Research on the Association Between Perceived Neighborhood Environment and Life Satisfaction Among Older Adults in Shanghai, China: Do Social Capital And Gender Matter?

This study examined the mediator role of social capital on the association between perceived neighborhood environment and life satisfaction among older adults in urban China, and further tested the moderating effect of gender in the above paths (i.e., from neighborhood environment to life satisfaction; from neighborhood environment to social capital; from social capital to life satisfaction). We used quota sampling approach to recruit 472 respondents aged 60 years old or older in Shanghai in 2020. From the perspective of structural equation modeling, multiple group analysis was conducted to examine the proposed hypotheses. The measurement model of social capital was well established in urban Chinese community contexts. Based on the whole sample, the results of the mediation model showed that social capital played a mediation role in the association between neighborhood environment and life satisfaction. Furthermore, the results of multiple group analysis showed that the association between neighborhood environment and cognitive social capital was only signicant among older women. The ndings highlight the role of neighborhood environment and social capital in building age-friendly communities.


Introduction
The global population aging process has accelerated rapidly, and this trend will continue in the next few decades. China has the largest older population in the world 1  Life satisfaction is frequently used to measure the cognitive dimension of subjective well-being in later life and therefore, is recognized as a crucial indicator of healthy aging 3 . Older adults who are satis ed with their lives are more likely to remain not only independently functional but also actively involved in social activities to pursue personal and collective interests. Policy and interventions around subjective well-being in later life therefore, should pay attention to modi able and culturally sensitive factors of life satisfaction in later life.
Previous literature has identi ed factors associated with life satisfaction, including socioeconomic status, physical health, living alone, interpersonal relationships, and access to health care 4,5 .
Furthermore, neighborhood environment was found to play an important role in promoting life satisfaction and other mental well-being indicators in later life [6][7][8][9] .
Neighborhood environment is a multi-dimensional concept and includes physical neighborhood environment and social neighborhood environment 10,11 . While the former includes transportation, outdoor spaces and buildings, and community services, the later includes social participation, volunteering, trust, and reciprocity 7,10,12−14 . In particular, community-based social capital (hereafter social capital), an important social supportive resource embedded from individuals' social connections, is recognized as a key indicator of social neighborhood environment. Social capital could sustain the wellbeing of older adults through promoting social support, information diffusion of health knowledge, health service utilization, and health behavior 8, 15 .
As compared with objective physical neighborhood environment indicators, older adults' perceived physical neighborhood environment was found to be a stronger factor of their mental well-being 16,17 . This might partially because that perceived neighborhood environment re ects individuals' variations in not only their access to neighborhood supportive resources, but also their capacities to utilize such resources. Therefore, in this research, we used older adults' subjective evaluations to assess physical neighborhood environment. In particular, community health care, public transportation, and community security play an important role in sustaining older residents' active involvement in community activities and ful lling their medical and daily care needs. These factors could further enhance their satisfaction levels with daily lives.
In summary, perceived neighborhood environment and social capital were found to be two crucial factors related to life satisfaction among older adults 8,18,19 . However, there is lack of studies examining how neighborhood environment affect social capital, the pathways linking these factors to life satisfaction, and the potential moderators in these associations, especially in the contexts of developing countries and regions. Therefore, this study tested the mediation effects of social capital on the relationship between neighborhood environment and life satisfaction and further examined the moderator role of gender in these associations in urban Chinese community contexts.

