Associations between human milk oligosaccharides and infant growth in a Bangladeshi mother-infant cohort

Miranda G Loutet; Arash Narimani; Huma Qamar; Chloe Yonemitsu; Lisa G Pell; Abdullah Al Mahmud; Tahmeed Ahmed; Lars Bode; Diego G Bassani; Daniel E Roth

doi:10.1038/s41390-023-02927-1

Associations between human milk oligosaccharides and infant growth in a Bangladeshi mother-infant cohort

Pediatr Res. 2023 Dec 5. doi: 10.1038/s41390-023-02927-1. Online ahead of print.

Authors

Miranda G Loutet^{1

2}, Arash Narimani³, Huma Qamar³, Chloe Yonemitsu⁴, Lisa G Pell³, Abdullah Al Mahmud⁵, Tahmeed Ahmed⁵, Lars Bode⁴, Diego G Bassani³, Daniel E Roth^{3

6}

Affiliations

¹ The Hospital for Sick Children, Toronto, ON, M5G 0A4, Canada. miranda.loutet@sickkids.ca.
² The University of Toronto, Toronto, ON, M5S 1A1, Canada. miranda.loutet@sickkids.ca.
³ The Hospital for Sick Children, Toronto, ON, M5G 0A4, Canada.
⁴ University of California San Diego, San Diego, CA, USA.
⁵ Nutrition and Clinical Services Division, icddr,b, Dhaka, Bangladesh.
⁶ The University of Toronto, Toronto, ON, M5S 1A1, Canada.

PMID: 38052861
DOI: 10.1038/s41390-023-02927-1

Abstract

Background: We aimed to estimate associations between human milk oligosaccharides (HMOs) and infant growth (length-for-age (LAZ) and weight-for-length (WLZ) z-scores) at 12 months postnatal age.

Methods: In this secondary analysis of data from a maternal vitamin D trial in Dhaka, Bangladesh (N = 192), absolute concentrations of HMOs were measured in 13 ± 1 week(s) postpartum milk samples, infant anthropometric measurements were obtained soon after birth and at 12 months postpartum, and infant feeding was classified during 6 months postpartum. Associations between individual HMOs or HMO groups and LAZ or WLZ were estimated by multivariable linear regression adjusting for infant feeding pattern, maternal secretor status, and other potential confounders.

Results: The concentrations of 6'sialyllactose, lacto-N-neotetraose, and the non-fucosylated non-sialylated HMOs were inversely associated with LAZ at 12 months of age, whereas the fucosylated non-sialylated HMO concentration was positively associated with LAZ at 12 months. These associations were robust in analyses restricted to infants who were primarily exclusively/predominantly fed human milk during the first 3 (or 6) months.

Conclusions: Since HMOs are both positively and negatively associated with postnatal growth, there is a need for randomized trials to estimate the causal benefits and risks of exogenously administered HMOs on infant growth and other health outcomes.

Impact: 6'sialyllactose, lacto-N-neotetraose, and the non-fucosylated non-sialylated human milk oligosaccharides (HMOs) were inversely associated with length-for-age z-scores (LAZ) at 12 months, whereas the fucosylated non-sialylated HMO concentration was positively associated with LAZ at 12 months among Bangladeshi infants. Associations between individual and grouped HMOs with infant length growth at 12 months were as strong or stronger in analyses restricted to infants who were exclusively or predominantly fed human milk up to 3 (or 6) months. Randomized trials are needed to characterize the effects of specific HMOs on infant growth, particularly in countries where postnatal linear growth faltering is common.