Evidence of ethnic variations in the relationships between routinely recorded clinical factors and T2D: a systematic review and meta-analysis

BackgroundEvidence on ethnic differences in factors associated with type 2 diabetes (T2D) is mixed. We aimed to systematically review evidence on ethnic variations in the relationships between routinely recorded demographic and clinical factors and T2D. MethodsWe searched Medline Complete and Embase for observational studies published between 1990 and 2023 investigating ethnic differences in factors routinely recorded in clinical encounters associated with T2D. We used random and fixed-effects meta-analysis to quantitatively summarise effect sizes across studies where possible. Risk of bias and study quality were assessed using the Newcastle-Ottawa Scale and Joanna Briggs Institute tool. PROSPERO registration: CRD42023394148. FindingsSearches identified 10 694 studies, of which, 54 (n=10 332 949 individuals) were eligible for inclusion, including 12 suitable for meta-analysis. Included studies reported ethnic differences in age at T2D diagnosis, anthropometric measures, and factors associated with womens health. Compared to individuals of White ethnicity, people of diverse ethnic backgrounds had 2-4-fold higher incidence and prevalence of T2D and younger age of onset. Waist-to-hip ratio (WHR) was a better discriminator of T2D across all ethnic groups compared to body mass index (BMI). While the association between overweight/obese BMI and T2D was strongest for people of White ethnicity (OR 4.85 CI 3.53-6.68) followed by Black (OR 3.27 CI 2.48-4.30) and East Asian ethnicities (OR 3.06 CI 2.29-4.16), the association between WHR and T2D was strongest for people of Black (OR 2.74, CI 2.22-3.39) than for White ethnicities (OR 2.51, CI 2.30-2.74). Included studies highlighted the emerging importance of women-health-associated factors such as index of parity, birth weight and breastfeeding, especially among women of diverse ethnicities. ConclusionRatio measures of central adiposity may better identify T2D in ethnically diverse populations than measures of overall adiposity. Sex-specific factors must be considered when assessing T2D risk. FundingWellcome Trust Grant GPPG1K9R/218584/Z/19/Z.