Background Adiposity is a major risk factor for T2D, and is associated with diabetes markers. Saudi Arabia is in the top ten countries as to the prevalence of obesity, with an overall prevalence of 35.6% among adults. Although obesity is a major health problem in Saudi, the associations between anthropometric measures and T2D have not been extensively investigated in population-based studies within the Saudi context. The aim of this study was to examine the cross-sectional associations between anthropometric measures such as body mass index (BMI), waist circumference (WC) and T2D prevalence in Saudi adults.

Methods This study was based on 2355 Saudis (≥18 years), part of a larger sample of 14,060 Saudis randomly selected from a roster of 150 primary health care centres (PHCC’s) of Riyadh city, Saudi Arabia. For each PHCC, a physician invited participants from randomly selected households (alternate houses). The number of subjects sampled from each PHCC was dependent on the total population assigned to the respective PHCC. Physicians confirmed diabetes diagnosis by either following World Health Organisation proposed cut-off values (fasting plasma glucose levels ≥7.0 mmol/L) and/or self-report of anti-diabetes medications. For the overall sample and for men and women separately, logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CI) of T2D prevalence across body mass index (BMI) groups and WC tertiles. The reference category was the lowest. Covariates included: age, sex, marital status, education level, family history of diabetes, physical activity, caloric intake, and history of gestational diabetes (GDM) in females.

Results The overall prevalence of T2D was 29.3% (25.4% in females versus 32.9% in males). In obese females (BMI ≥ 30 kg/m²) the OR for diagnosed diabetes was 3.39 (95% CI, 1.5–7.6, p = 0.001). While in males, a statistically significant association was observed for increased WC (OR 1.99, 95% CI, 1.2–3.1, p = 0.006) and T2D. In the overall sample, a direct and statistically significant (p = 0.014) association was observed between WC and the OR for diabetes within each group of BMI, whereas the associations of BMI with diabetes attenuated (p = 0.572) after accounting for WC.

Conclusion Results suggest that measures of central adiposity, such as WC, are associated with diabetes, independently of BMI. These results may have clinical and public health implications in terms of targeted preventive strategies to address the ongoing diabetes epidemic in this context.