). With greater age, men and women face increasing physical limitations that may progressively impair their quality of life (Freedman et al., 2013; Holmes et al., 2009). Longitudinal studies have reported that higher levels of social participation, an aspect of social integration, are associated with greater amounts of physical activity, better mobility, and lower physical disability among older adults (Buchman et al., 2009; Corbett et al., 2018; James et al., 2011), while other studies found no such differences (Avlund et al., 2004; Green et al., 2008). Though inconclusive, such studies highlight the potential of social integration to protect against age-related physical decline. Further longitudinal research is warranted to determine whether there are indeed benefits of social integration on physical functioning in later life as well as examine factors that underpin these associations. Moreover, it is not yet clear if the role of social integration for late-life physical functioning varies over time among men and women of different ages (i.e., young-old versus old-old). In an attempt to address this, the current study examines whether age and gender differences exist in the association between social integration and trajectories of physical functioning among older adults.

Social Integration and Physical Functioning

Various aspects of social relationships, ranging from the structure or function of social network, the quality of emotional support, the extent of social engagement and participation, or the investment in social roles, each play a role in well-being across the lifespan (Antonucci, 2001; Berkman et al., 2000). The concept of social integration incorporates multiple components of social ties to provide a more comprehensive or overarching picture of an individual’s engagement in social relationships. Social integration is a complex construct and takes various forms encompassing social networks such as family, friends, and other relations as well as engagement in social activities with groups and organizations (Fuller-Iglesias & Rajbhandari, 2016). Aging research often examines specific components of social integration, such as social networks, support quality, or community engagement (e.g., Cohen, 2004). For instance, prior research has assessed social network composition and contact frequency (e.g., Green et al., 2008); ratings of the perceived quality or functionality of support with interpersonal partners such as family and friends (e.g., Sharifian & Grühn, 2019); or the level of participation in community events or organizations (e.g., Corbett et al., 2018). Each of these separately represent important dimensions of an older adult’s social well-being; however, by focusing only on one of these dimensions, we miss the overall, multidimensional social integration of an older adult. The present study examines social integration comprehensively, as an umbrella concept representing the various social ties and activities with which older adults engage.

Theoretical models have linked social integration to well-being across the lifespan (Antonucci et al., 2014; Berkman et al., 2000; Thoits, 2011). Theoretically, there are various mechanisms that may explain this link, such as increasing instrumental support, fostering health-promoting behaviors, promoting coping, buffering stress, and enhancing feelings of self-worth and control, among others (Berkman et al., 2000; Thoits, 2011; Uchino, 2004). In addition to attempting to explain why social integration may influence well-being, theories also highlight personal and contextual factors that can account for variation in these links. The convoy model (Antonucci, 2001) emphasizes the important role that personal characteristics, such as gender or age, play in determining the influence of social connections, particularly in late adulthood (Fuller et al., 2020). Moreover, the convoy model highlights the importance of recognizing change over time across and within individuals (Fuller et al., 2020). The current study is guided by these theoretical frameworks in seeking to understand whether social integration predicts trajectories of physical functioning over time among older adults, and if such associations vary by gender and age.

The ways that social integration is linked to well-being likely vary depending on physical, cognitive, emotional, or functional well-being (Thoits, 2011). For example, research has shown increased social integration in later life to be associated with reduced risk of chronic disease (Crittenden et al., 2014), lower mortality (Barger, 2013; Holt-Lunstad et al., 2010), better psychological well-being (Sharifian & Grühn, 2019), better self-rated health and cardiovascular outcomes (Ertel et al., 2009), and decreased disability (Nilsson et al., 2010). While the impact on physical functioning may be less pronounced, there is evidence of implications of social integration for physical functioning (Buchman et al., 2009; James et al., 2011; Rosso et al., 2013). Moreover, social ties and social participation are associated with higher amounts of physical activity and better mobility, which can enhance physical functioning. While theoretical works and cross-sectional studies suggest that social integration may act as a potential protective factor against age-related functional decline, longitudinal studies may hold the key to determining the predictive nature of social integration.

Some longitudinal studies support the notion that social integration may be protective for older adults’ physical functioning and disability (Buchman et al., 2009; James et al., 2011; Kanamori et al., 2014; Unger et al., 1999), with some even finding that social integration has the potential to improve physical functioning among older adults (Corbett et al., 2018; Zunzunegui et al., 2005). However, in contrast, other longitudinal studies have determined that social integration did not predict older adults’ functional health or disability risk over time (Avlund et al., 2004; Green et al., 2008; Seeman & Chen, 2002). These contradictory findings warrant further investigation. Guided by the convoy model, we propose that variability in personal characteristics, such as age and gender, may help explain these contradictory findings, in that it may be the case that social integration is more predictive of trajectories of physical functioning among the oldest old as opposed to the young-old, or among women as opposed to men.

Age, Social Integration, and Physical Functioning

Research over the years has shown a marked decline in physical functioning with increasing age, as the prevalence of having one or more functional limitation or disability increases progressively with advancing age (Crimmins et al., 2016; Holmes et al., 2009). Indicative of these declines in physical functioning, older individuals are increasingly likely to experience challenges in their ability to carry out activities of daily living (ADLs), and the prevalence of overall ADL limitations becomes progressively greater with increasing age (Freedman et al., 2013; Martin & Schoeni, 2014; Seeman et al., 2010). Limitations in ADLs include difficulty with basic everyday tasks that are necessary for self-care (such as dressing, bathing, or eating) as well as more complex, instrumental household management tasks (such as financial management, cleaning, or shopping). Freedman et al. (2013) reported relatively low rates of ADL limitations for middle-aged men and women (aged 55–65) and the highest rates among the oldest old (those aged 85 and older). Moreover, research on trajectories of physical functioning indicates that the rate of physical decline increases with age (Liang et al., 2008). This age disparity in physical functioning is in large part accounted for by increased disease comorbidity with age (Chatterji et al., 2015; Freedman et al., 2007; Stenholm et al., 2015), yet research on exceptional longevity suggests that age does not strictly predict increased disability and functional decline (Christensen et al., 2008; Terry et al., 2008). For instance, if older adults maintain high social integration (i.e., having more social support, higher engagement in social activities, etc.), it may help reduce risks for decline in physical functioning (Buchman et al., 2009; James et al., 2011; Kanamori et al., 2014; Unger et al., 1999) despite their older age.

