Frontiers in Nutrition · Oct 31, 2026
Metabolic and Nutritional Profiles According to Glycemic Status in Obese Tunisian Adults: A Cross-Sectional Study
By Donia JERAB, NADIA KHERIJI, JMEL Haifa et al.
Background: Obesity and type 2 diabetes have reached epidemic proportions in Tunisia, yet metabolic and nutritional profiles across the glycemic spectrum remain poorly characterized among obese individuals. Objective: To compare anthropometric, metabolic, and dietary characteristics of obese Tunisian adults according to glycemic status. Methods: In this cross-sectional study, 210 obese adults (BMI ≥30 kg/m²) drawn from a survey based on stratified random sampling in Zaghouan, Tunisia, were classified as normoglycemic (n=34), prediabetic (n=90), or diabetic (n=86) according to ADA criteria. Dietary intake was assessed in 154 participants with a single 24-hour recall, and food consumption frequencies with a food frequency questionnaire. Micronutrient intakes were analyzed in absolute terms and after energy adjustment (density and residual methods), using non-parametric tests with false discovery rate (FDR) correction. Sensitivity analyses included adjustment for age. Results: Diabetic participants were older and more often illiterate than normoglycemic participants. BMI, systolic blood pressure, and triglycerides differed across groups independently of age, whereas differences in diastolic blood pressure and C-reactive protein were largely explained by age. Macronutrient distribution and food consumption frequencies did not differ after FDR correction. Absolute micronutrient intakes tended to be higher in normoglycemic participants, but none of these differences survived correction for multiple testing. After energy adjustment, riboflavin, vitamin B5, magnesium, phosphorus, and iron densities differed across groups (q<0.05), prediabetic participants having the lowest values, significantly lower than those of diabetic participants. Differences for riboflavin, vitamin B5, and iron persisted after adjustment for age. Nutrient densities did not differ significantly between newly diagnosed and known diabetic participants, and results were unchanged after exclusion of implausible energy reporters. Conclusions: In this sample of obese Tunisian adults, the tendency toward higher absolute micronutrient intakes in normoglycemic participants was consistent with their higher reported energy intake. After energy adjustment, prediabetic participants had the lowest densities of two B vitamins and three minerals, most of which were also lower than in diabetic participants unaware of their diagnosis. Diet quality may therefore deserve attention at the prediabetic stage, a hypothesis that should be confirmed in longitudinal studies with repeated dietary assessments and nutritional biomarkers.