The Lipid and Glyceride Profiles of Infant Formula Differ by Manufacturer, Region and Date Sold

Abstract

We tested the hypothesis that the lipid composition of infant formula is consistent between manufacturers, countries and target demographic. We developed techniques to profile the lipid and glyceride fraction of milk and formula in a high throughput fashion. Formula from principal brands in the UK (2017-2019; bovine-, caprine-, soya-based), the Netherlands (2018; bovine-based) and South Africa (2018; bovine-based) were profiled along with fresh British animal and soya milk and skimmed milk powder. We found that the lipid and glyceride composition of infant formula differed by region, manufacturer and date of manufacture. The formulations within some brands, aimed at different target age ranges, differed considerably where others were similar across the range. Soya lecithin and milk lipids had characteristic phospholipid profiles. Particular sources of fat, such as coconut oil, were also easy to distinguish. Docosahexaenoic acid is typically found in triglycerides rather than phospholipids in formula. The variety by region, manufacturer, date of manufacture and sub-type for target demographics lead to an array of lipid profiles in formula. This makes it impossible to predict its molecular profile. Without detailed profile of the formula fed to infants, it is difficult to characterise the relationship between infant nutrition and their growth and development.

Keywords: Lipidomics; infant development; infant nutrition; lipid metabolism.

Figure 1

Principal component analyses (PCAs) of infant formulae sold in the UK (2017–2019). Panels (A–D), triglycerides and diglycerides: (A), All brands; (B), British brand 1 (UKB1); British brand 2 (UKB2); British brand (UKB3). Panels E-H, phospholipids: (E), All brands; (F), British brand 1 (UKB1); (G), British brand 2 (UKB2); (H), British brand (UKB3). R² values for each component are shown in parentheses on each axis. H, ‘hungry’; N, ‘night’.

Figure 2

Principal Component Analyses and important loading variables of the phospholipid composition of formula UKB1 and reference materials. Panel (A), British brand 1 (UKB1) with subgroups that represent the hypothesised (*) dominant lipid source hypothesised marked (Red, cows’ milk lipids; Green, soya lecithin). (B), Plot of lipids from skimmed milk powder (SMP) and soya milk with loadings indicating which variables are the most important for distinguishing the two sample types. PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidylinositol. R² values for each component are shown in parentheses on each axis.

Figure 3

PCAs of bovine-based infant formula sold in South Africa and the Netherlands, caprine- and soya-based infant formulae sold in the UK and skimmed milk powder sold in the UK. Panel (A), phospholipids. Panel (B), triglycerides and diglycerides. Panel (C), Triglycerides and diglycerides for Dutch brand 2 (NLB2), with the three most abundant fatty ingredients marked in two subgroups. Panel (D), important loading variables for NLB2 that are consistent with independent profiling of coconut triglycerides. R² values for each component are shown in parentheses on each axis. Capr., caprine; Form, formula; G. m., Glycine max; NL, Netherlands; SA, South Africa; SMP, skimmed milk powder.

Figure 4

PCAs with important loading variables for Dutch and British brands that are produced by the same manufacturer. Panel (A), triglycerides; (B), phospholipids. DG, diglycerides; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PI, phosphatidylinositol; TG, triglyceride. 1, 0–6 mo; 2, 6–12 mo; 3, 12–24 mo; 24 mo+; H, 0–6 mo (hungry); NL, Dutch formula.