Associations between maternal adiposity and appetite-regulating hormones in human milk are mediated through maternal circulating concentrations and might affect infant outcomes

Authors: CHRISTENSEN, SOPHIE - University Of Copenhagen; LEWIS, JACK - University Of Copenhagen; LARNKJÆR, ANNI - University Of Copenhagen; FRØKIÆR, HANNE - University Of Copenhagen; Allen, Lindsay - A; MØLGAARD, CHRISTIAN - University Of Copenhagen; MICHAELSEN, KIM - University Of Copenhagen

Submitted to: Frontiers in Nutrition
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/13/2022
Publication Date: 11/4/2022
Citation: Christensen, S.H., Lewis, J.I., Larnkjær, A., Frøkiær, H., Allen, L.H., Mølgaard, C., Michaelsen, K.F. 2022. Associations between maternal adiposity and appetite-regulating hormones in human milk are mediated through maternal circulating concentrations and might affect infant outcomes. Frontiers in Nutrition. 9. Article 1025439. https://doi.org/10.3389/fnut.2022.1025439.
DOI: https://doi.org/10.3389/fnut.2022.1025439

Interpretive Summary: Appetite-regulating hormones (ARH) in mother's milk (HM) could affect the infants’ milk intake and its growth. These hormones may be increased in the milk when the mother is obese or overweight. Using the novel statistical method counterfactual-based mediation analysis we investigated direct and indirect effects between the ARHs and maternal and infant factors. This is a method to investigate relationships among the mother-milk-infant triangle. Therefore, our goals were to examine relationships between A) maternal adiposity and ARH in the milk, and B) milk-ARH, infant milk intake, and growth, and how they are affected by maternal and infant plasma ARH. Data were used from 204 mother-infant pairs, recruited for the Mothers, Infants and Lactation Quality study, at three visits from 1-8.49 months postpartum, including maternal and infant anthropometry and body composition, human milk and blood leptin, insulin and adiponectin. We found a positive correlation between maternal body-mass-index (BMI) and HM leptin, facilitated by maternal plasma leptin by 29% when using a BMI cutpoint of <25kg/m2, and by 51% when using a cutoff of BMI =25kg/m2 (p for nteraction <0.01). There was no mediated effect through plasma insulin in the association between BMI and HM insulin (p=0.068). Log-HM adiponectin was negatively associated with total milk intake (2.5 mL per 10% increase; ßlog=-26.1, p=0.046) with no mediated effect through infant plasma adiponectin (p=0.22). In conclusion, the association between maternal adiposity and HM leptin was mediated through circulating leptin to a stronger degree for overweight mothers compared those not overweight. This indicates that excess maternal weight and the resulting increased circulating leptin concentrations and concomitant low-grade inflammation, may be reflected in HM composition.

Technical Abstract: Background: Appetite-regulating hormones (ARH) in human milk (HM) are suggested to affect infants’ milk intake and possibly infant growth. Maternal adiposity might contribute to higher levels of ARH in HM, either directly from the adipose tissue or from raised circulating levels. Counterfactual-based mediation analysis can define direct and indirect effects between HM ARH and maternal and infant factors, and might be an important tool when investigating the mother-milk-infant triad. Objective: To test our hypotheses namely that potential associations between 1) maternal adiposity and HM ARH and 2) HM ARH and infant milk intake and growth are mediated through maternal and infant plasma ARH, respectively. Maternal and infant anthropometry and body composition, HM and blood samples were collected from 204 mother-infant dyads participating in the Mother, Infant and Lactation Quality study at three postpartum visits from 1-8.49 months. Leptin, insulin and adiponectin were analyzed using immunoassays. Mediation analyses using linear mixed-effect models were applied to investigate the direct and indirect effects through maternal and infant plasma hormone concentrations. Results: A positive association between maternal body-mass-index (BMI) and HM leptin was mediated by maternal plasma leptin by 29% when fixing BMI to <25kg/m2, and by 51% when fixing BMI to =25kg/m2 (p interaction<0.01). There was no mediated effect through plasma insulin in the association between BMI and HM insulin (p=0.068). We found negative and positive associations between HM insulin and total milk intake and infant weight, respectively, however these diminished in mediation analyses with reduced sample sizes. Conclusion: Our main results suggest that the association between maternal adiposity and HM leptin was mediated through circulating leptin to a stronger degree for mothers with compared to without overweight. This indicates that excess maternal adiposity, and the resulting rise of circulating leptin and concomitant low-grade inflammation, may be reflected in HM composition.