- Email: [email protected]
Chronic childhood adversity and stages of substance use involvement in adolescents
Drug and Alcohol Dependence 131 (2013) 85–91
Contents lists available at SciVerse ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Chronic childhood adversity and stages of substance use involvement in adolescents Corina Benjet a,∗ , Guilherme Borges a , María Elena Medina-Mora a , Enrique Méndez a,b a Department of Epidemiological and Psychosocial Research, National Institute of Psychiatry Ramón de la Fuente, Calzada México Xochimilco 101, Colonia San Lorenzo Huipulco, Mexico City 14370, Mexico b Institute for Applied Mathematical and Systems Research, National Autonomous University of Mexico, Ciudad Universitaria, Mexico City 04510, Mexico
a r t i c l e
i n f o
Article history: Received 18 September 2012 Received in revised form 30 November 2012 Accepted 2 December 2012 Available online 29 December 2012 Keywords: Alcohol Drugs Substance use Substance abuse Adversity Adolescence
a b s t r a c t Background: Studies have shown that those who experience chronic childhood adversity have a greater likelihood of substance abuse and dependence. However, substance use disorders are first preceded by substance use, and substance use is preceded by substance use opportunities. This study aims to estimate the association of chronic adversity with different stages of substance involvement: opportunities, use given the opportunity and abuse or dependence given use. Methods: 3005 adolescents aged 12–17 were interviewed in a stratified multistage general population probability survey of Mexico City, Mexico. Substance involvement and chronic childhood adversities were assessed with the World Mental Health Composite International Diagnostic Interview Adolescent Version (WMH-CIDI-A). Discrete-time survival models were performed; their survival coefficients and standard errors were exponentiated, and reported as odds-ratios (ORs). Results: Childhood adversities were associated with alcohol opportunity, alcohol use and alcohol abuse/dependence with significant ORs for individual adversities ranging from 1.4 to 4.1. Childhood adversities were also associated with illicit drug opportunity, drug use and drug abuse/dependence with significant ORs for individual adversities ranging from 1.6 to 17.3. Having more adversities was associated with greater incremental odds of substance involvement, particularly drug use given the opportunity. Conclusions: While adversities are mostly related to transitioning into use and disorder, a few are related to substance opportunities, particularly those which were likely to make substances available through parents. Attending to the needs of youth living in adversity, particularly adversities related to parental dysfunction and child abuse should be integral to addiction prevention efforts. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Epidemiological studies have shown that people who experience chronic childhood adversity have a greater likelihood of substance abuse and dependence as well as other psychiatric disorders (Benjet et al., 2010; Enoch et al., 2010; Kessler et al., 2010; Turner and Lloyd, 2003; Lloyd and Turner, 2008). Likewise, animal models have shown stress and early adversity to increase self-administration of substances (Becker et al., 2011; Goeders and Guerin, 1996; Lopez et al., 2011; Lynch et al., 1999; Goeders, 2002; Vengeliene et al., 2003; Koob and Kreek, 2007) while human models have shown that stress is related to substance abuse and dependence and to relapse and craving amongst
∗ Corresponding author at: National Institute of Psychiatry Ramón de la Fuente, Calzada México Xochimilco 101, Colonia San Lorenzo Huipulco, Mexico City 14370, Mexico. Tel.: +52 55 4160 5332; fax: +52 55 5513 3446. E-mail address: [email protected] (C. Benjet). 0376-8716/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.drugalcdep.2012.12.002
substance abusers (Higley et al., 2011; Sinha et al., 2011, 2006). Many theories have been proposed to explain this association. Among the most frequently mentioned in the addiction literature is the self-medication hypothesis, in which those who abuse substances do so to ease or relieve the pain of trauma, negative affect and psychiatric disorders (Garland et al., 2012; Khantzian, 1985; Robinson et al., 2011). Compatible with the self -medication hypothesis are theories which propose neurobiological pathways that alter learning, reward, craving, and impulsivity through stress allostasis. The theory of allostasis or allostatic load posits that the process of achieving stability through alteration of neural, neuroendocrine and immune mechanisms (allostasis), while adaptive in the short run, becomes “overloaded” (allostatic load) with chronic stress (McEwen, 2000). Frequent, multiple stressors, failure to habituate to chronic stressors, compensatory hyperactivity and delayed shutdown of the stress response alter key neurocircuitry that impacts upon craving, loss of control and compulsion and, thus, vulnerability to addiction (McEwen, 2000; Uhart and Wand, 2009).
86
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
The types of chronic childhood adversities typically included in epidemiological studies fall into one of four categories: family or parental dysfunction, abuse/neglect, interpersonal loss and socioeconomic disadvantage. Certain epidemiological questions regarding adversity and stage of drug involvement have yet to be addressed and may contribute to further understanding of this relationship. Substance use disorders are first preceded by substance use and substance use is preceded by substance use opportunities. Substance opportunities are generally considered a measure of availability of substances in a given context, although they may also reflect, to some degree, actively seeking out opportunities. Prior work has found that certain factors related to substance use and abuse are actually related to exposure to substance opportunities (Benjet et al., 2007a; Chen et al., 2004; Storr et al., 2011). For example, several studies have shown that males are more likely to have opportunities to use drugs than females, but females are equally as likely as males to try drugs when given the chance (Benjet et al., 2007a; Delva et al., 1999; Van Etten et al., 1999; Van Etten and Anthony, 1999). Another study found drug use opportunities to mediate the association between mental illness and drug use (Liang et al., 2011). An important question to address is whether the association between chronic childhood adversity and substance use disorders is due to those with childhood adversity having greater substance use opportunities, being more likely to use given the opportunity, being more likely to develop substance use problems given use, or all of the above. Given the reported associations of psychiatric disorder with both childhood adversity and substance involvement, it is also important to determine whether the association between adversity and substance involvement holds when controlling for psychiatric disorders. This study aims to address these questions. 2. Methods 2.1. Participants This report presents secondary analyses of data from the Mexican Adolescent Mental Health survey (Benjet et al., 2009a). The sample consists of 3005 adolescents (52.1% female) aged 12–17, selected from a stratified multistage area probability sample representative of the nearly two million adolescents residing in the Mexico City Metropolitan Area. In all strata, the primary sampling units were census count areas cartographically defined and updated for the XII Population and Housing Census in 2000. Secondary sampling units were city blocks (or groups of them), selected with probability proportional to size. All households within these city block units with adolescents in the age range were selected. One eligible member from each of these households was randomly selected using the Kish method of random number charts. The response rate of eligible respondents was 71%. 2.2. Procedures A verbal and written explanation of the study was given to both parents and adolescents, after which signed informed consent from a parent or legal guardian was obtained, as well as the assent of the adolescent. Interviews were conducted in the homes of the selected participants. All study participants and their families were offered information on local mental health services. The Internal Review Board of the National Institute of Psychiatry approved the recruitment, consent and field procedures. 2.3. Measures Substance involvement, adversity and psychiatric disorders were evaluated with the fully structured, computer assisted, World Mental Health adolescent version of the Composite International Diagnostic Interview (WMH-CIDI-A) the development and validity of which is described elsewhere (Kessler et al., 2009; Merikangas et al., 2009). The computer assisted version was administered, in which extensively trained lay interviewers read the questions to the participant directly from the computer screen, the questions are chosen by the computer based on previous responses of the participant and complex logical skip patterns. The interviewer inputs the respondent’s answers directly into the computer and consistency checks are programmed so that inconsistent information is probed and corrected. Diagnostic classification is based on meeting the diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; American Psychiatric Association, 1994). The substance use section includes questions
