Table 3. Factors associated with risk of all-cause mortality (univariate Cox model).
| Hazard ratio | 95% confidence interval | p-value | |
| Age (>60.6 years) | 5.464 | 1.2; 24.6 | 0.03 |
| Gender (female vs male) | 0.221 | 0.06; 0.8 | 0.02 |
| Time *BMI (≥39.4 kg.m−2 after 24 months) | 0.079 | 0.01; 0.6 | 0.01 |
| NIV initiation (Acute versus Chronic) | 3.24 | 1.1; 9.9 | 0.04 |
| VC (>72% predicted value) | 0.228 | 0.05; 1.0 | <0.1 |
| FEV1 (>66% of predicted value) | 0.2 | 0.04; 0.9 | 0.04 |
| FEV1/VC (>77.5%) | 0.332 | 0.09; 1.2 | <0.1 |
| AIH (>33/h) | 3.098 | 0.6; 15.3 | <0.2 |
| Mean nocturnal SpO2 (>91.4%) | 0.264 | 0.06; 1.2 | <0.1 |
| Medical History | |||
| Former smoker | 3.226 | 0.9; 11.7 | <0.1 |
| Type 2 Diabetes | 2.224 | 0.7; 6.6 | <0.2 |
| COPD | 3.112 | 1.0; 9.7 | 0.05 |
| Dyslipidemia | 1.218 | 0.4; 3.6 | >0.7 |
| Stroke | 1.964 | 0.4; 8.9 | 0.3801 |
| Sleep apnea | 1.648 | 0.4; 6.0 | >0.4 |
| Hypertension | – | – | >0.9 |
| Heart Failure | 2.36 | 0.8; 7.0 | <0.2 |
| coronary heart disease | 1.47 | 0.4; 5.4 | >0.5 |
| Pulmonary Hypertension | 1.986 | 0.5; 7.2 | >0.2 |
| Thrombo-embolism | 2.166 | 0.5; 9.8 | >0.3 |
| Asthma | 1.269 | 0.3; 5.7 | >0.7 |
| Depression | 0.686 | 0.2; 2.5 | >0.5 |
| Usual Medical treatment | |||
| ß-blokers, % | 3.418 | 1.1; 10.2 | 0.03 |
| Diuretics, % | 4.274 | 0.9; 19.3 | <0.1 |
| Non-thiazide diuretics, % | 6.257 | 1.7; 22.8 | 0.005 |
| CEI, % | 3.022 | 1.0; 9.0 | 0.05 |
| Combination of cardiovascular agents, % | 5.429 | 1.02; 24.5 | 0.03 |
| Oral antidiabetics, % | 2.568 | 0.9; 7.7 | <0.1 |
| Psychoacticve drugs, % | 0.368 | 0.1; 1.7 | <0.2 |
| Inhaled corticoids | 4.507 | 1.5; 13.8 | 0.01 |
| Calcium antagonists | 0.936 | 0.26; 3.4 | >0.9 |
| ARBs | 0.599 | 0.17; 2.2 | >0.4 |
| Anti-platelet drugs | 1.639 | 0.54; 5.0 | >0.3 |
| Anticoagulants | 0.682 | 0.09; 5.2 | >0.7 |
| Statins | 0.947 | 0.3; 2.9 | >0.9 |
| Insulin | 1.86 | 0.4; 8.4 | >0.4 |
| Follow-up | |||
| Additonal long term O2 | 2.208 | 0.7; 6.8 | <0.2 |
Time*BMI: Body Mass Index; VC: Vital Capacity expressed as percentage of predicted value; FEV1: Forced Expiratory Volume in one second; AIH: Apnea Hypopnea index expressed as number of events per hour of sleep; Combination of cardiovascular agents: Combination of a diuretic and at least one other cardiovascular agent among ß-Blokers, Calcium antagonists, Converting Enzime Inhibitors, Antagiotensin II receptor blockers; COPD: Chronic Obstructive Pulmonary Disease. Continuous data were converted to dichotomous data (> or<to the median value); Since BMI did not respect proportional risks hypothesis (time-dependant variable), it was converted to a new variable: BMI x time. This variable should be interpreted as follows: After 24 months, BMI above 39.4 kg m−2 is associated with lower risk of death than BMI bellow 39.4 kg.m−2.