A systematic review of collection and analysis of human milk for macronutrient composition

Abstract

BACKGROUND:As human milk (HM) composition varies by time and across even a single feed, methods of sample collection can significantly affect the results of compositional analyses and complicate comparisons between studies. OBJECTIVE:The aim was to compare the results obtained for HM macronutrient composition between studies utilizing different sampling methodologies. The results will be used as a basis to identify the most reliable HM sampling approach. METHODS:EMBASE, MEDLINE/PubMed, Cochrane Library, Scopus, Web of Science, and ProQuest databases were searched for relevant articles. Observational and interventional studies were included, and at least 2 authors screened studies and undertook data extraction. Quality assessment was conducted using the Newcastle-Ottawa scale and previously published pragmatic score. RESULTS:A total of 5301 publications were identified from our search, of which 101 studies were included (n = 5049 breastfeeding women). Methods used for HM collection were divided into 3 categories: collection of milk from all feeds over 24 h (32 studies, n = 1309 participants), collection at one time point (62 studies, n = 3432 participants), and "other methods" (7 studies, n = 308 participants). Fat and protein concentrations varied between collection methods within lactation stage, but there were no obvious differences in lactose concentrations. There was substantial variability between studies in other factors potentially impacting HM composition, including stage of lactation, gestational age, and analytical method, which complicated direct comparison of methods. CONCLUSIONS:This review describes the first systematic evaluation of sampling methodologies used in studies reporting HM composition and highlights the wide range of collection methods applied in the field. This information provides an important basis for developing recommendations for best practices for HM collection for compositional analysis, which will ultimately allow combination of information from different studies and thus strengthen the body of evidence relating to contemporary HM composition. This trial was registered at PROSPERO as CRD42017072563, https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42017072563.