Experimental and Clinical
Psychopharmacology
The Role of First Use of Inhalants Within Sequencing
Pattern of First Use of Drugs Among Brazilian University
Students
João Maurício Castaldelli-Maia, Sérgio Nicastri, Lúcio Garcia de Oliveira, Arthur Guerra de
Andrade, and Silvia S. Martins
Online First Publication, August 25, 2014. http://dx.doi.org/10.1037/a0037794
CITATION
Castaldelli-Maia, J. M., Nicastri, S., de Oliveira, L. G., de Andrade, A. G., & Martins, S. S.
(2014, August 25). The Role of First Use of Inhalants Within Sequencing Pattern of First Use
of Drugs Among Brazilian University Students . Experimental and Clinical
Psychopharmacology. Advance online publication. http://dx.doi.org/10.1037/a0037794
Experimental and Clinical Psychopharmacology
2014, Vol. 22, No. 5, 000
© 2014 American Psychological Association
1064-1297/14/$12.00 http://dx.doi.org/10.1037/a0037794
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
The Role of First Use of Inhalants Within Sequencing Pattern of First Use
of Drugs Among Brazilian University Students
João Maurício Castaldelli-Maia
Sérgio Nicastri
University of São Paulo and Fundação do ABC
University of São Paulo
Lúcio Garcia de Oliveira
Arthur Guerra de Andrade
University of São Paulo
University of São Paulo and Fundação do ABC
Silvia S. Martins
Columbia University
The present study investigated the role of first use of inhalants within a first drug sequencing pattern. In
a representative sample of university students from 27 Brazilian capitals (n ⫽ 12,711), we analyzed the
patterns of transition from/to first use of inhalants to/from the first use of alcohol, tobacco, cannabis,
cocaine, hallucinogens, ecstasy, amphetamines, prescription opioids, and tranquilizers. Cox proportional
hazards models were used to analyze data. Drugs that were not specified as the pair of drugs tested in
each model were included as time-varying covariates in all models. In this sample, first use of inhalants
was preceded only by the first use of alcohol and tobacco. However, first use of inhalants preceded first
use of cannabis, amphetamines, cocaine, and tranquilizers. First use of inhalants preceded the first use of
prescription opioids, and vice versa. This study highlights the need to intervene early with youths who
are at risk of or just beginning to use inhalants, because this class of drugs seems to be the first illegal
drug in Brazil to be experimented by respondents in our sample. There is also a call for attention to
individuals who have already first used inhalants because of their higher chance to experiment with other
drugs such as cannabis, cocaine, and prescription drugs. All these findings show an in-transition culture
of drug use, which should be tracked through time, because some classical models (i.e., gateway model)
might be outdated and might also not fit within different settings.
Keywords: Brazil, experimentation, gateway, inhalants, transition, university students
major consequences for abusers, their families, and society
(Balster et al., 2009), yet it is one of the least-understood and
poorly studied substance abuse problems (Balster et al., 2009).
Despite the substantial prevalence and serious toxicities of inhalant
use, it has been termed “the forgotten epidemic,” and remains as the
least studied form of substance abuse, although research on its epidemiology, neurobiology, treatment, and prevention has accelerated
in recent years (Howard et al., 2011). The high prevalence of lifetime
use of inhalants is also evident in the general population of Brazil, as
inhalants are the fourth most widely used drug (Fonseca, Galduroz,
Noto, & Carlini, 2010). Considering these statements, the present
study sought to investigate the role of first use of inhalants within
sequencing pattern of first use of drugs among Brazilian university
students.
Kandel (1975) and Kandel and Faust (1975) proposed a gateway
model that attempts to explain the passage from the first contact
with a drug to the first contact with a second drug. In this model,
the user would go through four stages of involvement with psychoactive substances. There was no special attention to inhalants in
this proposed model despite their high prevalence of use. At first,
usually individuals use beer and wine. Then, individuals could use
tobacco and/or strong drinks (liquor, spirits). From the second
stage the person could go to a third stage, in which use of cannabis
can happen. The gateway hypothesis implicitly assumes a causal
Inhalants are one of the most popular drug classes throughout
the world. Inhalant abuse, which is also referred to as volatile
substance abuse or solvent abuse, is a worldwide problem with
João Maurício Castaldelli-Maia, Department of Psychiatry, Medical School,
University of São Paulo, and Department of Neuroscience, Medical School,
Fundação do ABC, Brazil; Sérgio Nicastri, Department of Psychiatry, Medical
School, University of São Paulo; Lúcio Garcia de Oliveira, Department of
Psychiatry, Medical School, University of São Paulo; Arthur Guerra de Andrade, Department of Psychiatry, Medical School, University of São Paulo,
and Department of Neuroscience, Medical School, Fundação do ABC; Silvia
S. Martins, Department of Epidemiology, Columbia University Mailman
School of Public Health.
This work was supported by National Secretariat for Drug Policies
(SENAD), Brazil and São Paulo State Research Support Foundation FAPESP,
Brazil. Dr. Martins receives research support from the National Institute on
Drug Abuse (NIDA) Grant DA023434 and from the National Institute of
Child and Human Development (NICHD) Grant HD060072, while working on this article. Dr. Oliveira receives research support from FAPESP
Grant 2008/55550-7.
Correspondence concerning this article should be addressed to João
Maurício Castaldelli-Maia, Rua Dr Ovídio Pires de Campos, 785 - GREA,
05403-903, São Paulo, SP, Brazil. E-mail: jmcmaia2@gmail.com
1
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2
CASTALDELLI-MAIA ET AL.
chain sequence in which (a) alcohol and/or tobacco use precede
cannabis use, (b) cannabis is used before the onset of other illegal
drugs (including inhalants), and (c) the use of cannabis increases
the likelihood of using other illegal drugs (Fergusson et al., 2006).
Alcohol and tobacco are indeed the first drugs that adolescents
experiment with according to several international studies (Van
Etten et al., 1997; Wagner & Anthony, 2002; Herrera-Vázquez et
al., 2004; Wells & McGee, 2008; Caris et al., 2009; Mayet et al.,
2010), which may be attributable to the fact that these two drugs
have the highest prevalence of use throughout the life span in most
countries. However, there are a considerable number of individuals
who experiment with cannabis before alcohol and/or tobacco depending on neighborhood (Tarter et al., 2006) and ethnic factors
(Vaughn et al., 2008). Regarding sequence of first use of drugs
other than alcohol and tobacco, data are more controversial
(Castaldelli-Maia, Martins, de Oliveira, de Andrade, & Nicastri,
2014). In countries with a high prevalence of cannabis use, such as
Germany, Ukraine, and New Zealand, the first use of cannabis
precedes the first use of drugs other than alcohol and tobacco
(Degenhardt et al., 2010). This pattern was indeed the case for
Brazilian university students in a previous study (Castaldelli-Maia,
Martins, de Oliveira, van Laar, de Andrade, & Nicastri, 2014).
