Table 2.

Logistic Regressions Examining Differences in Reasons for Liking JUUL by JUUL Use Status (Past vs. Current Use)

Model including constructs (ref. current JUUL use)
Independent variables	AOR	95% CI		p-value
Female (ref. male)	1.53	1.13	2.06	.01
Age	0.94	0.83	1.06	.31
White (ref. non-white)	1.56	1.05	2.32	.03
Current non-e-cigarette tobacco use	0.55	0.38	0.80	.002
Current e-cigarette use (Other than JUUL)	0.13	0.09	0.18	<.001
Positive pharmacological effects (construct)	0.34	0.13	0.88	.03
Positive product characteristics (construct)	0.21	0.09	0.50	.001
Comparison to other e-cigarettes (construct)	0.95	0.46	1.95	.89
Concealability (construct)	1.19	0.62	2.30	.60
Peer Influence (construct)	0.70	0.51	0.95	.02
Other (construct)	0.69	0.29	1.64	.40
Post-hoc models to evaluate individual items from significant constructs (ref. current JUUL use)
Independent variables	AOR	95%CI		p-value
Model 1: Construct: Positive pharmacological characteristics
“It gives me energy”	1.53	0.87	2.70	.14
“It gives me a buzz”	0.46	0.33	0.63	<.001
“It helps me concentrate”	0.53	0.32	0.90	.02
“It helps me to lose weight”	1.70	0.63	4.58	.30
“It helps me control my appetite”	0.72	0.34	1.52	.39
Model 2: Construct: Positive product characteristics
“I like the shape of the JUUL”	1.07	0.51	2.21	.86
“I like the size of the JUUL”	0.54	0.28	1.04	.07
“I like the flavors”	0.69	0.49	0.95	.03
“I like the nicotine level”	0.65	0.32	1.32	.23
“It is easy to change pods”	0.70	0.39	1.27	.24
Model 3: Construct: Peer influence
“My friends use it”	0.66	0.49	0.89	.01

AOR = adjusted odds ratio.

Bolding indicates significant values (p < .05). For all models, all independent variables were entered into the multinomial logistic regression model simultaneously. Across all models, sex, race, age, use of non-e-cigarette other tobacco products, use of other e-cigarettes, and school (not shown) were included as covariates. Covariates did not differ across models, so values from the initial model are included to demonstrate directionality and significance.