Individual and province inequalities in health among older people in China: evidence and policy implications

Transcription

1 Individual and province inequalities in health among older people in China: evidence and policy implications Prof Maria Evandrou, Prof Jane Falkingham, Dr Zhixin Frank Feng, Dr Athina Vlachantoni Centre for Research on Ageing & ESRC Centre for Population Change, University of Southampton 1

2 Outline Background Aims and Research Question Conceptual Framework Data and Methods Results Summary Conclusions & Policy Implications

3 Background China Ageing population 2009 (167 million, 12%) 2030 (16%) 2050 (23%) Life expectancy: 67 years (1982) 71 years (2000) 85 years (2050) A developing country with tremendous economic development in the last 30 years. However, economic success has come hand-in-hand with the ever increasing social and economic disparities (individual economic status/province), especially between urban and rural, coastal and inland areas (provinces) Individual and province inequalities older people health outcomes

4 Aims and research question Aims To investigate the health outcomes of older people in China To examine variation in health outcomes according to individual and province level characteristics Research question How does the health of older people vary according to demographic characteristics, socio-economic indicators, health risk behaviours, household/family factors and provincial level factors? 4

5 Conceptual framework The determinants of health outcomes Demographic and Socio economic factors Health behaviour Health outcomes: Difficulty with ADLs Difficulty with IADLs Self reported health Life satisfaction Self reported memory Household/family factors Social security factors Provinces

6 Data Micro (individual level): The China Health and Retirement Longitudinal Study (CHARLS) (2011) biennial survey in China, conducted by the National School of Development (China Centre for Economic Research) at Peking University. Sample size: the national survey sample includes persons aged 45 years and older from 28 provinces. Analytical sample: persons aged 50 and over (N= 10,717) Macro (province level): Chinese Statistics Yearbook;

7 Data: CHARLS Survey areas

8 Map1: Percentage of Chinese population aged 65 and over, by province 8

9 Data: Health outcomes, CHARLS Table 1: Health of persons aged 50 and over, China, 2011 Response Difficulty with ADLs No difficulty (82.1%); Reports difficulty (17.9%) Difficulty with IADLs No difficulty (77.9%); Reports difficulty (22.1%) Self-reported health Positive (22.5%); Fair (47.1%); Negative (30.4%) Life Satisfaction Positive (22.9%); Fair (62%); Negative (15.2%) Self-reported memory Positive (19.1%); Fair (45.7%); Negative (35.2%) 9

10 Data: individual characteristics, CHARLS Individual characteristics, n=10,717 Age Ranging between 50-96, mean=62 Gender Male (50%); Female (50%) Marital status Married (86.1%); Other (separated, divorced and never married) (2.2%); Widowed (11.7%) Residence Urban (40.6%); Rural (59.4%) Highest educational High school and above (12%); Elementary and qualification middle (39.5%); Below primary (19.5%); Illiterate (29%) Income sources Wage (10.9%); Wage and pension (2.1%); Wage and others* (0.6%); Pension (22.6%); others* (4.6%); No income sources (59.3%) Medical insurance Yes (94%); No (6%) Smoking status Has never smoked (58%); Current smoker (31.8%); Ex-smoker (10.2%) 10

11 Data: Family/household characteristics, CHARLS Family/ household characteristics, n=10,717 Self-rated standard of family living Positive (3.2%); Fair (53.6%); Negative (43.2%) Receipt of Dibao by the household No (89.4%); Yes (10.6%) 11

12 Data: Macro level, province information Indicator Meaning Urban Percentage of urban population (based on household register) (%) GDPpc GDP per capita (1,000 Yuan) Migrants Percentage of migrants* (%) Income Mean income (1,000 Yuan) VATpc Value added tax per capita (1,000 Yuan) GovExpen Government expenditure per capita (1,000 Yuan) Fixedassets Fixed-asset investment per capita (1,000 Yuan) Stateown Share of state-owned enterprises in fixed-asset investment (%) ForInvest Share of foreign direct investment in fixed-asset investment (%) StateCom Percentage of state owned employees in the total of employees (%) PrivateCom Percentage of private owned employees in the total of employees (%) FandHCom Percentage of foreign and overseas Chinese employees in the total of employees (%) SocSec Percentage of expenditure for social safety net and employment incentives in total revenue of province (%) MedIns Percentage of expenditure for medical and health care in total revenue of province (%) Hosp Doctor Bed UrbEng RurEng Number of health care institutions per 10,000 population Number of medical technical personal in health care institutions per 1,000 population Number of beds in health care institutions per 1,000 population Urban Engel's coefficient (%) (% of food expenditure out of income consumption expenditure) Rural Engel's coefficient (%) (% of food expenditure out of income consumption expenditure)