However, there are few studies (Ding et al., 2009; Sanchez,
Noto, & Anthony, 2013) focusing on inhalants within drug first
use sequencing pattern. In the United States, Ding et al. (2009)
found that inhalants are more commonly used before other illicit
drugs in a large sample of inhalant users. Within the Brazilian
university student population, the prevalence of lifetime use of
inhalants was relatively high (24.6%) according to a survey conducted during this decade (Wagner et al., 2007). Based on these
data of relatively high prevalence of lifetime use of inhalants, one
can hypothesize that inhalants could be commonly used before
illegal drugs, occupying a similar role to cannabis. Sanchez, Ribeiro, Moura, Noto, & Martins (2013) analyzed data from a
representative sample of 8th to 12th grade students (middle
school ⫽ 8th to 9th grade; high school ⫽ 10th to 12th grade) from
private schools (n ⫽ 5,226) in São Paulo, Brazil, to test for
sequencing patterns among alcohol, tobacco, cannabis, and inhalants. In their study, inhalants were characterized as an intermediate
drug class between legal (alcohol and tobacco) and illegal (cannabis) drugs. But this previous study (Sanchez, Ribeiro, et al.,
2013) did not investigate the following: (a) the sequencing patterns
involving inhalants and other important drugs to the Brazilian
context (Castaldelli-Maia, Bhugra, de Andrade, & Lotufo-Neto,
2013), such as crack/cocaine (Moreira et al., 2014), prescription
drugs (Opaleye et al., 2014), hallucinogens (Castaldelli-Maia, de
Oliveira, de Andrade, Lotufo-Neto, & Bhugra, 2012), and ecstasy
(Remy et al., 2013); and (b) older students from college years in
private and public settings. Considering that adolescence is a
period of experimentation with various drugs (Malta et al., 2010),
the subsequent period (university or college years) may be of
interest to access the sequence of experimentation of with these
drugs. In a previous analysis using the sample of the present study,
the average age of first use of each of the nine drugs analyzed in
this study was between 15.2 and 21.2 years (Castaldelli-Maia,
Martins, de Oliveira, van Laar, de Andrade, & Nicastri, 2014).
This study investigated the role of first use of inhalants within a
drug use sequencing pattern in a large representative sample of
Brazilian university students. It is also important to consider the
first use of drugs other than inhalants as time-dependent variables,
as they may have a mediating role in the sequence of drug use.
Method
Data comes from an epidemiological study conducted in 27
Brazilian state capitals. The general objective of this epidemiological study was to evaluate the sociodemographic characteristics,
drug use and mental health aspects of a nationally representative
sample of university students (n ⫽ 12,711). The project was
previously evaluated and approved by the Ethics Committee for
the Analysis of Research Projects at the School of Medicine of the
University of São Paulo. Data collection was completed between
May and December of 2009.
Sample
The target population of this study were university students
enrolled in undergraduate programs in Higher Education Institutions (HEIs), both public and private, in the 27 Brazilian state
capitals (Brazil is composed by 27 states). The undergraduate
programs can last from 4 to 6 years in Brazil. A random sample
was stratified and recruited by unequal size clusters. Sampling was
conducted in two stages; a sample of HEIs was selected, followed
by the selection of classes of students was chosen from this
sample. Because the size of the HEIs and the classes (in terms of
number of universities) were not always the same, these conglomerates were of unequal sizes.
Because all 27 state capitals had to be covered, including public
and private HEIs, sample stratification was carried out based on
these two variables (public and private), namely, by capital and by
type of institution, a total of 54 levels. However, this stratification
was used only for operating purposes. At the data analysis stage,
only the five administrative regions (including the 27 state capitals) and the two HEI types were considered for stratification,
generating a total of 10 levels. To make the field work economically feasible, it was decided to select a sample of HEIs and, within
each one of these, select a sample of classes: the primary sampling
unit considered for this study was the HEI, and the secondary
sampling unit was the class. Thus, sampling selection was made in
two stages, in which the clusters considered included both HEIs
and student groups. Because the sizes (in relation to the number of
students) of the HEIs and the groups are not always the same, these
clusters are, in fact, of different sizes. Thus, in general, the sampling design consisted of a random sample, stratified by clusters of
different sizes, selected in two stages. However, this sampling
selection method posed a challenge because of the structure of the
population under review. In conventional sampling, the clusters
form a disconnected division, that is, the population elements are
related to a single cluster, which is not true for higher education.
Within a HEI, student groups are not disconnected, and there may
be an overlap between groups of students, because a single student
can be enrolled in more than one subject. To get around this issue,
we used the multiplicity-based sampling method, enabling the
individuals to be included in more than one cluster. However, it
was necessary to obtain information about the number of clusters
to which the elements selected for the sample were related, for
later use in the statistical analysis stage. At the end of the data
collection, 100 of 114 HEIs accepted to take part in this study
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
FIRST USE OF INHALANTS AND FIRST USE OF DRUGS
(88% of the estimated size), as well as 654 student classes (70.6%
of the estimated size), with a total participation of 12,721 college
students throughout Brazil. Although the response rate of participation was 95.6% among the college students who were inside
classes at the time of the interview, the final response rate for this
study was around 72.1% when the estimated size of the college
student sample was taken into consideration (12,721 of 17,651).
Finally, of these students (12,721), 10 were excluded because they
claimed to use Relevin (fictitious drug); thus, the data from 12,711
nationwide college students were analyzed. More details about this
survey can be found in Andrade et al. (2012).
Part of the 12,711 individuals was excluded from the present
study. Respondents older than 40 years (n ⫽ 548) were excluded
because of the high probability of memory bias. In addition, those
with missing data on the variables of interest—age (n ⫽ 141),
whether they used some of the drugs tested (n ⫽ 149 [cannabis];
183 [inhalants]; 228 [opioid analgesics]; 174 [tranquilizers]; 206
[ecstasy]; 191 [amphetamines]; 190 [cocaine]; 165 [hallucinogens]), alcohol (n ⫽ 3203) or tobacco (n ⫽ 0), those who did not
answer the age of first use of some of the drugs tested (n ⫽ 476
[cannabis]; 625 [inhalants]; 670 [opioid analgesics]; 602 [tranquilizers]; 316 [ecstasy]; 430 [amphetamines]; 284 [cocaine]; 298
[hallucinogens]); alcohol (n ⫽ 3,203) or tobacco (n ⫽ 985)—were
excluded. Sample size varied according to the transition considered.
