Introduction: Studies suggest that breastfeeding (BF) decreases the risk of T2DM in Gestational Diabetes Mellitus (GDM) women. The mechanisms by which this improvement occurs are poorly understood. Our objective was to analyze the association of BF with markers of insulin resistance (MIR) and the potential mediation effect of subclinical inflammation (SI) in this association, in women with overweight/obesity and GDM.
Methods: This prospective study enrolled a convenience sample of 99 pregnant women followed from the 1st trimester up until 2-6 months postpartum. Blood was drawn at postpartum. A latent variable of SI was inferred indirectly through a mathematical model from the observable variables: adiponectin, branched chain amino acids, E-selectin, copeptin, zonulin and lipopolysaccharide (LPS). Linear regression models showed the association of BF on MIR, adjusted for covariables identified by the software Daggity. Mediation analyses were performed to determine the total, direct and the indirect effect (via SI) of BF on MIR.
Results: Women who predominantly breastfed (BF group = 44) had lower levels of triglycerides (TG), fasting glucose and insulin, TG/HDL ratio, TyG index, and HOMA-IR. LPS levels were lower in the BF group [6.8 (4.2-10.6) vs 9.2 (6.9-11.5)ng/mL, p=0.048] while other variables were similar among groups. In linear regression models an inverse association was observed between BF and fasting glucose [-6,60(-11.06 to -2.14), p=0.004), HOMA-IR [-0.26(-0.49 to -0.03), p=0.026] and TyG index [-004(-0.06 to -0.01), p=0.003] adjusted for GDM, parity, pre-gestational BMI, scholarity, weight gain during pregnancy and mode of delivery. Mediation analysis showed that SI did not mediate the results seen on the linear regression models.
Conclusion: BF is associated with an improvement in MIR in postpartum obese/overweight women, regardless of GDM. SI did not mediate the association between BF and improvement in MIR.
J.M. Oliveira: None. A.A. Ferraro: None. P. Dualib: None. B. Almeida-Pititto: None.
