The cardiovascular consequences of impaired sleep

Introduction–Sleep is of critical importance to cardiovascular health, however data regarding sleep and several cardiovascular diseases are inconsistent or incomplete. This study aimed to investigate the association between two pairs of sleep-cardiovascular conditions: (1) short sleep duration and atrial fibrillation (AF), and (2) obstructive sleep apnea (OSA) and pulmonary hypertension (PH). Methods–From the population of patients undergoing diagnostic sleep study at one of six sleep centers within a large healthcare network, the 30,061 patients with a study between 1999-2005 were examined for prevalent AF. Logistic regression was used to model odds of AF for the primary exposure of total sleep time, after adjustment for demographics and comorbidities. The 27,589 patients with follow-up data and without AF at baseline were then examined for incident AF. Cox regression was used to model incident disease by total sleep time after adjustment. To examine the relation between OSA and PH, two separate cohorts were identified from the electronic record. The Echo Cohort contained 6,112 patients with both a sleep study and echocardiogram within 12 months; the right heart catheterization (RHC) Cohort matched 644 patients with a sleep study and RHC within 12 months. Linear models were created to assess the relation between apnea hypopnea index (AHI), as ln(AHI+1), and echocardiogram and RHC-derived hemodynamic outcomes after adjustment for age, sex, and body mass index. Results–In the AF analysis, 404 cases of prevalent AF were identified. After adjustment for patient characteristics, OSA, and cardiovascular risk factors, each 1-hr reduction in sleep duration was associated with a 1.17-fold [95% confidence interval [CI] 1.08, 1.27] increase in odds of AF in logistic models. In the longitudinal analysis, after a mean 5.2 years of follow-up, 1,820 incident AF cases were identified. Each 1-hour reduction in sleep duration was associated with a 1.09-fold [95% CI 1.05, 1.13] increase in risk for AF after adjustment in multivariable Cox models. For patients in the Echo Cohort, in a univariable linear model, each 1-unit increase in ln(AHI+1) was associated with a 0.6 mmHg [95% CI 0.2, 0.8] increase in RVSP. However, this was attenuated after adjustment for age, sex, and body mass index. In contrast, lower ejection fraction, higher left ventricular mass index, and worse of diastolic function were associated with ln(AHI+1) after adjustment. For patients in the RHC Cohort, mean pulmonary artery pressure was not associated with ln(AHI+1) in univariable or adjusted models. Among patients with PH and OSA and without another cause of pre-capillary PH, 90% had a post-capillary component to PH, suggesting that OSA is not a significant cause of isolated pre-capillary PH. Conclusions–Short sleep duration may be an important potentially modifiable risk factor for AF. Further work is needed to examine if interventions to prolong sleep may reduce AF risk. Findings from the OSA-PH inquiry suggest OSA is a mild and inconsistent cause of PH; clinicians caring for patients with PH and OSA should evaluate for heart failure as a cause of PH, rather than attributing PH to OSA alone. These combined findings point to important links between sleep health and cardiovascular disease, in particular AF and heart failure. Attention to healthy sleep may have broad public health implications for patients suffering from these common and morbid cardiovascular conditions