Background

Population ageing has been an ongoing trend globally in recent years. By the year 2015, there were more than 826 million people over 60 years of age in the world—a number projected to reach nearly 1.4 billion by 2030 [1,2,3]. The pace of population ageing is much rapid in China than many other high-income or low-and middle-income countries [4]. According to China’s sixth National Population Census, 13.3% of the nation’s entire population was aged ≥60 years in 2010 [5], and among these, 12.3% were aged ≥80 years [6]. In the next 25 years, the percentage of people in China aged 60 years or over is expected to be more than double, reach 28% (402 million) in 2040. In contrast, for the proportion of the population aged over 60 years to double from 7% to 14% took France, Sweden and the US 115 years, 85 years and 69 years respectively [4].

Shanghai is known for its dynamic economy and prolific resources. In 2016, its GDP was ¥2746.6 billion, an increment of ¥250.1 billion of the previous year, ranked No.1 among all cities of China. The annual GDP increase in Shanghai was 10.02%, while Beijing was 8.41% [7]. Shanghai was also the first city to enter ageing society in China. The average life expectancy in Shanghai reached 82.75 years of age by the end of 2015 (male 80.47, female 85.09). In the same year, 4.36 million (30.2%) of the city’s registered residents were aged ≥60 years, the highest proportion of older residents among China’s regional jurisdictions. Among them, 780, 500 were aged ≥80 years [8]. It is estimated that the proportion of elders among Shanghai residents could exceed 34% by the year 2020 [9], and 44.5% by the year 2050, which is higher than the estimated proportion of elders in Japan [10].

Population aging brings about major challenges to health and social care. Learn the health status of the elderly is vital for informing the development of disease prevention and health promotion programs, and healthcare and social services delivery that can enhance their life satisfaction and quality of life [11]. Self-rated health (SRH) is a useful assessment of health derived from asking individuals to rate their own general health on a five-point scale. SRH is considered by the World Health Organization to be an important indicator of population health and healthy life expectancy [12]. Numerous studies show that the age, gender, income, education [13], life style, psychological well-being [14], social support such as children’s accompany and emotional support [15], chronic diseases, physical and instrumental functioning [16, 17] are associated with self rate health [18].

According to health screening data, only about 20%–25% of elderly people have no vital diseases [19]. After over three decades of “one-child” policy, a period in which many young adults have sought opportunities far from their hometowns, the traditional intergenerational relationship between the elderly and their children has changed, as has the traditional family structure [20]. Thus, it is important to find ways in which society can meet the increasing needs of the elders. Employing SRH measures to investigate Shanghai elders’ view on their own health status will help to explore what are the determinants of SRH among Shanghai elders as well.

Objectives

1) to assess the SRH of community-dwelling elders in Shanghai; 2) to examine the effects on individual older persons’ SRH from demographic characteristics, health risk behaviours, physical/mental health condition, social support, and neighbourhood socioeconomic status, etc.; 3) to determine the main determinants affecting Shanghai elders’ SRH; and 4) to explore the subsequent public health and policy implications. This study will also extend the literature on SRH, particularly with regard to the ageing population in Shanghai.

Methods

Sampling

This study conducted a questionnaire survey that employed a multistage stratified cluster random sampling method. The sample was selected from Shanghai residents who were ≥60 years of age. In China, people who have reached 60 years of age are classified as elders [4]. Based on hypothesis testing, and considering the types I (0.05) and II (0.1) statistical error, the appropriate sample size for this project was considered to be 1756. This study obtained 2001 valid survey responses.

The multistage stratified cluster random sampling process included the following:

First, all of Shanghai’s 18 districts were divided into three groups according to their economic status – high, medium and low. Then, two districts each were randomly selected from the three groups, namely: Pudong and Changning (high), Hongkou and Putuo (medium), and Jinshan and Chongming (low). Based on the sample districts’ population size, 832, 199, 291, 274, 157, and 248 elders should be selected from the six districts respectively. Second, all neighbourhood communities in the six districts were ranked based on economic status (high, medium, or low), and one medium-level community was selected from each. Third, the selected communities provided the lists of their elder residents’ names, and participants were selected randomly from the lists based on the required sample size of the study.

