Table 3.
Unadjusted hazard ratios* | Adjusted hazard ratios† | |||||
---|---|---|---|---|---|---|
HR | CI 95% | P value | HR | CI 95% | P value | |
Male gender | 1.29 | 1.0–1.7 | 0.061 | 1.42 | 1.1–1.9 | 0.014 |
Age (years) | 1.08 | 1.1–1.1 | < 0.001 | 1.06 | 1.0–1.1 | < 0.001 |
BMI (kg/m2) | 1.01 | 1.0–1.0 | 0.272 | 1.01 | 1.0–1.0 | 0.304 |
AHI (events/h) | 1.01 | 1.0–1.0 | 0.024 | 1.00 | 1.0–1.0 | 0.728 |
Cardiovascular disease‡ | 8.07 | 6.4–10.3 | < 0.001 | 4.09 | 3.1–5.3 | < 0.001 |
Hypertension§ | 2.65 | 1.9–3.6 | < 0.001 | 1.58 | 1.1–2.2 | 0.008 |
IFG/T2D | 1.94 | 1.6–2.4 | < 0.001 | 1.47 | 1.2–1.8 | 0.001 |
COPD | 2.59 | 1.9–3.6 | < 0.001 | 1.36 | 1.0–1.9 | 0.071 |
CPAP treatment | 0.56 | 0.5–0.7 | < 0.001 | 0.64 | 0.5–0.8 | < 0.001 |
Current smoker|| | 0.99 | 0.8–1.3 | 0.954 | – | – | – |
Cholesterol# (mmol/l) | 0.72 | 0.6–0.8 | < 0.001 | – | – | – |
ESS score# | ||||||
Continuous | 0.98 | 1.0–1.0 | 0.186 | – | – | – |
ESS score ≥ 10 | 0.90 | 0.7–1.1 | 0.347 | – | – | – |
CVD cardiovascular disease, CPAP continuous positive airway pressure, HR hazard ratio, CI confidence interval, BMI body mass index, AHI apnea–hypopnea index, IFG impaired fasting glucose, T2D type 2 diabetes, COPD chronic obstructive pulmonary disease, ESS Epworth Sleepiness Scale (data available for 981 CPAP-treated and 986 control patients)
Significant values are shown in italics
*Each covariate was computed separately in the model
†The model was adjusted for gender, age, BMI, AHI, cardiovascular disease, hypertension, IFG/T2D, COPD, and CPAP treatment
‡Coronary artery disease, myocardial infarction, angina pectoris, stroke, intracranial atherosclerosis, or peripheral artery disease
§Blood pressure greater than 140/90 mmHg and/or use of antihypertensive medication
||History of COPD instead of smoking status was used in the final adjusted model in order to better elucidate the amount of exposure
#Data on serum cholesterol concentrations (data available for 845 CPAP-treated and 738 control patients) and ESS score at baseline did not affect the HR of the CPAP-treated patients and were thus omitted from the final adjusted model