Associations of ambient temperature exposure during pregnancy with the risk of miscarriage and the modification effects of greenness in Guangdong, China.

Miscarriage is one of the commonest complications of pregnancy. Although previous studies suggested that environmental factors were important causes of miscarriage, evidence is still inadequate. Here, we examined the association of maternal exposure to temperature with the risk of miscarriage and further assessed the modifying effects of surrounding residential greenness. A case-control study was conducted at a large hospital in Guangzhou, China. All participants' information was extracted from hospital records. An inverse distance weighted method was used to estimate the temperature exposure at each residential address, where the greenness was measured by Normalized Difference Vegetation Index (NDVI). A logistic regression model was applied to estimate the association of temperature exposure with the risk of miscarriage. A total of 2044 cases of miscarriage and 2285 controls were included in the present study. We observed a generally non-linear positive relationship between temperature exposure and the risk of miscarriage. More pronounced effects of high temperatures vs. low temperatures were found during the two months prior to hospitalization than in other periods. The odds ratio (OR) of 29.4 °C (95th centile) compared with 15 °C during the first month prior to hospitalization was 1.480 (95% CI: 1.021-2.145). Smaller effects of temperatures were seen on the risk of miscarriage among participants with moderately great surrounding greenness compared with those with less greenness. We concluded that maternal exposure to moderately high temperature during pregnancy may increase the risk of miscarriage, but the modifying effects of greenness on these associations need to be further tested in future studies.