Tobacco smoking and/or e-cigarette use abstinence or cessation, possibly related adverse events, difference in daily tobacco cigarette consumption and self-reported health: results of the multivariate analyses
| Outcomes at 24 months | Adjusted OR (95% CI) | p Value | Crude OR (95% CI) | p Value |
|---|---|---|---|---|
| Continuous tobacco abstinence from baseline or quit tobacco smoking during follow-up*,† | ||||
| Tobacco cigarettes only at baseline | 1 (ref cat) | – | 1 (ref cat) | – |
| E-cigarettes only at baseline | 5.56 (3.89 to 7.95) | <0.001 | 5.23 (3.72 to 7.35) | <0.001 |
| Both tobacco and e-cigarettes at baseline | 1.25 (0.85 to 1.84) | 0.2 | 1.17 (0.81 to 1.69) | 0.4 |
| Quit all cigarettes (electronic and/or traditional)* | ||||
| Tobacco cigarettes only at baseline | 1 (ref cat) | – | 1 (ref cat) | – |
| E-cigarettes only at baseline | 1.12 (0.73 to 1.72) | 0.6 | 1.09 (0.73 to 1.64) | 0.7 |
| Both tobacco and e-cigarettes at baseline | 0.94 (0.59 to 1.49) | 0.8 | 0.79 (0.51 to 1.23) | 0.3 |
| Safety–possibly related sAEs‡ | ||||
| Tobacco cigarettes only at baseline | 1 (ref cat) | – | 1 (ref cat) | – |
| E-cigarettes only at baseline | 1.48 (0.63 to 3.47) | 0.4 | 1.52 (0.66 to 3.48) | 0.3 |
| Both tobacco and e-cigarettes at baseline | 2.40 (1.09 to 5.26) | 0.029 | 2.23 (1.04 to 4.76) | 0.038 |
| Reduction of tobacco cigarette consumption by 50% or more from baseline to 24 months§ | ||||
| Tobacco cigarettes only at baseline | 1 (ref cat) | – | 1 (ref cat) | – |
| Both tobacco and e-cigarettes at baseline | 1.28 (0.90 to 1.82)¶ | 0.2 | 1.25 (0.89 to 1.76)¶ | 0.2 |
| 0.99 (0.57 to 1.71)** | 0.9 | 1.25 (0.76 to 2.05)** | 0.4 | |
| Reduction of tobacco cigarette consumption by 50% or more from baseline to 24 months†† | ||||
| Tobacco or dual users at baseline who switched to or continued with tobacco smoking only | 1 (ref cat) | – | 1 (ref cat) | – |
| Tobacco or dual users at baseline who started or continued dual use | 8.48 (4.05 to 17.8) | <0.001 | 7.47 (3.81 to 14.7) | <0.001 |
| Adjusted coefficient (95% CI) | Raw coefficient (95% CI) | |||
|---|---|---|---|---|
| Difference in the daily number of tobacco cigarettes from 24 months to baseline‡‡ | ||||
| Tobacco cigarettes only at baseline | 0 (ref cat) | – | 0 (ref cat) | – |
| Both tobacco and e-cigarettes at baseline | 0.44 (−0.88 to 1.75)¶ | 0.5 | 0.10 (−1.41 to 1.61)¶ | 0.9 |
| 0.75 (−1.13 to 2.63)** | 0.4 | 0.87 (−0.42 to 2.15)** | 0.2 | |
| Difference in the daily number of tobacco cigarettes from 24 months to baseline§§ | ||||
| Tobacco smokers who continued only smoking and dual users at baseline who switched to tobacco only smoking | 0 (ref cat) | – | 0 (ref cat) | – |
| Dual users at baseline who continued dual use and tobacco-only smokers who switched to dual use | −5.71 (−8.02 to −3.39) | <0.001 | −5.32 (−7.80 to −2.85) | <0.001 |
| Difference in the self-reported health score from 24 months to baseline¶¶ | ||||
| Tobacco cigarettes only at baseline | 0 (ref cat) | – | 0 (ref cat) | – |
| E-cigarettes only at baseline | 0.14 (−0.08 to 0.35) | 0.2 | −0.03 (−0.27 to 0.22) | 0.8 |
| Both tobacco cigarettes and e-cigarettes at baseline | −0.15 (−0.36 to 0.07) | 0.2 | −0.14 (−0.38 to 0.11) | 0.3 |
*Random-effect logistic regression with region as the cluster level, adjusting for the following baseline characteristics: age, gender, BMI, marital status, educational level, occupation, alcohol use, hypertension, hypercholesterolaemia, diabetes, self-reported health, years of tobacco smoking (former smoking for e-cigarette users), number of tobacco cigarettes smoked per day (or puffs per day for e-cigarette only smokers). A total of 902 participants were included in the final model due to 30 missing items in the self-reported health item at baseline.
