| Characteristic | BMI <30 kg·m−2 | BMI 30–35 kg·m−2 | BMI ≥35 kg·m−2 | p-value# |
| Subjects n | 186 | 210 | 217 | |
| Age years | 54.4±9.6 | 53.8±10.8 | 50.5±10.8 | <0.001 |
| Male | 82.3 | 81.0 | 78.3 | 0.597 |
| BMI kg·m−2 | 27.4±2.0 | 32.4±1.4 | 39.9±4.1 | <0.0001 |
| Current smoker | 24.3 | 20.5 | 22.7 | 0.654 |
| Excessive alcohol | 3.3 | 3.8 | 3.7 | 0.963 |
| Hypertension | 25.4 | 34.0 | 50.9 | <0.0001 |
| Cardiovascular disease | 3.2 | 4.8 | 4.6 | 0.769 |
| Diabetes mellitus | 2.2 | 2.4 | 6.5 | 0.045 |
| Participate in exercise | 73.6 | 57.6 | 51.9 | <0.0001 |
| Epworth Sleepiness Scale | 11.7±4.8 | 11.4±5.1 | 12.3±5.3 | 0.183 |
| AHI events·h−1 | 37.4±15.2 | 44.7±20.4 | 49.0±23.2 | <0.0001 |
| ODI events·h−1 | 26.4±13.4 | 35.1±18.9 | 41.9±23.5 | <0.0001 |
| SaO2 nadir | 79.0±6.6 | 76.9±7.5 | 73.2±8.9 | <0.0001 |
| Percentage of time SaO2<90% | 6.1±9.1 | 12.4±16.1 | 19.7±22.3 | <0.0001 |
| Total cholesterol mg·dL−1 | 212.6±37.3 | 207.2±42.4 | 200.9±41.5 | 0.016 |
| Total cholesterol ≥200¶ | 65.6 | 57.6 | 52.5 | 0.029 |
| LDL cholesterol mg·dL−1 | 156.1±35.2 | 152.1±38.2 | 145.4±36.9 | 0.012 |
| LDL cholesterol ≥130¶ | 76.3 | 71.9 | 65.0 | 0.040 |
| HDL cholesterol mg·dL−1 | 43.6±14.3 | 39.6±9.2 | 38.2±9.4 | <0.0001 |
| HDL cholesterol <40¶ | 48.4 | 57.1 | 67.3 | 0.001 |
| Triglycerides mg·dL−1 | 147.1±63.2 | 177.9±81.6 | 198.9±103.2 | <0.0001 |
| Triglycerides ≥150¶ | 43.0 | 59.5 | 68.7 | <0.0001 |
Data are presented as mean±sd or %, unless otherwise stated. All differences were significant except for the following characteristics: sex, current smoker, excessive alcohol use, Epworth Sleepiness Scale score and cardiovascular disease. AHI: apnoea/hypopnoea index; ODI: oxygen desaturation index; SaO2: arterial oxygen saturation; LDL: low-density lipoprotein; HDL: high-density lipoprotein. #: from ANOVA and Chi-squared or Fisher’s exact test; ¶: based on the National Cholesterol Education Programme (Adult Treatment Panel III) published criteria [28].