Risk of cognitive impairment in children after maternal exposure to the 1998 Yangtze River flood during pregnancy: analysis of data from China's second National Sample Survey on Disability.

BACKGROUND

More research is needed to understand the long-term effects of prenatal exposure to adverse events, such as floods and other natural disasters, on cognitive outcomes in childhood. We aimed to explore the risk of cognitive impairment in children following maternal exposure to the 1998 Yangtze River flood in China during pregnancy.

METHODS

For this study we obtained and analysed individual-level data from the second China National Sample Survey on Disability (CNSSD), which was done in 2006. We defined the flood period as June-August, 1998. The analytical sample comprised children from four birth cohorts, defined according to their month of birth: the post-partum exposed cohort (children born between June, 1997, and May, 1998), the prenatal exposed cohort (children born between June, 1998, and May, 1999), the preconception exposed cohort (children born between June 1999, and May, 2000), and the unexposed cohort (children born between June, 2000, and May, 2001). In the CNSSD, cognitive impairment was assessed and diagnosed by validated screening tools and procedures. Difference-in-difference models were used to examine variations in the effects of maternal flood exposure on cognitive impairment in childhood across the different birth cohorts and regions.

FINDINGS

108 175 children born between June, 1997, and May, 2001, and aged 4-8 years at the time of the survey, were included in our analysis. 1131 children had a cognitive impairment; the prevalence of cognitive impairment was 1·05% (95% CI 0·99-1·11). Maternal exposure to flood during pregnancy increased the risk of cognitive impairment among children (adjusted odds ratio [OR] 2·18 [95% CI 1·54-3·08]; p<0·0001). No significant sex-specific differences were observed, and the risk of cognitive impairment was especially high when maternal flood exposure occurred during the first trimester of pregnancy (adjusted OR 5·05 [95% CI 3·88-6·58]; p<0·0001). The risk of cognitive impairment also increased with longer durations of maternal flood exposure and with increasing severity of flooding; the risk was highest in the prenatal exposed cohort with 3 months of flood exposure in the most severely affected area (adjusted OR 5·56 [95% CI 1·58-19·54]; p=0·007).

INTERPRETATION

Prenatal flood exposure had a long-term negative effect on cognitive development of children. Greater maternal support and public health interventions during pregnancy and early life after a natural disaster are warranted to facilitate healthy cognitive development in later life.

FUNDING

National Social Science Foundation of China.