Data collection and ethics considerations

This study’s ethics review application was conducted and approved by the Research Ethics Board of Fudan University. The project team conducted questionnaire survey from June to August of 2011. All participants gave written informed consent before the survey took place. The investigators filled out the questionnaires during the survey on behalf of the respondents. The respondents were given the option of leaving the study at any time with no question asked; but none did.

Measurement

Dependent variable

SRH among Shanghai elders is the dependent variable in this study. SRH was assessed by asking the question: “How do you rate your health: very good, fairly good, average, fairly poor, or poor?” For cross-tabulations, SRH was categorized as good (very good and fairly good), average, and poor (fairly poor and poor); and for logistic regression analyses, it was categorized as good (very good and fairly good), and not good (average, fairly poor and poor).

Independent variables

There are 57 aspects in five sections in the questionnaire, and details are as follows:

Characteristics: residence location (urban, suburbs or outer suburbs); gender (male or female); age group (60–64, 65–69, 70–74, 75–79, and ≥80 years), and educational attainment (illiterate, primary, middle/high school, college or above).

Health risk behaviour: cigarette smoking (never, occasionally, often, has quit); alcohol drinking (never, occasionally, often, has quit); outdoor activities (yes, no).

Social support: marital status (widowed, married, single/divorced); source of income (pension, salary/savings, family allowance, other); living situation (non-empty nester, empty nester); meeting children (daily, weekly, monthly, yearly, <1 time/year); meeting neighbours (weekly, monthly, yearly, never); meeting relatives/friends (weekly, monthly, yearly, never); community activities (weekly, monthly, yearly, never).

Mental health status: feelings of loneliness (never, sometimes, always); nervousness/anxiety (never, sometimes, always); satisfaction with life (good, fair, poor).

Physical health status: changes in physical health status (better, unchanged, worse, variable); hearing (deaf, some hearing, able); ability to speak (yes, no); vision (blind, some vision, sighted). Furthermore, the Barthel index of activities of daily living (ADLs) was adopted for assessing the functional status of the respondents in the study. ADLs scores were classified as complete, partial, or almost no ability to perform activities.

The index consists of six measures: bathing and showering, dressing, self-feeding, functional mobility, personal hygiene and grooming, and toilet hygiene. The respondents were asked their degree of dependency in eating, personal hygiene, bathing/showering, dressing/undressing, walking, use of stairs, and toilet hygiene [21].

Chronic disease: Chronic diseases in the study were those diagnosed by physicians or symptoms described by the respondents, including hypertension, heart disease, diabetes, cataract, cerebrovascular disease, bronchitis, gastroenteritis, intervertebral disc disease, cardiovascular disease, and asthma. The survey responses were as follows: chronic disease (yes, no); hypertension (yes, no); heart disease (yes, no); diabetes (yes, no); cataracts (yes, no); cerebrovascular disease (yes, no); bronchitis (yes, no); gastroenteritis (yes, no); intervertebral disc disease (yes, no) cardiovascular disease (yes, no); asthma (yes, no); total number of chronic diseases (0, 1, 2, or ≥3).

Statistical analysis

The SPSS statistical package (version 16.0 for Windows) was used to analyse the survey data. Mean S.D. was used for descriptive statistics. The chi-squared test was used to determine the differences in socio-demographic characteristics among different groups in the data analysis. Significance was defined as p < 0.05. The relationship between SRH score and respondents’ socio-demographic characteristics, health risk behaviours, social support, mental health status and physical health status were tested using the chi-squared test. A series of logistic regression models were used to determine the independent association between SRH and its influencing factors.

