Diet as adjunctive therapy for sleep apnea risk: not only how much but also what to eat

Obstructive sleep apnea (OSA) affects 1 billion people worldwide, making it one of the most common chronic diseases in the world.¹ In the Western world, 34% of men and 17% of women experience OSA.² Importantly, the prevalence of OSA increases with adiposity and exceeds 50–80% among individuals with overweight and obesity.³ OSA triples the risk for cardiovascular diseases, including hypertension, coronary artery disease, stroke, heart failure, cardiac arrhythmias, and sudden cardiac death.⁴ Continuous positive airway pressure (CPAP), a standard therapy for OSA, is effective as a primary prevention for hypertension.⁵ However, CPAP failed to reduce cardiovascular risk in patients with OSA as a secondary prevention in 3 recent large randomized trials.^6–8 These disappointing results highlight the urgent need for alternative therapeutic approaches to OSA to reduce its cardiovascular complications. Lifestyle modifications, and dietary habits in particular, could be one such target. Considering that obesity is a major risk factor for OSA and that even modest weight loss is associated with reduction in OSA severity,⁹ targeting dietary patterns may be of particular importance in reducing cardiovascular risk associated with OSA. Accordingly, the American Heart Association recommends the inclusion of lifestyle interventions targeting weight loss to complement standard treatment of OSA.⁴

In this issue of the Journal of Clinical Sleep Medicine, Melaku and colleagues report on the association between diet quality and sleep risk using data from 4 cycles of the National Health and Nutrition Examination Survey (NHANES) occurring between 2005–2008 and 2015–2018.¹⁰ Diet and sleep apnea risk data, assessed using the USDA Automated Multiple-Pass method and the STOP-BANG questionnaire, respectively, were available for over 14,000 adults. From single-day dietary recalls, the authors calculated the Healthy Eating Index (HEI)¹¹ and Dietary Inflammatory Index (DII).¹² Half of the participants were at intermediate or high risk of sleep apnea. The mean HEI was 41.9 (range 2–90, out of 100; lower values indicate less healthy eating), reflecting low adherence to the Dietary Guidelines for Americans. The mean DII was 1.2 (range –5.39 to 5.38, theoretical maximum range of –8.87 to 7.98; greater values indicate greater inflammatory potential of the diet¹³). Although the DII is typically calculated using macronutrients, micronutrients, and intakes of various polyphenols, spices, and herbs associated with inflammatory status,¹³ in this case, the authors used only the 27 macronutrients and micronutrients available from NHANES.¹⁰

Participants with the healthiest diet quality based on HEI had 28% lower odds of being at high risk of sleep apnea compared with those in the lowest HEI quintile, while those with the highest DII score had 55% greater odds of being at high risk of sleep apnea compared with those in the lowest quintile. Sensitivity analyses combining intermediate and high sleep apnea risk scores on the STOP-BANG questionnaire or using OSA50 to define sleep apnea risk revealed similar findings. Moreover, both dietary indices were associated with snoring and tired components of the STOP-BANG questionnaire. Systemic inflammation, reflected by elevated C-reactive protein (CRP), and higher body mass index mediated the associations between diet quality and sleep apnea risk.

These findings are important because they are consistent with other studies showing associations between dietary quality and OSA risk. In a cohort of women recruited as part of the Go Red for Women Strategically Focused Research Network in New York City, we showed that higher adherence to a Mediterranean-style diet was associated with fewer sleep disturbances, a composite score of 9 factors including difficulty breathing and coughing or snoring at night, in both cross-sectional and longitudinal analyses over 1-year follow-up.¹⁴ Lopes and colleagues¹⁵ reported associations between the DII and daytime sleepiness in patients with mild to severe OSA. On the other hand, using the Berlin Questionnaire, Duan and colleagues¹⁶ reported an association between a healthy dietary index and reduced risk of sleep apnea in unadjusted analyses; those were no longer significant after adjusting for sociodemographics, diabetes, dyslipidemia, and other healthy lifestyle behaviors.

It is important to note that dietary quality indices were not associated with self-reported physician diagnosis of sleep apnea, which was available for almost 7,000 participants.¹⁰ This is unsurprising considering that more than 80% of patients with OSA remain undiagnosed in the general population.¹⁷ This staggering rate of underdiagnosis of OSA (4.9%) in a population where 50% individuals have an intermediate or high risk for OSA may underlie the lack of observed associations between physician-diagnosed OSA and the dietary quality indices. The low rate of diagnosis of OSA in the general population underscores the limitations of cross-sectional studies in investigating the associations of dietary patterns with the risk of OSA and its inflammatory and cardiovascular complications. The strength of this study is that at least the results from the STOP-BANG questionnaire were available. This OSA screening questionnaire has high sensitivity and specificity for the identification of OSA risk and is routinely used in clinical practice.¹⁸ Therefore, the results based on the STOP-BANG questionnaire scores likely reflect more accurately the prevalence of OSA in this study sample than self-reported, physician-made diagnosis.

Inflammatory markers such as CRP are reported to be increased in patients with OSA.^19,20 Mediation analysis cannot be conducted reliably in a cross-sectional sample where both the “cause” and “effect” are associated with the mediator (in this case, CRP) since it not possible to determine the temporal relationship between 2 variables (ie, whether unhealthy dietary patterns or OSA occurred first). Proinflammatory effects of OSA are well documented and may or may not be exacerbated by a proinflammatory diet.^19–21 It is cyclic hypoxemia in OSA and not obesity that promotes vascular inflammation in patients with obesity and OSA.²² In fact, CPAP is effective as a primary prevention for hypertension in patients with OSA despite weight gain that occurs with aging over a decade in patients with OSA.²³ In addition, the presence of OSA is also associated with unhealthy dietary patterns²⁴ and obesity, thereby confounding further its effects on the proinflammatory state observed in OSA.²⁵ Most studies reporting proinflammatory effects of dietary patterns do not account for possible OSA in their sample.^26,27 While this study did account for possible OSA using the STOP-BANG questionnaire, the directionality of the relationship between unhealthy dietary patterns, OSA, and inflammation cannot be determined.

Importantly, the extent to which diet quality was associated with sleep apnea risk in NHANES¹⁰ is on par with that of other lifestyle risk^16,28 and cardiometabolic risk factors.²⁸ Data from Melaku et al¹⁰ should be evaluated in clinical intervention studies to reveal the directionality of the association. Papandreou and colleagues²⁹ have demonstrated a benefit of following a healthy diet, in that case a Mediterranean diet, compared with a low-fat diet, on apnea-hypopnea index during rapid eye movement sleep. However, that study has confounding from differences between groups in change in waist circumference and physical activity. Nevertheless, until more robust data are available, patients at intermediate and high risk of sleep apnea would likely benefit, from multiple health standpoints, from recommendations on not only how much to eat but also on what to eat.

DISCLOSURE STATEMENT

Both authors have seen and approved the manuscript. Dr. St-Onge is supported in part by National Heart, Lung, and Blood Institute (NHLBI) R35HL155670 and R01HL142648. Dr. Jelic is supported in part by National Institutes of Health (NIH)/NHLBI R01HL106041 and R01HL137234. Dr. St-Onge has received consulting fees from Société des Produits Nestlé S.A. and PepsiCo, outside of the submitted manuscript. The authors report no conflicts of interest.