Meta-analyses of the Association of Sleep Apnea with Insulin Resistance, and the Effects of CPAP on HOMA-IR, Adiponectin, and Visceral Adipose Fat

Abstract

Objective: We sought to conduct an updated meta-analysis of randomized controlled trials (RCTs) on the effect of continuous positive airway pressure (CPAP) on insulin resistance, as measured by homeostasis model assessment of insulin resistance (HOMA-IR), visceral abdominal fat (VAF), and adiponectin. Additionally, we performed a separate meta-analysis and meta-regression of studies on the association of insulin resistance and obstructive sleep apnea (OSA).

Methods: All included studies were searched from PubMed (from conception to March 15, 2014). Data were pooled across all included RCTs as the mean difference in HOMA-IR and VAF, and as the standardized mean difference in the case of adiponectin analysis. From the included case-control studies, data on the difference of HOMA-IR between cases and controls were pooled across all studies, as the standardized mean difference (SMD).

Results: There was a significant difference in HOMA-IR (-0.43 [95% CIs: -0.75 to -0.11], p = 0.008) between CPAP treated and non CPAP treated participants. However, there was no significant difference in VAF or adiponectin; (-47.93 [95% CI: -112.58 to 16.72], p = 0.14) and (-0.06 [95% CI: -0.28 to 0.15], p = 0.56), respectively. Meta-analysis of 16 case-control studies showed a pooled SMD in HOMA-IR of 0.51 (95% CI: 0.28 to 0.75), p ≤ 0.001, between cases and controls.

Conclusions: The results of our meta-analyses show that CPAP has a favorable effect on insulin resistance. This effect is not associated with any significant changes in total adiponectin levels or amount of VAF. Our findings also confirm a significant association between OSA and insulin resistance.

Keywords: adiponectin; continuous positive airway pressure; insulin resistance; obstructive sleep apnea; visceral abdominal fat.

© 2015 American Academy of Sleep Medicine.

Figures

Figure 1. Flow diagram of articles identified and evaluated during the study selection process.

CPAP, continuous positive airway pressure; HOMA-IR, homeostatic model assessment index for insulin resistance; VAF, visceral abdominal fat; RCT, randomized controlled trials; PCOS, polycystic ovarian syndrome.

Figure 2. Pooled mean difference in HOMA-IR between CPAP and sham CPAP.

The diamond reflects the pooled estimate of mean difference with 95% confidence intervals (CIs). The squares represent the individual studies

Figure 3. Pooled mean difference in BMI and ESS for the studies analyzed in HOMA-IR meta-analysis.

BMI, body mass index. ESS, Epworth Sleepiness Scale. HOMA-IR, homeostatic model assessment index for insulin resistance. The diamond reflects the pooled estimate of mean difference with 95% confidence intervals (CIs). The squares represent the individual studies

Figure 4. Pooled mean difference in VAF between CPAP and sham CPAP.

VAF, visceral abdominal fat. The diamond reflects the pooled estimate of mean difference with 95% confidence intervals (CIs). The squares represent the individual studies

Figure 5. Subgroup analysis of the pooled mean difference in VAF between CPAP and sham CPAP based on the studies included in HOMA-IR analysis.

VAF, visceral abdominal fat. The diamond reflects the pooled estimate of mean difference with 95% confidence intervals (CIs). The squares represent the individual studies

Figure 6. Pooled standardized mean difference in adiponectin between CPAP and Sham CPAP.

The diamond reflects the pooled estimate of standardized mean difference with 95% confidence intervals (CIs). The squares represent the individual studies

Figure 7. Pooled standardized mean difference of HOMA-IR between cases and controls.

HOMA-IR, homeostatic model assessment index for insulin resistance. The diamond reflects the pooled estimate of standardized mean difference with 95% confidence intervals (CIs). The squares represent the individual studies

Figure 8. Assessment of publication bias.

(A) Funnel plot for detecting publication bias for the analysis of the effect of CPAP on HOMA-IR. Kendall's tau b = 0.00, p = 0.50. Intercept (B0) as −0.36 (95% CI: −3.03 to 2.30) p = 0.36. (B) Funnel plot for detecting publication bias for the analysis of the effect of CPAP on VAF. Kendall's tau b = −0.30, p = 0.23. Intercept (B0) as −4.15 (95% CI: −20.47 to 12.17) p = 0.23. (C) Funnel plot for detecting publication bias for the analysis of the effect of CPAP on Adiponectin. Kendall's tau b = 0.00, p = 0.50. Intercept (B0) as 1.04 (95% CI: 12.43 to 14.53) p = 0.25. (D) Funnel plot for detecting publication bias for the analysis of studies on the association between sleep apnea and insulin resistance. Kendall's tau b = 0.37, p = 0.02. Intercept (B0) as 2.48 (95% CI: −0.15 to 5.12) p = 0.03. HOMA-IR, homeostatic model assessment index for insulin resistance. VAF, visceral abdominal fat. CI, confidence intervals; CPAP, continuous positive airway pressure.