Table 4. Results of model fitting cubic function of log sIgA secretion to respiratory disease mortality.
| Model | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | |||||||||||
| Polynomial term | B | SE | p | B | SE | p | B | SE | p | B | SE | p | B | SE | p |
| Linear | 0.40 | 0.24 | .097 | 0.52 | 0.24 | .031 | 0.47 | 0.24 | .048 | 0.48 | 0.24 | .045 | 0.43 | 0.23 | .062 |
| Quadratic | -0.00 | 0.14 | .984 | -0.04 | 0.14 | .787 | 0.02 | 0.14 | .905 | 0.01 | 0.14 | .936 | 0.03 | 0.13 | .807 |
| Cubic | -0.13 | 0.06 | .033 | -0.15 | 0.06 | .011 | -0.12 | 0.06 | .035 | -0.12 | 0.06 | .039 | -0.11 | 0.05 | .047 |
Model1: adjusted for assay batch and sex. Model2: model1 + smoking status. Model3: model2 + Medications. Model4: model3 + Self-rated Health. Model5: model4 + manual social class.
Hazard ratios are per standard deviation of log sIgA secretion rate.