3
tionnaire was based on the World Health Organization research
questionnaire, which was previously adapted by Andrade et al.
(1997) and Stempliuk et al. (2005) for use with undergraduate
students.
The present study used part of these data, which focused on drug
use. This section was collected in a series of questions with
multiple response options. In the first part, the students reported
whether they had ever used the following drugs (Have you ever
tried NAME OF THE DRUG without a doctor’s prescription?
Answer: Yes or No): inhalants, alcohol, tobacco, cannabis, cocaine,
“merla” (a cocaine byproduct), crack, hallucinogens, ketamine,
ayahuasca (Santo Daime), ecstasy (3– 4 methylenodyoximethamphtamine [MDMA]), anabolic steroids, tranquilizers, prescription
opioids, anticholinergics, heroin, amphetamines, and synthetic
drugs (methamphetamines and GHB). In the second part of the
questionnaire, we asked for the age of the initial use of each of
these drugs (How old were you when you first tried NAME OF
THE DRUG? Answer: XX years). For this study, cocaine, “merla,”
and crack were combined in the same group (named cocaine).
Our survey instrument included questions that asked all respondents about the use of ‘inhalants or solvents,’ giving the following
as examples: “loló,” glue, paint thinner, benzene, enamel, gasoline,
and “lança-perfume.” All these examples of inhalants or solvents
cover the majority of types of substances of this class of drugs used
in Brazil (Mesquita et al., 1998; Hynes-Dowell et al., 2011; Sanchez, Noto, et al., 2013; Silva-Oliveira et al., 2014)
Measures
A structured, self-administered, and anonymous questionnaire
of 98 closed questions was developed with an emphasis on drug
use and related disorders, risky behaviors and the existence of
psychiatric comorbidity (depressive symptoms and psychotic and
nonspecific psychological complaints). The content of this ques-
Statistical Analysis
Time-dependent variables. Because the aim of this study
was to assess whether the transition from the first use of one drug
to the subsequent use of another drug occurred or not—always
involving inhalants in one of the two roles—these comparisons
Figure 1. Fluxogram presenting the division of the individuals between users and nonusers, among Brazilian
university students of the 27 state capitals, 2009. See the online article for the color version of this figure.
CASTALDELLI-MAIA ET AL.
4
Table 1
Sociodemographic Data (14 Variables Included in the Cox Survival
Models) of Brazilian University Students From 27 Capitals, 2009
Table 1 (continued)
Variable
%
95% CI
SE
28.67
26.42
17.00
14.33
9.64
2.13
21.83, 36.65
18.19, 36.70
11.22, 24.92
9.93, 20.23
3.75, 22.61
1.37, 3.31
0.03
0.04
0.03
0.02
0.04
0.01
15.05
84.95
8.27, 25.82
74.18, 91.73
0.04
0.04
92.03
7.53
88.58, 94.50
5.20, 10.80
0.01
0.01
c
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Variable
Gender
Female
Male
Agea
⬍22 years
⬎21 years
Religion
None
Catholic
Spiritualist
Umbanda/Candomblé
Jew
Reformed/Evangelical
Buddhist/Hindu
Santo Daime/Vegetal Union
Other
Practice of the religion
Special occasions only
Monthly or more
No
Socioeconomic statusa
Low
Low-average
Average
High-average
High
Head of the family education
Zero
Up to 4 years
5–8 years
9–11 years
12 years or more
Ethnic group
White
African-American
Mulato
Oriental
Indian
Otherb
Marital status
Single
Married/living together
Separated/divorced
Widow
Children
Yes
No
Employment
No
Up to 20 hr/week
⬎20 hr/week
Driver license
Yes
No
Area of undergraduate studies
Biological/health
Mathematics
Humanities
%
95% CI
SE
56.87
43.13
53.67, 60.00
40.00, 46.33
0.01
0.01
38.93
61.07
25.69, 54.04
45096, 74.31
0.07
0.07
14.85
50.00
8.90
1.77
0.67
17.42
0.64
0.20
4.97
12.25, 17.90
45.79, 54.22
7.40, 10.68
0.77, 4.04
0.37, 1.20
14.17, 21.24
0.40, 1.02
0.01, 0.44
3.39, 7.24
0.01
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
35.26
45.71
18.12
33.27, 37.30
43.26, 48.19
16.16, 20.26
0.01
0.01
0.01
0.01
1.75
25.23
62.84
10.09
0.01, 0.29
0.01, 2.37
21.64, 29.19
58.85, 66.66
7.76, 13.02
0.01
0.01
0.01
0.01
0.01
7.06
7.60
6.02
35.10
42.44
5.45, 9.10
6.01, 9.57
4.69, 7.69
31.66, 38.71
37.39, 47.66
0.01
0.01
0.01
0.01
0.02
62.27
6.50
24.71
2.45
0.11
2.96
66.05, 70.53
5.31, 7.94
21.44, 28.31
1.41, 4.22
0.57, 2.11
2.37, 3.71
0.01
0.01
0.01
0.01
0.01
0.01
75.68
20.10
3.57
0.65
63.89, 84.56
13.50, 28.84
2.03, 6.21
0.14, 2.87
0.05
0.03
0.01
0.01
19.78
79.56
11.82, 31.19
68.38, 87.51
0.04
0.04
39.65
17.85
41.85
34.95, 44.54
15.12, 20.94
35.50, 48.49
0.02
0.01
0.03
61.11
38.23
57.62, 64.49
34.97, 41.60
0.01
0.01
17.67
17.01
65.33
10.87, 27.40
11.60, 24.25
55.96, 73.64
0.04
0.03
0.04
were always performed between users and nonusers groups (e.g.,
investigating the transition from inhalants to cocaine—we tested
inhalants users vs. inhalants nonusers—these groups were defined
based on three options available in these two parts of the following
Current year in college/university
First
Second
Third
Fourth
Fifth
Sixth
All-day classes
Yes
No
Happy with the undergraduate
choice
Yes
No
a
Included in the Cox survival regression model as continuous measures. b The person does not include himself/herself within any of the
ethnic groups listed above. c We did not ask whether students had failed
in courses during their studies.
question: if the subject had already used inhalants, if the subject
had already used cocaine, and whether the subject had used the
inhalants before, in the same year, or after cocaine. Inhalants
users—if s/he just used cannabis or cocaine after inhalants. Inhalants nonusers—if s/he did not use both drugs, or used just cocaine,
or inhalants before cocaine. If s/he used both in the same year, s/he
was classified as a ‘drug seeker’). Users and nonusers were assembled for each transition to perform analyses of the transitions
from the first use of a drug to the first use of another drug.
