TABLE 1.
The Binary Logistic Regression Analysis for the Univariate and Multivariate Association Between Obesity and Hospital Mortality
| Model | Tested Category | Reference Category | Odds (Ref) | Probability (Ref) | OR | p | p for Interaction |
|---|---|---|---|---|---|---|---|
| Univariate model | Obesity (BMI ≥ 30) | BMI < 30, all ages | 0.32 (0.43) | 0.24 (0.30) | 0.74 (0.69–0.80) | < 0.001 | 0.12 |
| Obesity (BMI ≥ 30; < 45 yr) | BMI < 30, <45 yr | 0.04 (0.07) | 0.04 (0.07) | 0.58 (0.36–0.93) | 0.02 | ||
| Obesity (BMI ≥ 30; 45–65 yr) | BMI < 30, 45–65 yr | 0.19 (0.20) | 0.16 (0.17) | 0.97 (0.85–1.10) | 0.6 | ||
| Obesity (BMI ≥ 30; > 65 yr) | BMI < 30, > 65 yr | 0.68 (0.75) | 0.41 (0.43) | 0.91 (0.82–1.01) | 0.08 | ||
| Multivariate modela | Obesity (BMI ≥ 30) | BMI < 30, all ages | 0.98 (0.90–1.06) | 0.62 | 0.08 | ||
| Obesity (BMI ≥ 30; < 45 yr) | BMI < 30, < 45 yr | 0.59 (0.36–0.96) | 0.03 | ||||
| Obesity (BMI ≥ 30; 45–65 yr) | BMI < 30, 45–65 yr | 1.05 (0.91–1.20) | 0.52 | ||||
| Obesity (BMI ≥ 30; > 65 yr) | BMI < 30, > 65 yr | 0.97 (0.87–1.08) | 0.55 |
BMI = body mass index, OR = odds ratio.
a
The multivariate model was adjusted for age, gender, Acute Physiology and Chronic Health Evaluation IV mortality probability (quintiles), mechanical ventilation at ICU admission, Pao2/Fio2 ratio (quintiles), and the use of vasoactive drugs.
Estimates are ORs and 95% CIs. p is based on the post hoc Wald test for significance of the estimates. Bold values are significant at 5% alpha level.