隔代教養與兒童健康

Abstract

背景與目的：家庭教養影響兒童健康。鑒於台灣隔代教養家庭有逐年上升的趨勢，本研究旨在探討隔代教養的兒童是否健康較差。 方法：本研究資料來源為臺灣出生世代研究的第三波面訪調查，有效樣本為19,840 名3 歲兒童及其主要照顧者。本研究利用兒童居住安排定義隔代教養，其中主要與祖父母同住但非三代同堂者為隔代教養，並且進一步依父母是否偶爾與兒童同住再分為隔代A 與隔代B 兩組（前者為偶爾同住，後者則完全未同住），因此兒童依居住安排共有五組：父母（核心家庭）、三代同堂、隔代A，隔代B及其他。兒童健康變項為主要照顧者對兒童健康的評價。我們利用邏輯斯複迴歸分析隔代教養對兒童健康的影響，控制變項包括：兒童特性、家庭人口與社經特性。 結果：19,840 名樣本中，隔代教養組佔7.63%（隔代A 組 5.56%，隔代B 組 2.07%）。就兒童健康而言，主要照顧者評價為好的百分比分別是：父母組77.1%，三代組77.1%，隔代A 組 81.5%，隔代B 組 77.6%。雖然隔代教養組與非隔代教養組的兒童健康差異在雙變項分析時達統計顯著水準（p=0.014），但邏輯斯複迴歸分析結果則未達統計顯著水準（隔代A 組：OR=1.26, 95% CI 0.65-2.45 ;隔代B 組：OR=1.36, 95% CI 0.61-3.04）。值得注意的是，隔代B組非雙親家庭的比例特別高，達12.9%。 結論：兒童健康與隔代教養之間並無顯著相關。

Objective: Parenting matters to the health of children. In view of the increasing incidence of grandparents raising their grandchildren in Taiwan over the past decades, the study aims to examine whether children from grandparenting family have worse health outcomes. Method: Data for the analysis comes from the third-wave face-to-face interview survey of the Taiwan Birth Cohort Study, with an effective sample of 19,840 children aged 3 and their primary care givers. We used children’s living arrangement to define grandparenting - children lived with their grandparents and their parents seldom visited (grandparenting group A) or lived with their grandparents only (grandparenting group B); other types of living arrangement include living with parents only (nuclear family), living with parents and grandparents (three enerations family), and living with neither parents nor grandparents. Child health outcome variable is primary caregiver-rated health of the child. The multivariate logistic regression analysis was used to examine the effect of grandparenting on child health, with control variables including child characteristics, family demographics and socioeconomic status. Results: Among the 19,840 children, 7.63% were grandparenting (grandparenting group A, 5.56%; grandparenting group B, 2.07%). The proportion of better rated-health was 77.1% for the group of living with parents only, 77.1% for the roup of three generations family, 81.5% for grandparenting group A, and 77.6% for grandparenting group B. Although the health of children was significantly different between grandparenting groups and non-grandparenting groups in a Chi-square test (p-value=0.014), but it disappeared in a multivariate logistic regression analysis (grandparenting group A, OR=1.26, 95% CI 0.65-2.45; grandparenting group B, OR=1.36, 95% CI 0.61-3.04). Importantly, grandparenting group B had the highest percentage of single-parent families (12.9%). Conclusion: There is no significant association between grandparenting and child health.