regarding the lifetime use (defined as consumption of the substance at least once at any time in one’s life) of alcohol and illicit drugs, opportunity to use alcohol and drugs, and substance abuse and dependence. The illicit drugs included were marijuana, cocaine in any of its presentations, tranquilizers or stimulants used without a medical prescription, such as methamphetamine, and other substances (e.g., heroin, inhalants, LSD, etc.) which were grouped as “other drugs”. Participants were asked about each category of drugs openly and then presented with a list of numerous different street names for these drugs. The questions regarding opportunity to use alcohol and drugs were posed after the questions about alcohol and drug use so that opportunities to use alcohol and drugs referred to all the alcohol and drugs previously presented to the respondents. All participants were asked about opportunities regardless of whether they had previously endorsed using any substance. Opportunity to use alcohol and opportunity to use illicit drugs were asked about separately and defined as having the opportunity to use any substance, independently of whether or not the respondent did so. For example, someone offered the respondent drugs or the respondent was present when others were consuming and could have done so if he or she chose to. In order to cover a range of chronic childhood adversities and those for which previous research has found associations with substance abuse, we included different forms of parental loss, parental psychopathology, abuse, physical illness and economic adversity; these were evaluated from the childhood and post-traumatic stress disorder sections of the WMH-CIDI-A. Each of 12 chronic adversities was classified as present or not present using the same criteria as the World Mental Health Survey Initiative (Kessler et al., 2010). Physical abuse and parental violence were assessed with a modified version of the Conflict Tactics Scale (Strauss, 1979). Neglect was evaluated with questions often used in child welfare studies (Courtney et al., 1998). Sexual abuse was assessed by reading a definition of rape and then asking about other forms of abuse or molestation. In order to be consistent with the other World Mental Health Surveys; chronic sexual abuse was defined as reporting at least three episodes of sexual abuse thus representing chronic sexual abuse as opposed to acute one time trauma (Bruffaerts et al., 2010). To assess parental loss, the adolescents were asked whether they lived with both parents all of their lives. Those who did not were asked whether this was because their parents had separated or divorced, a parent had died or some other reason. Those mentioning separations of six months or more from either parent for some other reason were classified as other parental loss, with reasons ranging from having gone to boarding school, having left home, or that their parent was in prison. Parental pathology was evaluated using questions from the Family History Research Diagnostic Criteria Interview and included parental mental illness, substance problems, and criminal behavior (Endicott et al., 1978). Participants were considered to have experienced family economic adversity if the family ever received money from a government assistance program for poor families or by lack of parental employment most or all of the time during the participant’s childhood. Serious physical illness is based on the adolescent’s report of having experienced a life-threatening physical illness. Mood, anxiety and behavioral disorders are associated with childhood adversity and with substance involvement. In order to estimate the association of adversity with substance use and substance use disorder without the possible confounding of psychiatric disorders, the following psychiatric disorders were also evaluated with the WMH-CIDI according to DSM-IV criteria and were used as a covariate in the statistical models described below: major depression, dysthmia, bipolar I and II, generalized anxiety disorder, panic disorder, agoraphobia, specific phobia, social phobia, post-traumatic stress disorder, separation anxiety, eating disorders, attention deficit hyperactivity disorder, oppositional-defiant disorder, conduct disorder, and intermittent explosive disorder. The WMH-CIDI assessed age-of-onset of substance opportunity, use and disorder as well as for other psychiatric disorders retrospectively using a series of questions shown to improve accuracy of retrospect reports and avoid implausible response patterns (such as use of anchoring events and quality control programs to search for response inconsistencies; Knauper et al., 1999). 2.4. Statistical analysis Data were weighted to adjust for differential probabilities of selection and nonresponse as well as post-stratification to the total Mexico City Metropolitan Area adolescent population according to the year 2000 Census in the target age and sex range. The socio-demographic distribution of the sample closely approximates the target population, in that roughly half are female, there is an even distribution of ages, two-thirds live with both parents, and the socio-economic level of the parents represents the educational and income levels in Mexico; for example, one-fourth of parents have 6 years or less of formal education whereas only 13% have gone to college. More details of the weighted and un-weighted socio-demographic distribution of the sample is presented elsewhere (Benjet et al., 2007a, 2009a) as is the distribution of chronic adversity (Benjet et al., 2009b). First we present the prevalence estimates for each substance involvement stage. Then multivariate associations of the childhood adversities with substance opportunity, use and abuse/dependence were estimated using discrete-time survival analysis with person-years as the unit of analysis (Efron, 1988; Willett and Singer, 1993). Each model controlled for the adolescent’s age at interview, gender, psychiatric disorders prior to substance involvement and person-year. Psychiatric disorder and substance involvement are time varying variables whereas
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91 Table 1 Lifetime prevalence of substance involvement.
Alcohol opportunity Alcohol use amongst those w/opportunity Alcohol abuse/dependence amongst those w/use Drug opportunity Drug use amongst those w/opportunity Drug abuse/dependence amongst those w/use a
n (unweighted)
% (weighted)
SEa
2361 1693
79.9 73.4
0.6 0.9
88
6.4
0.8
784 134
28.7 18.1
0.5 1.1
41
32.1
5.1
Standard error.
adversities are assumed to predate substance involvement. The models for substance use represent the association of adversity with substance use only among those with opportunities for use. The models for substance abuse or dependence represent the association of adversity with substance abuse/dependence only among those with substance use. We initially tested the interaction of each individual adversity with gender on alcohol and drug stages, but only 7 of the 72 gender-byadversity interactions were significant at the p < 0.05 level and were inconsistent with regards to which gender had greater odds. Therefore the final models, while controlling for gender did not include adversity-by-gender interaction terms. The final models also included each of the 12 predictors for type of adversity and additional predictors for number of family dysfunction adversities (i.e., parental mental illness, parental substance problems, parental criminal behavior, witnessing family violence, physical abuse, sexual abuse and neglect) and number of other adversities (parental death, parental divorce, other parental loss, physical illness and economic adversity) so that the odds ratio for each adversity may be considered an estimation of the “pure effect” of that adversity controlling for other adversities and number of adversities (Kessler et al., 2010). To estimate the effects of number of adversities, the models used a series of dummy predictor variables for number of adversities (e.g., one such variable for respondents who experienced exactly one adversity, another for respondents who experienced two or three adversities and another for those who experienced four or more adversities) without information about the types of adversities experienced. Additional models with gender by number-of-adversity terms were also performed, but because only one model had significant interactions these are not presented on the tables. The survival coefficients and their standard errors were exponentiated and are reported in the form of odds-ratios (OR) and 95% confidence intervals (95% CI). As a result of this complex sample design and weighting, estimates of standard errors for proportions were obtained by the Taylor series linearization method using the SUDAAN Software (Research Triangle Institute, 2002).
