cpap does not reduce cardiovasc risk

A recent article looked at patients with moderate-to-severe obstructive sleep apnea (OSA) and documented cardiovascular disease (CVD), finding no reduced risk of adverse cardiovascular outcomes by using CPAP (see osa cpap not dec cv risk nejm2016 in dropbox, or DOI: 10.1056/NEJMoa1606599). details:

--2717 patients aged 45-75 who had moderate-to-severe OSA as well as coronary or cerebrovascular disease were randomized to receive CPAP treatment plus usual care (CPAP group) or usual care alone (usual-care group)

--mean age 61, 81% male, 64% Asian/25% white, 51% with coronary artery disease/49% cerebrovascular disease, 79% hypertensive, 44% stroke, 33% MI, 30% diabetic, 15% smokers, 78% on BP meds/57% statins/75% aspirin/27% diabetic meds, BMI 29, apnea-hypopnea index (AHI) 29 (moderate-to-severe obstructive sleep apnea), 84% snoring almost every day, but minimal daytime sleepiness

--primary composite end point was: death from cardiovascular causes, myocardial infarction, stroke, or hospitalization for unstable angina, heart failure, or transient ischemic attack

--secondary end points included other cardiovascular outcomes, health-related quality of life, snoring symptoms, daytime sleepiness, and mood.

--results:

    --In the CPAP group, the mean duration of adherence to CPAP therapy was 3.3 hours per night (in beginning 4.4, decreasing to 3.5 hours/night by 12 months), and the mean AHI decreased from 29.0 events per hour at baseline to 3.7 events per hour during follow-up, reflecting good control. 42% adhered to treatment for >4 hrs/night.

    --After a mean follow-up of 3.7 years:

        --primary end-point event occurred in 229 participants in the CPAP group (17.0%) and in 207 participants in the usual-care group (15.4%) [HR 1.10; (0.91 to 1.32) P = 0.34]. No significant effect on any individual or other composite cardiovascular end point was observed, and the trend was actually for more events in the CPAP group. 

        --CPAP significantly reduced snoring and daytime sleepiness and improved health-related quality of life and mood, with greater reductions in anxiety and depression (25-30% lower). no difference in road-traffic accidents or accidents caused by injury

    --no difference in subgroups: region (China vs other), age, sex, severity of OSA, BMI, daytime sleepiness, type of CVD, diabetes.

    --there were differences in patient characteristics in those who used CPAP > vs < 4 hours/night: more were Caucasian/European, men, had baseline CAD, hypertension, and fewer had TIA/stroke. Using propensity-score matching, there were 86 events in the CPAP group and 98 in usual-care: non-significant (ie, no benefit even in the subgroup who used CPAP for > 4 hours/night)

commentary:

--the indications for CPAP seem to continue to decrease over time. There is pretty good documentation that OSA is associated with hypoxemia and sympathetic stimulation, elevated blood pressure especially in the AM, inflammation, oxidative stress, metabolic syndrome/insulin resistance/type 2 diabetes (which seems to independent of the obesity common in both conditions), 2-3x increased risk of NAFLD (also apparently independent of obesity/diabetes), hypercoagulation, right-sided heart failure. There are several observational studies suggesting that there is an association with cardiovascular events (esp stroke). And there are studies finding that using CPAP lowers the blood pressure, but only a little (2-3 mmHg, but can be up to 6-7 mmHg in those with resistant hypertension, though a recent RCT found a 3-4mmHg decrease); and improves insulin sensitivity and endothelial function. And observational studies show that it lowers cardiovascular events in those adherent to treatment. Another study found a 64% relative and 28.5% absolute risk reduction in those using CPAP in a nonrandomized but prospective 6 year study, concluding “OSA treatment should be considered for primary and secondary cardiovascular prevention, even in milder OSA” (see Buchner NJ. Am J Crit Care Med 2007; 176 (12): 1274), advocating for more aggressive CPAP usage.

--the concerns about CPAP are: that OSA is remarkably common, including 20-30% of males and 10-15% of females, though this depends on how OSA is defined and what the cutpoints are for the AHI (these percentages are if one uses an AHI>5/hour cutpoint); OSA seems to be much more common in those with underlying cardiovascular disease, cited at 40-60%; but CPAP is a pretty significant intervention (very difficult for many patients to use CPAP machines/uncomfortable, and in this study, the average was only 3.3 hours/night, which is similar to general clinical practice). Using CPAP therefore only makes sense if there is real clinical improvement. 

--a prior study was done in Spain (see Barbe F. JAMA 2012; 307(20): 2161) which randomized 357 patients to CPAP, 366 control. All patients had AHI of  >= 20/hr and no significant daytime symptoms; and they excluded anyone with prior cardiovascular event (ie, much lower cardiovascular risk than those in the current study). In this 4-year study, as with the above study, there was no difference in cardiovascular events. Of note there was no difference in events when assessing those with the highest AHI levels or % of time with SaO2 <90%. Median CPAP usage was 5 hours, but those who used the CPAP >4 hours did have a 28% [HR 0-.72 (0.52-0.98), p=0.04)] lower incidence of hypertension or cardiovascular event. But looking specifically at cardiovascular events, there was a nonsignificant trend to lower CVD events (I should add here that 60% used the CPAP>4 hours, and this more adherent group may have done other healthy changes to decrease their risk of cardiovascular events. and these healthier things may have influenced the outcome more than the CPAP). similarly if look at amount of time with SaO2<90%, the less time the fewer events​.

--so, overall, this current study showed very little cardiovascular benefit in many patients at high risk for CVD events and with pretty severe OSA. Although it is true that people used the CPAP only for 3-4 hours/night, this seems to be pretty much the range of usual CPAP usage, and they did find a dramatic decrease in AHI. My sense is that the indications for CPAP, a pretty significant and difficult intervention for many patients, have decreased considerably over time (early commentators suggesting that we should use it to prevent right-heart failure, or improve hypertension control, or decrease cardiovascular events/metabolic syndrome, etc). Now, by far the major argument for CPAP for the vast majority of people, it seems to me, is for symptom control when patients are not sleeping well and having daytime somnolence or functional/psychological  impairment​ (depression/anxiety, etc) from inadequate sleep.