While social integration appears to be beneficial for various health behaviors and outcomes including physical functioning throughout adulthood (Thoits, 2011; Umberson & Montez, 2010), such health benefits of social integration may change with age. For example, though Elliot et al. (2018) found that social integration was not associated with lower interleukin-6 (i.e., inflammatory biomarker related to morbidity) overall for their sample aged 35–86, their additional analyses revealed that the association was significant for those aged 75 or older, indicating a shift with advanced age. Loneliness may be indicative of a lack of social integration, and research suggests that associations between loneliness and health may involve multiple pathways (e.g., health behaviors, neurobiological mechanisms, cognitive processes); moreover, the health impacts through such pathways may become more pronounced in later life though additional research on these age differences is needed (Ong et al., 2016). Specifically, for social integration and physical functioning, research on age differences also remains limited. Some longitudinal studies found no age differences among older adults in associations of social ties (e.g., Zunzunegui et al., 2005) or social activities and participation (e.g., James et al., 2011; Otsuka et al., 2018) with reduced risks of functional disability. However, the findings of Buchman et al. (2009) were more nuanced, indicating that social activity was associated with a lower rate of decline in global motor function. In other words, if older adults had lower levels of social activity, their degree of functional decline increased more rapidly over time (i.e., with age), which seems to suggest that maintaining social integration may become more important with age for reducing risks of functional decline. Thus, the implications of social integration for trajectories of physical functioning may change with age. Due to limited prior research, additional longitudinal studies are needed to investigate how age is related to the association of social integration with trajectories of physical functioning.

Gender, Social Integration, and Physical Functioning

Men and women experience different levels and types of physical limitations as they age (Holmes et al., 2009). Women are found to experience greater disability and increased physical functioning limitations than their male counterparts (Crimmins et al., 2011). For instance, more older women than men report difficulties with ADLs such as bathing, dressing, housework, and shopping (Hardy et al., 2008; Martin & Schoeni, 2014). Gender differences in physical limitations are amplified by physical and mental health, which likewise vary by gender, as women are likely to have more chronic physical (i.e., musculoskeletal, autoimmune, etc.) and mental (i.e., depression, anxiety, etc.) health conditions than their male counterparts (Bird & Rieker, 2008). Moreover, women tend to live longer than men (Crimmins et al., 2016; Leveille et al., 2000), and thus, gender differences in physical limitations over an extended period can be attributed to higher life expectancy in women as well as higher initial levels of disability among older women compared to men (Li et al., 2018). Generally speaking, older women therefore have the potential to be physically impaired or disabled for a longer amount of time than older men.

The trajectories of physical decline over time appear to be different for aging men and women (Gill et al., 2013; Liang et al., 2008; Romoren & Blekeseaune, 2003). Different patterns have emerged across studies, but general trends suggest that older women experience more prolonged trajectories of functional limitations, whereas men are more likely to experience death prior to a prolonged trajectory of functional decline (Deeg, 2005; Gill et al., 2013; Romoren & Blekeseaune, 2003). Multiple studies have found that men are more likely than women to have stable trajectories of minimal functional impairment or disability over time, sometimes, but not always, with a sharp decline before death (Rooth et al., 2016; Verbrugge et al., 2017). Women on the other hand appear to have greater variability in their trajectories of functional limitations (e.g., Lin, 2020); evidence suggests that older women are more likely than older men to have both gradually declining and improving physical functioning trajectories (Deeg, 2005; Liang et al., 2008; Rooth et al., 2016) as well as stable trajectories of moderate to severe disability over time (Gill et al., 2013). When adjusting for length-of-survival, one recent study found that older women’s self-reported disability levels declined at a faster rate than men (Botoseneanu et al., 2016). Taken together, these studies highlight gender differences in trajectories of physical functioning in later life. In addition, related to gender as well as age differences, Liang et al. (2008) identified greater gender gaps in physical functioning (i.e., greater functional impairment for women than men) for those at a more advanced age (e.g., age 75+) than younger individuals, and the gender gaps increased over a decade with different trajectories between genders.

Late-life social integration varies by gender, with women more socially integrated than men on average (James et al., 2011; Thomas, 2011). Despite evidence for women’s greater social engagement and increased social ties, conclusions on gender differences in how social integration is related to older adults’ physical functioning are not straightforward. Cross-sectional research on just men or women suggest both genders have fewer physical limitations when they are more socially integrated (e.g., Davidson et al., 2003; Li et al., 2018). However, longitudinal studies focused on comparing gender differences paint a more complex picture. Initially, studies suggested that the protective effects of social ties (i.e., social network size, function, and quality) for physical functioning were stronger for older men than for older women (e.g., Seeman, 1996; Unger et al., 1999). Later, contradictory findings emerged, suggesting that social participation and contact were more strongly associated with reduction in disability risk for women as compared to men (Avlund et al., 2004), whereas others found no gender differences at all (Buchman et al., 2009; Zunzunegui et al., 2005). More recently, James et al. (2011) found that the association between social activity and risk of ADL disability was stronger for men than for women. Taken together, these studies present contradictory findings, yet they also examine different cohorts and varying aspects of both social integration and physical functioning. For example, Avlund et al. (2004) operationally defined social participation in terms of visiting others and attending a group activity and physical functioning in terms of first onset of disability (i.e., needing help with one of six mobility items). In contrast, James et al. (2011) operationally defined social activity with a measure of six items including attending events, meetings, religious practices, visits with others, volunteer work, and trips and physical functioning with an ADL scale. Thus, despite their contradicting conclusions, because these studies differ in their operational definitions of social integration, it is not possible to draw clear conclusions based on the existing literature as to whether the protective nature of social integration for physical functioning varies between older men and women, especially with respect to trajectories of physical functioning over time.

Objectives

The present study aims to better understand how age and gender are related to longitudinal trajectories of physical functioning and their associations with social integration. The first objective of this study is to examine the relationship between longitudinal patterns of ADLs and gender among older individuals of varying ages. The first research question asks: Do trajectories of physical functioning (i.e., overall ADLs) differ among older men and women of varying ages (research question 1)? Based on the aforementioned background, we hypothesized that trajectories of ADLs would differ by age and gender in that those of more advanced age and women would experience greater increases in ADL limitations than their younger and male counterparts, respectively. Particularly, the oldest women were expected to experience the greatest ADL limitations and steepest functional decline, aligned with previous findings (Liang et al., 2008).

Our second research question concerns the longitudinal association between social integration and physical functioning. Though prior literature was inconclusive on the association (Avlund et al., 2004; Buchman et al., 2009; Green et al., 2008; James et al., 2011), the contradicting evidence may be a result of assessing only specific aspects of social integration (such as social support or social participation). Given that we employ a comprehensive measure of social integration, we anticipated that a broad assessment of social integration would best capture any potential protective effects; thus, consistent with the prior longitudinal literature reporting positive implications of social integration (Buchman et al., 2009; James et al., 2011; Kanamori et al., 2014), we anticipated that social integration is associated with physical functioning overall among older adults. In the present study, we additionally addressed age and gender differences in the association between social integration and physical functioning, which is not yet well understood. This study seeks to fill the gap in the literature by specifically addressing the following question: Do the associations of social integration with trajectories of ADLs over time vary among those of different ages (research question 2A) and between older men and women (research question 2B)? Considering the aforementioned (limited) research findings on age differences, we hypothesized that the association of social integration with ADLs would increase with age and over time among older adults. In contrast, we addressed gender differences in an exploratory manner without making any directional hypothesis due to the contradictory findings of the existing literature discussed earlier, while anticipating some gender differences in the association of social integration with trajectories of ADLs.

