Dear editor
I read the recent publication by Myllylä and colleagues [1] with great interest. Their matched case:control analysis analyzed the impact of OSA adherence on CVD events. The study was performed at only one center, but benefitted from its large sample size (1,030 matched pairs for a total of 2,060 patients) and impressively long follow-up time in both cases (who had used CPAP for a median of 6.4 hours/day over 8.7 years of follow-up) and controls (who had abandoned CPAP despite recommendations to continue CPAP; these patients had used CPAP for a median of only 4 months over 6.2 years of follow-up). CPAP-adherent patients demonstrated statistically and clinically significant reductions in CVD event risk compared to controls matched on gender, age, and apnea-hypopnea index.
Although these results might indicate that optimally used CPAP reduces CVD risk, I suspect the results are as likely explained by the ‘healthy adherer’ effect. This effect is best demonstrated in randomized controlled trials (RCTs) with a placebo arm.
For example, in an RCT of candesartan or placebo for heart failure patients, high adherence to candesartan was associated with a 35% reduction in risk of death compared to those with low adherence to candesartan [2]. On the surface, such data would suggest a dose-response relationship between candesartan and its clinical benefit. However, when the same analytic approach was applied to those assigned to the placebo arm, the risk of death was again much lower in those with high adherence to placebo. In fact, high adherence to placebo had a several-fold larger effect size on reduced mortality than candesartan itself. Given the lack of biologic plausibility that placebo reduces the risk of death in heart failure, the more likely explanation is that those who adhere to study drug (whether candesartan or placebo) have other healthy behaviors, socioeconomic factors, or other unmeasurable confounders that affect outcomes.
Similar results of high placebo adherence being associated with a lower risk of death were seen in the Women’s Health Initiative RCT of hormone replacement or placebo [3] and in a large multicenter RCT of chronic lung disease patients comparing active to placebo inhalers [4]. In summary, high adherence (whether to small molecule drugs, hormone replacement, inhalers, CPAP, or even placebo) may simply indicate the presence of health-promoting factors in adherers and provides little information regarding the efficacy of the actual treatment adhered to or not adhered to.
These and other examples demonstrate the pitfalls of attempting to use CPAP adherence to gain insight into the impact of CPAP on clinical outcomes. The results of Myllylä and colleagues’ study clearly provide important data that reinforce the notion that patients who abandon CPAP are at higher risk of poor outcomes. However, the results provide little support that CPAP treatment itself has any impact on CVD risk.
Acknowledgements
The views in this letter are solely those of the author and do not reflect the views of the United States Government, Department of Veterans Affairs, Minneapolis Veterans Affairs Health Care System, or the University of Minnesota.
Footnotes
Conflicts of Interest:
K. Kunisaki reports having previously served as a consultant to GlaxoSmithKline in 2018, and current receipt of funding from the United States National Institutes of Health for sleep-related research (R01 HL131049).
Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of a an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.
References
- 1.: Myllylä M, Hammais A, Stepanov M, Anttalainen U, Saaresranta T, Laitinen T (2019) Nonfatal and fatal cardiovascular disease events in CPAP compliant obstructive sleep apnea patients. Sleep Breath 2019 March 8. doi: 10.1007/s11325-019-01808-4. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.: Granger BB, Swedberg K, Ekman I, Granger CB, Olofsson B, McMurray JJ, Yusuf S, Michelson EL, Pfeffer MA; CHARM investigators (2005) Adherence to candesartan and placebo and outcomes in chronic heart failure in the CHARM programme: double-blind, randomised, controlled clinical trial. Lancet 366:2005–2011. [DOI] [PubMed] [Google Scholar]
- 3.: Curtis JR, Larson JC, Delzell E, Brookhart MA, Cadarette SM, Chlebowski R, Judd S, Safford M, Solomon DH, Lacroix AZ (2011) Placebo adherence, clinical outcomes, and mortality in the women’s health initiative randomized hormone therapy trials. Med Care 49:427–435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.: Vestbo J, Anderson JA, Calverley PM, Celli B, Ferguson GT, Jenkins C, Knobil K, Willits LR, Yates JC, Jones PW (2009) Adherence to inhaled therapy, mortality and hospital admission in COPD. Thorax 64:939–943. [DOI] [PubMed] [Google Scholar]