Results

This study included 2001 respondents from different residential communities of Shanghai (Table 1). The mean age was 71.6 years. 324 of them or 16.2% of the total in the sample were aged ≥80 years. The gender ratio was 2:3 (841 male to 1153 female). The ratio of dwelling locations was 2:2:1 for downtown, suburbs and outer suburbs respectively. 39.4% of them received high school or higher education, and 28.7% were illiterate (see Table 2).

Table 1 Selected communities and characteristics of respondents
Full size table
Table 2 Socio-demographic characteristics and health risk behaviours of respondents
Full size table

Health risk behaviours have a positive association on SRH

Most of the respondents were non-smokers (77.9%) and non-drinkers (78.4%); and more than a half (58.4%) took part in regular outdoor physical exercise. There was a linear relationship between age and SRH. Table 2 shows that younger age (χ2 = 27.5, p < 0.05), male sex (χ2 = 11.5, p < 0.1), and living in suburban areas (χ2 = 55.1, p < 0.05) were associated with better SRH scores.

Another group with better SRH scores was those who smoke cigarettes (χ2 = 18.0, p < 0.1) or drink alcohol (χ2 = 18.6, p < 0.1). These scores were better regardless of whether they smoked/drank occasionally or regularly, or frequency of participation in outdoor activities (χ2 = 69.3, p < 0.05). Only 10% of regular cigarette smokers, 3.7% of regular alcohol drinkers, and 9.3% of physically active respondents rated their health status as “poor”.

The stronger the social support, the higher the SRH score

The association between social support and mental condition and SRH scores was cross-tabulated and tested using the chi-squared test. The results show that the majority (75.9%) of the respondents were married, and 21.6% were widowed. The main source of income for 86.1% of respondents was a government pension. More than three-fifths of them (65.3%) were empty nesters. A similar proportion, 65.2%, reported meeting with their children daily and nine out of ten interacted with their neighbours weekly, and 22.4% met with their relatives and friends on a weekly basis. In terms of mental health, 83% of the respondents reported that they never felt lonely and 89.2% of them had never felt nervous or anxious. Only 5.1% reported that they were not very satisfied with their lives.

As shown in Table 3, respondents with better social support tended to have better SRH scores. For example, those who received a pension (χ2 = 28.8, p < 0.05) and met more frequently with their children (χ2 = 21.1, p < 0.1), neighbours (χ2 = 25.3, p < 0.05), and relatives and friends (χ2 = 41.6, p < 0.05) tended to have significantly better SRH scores.

Table 3 Social support and mental health status of respondents
Full size table

Better physical health is positively associated with better SRH scores

Most (98.1%) of the respondents in this study were able to live independently. Over the past year, 64% reported had no change in physical condition and less than 29.2% of them felt some decline of health in the past year. One fifth noticed a decline in their hearing ability and 5% were unable to hear. 6.5% of the respondents had difficulty in speaking and 28.5% of them had vision issues (see Table 4).

Table 4 Physical health condition of the respondents
Full size table

The overall prevalence of chronic diseases among the respondents was quite high (77.2%), however. The 10 most commonly reported chronic diseases were: hypertension (51.1%), heart disease (21.8%), diabetes (14.5%), cataracts (8.7%), cerebrovascular disease (7.5%), bronchitis (6.0%), gastroenteritis (5.7%), intervertebral disc disease (5.6%), cardiovascular disease (2.9%), and asthma (2.6%). The percentage of respondents reporting having three or more of the diseases was 22.7%, and a similar percentage have two diseases.

Chi-square analysis found that the respondents’ ability in hearing (χ2 = 38.7, p < 0.05), speaking (χ2 = 16.1, p < 0.05) or vision (χ2 = 78.3, p < 0.05) affects their SRH negatively. Chronic illnesses (except asthma) also impact SRH significantly; and the degree of influence on SRH was associated with the number of reported diseases: more diseases also resulted in a poorer SRH score.