Methods

Sample and Data

For the present study, three waves of data from the Social Integration and Aging Study (Fuller-Iglesias & Rajbhandari, 2016; Toyama & Fuller, 2020) were used. The study was a community-based study of older adults in a small metropolitan area in the Midwest of the United States, which consisted of written surveys aimed at investigating various dimensions of social integration and their associations with health and well-being among older adults. A convenience sampling method was used to recruit participants from senior-focused organizations such as senior centers, community programs, and retirement and assisted living communities. At Wave 1 (2013), 413 individuals aged 60 or older were recruited via mail (68%) and in person (32%). The response rate was 34% for mail recruitment, and an estimated participation was 70% for in-person recruitment. After 2 years, at Wave 2 (2015), a follow-up survey was mailed to all original respondents, and 307 out of them returned the Wave 2 surveys. The Wave 2 response rate was 81% (when excluding 32 participants who died between waves). Participants were again invited to complete a follow-up survey by mail 2 years later, at Wave 3 (2017). Among the Wave 2 participants, 245 completed the Wave 3 surveys. The Wave 3 response rate was 90% (when excluding 33 who passed away between waves). Those who completed all three waves (i.e., Wave 3 participants) differed in: age, F(2, 410) = 19.01, p < .001; level of education, F(2, 408) = 7.14, p < .001; functional limitations, F(2, 397) = 29.47, p < .001; and social integration, F(2, 386) = 14.73, p < .001. Specifically, Wave 3 participants were younger, had fewer functional limitations, and higher social integration than those who participated in only one or two surveys; Wave 3 participants also had higher education than those who participated in only the first survey (but did not differ in education from those who participated in the previous two surveys).¹ Further, sex did not differ among the Wave 3 participants and those who participated in only the previous one or two surveys (χ² = 4.52, df = 2, p > .10).

The present study analyzed data from 400 participants who responded to demographic, social integration, and functional limitations measures at one or more waves. Their overall mean age was 80.3 years, and a majority (72%) were female. Descriptive statistics are summarized in Table 1.

Table 1

Descriptive Statistics for Respondents Included in the Present Study.

Number of respondents (number of women)	Wave 1400 (289)	Wave 2303 (223)	Wave 3244 (183)
Variable/Measure	Mean (SD) [Range of scores]
Variable/Measure	Wave 1	Wave 2	Wave 3
Age (in years)
Women	80.2 (8.2) [61–98]	81.3 (7.9) [62–100]	82.2 (7.4) [64–102]
Men	80.4 (8.5) [60–94]	81.9 (8.6) [61–96]	82.5 (8.8) [63–96]
Overall	80.3 (8.3) [60–98]	81.5 (8.1) [61–100]	82.3 (7.8) [63–102]
Education (in years)
Women	13.0 (2.0) [6–17]	13.1 (2.1) [6–17]	13.3 (2.1) [6–17]
Men	12.8 (3.4) [0–17]	13.4 (2.6) [6–17]	13.3 (2.7) [6–17]
Overall	13.0 (2.5) [0–17]	13.2 (2.2) [6–17]	13.3 (2.2) [6–17]
Social integration (SILLS)
Women	14.8 (2.0) [7.2–19.5]	15.2 (2.1) [7.2–20.0]	15.1 (2.0) [8.5–19.4]
Men	14.0 (2.3) [5.4–18.4]	14.4 (2.5) [9.0–20.0]	14.7 (2.3) [8.2–19.0]
Overall	14.6 (2.1) [5.4–19.5]	15.0 (2.2) [7.2–20.0]	15.0 (2.1) [8.2–19.4]
Functional (ADL) limitations
Women	2.1 (3.0) [0–16]	2.2 (3.4) [0–18]	2.6 (3.6) [0–18]
Men	2.1 (2.9) [0–12]	2.2 (3.5) [0–17]	1.9 (2.9) [0–12]
Overall	2.1 (3.0) [0–16]	2.2 (3.4) [0–18]	2.4 (3.4) [0–18]

Note. “Overall” refers to the entire sample including all female and male respondents. Higher scores of social integration (SILLS) and functional (ADL) limitations indicate being more socially integrated and having greater limitations in ADLs, respectively. For education, Supplemental Table S2 shows the frequency of each educational level. ADLs = activities of daily living; SILLS = Social Integration in Later Life Scale.

Measures

Time Variable

A variable that would indicate the wave of the survey was created and centered at Wave 1 (Wave 1 = 0; Wave 2 = 1, Wave 3 = 2). One unit of this variable represents 2 years, which was the period of time between two consecutive waves.

Demographic Variables

In the surveys, respondents reported their age, sex, and educational level. For the present study, baseline age at Wave 1 (in years; centered at age 80) and sex (female = 0 and male = 1) were used. Educational level was coded into the number of years and then treated as a continuous variable (e.g., high school graduate = 12, college graduate = 16, graduate degree or professional degree = 17), which was centered at 12 years.

Functional Limitations

All three waves of surveys included the assessment of ADLs utilizing the Older Americans Resources and Service scale (Fillenbaum, 2013). Respondents were asked to indicate how much assistance they need for nine activities including instrumental ADLs (e.g., shopping, handling money) and basic ADLs (e.g., eating, dressing). For each item, they were given response options including “without help” (0), “with some help” (1), and “someone must do this for me” (2). The overall scale for ADL limitations was created by summing the scores for the nine items. If participants did not complete all nine items but responded to five or more items (i.e., more than half of the nine items), the average score of their completed items was imputed for each of their missing items, which was added to their overall score. These imputed overall scores were 6.9% of all valid cases. Five cases with four or fewer items (i.e., less than half of the nine items) completed were dropped from the study, which represented 0.5% of all cases with at least one of the nine items completed.

Social Integration

The three waves of surveys included items for the comprehensive Social Integration in Later Life Scale (SILLS; Fuller-Iglesias & Rajbhandari, 2016). The SILLS was designed to assess four dimensions of social integration. The subscales of the four dimensions had (a) five items for frequency of interactions with social ties including family, friends, and neighbors; (b) five items for frequency of social activities (i.e., attending meetings of a group, attending a religious service, attending a community event, volunteering, going on an outing); (c) four items for satisfaction with social ties; and (d) four items for satisfaction with social activities (i.e., recreation/leisure activities, social gatherings, religious or spiritual activities, connection to their community). Responses were given on a five-point Likert-type scale ranging from never (1) to very frequently (5) for the frequency items and from very dissatisfied (1) to very satisfied (5) for the satisfaction items. The average scores of each of the four subscales were summed to create an overall score. The scale αs were .73 at Wave 1, .77 at Wave 2, and .78 at Wave 3. This variable was centered at 14.5 based on the overall mean at Wave 1.