Analyses focused on transitions of first use of inhalants with first
use of alcohol, amphetamines, cannabis, cocaine, ecstasy, hallucinogens, prescription opioids, tobacco, and tranquilizers. Figure 1
depicts the division between users and nonusers. The drug seekers
were considered drug users with the response (the use of the
subsequent drug tested in the model) of event censored. With this
we have tried to avoid loss of important data on other sociodemographic variables included in the regression models.
Time-dependent covariates. Drugs other than the pair of
drugs tested in each survival model were included as time-varying
dependent covariates in all models. (e.g., in the model testing the
pathway from cannabis to inhalants—if the individual first used
alcohol after first using cannabis and before first trying inhalants,
s/he was classified as alcohol user in the alcohol use time dependent variable of this model).
All analyses were performed using the survey settings commands with weights, strata, and primary sampling units. The
analysis of the survey data complied with the following characteristics of the sampling plan: (a) a complex sample, (b) the use of
stratification, (c) clustering, and (d) dissimilar selection probabilities. Survival analysis methods were used to compare between one
type of drug users and nonusers the cumulative prevalence of other
type of drug use. We chose to use Cox Regression Survival Models
(Cox, 1972), following an important framework (Singer & Willett,
1994). We used the statistical package STATA version 11 to run
the models with time dependent covariates (as described above)
and non–time-dependent covariates (these 14 sociodemographic
covariates are presented in Table 1) adjustment. The Kaplan–
Meier curves were plotted without any adjustment for other vari-
FIRST USE OF INHALANTS AND FIRST USE OF DRUGS
ables, and without the exclusion of individuals aged more than 40
years. We chose to use a link test to test for proportional-hazards
assumption (Cleves et al., 2010). These re-estimation tests were
carried out to test for the validity of each Cox model. It searches
for variables to add to the model. Under the assumption that the
Cox model is correctly specified, the added variables will add little
or no explanatory power, so it tests whether these variables are
insignificant. The Cox model reaches validity with p ⬎ .05 in this
link test. Cox models with p ⬍ .05 in link test were not considered
for discussion.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Results
Table 1 depicts the sociodemographic information of the sample. Table 2 presents the results of the transitions analyzed, and
Figure 2 highlights all the transitions that did not violate the Linktest of
proportional hazard-assumptions (LPHA) and could be taken into
account. Figure 3 depicts all Kaplan–Meier curves of the transitions that had significant differences that did not violate the LPHA.
Overall, the first use of inhalants was subsequent to first use of
alcohol (n ⫽ 1,181), tobacco (n ⫽ 655), and prescription opioids
(n ⫽ 64), and preceded the subsequent first use of prescription
opioids (n ⫽ 181), cannabis (n ⫽ 409), cocaine (n ⫽ 271),
tranquilizers (n ⫽ 165), and amphetamines (n ⫽ 363).
Transitions From First Use of Inhalants to First Use
of Other Drugs
Seven of nine models did not violate the LPHA. Within the valid
models, the cumulative prevalence of first use of amphetamines
(adjusted hazard ratios [aHR] ⫽ 2.59, 95% CI ⫽ 1.65– 4.07, p ⬍
.001), cannabis (aHR ⫽ 1.56, 95% CI ⫽ 1.26 –1.93, p ⬍ .001),
cocaine (aHR ⫽ 4.33, 95% CI ⫽ 3.08 – 6.09, p ⬍ .001), prescrip-
5
tion opioids (aHR ⫽ 29.39, 95% CI ⫽ 7.17–120.54, p ⬍ .01), and
tranquilizers (aHR ⫽ 2.10, 95% CI ⫽ 1.05– 4.18, p ⬍ .05) were
significantly higher among those that had first initiated inhalant
use as compared to inhalants nonusers. Regarding gender, there
was higher cumulative prevalence of alcohol and cannabis first
use among men as compared with women (aHR ⫽ 0.70, 95%
CI ⫽ 0.53– 0.93, p ⬍ .05, and aHR ⫽ 0.56, 95% CI ⫽
0.39 – 0.80, p ⬍ .01, respectively), adjusting for previous use of
inhalants. However, regarding amphetamine and tranquilizer
first use, an opposite situation occurred. The cumulative prevalence of first use in these cases was higher among women
compared with men (aHR ⫽ 5.51, 95%CI ⫽ 3.72– 8.15, p ⬍
.001, and aHR ⫽ 2.28, 95%CI ⫽ 1.48 –3.50, p ⬍ .001, respectively), adjusting for previous use of inhalants. There were no
gender differences with regard to cocaine, prescription opioid,
and tobacco first use.
Transitions From First Use of Other Drugs to
First Use of Inhalants
All models did not violate the LPHA. The cumulative prevalence of first use of inhalants was significantly higher among
alcohol (aHR ⫽ 1.41, 95% CI ⫽ 1.09 –1.81, p ⬍ .01), prescription
opioids (aHR ⫽ 3.51, 95% CI ⫽ 1.14 –10.82, p ⬍ .05), and
tobacco (aHR ⫽ 2.70, 95% CI ⫽ 1.92–3.79, p ⬍ .001) first users
as compared with nonusers of these drugs. In addition, the cumulative prevalence of first use of inhalants was significantly lower in
cocaine (aHR ⫽ 0.06, 95% CI ⫽ 0.01– 0.23, p ⬍ .001) first users
as compared with cocaine nonusers. Within all the transitions,
there was a higher cumulative prevalence of inhalant first use
among men as compared with women, adjusting for other drugs
previous use.
