Methods: We conducted multivariate regression analysis using the WHO Study on Global Ageing and Adult Health (SAGE) of respondents living in China over the age of 50 (n=13,367) in 2007. Data were collected on socio-demographic backgrounds, work histories, household membership, health, social cohesion, and subjective well-being. The majority of our sample was born in a rural region (51% rural hukou status), had completed less than six years of formal education (44%), was married (83%), and had an average of 2.75 people in their household. Using household roster, we categorized living arrangement into six groups: living with spouse only (n =5,481), living with just adult children (two generation, n =3,379), living with adult children and grandchildren (three generation, n = 1,868), living with just grandchildren (skipped-generation, n =522), living alone (n =1,594), and all other household arrangements (n =523). We defined subjective well-being using the following measures: satisfaction with health, overall life quality, life satisfaction, and happiness. A rich set of sociodemographic characteristics (e.g., age, gender, education, income, social connection) were included in all analyses to address at least partially the selection bias.
Findings: Descriptive statistics indicate that two- and three-generational households tended to have more resources (e.g., higher family income), live in urban areas, and have more social connections than other arrangements. In contrast, those in skipped-generational households had fewer resources, were working and engaged more in risky health behaviors (drinking alcohol and smoking). Regression results suggest that two-generational living arrangements were significantly associated with lower subjective well-being than were those living with spouse only. In contrast, those living in three-generational living arrangements had significantly better subjective well-being than those in two-generational living arrangements. Interaction analysis further indicate that the lower subjective well-being found for those living in two-generation households were particularly true among those of low-income.
Conclusions and Implications: This study highlights important protective and risk factors associated with various multigenerational household arrangements for older adults in China and identifies not only that rural skipped-generation households as a vulnerable social subgroup in Chinese society in our data but also that elders in low-income two-generational households worth particular attention. The many permutations of multigenerational households in China appear to have diverse vulnerabilities that policy and clinical interventions must consider going forward.