Analysis Strategy

In order to answer the research questions, a two-level hierarchical linear modeling analysis was conducted by using IBM SPSS version 27 (IBM, 2020) with maximum likelihood, which allowed all available data at any of the three waves to be included to produce less biased estimates than listwise deletion (i.e., using only data of those who completed all items of interest at all waves) and other traditional approaches to dealing with missing data (Heck et al., 2014). To examine levels and trajectories of functional limitations, the model included level-1 time-variant variables (i.e., time/wave, social integration) and level-2 time-invariant or individual-level variables (i.e., baseline age, sex, education) as predictor or covariates as well as their interaction terms.

In the model, the fixed effect of the individual-level intercept, which was allowed to vary among individuals with its random effect, would indicate the average level of functional limitations at Wave 1 with the reference conditions (i.e., baseline age 80, female, 12 years of education, SILLS = 14.5). The time slope (i.e., the effect of wave) would indicate a change rate over time in functional limitations while it was also treated as random and allowed to vary among individuals. As results from preliminary analysis with linear and quadratic components of time slopes showed that the quadratic slope was not significant, only the linear time slope was included. In addition to the main effects of the predictors and covariates, three three-way interaction terms were entered into the model consisting of: (a) wave, age, and sex (to address research question 1); (b) wave, age, and social integration (to address research question 2A); and (c) wave, sex, and social integration (to address research question 2B).

In the final model, nonsignificant interactions were removed if they were not a part of significant three-way interactions and if removing them improved the criterion in comparing models (i.e., Akaike’s information criterion). As seen in Table 2, the hierarchical linear model with all three three-way interactions (“full interaction model”) indicated that the interaction of wave, sex, and social integration was not significant. After removing this three-way interaction, the lower two-way interaction of sex and social interaction remained nonsignificant, which was thus also removed from the final model (see the final model in Table 2). Although the main effect of education was not significant, removing it would result in having a larger (i.e., worse) value of Akaike’s information criterion, so, education was kept in the final model.

Table 2

Estimates of Effects of Two-Level Hierarchical Linear Models for Functional Limitations: Full Interaction Model (With All Three Three-Way Interactions) and Final Model.

Parameter/Predictor	Full interaction model	Final model
Parameter/Predictor	Unstandardized coefficient(standard error)	Unstandardized coefficient(standard error)
Fixed effect:
Intercept	2.126 (0.169)***	2.145 (0.169)***
Wave	0.738 (0.086)***	0.748 (0.086)***
Age at Wave 0 (centered at 80)	0.174 (0.020)***	0.173 (0.020)***
Sex (female = 0, male = 1)	−0.201 (0.306)	−0.168 (0.306)
Education (centered at 12 years)	0.029 (0.056)	0.020 (0.055)
Social integration (SI; centered at 14.5)	−0.167 (0.061)**	−0.202 (0.052)***
Wave × age	0.057 (0.011)***	0.056 (0.011)***
Wave × sex	−0.183 (0.164)	−0.201 (0.164)
Wave × SI	−0.098 (0.043)*	−0.115 (0.036)**
Age × sex	−0.087 (0.037)*	−0.089 (0.037)*
Age × SI	−0.004 (0.006)	−0.004 (0.006)
Sex × SI	−0.118 (0.111)	(removed)
Wave × age × sex	−0.048 (0.020)*	−0.047 (0.020)*
Wave × age × SI	−0.014 (0.005)**	−0.014 (0.005)**
Wave × sex × SI	−0.054 (0.075)	(removed)
Random effect:
Intercept	6.646 (0.634)***	6.735 (0.638)***
Wave	0.876 (0.279)**	0.937 (0.282)***
Akaike’s information criterion	4173.8	4172.9

Note. ***p < .001, **p < .01, *p < .05; level-1 covariance structure: diagonal.

Using estimates for main effects and interactions in the final model, trajectories of functional limitations were depicted for those with different conditions (i.e., age, sex, social integration) for further investigations of the interactions. Post-hoc analyses were conducted using alternative hierarchical linear models that contained re-centered variables of interest (as well as all other variables included in the final model) so that the effects of covariates or predictors could be examined with the specific value or condition used in re-centering the variables (e.g., Wave 3, age 75 or 85; Hoffman, 2015).

Results

Results for the final model are shown in Table 2. The main effect of wave was significant and positive (0.748, p < .001), which indicates that the slope or trajectory of functional limitations increased over time. In addition, the main effect of age was significant (0.173, p < .001), which shows that older age predicted greater levels of functional limitations. The effect of social integration was also significant (−0.202, p < .001), indicating that reporting higher social integration predicted fewer functional limitations. There were no significant main effects of sex or education, which suggests that neither sex nor education predicted levels of functional limitations. The significant random effects of intercept (6.735, p < .001) and wave (0.937, p < .001) indicated that the initial levels and slopes/trajectories over time of functional limitations varied among individuals. In addition, the two significant three-way interactions of (a) wave with age and sex and (b) wave with age and social integration indicated that there were systematically varying effects on the trajectories depending on combinations of age, sex, and levels of social integration as follows.

Interaction of Wave, Age, and Sex (for Research Question 1)

The effect of the three-way interaction of wave, age, and sex is plotted in Figure 1, depicting the predicted trajectories for hypothetical men and women of two specific ages, 75 and 85 (at Wave 1), with the reference conditions (i.e., the centered values) of education and social integration based on the estimates of effects in the final model (Hoffman, 2015).

Figure 1

Trajectories of ADL limitations for women and men of different ages with the reference conditions of the other covariate and predictor (i.e., 12 years of education, SILLS = 14.5). Values next to each line indicate ADL scores at Wave 1 or Wave 3, which are estimated using coefficients in the final model in Table 2. Symbols of significance levels for pairs of ADL scores of women and men of each age indicate that the ADL scores of each pair at Wave 1 or Wave 3 differed significantly (***p < .001) or marginally (†p < .10) or did not differ (n.s. p > .10), as these significant levels are for the effect of sex in the post-hoc hierarchical linear models using the age variable recentered at 75 or 85 as well as the wave variable originally centered at Wave 1 or recentered at Wave 3. In addition, ADL scores at Wave 1 and those at Wave 3 significantly differed between women aged 75 and 85 and between men aged 75 and 85, as the effect of age is significant in all these post-hoc models (p < .001) as well as those using the sex variable recentered at male = 0 (p < .05). (More detailed information of these assessments for different ages and sexes will be provided upon request.). ADLs = activities of daily living; SILLS = Social Integration in Later Life Scale.

As shown in Figure 1, while trajectories were similar between the younger men and women (aged 75 at Wave 1), the trajectories appeared to differ between older men and women (aged 85 at Wave 1). This gender difference for those aged 85 was found in post-hoc analyses with two alternative hierarchical linear models using the age variable centered at age 85 while either using the original wave variable centered at Wave 1 or re-centering it at Wave 3. These results indicated that for those aged 85 at Wave 1, the main effect of sex was marginal (−0.614, p = .079) at Wave 1 but became greater and significant at Wave 3 (−1.535, p < .001) with the time variable centered at Wave 3 (treated as fixed as the model with its random effect did not converge; detailed results available upon request). In addition, the two-way interaction of wave and sex was significant for those aged 85 at Wave 1 (−.434, p < .05). These results indicate that for the older (initially aged 85) individuals, the gender difference of men having lower functional limitations than women became more pronounced over time, as women’s functional limitations increased more rapidly than men. This is contrasted with no significant effect of sex or two-way interaction of wave and sex for those originally aged 75 over the same period of time.