Table 2
Cox Survival Regression Models Results of Transitions From/to Inhalants Among University Students of 27 Brazilian Capitals, 2009
Transitiona
Genderd
From
To
Usersb
aHR
95% CI
p
aHR
95% CI
p
p (TPHA)c
Alcohol
Amphetamines
Cocaine
Ecstasy
Hallucinogens
Cannabis
Prescription opioids
Tobacco
Tranquilizers
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Inhalants
Alcohol
Amphetamines
Cocaine
Ecstasy
Hallucinogens
Cannabis
Prescription Opioids
Tobacco
Tranquilizers
1181
29
20
23
24
283
64
655
37
115
363
271
417
338
409
181
274
165
1.41
0.88
0.06
0.63
0.96
0.95
3.51
2.70
1.07
0.78
2.59
4.33
2.12
6.88
1.56
29.39
1.30
2.10
1.09, 1.81
0.47, 1.64
0.01, 0.23
0.32, 1.23
0.59, 1.56
0.72, 1.24
1.14, 10.82
1.92, 3.79
0.37, 3.10
0.43, 1.43
1.65, 4.07
3.08, 6.09
0.93, 4.84
4.14, 11.42
1.26, 1.93
7.17, 120.54
0.84, 1.99
1.05, 4.18
.007
.700
<.001
.181
.881
.729
.029
<.001
.892
.426
<.001
<.001
.071
⬍.001
<.001
<.001
.224
.035
0.58
0.60
0.67
0.61
0.61
0.56
0.61
0.66
0.67
0.70
5.51
0.69
0.51
0.52
0.56
1.59
0.85
2.28
0.45, 0.75
0.47, 0.77
0.56, 0.81
0.48, 0.78
0.48, 0.78
0.44, 0.72
0.48, 0.78
0.52, 0.83
2.55, 13.15
0.53, 0.93
3.72, 8.15
0.43, 1.11
0.26, 1.01
0.30, 0.93
0.39, 0.80
0.96, 1.86
0.69, 1.04
1.48, 3.50
<.001
<.001
<.001
<.001
<.001
<.001
<.001
.001
<.001
.017
<.001
.131
.052
.028
.002
.068
.130
<.001
.236
.177
.077
.169
.159
.141
.160
.070
.080
.308
.354
.461
.005
.006
.304
.246
.616
.700
a
All the transitions were adjusted for the intermediate use of the other eight drugs and to 14 sociodemographic variables. b Individuals who used the drug
in the second column before the drug in the first column. c Pearson coefficient of Test of Proportional-Hazard Assumption (link test). In bold are all the
transitions that did not violate the test. d Analysis of the association between the variable gender and the second drug use prevalence, adjusted for each
transition. The reference category was male.
CASTALDELLI-MAIA ET AL.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
6
Figure 2. The scheme highlights all the transitions that did not violate the linktest of proportional hazardassumptions (LPHA) and could be taken into account (see details in Table 2). The arrows were drawn every time
the Cox survival regression model reached significance (ⴱ p ⬍ .05). If the arrow is blue (gray), the cumulative
prevalence of first use of the drug in the lefthead of the arrow was higher in users than nonusers of the drug in
the righthead of the arrow (aHR ⬎ 1). If the arrow is red (gray), the cumulative prevalence of first use of the
drug in the lefthead of the arrow was lower in users than nonusers of the drug in the righthead of the arrow
(aHR ⬍ 1). See the online article for the color version of this figure.
Discussion
The main aim of the present study was to conduct an analysis of
the role of first use of inhalants within the sequence of first drug
use, and we found that it was preceded by the first use of alcohol,
tobacco, and prescription opioids. However, the first use of inhalants preceded the first use of amphetamines, cannabis, cocaine,
prescription opioids, and tranquilizers. These findings place inhalants into the position that is typically ascribed to cannabis, as its
first use is more common after the first use of legal drugs and
before the first use of illegal drugs. Additionally, cannabis first use
did not precede inhalants first use; rather, the opposite pattern was
evident.
A global debate on the validity of studying the transition of
various drug use (Van Etten et al., 1997; Wagner & Anthony,
2002; Herrera-Vázquez et al., 2004; Wells & McGee, 2008;
Caris et al., 2009; Mayet et al., 2010) was revived because of
the results of a study investigating epigenetic alterations related
to drug experimentation in laboratory animals (Levine et al.,
2011). Pretreatment of mice with nicotine increased their responses to cocaine, as assessed by addiction-related behaviors
and synaptic plasticity in the striatum, a brain region critical for
addiction-related rewards. The responses to cocaine were altered only when nicotine was administered first, and then nicotine and cocaine were administered concurrently. Reversing
the order of the drug administration was ineffective, as cocaine
had no effect on nicotine-induced behaviors and synaptic plasticity (Levine et al., 2011). There is also fresh epidemiologic
data on this field. A recent multicenter study revealed that the
sequence of drug experimentation may be linked to particular
issues related to alcohol and drug use in a specific country
(Degenhardt et al., 2010) and that violations of the classical
sequence of drug experimentation (i.e., alcohol/tobacco ¡ cannabis ¡ other drugs) seem to be more common in countries
with low lifetime use of alcohol, tobacco, and cannabis. Thus,
it is important to study first use sequencing patterns in other less
developed settings, such as Brazil, where there is different
prevalence of lifetime drug use (e.g., higher levels of lifetime
prevalence of inhalant use).
Ding et al. (2009) studied the position of inhalants in the
sequence of drug use in United States using data from the 2003
National Survey on Drug Use and Health (NSDUH). The age of
first inhalant use was compared with the age of onset of other
drug use among 6,466 inhalant users who also used at least one
of 14 other drugs. The findings indicated that only 4.2% of
multiple drug users had used inhalants before using other drugs,
in particular, alcohol, tobacco, and cannabis. Thus, the theory
that first use of inhalants is more common before the first use
of illegal drugs was not supported in this study. Comparing this
study (Ding et al., 2009) design with the current one, we have
found many differences. In the present study, we used survival
analysis methods to compare drug users and nonusers while
taking into account the cumulative prevalence of other drug use
across time. We chose to use Cox regression survival models
following an important framework (Singer & Willett, 1994;
7
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
FIRST USE OF INHALANTS AND FIRST USE OF DRUGS
Figure 3. Kaplan–Meier curves of all transitions which involved inhalants and did not violate linktest of
proportional hazard-assumptions (LPHA). See the online article for the color version of this figure.
Wagner & Anthony, 2002). Another explanation for the differences between the two studies could be trends over time. That
is, inhalants could be more commonly used in 2009 in Brazil as
compared with the United States in 2003. However, recent U.S. data
show that 17% of youths initiated inhalants during middle school
(Ober et al., 2013)—a considerably high rate of initiation.
Our findings are in line with those from Sanchez, Ribeiro, et
al. (2013) in their study with a younger sample of private school
high- and middle-school students in São Paulo, Brazil. We also
found that inhalants use preceded the cannabis use. In addition
to that transition, our study expand this role of inhalant use—
preceding the use of other illicit drugs—in a large and repre-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
8
CASTALDELLI-MAIA ET AL.
sentative sample of Brazilian urban university students. Considering that, a number of classical statements should be
reconsidered in light of findings such that in our country. First,
if there is a more common sequence of drug experimentation,
then inhalant use should be placed in the same step as (or even
before) cannabis—after legal drug use and before illegal drug
use. There are few studies regarding drug transitions that take
the position of inhalants in this sequence into account.
