Ethnicity and antenatal maternal anxiety: associations with the human milk lipidome in a multi-ethnic Asian cohort

INTRODUCTION: Human milk (HM) lipid composition varies considerably among women, but factors associated with variation in complex lipids beyond fatty acid constituents are less well-described. We investigated maternal and infant factors associated with the intact human milk lipidome at 3 weeks and 3 months postpartum in a multi-ethnic Asian population. METHODS: In this observational cohort study, 237 lipids across 12 lipid subclasses were quantified in HM samples collected at 3 weeks (n = 189) and 3 months (n = 104) postpartum via liquid chromatography-mass spectrometry and direct-infusion mass spectrometry. Principal component analysis captured major axes of lipid variation and time-specific regression models examined associations between predictors (including ethnicity, parity, pre-pregnancy BMI, tobacco smoke exposure, maternal mental health, breastfeeding exclusivity/predominance) and top principal components, with correction for multiple comparisons. RESULTS: Ethnicity emerged as the strongest factor at both timepoints, explaining 5.6%-10.4% of the variance at 3 weeks and 2.6%-15.6% at 3 months. Chinese had higher polyunsaturated-fatty-acid (PUFA) lipids, especially very long-chain PUFA triacylglycerols (TGs) and n-3 PUFA phospholipids (PLs), than Indians and Malays. Indians had higher odd-chain lipids (saturated and monounsaturated) and specific PLs compared with Chinese. Together, Indians and Malays had a more saturated- and monounsaturated lipid composition, as well as long-chain sphingolipids compared with Chinese. Antenatal maternal anxiety was independently associated with an altered lipid profile - higher ceramides, monounsaturated lipids and lower PUFA TGs - at 3 weeks postpartum. CONCLUSION: Ethnic differences may partly reflect culturally distinct postpartum confinement diets and fatty acid desaturase (FADS1/FADS2) genotypic variants. Further investigation into culturally tailored nutritional guidance and perinatal psychological support as modifiable factors associated with the HM lipidome is warranted.