Interaction of Wave, Age, and Social Integration (for Research Question 2A)

The effect of the significant three-way interaction of wave, age, and social integration is plotted in Figure 2. This figure depicts the predicted trajectories for those aged 75 and 85 (at Wave 1) having two specific levels of social integration (i.e., approximately one standard deviation below and above the centered value of SILLS) based on the estimates of effects in the final model.

Figure 2

Trajectories of ADL limitations for those of different ages having different levels of social integration (SI): low SI (SILLS = 12.5) and high SI (SILLS = 16.5); with 12 years of education (reference condition); using the average scores between sexes. Values next to each line indicate ADL scores at Wave 1 or Wave 3, which are estimated using coefficients in the final model in Table 2. Symbols of significance levels for pairs of ADL scores of those of each age with low and high SI indicate that the ADL scores of each pair at Wave 1 or Wave 3 differed significantly (***p < .001; **p < .01), as these significant levels are for the effect of SI in the post-hoc hierarchical linear models using the age variable recentered at 75 or 85 as well as the wave variable originally centered at Wave 1 or recentered at Wave 3. In addition, ADL scores at Wave 1 and those at Wave 3 significantly differed between those aged 75 and 85 with low SI and between those aged 75 and 85 with high SI, as the effect of age is significant (p < .001) in post-hoc hierarchical linear models using the SI variable recentered at 12.5 (low) and 16.5 (high) as well as the original and recentered wave variables. (More detailed information of these assessments for different ages and levels of SI will be provided upon request.). ADLs = activities of daily living; SILLS = Social Integration in Later Life Scale.

As seen in Figure 2, while functional limitations appeared greater for both those aged 75 and 85 with low social integration as compared to their counterparts with high social integration at Wave 1, the effect of social integration (i.e., difference between those with low and high social integration) appeared to increase more substantially over time for older individuals (aged 85 at Wave 1). This difference was indicated in the results from two post-hoc hierarchical linear models that contained the age variable re-centered at either 75 or 85 as well as the other variables included in the final model, which indicated that the two-way interaction of wave and social integration was significant only for those aged 85 (–0.184, p < .001; detailed results available upon request), not for those aged 75. Thus, these results indicate that social integration moderated the association between age and trajectory of functional limitations; social integration appeared to have stronger effects over time for older individuals.

Interaction of Wave, Sex, and Social Integration (for Research Question 2B)

The three-way interaction of wave, sex, and social interaction was not significant, which indicated that social integration did not moderate the association between sex and trajectory of functional limitations. In other words, the effect of social integration on trajectories of functional limitations did not differ between sexes.

Discussion

Using a longitudinal design with a relatively old sample, the present study examined potential age and gender differences in the trajectories of ADL limitations and the association between social integration and those ADL trajectories among older adults. The present study makes a unique contribution to furthering the literature on the potential health benefits of social integration by highlighting the nuance of associations between social integration and trajectories of physical functioning for older adults as discussed below.

While prior research has denoted changes in physical functioning among older adults over time, the current findings additionally identified gender and age differences in the trajectories of physical limitations among a late-life sample. As hypothesized, current findings indicated age differences in physical limitations. For example, when comparing those aged 75 (i.e., young-old) and aged 85 (i.e., old-old) at baseline, as shown in Figure 1, there was a clear distinction with old-old individuals reporting more ADL limitations as well as greater increases in their average level of ADL limitations over 4 years as compared to young-old individuals. In contrast to prior research (e.g., Crimmins et al., 2011) and our expectations, physical limitations did not vary by gender on average (at age 80); however, there were gender differences once age was taken into account. Specifically, while the trajectories of ADL limitations appeared similar for men and women at younger ages (i.e., young-old), the ADL limitations of old-old women increased more rapidly over 4 years than their male counterparts. In contrast to previous findings documenting gender differences in levels of physical functioning (Holmes et al., 2009), which may be greater for older cohorts than younger individuals (Liang et al., 2008), the present findings are unique in showing increasing gender gaps in the trajectories of functional limitations over time for old-old individuals. Such gender gaps in increasing functional limitations may be related to older women’s greater risk of physical disadvantages such as losses of muscle and bone (i.e., sarcopenia and osteoporosis) that tend to increase with age, as compared to older men (Tarantino et al., 2013), but the mechanisms of gender gaps at the oldest ages warrants further investigation.

A second contribution of this study was examining the varying implications of social integration for physical functioning among older men and women. The present findings show that social integration was associated with levels and trajectories of late-life physical functioning. Our hypotheses were partially supported in that age differences in the association between social integration and ADLs increased over time; however, anticipated gender differences were not found. Specifically, as seen in Figure 2, for old-old individuals, unlike their young-old counterparts, disparities in physical functioning appeared to increase over time between those who were more and less socially integrated. Essentially, among old-old adults, those reporting better social integration had trajectories of physical functioning (i.e., rates of functional decline) more similar to their young-old counterparts; whereas those old-old individuals reporting poorer social integration had steeper increases in their trajectories of ADL limitations indicating more rapid declines in physical functioning over time. These findings suggest that social integration may become even more important with age for maintaining late-life functional health. Social ties and engagement are associated with better mental and physical/physiological health and health behaviors (e.g., exercising, not smoking; Berkman et al., 2000; Corbett et al., 2018; Thoits, 2011), which are all factors that could help aging adults reduce or delay functional decline. In other words, social integration may play a key role for promoting successful aging (Rowe & Kahn, 1997) by reducing risks of disability and helping maintain high functioning as well as promoting engagement with life. The unique conclusion of these findings is that the positive implications of social integration may increase particularly for the old-old who have greater risks of functional decline.

In contrast, no gender differences in the association of social integration with ADLs were found, suggesting that social integration may be similarly beneficial for the functional health of older men and women. Despite the apparent similarity, it is still possible that different issues may need to be considered for men and women. Given that the present findings suggest that women at more advanced ages face greater risk of functional decline as compared to men and that the positive implications of social integration for physical functioning may increase with age regardless of gender, staying socially active may be especially important for old-old women to maintain their functional health. On the other hand, despite not experiencing the steeper physical functioning decline of older women, older men may still be particularly vulnerable in terms of morbidity and mortality due to their tendency to have smaller social networks and be less socially engaged than older women (Antonucci et al., 2014; McLaughlin et al., 2010; Thomas, 2011). It is thus likely increasingly important for men to maintain high social integration during their later lives. In sum, social integration seems to be a key element for promoting functional health as well as other aspects of well-being for both older men and women while different approaches to helping them maintain social integration may be required considering their potentially distinct needs.

