Prevalence and predictors of abnormal dyspnea during exercise in persons with type 2 diabetes: a cross-sectional study (PROTECTION trial)

Abstract

Abstract Aim To determine the prevalence, characteristics, and predictors of abnormal exertional dyspnea in T2D during a maximal exercise test. Methods 97 adults with non-insulin-dependent T2D underwent spirometry, maximal upright cardiopulmonary exercise test with Borg 0–10 dyspnea rating, cardiopulmonary exercise echocardiography, assessments of physical activity, body composition, blood pressure, bloodwork, and handgrip strength. Dyspnea scores above the 95th percentile of predicted were considered abnormal. Between-group differences were assessed with ANOVA, Mann-Whitney U, or χ² tests, and a final predictive model for abnormal dyspnea was created with a binary logistic regression. Results The prevalence of abnormal exertional dyspnea was 16.5%. These persons had lower lung function, peak oxygen uptake, higher semisupine left atrial volume index at rest, and higher semisupine estimated left ventricle (LV) filling pressures with lower biventricular systolic indices at rest and low-intense exercise ( p < 0.05). After exploratory analysis on 19 potential predictors, the final logistic regression model included FEV1(%predicted) and E/e′avgrest. Each unit increase in E/e′avgrest raised the odds of abnormal dyspnea by 44% (OR 1.44, 95% CI 1.10–1.90; p = 0.009), and each percent increase in FEV1(%predicted) reduced the odds by 5% (OR 0.95, 95% CI 0.91–0.99; p = 0.023). The model explained 15–26% of variance (Nagelkerke R²=0.255). Conclusion Approximately one in six persons with well-controlled, non-insulin-dependent T2D had abnormal exertional dyspnea during maximal exercise testing. This was associated with worse forced expiratory volumes and indices of estimated LV filling pressures. Assessment of dyspnea during exercise using the Borg 0–10 scale is a simple, low-cost method for identifying persons who may benefit from targeted cardiac and pulmonary evaluation. Trial registration PROTECTION trial (NCT05023538).

The authors' abstract, as published at the source. Cardiovascular Diabetology, 2026 · DOI ↗