13 Methods 1. Factor analysis for the development of derived indices for province characteristics 2. Hierarchical analysis with multilevel logistic model to investigate the health outcomes of older people (50 plus) 2.1 Binomial multilevel regression 2.2 Multinomial multilevel regression 13

14 Results I Table 2: Factor loading of province characteristics Factor1 (D&M) Factor2 (GLQ) Factor3 (SSI) Communality Urban GDPpc Migrants Income VATpc GovExpen Fixedassest Stateown ForInvest StateCom PrivateCom FandHCom SocSec MedIns Hosp Doctor Bed UrbEng RurEng %Var

15 Results I Regional distribution of factor scores of provinces Factor 1: D&M (Developed, marketized and lower level of spending on health care and lower provision of health facilities) Factor 2: HLQ (Higher level of health facilities and quality of life) 15

16 Results I Regional distribution of factor scores of provinces Factor 3: SSI (Strong state influence and higher social security expenditure) 16

17 Results II Table 3: Binomial multilevel regression estimates for reporting difficulty with Activities of Daily Living (ADLs), aged 50+ Reporting difficulty with ADLs Individual characteristics Province characteristics ORs ORs Fixed Part CONS 0.03*** 0.04*** Age 1.05*** 1.05*** Female (ref: Male) 1.35*** 1.34*** Marital status (ref: Married) Others Widowed Rural (ref: Urban) 1.44*** 1.43*** Education (ref: High school and above) Illiterate 1.82*** 1.82*** Below Primary 1.91*** 1.92*** Elementary and middle 1.44*** 1.44*** Income sources (ref: Wage) Wage and pension Wage and other 1.87* 1.83 Pension 1.45*** 1.45*** Others 2.31*** 2.29*** No 1.69*** 1.68*** *** p<0.01, **p<0.05, * p<0.1 17

18 Results II Table 3 (Continue) Binomial multilevel regression estimates for reporting difficulty with Activities of Daily Living (ADLs), aged 50+ Reporting difficulty with ADLs Individual characteristics Province characteristics ORs ORs Fixed Part Have not medical insurance 0.89 (ref: Yes) 0.88 Smoke (ref: Never) Current smoking Smoked but quit 1.45*** 1.44*** Self-rated living standard (ref: Positive) Fair Negative 1.71*** 1.71*** Dibao household (ref: No) 1.80*** 1.80*** (Factor Scores) Developed, marketized and lower health care 0.82* Higher health facilities and quality of life 0.91 Strong state influence and higher social security expenditure 0.85 *** p<0.01, **p<0.05, * p<0.1 Random Part Level 2: Province 0.23(0.09) 0.22(0.08) 18

19 Results II Table 4: Summary of significant effects of predictors on the health outcomes ( = significant at p<0.1) Difficulty with ADLs Difficulty with IADLs Selfreported health Life Satisfaction Selfreported memory Individual characteristics Age - - Gender Marital status Residence - ( ) Highest educ. qualification - Income sources - ( ) Medical Insurance Smoking status ( ) Self-rated living standards - - Hhold receipt of Dibao - - Province characteristics (Factor Scores) Difficulty with Mobility* D&M HLQ SSI

20 Results III Figure1: Differential relative odds of reporting difficulty with ADLs for provinces, derived from individual characteristics compared to the national average (1) 20

21 Results III Figure 2: Differential relative odds of reporting difficulty with IADLs for provinces derived from individual characteristics compared to the national average (1) 21

22 Results III Figure 3: Differential relative odds of reporting fair SRH (A) and negative SRH (B) for provinces derived from individual characteristics compared to the national average (1) 22

23 Results III Figure 4: Differential relative odds of reporting fair life satisfaction (A) and negative life satisfaction (B) for provinces derived from individual characteristics compared to the national average (1) 23

24 Results III Figure 5: Differential relative odds of reporting fair memory (A) and negative Memory (B) for provinces derived from individual characteristics compared to the national average (1) 24

25 Summary: Factors affecting health amongst older people (50+) in China, 2011 (a)individual and family/household level Being female, living in rural areas, having low educational qualifications, having no income sources or receiving income from other sources, being an ex-smoker, reporting relatively low living standards and living in a household which receives the Dibao, are characteristics associated with: reporting of difficulty with ADLs or IADLs, fair/poor SRH, fair/poor life satisfaction and fair/poor memory. 25

26 Summary: Factors affecting health amongst older people (50+) in China, 2011 (b) Province level Individuals living in provinces with better health facilities and better quality of life are less likely to report difficulty with ADLs or poor life satisfaction. 26

