Dietary adherence and glycemic control in an exploratory cross-sectional study of patients with diabetes in Nouakchott, Mauritania

Dietary behaviors of people with diabetes in Mauritania have never been described and no brief adherence instrument reflecting the local food environment exists. This exploratory study described those behaviors and their association with glycemic control. One hundred patients were recruited consecutively from the follow-up register of the SOS Diabetes Association in Nouakchott, November 2023 to April 2024. The study was association-based rather than population-based and 98% of participants were urban. An 11-item adherence score was calculated, with good adherence defined a priori as ≥ 7/11. The primary outcome was fasting blood glucose ≥ 1.30 g/L, the upper limit of the American Diabetes Association preprandial target. Modified Poisson regression with robust variance estimated prevalence ratios (PR) adjusted for age, sex, body mass index and diabetes type. Item-level analyses were exploratory, with Benjamini–Hochberg correction. Poor glycemic control affected 53.0% of participants and good adherence 51.0%. Good adherence was not significantly associated with poor control (PR 0.81; 95% CI 0.55–1.20; p = 0.301). No alternative specification reached significance, including the continuous score (PR 0.90 per point; 95% CI 0.79–1.03) and HbA1c ≥ 7% as outcome (PR 0.70; 95% CI 0.46–1.07; n = 51). Post hoc power was 23%. No behavior survived correction. Item-level adherence ranged from 89.0% to 12.0%. No analysis identified a statistically conclusive association. Estimates lay consistently below the null but every interval included it, and the study was underpowered. The findings are hypothesis-generating and provide the first local baseline for the dietary dimension of diabetes care in Mauritania.