Response to Hudgel: Poor diet, poor sleep in sleep apnea, which is the cart and which is the horse?

We appreciate Dr. Hudgel’s detailed critique of our article. We wholeheartedly agree with the importance of helping patients with obstructive sleep apnea (OSA) meet optimal weight targets—our paper was never intended to discourage weight management, nor did we suggest that CPAP be used as a sole treatment for OSA. Rather, our analyses suggest the need for better health promotion programs for patients with OSA given the association of OSA with a poorer diet.

In considering Dr. Hudgel’s comments, it is important to note that our paper did not address obesity as an outcome, or CPAP as an intervention. Rather, we addressed the inter-relationships among OSA, sleep quality, and diet—the latter a health behavior that associates with health outcomes even after considering the effects of body weight. Through a systematic analysis of data from a large and well-characterized community cohort, we found that (1) individuals with OSA are more likely to eat diets enriched with red/processed meat and low in whole grains; (2) that a portion of this association is explained by reductions in N3 sleep; and (3) findings persisted even after adjusting for multiple confounders, including physical activity, socioeconomic variables, and even waist-hip ratio. These findings, although limited by the study’s cross-sectional design, supported the importance of improving low-quality diets in individuals with OSA. The results also led us to suggest that efforts at improving sleep quality (rather than targeting the AHI only) may contribute to improvements in diet and related cardiometabolic health. The need to identify modifiable factors that facilitate a healthy diet is supported by the well-known challenges in achieving optimal weight goals, with only a small minority of individuals in medical or behavioral weight reduction programs experiencing sustained weight loss. This work lays the groundwork for randomized controlled trials to test the impact of interventions addressing diet and sleep quality in patients with OSA.

Dr. Hudgel questioned the scientific premise of our study. However, as reviewed in our paper, there is a well-established literature that demonstrates that inadequate sleep can alter eating behaviors via neuro-humoral and/or hedonic pathways. Patients with OSA frequently have disturbed sleep but the association of disturbed sleep and diet in this population previously had not been investigated. Our interest in N3 sleep was based on experimental research demonstrating that reductions in N3 sleep lead to adverse metabolic and autonomic nervous system responses [1], and from observational studies showing that reduced N3 sleep is associated with higher adiposity [2], hypertension [3, 4], coronary artery calcification [5], and cardiac arrhythmias [6]. Dr. Hudgel cites the failures of several studies to show that CPAP improves inflammatory markers or CVD and can result in weight gain. However, these findings do not contradict the observations we made in an untreated population, but rather underscore the complexity of interpreting CPAP intervention studies due to variable adherence, duration of intervention, and potential off-target responses. In contradiction to the studies Hudgel quote, in patients with OSA and cardiovascular disease, inflammation, as measured by hs-CRP, decreased more after 12 weeks of CPAP compared with a controlled intervention [7]. Experimental studies performed under carefully controlled conditions also demonstrated that CPAP improved several markers of metabolic function [8, 9]. The basis for modest weight gain after CPAP use is poorly understood but unlikely relates to changes in sleep quality but rather to changes in energy expenditure during sleep.

Several methodological issues were also identified by Hudgel. He raised concerns over selection bias. As reported, the population was a large diverse sample of older adults, representative of the larger MESA cohort. Of course, associations in this community cohort may differ from that in a clinical sample with a higher proportion of individuals with symptomatic OSA, and we described limits to generalizability. He also raised concern over multiple statistical testing. Although we were reluctant to apply a simple Bonferroni correction factor due to the high degree of correlated outcome data, we feel confident that our key findings met conventional levels of significance. For example, the association with red/processed meat score had a p value of 0.0017, which is equivalent to a Bonferroni correction of 29 comparisons, which exceeded the number of statistical tests we performed. We agree with the limitations of the cross-sectional design, although note several factors that support a causal association among OSA, reduced N3 and poor diet: (1) The experimental literature strongly supports a causal association between inadequate sleep and unhealthy diet, with less data in the reverse direction; and (2) reduced N3 sleep is prospectively associated with incident hypertension [3, 10]. Nonetheless, we were careful to discuss the potential for reverse causality, which itself would be an important observation. In particular, if poor diet contributes to reduced N3 sleep, then interventions that improve diet may reduce severity of OSA by increasing time during deep sleep, when breathing tends to be more stable. We also note that our study outcome was not obesity, a risk factor for OSA. Rather, we focused on diet quality, which did not correlate with body mass index. Our results persisted after adjusting for adiposity.

In summary, we agree with many of Dr Hudgel’s points, and specifically with the need to provide patients with OSA comprehensive weight management tools, which we suggest should explicitly address healthy diet. We also suggest that sleep apnea management should include efforts to improve not only the AHI but also sleep quality, which likely influences multiple health outcomes, including dietary behaviors. Sleep apnea and diet are probably inter-related in more complex ways than is routinely considered. Additional studies are needed to fully understand the impact of CPAP on dietary behaviors and the interaction of CPAP and diet quality on health outcomes in patients with OSA. We hope our article will stimulate further research investigating the neurophysiological pathways that may mediate associations among OSA, sleep quality, and cardiometabolic health and support the development of clinical trials that test multimodal interventions in patients with OSA.

Sincerely,

Funding

MESA is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with MESA investigators. Support for MESA is provided by contracts HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, N01-HC-95169, UL1-TR-000040, UL1-TR-001079, UL1-TR-001881, and DK06349. Funding support for the Sleep Polysomnography dataset was provided by grant HL56984. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health. Dr. Susan Redline is partly funded through the National Heart, Lung, and Blood Institute (1R35 HL135818-01). Drs. Johnson-Morgan and Redline report grants from NIH, during the conduct of this study. Dr. Shea reports grants from NHLBI, during the conduct of the study. Dr. St-Onge report grants from NHLBI and the American Heart Association Go Red for Women Strategically Focused Research Network during the conduct of this study.