3. Results The prevalence of lifetime substance involvement in this representative sample of Mexican adolescents is provided in Table 1. Most adolescents (79.9%) have had alcohol opportunities, of those, a majority (73.4%) have used alcohol and few (6.4%) of those who have used, have developed alcohol abuse or dependence. With regards to illicit drugs, less than a third have had drug use opportunities. Of those with drug opportunities, 18.1% have tried drugs and of those, 32.1% have developed abuse or dependence. Table 2 presents the multivariate models for the association of childhood adversity and alcohol opportunities, alcohol use given the opportunity and alcohol abuse/dependence given use. When controlling for other adversities, number of adversities, and psychiatric disorders, parental criminal behavior, family violence and sexual abuse show significantly elevated odds (OR = 1.5, 95% CI = 1.1–1.9; OR = 1.6, 95% CI = 1.2–1.9 and OR = 1.9, 95% CI = 1.1–3.1, respectively) for alcohol opportunities. Given the opportunity to use, parent criminal behavior and sexual abuse have 1.4 (95% CI = 1.1–1.9) and 2.2 (95% CI = 1.5–3.2) times the odds of alcohol use, respectively. Given alcohol use, parental criminal behavior has four times the odds (95% CI = 1.9–9.2) and other parental loss twice the odds (95% CI = 1.2–3.9) of transitioning to alcohol abuse or dependence. Multivariate models for the association of childhood adversity and drug opportunities, use given the opportunity and drug abuse
87
or dependence are shown in Table 3. Parental substance problems, parental criminal behavior, and physical abuse all have significantly greater odds of drug opportunities (OR = 1.6, 95% CI = 1.0–2.5; OR = 1.7, 95% CI = 1.0–2.7 and OR = 1.6, 95% CI = 1.2–2.1, respectively). For drug use given the opportunity, these same adversities, plus family violence and neglect, are associated with larger odds ratios from 2.1 (95% CI = 1.0–4.2) for physical abuse to 3.5 (95% CI = 1.1–10.7) for neglect. Amongst drug users, parental criminal behavior, physical abuse and parental death are associated with elevated odds of transitioning to drug abuse or dependence (OR = 17.3, 95% CI = 2.7–110.8; OR = 9.1, 95% CI = 1.2–66.8 and OR = 6.7, 95% CI = 1.4–31.2, respectively). Table 4 shows the association of number of adversities with alcohol opportunity, alcohol use given the opportunity and alcohol abuse/dependence given use. A greater number of adversities is associated with increasingly greater odds of alcohol use given the opportunity, but is not associated with other stages of alcohol involvement. Table 5 shows incremental odds of drug opportunity and transition to drug use with increasing number of adversities. Odds ratios for drug opportunity range from 1.4 (95% CI = 1.2–1.8) for experiencing two or three adversities to 2.7 (95% CI = 1.8–4.0) for reporting four or more adversities. Odds ratios for drug use given the opportunity range from 3.9 for one adversity (95% CI = 1.9–7.8) to 5.1 (95% CI = 2.2–11.5) for four or more adversities. Number of adversities is not associated to drug abuse/dependence overall, but there is a significant interaction with sex (data not shown but available upon request), such that the number of adversities is associated with the transition to drug abuse/dependence for females only (p < 0.001).
4. Discussion In the Mexico City Metropolitan Area, alcohol use among adolescents is normative and accepted, as evidenced by the majority of adolescents having had exposure to alcohol opportunity and having used alcohol given the opportunity; drugs are less available and drug use is less tolerated than alcohol, as suggested by the fact that less than a third have had drug opportunities and less than a fifth of those with the opportunity have tried drugs. All models estimating the association of adversity with substance involvement stages were significant, however, not all individual adversities were independently associated to substance involvement. One particular adversity, parental criminal behavior, stands out for its consistent association in all models, from a 40% increased odds of alcohol use given the opportunity, to 17 times the odds for drug abuse or dependence given use. One explanation might be that this association reflects a genetic predisposition, such as risk taking, impulsivity, novelty seeking, or antisocial personality (Kreek et al., 2005), but cannot be explained by a genetic predisposition toward addiction as these models control for parental substance problems. Another explanation might be that parental criminal behavior involves a greater degree of chronic stress. It may also be that those willing to disclose parental criminal behavior are also more willing to disclose their own substance involvement. In general, family dysfunction adversities, such as witnessing family violence, physical and sexual abuse and neglect, have odds ratios above one, though not all are statistically significant in all models. Adversities were not only related to substance use and the development of abuse or dependence, but were also associated to substance opportunities: parental criminal behavior, witnessing family violence and sexual abuse, in the case of alcohol opportunities, and parental criminal behavior, parental substance problems and physical abuse, in the case of drug opportunities. Opportunity is usually considered an indirect measure of availability in a given context, but may also reflect, to some extent, those who search for
88
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
Table 2 Estimated effects (odds ratios) of chronic adversity on risk for alcohol opportunity, alcohol use given the opportunity and alcohol abuse/dependence given use in multivariate survival models. Alcohol opportunity
Alcohol use given opportunity
Alcohol abuse/dependence given use
Multivariate modelb
Multivariate modelb
Multivariate modelb
n
OR
OR
OR
352 162 188 582 413 50 173 169 413 413 205 738
1.01 (0.77–1.32) 1.09 (0.75–1.59) 1.46 (1.12–1.90) 1.55 (1.24–1.93) 0.99 (0.69–1.42) 1.85 (1.12–3.05) 1.01 (0.71–1.46) 1.06 (0.82–1.38) 0.97 (0.77–1.22) 0.98 (0.77–1.24) 1.38 (0.99–1.91) 1.07 (0.92–1.25) 2 12 = 44.18 (p < 0.0001)
a
Parental mental illness Parental substance problems Parental criminal behavior Witnessing family violence Physical abuse Sexual abuse Neglect Parent died Parent divorce Other parent loss Physical illness Economic adversity
(95% CI)
(95% CI)
1.25 (0.97–1.61) 1.32 (0.96–1.82) 1.37 (1.02–1.85) 1.24 (0.91–1.68) 1.26 (0.90–1.75) 2.21 (1.52–3.23) 1.14 (0.73–1.77) 0.96 (0.62–1.51) 1.08 (0.82–1.41) 1.14 (0.85–1.53) 0.93 (0.58–1.49) 0.92 (0.74–1.15) 2 12 = 35.84 (p < 0.0003)
(95% CI)
1.37 (0.58–3.24) 0.86 (0.30–2.48) 4.14 (1.86–9.21) 1.84 (0.83–4.09) 1.35 (0.72–2.53) 0.42 (0.13–1.34) 1.24 (0.48–3.25) 0.21 (0.04–1.18) 0.68 (0.20–2.30) 2.16 (1.18–3.94) 1.21 (0.38–3.87) 1.54 (0.85–2.78) 2 12 = 127.22 (p < 0.0001)
a
Weighted n for each individual adversity. Discrete-time survival model with person-year the unit of analysis controlling for age, gender, person-year, psychiatric diagnosis prior to substance involvement, type of adversity, number of family dysfunction adversities (parental mental illness, parental substance problems, parental criminal behavior, witnessing family violence, physical abuse, sexual abuse and neglect), and number of other adversities (parent died, parental divorce, other parental loss, physical illness and economic adversity). b