Future Research

The current study has some strengths as well as limitations that highlight directions for future research. A key strength of the present study is the use of multiple time points of measures, which allowed longitudinal investigations on trajectories of functional limitations. Our findings suggest that by addressing questions of change over time using multiple waves, more nuanced patterns of physical functioning emerge by gender and age. Moreover, this study also provided some evidence for possible changes with age in the association of social integration with functional health, which has not been demonstrated in previous research. While the health benefits of social integration have been well documented, the focus of the present study on trajectories of physical functioning and age difference is unique and makes an important contribution to the existing literature. Future research should continue to employ the use of longitudinal designs to answer similar questions. In particular, while this study follows participants over 4 years, studies over even long time periods would help provide a fuller picture of the protective nature of social integration for functional health.

Another strength of the current study is the use of an older sample (aged 80 at baseline on average). Many prior studies examined differences in physical functioning between midlife and older adulthood; by using an older sample including young-old and old-old adults, the current study focused on identifying age patterns within older adulthood. While the advanced age of this sample allowed for the examination of unique age differences in late life, the sample was a convenience sample and was somewhat lacking in gender diversity, as is common amongst studies of older adults. Given that a majority of the participants of this study were female, a future research direction would be to examine issues of gender, physical functioning, and social integration within a more gender-balanced sample. In addition, a limitation of the current study was that the participants were predominantly White (98%). This reflected the population in the specific geographic area to some extent, but it was not representative for those in more racially diverse areas as well as the national population of the United States. Thus, the generalizability of the present findings is limited and may not be applicable to people with various racial/ethnic backgrounds. Future research should replicate these findings using more diverse and representative samples of the general older population.

Finally, a strength of this study was the use of a multidimensional measure of social integration, which included both satisfaction with and frequency of social support and engagement. The benefit of using such a multidimensional measure is that it provides a more holistic picture of social integration on a broad level. On the other hand, there are benefits in examining the various specific components of social integration in depth, so a future direction for this line of research is to determine whether specific aspects of social integration (e.g., social participation) play a larger role in protecting against decline in physical functioning in late life. Moreover, future research should examine the potential mechanisms of how various aspects of social integration can alter trajectories of functional health particularly for old-old individuals.

Practical Implications

While further investigations are warranted, the present findings can inform practice. The findings suggest that aging adults have higher risks for having greater functional limitations as they became older, and women at more advanced ages have greater risks for functional decline over time. These findings also highlight that enhancing social integration should be considered especially for old-old individuals as a potential avenue to reducing such functional decline with age. While the potential benefits of social integration for functional health may be similar for both genders, interventions to promote social integration and functional health may need to be tailored to older women and men. For example, for older women, it could be beneficial to integrate strength training (Correa et al., 2012) into existing group social settings as an intervention to promote both social integration and physical functioning. For older men, while such physical training in a group may also be valuable, organizing social gatherings targeting older men’s specific hobbies and interests may better help them stay socially connected and bolster their social integration, given their tendency to be less socially integrated than older women. These potential applied approaches to enhancing social integration and functional health should be examined in future research.

Conclusions

The present study aimed to investigate age and gender differences in the trajectories of late-life functional limitations and the association of social integration with those trajectories. In addition to providing additional evidence for greater functional decline particularly for old-old women as compared to men, the findings suggest that the link between social integration and functional health increasingly strengthens over time for old-old adults as compared to their younger counterparts. The present findings suggest future directions of research and highlights the potential benefits of enhancing social integration for older men and women to maintain their functional health.

Supplemental Material

FIGURE S1. - Supplemental material for Longitudinal Implications of Social Integration for Age and Gender Differences in Late-Life Physical Functioning

Supplemental material, FIGURE S1., for Longitudinal Implications of Social Integration for Age and Gender Differences in Late-Life Physical Functioning by Masahiro Toyama, Heather R. Fuller and Jonix Owino in The International Journal of Aging and Human Development

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Masahiro Toyama

Note

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Masahiro Toyama (PhD in Developmental Science and Gerontology [dual-major], North Dakota State University) is an assistant professor of Psychology at the University of the Ozarks.

Heather R. Fuller (PhD in Developmental Psychology, University of Michigan) is an associate professor of Human Development and Family Science at North Dakota State University.

Jonix Owino (MS in Sociology, North Dakota State University) is a doctoral candidate in Developmental Science at North Dakota State University.

References

Antonucci

T. C

. (2001). Social relations an examination of social networks, social support. In Birren

Schaie

(Eds.), Handbook of the psychology of aging (pp. 427–453). Academic Press.

Antonucci

T. C.

Ajrouch

K. J.

Birditt

K. S

. (2014). The convoy model: Explaining social relations from a multidisciplinary perspective. The Gerontologist, 54(1), 82–92.doi:10.1093/geront/gnt118

Avlund

Lund

Holstein

B. E.

Due

. (2004). Social relations as determinant of onset of disability in aging. Archives of Gerontology and Geriatrics, 38(1), 85–99.doi:10.1016/j.archger.2003.08.003

Barger

S. D

. (2013). Social integration, social support and mortality in the US National health interview survey. Psychosomatic Medicine, 75(5), 510–517.doi:10.1097/PSY.0b013e318292ad99

Berkman

L. F.

Glass

Brissette

Seeman

T. E

. (2000). From social integration to health: Durkheim in the new millennium. Social Science & Medicine, 51(6), 843–857.doi:10.1016/S0277-9536(00)00065-4

Bird

Rieker

P. P

. (2008). Gender and health: The effects of constrained choices and social policies. Cambridge University Press.

Botoseneanu

Allore

H. G.

Mendes de Leon

C. F.

Gahbauer

E. A.

Gill

T. M

. (2016). Sex differences in concomitant trajectories of self-reported disability and measured physical capacity in older adults. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 71(8), 1056–1062.doi:10.1093/gerona/glw038

http://www.ncbi.nlm.nih.gov/pubmed/27071781

Buchman

A. S.

Boyle

P. A.

Wilson

R. S.

Fleischman

D. A.

Leurgans

Bennett

D. A

. (2009). Association between late-life social activity and motor decline in older adults. Archives of Internal Medicine, 169(12), 1139–1146.doi:10.1001/archinternmed.2009.135

Chatterji

Byles

Cutler

Seeman

Verdes

. (2015). Health, functioning, and disability in older adults—present status and future implications. The Lancet, 385(9967), 563–575.doi:10.1016/S0140-6736(14)61462-8

10.

Christensen

McGue

Petersen

Jeune

Vaupel

J. W

. (2008). Exceptional longevity does not result in excessive levels of disability. Proceedings of the National Academy of Sciences, 105(36), 13274–13279.doi:10.1073/pnas.0804931105

11.

Cohen

. (2004). Social relationships and health. American Psychologist, 59(8), 676–684.doi:10.1037/0003-066X.59.8.676

12.

