ObjectiveImpairment of circadian blood pressure (BP) patterns has been associated with cardiovascular risks and events in individuals with hypertension and in general populations, which are more likely to be found in obstructive sleep apnea (OSA). The aim of this study was to investigate the association of non-dipping BP pattern with new-onset diabetes in hypertensive patients with OSA, based on Urumqi Research on Sleep Apnea and Hypertension (UROSAH) data.Materials and methodsThis retrospective cohort study included 1841 hypertensive patients at least 18 years of age, who were diagnosed with OSA without baseline diabetes and had adequate ambulatory blood pressure monitoring (ABPM) data at enrollment. The exposure of interest for the present study was the circadian BP patterns, including non-dipping and dipping BP pattern, and the study outcome was defined as the time from baseline to new-onset diabetes. The associations between circadian BP patterns and new-onset diabetes were assessed using Cox proportional hazard models.ResultsAmong 1841 participants (mean age: 48.8 ± 10.5 years, 69.1% male), during the total follow-up of 12172 person-years with a median follow-up of 6.9 (inter quartile range: 6.0-8.0) years, 217 participants developed new-onset diabetes with an incidence rate of 17.8 per 1000 person-years. The proportion of non-dippers and dippers at enrollment in this cohort was 58.8% and 41.2%, respectively. Non-dippers were associated with higher risk of new-onset diabetes compared with dippers (full adjusted hazard ratio [HR]=1.53, 95% confidence interval [CI]: 1.14-2.06, P=0.005). Multiple subgroup and sensitivity analyses yielded similar results. We further explored the association of systolic and diastolic BP patterns with new-onset diabetes separately, and found that diastolic BP non-dippers were associated with higher risk of new-onset diabetes (full adjusted HR=1.54, 95% CI: 1.12-2.10, P=0.008), whereas for systolic BP non-dippers, the association was nonsignificant after adjusted the confounding covariates (full adjusted HR=1.35, 95% CI: 0.98-1.86, P=0.070).ConclusionsNon-dipping BP pattern is associated with an approximately 1.5-fold higher risk of new-onset diabetes in hypertensive patients with OSA, suggesting that non-dipping BP pattern may be an important clinical implication for the early prevention of diabetes in hypertensive patients with OSA.