†Including (1) users of e-cigarettes only at baseline who quit e-cigarette use and did not start tobacco smoking; (2) smokers of tobacco cigarettes only at baseline who quit tobacco cigarette use and did not start e-cigarette use; (3) dual users at baseline who quit both tobacco smoking and e-cigarette use; (4) e-cigarette users at baseline who continued to use only e-cigarettes; (5) tobacco cigarettes smokers at baseline who quit tobacco smoking and started using e-cigarettes; (6) dual users at baseline who quit tobacco smoking and continued to use e-cigarettes only.
‡Random-effect logistic regression with region as the cluster level, adjusting for age, baseline self-reported health and quitting all smoking during the 24-month follow-up.
§E-cigarette-only users at baseline were not included. Random-effect logistic regression with region as the cluster level, adjusting for the following baseline characteristics: age, gender, BMI, marital status, educational level, occupation, alcohol use, hypertension, hypercholesterolaemia, diabetes, self-reported health, years of tobacco smoking, number of tobacco cigarettes smoked per day. A total of 682 participants were included in the final model due to 21 missing items in the self-reported health item at baseline.
¶Computed from a model that included the whole sample of tobacco and dual smokers at baseline.
**Computed from a model that included only the subsample of baseline smokers who continued smoking at 24 months (excluding those who had quit and whose cigarette consumption was zero at follow-up).
††Restricted to tobacco only smokers or dual users at baseline who did not quit or switch to e-cigarettes only (n=529). Random-effect logistic regression with region as the cluster level, adjusting for the following baseline characteristics: age, gender, self-reported health, years of tobacco smoking, number of tobacco cigarettes smoked per day.
‡‡E-cigarette-only users at baseline were not included. Random-effect linear regression with region as the cluster level, adjusting for the following baseline characteristics: age, gender, BMI, marital status, educational level, occupation, alcohol use, hypertension, hypercholesterolaemia, diabetes, self-reported health, years of tobacco smoking, number of tobacco cigarettes smoked per day. A total of 682 participants were included in the final model due to 21 missing items in the self-reported health item at baseline.
§§Restricted to tobacco-only smokers or dual users at baseline who did not quit or switch to e-cigarettes only (n=529). Random-effect linear regression with region as the cluster level, adjusting for the following baseline characteristics: age, gender, self-reported health, years of tobacco smoking, number of tobacco cigarettes smoked per day.
¶¶Random-effect linear regression with region as the cluster level, adjusting for the following baseline characteristics: age, gender, BMI, marital status, educational level, occupation, alcohol use, hypertension, hypercholesterolaemia, diabetes, self-reported health, years of tobacco smoking (former smoking for e-cigarette users), number of tobacco cigarettes smoked per day (or puffs per day for e-cigarette only smokers). A total of 899 participants were included in the final model due to 30 missing items in the self-reported health item at baseline and 3 missing items in the self-reported health at 24 months.
BMI, body mass index; sAEs, serious adverse events.