Social and health risk factors for miscarriage: a population-based study using French healthcare records

Objective: To study social and clinical factors associated with (early or late) miscarriage, while accounting for births and induced abortions. Design: Population-based analysis of French healthcare records from 2015 to 2023. Subjects : Pregnancies among women aged 15-49 that resulted in birth, miscarriage or induced abortion. Main Outcome Measures: Predicted probabilities for a pregnancy to result in birth, induced abortion, or miscarriage, including early (<14 gestational weeks) or late (14-22 gestational weeks) miscarriage. The probabilities are estimated using multinomial regression models controlling for key socio-demographic and health factors. Results: Eight percent of the studied pregnancies were miscarriages and among them, seven percent occurred at a late gestational age. Adjusted results for socio-demographic factors show age below 19 or above 40 were associated with higher predicted probabilities for miscarriage: while the predicted probabilities were below 10% for the age groups in the middle, they were 14.0% [13.9%;14.2%] for adolescents and 16.0% [15.9%;16.1%] for those in their early 40s. There were no differences in overall miscarriage risk between rural, peri-urban and urban residential areas, or by financial precarity status, except for late miscarriages where risk was higher among those in a precarious financial situation. Regarding past reproductive history, miscarriage was more likely after the use of medically assisted reproduction, with the predicted probabilities for this group being 14.9% [14.7%;15.0%] compared to 7.5% [7.4%;7.5%] for those without. Previous births or induced abortions were associated with lower predicted probabilities of miscarriage, whereas previous miscarriages were associated with an increased probability of miscarriage (e.g. 12.5% [12.3%;12.8%] after multiple miscarriages and 7.3% [7.3%;7.3%] if no previous miscarriage). Chronic diseases were each associated with slightly higher predicted probabilities of miscarriage, with exceptionally high association for kidney failure (17.0% [15.1%;19.1%]). Conclusion: No social inequalities were observed for early miscarriage risk, but the risk of late miscarriage was more important for financially precarious than other women. Regarding health factors, chronic kidney disease was particularly strongly associated with elevated miscarriage risk, while all included chronic diseases increased the risk somewhat.