Obstructive sleep apnea and incident cardiometabolic outcomes in Chinese

Abstract

Despite cumulating evidence suggesting an association between obstructive sleep apnea (OSA) and cardiometabolic consequences, the strength and significance of this association remain unclear. The treatment of OSA by CPAP has so far produced inconsistent results on the prevention of these adverse outcomes.

This thesis (comprising of four studies) aimed to investigate OSA and the effect of continuous positive airway pressure (CPAP) on long-term consequences, including incident type 2 diabetes (T2D), hypertension and major adverse cardiovascular events (MACEs).

This retrospective clinic-based cohort study was conducted using clinical and territory-wide electronic healthcare data in Hong Kong. For study one and two, I excluded the subjects with known T2D at baseline. For study three, I assembled the cohort free of known hypertension. The association with cardiometabolic outcomes and effect of CPAP therapy were examined by Cox-regression. The risk over the continuum of OSA-related indices was examined by cubic spline analysis.

A total of 1860 Chinese adults were included. Study one demonstrated that 152 out of 1206 subjects developed T2D at 7.3-year follow-up. OSA predicted T2D, independent of confounders including baseline obesity and body weight change. However, the relationship was non-linear. The rising trajectory of T2D risk attenuated albeit increase in apnea-hypopnea index (AHI). Regular long-term CPAP use in moderate-severe OSA patients significantly reduced the risk to that of subjects with no OSA.

Study two is the sequel of study one, which reviewed 1206 subjects and screened out seven determinants (gender, OSA severity, body mass index, waist circumference, comorbidities, smoking and family history) to build a T2D prognostic model. The developed nomogram effectively predicted the 3-, 5- and 10-year T2D risk. This nomogram was internally validated and yielded good performance. It is pending for external validation.

In study three, of 673 initially normotensive subjects, 174 developed hypertension within 7.7-year follow-up. Severe OSA was significantly associated with incident hypertension. Similar results were seen in the analysis of rapid eye movement-AHI, oxygen desaturation index and sleep time with oxygen saturation<90% (TST90). Regular CPAP use was associated with the reduction of hypertension incidence rate from 6.18 to 5.14 per 100 person-years. However, the risk in CPAP treated group was not significantly different from the untreated patients with moderate-severe OSA.

Study four with 8.3-year follow-up found that 278 of 1860 subjects experienced incident MACEs. AHI was associated with incident MACEs in univariate analysis, but this was no longer significant after adjustment. Among other OSA-related variables, TST90 was a significant predictor of MACEs, as was mean heart rate. A non-linear relationship was observed between TST90 and its HR of MACEs. Regular CPAP use was not associated with a lower rate of MACEs significantly.

In conclusion, OSA severity (measured by AHI) independently predicted incident diabetes and hypertension. OSA-related predictors (TST90 and mean heart rate) but not AHI significantly contributed to the risk for MACEs. The association between OSA and cardiometabolic outcomes might be non-linear. The long duration of regular treatment of OSA with CPAP was associated with a lower risk of incident T2D, but not incident hypertension or MACEs.