A Dietary and Lifestyle Intervention Improves Treatment Adherence and Clinical Outcomes in Overweight and Obese Patients with Obstructive Sleep Apnea: A Randomized, Controlled Trial

We conducted a parallel, randomized, controlled, follow-up study of consecutive patients who were admitted to the Sleep Disorders Center, Department of Respiratory Medicine, University of Crete Medical School, between December 2021 and March 2022. The inclusion criteria were (a) patients aged >18 years with newly diagnosed moderate to severe OSA (apnea-hypopnea index (AHI) ≥ 15 events/h) through an attended overnight polysomnography according to standard criteria, (b) overweight and obese (BMI > 25 kg/m²), (c) eligible for PAP treatment with a follow-up in-laboratory PAP titration with full polysomnography to establish the appropriate PAP settings, (d) with adherence data available in the 6 months after initiation of treatment, and (e) with an above-elementary school education. The exclusion criteria were refusal to participate, patients on PAP treatment, current participation in a weight loss program, central sleep apnea syndromes, obesity hypoventilation syndrome, restrictive ventilator syndromes, severe congestive heart failure, a history of life-threatening arrhythmias, severe cardiomyopathy, long-term oxygen therapy, chronic kidney disease, family or personal history of mental illness, drug or alcohol abuse, severe cognitive impairment, concurrent oncological diseases, pregnancy or lactation, recent hospitalization for acute or chronic respiratory disease, history of narcolepsy, or restless leg syndrome. All subjects provided written informed consent and ethical approval was provided by the University of Crete Research Ethics Committee (REC-UOC) (approval number: 158/29.11.2021). The trial was also registered on ClinicalTrials.gov with Trial registry: ClinicalTrials.gov; No.: NCT05881824.

Individuals were assigned (1:1) following simple randomization procedures (computerized random numbers) to a usual (standard) care group (SCG, n = 39) receiving usual follow-up care or an intervention group—Mediterranean diet group (MDG, n = 37) with follow-up care based on an additional behavioral intervention aiming at weight loss and increasing adherence to the Mediterranean diet (Figure 1). Following randomization, sequentially numbered envelopes that were opaque and sealed were utilized in the allocation concealment process, prepared by an individual not involved in the trial.

Blinding the patients or care team was not feasible since all patients were informed in the consent form that they would be randomized to either the MDG or SCG. Study investigators and the statistician involved in data analysis were blinded to the intervention.

Patients were followed for 6 months.

All patients underwent a detailed evaluation that included anthropometric parameters including BMI, and medical and sleep history, and comorbidities including physician-based diagnosis for depression, smoking history, and alcohol intake. Subjective daytime sleepiness, reflected by the Epworth Sleepiness Scale (ESS), and the patient’s level of depression reflected by Beck’s depression scale (BDI) were also recorded before PAP initiation. ESS, PAP adherence (hours of device use), BMI, and arterial blood pressure measurements were also evaluated pre- and post-intervention.

All patients attended a PAP clinic before treatment initiation, where they were given specific counseling and education on the proper use and maintenance of PAP and underwent personalized, formal mask fitting by a specialized nurse. The total time for the appointment in the PAP clinic was 20 min/patient. Once PAP was started, patients were reviewed in the outpatient sleep clinic at 1-month and at 3-month intervals during the first year, and every 6 months thereafter. During these appointments, a clinical assessment was made, and patients were further encouraged to use the device. In addition, all patients received oral healthy lifestyle advice and counseling on physical activity and sleep habits and had the opportunity to discuss other health issues related to the condition, such as weight reduction and smoking cessation. At each visit, the compliance data were downloaded from the PAP device and reviewed by the PAP clinic nurse together with the patients. Any concerns or questions, such as pressure sores, persistent air leakage, claustrophobia, nasal congestion, and other side effects resulting from the nasal mask interface that might lead to suboptimal compliance were addressed immediately by the PAP clinic nurse. Changes in the PAP setting, nose/face mask, or circuit were made after consultation with the responsible sleep physician if necessary. In every follow-up visit, residual symptoms, including residual sleepiness, or change in patient’s overall health status were recorded by the sleep nurse and sleep physician. This format adhered to a standardized approach according to our PAP clinic’s procedures [21].