Neighborhood Environment and Life Satisfaction
According to ecological model of ageing, both physical and social neighborhood environments play important roles in promoting well-being in later life through person-environment interactions 6,7,20,21 . For older adults, good-quality physical neighborhood environment (e.g., adequate access to health care, secure living environment, and good transportation conditions) could allow them to be more adaptive to their age-related social and health changes, which could further enhance their satisfaction levels with their lives. Proponents of ecological model of ageing also put great emphasis on the potential in uence of physical environment on social environment, and their effects on health outcomes in later life 22 . For example, good-quality of physical neighborhood environment (e.g. accessible public transportation and good community security) could lead to the increase in social capital (e.g., high levels of sense of belonging to local community, active participation in community activities and volunteering) 11 . Therefore, we argue that physical neighborhood environment could affect life satisfaction of older adults through community-based social capital.
As the most economically developed city in China, Shanghai aims to develop age friendly communities and promote the quality of neighborhood environment in all local rural and urban communities by 2035. For example, the city plans to develop community health service centers and day care centers that deliver regular physical examinations, need assessments, primary care, prevention, long-term care, and hospice care for older residents in all communities. Walkable environments and the accessibility and safety of transportation are other development goals. Moreover, Shanghai also plans to develop safe, livable environment, educational programs preventing fraudulent activities, and neighborhood watch for older adults living alone and those with disabilities. Furthermore, it plans to develop community organizations and activities to promote older adults' participation in community activities, peer support groups, and volunteering. Older adults are also encouraged to participate in community development and administration. However, these programs are in the process of being developed, and it is important to provide updated evidence on whether these activities have an impact on well-being among older adults.
The majority of research on neighborhood environment and life satisfaction among older adults has been conducted in developed and Western countries and regions 7,10,11 . Although the literature supports the role of neighborhood environment in enhancing life satisfaction in later life, ndings on the associations between speci c neighborhood environment indicators and life satisfaction are mixed 6,19,23−25 . For example, transportation, community support, health services, and social environment were also found to be signi cantly associated with life satisfaction among Canadian older adults 25 . Park and Lee 6 found that perceived neighborhood environment indicators such as transportation, social services, social participation, and social inclusion were signi cantly associated with life satisfaction among older adults in Seoul, South Korea. In a recent study conducted in Hong Kong, transportation and social participation were associated with life satisfaction 24 . However, community and health services were not associated with life satisfaction among adults aged 75 years or older 24 . Based on a nationally representative sample, Xie 19 found that older adults' perception of local amenities was signi cantly associated with life satisfaction among older adults in urban China. However, the association between community services and life satisfaction was nonsigni cant.
Presumably, these mixed results might partially result from the lack of consensus on the conceptualization and measurement of social and physical aspects of the neighborhood environment 7,20 . Different measurements have been used in empirical studies, making the comparison of the ndings even more di cult. Furthermore, there is a lack of examination of the interplay between physical neighborhood environment and social capital, especially in developing country contexts.

Page 5/18
The conceptualization and operationalization of social capital can be examined from multiple perspectives (e.g., collectivist and individualist) and levels (e.g., micro and macro) 3,26 . As a commonly used de nition of social capital in the health-related eld, Putnam conceptualized social capital from a collectivist perspective and described it as "features of social organization, such as trust, norms, and networks, that can improve the e ciency of society by facilitating coordinated actions" 27 . Social capital also refers to individuals' supportive resources based on their social connections in local communities, where people share cultural values and social norms and organization memberships 28 .
Regarding the measurement of social capital, it is considered a multidimensional concept, including cognitive and structural dimensions 3 . Cognitive social capital is re ected by subjective appraisals of social trust and reciprocity in the community. Structural social capital, on the other hand, refers to objective aspects of social capital and is often measured by memberships in community organizations, volunteering, social participation in community activities, and citizenship activities to address common problems in the community 26 . Community social capital can not only provide older adults with social support and a social network, but also promote collective action in communities and enhance a sense of belonging and meaning in life, which is important in older age 3,8 .
Research on social capital and well-being (including life satisfaction) in later life has been mainly conducted in developed and Western contexts 8 . Systematic reviews found that individual-level social capital was a stronger factor linked to life satisfaction than collective social capital (e.g., community and country levels) 3,8 . Furthermore, the association between cognitive social capital and life satisfaction was found to be stronger than that between structural social capital and life satisfaction 3,8 .
Whereas social capital was recognized as an important factor of life satisfaction, research ndings on the relationships between speci c domains of social capital and life satisfaction were inconclusive. For example, both cognitive and structural social capital were found to be signi cant factors related to life satisfaction among older adults in developed countries 8, 29,30 . Speci cally, trust, reciprocity, organization memberships, and citizenship activities were signi cantly associated with life satisfaction 29,31,32 .
However, ndings on the associations between country-level social capital indicators such as trust, volunteering, and organization memberships and life satisfaction were mixed in developed countries 33,34 . Furthermore, the association between structural social capital indicators and life satisfaction was found to be nonsigni cant in Chinese contexts 35,36 .
Furthermore, low social trust and reciprocity were found in poor neighborhood with high crime prevalence rates 37 . The literature has found signi cant association between the perception of neighborhood characteristics and cognitive social capital indicators (e.g., a sense of communities) 38, 39 . Perceived community security was found to be associated with social trust and support among local residents 37 .
Security and transportation were found to be related to the level of physical activities among residents 9 . While the neighborhood environment refers to access to amenities and services in local communities, social capital can be used to diffuse health-related information, promote social participation, and enhance service utilization rates. In this case, the neighborhood environment provides the foundation for social capital to develop. We hypothesized that social capital would play a signi cant mediator role in the association between neighborhood environment and life satisfaction (Hypothesis 1).