Corbett

D. B

Rejeski

W. J

Tudor-Locke

Glynn

N. W

Kritchevsky

S. B

McDermott

M. M

Church

T. S

Fielding

R. A

Gill

T. M

King

A. C

Miller

M. E

Chen

Pahor

Manini

T. M

Pahor

Guralnik

J. M

Leeuwenburgh

Caudle

Crump

Kaplan

R. M

LIFE Study Investigators . (2018). Social participation modifies the effect of a structured physical activity program on major mobility disability among older adults: Results from the life study. The Journals of Gerontology: Series B, 73(8), 1501–1513.doi:10.1093/geronb/gbx051

http://www.ncbi.nlm.nih.gov/pubmed/28482106

13.

Correa

C. S.

LaRoche

D. P.

Cadore

E. L.

Reischak-Oliveira

Bottaro

Kruel

L. F. M.

Tartaruga

M. P.

Radaelli

Wilhelm

E. N.

Lacerda

F. C.

Gaya

A. R.

Pinto

R. S

. (2012). 3 different types of strength training in older women. International Journal of Sports Medicine, 33(12), 962–969.doi:10.1055/s-0032-1312648

http://www.ncbi.nlm.nih.gov/pubmed/22782384

14.

Crimmins

E. M.

Kim

J. K.

Solé-Auró

. (2011). Gender differences in health: Results from share, ELSA and HRS. European Journal of Public Health, 21(1), 81–91.doi:10.1093/eurpub/ckq022

20237171

15.

Crimmins

E. M.

Zhang

Saito

. (2016). Trends over 4 decades in disability-free life expectancy in the United States. American Journal of Public Health, 106(7), 1287–1293.doi:10.2105/AJPH.2016.303120

16.

Crittenden

C. N.

Pressman

S. D.

Cohen

Janicki-Deverts

Smith

B. W.

Seeman

T. E

. (2014). Social integration and pulmonary function in the elderly. Health Psychology, 33(6), 535–543.doi:10.1037/hea0000029

17.

Davidson

Daly

Arber

. (2003). Older men, social integration and organisational activities. Social Policy and Society, 2(2), 81–89.doi:10.1017/S1474746403001118

18.

Deeg

D. J. H

. (2005). Longitudinal characterization of course types of functional limitations. Disability and Rehabilitation, 27(5), 253–261.doi:10.1080/09638280400006507

19.

Elliot

A. J.

Heffner

K. L.

Mooney

C. J.

Moynihan

J. A.

Chapman

B. P

. (2018). Social relationships and inflammatory markers in the MIDUS cohort: The role of age and gender differences. Journal of Aging and Health, 30(6), 904–923.doi:10.1177/0898264317698551

http://www.ncbi.nlm.nih.gov/pubmed/28553813

20.

Ertel

K. A.

Glymour

M. M.

Berkman

L. F

. (2009). Social networks and health: A life course perspective integrating observational and experimental evidence. Journal of Social and Personal Relationships, 26(1), 73–92.doi:10.1177/0265407509105523

21.

Fillenbaum

G. G

. (2013). Multidimensional functional assessment of older adults: The Duke older Americans resources and services procedures. Psychology Press.

22.

Freedman

V. A.

Schoeni

R. F.

Martin

L. G.

Cornman

J. C

. (2007). Chronic conditions and the decline in late-life disability. Demography, 44(3), 459–477.doi:10.1353/dem.2007.0026

23.

Freedman

V. A.

Spillman

B. C.

Andreski

P. M.

Cornman

J. C.

Crimmins

E. M.

Kramarow

Lubitz

Martin

L. G.

Merkin

S. S.

Schoeni

R. F.

Seeman

T. E.

Waidmann

T. A

. (2013). Trends in late-life activity limitations in the United States: An update from five national surveys. Demography, 50(2), 661–671.doi:10.1007/s13524-012-0167-z

http://www.ncbi.nlm.nih.gov/pubmed/23104207

24.

Fuller

H. R.

Ajrouch

K. J.

Antonucci

T. C

. (2020). Original voices the convoy model and later‐life family relationships. Journal of Family Theory & Review, 12, 126–146.doi:10.1111/jftr.12376

25.

Fuller-Iglesias

H. R.

Rajbhandari

. (2016). Development of a multidimensional scale of social integration in later life. Research on Aging, 38(1), 3–25.doi:10.1177/0164027514568324

26.

Gill

T. M.

Gahbauer

E. A.

Lin

Han

Allore

H. G

. (2013). Comparisons between older men and women in the trajectory and burden of disability over the course of nearly 14 years. Journal of the American Medical Directors Association, 14(4), 280–286.doi:10.1016/j.jamda.2012.11.011

27.

Green

A. F.

Rebok

Lyketsos

C. G

. (2008). Influence of social network characteristics on cognition and functional status with aging. International Journal of Geriatric Psychiatry, 23(9), 972–978.doi:10.1002/gps.2023

28.

Hardy

S. E.

Allore

H. G.

Guo

Gill

T. M

. (2008). Explaining the effect of gender on functional transitions in older persons. Gerontology, 54(2), 79–86.doi:10.1159/000115004

29.

Heck

R. H.

Thomas

S. L.

Tabata

L. N

. (2014). Multilevel and longitudinal modeling with IBM SPSS (2nd ed.). Routledge.

30.

Hoffman

. (2015). Longitudinal analysis: Modeling within-person fluctuation and change. Routledge.

31.

Holmes

Powell-Griner

Lethbridge-Cejku

Heyman

. (2009). Aging differently: physical limitations among adults aged 50 years and over: United States, 2001–2007 (NCHS data brief 20). National center for health statistics. http://www.cdc.gov/nchs/data/databriefs/db20.htm

32.

Holt-Lunstad

Smith

T. B.

Layton

J. B

. (2010). Social relationships and mortality risk: A meta-analytic review. doi:10.1371/journal.pmed.1000316

33.

IBM . (2020). IBM SPSS software. https://www.ibm.com/products/spss-statistics

34.

James

B. D.

Boyle

P. A.

Buchman

A. S.

Bennett

D. A

. (2011). Relation of late-life social activity with incident disability among community-dwelling older adults. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 66(4), 467–473.doi:10.1093/gerona/glq231

http://www.ncbi.nlm.nih.gov/pubmed/21300745

35.

Kanamori

Kai

Aida

Kondo

Kawachi

Hirai

Shirai

Ishikawa

Suzuki

JAGES Group . (2014). Social participation and the prevention of functional disability in older Japanese: The JAGES cohort study. PLoS ONE, 9(6), e99638.doi:10.1371/journal.pone.0099638

http://www.ncbi.nlm.nih.gov/pubmed/24923270

36.

Leveille

S. G.

Resnick

H. E.

Balfour

. (2000). Gender differences in disability: Evidence and underlying reasons. Aging Clinical and Experimental Research, 12(2), 106–112.doi:10.1007/BF03339897

37.

Hagan

Grodstein

VanderWeele

T. J

. (2018). Social integration and healthy aging among U.S. women. Preventive Medicine Reports, 9, 144–148.doi:10.1016/j.pmedr.2018.01.013

38.