Models’ testing on the determinants of SRH

Table 5 reports a series of logistic regression models of predicted odds ratios of SRH.

Table 5 Multiple Logistic Regression of Factors Influencing SRH of Respondents
Full size table

Model 1 shows that the location of residence is a significant determinant of the elders’ SRH. Respondents who are living in outer suburbs are more likely to have better SRH score. Health risk behaviours such as smoking, drinking and outdoor exercise were significantly and positively associated with their SRH. Those who had quit smoking were 2.3 times as likely to have poor SRH than the ones who had never smoked.

Model 2 indicates that respondents who enjoy better social support are more likely to have higher SRH score. Meeting with family, relatives and friends more often and taking part in more community activities was associated with higher SRH score. Those who receiving a pension are 2.2 times as likely to have higher SRH score than those living on savings or allowances from their children. However, marital status and living arrangements have no significant association with SRH.

Model 3 shows that the elders who are satisfied with their lives are twice as likely to have higher SRH scores. Meanwhile, feelings of loneliness are 3.2 times as likely to have low SRH, and those who reported feeling nervous or anxious are 6.2 times as likely to have low SRH score.

Model 4 shows that ADLs disability and using assistive tools in daily lives impact SRH score. The respondents who have physical functional challenges and required assistance in their daily lives are 16.1 times as likely to have low SRH score than those who live independently and without physical disability.

Model 5 includes the variables of 10 chronic diseases. It shows that these diseases are all significantly associated with SRH. Those with one or more chronic diseases are more likely to have lower SRH score.

Model 6 includes all of the variables in the survey. Overall, the respondents with better SRH scores are 1.7 times as likely to be living in the outer suburbs, 2.2 times as likely to be non-smokers, 1.4 times as likely to have had regular outdoor exercises, and those who meet with family, relatives or friends more often. While respondents with lower SRH scores are 1.9 times as likely to feel nervous or anxious, 4.4 times as likely to be dissatisfied with their lives, and 11.1 times as likely to be ADLs disabled and/or have been diagnosed with chronic illnesses such as hypertension, cardiovascular disease, diabetes, cerebrovascular disease and gastroenteritis.

The results of multiple logistic regressions show that socio-demographic and health risk behaviours only made up 6.6% of the total contribution to SRH. When combined with mental health variables (Model 3), it increased to 15.1%. Adding chronic diseases (Model 5), it became 19.6%. And all socio-demographic and health risk variables (Model 6) contributed 20.6% to SRH.

The models above show that among various determinants of SRH of the Shanghai elders, 6.6% is from socio-demographic and health risk behaviours, 2.4% is from social support, 8.5% is from mental health, 20.0% is from physical conditions, and 13.0% from chronic diseases.

Discussion

SRH was initially used by Mossey and Shapiro for predicting 7-year survival odds; and it became widely used as an indicator of population health afterwards [22]. According to Jylha, SRH is an individual and subjective conception that constitutes a crossroad between the social world and psychological experience on one hand, and the biological world on the other [23]. Scholars in different nations investigated the correlates of SRH based on Jylha’s theory; and French and Browning concluded that correlates of SRH are similar across continents [24]. This study examined the determinants of SRH among elders in Shanghai, which provides new evidence on whether the correlates of SRH similarities are greater than the differences between regions.

Multiple cross-sectional studies have been done previously on associations between SRH and health status as well as other factors. Most of these studies found that socio-demographic characteristics such as age, gender, education level, location of residency are related to SRH [25, 26]. Lower educational attainment is associated with lower SRH score [27], physical inactivity is associated with a decline in SRH score [28], and smoking is associated with lower SRH score. Modest alcohol intake and more social support are associated with higher SRH score [29,30,31]; while deterioration in physical [32, 33] or mental [32] health would significantly lower one’s SRH score.