The Moderating Role of Gender
Although older men and women tend to have different social roles in their community lives, limited studies have examined gender differences in the relationships among neighborhood environment, social capital, and life satisfaction, especially in the contexts of developing countries and regions. For example, although older women tend to bene t more from trust and reciprocity among neighbors 18,40 , older men bene t more from formal activities such as citizenship activities (e.g., local residents working together to solve common problems in the communities) 18 . Empirical research also identi ed gender differences in neighborhood walking 41 . Compared with older men, older women tend to have higher levels of involvement in informal reciprocity, caring activities, and other activities held by community organizations. In contrast, older men are more likely to get involved in relatively formal social activities. Furthermore, there is a lack of research on the potential moderator role of gender in the relationship between neighborhood environment and social capital. Therefore, we proposed that gender would play a moderator role in the pathways linking neighborhood environment, social capital, and life satisfaction (Hypothesis 2).

Sampling
Data were based on a social survey titled "Social Capital, Intergenerational Solidarity, and Mental Health among Older Chinese Adults" (SCIENCE). Quota sampling was used to select respondents from Siping Street, Yangpu District, Shanghai, China, in summer 2020. As the most economically developed city in China, Shanghai has seen the aging process of its local population in Shanghai accelerate rapidlyaround 35% of residents were aged 60 years or older in 2019 (the national average is 18.1%). Yangpu is in the northeast of Shanghai's central area. Its economic development is at the medium level among the 16 districts in Shanghai. Siping Street consists of 23 communities. Tongji University is in this area. If the university staff and students were not counted, the proportion of residents aged 60 years or older is above 40%. Furthermore, Shanghai has initiated a second round of pilot projects to build age friendly communities for older adults with cognitive impairments. Siping Street is one of the participating streets in these projects. Therefore, this area is suitable for studying neighborhood environment, social capital and life satisfaction among older adults.
The sampling procedures were as follows: In the rst step, we included all 23 communities from Siping Street. In the second step, we recruited 20 respondents from each community, with the support of local community centers. We controlled for the age and gender ratios of the respondents in each community to ensure consistency with those from a representative sample (collected in 2019) surveyed by the Shanghai Municipal Health Commission (e.g., age group 60-64: 48% men and 52% women; age group 75 or older: 41% men and 59% women). The inclusion criteria of the sample were: (a) aged 60 years or older; (b) had local household registration status; (c) lived in local communities for more than 6 months in the past 12 months; and (d) passed a cognitive screening test based on the Short Portable Mental Status Questionnaire 42 . The cutoff point was 6 for those with a high school education or lower and 7 for those with a college education or higher.
Six trained interviewers conducted face-to-face interviews with respondents at community centers. For those with limited functional health, interviews were conducted at their home. The survey included rich information in terms of older adults' sociodemographic characteristics, depressive symptoms, loneliness, life satisfaction, neighborhood environment, and social capital. Of 498 respondents, 476 agreed to participate in the survey and successfully completed the interview. Four respondents did not pass the cognitive screening test. The response rate was 94.77%. Ethics approval was obtained from the University Ethics Committee. All methods were performed in accordance with the relevant guidelines and regulations. Informed consent forms were obtained before collecting the data. The nal sample size was 472.

