Analysis of Recent Papers in Hypertension

Treatment of Obstructive Sleep Apnea With Continuous Positive Airway Pressure Appears to Decrease the Incidence of Incident Hypertension


  • Michael J. Bloch MD,

    1. From the Seinsheimer Cardiovascular Health Program, Division of General Internal Medicine/Geriatrics, Medical University of South Carolina, Ralph H. Johnson VA Medical Center, Charleston, SC;
    Search for more papers by this author
  • Jan N. Basile MD

    1. Department of Internal Medicine, University of Nevada School of Medicine, Reno, NV
    Search for more papers by this author

Michael J. Bloch, MD, MedSchool Associates North, University of Nevada School of Medicine, 1500 East Second Street, Suite 302 Reno, NV 89520

Studies Reviewed

Marin JM, Augusti A, Villar I, et al. Association between treated and untreated obstructive sleep apnea and risk of hypertension. JAMA. 2012;307(20):2169–2176.1

Barbe F, Duran-Cantolla J, Sanchez-de-la-Torre M, et al. Effect of continuous positive airway pressure on the incidence of hypertension and cardiovascular events in nonsleepy patients with obstructive sleep apnea: a randomized trial. JAMA. 2012;307(20):2161–2168.2


Obstructive sleep apnea (OSA) is a common condition that has been associated with an increased risk of hypertension in observational studies. The Wisconsin Sleep Cohort Study1 found a strong association between the severity of OSA and the risk of developing hypertension 4 years later. The Sleep Heart Health Study2 also demonstrated an association between OSA severity and incident hypertension, but the effect was attenuated when fully adjusted for other risk factors.

A number of studies have previously attempted to determine the effect of continuous positive airway pressure (CPAP) on blood pressure (BP). These previous attempts found a modest (around 2 mmHg) effect on BP with CPAP, but were limited by a relatively small sample size and short follow-up.3,4 Accordingly, it has been difficult for clinicians to determine whether CPAP use can decrease the risk of developing hypertension in patients with documented OSA. Designing studies to answer this question has been hampered by the known benefits of CPAP therapy on daytime sleepiness and quality of life in people with symptomatic OSA, essentially precluding the clinical equipose required to conduct randomized controlled clinical trials in affected patients.

While certainly not definitive, two recent reports from Spain shed some light on this important clinical question. The first study from the Zagaoza Sleep Cohort, after adjusting for potential confounders, observed that after 12.2 years of follow-up there was a strong dose-response relationship between OSA severity and incident hypertension as well as a strong relationship between CPAP and a reduced incidence of newly diagnosed hypertension.

The second study, a prospective randomized trial of patients with OSA who did not have significant daytime sleepiness, found a trend toward decreased incidence of hypertension and cardiovascular events in patients randomized to CPAP. While the effect on the entire study cohort was not significant, largely owing to insufficient power, there was a statistically significant reduction in both incident hypertension and cardiovascular events in subgroups who had more severe disease and who were more adherent with CPAP therapy.

Discussed below, both of these studies have significant methodological shortcomings that limit our ability to draw firm conclusions for clinical practice. They do, however, support the hypothesis that treatment of OSA in patients who are not yet hypertensive may prevent its development. Given the high and overlapping prevalence of OSA and hypertension, further research is indicated to determine whether (1) the threat of incident hypertension is high enough to warrant treatment of OSA in patients without significant daytime sleepiness or other clinical characteristics that would otherwise dictate using CPAP, and (2) the magnitude of overall clinical benefit is high enough that we should be screening for and potentially treating OSA in patients with hypertension.

While we would like to see good-quality prospective randomized clinical trials designed to answer these key clinical questions, such studies will be difficult to perform. Because the presence of daytime sleepiness and other symptoms of OSA are improved with effective treatment, it is difficult to design a clinical trial where these patients can ethically be randomized to placebo. Additionally, given the known detrimental effects of uncontrolled hypertension, it is difficult not to intensify pharmacologic antihypertensive therapy in patients with OSA enrolled in a clinical trial.

In the absence of more definitive data, it seems premature to prescribe CPAP only for BP control or the prevention of hypertension in patients with OSA. It does, however, seem reasonable for clinicians to continue to promote the potential benefits of CPAP on BP when prescribing it for other well-established clinical indications, such as daytime sleepiness. As currently recommended, clinicians should screen for daytime sleepiness using standardized tools such as the Epworth Sleepiness Scale in patients with hypertension and prehypertension.

