Association between body mass index and effectiveness of continuous positive airway pressure in patients with obstructive sleep apnea: a retrospective study

Xinmei Huang,Liyan Bao,Xuxia Tang,Jun Shen,Xupei Ni,Yanfei Shen
DOI: https://doi.org/10.1007/s11325-019-01960-x
2019-11-18
Sleep And Breathing
Abstract:PurposeIneffective use of continuous positive airway pressure (CPAP) therapy can result in inconvenience and additional costs in patients with obstructive sleep apnea (OSA). This study investigated the predictive value of body mass index (BMI) to assess the efficacy of CPAP in patients with OSA.MethodsData were extracted from a retrospective study performed in Silkeborg Hospital. The primary outcome was the improvement of Apnea-Hypopnea Index (AHI) after CPAP treatment. Association between BMI and improvement of AHI was assessed by multivariable linear regression. Interactions between BMI, baseline AHI severity (≥ 30 or < 30), and diabetes were also evaluated.ResultsFour hundred eighty-one patients were included in the study. After adjusting for confounders, high BMI (coefficient [coef], 0.80; 95% confidence interval [CI], 0.59–1.00; p < 0.001) and high AHI severity (AHI ≥ 30) (coef, 29.2; 95% CI, 26.7–31.7; p < 0.001) were associated with greater improvement of AHI after CPAP treatment, while diabetes was associated with less improvement of AHI (coef, − 4.91; 95% CI, − 9.40 to − 0.42; p = 0.032). Baseline AHI severity, diabetes, and BMI showed significant interactions (p < 0.001). On subgroup analysis, the association between BMI and improvement of AHI remained significant only in patients belonging to high AHI severity subgroup (coef, 1.18; 95% CI, 0.8–1.49; p < 0.001) and that without diabetes (coef, 1.42; 95% CI, 1.11–1.72; p < 0.001).ConclusionsPatients with OSA having high BMI, without diabetes, are more likely to benefit from CPAP therapy. Future studies should explore the predictors of the efficacy of CPAP in more depth.
clinical neurology,respiratory system
What problem does this paper attempt to address?