Measurement
Life satisfaction variable. Life satisfaction was measured by the 8-item Life Satisfaction Scale-Chinese, which was designed to measure life satisfaction among older Chinese populations 43  The other four indicators were used to assess structural social capital. The number of organization memberships for each respondent in the past year was calculated based on the following list: political organizations, religious groups, work unions, women's groups, community associations, sports groups, charity groups, professional associations, neighborhood committees, and community colleges for older adults.
The frequency of the respondents' participation in social activities organized by these social organizations in the past year were assessed based on a 6-point Likert scale, ranging from 1 (never) to 3 (several times per year) to 6 (twice or more per week). The respondents were also asked whether they worked with other residents to solve common problems in local communities in the last year (0 = never, 2 = occasionally, 4 = almost always) and whether they engaged in volunteering activity in the past 30 days (0 = no, 1 = yes).
Perceived neighborhood environment variable. Instead of objective measure of neighborhood, perceived indicators of neighborhood environment were found to be a stronger factor linked to mental well-being 16 ; thus, neighborhood environment was measured by older residents' subjective appraisals in this study. The respondents were asked about their satisfaction with three domains of the neighborhood environment: community health care, security, and transportation conditions. Responses were measured on a 5-point Likert scale (0 = strongly disagree, 2 = neutral, 4 = strongly agree).
Moderator and covariates. The moderator was gender, assessed by a binary variable (0 = male, 1 = female). Regarding covariates, age was measured in years. Marital status, living arrangement, and educational attainment were assessed by binary variables (1 = married, 0 = other; 1 = living alone, 0 = living with others; 1 = secondary school education or higher, 0 = primary school education or lower). The log value of household income per year was calculated. Furthermore, the respondents were asked whether they have any of 14 doctor-diagnosed chronic diseases, including diabetes, arthritis, hypertension, stroke, and cardiovascular diseases. Summed scores were calculated to represent the number of chronic diseases. Finally, cognitive function was assessed by Short Portable Mental Status Questionnaire 42 , with a range of 0-10. Higher scores indicate higher cognitive function.

Data Analysis
We used structural equation modeling (SEM) in to examine the mediation model of perceived neighborhood environment, social capital, and life satisfaction. From a SEM perspective, a multiple-group analysis was further conducted to test the gender difference in the mediation model. Mplus 7.0 was used 44 . This approach, widely used in empirical studies, allowed us to test whether gender moderates the mediation model 46,47 .
The analytic procedures were as follows: First, we conducted a measurement model of cognitive and structural social capital using con rmatory factor analysis. The following t indexes and cutoff points were used to determine whether the model adequately t the data: the chi-square test statistic (nonsigni cant estimate indicates good model t), weighted root mean square residual (WRMR; less than 1 indicates good t); Tucker-Lewis index (TLI; > .90 indicates good t), comparative t index (CFI; > .90 indicates good t), root mean square error of approximation (RMSEA; < .05 indicates good t), and standardized root mean square residual (SRMR; < .08 indicates good t) 48 .
Second, we conducted a structural model to regress life satisfaction on perceived neighborhood environment while controlling covariates. Furthermore, structural and cognitive social capital were entered in the structural model to test the indirect effect of perceived neighborhood environment on life satisfaction through social capital.
Finally, we built multiple-group models to test the moderation role of gender in the mediation model. The speci c procedures were as follows: First, we tested the measurement model in each gender group separately. Second, factor loadings of latent variables were held equal between the two gender groups.
This allowed us to conduct meaningful comparisons between the groups 49 . Third, outcome variable and covariates were added to the structural model. We tested the moderation effects of gender on the relationships among neighborhood environment, social capital, and life satisfaction using Wald tests. All missingness in key variables was less than 3%. Listwise deletion was used in the analysis. Table 1 shows the sociodemographic characteristics of older respondents. The respondents' average age was 68.37 years, 56.4% were women, more than 80% were married, and only 12.7% lived alone at the time of the survey. Around half of the respondents reported their monthly household income was 9,000 RMB or less, and 57.8% completed a high school education or more. The average number of chronic diseases was 1.37. The most common chronic diseases were hypertension (47.5%), heart disease (18.4%), diabetes or high blood sugar (16.3%), and arthritis (9.7%). On average, around 75% of the respondents were satis ed or very satis ed with their lives (average score of 4 or higher).

Multiple-Group Analysis
We further examined gender differences in the relationships between neighborhood environment, cognitive social capital, and life satisfaction. Because the association of structural social capital and life satisfaction was nonsigni cant for both men and women, structural social capital was not included in the multiple-group analysis.
First, measurement models of cognitive social capital were established in both gender groups, even after factor loadings of cognitive social capital were held equal between the two gender groups, χ 2 (6) = 6.167, p = .4048, RMSEA = .011, CFI = 1.000, TLI = 1.000, WRMR = .049. In other words, we established factor loading invariance in this latent variable.
Second, life satisfaction, neighborhood environment, and covariates were entered in the nal model,  Figure 1).

