Background
The world’s population is rapidly aging. The population aged 60 years and older is expected to rise to 2 billion by 2050 [
1]. People worldwide are living longer, and now the oldest-old (aged 80 and above) are the fastest-growing age group globally. The number of the oldest-old will have quadrupled between 2000 and 2050 to 434 million [
1]. China is one of the fastest aging countries and has the largest oldest-old population in the world [
2]. There is growing evidence that a person’s dynamic psychological aspects might be related to long-term survival [
3,
4]. Substantial evidence indicates that subjective well-being (SWB) as an important predictor of longevity was associated with a decreased risk of all-cause mortality [
5,
6].
There are two recognized theoretical perspectives in well-being research: SWB emphasizes hedonism or happiness, and psychological well-being (PWB) focuses on eudaemonia or self-realization [
7]. Although SWB is highly correlated to PWB, they belong to different constructs in terms of positive psychological function [
8]. SWB is defined as the cognitive and affective evaluations of one’s life [
9]. The cognitive aspect usually refers to one’s judgments regarding life satisfaction [
9,
10]. The affective aspect typically refers to one’s moods, emotions, and feelings [
9], which is measured by the frequency and intensity of positive affect and negative affect [
11]. In brief, SWB involves three components: life satisfaction, positive affect, and negative affect. PWB is considered to be a process of realizing values that make us feel alive and real, give us a sense of life, and seek to develop personal potential and to achieve full function [
7,
12,
13]. PWB was composed of self-acceptance, autonomy, purpose in life, personal growth, positive relationships, and environmental mastery [
13]. To some extent, it is difficult to distinguish between SWB and PWB. In previous studies, more focused on PWB than on SWB, and some mistook SWB for PWB [
14,
15]. Our study explores the predictors of SWB from the real meaning of it.
In recent years, psychologists have paid more attention to the SWB of older people (aged 60 and above) and its related factors [
16]. However, most studies focused on SWB and its related factors among older people and seldom considered those among the oldest-old [
17]. Life span theories have proposed that older people can regulate age-related losses (e.g., loss of family or friends, decreased physical or cognitive functioning) and adapt their old life [
18,
19], but researchers have assumed that age-related losses can severely limit this adaptability in oldest-old age [
17,
20,
21]. Previous researches have found a downward trend in the SWB level with increasing age among older people [
22]. ‘A common stereotype about aging’ showed that a decrease in gains and an increase in losses, such as poor health condition and physical disability, which are more common in oldest-old adults, may lead to their worse SWB [
23,
24]. Previous studies have suggested that the predictors of SWB can differ across age groups [
17,
25‐
27]. A meta-analysis suggested that the association of physical and social resources with SWB was stronger in the oldest-old than in the young-old (aged 65–79) [
17,
26]. Subsequent studies pointed out that health factors played a particularly strong role in the young-old, while specific social factors became more important in the oldest-old [
27]. In the previous literature, the Georgia Centenarian Study found significant direct or indirect effects of physical health impairment, social resources, cognitive functioning, and education on positive aspects among oldest-old adults [
28]. Nakagawa et al. reported that cognitive function, hearing problems, and activities of daily living (ADL) were strong predictors of well-being in both Japanese and American centenarians [
17]. A Spanish study suggested that SWB among the oldest-old was mainly predicted by personality traits [
29]. In a Swedish study, social network quality, self-rated overall health, sense of controlling one’s life, and depression were significantly associated with life satisfaction [
30]. The limitations of these existing literature are that the sample size is too small (less than 350) and not representative enough.
There have been a few studies of the SWB among the oldest-old in developed countries, but the results should not be extended to developing countries or different cultures [
16]. In developed and developing economies, the relationships between the same factor and SWB may be inconsistent. For example, some researchers have reported the strongest correlations between economic status and SWB in poor developing countries but the weakest correlations in the wealthier developed countries [
31]. Diener pointed out the cultural differences in SWB that SWB was influenced by the environment and social circumstances [
32]. There are both universal and culture-specific predictors of well-being [
32]. One theory of culture and self has proposed that individuals learn cultural values through their cultural background and internalize them into their personal beliefs [
33]. Individuals in independent societies such as American and Western countries attach importance to self-esteem and personal feelings in judging SWB, whereas individuals in interdependent societies such as Asia and Eastern countries value relational harmony and social norms [
33,
34].
