Prevalence of High Predicted 10‑Year Cardiovascular Disease Risk Among Hypertensive Asymptomatic Diabetic Outpatients in Dar es Salaam, Tanzania Using a Locally Adapted Framingham Risk Score

Background: Cardiovascular disease (CVD) risk is elevated in people with diabetes, but age‑ and sex‑specific distributions and modifiable drivers in Tanzanian clinic populations are incompletely described. Objective: To estimate the distribution of Framingham 10‑year CVD risk and identify independent predictors of high predicted risk (>20%) among adults with diabetes attending a tertiary clinic in Dar es Salaam. Methods: Cross‑sectional study of 249 adults with diabetes (163 women, 86 men). Framingham 10‑year CVD risk was categorized as Low (<10%), Intermediate (10–20%), and High (>20%). We summarized observed proportions and fitted multivariable multinomial logistic regression (reference = Low) to estimate relative risk ratios for Intermediate and High categories; a binary logistic model (High vs Not high) was used for clinical interpretability. Candidate predictors were prespecified and supplemented by bivariate screening; continuous variables were modelled on appropriate scales, interactions tested, and multicollinearity assessed (VIFs). Missing laboratory data (6–8%) were addressed with multiple imputation in sensitivity analyses. Model discrimination and calibration were evaluated. Results: Overall, 167/249 (67.1%, 95% CI 61.1–72.7) were classified as high 10‑year CVD risk. High‑risk prevalence was higher in men, 66/86 (76.7%, 95% CI 66.9–84.4), than in women, 101/163 (62.0%, 95% CI 54.7–68.8) (p=0.01). In multivariable analyses (High vs Low), independent predictors were older age (adjusted OR per year 1.07; 95% CI 1.04–1.10), waist circumference (per 10 cm 1.36; 95% CI 1.10–1.69), systolic blood pressure (per 10 mmHg 1.18; 95% CI 1.03–1.35), and LDL‑C (per 1 mmol/L 1.31; 95% CI 1.01–1.70). Regular physical activity was protective (adjusted OR 0.56; 95% CI 0.33–0.95). Sex was associated in univariable analysis but not after adjustment. Model diagnostics showed acceptable collinearity (all VIFs <3) and good discrimination for the binary contrast (AUC 0.78; 95% CI 0.73–0.83). Findings were robust to multiple imputation and sensitivity analyses. Conclusions: Two‑thirds of adults with diabetes in this clinic cohort have high predicted 10‑year CVD risk, with men transitioning to high risk earlier than women. Central adiposity, systolic blood pressure and LDL‑C are key modifiable drivers; physical activity is protective.