Aims/hypothesis: We aimed to characterise the immunogenic background of insulin-dependent diabetes in a resource-poor rural African community. The study was initiated because reports of low autoantibody prevalence and phenotypic differences from European-origin cases with type 1 diabetes have raised doubts as to the role of autoimmunity in this and similar populations. Methods: A study of consecutive, unselected cases of recently diagnosed, insulin-dependent diabetes (n = 236, ≤35 years) and control participants (n = 200) was carried out in the ethnic Amhara of rural North-West Ethiopia. We assessed their demographic and socioeconomic characteristics, and measured non-fasting C-peptide, diabetes-associated autoantibodies and HLA-DRB1 alleles. Leveraging genome-wide genotyping, we performed both a principal component analysis and, given the relatively modest sample size, a provisional genome-wide association study. Type 1 diabetes genetic risk scores were calculated to compare their genetic background with known European type 1 diabetes determinants. Results: Patients presented with stunted growth and low BMI, and were insulin sensitive; only 15.3% had diabetes onset at ≤15 years. C-peptide levels were low but not absent. With clinical diabetes onset at ≤15, 16–25 and 26–35 years, 86.1%, 59.7% and 50.0% were autoantibody positive, respectively. Most had autoantibodies to GAD (GADA) as a single antibody; the prevalence of positivity for autoantibodies to IA-2 (IA-2A) and ZnT8 (ZnT8A) was low in all age groups. Principal component analysis showed that the Amhara genomes were distinct from modern European and other African genomes. HLA-DRB1*03:01 (p = 0.0014) and HLA-DRB1*04 (p = 0.0001) were positively associated with this form of diabetes, while HLA-DRB1*15 was protective (p < 0.0001). The mean type 1 diabetes genetic risk score (derived from European data) was higher in patients than control participants (p = 1.60 × 10−7). Interestingly, despite the modest sample size, autoantibody-positive patients revealed evidence of association with SNPs in the well-characterised MHC region, already known to explain half of type 1 diabetes heritability in Europeans. Conclusions/interpretation: The majority of patients with insulin-dependent diabetes in rural North-West Ethiopia have the immunogenetic characteristics of autoimmune type 1 diabetes. Phenotypic differences between type 1 diabetes in rural North-West Ethiopia and the industrialised world remain unexplained.
Bibliographical noteFunding Information:
The authors thank C. Self, research student, for assistance with computing and M. McDonnell and A. Magill, Department of Clinical Biochemistry, Royal Victoria Hospital, Belfast for laboratory support. We are grateful to V. Lampasona, San Raffaele Scientific Institute, Milan and A. Lernmark, Lund University, Sweden for the plasmids used for islet autoantibody measurement. The authors declare that there are no relationships or activities that might bias, or be perceived to bias, their work.
DIWP and ERT received a grant from the Association of Physicians of Great Britain and Ireland; DLC is supported by a grant from the NIH, NIH1K99HD099330-01; SFAG is supported by the NIH (RO1 DK085212) and the Daniel B. Burke Endowed Chair for Diabetes Research; STJ and RDL are funded by Type 1 Diabetes UK (DUK 19/0005951); RDL is funded by the EFSD (MMBP1C3R), the EU (EU-FP7:282510) and St Bartholomew’s Hospital, London, Charity (OGA011582); TV is funded by the British Diabetic Twin Trust. Acknowledgements Authors’ relationships and activities
© 2020, The Author(s).
Copyright 2020 Elsevier B.V., All rights reserved.
- Type 1 diabetes
ASJC Scopus subject areas
- Internal Medicine
- Endocrinology, Diabetes and Metabolism