27 Conclusion I: Implications for Policy Both individual and province level factors matter Improving the environment in which older people live, both their household circumstances and the area in which they live can impact positively on health Lower socio-economic status (individual or household indicators) is clearly and strongly associated with the report of poorer health status. Improving the socio-economic status of older people and their families can improve health outcomes Health-risk behaviour (smoking status) is associated with poorer health status. Public health campaigns that reduce smoking may improve health outcomes both amongst today s elders and future cohorts 27

28 Conclusion II: Implications for Policy There are marked differences between provinces in a range of indicators, such as modernisation and the amount of expenditure on health services Improving the level of spending on health care and the provision of health care facilities at the province level can improve health outcomes at the individual level, which has implication of the redistribution of resources between richer and poorer provinces. Statutory assistance in the form of cash benefits and health services for individuals at the lowest part of the income distribution may also add to a comprehensive set of social policies protecting the most vulnerable groups in society. Health outcomes are poorer in rural areas than in urban areas Policies need to target resources to rural areas to ensure rural older people are not disadvantaged compared to their urban counter parts. 28

29 References Cai, F. et al (2012) The Elderly and Old Age Support in Rural China: Challenges and Prospects. Washington DC: The World Bank. Feng, Z., Wang, W., Jones, K., & Li, Y. (2012). An exploratory multilevel analysis of income, income inequality and self-rated health of the elderly in China. Social Science & Medicine, 75 (12): Feng, Z, Wang, W, & Jones, K.(2013) A multilevel analysis of the role of the family and the state in self-rated health of elderly Chinese. Health & Place, 23, Goldstein H, Carpenter J, Kenward M G, Levin K A, 2009, "Multilevel models with multivariate mixed response types" Statistical Modelling Hay,D. I. (1988). Socioeconomic status and health status: A study of males in the Canada Health Survey. Social Science and Medicine, 27, House, J. S., Kessler, R. C., Herzog, A. R., Kinney, A. M., Mero, R. P.,&Breslow, M. F. (1990). Age, socioeconomic status, and health. Milbank Quarterly, 68, Hox, J. (2002). Multilevel analysis : techniques and applications. London: Psychology Press. Jarvis, M. and Wardle, J. (1999) Social patterning of individual health behaviours: the case of cigarette smoking. In: Marmot M, Wilkinson R, eds. Social determinants of health. Oxford: Oxford University Press.

30 30

31 Additional slides 31

32 Results Table 5: Binomial multilevel regression estimates for reporting difficulty with IADLs Reporting difficulty with IADLs Individual characteristics Province characteristic ORs ORs Fixed Part CONS 0.29*** 0.29*** Age 1 1 Female (ref: Male) Marital status (ref: Married) Others Widowed Rural (ref: Urban) Education (ref: High school and above) Illiterate 0.86* 0.85* Below Primary Elementary and middle Income sources (ref: Wage) Wage and pension Wage and other Pension Others No *** p<0.01, **p<0.05, * p<0.1 32

33 Results (Continued) Table 5: Binomial multilevel regression estimates for reporting difficulty with IADLs Fixed Part Have not medical insurance (ref: Yes) Reporting difficulty with IADLs Individual characteristics Province characteristic ORs ORs 1.37*** 1.38*** Smoke (ref: Never) Current smoking Smoked but quit 0.84* 0.84* Self-rated living standard (ref: Positive) Fair Negative Dibao household (ref: No) (Factor Scores) Developed, marketized and lower health care 1.1 Higher health facilities and quality of life 1 Strong state influence and higher social security expenditure 1.02 Random Part Level 2: Province 0.07(0.03) 0.07(0.04) *** p<0.01, **p<0.05, * p<0.1 33

34 Results: Table 6 Multinomial multilevel regression estimates for reporting fair or negative self-rated health *** p<0.01,**p<0.05,* p<0.1 Reporting fair or negative self-reported health Individual characteristics Province characteristics Fair Negative Fair Negative ORs ORs ORs ORs Fixed Part CONS *** *** Age *** *** Female (ref: Male) 1.17** 1.54*** 1.18** 1.54*** Marital status (ref: Married) Others Widowed *** *** Rural (ref: Urban) *** *** Education (ref: High school and above) Illiterate *** *** Below Primary 1.36*** 1.97*** 1.36*** 1.98*** Elementary and 1.23** 1.42*** 1.23** 1.43*** middle Income sources (ref: Wage) Wage and pension Wage and other Pension 1.30*** 1.94*** 1.30*** 1.92*** Others *** *** No 1.24** 2.06*** 1.24*** 2.05*** 34

