Background and aims: Pregnancy drives marked physiological adaptations, including altered insulin sensitivity, linked to the risk to develop gestational diabetes. However, metabolic changes

and their relationship with insulin sensitivity remains poorly understood. This study aimed to investigate longitudinal changes in metabolite profiles in fasting and fed states during and after pregnancy, and the possible associations with insulin sensitivity.

Material and methods: We included data from 26 pregnant women with overweight or obesity from the Pregnancy Outcome and Maternal Insulin Sensitivity (PROMIS) study. Targeted metabolomics was analyzed by LC–HRMS (Orbitrap Exploris 120). Metabolites were measured in the fasting state and 2 hours following a meal tolerance test (MTT) at weeks 12-16 and 24-26 of gestation, and 3 months postpartum. Insulin sensitivity was calculated based on HOMA-IR and the MTT-derived Matsuda index (MI).

Result: A total of 156 metabolites were annotated and semi-quantified. In the early second trimester, 11 fasting and 25 fed metabolites were associated with insulin sensitivity (HOMA-IR or MI). These numbers increased to 61 fasting and 41 fed metabolites in the late second trimester, while 19 fasting and 20 fed metabolites were associated with these indices postpartum. Fructose in the fasting state was consistently correlated with MI (r=-0.502, p=0.040 and r=-0.701, p=0.005 at the late second trimester and postpartum, respectively), whereas pyruvic acid in the fed state was consistently correlated with HOMA-IR (r=0.542, p=0.016 and r=0.069, p=0.006, respectively). Early-pregnancy fasting state pyruvic acid and fed state aminoadipic acid showed the strongest association with decreased insulin sensitivity in the late second trimester (pyruvic acid vs MI: r=-0.828, r=0.006; aminoadipic acid vs HOMA-IR: r=0.516, p=0.020).

Conclusion: We observed time-specific associations between specific metabolites, fasting and fed state and insulin sensitivity. These exploratory findings underscore the importance of accounting for both metabolic status and gestational age when developing and evaluating metabolomic biomarker panels for monitoring maternal metabolic health.

P. Arum¹, P de Blaauw², J Groen², M.R Heiner-Fokkema², H. L. Lutgers³, E. M. van der Beek^1,4, J. K. Kruit¹

 

¹Department of Pediatrics, UMCG, Groningen, the Netherlands, ²Laboratory of Metabolic Diseases, Department of Laboratory Medicine, UMCG, Groningen, the Netherlands, ³Department of Endocrinology, UMCG, Groningen, the Netherlands, ⁴Nestlé Institute of Health Sciences, Nestlé Research, Societé du Produits Nestlé Lausanne, Lausanne, Switzerland, parum@umcg.nl