All patients in the intervention group attended individual weekly 60–90 min sessions led by a dietitian in the first month and twice/month thereafter. In this group, all the features described above for the standard group were included, plus additional visits involving intensive dietitian-led behavioral intervention aiming at weight loss and increasing adherence to the Mediterranean diet [22]. Dietary behavior was assessed through the Food Frequency Questionnaire [23] and Mediterranean Diet Score [24] before PAP initiation. The dietitian conducted a 24 h recall for the participant, aiming to gather data on his customary dietary program, meal timings, preferred food quantities, and quality. At this point, questions were also posed regarding the consumption of water, alcohol, smoking, any allergies, special preferences or aversions, and about physical activity. In addition, a comprehensive approach was adopted by obtaining data on family, cultural, and professional background in order to guide subsequent steps towards the maximum possible outcome.

Subsequently, the dietitian provided every participant in this group with a document containing comprehensive suggestions for a nutritious diet, illustrating all food categories and outlining the specific amounts for each, clarifying which options are healthier for each group (e.g., lean dairy). The dietitian engaged in a dialogue with the participant to gather further details or inquiries about the diet, following which the former devised a personalized diet plan based on the latter’s requirements, inclinations, and unique characteristics.

Accordingly, this intervention was adjusted to fulfill the specific needs of these patients. Guidance in physical exercise, optimal sleep length, and sleep hygiene education were also given. Ultimately, a personalized therapeutic diet plan was implemented in this group.

PAP usage data included mask type (nasal or full face), number of nights on PAP, average use per night (hours), air leakage, and air pressure delivered. The usage of PAP, effective pressure, and residual AHI was monitored at 1, 6, 12 months after initiation, and patients were contacted by a trained sleep nurse. Patients were also encouraged to contact the telephone helpline during working hours. The humidification was defined based on patients’ feedback and was adjusted during the follow-up if necessary. In order to optimize PAP adherence unplanned visits were immediately arranged in case of low adherence to PAP therapy. Regular PAP compliance was defined as using the therapy for an average of 4 h a night for at least 70% of the nights [25]. However, in our study for optimal PAP adherence we used the cut-off point of six hours of PAP use [21,26].

A pilot study was conducted with 49 individuals to determine the sample size. With the data obtained from the pilot study, the sample size was determined as at least 63 individuals, to obtain at least 80% power to detect a significant difference in the follow-up mean hours of PAP use values between the MDG and the SCG, allowing for a type-I error rate of 0.05.

Results are presented as mean ± standard deviation (SD) for continuous variables if normally distributed and as median (25th–75th percentile) if not. Qualitative variables are presented as absolute number (percentage). For comparisons between groups, a two-tailed t-test for independent samples (for normally distributed data) or a Mann–Whitney U test (for non-normally distributed data) was utilized for continuous variables and the chi-square test for categorical variables. The analysis of covariance was used to test adjusted between-group differences at the end of the 6-month intervention. All models were adjusted for basic confounders, namely age and gender, baseline BMI, smoking status, questionnaires scores, OSA severity indices, and co-morbidities. Analyses were additionally adjusted for weight loss (expressed as BMI difference) to test the weight-loss independent impact of the dietary/lifestyle intervention implemented on PAP adherence. Factors associated with optimal adherence (use of the device ≥6 h) at the end of the 6 months’ follow-up were analyzed with bivariate logistic regression after adjustment for various potential basic explanatory confounders. We checked multicollinearity among the predictors using collinearity statistics to ensure that collinearity between predictor variables was in the acceptable range as indicated by the tolerance value variance inflation factor. Age was considered continuously and categorically, as age groups of 18–59 and >60 years, BMI was also considered continuously and categorically, as BMI groups of <30 and ≥30 kg/m². For the purpose of this analysis, the term cardiovascular disease (CVD) used, referred to any of the following conditions: coronary disease, stroke, atrial fibrillation, and heart failure. Results were considered significant when p values were <0.05. Data were analyzed using SPSS software (version 25, SPSS Inc., Chicago, IL, USA).