35 Results: Table 6 (Continued) Multinomial multilevel regression estimates for reporting fair or negative self-rated health Reporting fair or negative self-reported health Individual characteristics Province characteristics Fair Negative Fair Negative ORs ORs ORs ORs Fixed Part Have not medical insurance (ref: Yes) *** p<0.01,**p<0.05,* p< *** *** Smoke (ref: Never) Current smoking Smoked but quit *** *** Self-rated living standard (ref: Positive) Fair 1.66*** 1.60*** 1.65*** 1.60*** Negative 2.32*** 4.14*** 2.31*** 4.12*** Dibao household 1.19* 1.93*** 1.19* 1.92*** (ref: No) (Factor Scores) Developed, marketized and lower health care Higher health facilities and quality of life Strong state influence and higher social security expenditure Random Part Level 2: Province 0.19(0.06) 0.15(0.05) 0.20(0.07) 0.15(0.05) 35

36 Results: Table 7 Multinomial multilevel regression estimates for reporting fair or negative life satisfaction *** p<0.01,**p<0.05,* p<0.1 Reporting fair or negative life satisfaction Individual characteristics Province characteristics Fair Negative Fair Negative ORs ORs ORs ORs Fixed Part CONS *** *** Age 0.99** 0.97*** 0.99** 0.97*** Female (ref: Male) Marital status (ref: Married) Others *** *** Widowed 0.79*** *** 1.04 Rural (ref: Urban) ** ** Education (ref: High school and above) Illiterate 0.62*** *** 1.06 Below Primary 0.76*** *** 1.21 Elementary and middle Income sources (ref: Wage) Wage and pension Wage and other Pension 1.26** ** 1.16 Others * 1.36 No 1.28*** 1.72*** 1.29*** 1.74*** 36

37 Results: Table 7 (Continued) Multinomial multilevel regression estimates for reporting fair or negative life satisfaction Reporting fair or negative life satisfaction Individual characteristics Province characteristics Fair Negative Fair Negative ORs ORs ORs ORs Fixed Part Have not medical 0.87 insurance (ref: Yes) Smoke (ref: Never) Current smoking Smoked but quit ** ** Self-rated living standard (ref: Positive) Fair 3.16*** 4.25*** 3.16*** 4.34*** Negative 5.47*** 32.33*** 5.47*** 33.08*** Dibao household (ref: No) *** *** (Factor Scores) Developed, marketized and lower health care Higher health facilities and quality of life 0.81** 0.66*** Strong state influence and higher social security expenditure *** p<0.01,**p<0.05,* p<0.1 Random Part Level 2: Province 0.18(0.06) 0.45(0.16) 0.13(0.05) 0.24(0.09) 37

38 Results: Table 8 Multinomial multilevel regression estimates for reporting fair or negative self-reported memory *** p<0.01,**p<0.05,* p<0.1 Reporting fair or negative self-reported memory Individual characteristics Province characteristics Fair Negative Fair Negative ORs ORs ORs ORs Fixed Part CONS 0.69** 0.14*** 0.74* 0.15*** Age ** ** Female (ref: Male) 1.18** 1.65*** 1.19** 1.67*** Marital status (ref: Married) Others * * Widowed 0.85* 0.78** 0.85* 0.78** Rural (ref: Urban) 1.17** 1.60*** 1.17** 1.60*** Education (ref: High school and above) Illiterate 1.29** 3.46*** 1.29** 3.45*** Below Primary 1.45*** 3.11*** 1.45*** 3.10*** Elementary and 1.28*** 1.82*** 1.27*** 1.81*** middle Income sources (ref: Wage) Wage and pension Wage and other Pension * * Others 1.34* 1.72*** 1.30* 1.66*** No 1.24** 1.64*** 1.23** 1.62*** 38

39 Results: Table 8 (Continued) Multinomial multilevel regression estimates for reporting fair or negative self-reported memory Fixed Part Have not medical insurance (ref: Yes) Reporting fair or negative self-reported memory Individual characteristics Province characteristics Fair Negative Fair Negative ORs ORs ORs ORs Smoke (ref: Never) Current smoking Smoked but quit 1.28** 1.47*** 1.28** 1.48*** Self-rated living standard (ref: Positive) Fair 1.77*** 1.69*** 1.76*** 1.68*** Negative 2.43*** 3.18*** 2.42*** 3.17*** Dibao household ** ** (ref: No) (Factor Scores) Developed, marketized and lower health care * Higher health facilities and quality of life Strong state influence and higher social security expenditure *** p<0.01,**p<0.05,* p<0.1 Random Part Level 2: Province 0.22(0.07) 0.42(0.14) 0.21(0.07) 0.45(0.15) 39