Evolution in intergenerational exchanges between elderly people and their grandchildren in Taiwan; data from a multiple round cross-sectional study from 1993 to 2007.

Background: This study aimed to evaluate social evolution in Taiwan in recent decades using the changing pattern of care provided by grandparents for their grandchildren as an indicator.
Methods: Data from the second, fourth and sixth wave surveys of the Study of Health and Living Status of the Middle-Aged and Elderly in Taiwan were used for the analysis. This survey collected individual characteristics, including age, gender, education, ethnicity, dwelling place, living with partners, co-resident with children, employment status, self-reported health status and their provision of care for their grandchildren. Information about the attitudes toward National Health Insurance (NHI) was further collected in a questionnaire of 1999 following the implementation of NHI in 1995. By elders, we mean persons 60 or more years old. By grandchildren, we mean persons under 16 years of age. First, changes in individual characteristics were compared during these study periods (chi-square test). Then the logistic regression was performed to determine how significantly elders' grandchild-care behavior was associated with their individual characteristics.
Results: The percentage of elders providing grandchild care increased from 7.7% in 1993 to 13.6% in 1999, and then to 19.4% in 2007. By analysis, significant association was found between behavior in taking care of grandchildren and individuals of lower age, grandmothers, those living with partners or co-residing with children, those unemployed and those with better self-reported health status. And the effect of year was confirmed in the multivariable analysis.
Conclusions: This study pointed out the changing pattern of elders' behavior in taking care of their grandchildren as the main indicator and their related individual characteristics. We argue the need for improving social security policies in an ageing society. We suggest that the interaction between population policies and those of social welfare, including policies for health care and childcare, should be carefully evaluated.