Table 3 Estimated effects (odds ratios) of chronic adversity on risk for drug opportunity, drug use given the opportunity and drug abuse/dependence given use in multivariate survival models. Drug opportunity
Drug use given the opportunity
Drug abuse/dependence given use
Multivariate modelb
Multivariate modelb
Multivariate modelb
n
OR
OR
OR
352 162 188 582 413 50 173 169 413 413 205 738
1.14 (0.77–1.68) 1.62 (1.04–2.52) 1.65 (1.01–2.71) 1.19 (0.84–1.69) 1.56 (1.17–2.09) 1.60 (0.88–2.90) 1.19 (0.78–1.83) 1.01 (0.72–1.41) 1.10 (0.82–1.47) 1.34 (0.98–1.84) 1.19 (0.85–1.67) 0.95 (0.72–1.24) 2 12 = 29.80 (p < 0.0030)
a
Parental mental illness Parental substance problems Parental criminal behavior Witnessing family violence Physical abuse Sexual abuse Neglect Parent died Parent divorce Other parent loss Physical illness Economic adversity
(95% CI)
(95% CI)
1.50 (0.74–3.01) 2.36 (1.02–5.42) 2.61 (1.26–5.43) 3.11 (1.51–6.42) 2.07 (1.02–4.23) 1.68 (0.52–5.42) 3.45 (1.11–10.74) 1.64 (0.41–6.59) 1.54 (0.76–3.11) 1.35 (0.67–2.72) 0.64 (0.26–1.59) 1.44 (0.67–3.10) 2 12 = 38.87 (p < 0.0001)
(95% CI)
6.94 (0.76–62.94) 0.75 (0.10–5.44) 17.33 (2.71–110.81) 0.62 (0.12–3.21) 9.06 (1.23–66.82) 0.30 (0.03–2.97) 1.24 (0.11–13.85) 6.71 (1.44–31.20) 1.14 (0.19–6.67) 0.52 (0.06–4.73) 0.61 (0.19–1.95) 1.74 (0.25–12.16) 2 12 = 32.27 (p < 0.0013)
a
Weighted n for each individual adversity. Discrete-time survival model with person-year the unit of analysis controlling for age, gender, person-year, psychiatric disorder prior to substance involvement, type of adversity, number of family dysfunction adversities (parental mental illness, parental substance problems, parental criminal behavior, witnessing family violence, physical abuse, sexual abuse and neglect), and number of other adversities (parent died, parental divorce, other parental loss, physical illness and economic adversity). b
opportunities. The association of opportunity with these particular adversities (parental criminal behavior and parental substance problems) may reflect availability in the home via the parents, while the association of opportunity with physical and sexual abuse and witnessing family violence may also suggest parental (or the
perpetrator of the abuse or violence) using of substances during abuse which are available to the adolescent. Additionally this study did not find consistent gender differences for the association between adversity and substance involvement, with the exception of the finding that the cumulative
Table 4 Estimated effects (odds ratios) of number of chronic adversities on risk for alcohol opportunity, alcohol use given the opportunity and alcohol abuse/dependence given use in multivariate survival models. Alcohol opportunityb
Alcohol use given opportunityb
Alcohol abuse/dependence given useb
OR
OR
Number of adversities
na
OR
0 1 2 or 3 4+
969 1008 822 205
1.00 – 1.07 (0.89–1.28) 1.17 (0.97–1.42) 1.10 (0.82–1.48) 2 3 = 2.85 (p < 0.4150)
a
(95% CI)
(95% CI)
1.00 – 1.16 (0.96–1.40) 1.33 (1.06–1.66) 1.55 (1.11–2.17) 2 3 = 12.32 (p < 0.0064)
(95% CI)
1.00 – 0.83 (0.41–1.71) 1.70 (0.91–3.17) 2.61 (1.00–6.84) 2 3 = 6.78 (p < 0.0794)
Weighted n of participants with specified number of adversities. Discrete-time survival model with person-year the unit of analysis estimated with dummy predictors for number of adversities and controlling for age, gender, person-year, and psychiatric disorder prior to substance involvement. b
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
89
Table 5 Estimated effects (odds ratios) of number of chronic adversities on risk for drug opportunity, drug use given the opportunity and drug abuse/dependence given use in multivariate survival models. Drug opportunityb
Drug use given opportunityb
Drug abuse/dependence given useb
OR
OR
Number of adversities
na
OR
0 1 2 or 3 4+
969 1008 822 205
1.00 – 1.10 (0.87–1.39) 1.42 (1.15–1.76) 2.66 (1.78–3.97) 23 = 31.01 (p < 0.0001)
(95% CI)
(95% CI)
1.00 – 3.85 (1.90–7.82) 3.27 (1.45–7.39) 5.05 (2.22–11.50) 23 = 21.81 (p < 0.0001)
(95% CI)
1.00 – 1.14 (0.16–8.02) 1.33 (0.23–7.86) 1.16 (0.18–7.52) 23 = 0.16 (p < 0.9835)
a
Weighted n of participants with specified number of adversities. Discrete-time survival model with person-year the unit of analysis estimated with dummy predictors for number of adversities and controlling for age, gender, person-year, and psychiatric disorder prior to substance involvement. b
number of adversities is related to greater odds of transitioning to drug abuse/dependence among females only. Prior research has also found mixed results regarding gender differences in the impact of adversity on substance involvement, some finding no differences (Greenfield et al., 2002; Lloyd and Turner, 2008) and others finding a greater impact for females (Heffner et al., 2011; Hyman et al., 2008). The inconsistency of these results may be due to differences in the substances, substance involvement stages or populations studied. False positive gender interactions could be due to differences in the prevalence of substance involvement at different stages for males and females as suggested by Lloyd and Turner (2008), whereas real gender differences in the transition to abuse/dependence could be due in part to gender differences in the frequency and types of adversity experienced. For example, females experience greater sexual abuse and research has shown that sexual abuse, in comparison to other stressors, has a particularly high risk for developing psychopathology. However, meta-analyses of studies regarding trauma and posttraumatic stress disorder have shown that sex differences in the frequency and type of adversities only partly account for the greater vulnerability of females for that particular response to trauma (Tolin and Foa, 2006). Biological explanations have also been proposed and recent evidence suggests that gonadal hormones might contribute to a greater sensitivity of females to the rewarding properties of substances (Anker and Carroll, 2011). The implications of these findings should be considered in light of the limitations of the methodology. First, the cross-sectional retrospective design does not allow for conclusions regarding causality. We have assumed, but cannot confirm, that the adversities predate substance involvement; this assumption is made given that most of the adversities considered, if endorsed, do not have discreet onsets and are likely to have been present throughout childhood (for example, economic adversity) and even events which have discreet onsets, like parental divorce, are likely to involve stress that both predate and follow the actual event for years. This assumption, therefore, should temper the interpretation of directionality; a prospective longitudinal design is necessary to address directionality and causality. Additionally, self-reports rely on the willingness of participants to disclose information on unlawful and stigmatizing behavior. Despite assurances of confidentiality, adolescents might be hesitant to disclose such information, which, undoubtedly, leads to an underestimation. Adolescent reports of parental mental illness, substance use problems and criminal behavior are also limited by the reliability with which adolescents can make such reports and are likely to be underestimated. Another factor that probably leads to underestimation of prevalence and association is that this survey only includes adolescents with a fixed residence and does not include homeless or institutionalized adolescents who are likely to have both a greater number of adversities as well as substance involvement.