In our examination of the relationships between SRH and socio-demographic factors, health risk behaviours, physical and mental health condition, and social support, nearly 40% of the respondents reported good SRH. This ratio is very close to the findings from some developed countries (i.e. 39.0% in Finland [34] and 39.6% in Spain [35]), and is considerably higher than similar studies’ findings from other parts of China (36.4%; 35.3%) [36, 37]. Shanghai’s relatively better socio-economic conditions and healthcare access may have contributed to the elders’ better SRH outcome [38].

Examining the extent of various determinants and their association with SRH can help to assess their relative importance in people’s lives, and develop more effective strategies to improve population health. Consistent with prior studies’ findings, multiple logistic regression analysis in this study indicated that physical health matters more than socio-demographic and risk health behaviours, which means that the SRH scores of the respondents, to a great degree, was affected by their physical health, which reflect their true health conditions. [39, 40].

Both single factor analysis and multiple regression analysis show that those respondents who lived in outer suburbs have higher SRH scores, while those who live in downtown have the poorest SRH. Meanwhile, the elders who had quit smoking, but not those who are still smoking, are more likely to have lower SRH scores. It may be assumed that those who quit smoking was due to declining of physical health. These new findings, as well as those of previous studies, indicate that health risk behaviour such as drinking and smoking have positive effects on SRH. Respondents who participated in outdoor activities regularly are also more likely to have higher SRH scores, which means that increasing the elders’ physical activities participation improves their SRH [41].

Further analysis found that social support, health risk behaviours, health status and neighbourhood economic status all impact on Shanghai elders’ SRH.

Social support

Respondents’ social support situation has a direct impact on their SRH. Generally, those with stronger social support networks tend to have better SRH scores. This demonstrates that being involved in more social function (such as interpersonal communication and emotional attachment with others) results in a better SRH score. It also suggests that SRH may be closely tied with mental health and overall wellbeing. These findings are also quite consistent with some previous studies [42, 43].

However, financial support from family members did not seem to have positive association with SRH. This finding is very different from previous studies’ findings, which found that financial support from family members tended to have a positive association with SRH [44]. While this new finding deserving further investigation, some assumptions could be made as spring board for such new studies: 1) they did not need financial support, 2) they did not like to be a burden of their respective families, or 3) they value emotional support more than financial assistance. Shanghai is a relatively affluent city and its residents are relatively well-off. According to 2016 Pension policy launched by Social Security Bureau in Shanghai, people over 60 years old can get at least ¥750 per month as Supplemental Security Income (SSI) [45]. Considering this generation of elders had experienced full employment period of China, it is likely that most of them are entitled to employment pension, which would be much more than the SSI. Therefore, money from children matters less. It also bears mentioning that, in Chinese cultural tradition, children are obligated to give their parents money and gifts as a way of showing respect. Therefore, money from family members alone may not make a difference in an elder’s SRH.

Neighbourhood economic status

Neighbourhood economic status appears to be an important determinant of SRH. Respondents in wealthier neighbourhoods tend to have better SRH, while those in poorer neighbourhoods tend to have poorer SRH. Respondents from neighbourhoods with high economic status tend to meet with their adult children more often (72.3% meet with their children everyday vs. 56.5% and 58.6% in communities with medium and low economic status, respectively). They are also contacting with their neighbours more often (93.5% reported weekly contact vs. 84.6% and 87.9% in communities with medium and low economic status, respectively). Residents of wealthier communities attend more gatherings with their relatives and friends (32.7% do so weekly vs. 12.4% and 10.1% in communities with medium and low economic status, respectively). Meanwhile, only 4.4% of the respondents in wealthier communities responded that they have negative feelings about their current lives vs. 4.8% and 7.4% in communities with medium and low economic status, respectively. Finally, this group also has better SRH—only 9% of them rated “poor” as their health status vs. 15.4% and 17.4% from communities with medium and low economic status, respectively.