The first use of inhalants preceded the first use of amphetamines, cannabis, cocaine, prescription opioids, and tranquilizers. However, the magnitude of the association between prior
use of inhalants and subsequent use of these drugs varied a lot,
which leads us to believe that the chance of using some drugs
may be greater than others after using inhalants. Surely this is
the case of prescription opioids and cocaine, with the highest
aHRs. However, the absolute number of individuals who first
used a drug after using inhalants was highest for cannabis and
amphetamines, which should also call for the attention of policy
makers.
The transition to/from inhalants use seems to be a confounding factor in the association between gender and drug use. In all
transitions tested from other drugs first use to inhalants first
use, there was higher prevalence of inhalants first use in men
compared with women, which is in line with previous findings
of the prevalence of inhalants use in this sample (Andrade et al.,
2012). Regarding the use of other drugs, there was higher
cumulative prevalence of alcohol and cannabis first use among
men, and of amphetamines and tranquilizers first use among
women, which is also in line with previous findings (Andrade et
al., 2012). However, the present study did not confirm higher
prevalence of tobacco and cocaine first use among men, and
first use of tranquilizers among women, as previously found in
this sample (Andrade et al., 2012). This leads us to believe that
adjusting for previous inhalant first use made the gender differences in cocaine and tranquilizers first use disappear, positing first use of inhalants as a confounding factor. Other studies
could investigate the role of inhalants previous use as a confounding factor for associations between gender and drug
use.
There is a consensus among the few studies regarding inhalants from across all of the areas in the addiction field (Howard
et al., 2011; Konghom et al., 2010), and our results support the
implication that more attention to inhalant use and in its consequences and correlates are needed, especially in Brazil. Nonnemaker et al. (2011) suggested a strong relationship between
inhalant use, other problem behaviors, and sensation-seeking.
These results highlight the need to intervene early with youths
who are at risk of or just beginning to engage in inhalant use,
given that low-frequency inhalant users report predominately
hedonic experiences during inhalant intoxication that are not
aversive (Garland & Howard, 2010). Even nonpersistent use of
inhalants is associated with a greater risk for nonmedical opioid
and stimulant use and is a strong predictor of nonpersistent
opioid and stimulant use (Nakawaki & Crano, 2012).
Limitations
An important limitation of this study is the recall bias, which
is inherent in cross-sectional studies (Raphael, 1987). In addi-
tion, it is also difficult to extrapolate from data obtained from
university students to the Brazilian general population. College
students in this country are not representative of the general
population, as only 13.9% of young adults have access to higher
education. Almost 50% of university students study in private
institutions, which sets them apart from the general population.
However, we did use a representative sample from all of the
Brazilian capitals, thus providing balanced data from this large
country, with various social, cultural, and economic differences
(Victora et al., 2010; Parahyba & Veras, 2008). In addition, 8 of
36 transitions tested violated the LPHA), which implies the
absence of important transitions (i.e., between cannabis and
tobacco).
Another limitation is that if a respondent did not provide the
age of first use of a drug, we could not use this individual’s data
to perform any analysis of a transition that involved this drug.
In view of the large number of people who did not remember at
least one of the ages of experimenting with some of the drugs,
we decided to carry out a comparative analysis between individuals who remembered all the ages of experimentation with
drugs and those who did not remember at least one of the ages
(missing group). To this end, we analyzed the subsample of
those individuals who used drugs at least 1 drug of all the
following three groups: (i) alcohol and/or tobacco; (ii) marijuana and/or inhalants; (iii) cocaine and/or hallucinogens and/or
ecstasy and/or amphetamine and/or prescription opioids and/or
tranquilizers. From this analysis (see the Appendix), individuals
who were older, had a religion, not practiced their religion, used
hallucinogens, used ketamine, did not use tranquilizers, did not
use barbiturates, and did not use codeine syrup were associated
with the missing group. Therefore, one should take care when
extrapolating our findings to these subpopulations. Future analyses on first drug use sequencing patterns involving hallucinogens and tranquilizers would be interesting to better evaluate
these transitions.
Conclusion
This study highlights the need to intervene early with youths
who are at risk of or are just beginning to engage in inhalant
use, because this class of drugs seems to be the first to be
experimented with in our sample. There is also a call for
attention to individuals who have already first used inhalants
because of their higher chance to experiment with other drugs
such as cannabis, cocaine, and prescription drugs. It was also
interesting to find that cannabis use was subsequent compared
with inhalants within first use sequencing pattern. All these
findings show an in-transition culture of drug use among young
adults in Brazil.
References
Andrade, A. G., Duarte, P. do C., Barroso, L. P., Nishimura, R., Alberghini, D. G., & Oliveira, L. G. (2012). Use of alcohol and other drugs
among Brazilian college students: Effects of gender and age. Revista
Brasileira de Psiquiatria, 34, 294 –305. doi:10.1016/j.rbp.2012.02.002
Andrade, A. G., Queiroz, S., Villaboim, R. C. M., César, C. L. G., Alves,
M. C. G. P., & Bassit, A. Z. (1997). Uso de álcool e drogas entre alunos de
graduação da Universidade de São Paulo. Revista ABP-APAL, 19, 53–59.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
FIRST USE OF INHALANTS AND FIRST USE OF DRUGS
Balster, R. L., Cruz, S. L., Howard, M. O., Dell, C. A., & Cottler, L. B.
(2009). Classification of abused inhalants. Addiction, 104, 878 – 882.
doi:10.1111/j.1360-0443.2008.02494.x
Caris, L., Wagner, F. A., Rios-Bedoya, C. F., & Anthony, J. C. (2009).
Opportunities to use drugs and stages of drug involvement outside the
United States: Evidence from Republic of Chile. Drug and Alcohol
Dependence, 102, 30 –34. doi:10.1016/j.drugalcdep.2008.12.004
Castaldelli-Maia, J. M., Bhugra, D., de Andrade, A. G., & Lotufo-Neto, F.
(2013). Substance use and misuse in Brazilian Movies (2000 –2008).
Substance Use & Misuse, 48, 248 –257. doi:10.3109/10826084.2012
.753549
Castaldelli-Maia, J. M., Martins, S. S., de Oliveira, L. G., de Andrade,
A. G., & Nicastri, S. (2014). The role of drug use sequencing pattern in
further problematic use of alcohol, tobacco, cannabis, and other drugs.
Journal of Mental Health [in press].