The Chinese Longitudinal Healthy Longevity Survey (CLHLS) is a dynamic, prospective, and national cohort of Chinese older people, which is the largest longitudinal study of the oldest-old in the world [
2,
16]. Some studies have obtained related results about factors of SWB using CLHLS data. Zhang et al. reported that better intergenerational relations promoted older people’s positive affect and reduced their negative affect [
35]. Chen et al. found that coresidence with spouse or children was associated with positive well-being [
36]. Brown et al. reported a strong negative relationship between religious participation and SWB [
37]. Li et al. observed that life satisfaction and affective aspect were both influenced by demographic variables and social supports [
16]. One limitation of this study is that the analysis of variance was used to examine the influences without adjusting for any confounders. To sum up, there is still a lack of systematic investigation of comprehensive contextual factors of SWB among Chinese oldest-old. Based on the bottom-up theories which explain and predict SWB by focusing on objective life circumstances [
38], the present study could capture contextual factors of SWB comprehensively.
To fill the gap in the existing research literature, we explore the sociodemographic, health-related, and social predictors of SWB among oldest-old people using a large and representative population from CLHLS conducted from 1998 to 2014. The prespecified hypothesis is that the oldest-old with better health conditions and social resources will have better SWB. We hope that our research findings will contribute to developing the improvement and promotion strategies of the oldest-old’s well-being.
Discussion
In this large prospective community-based cohort study in China, we found that age, gender, ethnic group, education, primary occupation before retirement, current marital status, and place of residence were sociodemographic predictors of SWB among the Chinese oldest-old. Then, the health-related predictors of SWB included self-rated health, visual function, hearing function, diet quality, smoking status, drinking status, and exercise status. Moreover, SWB was influenced by some social factors, such as the number of biological siblings, the number of children, leisure activities, financial independence, and access to adequate medical service. In particular, after adjustment for possible interactions in the multivariable-adjusted model, self-rated health, access to adequate medical services, exercise status, and place of residence still exert a stronger effect on SWB among the Chinese oldest-old than other factors.
Two variables deserve attention in our study, that is, age and gender. Their promotive and inhibitory effects on SWB were reversed in unadjusted analysis and multivariate analysis. Combined with the basic characteristics (Table
1), we think that the unadjusted effect is correct. Maybe because some of the predictors are moderators of the relationships between age or gender and SWB. Consistent with an American study [
24], our findings showed that relatively younger octogenarians had better mental health than centenarians. A longitudinal study demonstrated that well-being declined with impending death among the national samples from Germany, the United Kingdom, and America [
49]. But Italian centenarians reported greater satisfaction with life and less anxiety and depression than younger people [
50]. A general theoretical perspective revealed that the aging process, such as the declines in cognitive and physical functioning, reduced the capacity of individuals to adapt to stressful events [
24,
51]. Inconsistent with previous studies in developed countries [
17], we observed the gender difference in SWB among the oldest-old. Besides, there are significant gender differences in some predictors of SWB among the oldest-old. Compared to older women, older men had a higher proportion of being octogenarians and white-collar workers. Instead, older women were more likely to be illiterate and widowed. It is well known that women live longer than men. That may be why men are more likely to be octogenarian and women have lost their husbands. In traditional China, parents usually placed more emphasis on the education and development of their sons than their daughters [
43]. Our research is informative to gender-tailored interventions for a better SWB among the oldest-old.
Ethnic differences have been found in life satisfaction among older people from Nepal [
52], but not found in affect and loneliness among the oldest-old from American [
25]. In China, the special cultural traditions, different values, and even strong religious beliefs of ethnic minorities may cause differences in SWB between ethnic minorities and Han nationality. Ethnicity also affects other key predictors of SWB, such as education, occupation, and place of residence. Among older people from Vietnam and Israel [
53,
54], higher levels of education were also associated with a better SWB. But it was not found among the oldest-old from Japan, America, and Sweden [
17,
25,
30]. Empirical evidence shows that education is positively related to psychological resilience, which can play a role in buffering various stressors [
44,
55]. Primary occupation before retirement can reflect socioeconomic status to some extent, and it is also closely related to some predictive factors of SWB, such as financial independence, access to adequate medical service, and even diet quality. The accumulated savings, more pensions, and more social security can better support the life and psychological needs of older people. Compared to rural areas, the advantages of living in cities for better SWB are easy to explain. Residents in urban areas have easier access to better health care services and medical technologies [
56], as well as a broader social network, more social activities, and more community services [
43,
56]. Widowed has been proved to be a predictor of loneliness and depression among older people in China [
44,
57], as well as life satisfaction among the oldest-old in Sweden [
30]. The death of a spouse means the loss of some psychological support and companionship in daily life, instead of emotional sadness [
44]. Our research recommended that the mental health of the widowed oldest-old need to be highly valued in the social support system.
