Association between social capital and the prevalence of gestational diabetes mellitus: An interim report of the Japan Environment and Children's Study

Background Social capital is generally defined as the quality and frequency of social interactions with relatives, neighbors, and society. Social capital refers to broad concepts of social interactions and structures in individuals, communities and societies such as trust (e.g., neighborhood trust, which is social cohesion with neighbors) and social support (e.g., emotional support, which is the level of the feeling of being loved). Studies during the last few decades have shown that there is a significant inverse association between social capital and the prevalences of diseases such as depression and acute coronary syndrome. Significant inverse associations between trust, social support and the prevalence of diabetes have also been shown. However, associations between social capital and the prevalence of gestational diabetes mellitus (GDM) are less clear. Methods We used the primary dataset of the Japan Environment and Children's Study (JECS), including 10,228 mothers with recorded obstetric outcomes from January to December 2011. In this study, we included 8874 the 10,228 participants. Diagnosis of GDM was determined using the criteria of the Japan Diabetes Society (JDS). The quality and quantity of social capital were measured with nine questions on a self-administered questionnaire during the second or third trimester. Using principal component analysis (PCA), we distinguished the following three components (indices) of social capital: (A) index of all nine questions about social capital, (B) index of emotional support and neighborhood trust and (C) index of generalized trust. The high factor loading variants of indices were as follows; (A) all nine variants, (B) five variants about emotional support and neighborhood trust and (C) two variants about generalized trust. Multivariate random effect modeling was used to calculate the odd ratios (ORs) for evaluating the association between these indices of social capital and the prevalence of GDM. This model was adjusted for baseline characteristics such as family income, BMI before pregnancy and smoking during pregnancy. Results Of the 8874 participants, 204 women developed GDM (2.30%). Multivariable logistic regression analysis showed that index B, the index of emotional support and neighborhood trust (adjusted OR: 0.651, 95% CI: 0.429, 0.987) was significantly and independently associated with the prevalence of GDM. Conclusions We found that the 5-question index of emotional support and neighborhood trust is significantly associated with the prevalence of GDM during pregnancy.