While we were able to evaluate the association of adversity with alcohol and drug involvement separately, our sample size did not allow for evaluating individual drugs separately. Most drug use in this sample was with marijuana (3.2%) and to a lesser extent tranquilizers and stimulants used without a medical prescription (2%) and cocaine (1.6%; Benjet et al., 2007b). Even evaluating alcohol and drug involvement separately does not allow for teasing apart these substances entirely, as some youth who use alcohol also use drugs and many youth who use drugs also use alcohol. Alcohol is considered a gateway substance for other drugs. For example, in Mexico only 21.5% of those having used cannabis had not already used either alcohol or tobacco before cannabis use (Degenhardt et al., 2010). Further research should address whether the association between adversity and illicit drug involvement can be explained by first use of alcohol or whether adversity increases risk for trying other types of drugs beyond the increased risk for alcohol involvement. Because of the low prevalence of abuse or dependence, the confidence intervals for these models are large, limiting the precision of these estimations and their interpretation. Finally, because many of these adolescents are still experiencing adversity, or little time has passed for the development of a substance disorder, these data may represent only the acute effect of adversity upon substance involvement. Only longitudinal data can address conclusively the persistence of substance involvement or the transition to further stages of substance involvement later in life. Age is likely to play an important role in the transition to abuse or dependence, as is length of time since first use. More complex understanding of the association between adversity and the transition between substance involvement stages would benefit from future research into how age might moderate this association. Despite these limitations, this study has several important novel strengths. First, while most surveys of adolescents are school-based samples, this is a general population study, which includes youth who no longer attend school (18.8% of these youth); a group that has greater substance involvement than school attending youth (Benjet et al., 2012). Secondly, this survey, which was designed to evaluate overall mental health in adolescents, allowed us to control in our models for psychiatric disorders such as mood, anxiety and behavioral disorders, which are highly related to both substance involvement and adversity, and thus possible confounding factors. And finally, while many studies have reported an association of adversity with substance abuse and or dependence (Enoch et al., 2010; Turner and Lloyd, 2003; Lloyd and Turner, 2008; Veijola et al., 2008), this study is novel in evaluating the impact of adversity on the transition to different stages of substance involvement. In conclusion, we addressed the question of whether the association between chronic childhood adversity and substance use disorders found in prior research is due to those with childhood adversity having greater substance use opportunities, being more likely to use given the opportunity, being more likely to develop substance use problems given use, or all of the above. These
90
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
findings suggest that the answer is all of the above. Chronic adversity is associated with both alcohol and drug involvement at all stages including opportunities for use of substances, not just use given the opportunity or the development of abuse or dependence given use. Attending to the needs of youth living in adversity, particularly adversities related to parental dysfunction and child abuse should be a part of strategies to prevent substance involvement and addiction. Role of Funding Source Funding for this study was provided by the National Council on Science and Technology (CONACyT) Grant SSEDF-2003-CO1-22 and Grant CB-2006-01-60688 with complementary funds from Fundación Miguel Alemán AC. Neither funding source had any further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. Contributors Corina Benjet was responsible for obtaining funding, implementation of the survey, and wrote the first draft of the manuscript. Guilherme Borges was responsible for quality control during fieldwork and made a significant intellectual contribution to the study design, interpretation of data, and drafting of the manuscript. María Elena Medina-Mora contributed to the study protocol, design and implementation. Enrique Méndez undertook the statistical analysis. All authors contributed to and have approved the final manuscript. Conflict of Interest No conflict declared. Acknowledgements The survey was carried out in conjunction with the World Health Organization World Mental Health (WMH) Survey Initiative. We thank the WMH staff for assistance with instrumentation and fieldwork. References Anker, J.J., Carroll, 2011. Females are more vulnerable to drug abuse than males: evidence from preclinical studies and the role of ovarian hormones. Curr. Top. Behav. Neurosci. 8, 73–96. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, (DSM-IV), 4th ed. American Psychiatric Association, Washington, DC. Becker, H.C., Lopez, M.F., Doremus-Fitzwater, T.L., 2011. Effects of stress on alcohol drinking: a review of animal studies. Psychopharmacology (Berl.) 218, 131–156. Benjet, C., Borges, G., Medina-Mora, M.E., 2010. Chronic childhood adversity and onset of psychopathology during three life stages: childhood, adolescence and adulthood. J. Psychiatr. Res. 44, 732–740. Benjet, C., Borges, G., Medina-Mora, M.E., Blanco, J., Zambrano, J., Orozco, R., Fleiz, C., Rojas, E., 2007a. Drug use opportunities and the transition to drug use among adolescents in the Mexico Metropolitan Area. Drug Alcohol Depend. 90, 128–134. Benjet, C., Borges, G., Medina-Mora, M.E., Fleiz, C., Blanco, J., Zambrano, J., Rojas, E., Ramirez, M., 2007b. Prevalence and socio-demographic correlates of drug use among adolescents: results from the Mexican Adolescent Mental Health Survey. Addiction 102, 1261–1268. Benjet, C., Borges, G., Medina-Mora, M.E., Zambrano, J., Cruz, C., Méndez, E., 2009b. Descriptive epidemiology of chronic childhood adversity in Mexican adolescents. J. Adolesc. Health 45, 483–489. Benjet, C., Hernandez-Montoya, D., Borges, G., Mendez, E., Medina-Mora, M.E., Aguilar-Gaxiola, S., 2012. Youth who neither study nor work: mental health, education and employment. Salud Publica Mex. 54, 410–417. Benjet, C., Medina-Mora, M.E., Borges, G., Zambrano, J., Aguilar-Gaxiola, S., 2009a. Youth mental health in a populous city of the developing world: results from the Mexican Adolescent Mental Health Survey. J. Child Psychol. Psychiatry 50, 386–395.
