Table 4:
Serum levels of 25(OH)D and overall survivala among EA patients, stratified by BMI (N=469)
| 25(OH)D Quartilesb | N deaths/patients | HRc | 95% Confidence Limits | ||
|---|---|---|---|---|---|
| BMI (≥18.5 and <25) | Quartile 1 (Highest) | 30/37 | REF | ||
| Quartile 2 | 26/38 | 0.68 | 0.40 | 1.17 | |
| Quartile 3 | 34/38 | 1.08 | 0.65 | 1.79 | |
| Quartile 4 (lowest) | 33/36 | 0.74 | 0.44 | 1.25 | |
| BMI (≥25 and <30) | Quartile 1 (Highest) | 45/56 | REF | ||
| Quartile 2 | 37/51 | 0.98 | 0.62 | 1.54 | |
| Quartile 3 | 37/50 | 0.83 | 0.53 | 1.29 | |
| Quartile 4 (lowest) | 32/45 | 1.14 | 0.71 | 1.82 | |
| BMI (≥30) | Quartile 1 (Highest) | 15/21 | REF | ||
| Quartile 2 | 18/29 | 1.32 | 0.65 | 2.67 | |
| Quartile 3 | 23/34 | 1.51 | 0.77 | 2.96 | |
| Quartile 4 (lowest) | 25/34 | 1.22 | 0.63 | 2.36 | |
Overall survival was calculated as time between date of blood draw and date of death or date last known to be alive.
Quartiles of vitamin D were determined accounting for month of blood draw.
Model estimates adjusted for the main effect of vitamin D, age, sex, smoking status, the main effect of BMI categories, year of diagnosis, chemotherapy, radiation, time-dependent surgery, and baseline hazard was stratified by tumor stage by lymph node status.