Discussion
The present study aimed to test a mediation model of neighborhood environment, social capital, and life satisfaction among older adults living in urban China, as well as the moderator role of gender in the paths of the above mediation model. The ndings represent new contributions by highlighting the role of social capital in the relationship between neighborhood environment and life satisfaction, and also further examining gender differences in these associations. This study also adopted social capital theory to examine the in uence of physical environment on social environment in the framework of the ecological model of aging.
Consistent with the ndings of previous research, cognitive social capital, rather than structural social capital, was signi cantly related to life satisfaction when controlling for neighborhood environment.
Furthermore, perceived neighborhood environment was signi cantly associated with life satisfaction in the nal model. This nding is also consistent with those of previous studies. It provides further evidence on the importance of neighborhood environment and its impact on life satisfaction. Furthermore, cognitive social capital played a partial mediation role in the relationship between community health care, security, and life satisfaction. Transportation, in contrast, tended to have a direct effect on life satisfaction. For example, in our data collection site (i.e., Siping Street), most residents have adequate access to public transportation, and many shopping areas are easily accessible. It is likely that a goodquality neighborhood environment in terms of health care and security could play an important role in fostering cognitive social capital (i.e., trust and reciprocity). Furthermore, social capital might enhance life satisfaction by enhancing supportive resources, sharing information, promoting health-related behavior, and promoting use of home-and community-based services. Finally, the ndings add new empirical evidence by identifying gender differences in the relationship between community health care service and cognitive social capital. The above relationship was signi cant among women but not men. In Chinese culture, women are more likely to take on caregiving roles than men. Community health care services might provide opportunities for older women to not only ful ll their social role as caregivers, but also facilitate their social participation, information sharing, and reciprocity among neighbors. In other words, community health care services are likely to affect life satisfaction among older women both directly and indirectly. For men, health care services are more likely to be a crucial source of health support, rather than important opportunities for socializing.
The ndings have the following policy and intervention implications. First, both neighborhood environment and social capital should be included in older adults' needs assessment and used to identify older adults who are at risk of poor life satisfaction and subjective well-being. Second, social capital policies and interventions should focus on not only promoting social trust and reciprocity in the neighborhood, but also helping older adults utilize available community-and home-based health services and outdoor public spaces. In doing so, the resources embedded in the neighborhood environment could be used in a more e cient manner, which could further promote a sense of life satisfaction. Third, older women' participation in health care services could also be important opportunities for them to socialize. This could be used to enhance both health service utilization and social capital (e.g., social participation and reciprocity). Fourth, the development of neighborhood environment should emphasize older residents' feelings of bonding and belonging to local communities. Both subjective and objective indicators should be used as part of the outcome evaluation of the neighborhood environment development projects.

Limitations
The limitations of the study are as follows. First, given that the data were cross-sectional, we could not examine the causality between neighborhood environment, social capital, and life satisfaction. Second, objective indicators of physical environment and community services such as road network, public open space, and the availability of daycare centers were not included in this study. Moreover, in this study, we used three indicators, rather than latent construct to assess neighborhood environment. This allows us to test the gender difference in each pathway from neighborhood environment indicators to social capital and life satisfaction. Future studies should use a more comprehensive measurement of both physical and social environments of neighborhood environment, and further develop latent constructs of subdimensions of neighborhood environment in both rural and urban Chinese settings. Furthermore, given that we did not have access to the full list of residents living in Siping Street, we did not use random sampling to collect the data. Quota sampling was used to recruit the respondents. Although we matched the age and gender ratio of the sample to the population of the street, the sample may not be representative of the population of the area and the city of Shanghai. Future studies with larger samples are needed to further address this issue.

Conclusion
In conclusion, the ndings show that perceived neighborhood environment not only directly affected life satisfaction, but also indirectly affected life satisfaction through community-based cognitive social capital. The ndings also show that the relationship between health care services in the community and cognitive social capital was signi cant among older women but not older men. On one hand, the social and physical dimensions of neighborhood environment matter for life satisfaction. On the other hand, indepth understanding of the pathway linking these factors to life satisfaction is needed to comprehensively evaluate the impact of neighborhood environment on subjective well-being among older adults. Declarations