Bruffaerts, R., Demyttenaere, K., Borges, G., Haro, J.M., Chiu, W.T., Hwang, I., Karam, E.G., Kessler, R.C., Sampson, N., Alsonso, J., Andrade, L.H., Angermeyer, M., Benjet, C., Bromet, E., de Girolamo, G., de Graaf, R., Florescu, S., Gureje, O., Horiguchi, I., Hu, C., Kovess, V., Levinson, D., Posada-Villa, J., Sagar, R., Scott, K., Tsang, A., Vassilev, S.M., Williams, D.R., Nock, M.K., 2010. Childhood adversities as risk factors for onset and persistence of suicidal behavior. Br. J. Psychiatry 197, 20–27. Chen, C.Y., Dormitzer, C.M., Bejarano, J., Anthony, J.C., 2004. Religiosity and the earliest stages of adolescent drug involvement in seven countries of Latin America. Am. J. Epidemiol. 159, 1180–1188. Courtney, M.E., Piliavin, I., Grogan-Kaylor, A., Nesmith, A., 1998. Foster youth transitions to adulthood: a longitudinal view of youth leaving care. Child Welfare 80, 685–717. Degenhardt, L., Dierker, L., Chiu, W.T., Medina-Mora, M.E., Neumark, Y., Sampson, N., Alonso, J., Angermeyer, M., Anthony, J.C., Bruffaerts, R., de Girolamo, G., de Graaf, R., Gureje, O., Karam, A.N., Kostyuchenko, S., Lee, S., Lépine, J.P., Levinson, D., Nakamura, Y., Posada-Villa, J., Stein, D., Wells, J.E., 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 Alcohol Depend. 108, 84–97. Delva, J., Van Etten, M.L., Gonzalez, G.B., Cedeno, M.A., Penna, M., Caris, L.H., Anthony, J.C., 1999. First opportunities to try drugs and the transition to first drug use: evidence from a national school survey in Panama. Subst. Use Misuse 34, 1451–1467. Efron, B., 1988. Logistic regression, survival analysis and the Kaplan–Meier curve. J. Am. Stat. Assoc. 83, 414–425. Endicott, J., Andreasen, N., Spitzer, R.L., 1978. Family History Research Diagnostic Criteria. New York State Psychiatric Institute, New York, NY. Enoch, M.A., Hodgkinson, C.A., Yuan, Q., Shen, P.H., Goldman, D., Roy, A., 2010. The influence of GABRA2, childhood trauma, and their interaction on alcohol, heroin, and cocaine dependence. Biol. Psychiatry 67, 20–27. Garland, E.L., Pettus-Davis, C., Howard, M.O., 2012. Self-medication among traumatized youth: structure equation modeling of pathways between trauma history, substance misuse, and psychological distress. J. Behav. Med. (Epub ahead of print). Goeders, N.E., Guerin, G.F., 1996. Effects of surgical and pharmacological adrenalectomy on the initiation and maintenance of intravenous cocaine selfadministration in rats. Brain Res. 722, 145–152. Goeders, N.E., 2002. Stress and cocaine addiction. J. Pharmacol. Exp. Ther. 30, 785–789. Greenfield, S.F., Kolodziej, M.E., Sugarman, D.E., Muenz, L.R., Vagge, L.M., He, D.Y., Weiss, R.D., 2002. History of abuse and drinking outcomes following inpatient alcohol treatment: a prospective study. Drug Alcohol Depend. 67, 227–234. Heffner, L.J., Blom, T.J., Anthenelli, R.M., 2011. Gender differences in trauma history and symptoms as predictors of relapse to alcohol and drug use. Am. J. Addict. 20, 307–311. Higley, A.E., Crane, N.A., Spadoni, A.D., Quello, S.B., Goodell, V., Mason, B.J., 2011. Craving in response to stress induction in a human laboratory paradigm predicts treatment outcome in alcohol-dependent individuals. Psychopharmacology (Berl.) 218, 121–129. Hyman, S.M., Paliwal, P., Chaplin, T.M., Mazure, C.M., Rounsaville, B., Sinha, R., 2008. Severity of childhood trauma is predictive of cocaine relapse outcomes in women but not men. Drug Alcohol Depend. 92, 208–216. Kessler, R.C., Avenevoli, S., Green, J., Gruber, M.J., Guyer, M., He, Y., Jin, R., Kaufman, J., Sampson, N.A., Zaslavsky, A.M., 2009. National Comorbidity Survey Replication Adolescent Supplement: III. Concordance of DSM.IV/CIDI diagnoses with clinical reassessments. J Am. Acad. Child Adolesc. Psychiatry 48, 386–399. Kessler, R.C., McLaughlin, K.A., Green, J.G., Gruber, M.J., Sampson, N.A., Zaslavsky, A.M., Aguilar-Gaxiola, S., Alhamzawi, A.O., Alonso, J., Angermeyer, M., Benjet, C., Bromet, E., Chatterji, S., de Girolamo, G., Demyttenaere, K., Fayyad, J., Florescu, S., Gal, G., Gureje, O., Haro, J.M., Hu, C., Karam, E.G., Kawakami, N., Lee, S., Lépine, J.P., Ormel, J., Posada-Villa, J., Sagar, R., Tsang, A., Üstün, B., Vassilev, S., Viana, M.C., Williams, D.R., 2010. Childhood adversities and adult psychopathology in the WHO World Mental Health Surveys. Br. J. Psychiatry 197, 378–385. Khantzian, E.J., 1985. The self-medication hypothesis of addictive disorders: focus on heroin and cocaine dependence. Am. J. Psychiatry 142, 1259–1264. Knauper, B., Cannell, C.F., Schwarz, N., Bruce, M.L., Kessler, R.C., 1999. Improving the accuracy of major depression age of onset reports in the US National Comorbidity Survey. Int. J. Methods Psychiatr. Res. 8, 39–48. Koob, G., Kreek, M.J., 2007. Stress, dysregulation of drug reward pathways and the transition to drug dependence. Am. J. Psychiatry 164, 1149–1159. Kreek, M.J., Nielson, D.A., Butelman, E.R., LaForge, K.S., 2005. Genetic influences on impulsivity, risk taking, stress responsivity and vulnerability to drug abuse and addiction. Nat. Neurosci. 8, 1450–1457. Liang, W., Lenton, S., Allsop, S., Chikritzhs, R., 2011. Does availability of illicit drugs mediate the association between mental illness and substance use? Subst. Use Misuse 46, 1304–1308. Lloyd, D.A., Turner, R.J., 2008. Cumulative life-time adversities and alcohol dependence in adolescence and young adulthood. Drug Alcohol Depend. 93, 217–226. Lopez, M.F., Doremus-Fitzwater, T.L., Becker, H.C., 2011. Chronic social isolation and chronic variable stress during early development induce later elevated ethanol intake in adult C57BL/6J mice. Alcohol 45, 355–364. Lynch, W.J., Kushner, M.G., Rawleigh, J.M., Fisdon, J., Caroll, M.E., 1999. The effects of restraint stress on voluntary ethanol consumption in rats. Exp. Clin. Psychopharmacol. 7, 318–323. McEwen, B.S., 2000. Allostatis and allostatic load: implications for neuropsychopharmacology. Neuropsychopharmacology 22, 1083–1124.
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91 Merikangas, K.R., Avenevolli, S., Costello, E.J., Koretz, D., Kessler, R.C., 2009. National Comorbidity Survey Replication adolescent supplement: I. Background and measures. J. Am. Acad. Child Adolesc. Psychiatry 48, 367–379. Research Triangle Institute, 2002. Sudaan Release 8.0.1. Research Triangle Institute, Research Triangle Park, NC. Robinson, J., Sareen, J., Cox, B.J., Bolton, J.M., 2011. Role of self-medication in the development of comorbid anxiety and substance use disorders: a longitudinal investigation. Arch. Gen. Psychiatry 68, 800–807. Sinha, R., Fox, H.C., Hong, K.I., Hansen, J., Tuit, K., Kreek, M.J., 2011. Effects of adrenal sensitivity, stress- and cue-induced craving, and anxiety on subsequent alcohol relapse and treatment outcomes. Arch. Gen. Psychiatry 68, 942–952. Sinha, R., Garcia, M., Paliwal, P., Kreek, M.J., Rounsaville, B.J., 2006. Stress-induced cocaine craving and hypothalamic-pituitary-adrenal responses are predictive of cocaine relapse outcomes. Arch. Gen. Psychiatry 63, 324–331. Storr, C.L., Wagner, F.A., Chen, C.Y., Anthony, J.C., 2011. Childhood predictors of first chance to use and use of cannabis by young adulthood. Drug Alcohol Depend. 117, 7–15. Strauss, M.A., 1979. Measuring intrafamily conflict and violence: the conflict tactics (CT) scales. J. Marriage Fam. 41, 75–88. Tolin, D.F., Foa, E.B., 2006. Sex differences in trauma and posttraumatic stress disorder: a quantitative review of 25 years of research. Psychol. Bull. 132, 959–992.