Based on previous studies that reported the strong associations between self-rated health and one or more aspects of SWB among both the young-old and the oldest-old in developed countries [
11,
30,
58], our study shows that self-rated health is a strong predictor of SWB among Chinese oldest-old. However, the association between any kind of self-reported diseases and SWB among the oldest-old was weakened and became nonsignificant. Self-rated health measures something different from physician’s ratings but depending upon one’s hypothesis, which is ‘subjective’ or ‘perceived’ as opposed to ‘objective’ or ‘actual’ [
59]. Self-rated health does not decline with increasing age, to the same extent as chronic diseases and disabilities increase [
59,
60]. Older people usually more positively assess their health than the middle-aged [
59,
61]. Published researches have illustrated that there is a causal relationship between self-rated health and a series of psychological factors, such as self-esteem [
58], depression [
58,
62], and loneliness [
43,
63]. The self-concept hypothesis provides evidence for the stability of self-rated health, and it reflects one’s established beliefs about their health [
64,
65]. Thus, our research advocates that social services should focus more on the oldest-old’s subjective perception of their health.
Similar to related studies in developed countries [
17,
66], our findings suggest that visual and hearing impairment has a significant negative impact on the SWB of the oldest-old. Visual and hearing disabilities will adversely affect the interpersonal communication and activity of the daily life of older people [
67,
68]. These disabilities can also cause older people to be fearful of or anxious about the unknown world that they can’t see or hear. Our research reminds us to pay more attention to the mental health of disabled older people. Similar to our research, the previous study among Canadian older men has observed that frequent consumption of vegetables and fruit is associated with greater life satisfaction [
69]. Nutrition-related health problems include frailty, depression, visual function, chronic non-communicable diseases, and so on [
70]. Proper nutrition is a modifiable factor that ultimately improves health, prevents functional disability, and promotes one’s well being [
69]. To some extent, smoking and drinking can indeed make people relaxed and release some psychological stress. However, the damage of smoking and drinking to physical functions, such as increased risks of lung cancer, atherosclerosis, and liver cirrhosis [
71,
72], makes it not recommended. Previous researches have indicated that physical activity is a significant and robust predictor of SWB among older people in developed countries [
10,
22,
73]. Our study shows that this prediction is still applicable for the oldest-old. Physical activity is linked with the release of emotion-related neurotransmitters, including norepinephrine, dopamine, serotonin, and endorphins [
10,
74]. Under the condition that the oldest-old physical function permits, close relatives and social workers should help them to participate in physical activities.
Siblings relationship is almost the longest lasting in a person’s lives, which plays an important role in the social and emotional support of older people [
75,
76]. There is robust evidence indicating that the sibling relationship is significantly related to depression, anxiety, loneliness, and life satisfaction among young-old adults both in developed and developing countries [
75‐
77]. As observed in a recent paper, children are related to better SWB and lack of depression among older people in European countries [
78]. The roles of children in providing instrumental, emotional, and economic support for older people make them indispensable in the later stages of one’s life. In agreement with published articles focusing on older people from Croatia, our findings showed that engagement in leisure activities contributed to better SWB [
79,
80]. Through participating in leisure activities, the oldest-old can meet life values and needs, build social relationships, feel positive emotions, and therefore enhance the well-being [
80].
