Table 3.
Moderation Analysis of the Association Between Pulmonary Function and Cognition
| Participants | β (95% CI)* | p Value | p Interaction† | |
|---|---|---|---|---|
| Gender | .0334 | |||
| Men | 4134 | 0.43 (0.32, 0.53) | <.0001 | |
| Women | 4122 | 0.59 (0.44, 0.74) | <.0001 | |
| Education | .0002 | |||
| Illiterate (0 year) | 2369 | 0.45 (0.24, 0.65) | <.0001 | |
| Low (1–9 years) | 3629 | 0.52 (0.39, 0.65) | <.0001 | |
| High (≥10 years) | 2258 | 0.37 (0.21, 0.54) | <.0001 | |
| Smoking | .0412 | |||
| Current | 2684 | 0.38 (0.24, 0.53) | <.0001 | |
| Former | 778 | 0.33 (0.10, 0.55) | .0043 | |
| Never | 4794 | 0.55 (0.42, 0.68) | <.0001 |
Notes: CI = confidence interval; SD = standard deviation.
*Repeated-measures generalized estimating equations were used to test cognitive change associated with each SD increment of pulmonary function over time.
†We then examined whether the association between pulmonary function and cognitive function over time was modified by age, gender, education, geographic region, smoking, physical function limitation, or lung disease. Stratified analysis was conducted for those with significant interaction.