91
Turner, R.J., Lloyd, D.A., 2003. Cumulative adversity and drug dependence in young adults: racial/ethnic contrasts. Addiction 98, 305–315. Uhart, M., Wand, G.S., 2009. Stress, alcohol and drug interaction: an update of human research. Addict. Biol. 14, 43–64. Van Etten, M.L., Anthony, J.C., 1999. Comparative epidemiology of initial drug opportunities and transitions to first use: marijuana, cocaine, hallucinogens, and heroin. Drug Alcohol Depend. 54, 117–125. Van Etten, M.L., Neumark, Y.D., Anthony, J.C., 1999. Male–female differences in the earliest stages of drug involvement. Addiction 94, 1413–1419. Veijola, J., Laara, E., Joukamaa, M., Isohanni, M., Hakko, H., Haapea, M., Pirkola, S., Maki, P., 2008. Temporary parental separation at birth and substance use disorder in adulthood. A long-term follow-up of the Finnish Christmas Seal Home Children. Soc. Psychiatry Psychiatr. Epidemiol. 43, 11–17. Vengeliene, V., Siegmund, S., Singer, M.V., Sinclair, J.D., Li, T.K., Spanagel, R., 2003. A comparative study on alcohol-preferring rat lines: effects of deprivation and stress phases on voluntary alcohol intake. Alcohol. Clin. Exp. Res. 27, 1048–1054. Willett, J.B., Singer, J.D., 1993. Investigating onset, cessation, relapse, and recovery: why you should, and how you can, use discrete-time survival analysis to examine event occurrence. J. Consult. Clin. Psychol. 61, 952–965.
Contents lists available at SciVerse ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Chronic childhood adversity and stages of substance use involvement in adolescents Corina Benjet a,∗ , Guilherme Borges a , María Elena Medina-Mora a , Enrique Méndez a,b a Department of Epidemiological and Psychosocial Research, National Institute of Psychiatry Ramón de la Fuente, Calzada México Xochimilco 101, Colonia San Lorenzo Huipulco, Mexico City 14370, Mexico b Institute for Applied Mathematical and Systems Research, National Autonomous University of Mexico, Ciudad Universitaria, Mexico City 04510, Mexico
a r t i c l e
i n f o
Article history: Received 18 September 2012 Received in revised form 30 November 2012 Accepted 2 December 2012 Available online 29 December 2012 Keywords: Alcohol Drugs Substance use Substance abuse Adversity Adolescence
a b s t r a c t Background: Studies have shown that those who experience chronic childhood adversity have a greater likelihood of substance abuse and dependence. However, substance use disorders are first preceded by substance use, and substance use is preceded by substance use opportunities. This study aims to estimate the association of chronic adversity with different stages of substance involvement: opportunities, use given the opportunity and abuse or dependence given use. Methods: 3005 adolescents aged 12–17 were interviewed in a stratified multistage general population probability survey of Mexico City, Mexico. Substance involvement and chronic childhood adversities were assessed with the World Mental Health Composite International Diagnostic Interview Adolescent Version (WMH-CIDI-A). Discrete-time survival models were performed; their survival coefficients and standard errors were exponentiated, and reported as odds-ratios (ORs). Results: Childhood adversities were associated with alcohol opportunity, alcohol use and alcohol abuse/dependence with significant ORs for individual adversities ranging from 1.4 to 4.1. Childhood adversities were also associated with illicit drug opportunity, drug use and drug abuse/dependence with significant ORs for individual adversities ranging from 1.6 to 17.3. Having more adversities was associated with greater incremental odds of substance involvement, particularly drug use given the opportunity. Conclusions: While adversities are mostly related to transitioning into use and disorder, a few are related to substance opportunities, particularly those which were likely to make substances available through parents. Attending to the needs of youth living in adversity, particularly adversities related to parental dysfunction and child abuse should be integral to addiction prevention efforts. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Epidemiological studies have shown that people who experience chronic childhood adversity have a greater likelihood of substance abuse and dependence as well as other psychiatric disorders (Benjet et al., 2010; Enoch et al., 2010; Kessler et al., 2010; Turner and Lloyd, 2003; Lloyd and Turner, 2008). Likewise, animal models have shown stress and early adversity to increase self-administration of substances (Becker et al., 2011; Goeders and Guerin, 1996; Lopez et al., 2011; Lynch et al., 1999; Goeders, 2002; Vengeliene et al., 2003; Koob and Kreek, 2007) while human models have shown that stress is related to substance abuse and dependence and to relapse and craving amongst
∗ Corresponding author at: National Institute of Psychiatry Ramón de la Fuente, Calzada México Xochimilco 101, Colonia San Lorenzo Huipulco, Mexico City 14370, Mexico. Tel.: +52 55 4160 5332; fax: +52 55 5513 3446. E-mail address: [email protected] (C. Benjet). 0376-8716/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.drugalcdep.2012.12.002
substance abusers (Higley et al., 2011; Sinha et al., 2011, 2006). Many theories have been proposed to explain this association. Among the most frequently mentioned in the addiction literature is the self-medication hypothesis, in which those who abuse substances do so to ease or relieve the pain of trauma, negative affect and psychiatric disorders (Garland et al., 2012; Khantzian, 1985; Robinson et al., 2011). Compatible with the self -medication hypothesis are theories which propose neurobiological pathways that alter learning, reward, craving, and impulsivity through stress allostasis. The theory of allostasis or allostatic load posits that the process of achieving stability through alteration of neural, neuroendocrine and immune mechanisms (allostasis), while adaptive in the short run, becomes “overloaded” (allostatic load) with chronic stress (McEwen, 2000). Frequent, multiple stressors, failure to habituate to chronic stressors, compensatory hyperactivity and delayed shutdown of the stress response alter key neurocircuitry that impacts upon craving, loss of control and compulsion and, thus, vulnerability to addiction (McEwen, 2000; Uhart and Wand, 2009).
