Anthropometric indices and their cut-off points in relation to type 2 diabetes among Ghanaian migrants and non-migrants: The RODAM study

•WHR performed better than WC and BMI as anthropometric indices for assessing the burden of T2DM among Ghanaians.•Optimal cut-offs for WC, BMI and WHR among Ghanaians varied based on sex and level of urbanization.•Cut-offs for BMI and WC were lower than WHO established standards among Ghanaians. To compare body mass index (BMI), waist circumference (WC) and waist-to-hip ratio (WHR) as determinants of type 2 diabetes (T2DM) and determine optimal cut-offs in a sub-Saharan African population. Data from the RODAM study including Ghanaians aged 25–70 living in rural Ghana, urban Ghana and Europe were used. Logistic regression was used to assess associations between BMI, WC, WHR and T2DM status, by sex and site. Area under the curve (AUC) were constructed to discriminate between indices and establish performance and cut-off values. WHR had the strongest association with T2DM in men and women across sites, except for rural men. The highest adjusted odds ratio (aOR) and AUC were in rural women for WHR (aOR = 2.09, 95%CI = 1.47–2.99; AUC = 0.71). Among migrants, WHR had higher AUCs compared with BMI (p