Observational Study of Association Between Treated and Untreated OSA and the Risk of Developing Hypertension

This observational study included 1889 patients referred to a single sleep center in Zaragoza, Spain, for evaluation of suspected sleep-disordered breathing. Exclusion criteria included a diagnosis of hypertension, treatment for hypertension, or history of cardiovascular or other significant comorbid conditions. Patients with an apnea-hypopnea index (AHI) of at least 5 were diagnosed with OSA and, consistent with Spanish guidelines, CPAP therapy was offered to all patients with an AHI of at least 30 or an AHI between 5 and 29 with lifestyle limiting daytime sleepiness.

Patients were divided into 5 groups based on their sleep study results and their treatment status (1) AHI <5 (controls without OSA), (2) presence of OSA but without meeting criteria for CPAP, (3) eligible for CPAP but declined, (4) started CPAP but not adherent, and (5) treated and adherent with CPAP therapy. Clinical follow-up continued for a mean of 12.2 years, with incident hypertension (defined as a diagnosis of hypertension by primary physician, treatment for hypertension initiated, or BP >140/90 mm Hg) as the primary endpoint.

The first important finding was that the incidence of new-onset hypertension correlated with the severity of OSA as measured by AHI. Approximately 80% of patients with severe OSA (AHI >30) developed incident hypertension as compared with only about 50% without OSA (AHI <5).

In multivariate models that corrected for recognized potential confounders, patients with OSA who declined CPAP or were nonadherent with CPAP were nearly twice as likely to develop incident hypertension as controls without OSA. Patients who had OSA not severe enough to warrant treatment with CPAP had an intermediate risk of developing hypertension. Interestingly, in most models, patients who were treated and adherent with CPAP therapy actually had no increased risk of developing hypertension as compared with controls without OSA.

While there certainly may be unrecognized confounding variables in this analysis, these data add strength to the still unproven hypothesis that treatment of more severe OSA with CPAP decreases the incidence of hypertension in patients who are not yet hypertensive.

Randomized Controlled Trial of CPAP on the Incidence of Hypertension and Cardiovascular Events in Nonsleepy Patients With OSA

As part of this multicenter, parallel-group, randomized controlled trial at 14 teaching hospitals in Spain between May 2004 and May 2006, 725 patients with evidence of OSA (defined as an AHI of at least 20) without excessive daytime sleepiness (defined as a score on the Epworth Sleepiness Scale of <10) were randomly allocated to receive CPAP or no active intervention. The main outcome measure was either the development of hypertension in patients who did not already have hypertension or incident cardiovascular events. Follow-up was for a median of 4.0 years.

Unfortunately, 50% of patients already had baseline hypertension and were not eligible for the endpoint of incident hypertension, and cardiovascular events were uncommon, both of which significantly decreased the power to detect a statistical difference with CPAP therapy. In the group allocated to CPAP, there were 68 cases of incident hypertension and 28 cardiovascular events during follow-up compared with 79 cases of incident hypertension and 31 cardiovascular events in the control group. These differences were not significant.

In addition, the concept of clinical equipose limited enrollment to patients with only minimal symptoms and relatively mild OSA severity. The issue of adherence to CPAP therapy may have also influenced the findings. An intriguing post hoc analysis suggested that: (1) patients who were adherent with CPAP therapy for at least 4 hours per night had a significant decrease in the incidence of the combined endpoint of incident hypertension or cardiovascular events, and (2) patients with evidence of more severe OSA (defined as worse nocturnal oxygen desaturation) who were adherent with CPAP for less than 4 hours per night showed a higher risk of incident hypertension and cardiovascular disease than the control group.

Final Thoughts

While the results of these two clinical trials are far from definitive, they lend increased support to the hypothesis that CPAP, when used regularly, in patients with OSA may attenuate the risk of developing hypertension, especially with more severe OSA. While these findings should not lead to any change in the current indications for CPAP therapy in patients with OSA, they do remind us that we should be screening for OSA in patients at risk for hypertension while continuing to promote better adherence to CPAP therapy when it is indicated based on other clinical criteria.

A related important clinical question that is not addressed by these data but needs clarification in future clinical trials is the magnitude of the effect of BP reduction and/or reduction in cardiovascular events in patients with concomitant established hypertension and OSA who are adherent to CPAP therapy. Specifically, it is important to know whether the use of CPAP in patients with OSA who do not have established indications for its use offers enough meaningful reduction in BP that it be specifically recommended for that purpose. Furthermore, it would be interesting to determine whether the BP-lowering effect of CPAP offers equivalent or greater benefit in reducing major cardiovascular events compared with pharmacologic therapy for hypertension.

The types of studies to more definitively determine the effects of CPAP therapy on BP and cardiovascular events are difficult to design and administer. But with the increasing prevalence of OSA and hypertension, and the overlap of these two conditions in our patient population, the planning of such studies should be a priority.