Effect of Weight Loss and Continuous Positive Airway Pressure on Obstructive Sleep Apnea and Metabolic Profile Stratified by Craniofacial Phenotype: A Randomized Clinical Trial

Rationale: Craniofacial structure is believed to modulate the effect of weight loss on obstructive sleep apnea (OSA), but whether this affects metabolic profile after weight loss compared with continuous positive airway pressure (CPAP) is unknown among obese Chinese patients with OSA. Objectives: To compare the change in metabolic profile between a lifestyle modification program (LMP), stratified by craniofacial phenotype, and CPAP therapy for 6 months. Methods: We randomly assigned 194 patients with body mass index ⩾ 25 kg/m² and moderate to severe OSA to participate in the LMP or receive CPAP therapy for 6 months in a 2:1 ratio. Assessments included computed tomography for assessing maxillomandibular volume (MMV), hsCRP (high-sensitivity C-reactive protein), and insulin sensitivity. Measurements and Main Results: Among 128 and 66 subjects in the LMP and CPAP groups, respectively, hsCRP was reduced more in the LMP group than the CPAP group (median [interquartile range], -0.7 [-1.4 to -0.0] vs. -0.3 [-0.9 to 0.4] mg/L; P = 0.012). More patients in the LMP group achieved low hsCRP (<1 mg/L) than the CPAP group (21.1% vs. 9.1%; P = 0.04). Insulin sensitivity improved only in the LMP group, with 3.1 (95% confidence interval, 1.5-6.6) times more patients with normal glucose regulation after intervention. The LMP group was stratified into LMP-small MMV (n = 64) and LMP-large MMV (n = 64) groups according to the median MMV value of 233.2 cm³. There was no significant difference in hsCRP (median [interquartile range], -0.7 [-1.3 to 0.1] vs. -0.7 [-1.5 to -0.2] mg/L; P = 0.884) and insulin sensitivity (median [interquartile range], 0.5 [-0.2 to 1.9] vs. 0.6 [0.1 to 2.0]; P = 0.4860) between the LMP-small MMV and LMP-large MMV groups. Conclusions: Weight reduction alleviated subclinical inflammation and improved insulin sensitivity more than CPAP among obese Chinese patients with moderate to severe OSA, and this effect was not influenced by craniofacial structure. Clinical trial registered with www.clinicaltrials.gov (NCT03287973).