Skip to main content
. Author manuscript; available in PMC: 2017 Aug 1.
Published in final edited form as: Am J Addict. 2016 Jul 15;25(5):400–407. doi: 10.1111/ajad.12403

TABLE 3.

Binary logistic models delineating correlates of use of the five most common NPS classes

Psychedelic phenethylamines
Euphoric stimulants
Tryptamines
Synthetic cathinones
Synthetic cannabinoids
AOR 99% CI AOR 99% CI AOR 99% CI AOR 99% CI AOR 99% CI
Sex
 Male 1.00 1.00 1.00 1.00 1.00
 Female .51*** (.40, .65) .61*** (.47, .80) .40*** (.31, .52) .61** (.44, .84) .53*** (.42, .67)
Age (quartiles)
 16–21 1.00 1.00 1.00 1.00 1.00
 22–60 .59*** (.46, .74) 1.23 (.93, 1.62) .65*** (.51, .81) .95 (.69, 1.31) .82 (.66, 1.03)
Race/ethnicity
 White 1.00 1.00 1.00 1.00 1.00
 Black .22* (.07, .72) .20* (.05, .85) .47 (.20, 1.13) 1.00 .48 (.21, 1.09)
 Hispanic .79 (.49, 1.28) .63 (.35, 1.13) 1.23 (.80, 1.88) .68 (.35, 1.32) .71 (.44, 1.12)
 Asian .9 (.43, 1.90) .49 (.17, 1.42) .39 (.15, 1.01) .73 (.25, 2.11) .7 (.31, 1.54)
 Other .84 (.53, 1.34) 1.03 (.63, 1.67) .8 (.51, 1.28) .44* (.20, .95) 1.21 (.81, 1.82)
Employment
 Employed 1.00 1.00 1.00 1.00 1.00
 Unemployed 1.05 (.82, 1.33) 1.13 (.86, 1.48) .99 (.78, 1.25) 1.38* (1.01, 1.88) 1.36** (1.09, 1.70)
Sexual orientation
 Heterosexual 1.00 1.00 1.00 1.00 1.00
 Bisexual 1.24 (.92, 1.68) 1.23 (.88, 1.72) 1.69*** (1.26, 2.26) 1.06 (.70, 1.60) 1.76*** (1.34, 2.31)
 Gay/lesbian .44*** (.28, .69) .63* (.40, .99) .55** (.37, .83) .57 (.32, 1.00) .54** (.36, .80)
Nightclub attendance (past year)
 Less than once a month 1.00 1.00 1.00 1.00 1.00
 Once a month 1.70*** (1.29, 2.22) 2.16*** (1.60, 2.90) 1.74*** (1.34, 2.26) 1.67** (1.17, 2.40) 1.17 (.91, 1.52)
 Every other week 2.62*** (1.90, 3.60) 2.61*** (1.83, 3.72) 2.01*** (1.45, 2.77) 2.36*** (1.57, 3.57) 1.05 (.75, 1.47)
 Every week 2.24*** (1.58, 3.15) 3.03*** (2.11, 4.36) 2.11*** (1.50, 2.97) 2.72*** (1.79, 4.15) 1.30 (.92, 1.83)

NPS, new psychoactive substance; AOR, adjusted odds ratio (controlling for all other variables in the model); CI, confidence interval.

Although the mean number of NPS classes used in the sample was 1.04 (SD =1.55), the maximum used was eight classes with 25.4% reporting use of more than one class; therefore, we utilized a Bonferroni correction (α =.05/5=.01) to correct for potential Type I Error associated with examining non-independent outcomes.

*

p <.01,

**

p <.001,

***

p <.0001.