Table 2.
Study | Country | Population | OSA | PCOS | Controls | P (OSA %) | Notes | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Diagnosis method | N | Age (yr) | BMI (kg/m2) | OSA (%) | N | Age (yr) | BMI (kg/m2) | OSA (%) | PCOS versus controls | ||||
de Sousa 2012 [50] | Germany | Adolescents | PSG | 35 | 15.2 ± 1.0 | 33.2 ± 6.8 | 0.0% | 19 | 15.2 ± 1.1 | 32.4 ± 3.9 | 0.0% | NS | HOMA-IR and androgen levels were higher in subgroups of girls with POCS than controls. |
Fogel 2001 [35] | USA | Adults | PSG | 18 | 31.1 ± 5.5 | 36.9 ± 5.5 | 72% | 18 | 32.3 ± 5.5 | 36.9 ± 5.9 | 39% | 0.10 | Women with PCOS had higher waist-to-hip ratio and testosterone levels. |
Nandalike 2012 [54] | USA | Adolescents | PSG | 28 | 16.8 ± 1.9 | 44.8 ± 8.8 | 57.1% | 28 | 17.1 ± 1.8 | 40.2 ± 4.7 | 14.3% | < 0.01 | Only girls with symptoms were tested. No difference in BMI z-scores between groups. More girls in the PCOS group (32%) had a history of adenotonsillectomy (treatment of choice for OSA in children) prior to PSG compared controls (10.7%). |
Tasali 2008 [56] | USA | Adults | PSG | 52 | 29.7 ± 5.1 | 39.2 ± 7.2 | 55.8% | 21 | 30.7 ± 5.0 | 36.0 ± 6.9 | 19% | 0.01 | The authors adjusted for BMI and ethnicity in their analysis of OSA prevalence. |
Temple 2013 [57] | USA | Adults | PSG | 129 | 28.2 ± 5.7 | 38.6 ± 6.8 | 48.1% | 46 | 31.0 ± 6.1 | 39.4 ± 7.5 | 41% | 0.03 | The authors adjusted for age, BMI, and race in their analysis of OSA prevalence. |
Vgontzas 2001 [16] | USA | Mixed (16–45 yr) | PSG | 53 | 30.4 ± 6.6 | 38.7 ± 8.0 | 11.3% | 452 | 32.1 ± 6.4 | 26.4 ± 6.4 | 0.4% | < 0.0001 | Despite the authors adjusting for BMI, statistical adjustment is unlikely to fully exclude an obesity-related effect on OSA prevalence as the difference in BMI was high (12.3 kg/m2). |
Wootton 2017 [59] | USA | Adolescents | PSG | 16 | NR | NR | 50% | 14 | NR | NR | 57.1% | NS | Total N = 30 girls, age 14–18 years; and mean BMI 37.2 kg/m2 (range 22–55 kg/m2). No information on how participants were selected, or differences between subgroups in age, BMI, etc. |
Yang 2009 [37] | Taiwan | Adults | PSG | 18 | 29.1 ± 6.1 | 21.7 ± 2.4 | 0.0% | 10 | 31.6 ± 12.3 | 20.9 ± 1.8 | 0.0% | NS | PCOS group had higher testosterone levels and larger waist circumference and WHR compared to controls. |
Data presented as mean ± standard deviation
BMI body mass index, HOMA-IR homeostatic model assessment of insulin resistance, WHR waist-to-hip ratio, N sample size, NR not reported, NS not significant, OSA obstructive sleep apnoea, OSA % OSA prevalence, PCOS polycystic ovary syndrome, PSG polysomnography, yr years