Table 2:
Incremental Models for Assessing Association of Baseline Smoking Status with Longitudinal Trajectory of Muscle Attenuation
| Primary predictor, model | Male |
Female |
||
|---|---|---|---|---|
| Main effect | Linear trend | Main effect | Liner trend | |
|
| ||||
| Smoking status (current vs. former) | ||||
| Model 1 | 0.879 (0.106) [<.001] | −0.108 (0.028) [<.001] | 0.721 (0.128) [<.001] | −0.077 (0.034) [=.02] |
| Model 2 | 0.544 (0.103) [<.001] | −0.105 (0.028) [<.001] | 0.493 (0.127) [<.001] | −0.069 (0.034) [=.043] |
| Model 3 | 0.329 (0.101) [<.001] | −0.114 (0.028) [<.001] | 0.416 (0.123) [<.001] | −0.069 (0.034) [=.043] |
| Model 4 | −0.695 (0.091) [<.001] | −0.094 (0.029) [=.001] | −0.622 (0.110) [<.001] | −0.051 (0.035) [=.15] |
| Model 5 | −0.681 (0.090) [<.001] | −0.094 (0.029) [=.002] | −0.607 (0.109) [<.001] | −0.050 (0.035) [=.16] |
| Model 6 | −0.882 (0.091) [<.001] | −0.079 (0.028) [=.005] | −0.688 (0.110) [<.001] | −0.045 (0.035) [=.2] |
|
| ||||
| † Smoking quit time (y) | ||||
| Model 1 | −0.023 (0.015) [=.12] | 0.001 (0.004) [=.89] | −0.038 (0.019) [=.051] | 0.003 (0.005) [=.62] |
| Model 2 | 0.029 (0.015) [=.051] | 0.002 (0.004) [=.57] | 0.005 (0.019) [=.78] | 0.002 (0.005) [=.7] |
| Model 3 | 0.026 (0.014) [=.07] | 0.003 (0.004) [=.51] | 0.011 (0.019) [=.56] | 0.002 (0.005) [=.74] |
| Model 4 | 0.051 (0.012) [<.001] | 0.002 (0.004) [=.64] | 0.040 (0.016) [=.01] | 0.001 (0.005) [=.88] |
| Model 5 | 0.044 (0.012) [<.001] | 0.003 (0.004) [=.52] | 0.034 (0.016) [=.04] | 0.001 (0.005) [=.87] |
| Model 6 | 0.045 (0.012) [<.001] | 0.002 (0.004) [=.63] | 0.035 (0.016) [=.03] | 0.000 (0.005) [=.97] |
Note: Displayed numbers are estimated regression coefficients, standard errors in parentheses, and P values in brackets. The main effect terms assess the association with baseline muscle attenuation, while the linear trend terms (by interaction with time) assess the association with linear longitudinal change rate of muscle attenuation.
Analyses with smoking quit time as primary predictor were performed among those who had quit smoking at baseline.
Model 1: Unadjusted
Model 2: Adjusted for age, race, education level, and marital status
Model 3:Model 2 + Diagnostic history of cardiovascular diseases, diabetes, and COPD
Model 4: Model 3 + Body mass index
Model 5: Model 4 + Pack-year
Model 6: Model 5 + Percent emphysema