Castaldelli-Maia, J. M., Martins, S. S., de Oliveira, L. G., van Laar, M., de
Andrade, A. G., & Nicastri, S. (2014). Use transitions between illegal
drugs among Brazilian university students. Social Psychiatry and Psychiatric Epidemiology, 49, 385–394. doi:10.1007/s00127-013-0748-1
Castaldelli-Maia, J. M., Oliveira, H. P., Andrade, A. G., Lotufo-Neto, F.,
& Bhugra, D. (2012). Using selected scenes from Brazilian films to
teach about substance use disorders, within medical education. Sao
Paulo Medical Journal, 130, 380 –391. doi:10.1590/S1516-318020
12000600005
Cleves, M., Gutierrez, R. G., Gould, W., & Marchenko, Y. V. (2010). An
introduction to survival analysis using Stata (3rd ed.). College Station,
TX: Stata Press.
Cox, D. R. (1972). Regression models and life time tables. Journal of the
Royal Statistical Society (Series B), 34, 187–202.
Degenhardt, L., Dierker, L., Chiu, W. T., Medina-Mora, M. E., Neumark,
Y., Sampson, N., . . . Kessler, R. C. (2010). Evaluating the drug use
“gateway” theory using cross- national data: Consistency and associations of the order of initiation of drug use among participants in the
WHO World Mental Health Surveys. Drug and Alcohol Dependence,
108, 84 –97. doi:10.1016/j.drugalcdep.2009.12.001
Ding, K., Chang, G. A., & Southerland, R. (2009). Age of inhalant first
time use and its association to the use of other drugs. Journal of Drug
Education, 39, 261–272. doi:10.2190/DE.39.3.c
Fergusson, D. M., Boden, J. M., & Horwood, L. J. (2006). Cannabis use
and other illicit drug use: Testing the cannabis gateway hypothesis.
Addiction, 101, 556 –569. doi:10.1111/j.1360-0443.2005.01322.x
Fonseca, A. M., Galduroz, J. C., Noto, A. R., & Carlini, E. L. (2010).
Comparison between two household surveys on psychotropic drug use in
Brazil: 2001 and 2004. Ciencia e Saude Coletiva, 15, 663– 670.
Garland, E. L., & Howard, M. O. (2010). Phenomenology of adolescent
inhalant intoxication. Experimental and Clinical Psychopharmacology,
18, 498 –509. doi:10.1037/a0021737
Herrera-Vázquez, M., Wagner, F. A., Velasco-Mondragon, E., Borges, G., &
Lazcano-Ponce, E. (2004). Inicio en el consumo de alcohol y tabaco y
transición a otras drogas en estudiantes de Morelos, México. Salud Publica
de México, 46, 132–140. doi:10.1590/S0036-36342004000200007
Howard, M. O., Bowen, S. E., Garland, E. L., Perron, B. E., & Vaughn,
M. G. (2011). Inhalant use and inhalant use disorders in the United
States. Addiction Science and Clinical Practice, 6, 18 –31.
Hynes-Dowell, M., Mateu-Gelabert, P., Barros, H. M. T., & Delva, J.
(2011). Volatile substance misuse among high school students in South
America. Substance Use & Misuse, 46, 27–34. doi:10.3109/10826084
.2011.580192
Kandel, D. B. (1975). Stages in adolescent involvement in drug use.
Science, 190, 912–914. doi:10.1126/science.1188374
Kandel, D., & Faust, R. (1975). Sequence and stages in patterns of
adolescent drug use. Archives of General Psychiatry, 32, 923–932.
doi:10.1001/archpsyc.1975.01760250115013
9
Levine, A., Huang, Y., Drisaldi, B., Griffin, E. A.Jr., Pollak, D. D., Xu, S.,
. . . Kandel, E. R. (2011). Molecular mechanism for a gateway drug:
Epigenetic changes initiated by nicotine prime gene expression by
cocaine. Science Translational Medicine, 3, 107–109. doi:10.1126/
scitranslmed.3003062
Malta, D. C., Sardinha, L. M., Mendes, I., Barreto, S. M., Giatti, L., Castro,
I. R., . . . Crespo, C. (2010). Prevalence of risk health behavior among
adolescents: Results from the 2009 National Adolescent School-based
Health Survey (PeNSE). Ciencia e Saude Coletiva, 15, 3009 –3019.
doi:10.1590/S1413-81232010000800002
Mayet, A., Lagleye, S., Chau, N., & Falissard, B. (2010). The mediation
role of licit drugs in the influence of socializing on cannabis use among
adolescents: A quantitative approach. Addictive Behaviors, 35, 890 –
895. doi:10.1016/j.addbeh.2010.06.001
Mesquita, A. M., de Andrade, A. G., & Anthony, J. C. (1998). Use of the
inhalant lança by Brazilian medical students. Substance Use & Misuse,
33, 1667–1680. doi:10.3109/10826089809058949
Moreira, M. M., Barbosa, G. L., Laranjeira, R., & Mitsuhiro, S. S. (2014).
Alcohol and crack/cocaine use in women: A 14-year cross-sectional
study. Journal of Addiction Diseases. [Epub ahead of print]
Nakawaki, B., & Crano, W. D. (2012). Predicting adolescents’ persistence,
non-persistence, and recent onset of nonmedical use of opioids and
stimulants. Addictive Behaviors, 37, 716 –721. doi:10.1016/j.addbeh
.2012.02.011
Nonnemaker, J. M., Crankshaw, E. C., Shive, D. R., Hussin, A. H., &
Farrelly, M. C. (2011). Inhalant use initiation among U.S. adolescents:
Evidence from the National Survey of Parents and Youth using discretetime survival analysis. Addictive Behaviors, 36, 878 – 881. doi:10.1016/
j.addbeh.2011.03.009
Ober, A. J., Miles, J. N., Ewing, B., Tucker, J. S., & D’Amico, E. J. (2013).
Risk for inhalant initiation among middle school students: Understanding individual, family, and peer risk and protective factors. Journal of
Studies on Alcohol and Drugs, 74, 835– 840.
Opaleye, E. S., Ferri, C. P., Locatelli, D. P., Amato, T. C., & Noto, A. R.
(2014). Nonprescribed use of tranquilizers and use of other drugs among
Brazilian students. Revista Brasileira de Psiquiatria, 36, 16 –23. doi:
10.1590/1516-4446-2013-1180
Parahyba, M. I., & Veras, R. (2008). Socio-demographic differentials in the
functional decline among the elderly in Brazil. Ciencia & Saude Coletiva, 13, 1257–1264. doi:10.1590/S1413-81232008000400022
Raphael, K. (1987). Recall bias: A proposal for assessment and control.
