Association between the 2026 IDF-DAR risk category and Ramadan fasting intention among adults with diabetes in primary care: a cross-sectional study

Background The 2026 International Diabetes Federation–Diabetes and Ramadan (IDF-DAR) Risk Calculator supports Ramadan-related diabetes risk assessment, but its association with fasting intention in primary care is unclear. We examined the association between risk category and fasting intention and explored illness-perception and decision-related variables. Methods This single-centre cross-sectional study included 488 adults with diabetes assessed in a hospital-based family medicine clinic in Türkiye from 19 February to 19 March 2026, during Ramadan. The outcome was self-reported intention to fast during the remainder of Ramadan. Risk-category associations were examined using crude and sociodemographic-adjusted logistic regression. Age and four five-point decision-related indicators were modelled nonlinearly in an exploratory model that also included the Brief Illness Perception Questionnaire (B-IPQ) total score. An extended model added previous-year fasting experience. Results Overall, 178 participants (36.5%) intended to fast. Of 95 low-, 230 moderate-, and 163 high-risk participants, 38.9%, 46.5%, and 20.9%, respectively, intended to fast. Compared with low risk, high-risk classification was associated with lower odds of fasting intention in the crude model (OR 0.41, 95% CI 0.24–0.72) and the sociodemographic-adjusted model (OR 0.44, 95% CI 0.25–0.79). In exploratory Model 3, the high-risk estimate was imprecise (OR 1.16, 95% CI 0.36–3.72), whereas the moderate-risk estimate was elevated but sensitive to model specification (OR 3.77, 95% CI 1.34–10.56). The B-IPQ total score was not associated with fasting intention (OR 1.00, 95% CI 0.96–1.04). Perceived personal fasting risk and confidence in glycaemic control showed overall nonlinear associations (both p < 0.001). In extended Model 4, previous-year fasting was strongly associated with current fasting intention (yes vs. no: OR 25.00, 95% CI 7.92–78.95), while the high-risk estimate remained imprecise (OR 0.55, 95% CI 0.13–2.42). Conclusions In this hospital-based sample, approximately one in five high-risk participants intended to fast. High-risk classification was associated with lower fasting intention before exploratory contemporaneous variables were added; thereafter, the estimate was attenuated and imprecise. Prospective studies should examine actual fasting behaviour and clinical outcomes.