To compare the effects of continuous positive airway pressure (CPAP), glucagon-like peptide-1 receptor agonists (GLP-1 RAs), their combination, and no active intervention on respiratory, sleepiness, and metabolic outcomes in adults with obstructive sleep apnea (OSA). We searched PubMed, Embase, and CENTRAL through August 2025 for randomised trials of CPAP, exenatide, liraglutide, tirzepatide, or their combinations. The primary endpoint was apnea-hypopnea index (AHI). Secondary endpoints were Epworth Sleepiness Scale (ESS), body mass index (BMI), systolic and diastolic blood pressure (SBP, DBP), fasting glucose, and glycated haemoglobin (HbA1c). Random-effects network meta-analyses estimated mean differences (MDs) with 95% confidence intervals (CIs). Treatments were ranked using SUCRA, and certainty of evidence was assessed with GRADE. Thirty-four trials including 3964 participants were eligible. CPAP produced the largest reduction in AHI versus no active intervention (MD -22.17 events/h; 95% CI -38.01 to -6.33) and improved ESS (MD -2.75; 95% CI -3.71 to -1.79). Liraglutide reduced BMI (MD -1.60 kg/m2; 95% CI -2.04 to -1.16) and HbA1c (MD -0.19%; 95% CI -0.25 to -0.13), whereas CPAP showed no meaningful metabolic effect. Liraglutide plus CPAP achieved the greatest BMI reduction (MD -2.00 kg/m2; 95% CI -3.49 to -0.51). No intervention significantly changed SBP, DBP, or fasting glucose. According to GRADE, certainty of evidence was moderate for CPAP effects on respiratory and sleepiness outcomes and for GLP-1 receptor agonists on BMI and HbA1c, and low for blood pressure and fasting glucose. CPAP is the most effective therapy for respiratory control, while GLP-1 receptor agonists primarily improve weight and glycaemic indices, supporting an integrated airway-metabolic approach to OSA management.