Table 2.
Logistic regression for prediction of diabetes in the UK Biobank dataset.
| Variable | Odds ratio | 95% CI | p-value | |
|---|---|---|---|---|
| Model 1 | PAT Area (cm2) | 1.44 | 1.40, 1.47 | 8.43 ×10−164 |
| Model 2 | PAT Area (cm2) | 1.36 | 1.32, 1.40 | 1.26 ×10−92 |
| Male sex | 1.32 | 1.20, 1.44 | 8.97 ×10−9 | |
| Age (years) | 1.03 | 1.03, 1.04 | 4.66 ×10−30 | |
| Model 3 | PAT Area (cm2) | 1.07 | 1.03, 1.10 | 1.41 ×10−4 |
| Male sex | 1.80 | 1.63, 1.98 | 2.43 ×10−32 | |
| Age (years) | 1.05 | 1.04, 1.05 | 7.91 ×10−54 | |
| BMI | 1.18 | 1.17, 1.19 | 4.27 ×10−305 |
The dataset includes 2,529 diabetics and 40,399 non-diabetics. Odds ratios for PAT area are indicated for an increase of 10 cm2. BMI, body mass index; CI, confidence interval; PAT, pericardial adipose tissue.