TABLE 1

Reasons for First Trying E-Cigarettes (Fall 2013) Predicting Continued Use of E-Cigarettes Over Time (Spring 2014)

Reasons for Trying E-CigarettesIndependent ModelsMultivariable Models
OR (95% CI)aOR (95% CI)OR (95% CI)aOR (95% CI)
Interest
 Curiosity0.77 (0.50–1.20)0.86 (0.54–1.37)1.02 (0.60–1.70)0.97 (0.57–1.64)
 It is cool2.14 (0.80–5.73)1.64 (0.72–3.74)1.62 (0.69–3.78)1.52 (0.64–3.61)
Desirable attributes
 Good flavors1.66 (1.06–2.61)a1.39 (0.86–2.22)1.12 (0.64–2.00)0.98 (0.54–1.78)
 Does not smell bad2.46 (1.38–4.36)a2.23 (1.23–4.04)a,b1.36 (0.60–3.06)1.48 (0.64–3.39)
 Hide from adults2.16 (1.10–4.28)a1.86 (0.92–3.78)0.80 (0.32–2.01)0.80 (0.31–2.04)
 Low cost3.08 (1.34–7.08)a2.56 (1.08–6.06)a,b1.33 (0.48–3.64)1.48 (0.53–4.10)
Social norms
 Friends use1.69 (1.05–2.72)a1.67 (1.01–2.76)a,b1.41 (0.82–2.42)1.36 (0.78–2.38)
 Parents/family use0.96 (0.43–2.11)1.16 (0.50–2.68)0.90 (0.36–2.22)0.94 (0.37–2.40)
 Can use anywhere3.02 (1.66–5.49)a2.24 (1.18–4.26)a2.06 (0.95–4.48)1.56 (0.68–3.58)
Goal-directed
 To quit smoking cigarettes20.24 (2.67–153.30)a13.42 (1.72–104.38)a20.94 (2.64–166.46)a14.54 (1.80–117.80)a
 Healthier than cigarettes1.80 (1.08–3.00)a1.60 (0.94–2.76)1.06 (0.57–1.99)1.14 (0.60–2.16)
  • Reasons are coded yes = 1, no = 0. All models controlled for school status (middle school versus high school). aOR, adjusted odds ratio including covariates; OR, odds ratio.

  • Independent models reflect estimates when reasons are entered into separate logistic regression models. Multivariable models reflect estimates when reasons are entered simultaneously into the same logistic regression model. Results for covariates in final adjusted model: middle school (versus high school) OR, 2.86 (95% CI, 1.01–8.13); female (versus male) OR, 1.16 (95% CI, 0.70–1.92); white (versus nonwhite) OR, 0.52 (95% CI, 0.24–1.14); traditional cigarette use in the past month at wave 1 OR, 2.07 (95% CI, 1.12–3.82); other tobacco use in the past month at wave 1 OR, 1.38 (95% CI, 0.78–2.47).

  • Similar results are seen if including age instead of school status (middle school versus high school) in the model; age is negatively related to the odds of continuing e-cigarette use; OR, 0.72 (95% CI, 0.58–0.89).

  • a Confidence intervals (CIs) that do not overlap 1.00, signifying statistical significance, P < .05.

  • b Indicates results were no longer significant in sensitivity analyses that did not impute missing data.