To investigate effects of maternal diabetes and metformin treatment on metabolic newborn screening (NBS) results of infants born to mothers with hyperglycemia during pregnancy.


Retrospective case-control study. NBS results of infants born to mothers treated with metformin for hyperglycemia during pregnancy were compared with diet-treated subjects with diabetes and matched normal control subjects. Exclusions: maternal type 1 diabetes, major fetal anomalies, and incomplete infant data. Inclusions: maternal hyperglycemia in pregnancy treated with diet alone or diet plus metformin. Results from the New South Wales Newborn Screening Program (dried infant blood spot sample, 24–72 h after birth) for 25 routinely studied analytes were measured using mass spectrometry. Data from metformin-exposed and control infants were compared using nonparametric methods and multiples of the median for each analyte.


A total of 574 case subjects were compared with 952 diet-treated case subjects with diabetes and 979 control subjects. Metformin-exposed infants had shorter gestational age (266 ± 7 vs. 272 ± 10 vs. 274 ± 9 days) (P < 0.001) and lower birth weights (3.28 ± 0.51 vs. 3.29 ± 0.49 vs. 3.33 ± 0.43 kg) (P = 0.008). Short-, medium-, and one long-chain acylcarntine (tetradecanoylcarnitine [C14]) concentrations were higher in the metformin-exposed group compared with normal control subjects. Comparison with diet-treated control subjects with diabetes (to eliminate confounding by hyperglycemia) continued to show raised butyrylcarnitine (C4), isovalerylcarnitine (C5), and glutarylcarnitine (C5D) in the metformin-exposed group. There was no evidence of vitamin B12 deficiency (low methionine and elevated propionylcarnitine [C3]) in metformin-exposed infants. All results were within normal population limits.


We have identified subtle (nonpathological) changes in neonatal metabolism that represent a signature effect of fetal metformin exposure.

V.W., D.S., and M.M. contributed equally to the study.

This article contains supplementary material online at https://doi.org/10.2337/figshare.15148788.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/content/license.
You do not currently have access to this content.