Alterations in the metabolism of phospholipids, bile acids and branched-chain amino acids predicts development of type 2 diabetes in black South African women: a prospective cohort study

Abstract

BackgroundSouth Africa (SA) has the highest global projected increase in diabetes risk. Factors typically associated with insulin resistance and type 2 diabetes risk in Caucasians are not significant correlates in black African populations. Therefore, we aimed to identify circulating metabolite patterns that predict type 2 diabetes development in this high-risk, yet understudied SA population. MethodsWe conducted a prospective cohort study in black SA women with normal glucose tolerance (NGT). Participants were followed for 13 years and developed (i) type 2 diabetes (n = 20, NGT-T2D), (ii) impaired glucose tolerance (IGT) (n = 27, NGT-IGT), or (iii) remained NGT (n = 28, NGT-NGT). Mass-spectrometry based metabolomics and multivariate analyses were used to elucidate metabolite patterns at baseline and at follow-up that were associated with type 2 diabetes development. ResultsMetabolites of phospholipid, bile acid and branched-chain amino acid (BCAA) metabolism, differed significantly between the NGT-T2D and NGT-NGT groups. At baseline: the NGT-T2D group had i) a higher lysophosphatidylcholine:lysophosphatidylethanolamine ratio containing linoleic acid (LPC(C18:2):LPE(C18:2)), ii) lower proliferation-related bile acids (ursodeoxycholic- and chenodeoxycholic acid), iii) higher levels of leucine and its catabolic intermediates (ketoleucine and C5-carnitine), compared to the NGT-NGT group. At follow-up: the NGT-T2D group had i) lower LPC(C18:2) levels, ii) higher apoptosis-related bile acids (deoxycholic- and glycodeoxycholic acid), and iii) higher levels of all BCAAs and their catabolic intermediates. ConclusionsChanges in lysophospholipid metabolism and the bile acid pool occur during the development of type 2 diabetes in black South African women. Further, impaired leucine catabolism precedes valine and isoleucine catabolism in the development of type 2 diabetes. These metabolite patterns can be useful to identify and monitor type 2 diabetes risk >10 years prior to disease onset and provide insight into the pathophysiology of type 2 diabetes in this high risk, but under-studied population.