Undoubtedly respondents from wealthier neighbourhoods enjoy better living environment, which ensures the quality of life of the residents. A better living environment in turn enables them to host family, relatives and friends; and these interactions have positive association with their health and social life. Since elders living in wealthier communities tend to be better educated, they are likely to have had better income jobs and have better pension; thus, they tend to be more optimistic about their lives as well as their overall wellbeing, and these positive attitudes influenced their SRH outcomes. This finding is not surprising because health, and in particular SRH, is strongly associated with socio-economic conditions one is in. Similarly, prior studies have explored the link between neighborhood and health. They all found neighborhood is an important element of living condition, which may influence health in two ways: first, through health behaviors, attitudes and healthcare utilization, and influencing health status; second, through environmental quality and community resources. Thereby, future policymaking may focus on community environment for improve the elders’ heath status [46,47,48].

Physical and mental health status

With regard to physical health, the overall prevalence of chronic disease among respondents was nearly 80%, which is 20% higher than the rate reported in the Fourth Household Health Survey in China. The five most commonly reported chronic diseases in this study were hypertension, heart disease, diabetes, cataracts and cerebrovascular disease. All of these diseases are a serious issue for public health in China. Chinese scholars have identified hypertension and coronary heart disease as main threats to the health of the elders [49].

In this study, only 1.9% respondents had poor ADLs, which is a much lower rate than that of the elders in other parts of China [50]. However, consistent with previous studies, both uninvariate and multivariate analyses show that respondents with lower ADLs scores tended to have lower SRH scores [51]. Chronic disease is a key factor leading ADLs dependence, and thus, chronic disease prevention and control are crucial for improving the elders’ health status [52].

Multiple regression analysis found limited influence of demographic and health risk behaviour variables on SRH. The most influential variables are mental and physical health. Respondents who reported feelings of loneliness tended to be widowed women over 80 years of age, with low educational attainment, and no outdoor activities. These respondents rarely meet with their children (less than once a year), never take part in community activities, feel nervous constantly, have lower ADLs scores, and tend to have low SRH scores.

This study’s findings are consistent with strong epidemiological and physiological evidences in the literature that social isolation is a particularly important risk factor affecting elders’ SRH. Higher levels of depression among elders is associated to poorer health-related quality of life [53], and loneliness have a significant impact on physical health [54, 55].

Despite the elders’ relatively poorer health condition, this study shows that most of elders residents in Shanghai are satisfied with their lives.

Limitations

This study should be interpreted in light of the following limitations. First, this cross-sectional study established correlations but not causal relations among social support, health condition and SRH. We anticipate future research in the area to further elucidate these relationships. Second, there was no meaningful gender analysis conducted due to the gender imbalance in survey respondents.

Conclusions

This is the first study on the SRH of Shanghai elders and its determinants such as socio-demographics, lifestyle, neighbourhood economic condition and health status. The main findings of this study include: nearly two-fifths of the respondents reported their own health as good; respondents from wealthier neighbourhoods tend to enjoy more social support, are more optimistic about their lives, and their SRH scores tend to be higher; respondents who receive less social support and have limited contacts with their children tend to feel lonely, and tend to have lower SRH scores; respondents with low ADLs scores tend to also have low SRH scores, as well as those with mental health concerns. Unlike in previous studies, financial support from family members did not translate into higher SRH score in Shanghai, which we tribute to Shanghai’s social security system and the Chinese cultural tradition that support parents is children’s obligation.

These findings illustrate the main determinants of SRH, and highlight the important elements that keeping elders healthy—both physically and mentally. This study sheds light on the general condition of the older population of Shanghai as reflected by their SRH. Our findings will also be valuable to policymakers in planning and allocating appropriate resources for the wellbeing of Shanghai’s ever-growing older population, and they will also help other scholars to further explore relevant social issues and help practitioners develop effective intervention strategies. Further studies adding to these findings are anticipated.