Participants’ socio-demographic and health status characteristics are outlined in Table 1. Most of the participants were men (78%), obese (77%), and current or former smokers (64%) with a medium educational level. The most prevalent diseases were hypertension (49%), followed by dyslipidemia (39%), CVD (27%), and COPD (22%). There was a significantly higher proportion of men in the intervention group compared to the control group. Other evaluated features remained relatively insignificant between both groups, such as age, other comorbidities, and smoking status (all p > 0.05), except for the presence of cardiovascular disease (CVD; 41 vs. 14%, p = 0.009).

Comparison of PSG parameters of the SCG and the MDG population showed no significant differences between groups (Table 2). Although no significant difference was noted between nocturnal and diurnal symptoms (Table 3), frequent awakenings reported were significantly higher in the MDG compared to the SCG (65 vs. 41%, p = 0.04).

Regarding lifestyle habits, the MDG exhibited a moderate level of adherence to the Mediterranean diet as assessed by MedDietScore (29 ± 5).

All patients continued to use their PAP at the end of the follow-up period. Auto-PAP was prescribed to the majority of participants with final levels at the end of the follow-up of PAP pressure of 9.2 for the SCG and 8.4 for the MDG (p = 0.04). Post-intervention PAP use was significantly higher in the MDG compared to the SCG (6.1 ± 1.2 vs. 5.4 ± 1.4, p = 0.02). Further analysis showed that this difference persisted after adjustments for age, gender, BMI, difference in BMI, ESS, difference in ESS, BDI score, OSA severity, and comorbidities (5.2 vs. 6.1, p = 0.03). Moreover, diet/lifestyle intervention was identified as one of the most significant predictive factors for optimal PAP adherence (OR = 5.458, 95% CI = 1.144–26.036, p = 0.03) (Table 4).

Regarding BMI, an increase was noted in the SCG, whereas a decrease (improvement) was observed in the MDG, (0.41 ± 1.8 vs. −0.75 ± 1.3, p = 0.02) (Figure 2). Specifically, during the 6-month follow-up, patients in the SCG demonstrated an average weight gain of 1.6% of their baseline body weight, while those in the MDG showcased a loss equivalent to 1.5% (p = 0.04). BMI difference, although attenuated, persisted after adjustments for age, gender, BMI, ESS, BDI score, OSA severity, PAP adherence, and comorbidities (0.23 vs. −0.55, p = 0.31).

In terms of blood pressure measurements a decrease was noticed only in the MDG (SBP −2.5 ± 11, DBP −1.1 ± 7.6) compared to the SCG (SBP 0.14 ± 9.2, DBP 0.40 ± 9.2) after the 6 months of the follow-up period. Nevertheless, the aforementioned changes did not exhibit statistical significance (p = 0.28 and p = 0.45, respectively) (Figure 3). However, after accounting for confounding variables, a substantial decline in SBP (−5.5 vs. 2.8, p = 0.014) and DBP (−4.0 vs. 2.5, p = 0.01) was specifically detected in the MDG versus the SCG.

A significant decrease in ESS was also noted in both groups. Nonetheless, no group manifested a significant predominance in this improvement (−4.5 vs. −3.5, p = 0.19), even following adjustment of confounders (Figure 4). Additionally, both groups displayed a noteworthy decline in the proportion of patients experiencing excessive daytime sleepiness (SCG: 19% vs. 62%, p < 0.001, MDG: 16% vs. 54%, p < 0.001). The proportion of residual sleepiness at the end of the follow-up period was comparable between the groups (16% versus 19%, p = 0.76).