Non-autoimmune, insulin-deficient diabetes in children and young adults in Africa: evidence from the Young-Onset Diabetes in sub-Saharan Africa (YODA) cross-sectional study

Background: Studies of type 1 diabetes in sub-Saharan Africa have suggested that the clinical phenotype might differ from phenotypes reported elsewhere. We aimed to establish whether type 1 diabetes diagnosed in children and young adults in three countries across sub-Saharan Africa is of autoimmune origin. Methods: In this observational, cross-sectional study, we identified participants without obesity from outpatient clinics in government and private hospitals in Cameroon, Uganda, and South Africa who were of self-reported Black African ethnicity with young-onset (age <30 years), insulin-treated, clinically diagnosed type 1 diabetes. We measured islet autoantibodies to GADA, IA-2A, and ZnT8A, and calculated a genetic risk score (GRS) for type 1 diabetes, which we compared with control populations without diabetes derived from the Uganda Genome Resource databank and other studies. Endogenous insulin secretion was assessed using plasma C-peptide. We compared findings with those for participants with self-reported Black (n=429) and White (n=2602) ancestry with type 1 diabetes from the SEARCH for Diabetes in Youth (SEARCH) study in the USA. Findings: Of 1072 participants identified between Aug 28, 2019, and March 31, 2022 (Cameroon and Uganda), and Oct 3, 2007, to Sept 14, 2015 (South Africa), 894 were included in our analysis (454 [50·8%] were male and 440 [49·2%] were female): 248 participants were from Cameroon, 370 from Uganda, and 276 from South Africa. Participants from sub-Saharan Africa were diagnosed with diabetes at a median age of 15 years (IQR 11–19), with a median diabetes duration of 5 years (2–10), and a BMI of 21·7 kg/m² (19·5–24·1). Only 312 (34·9%) of 894 participants were positive for islet autoantibodies; these participants had classic features of type 1 diabetes, including 225 (82·7%) of 272 with plasma C-peptide <200 pmol/L, and high type 1 diabetes GRS. Those without islet autoantibodies (582 [65·1%] of 894) had significantly lower median type 1 diabetes GRS than those with autoantibodies (9·66 [IQR 7·77–11·33] vs 11·76 [10·49–12·91]; p<0·0001), suggesting a subgroup with a non-autoimmune diabetes subtype, with clinical features and C-peptide concentrations not consistent with type 2 diabetes. Among participants diagnosed younger than 20 years, autoantibody-negative diabetes was also observed in 65 (15·1%) of 429 participants with Black ancestry in SEARCH (although less frequently than in sub-Saharan Africa [59 (55·1%) of 107]), and these participants also had a low type 1 diabetes GRS (median 10·41 [IQR 8·65–12·22] in autoantibody-negative subgroup). No such pattern was observed in White participants in SEARCH: 241 (9·3%) of 2602 were autoantibody negative and median GRS for type 1 diabetes was similar in autoantibody-negative and autoantibody-positive participants (median 13·42 [IQR 11·80–14·61] vs 13·49 [12·29–14·58]). Interpretation: In sub-Saharan Africa, clinically diagnosed type 1 diabetes is heterogeneous, comprising classic autoimmune type 1 diabetes and a novel, non-autoimmune, insulin-deficient diabetes subtype. There is evidence of this subtype in Black but not White individuals in the USA. Therefore, alternative causes must be considered in this group of individuals, and understanding the drivers of this subtype might offer new insights into prevention and treatment. Funding: UK National Institute of Health and Care Research. Translation: For the French translation of the abstract see Supplementary Materials section.