Table 2.
Associations between smoking status, duration of smoking cessation, cumulative and current cigarette consumption, and FEV1 decline (mL/year).
| Total N = 25,352 | No. Participants* (Observations) | Unadjusted FEV1 decline in mL/year† (95% CI) | Adjusted difference in FEV1 decline in mL/year‡ (95% CI) | P-value | |
|---|---|---|---|---|---|
| Smoking status | |||||
| Never-smokers | 10,087 (29,015) | 31·01 (30·66, 31·37) | Referent | ||
| Former smokers | 6,989 (17,915) | 34·97 (34·36, 35·57) | 1·82 (1·24, 2·40) | <0·0001 | |
| Current smokers | 2,462 (6,203) | 39·92 (38·92, 40·92) | 9·21 (8·35, 10·08) | <0·0001 | |
| Variable smoking status | 5,814 (17,095) | 32·90 (32·37, 33·43) | 2·56 (2·03, 3·08) | <0·0001 | |
| Duration of smoking cessation | |||||
| Never-smokers | 10,087 (29,015) | 31·01 (30·66, 31·37) | Referent | ||
| Former smokers, by duration of cessation | |||||
| 30+ years | 3,072 (8,667) | 35·17 (34·21, 36·14) | 0·93 (0·22, 1·64) | 0·0104 | |
| 20 to <30 years | 1,841 (4,744) | 35·23 (34·03, 36·44) | 2·50 (1·61, 3·40) | <0·0001 | |
| 10 to <20 years | 1,252 (2,800) | 38·16 (35·51, 40·80) | 5·67 (4·15, 7·19) | <0·0001 | |
| < 10 years | 814 (1,682) | 46·33 (42·58, 50·08) | 6·75 (4·48, 9·02) | <0·0001 | |
| Current smokers | 2,462 (6,203) | 39·92 (38·92, 40·92) | 9·42 (8·56, 10·28) | <0·0001 | |
| Cumulative cigarette consumption | |||||
| Never-smokers | 10,087 (29,015) | 31·01 (30·66, 31·37) | Referent | ||
| Ever smokers, by baseline pack-years | |||||
| <1 pack-year | 546 (1,652) | 31·29 (29·82, 32·76) | −0·08 (−1·35, 1·19) | 0·90 | |
| 1 to <10 pack-years | 2,475 (6,884) | 31·38 (30·72, 32·05) | 0·87 (0·17, 1·57) | 0·0153 | |
| 10 to <20 pack-years | 1,779 (4,601) | 33·28 (32·37, 34·18) | 1·76 (0·85, 2·68) | 0·0002 | |
| 20+ pack-years | 4,476 (10,556) | 38·50 (37·48, 39·52) | 1·53 (0·74, 2·32) | 0·0002 | |
| Current cigarette consumption | |||||
| Never-smokers | 10,087 (29,015) | 31·01 (30·66, 31·37) | |||
| Former smokers | 6,989 (17,915) | 34·97 (34·36, 35·58) | 1·57 (1·00, 2·14) | <0·0001 | |
| Current smokers, by cigarettes/day | |||||
| <5 cigarettes/day | 118 (426) | 33·21 (30·26, 36·17) | 7·65 (6·21, 9·09) | <0·0001 | |
| 5 to <20 cigarettes/day | 989 (2,498) | 36·72 (35·22, 38·22) | 8·93 (7·98, 9·89) | <0·0001 | |
| 20 to <30 cigarettes/day | 975 (2,223) | 40·96 (39·28, 42·64) | 9·98 (8·95, 11·00) | <0·0001 | |
| 30+ cigarettes/day | 452 (1,035) | 43·98 (41·36, 46·60) | 11·24 (9·86, 12·62) | <0·0001 | |
CI = confidence interval.
Linear mixed models with cohort-specific unstructured covariance matrices were used to test associations with repeated measures of the FEV1. Participants with variable smoking status were excluded from analyses of duration of smoking cessation and of cumulative and current cigarette consumption.
Number of participants in group at baseline. Time-variant cigarettes per day was used to classify sustained current smokers, such that certain current smokers changed cigarette-per-day groups over follow-up.
Unadjusted mean FEV1 decline was estimated from a model including only age and age2 as predictors. Unadjusted models were performed separately for each stratum of the primary exposures. Since age distribution differed by stratum, unadjusted FEV1 declines were consistently estimated at 57 years, the overall median age at exam.
Adjusted effect estimates for smoking exposures, relative to never-smoking, were generated using models adjusted for the smoking parameter, age, age2, height, height2, sex, weight, race/ethnicity, birth year, site, study, and educational attainment. Multiplicative interactions with age were modeled for all covariates. The effect estimate for (smoking-exposure*age) was interpreted as the association of the smoking exposure with annualized lung function decline.