Hypertension and obstructive sleep apnea are both highly prevalent, and share risk factors, so unsurprisingly, most people with OSA have HTN. This overlap has made it fiendishly hard to determine if OSA actually causes HTN. The widespread belief that OSA causes HTN was creeping close to dogma, egged on by cross-sectional studies that suggested an independent association (after controlling for age, BMI, etc), and therapy trials that showed a ~2 mm Hg reduction in blood pressure with treatment of OSA with continuous positive airway pressure (CPAP).
As Irene Cano-Pumarega et al point out, though, even an independent association between OSA and HTN does not prove causality. There could still be a hidden, unmeasured factor that causes both HTN and OSA. Longitudinal cohort studies are the gold standard for this type of question, and they’ve published theirs in the December AJRCCM.
They enrolled Spaniards selected randomly from the community who did not have sleep apnea by initial polysomnogram, and followed them for ~7.5 years (the Vitoria cohort). At the end, subjects underwent home polygraphy (not polysomnography). They had a cohort of 1,180 subjects left to analyze after excluding those with OSA at study onset or who were lost to followup.
Results: Increasing respiratory disturbance index was directly associated with risk for hypertension (p=0.001). However, after adjusting for age, sex, BMI, neck circumference, caffeine use and fitness level, the association disappeared.
Previous trials demonstrating the association between hypertension and OSA were recruited from clinic populations — where there was a high prevalence of both conditions. In contrast, this was a population study, with people recruited randomly from the community. As would be expected, people were healthier and thinner overall, with BMI ~26, and with a lower incidence of OSA than in previous positive studies. Here, the incidence of more severe OSA (as defined by RDI > 14) was only 13% (149 of 1180). That should make it harder to detect an association between OSA and HTN. To identify OSA and grade its severity at the end of the study, they used home polygraphy, which tends to underestimate the severity of OSA as compared to polysomnography. It’s a flaw, and means the incidence of OSA could have been higher, and more generally, we don’t know what the real incidence of OSA was.
Interestingly, they show the same dose response relationship (worse OSA = higher risk for HTN) as Peppard’s longitudinal study in NEJM. Peppard’s did use gold standard polysomnography (not home polygraphy), but was a smaller cohort. In the current study, though, the dose-response relationship disappeared after covariate adjustment.
Clinical Takeaway: How important is it if obstructive sleep apnea causes hypertension, or some risk factor(s) contributes to both? Authors suggest causality would be a rationale to treat asymptomatic people with mild-to-moderate OSA with CPAP; many would disagree (why not just check blood pressure frequently?). Regardless of whether there’s causality or association, the independent relationship seems relatively weak, as shown in this and previous studies. And CPAP’s effect on blood pressure is too small for most people to forgo additional medication for hypertension. Practically speaking, the question of causality vs. association for OSA and hypertension is a tempest in a teapot.
Cano-Pumarega I et al. Obstructive Sleep Apnea and Systemic Hypertension. Longitudinal Study in the General Population: the Vitoria Sleep Cohort. Am J Respir Crit Care Med 2011;184:1299-1304.