International Journal of Epidemiology, 16, 167–170. doi:10.1093/ije/16
.2.167
Remy, L., Narvaez, J., Sordi, A., Guimarães, L. S., Von Diemen, L.,
Surratt, H., . . . Pechansky, F. (2013). Correlates of unprotected sex in a
sample of young club drug users. Clinics (Sao Paulo), 68, 1384 –1391.
doi:10.6061/clinics/2013(11)01
Sanchez, Z. M., Noto, A. R., & Anthony, J. C. (2013). Social rank and
inhalant drug use: The case of lança perfume use in São Paulo, Brazil.
Drug and Alcohol Dependence, 131, 92–99. doi:10.1016/j.drugalcdep
.2012.12.001
Sanchez, Z. M., Ribeiro, L. A., Moura, Y. G., Noto, A. R., & Martins, S. S.
(2013). Inhalants as intermediate drugs between legal and illegal drugs
among middle and high school students. Journal of Addictive Diseases,
32, 217–226. doi:10.1080/10550887.2013.795472
Silva-Oliveira, F., Jorge, K. O., Ferreira e Ferreira, E., Vale, M. P.,
Kawachi, I., & Zarzar, P. M. (2014). The prevalence of inhalant use and
associated factors among adolescents in Belo Horizonte, Brazil. Ciencia
e Saude Coletiva, 19, 881– 890. doi:10.1590/1413-81232014193
.07542013
Singer, J. D., & Willett, J. B. (1994). Designing and analyzing studies of
onset, cessation, and relapse: Using survival analysis in drug abuse
research. NIDA Research Monograph, 142, 196 –263.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
10
CASTALDELLI-MAIA ET AL.
Stempliuk, V. A., Barroso, L. P., Andrade, A. G., Nicastri, S., & Malbergier, A. (2005). Comparative study of drug use among undergraduate
students at the University of São Paulo-São Paulo campus in 1996 and
2001. Revista Brasileira de Psiquiatria, 27, 185–193. doi:10.1590/
S1516-44462005000300006
Tarter, R. E., Vanyukov, M., Kirisci, L., Reynolds, M., & Clark, D. B.
(2006). Predictors of marijuana use in adolescents before and after licit
drug use: Examination of the gateway hypothesis. The American Journal
of Psychiatry, 163, 2134 –2140. doi:10.1176/appi.ajp.163.12.2134
Van Etten, M. L., Neumark, Y. D., & Anthony, J. C. (1997). Initial
opportunity to use marijuana and the transition to first use: United States,
1979 –1994. Drug and Alcohol Dependence, 49, 1–7. doi:10.1016/
S0376-8716(97)00127-0
Vaughn, M. G., Wallace, J., Perron, B. E., Copeland, V., & Howard, M. O.
(2008). Does marijuana use serve as a gateway to cigarette use for
African-Americans? The American Journal of Drug and Alcohol Abuse,
34, 782–791. doi:10.1080/00952990802455477
Victora, C. G., Matijasevich, A., Silveira, M., Santos, I., Barros, A. J., &
Barros, F. C. (2010). Socio-economic and ethnic group inequities in
antenatal care quality in the public and private sector in Brazil. Health
Policy and Planning, 25, 253–261. doi:10.1093/heapol/czp065
Wagner, F. A., & Anthony, J. C. (2002). Into the world of illegal drug use:
Exposure opportunity and other mechanisms linking the use of alcohol,
tobacco, marijuana, and cocaine. American Journal of Epidemiology,
155, 918 –925. doi:10.1093/aje/155.10.918
Wagner, G. A., Stempliuk, V. A., Zilberman, M. L., Barroso, L. P., &
Andrade, A. G. (2007). Alcohol and drug use among university students:
Gender differences. Revista Brasileira de Psiquiatria, 29, 123–129.
doi:10.1590/S1516-44462006005000033
Wells, J. E., & McGee, M. A. (2008). Violations of the usual sequence of drug
initiation: Prevalence and associations with the development of dependence in
the New Zealand Mental Health Survey. Journal of Studies on Alcohol
and Drugs, 69, 789 –795.
(Appendix follows)
FIRST USE OF INHALANTS AND FIRST USE OF DRUGS
11
Appendix
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Results of Logistic Regression Comparing Individuals Who Used at Least 3 Drugs and Were Reminded of All the
Ages of Experimentation of These Drugs and Those Who Any Age of Experimentation Was Missing Among
Brazilian University Students, 2009
Variable
aOR
SE
p
Age
Gender
Type of religion
Practicing the religion
Scholarity of the head of the family
Socioeconomic status
Ethnic group
Civil status
Children
Currently working
Driver’s license
University course area
Semester of the course
All-day course
Happy with the course choice
Alcohol use
Tobacco use
Cannabis use
Inhalants use
Cocaine use
Hallucinogens use
Ketamine use
Ayahuasca use
Ecstasy use
Steroids use
Tranquilizers use
Sedative/barbiturates use
Opioid analgesics use
Codeine syrup use
Anticholinergics use
Heroin use
Amphetamine use
Synthetic drugs use
1.92
1.01
1.27
0.64
0.81
1.04
0.69
0.78
0.56
1.22
0.56
1.18
0.96
0.98
0.81
0.90
0.83
1.19
0.99
1.23
2.59
20.85
1.49
0.62
0.59
0.54
0.17
0.45
0.05
2.00
6.63
0.69
1.97
0.51
0.23
0.11
0.12
0.12
0.03
0.16
0.20
0.27
0.17
0.16
0.17
0.09
0.31
0.33
0.86
0.15
0.38
0.33
0.38
1.20
25.32
0.85
0.23
0.27
0.11
0.13
0.19
0.04
1.83
11.47
0.28
1.93
.018
.936
.008
.023
.177
.238
.125
.364
.250
.166
.056
.244
.735
.954
.607
.918
.330
.591
.979
.498
.044
.014
.486
.217
.259
.005
.028
.070
.002
.452
.277
.391
.490
[95% CI]
1.12,
0.64,
1.06,
0.44,
0.60,
0.97,
0.43,
0.46,
0.21,
0.91,
0.31,
0.88,
0.79,
0.51,
0.36,
0.13,
0.57,
0.62,
0.50,
0.66,
1.02,
1.86,
0.47,
0.29,
0.24,
0.35,
0.03,
0.18,
0.01,
0.32,
0.21,
0.30,
0.28,
3.28
1.61
1.52
0.94
1.09
1.11
1.10
1.32
1.50
1.63
1.01
1.58
1.17
1.86
1.82
6.06
1.20
2.27
1.95
2.30
6.55
233.25
4.65
1.32
1.46
0.83
0.82
1.06
0.33
1.23
206.65
1.59
1.39
Received April 2, 2014
Revision received July 10, 2014
Accepted July 11, 2014 䡲