Liang

Bennett

J. M.

Shaw

B. A.

Quiñones

A. R.

Ofstedal

M. B

. (2008). Gender differences in functional status in middle and older age: Are there any age variations? The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 63(5), S282–S292.doi:10.1093/geronb/63.5.S282

http://www.ncbi.nlm.nih.gov/pubmed/18818448

39.

Lin

. (2020). Inter-individual variability in trajectories of functional limitations by race/gender. The Journals of Gerontology: Series B, 75(5), 1082–1092.doi:10.1093/geronb/gby156

40.

Martin

L. G.

Schoeni

R. F

. (2014). Trends in disability and related chronic conditions among the forty-and-over population: 1997–2010. Disability and Health Journal, 7(1), S4–S14.doi:10.1016/j.dhjo.2013.06.007

41.

McLaughlin

Vagenas

Pachana

N. A.

Begum

Dobson

. (2010). Gender differences in social network size and satisfaction in adults in their 70s. Journal of Health Psychology, 15(5), 671–679.doi:10.1177/1359105310368177

42.

Nilsson

C. J.

Avlund

Lund

. (2010). Mobility disability in midlife: A longitudinal study of the role of anticipated instrumental support and social class. Archives of Gerontology and Geriatrics, 51(2), 152–158.doi:10.1016/j.archger.2009.09.043

43.

Ong

A. D.

Uchino

B. N.

Wethington

. (2016). Loneliness and health in older adults: A mini-review and synthesis. Gerontology, 62(4), 443–449.doi:10.1159/000441651

44.

Otsuka

Tomata

Zhang

Sugiyama

Tanji

Sugawara

Tsuji

. (2018). Association between social participation and incident risk of functional disability in elderly Japanese: The Ohsaki cohort 2006. Journal of Psychosomatic Research, 111, 36–41.doi:10.1016/j.jpsychores.2018.05.004

45.

Romoren

Blekeseaune

. (2003). Trajectories of disability among the oldest old. Journal of Aging and Health, 15(3), 548–566.doi:10.1177/0898264303253633

46.

Rooth

van Oostrom

S. H.

Deeg

D. J. H.

Verschuren

W. M. M.

Picavet

H. S. J

. (2016). Common trajectories of physical functioning in the Doetinchem cohort study. Age and Ageing, 45(3), 382–388.doi:10.1093/ageing/afw018

http://www.ncbi.nlm.nih.gov/pubmed/26892339

47.

Rosso

A. L.

Taylor

J. A.

Tabb

L. P.

Michael

Y. L

. (2013). Mobility, disability, and social engagement in older adults. Journal of Aging and Health, 25(4), 617–637.doi:10.1177/0898264313482489

48.

Rowe

J. W.

Kahn

R. L

. (1997). Successful aging. The Gerontologist, 37(4), 433–440.doi:10.1093/geront/37.4.433

49.

Seeman

Chen

. (2002). Risk and protective factors for physical functioning in older adults with and without chronic conditions: MacArthur studies of successful aging. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 57(3), S135–S144.doi:10.1093/geronb/57.3.S135

http://www.ncbi.nlm.nih.gov/pubmed/11983740

50.

Seeman

T. E

. (1996). Social ties and health: The benefits of social integration. Annals of Epidemiology, 6(5), 442–451.doi:10.1016/S1047-2797(96)00095-6

51.

Seeman

T. E.

Merkin

S. S.

Crimmins

E. M.

Karlamangla

A. S

. (2010). Disability trends among older Americans: National health and nutrition examination surveys, 1988–1994 and 1999–2004. American Journal of Public Health, 100(1), 100–107.doi:10.2105/AJPH.2008.157388

http://www.ncbi.nlm.nih.gov/pubmed/19910350

52.

Sharifian

Grühn

. (2019). The differential impact of social participation and social support on psychological well-being: Evidence from the Wisconsin longitudinal study. The International Journal of Aging and Human Development, 88(2), 107–126.doi:10.1177/0091415018757213

53.

Stenholm

Westerlund

Head

Hyde

Kawachi

Pentti

Kivimäki

Vahtera

. (2015). Comorbidity and functional trajectories from midlife to old age: The health and retirement study. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 70(3), 332–338.doi:10.1093/gerona/glu113

http://www.ncbi.nlm.nih.gov/pubmed/25060316

54.

Stephens

Noone

Alpass

. (2014). Upstream and downstream correlates of older people’s engagement in social networks: What are their effects on health over time? The International Journal of Aging and Human Development, 78(2), 149–169.doi:10.2190/AG.78.2.d

http://www.ncbi.nlm.nih.gov/pubmed/24956923

55.

Tarantino

Baldi

Celi

Rao

Liuni

F. M.

Iundusi

Gasbarra

. (2013). Osteoporosis and sarcopenia: The connections. Aging Clinical and Experimental Research, 25(S1), 93–95.doi:10.1007/s40520-013-0097-7

56.

Terry

D. F.

Sebastiani

Andersen

S. L.

Perls

T. T

. (2008). Disentangling the roles of disability and morbidity in survival to exceptional old age. Archives of Internal Medicine, 168(3), 277–283.doi:10.1001/archinternmed.2007.75

57.

Thoits

P. A

. (2011). Mechanisms linking social ties and support to physical and mental health. Journal of Health and Social Behavior, 52(2), 145–161.doi:10.1177/0022146510395592

58.

Thomas

P. A

. (2011). Gender, social engagement, and limitations in late life. Social Science & Medicine, 73(9), 1428–1435.doi:10.1016/j.socscimed.2011.07.035

59.

Toyama

Fuller

H. R

. (2020). Longitudinal stress-buffering effects of social integration for late-life functional health. The International Journal of Aging and Human Development, 91(4), 501–519.doi:10.1177/0091415019871196

60.

Uchino

B. N

. (2004). Social support and physical health: Understanding the health consequences of relationships. Yale University Press.

61.

Umberson

Montez

J. K

. (2010). Social relationships and health: A flashpoint for health policy. Journal of Health and Social Behavior, 51(1_suppl), S54–S66.doi:10.1177/0022146510383501

http://www.ncbi.nlm.nih.gov/pubmed/20943583

62.

Unger

J. B.

McAvay

Bruce

M. L.

Berkman

Seeman

. (1999). Variation in the impact of social network characteristics on physical functioning in elderly persons: MacArthur studies of successful aging. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 54(5), S245–S251.doi:10.1093/geronb/54B.5.S245

http://www.ncbi.nlm.nih.gov/pubmed/10542826

63.

Verbrugge

L. M.

Latham

Clarke

P. J

. (2017). Aging with disability for midlife and older adults. Research on Aging, 39(6), 741–777.doi:10.1177/0164027516681051

64.

Zunzunegui

Rodriguez-Laso

Otero

Pluijm

S. M. F

Nikula

Blumstein

Jylhä

Minicuci

Deeg

D. J. H

CLESA Working Group . (2005). Disability and social ties: Comparative findings of the CLESA study. European Journal of Ageing, 2(1), 40–47.doi:10.1007/s10433-005-0021-x

http://www.ncbi.nlm.nih.gov/pubmed/28794715

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.16 MB