Maternal serum fructosamine levels and stillbirth: a case-control study of the Stillbirth Collaborative Research Network

Objective To evaluate the association between maternal fructosamine levels at the time of delivery and stillbirth. Design Secondary analysis of a case-control study. Setting Multicentre study of five geographic catchment areas in the USA. Population All singleton stillbirths with known diabetes status and fructosamine measurement, and representative live birth controls. Main outcome measures Fructosamine levels in stillbirths and live births among groups were adjusted for potential confounding factors, including diabetes. Optimal thresholds of fructosamine to discriminate stillbirth and live birth. Results A total of 529 women with a stillbirth and 1499 women with a live birth were included in the analysis. Mean fructosamine levels were significantly higher in women with a stillbirth than in women with a live birth after adjustment (177 +/- 3.05 versus 165 +/- 2.89 mu mol/L, P < 0.001). The difference in fructosamine levels between stillbirths and live births was greater among women with diabetes (194 +/- 8.54 versus 162 +/- 3.21 mu mol/L), compared with women without diabetes (171 +/- 2.50 versus 162 +/- 2.56 mu mol/L). The area under the curve (AUC) for fructosamine level and stillbirth was 0.634 (0.605-0.663) overall, 0.713 (0.624-0.802) with diabetes and 0.625 (0.595-0.656) with no diabetes. Conclusions Maternal fructosamine levels at the time of delivery were higher in women with stillbirth compared with women with live birth. Differences were substantial in women with diabetes, suggesting a potential benefit of glycaemic control in women with diabetes during pregnancy. The small differences noted in women without diabetes are not likely to justify routine screening in all cases of stillbirth. Tweetable abstract Maternal serum fructosamine levels are higher in women with stillbirth than in women with live birth, especially in women with diabetes.