Background and purposeAbdominal obesity (AO) is a well-known independent risk factor for stroke in the general population although it remains unclear in the case of the elderly, especially in Chinese older patients with obstructive sleep apnea (OSA), considering the obesity paradox. This study aimed to investigate the association between AO and stroke among Chinese older patients with OSA.MethodsData were collected from January 2015 to October 2017, and 1,290 older patients (age 60–96 years) with OSA (apnea–hypopnea index ≥ 5 events/h on polysomnography) were consecutively enrolled from sleep centers at six hospitals, evaluated for AO defined as waist circumference (WC) using the standardized criteria for the Chinese population, and followed up prospectively for a median period of 42 months. Logistic regression and Cox regression analyses were used to determine the cross-sectional and longitudinal associations between AO and stroke risk in these participants and different groups of the severity of OSA.ResultsParticipants with AO had a higher prevalence of stroke at baseline. A higher incidence of stroke during a median follow-up period of 42 months in participants with AO than in participants without AO (12.4% vs. 6.8% and 8.3% vs. 2.4%, respectively; both P < 0.05) was predicted. Cross-sectional analysis revealed an association between AO and stroke (odds ratio [OR]1.96, 95% confidence interval [CI] 1.31–2.91), which was stronger among participants with moderate OSA only (OR 2.16, 95%CI 1.05–4.43). Cox regression analysis showed that, compared to participants without AO, participants with AO had a higher cumulative incidence of stroke (hazard ratio [HR] 2.16, 95% CI 1.12–4.04) during a median follow-up of 42 months, and this association was observed in patients with severe OSA only (HR 3.67, 95% CI 1.41–9.87) but not for individuals with mild OSA (HR = 1.84, 95% CI 0.43–6.23) and moderate OSA (HR = 1.98, 95% CI 0.73–6.45).ConclusionThe risk of stroke is associated with AO among Chinese older patients who have OSA, both at baseline and during follow-up, and the strength of the association varied by OSA severity. Active surveillance for early detection of AO could facilitate the implementation of stroke-preventive interventions in the Chinese older OSA population.