One novel finding from our study is that the oldest-old with financial dependence more likely to face worse SWB than others with financial independence. Related research has revealed that receiving financial support from adult children considerably increased male older people’s negative aspects [
35]. It was guessed that financial independence probably means autonomy, which is directly related to their SWB. Besides, for the oldest-old, their children are mostly at or near retirement age, thus they can only get little financial support from their children. One finding which needs to be taken seriously is that access to adequate medical service makes a powerful impact on the SWB of the oldest-old. Older people in remote, poor, rural areas and those left behind cannot get adequate medical service. The social phenomenon of “the difficulty in seeing a doctor” among older people is prevalent in China. Because of travel inconvenience and medical procedures cumbersome, it is difficult for older people to timely and quick access to medical services. Medical treatment may become a major mental distress for older people. Because of this, relevant departments should actively provide convenient, low-cost, and effective medical and health services for older people, especially the oldest-old.
Given the differences compared with studies in other countries, the current study suggested that gender, ethnicity, and education may be culture-specific factors. These factors are endowed with unique cultural connotations in different countries due to history or tradition. This study found that self-rated health, access to adequate medical services, exercise status, and place of residence were more strongly associated with SWB among the Chinese oldest-old than other factors. From an American study, well-being among adults aged 75 and over was found to be most influenced by friendships, spouse, and financial security, while that among adults aged 65–74 was most influenced by satisfaction with spouse, friendships, and government services [
81]. Moreover, there may be consistency across age groups regarding the associations of self-rated health and sibling relationship with SWB. These comparisons may not be accurate because of inconsistencies in measurement methods and the lack of large samples of the oldest-old. Additional cross-country and cross-age researches are required to ascertain the country-specific and age-specific differences and mechanisms regarding the predictors of SWB.
There are several strengths to our study. First, to the best of our knowledge, the present study is the first to carry out a systematic investigation of comprehensive contextual factors of SWB among the oldest-old in China. Second, the current study included a large and representative sample of the Chinese oldest-old, allowing a robust assessment for the predictors of SWB among this group. Third, this article carried on a preliminary exploration of all available and possible sociodemographic, health-related, and social factors, which were obtained through face-to-face surveys. Fourth, our results supported that SWB was significantly correlated with self-rated life satisfaction, optimism, happiness, personal control, conscientiousness, anxiety, loneliness, and uselessness. It is reasonable to infer that the predictors of SWB may have a similar effect on these dimensions. Finally, our findings may also apply to other populations in transition, particularly in East Asia and Southeast Asia (e.g. Malaysia, Nepal), where contextual background and cultures are similar to those in China, such as the breakdown of traditional family structures and the imperfection of social security systems [
43].
This study has several limitations. First, those strong predictors of SWB among the oldest-old may interact with other variables, and we overestimate their effects on SWB by not adjusting the interactions. Continued efforts to explore these relevant interactions further will yield more robust and realistic predictive effects. Second, during the follow-up period from 1998 to 2014, great changes have taken place in China, such as the rise of the Internet and electronic information technology, the government’s greater concern for people’s livelihood, and the improvement of people’s quality of life. These changes will more or less have an impact on the content of this study. Subsequent studies suggest that study populations with a small-time span be selected. Third, in our study, SWB was not measured by other more valid and reliable scales, such as the Philadelphia Geriatric Center Morale Scale and the Well-Being subscale of the Differential Personality Questionnaire. The Cronbach’s α of our SWB scale was only 0.68, thus the reliability of the scale was defective. Fourth, our study used a single-factor structure of the SWB indicator, which varied from previous studies that distinguished life satisfaction and affective aspects [
16,
36]. The factor structure of SWB in oldest-old people should be further examined. Fifth, detection bias might exist because most information of variables came from the interviewee’s report and was not verified again. Finally, because of massive missing or unavailable data, the study did not explore other possible predictors, such as household income, community services, and personality traits. The effects of these factors can be discussed in future studies.
Conclusions
In this cohort study of over 30,000 oldest-old people in China, we observed that SWB in the oldest-old is influenced by a large number of complex sociodemographic, health-related, and social factors, including age, gender, ethnic group, education, primary occupation before retirement, current marital status, place of residence, self-rated health, visual function, hearing function, diet quality, smoking status, drinking status, exercise status, the number of biological siblings, the number of children, leisure activities, financial independence, and access to adequate medical service. Our research advocates that special attention should be paid to the mental health of centenarians, women, rural residents, widowed, physically disabled, and childless oldest-old people. Relevant agencies can improve physical activities, leisure activities, financial support, and medical services to promote the well-being of the oldest-old. Future study is expected to explore other social factors that can be intervened, such as community services and social support.