86
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
The types of chronic childhood adversities typically included in epidemiological studies fall into one of four categories: family or parental dysfunction, abuse/neglect, interpersonal loss and socioeconomic disadvantage. Certain epidemiological questions regarding adversity and stage of drug involvement have yet to be addressed and may contribute to further understanding of this relationship. Substance use disorders are first preceded by substance use and substance use is preceded by substance use opportunities. Substance opportunities are generally considered a measure of availability of substances in a given context, although they may also reflect, to some degree, actively seeking out opportunities. Prior work has found that certain factors related to substance use and abuse are actually related to exposure to substance opportunities (Benjet et al., 2007a; Chen et al., 2004; Storr et al., 2011). For example, several studies have shown that males are more likely to have opportunities to use drugs than females, but females are equally as likely as males to try drugs when given the chance (Benjet et al., 2007a; Delva et al., 1999; Van Etten et al., 1999; Van Etten and Anthony, 1999). Another study found drug use opportunities to mediate the association between mental illness and drug use (Liang et al., 2011). An important question to address is whether the association between chronic childhood adversity and substance use disorders is due to those with childhood adversity having greater substance use opportunities, being more likely to use given the opportunity, being more likely to develop substance use problems given use, or all of the above. Given the reported associations of psychiatric disorder with both childhood adversity and substance involvement, it is also important to determine whether the association between adversity and substance involvement holds when controlling for psychiatric disorders. This study aims to address these questions. 2. Methods 2.1. Participants This report presents secondary analyses of data from the Mexican Adolescent Mental Health survey (Benjet et al., 2009a). The sample consists of 3005 adolescents (52.1% female) aged 12–17, selected from a stratified multistage area probability sample representative of the nearly two million adolescents residing in the Mexico City Metropolitan Area. In all strata, the primary sampling units were census count areas cartographically defined and updated for the XII Population and Housing Census in 2000. Secondary sampling units were city blocks (or groups of them), selected with probability proportional to size. All households within these city block units with adolescents in the age range were selected. One eligible member from each of these households was randomly selected using the Kish method of random number charts. The response rate of eligible respondents was 71%. 2.2. Procedures A verbal and written explanation of the study was given to both parents and adolescents, after which signed informed consent from a parent or legal guardian was obtained, as well as the assent of the adolescent. Interviews were conducted in the homes of the selected participants. All study participants and their families were offered information on local mental health services. The Internal Review Board of the National Institute of Psychiatry approved the recruitment, consent and field procedures. 2.3. Measures Substance involvement, adversity and psychiatric disorders were evaluated with the fully structured, computer assisted, World Mental Health adolescent version of the Composite International Diagnostic Interview (WMH-CIDI-A) the development and validity of which is described elsewhere (Kessler et al., 2009; Merikangas et al., 2009). The computer assisted version was administered, in which extensively trained lay interviewers read the questions to the participant directly from the computer screen, the questions are chosen by the computer based on previous responses of the participant and complex logical skip patterns. The interviewer inputs the respondent’s answers directly into the computer and consistency checks are programmed so that inconsistent information is probed and corrected. Diagnostic classification is based on meeting the diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; American Psychiatric Association, 1994). The substance use section includes questions
regarding the lifetime use (defined as consumption of the substance at least once at any time in one’s life) of alcohol and illicit drugs, opportunity to use alcohol and drugs, and substance abuse and dependence. The illicit drugs included were marijuana, cocaine in any of its presentations, tranquilizers or stimulants used without a medical prescription, such as methamphetamine, and other substances (e.g., heroin, inhalants, LSD, etc.) which were grouped as “other drugs”. Participants were asked about each category of drugs openly and then presented with a list of numerous different street names for these drugs. The questions regarding opportunity to use alcohol and drugs were posed after the questions about alcohol and drug use so that opportunities to use alcohol and drugs referred to all the alcohol and drugs previously presented to the respondents. All participants were asked about opportunities regardless of whether they had previously endorsed using any substance. Opportunity to use alcohol and opportunity to use illicit drugs were asked about separately and defined as having the opportunity to use any substance, independently of whether or not the respondent did so. For example, someone offered the respondent drugs or the respondent was present when others were consuming and could have done so if he or she chose to. In order to cover a range of chronic childhood adversities and those for which previous research has found associations with substance abuse, we included different forms of parental loss, parental psychopathology, abuse, physical illness and economic adversity; these were evaluated from the childhood and post-traumatic stress disorder sections of the WMH-CIDI-A. Each of 12 chronic adversities was classified as present or not present using the same criteria as the World Mental Health Survey Initiative (Kessler et al., 2010). Physical abuse and parental violence were assessed with a modified version of the Conflict Tactics Scale (Strauss, 1979). Neglect was evaluated with questions often used in child welfare studies (Courtney et al., 1998). Sexual abuse was assessed by reading a definition of rape and then asking about other forms of abuse or molestation. In order to be consistent with the other World Mental Health Surveys; chronic sexual abuse was defined as reporting at least three episodes of sexual abuse thus representing chronic sexual abuse as opposed to acute one time trauma (Bruffaerts et al., 2010). To assess parental loss, the adolescents were asked whether they lived with both parents all of their lives. Those who did not were asked whether this was because their parents had separated or divorced, a parent had died or some other reason. Those mentioning separations of six months or more from either parent for some other reason were classified as other parental loss, with reasons ranging from having gone to boarding school, having left home, or that their parent was in prison. Parental pathology was evaluated using questions from the Family History Research Diagnostic Criteria Interview and included parental mental illness, substance problems, and criminal behavior (Endicott et al., 1978). Participants were considered to have experienced family economic adversity if the family ever received money from a government assistance program for poor families or by lack of parental employment most or all of the time during the participant’s childhood. Serious physical illness is based on the adolescent’s report of having experienced a life-threatening physical illness. Mood, anxiety and behavioral disorders are associated with childhood adversity and with substance involvement. In order to estimate the association of adversity with substance use and substance use disorder without the possible confounding of psychiatric disorders, the following psychiatric disorders were also evaluated with the WMH-CIDI according to DSM-IV criteria and were used as a covariate in the statistical models described below: major depression, dysthmia, bipolar I and II, generalized anxiety disorder, panic disorder, agoraphobia, specific phobia, social phobia, post-traumatic stress disorder, separation anxiety, eating disorders, attention deficit hyperactivity disorder, oppositional-defiant disorder, conduct disorder, and intermittent explosive disorder. The WMH-CIDI assessed age-of-onset of substance opportunity, use and disorder as well as for other psychiatric disorders retrospectively using a series of questions shown to improve accuracy of retrospect reports and avoid implausible response patterns (such as use of anchoring events and quality control programs to search for response inconsistencies; Knauper et al., 1999). 2.4. Statistical analysis Data were weighted to adjust for differential probabilities of selection and nonresponse as well as post-stratification to the total Mexico City Metropolitan Area adolescent population according to the year 2000 Census in the target age and sex range. The socio-demographic distribution of the sample closely approximates the target population, in that roughly half are female, there is an even distribution of ages, two-thirds live with both parents, and the socio-economic level of the parents represents the educational and income levels in Mexico; for example, one-fourth of parents have 6 years or less of formal education whereas only 13% have gone to college. More details of the weighted and un-weighted socio-demographic distribution of the sample is presented elsewhere (Benjet et al., 2007a, 2009a) as is the distribution of chronic adversity (Benjet et al., 2009b). First we present the prevalence estimates for each substance involvement stage. Then multivariate associations of the childhood adversities with substance opportunity, use and abuse/dependence were estimated using discrete-time survival analysis with person-years as the unit of analysis (Efron, 1988; Willett and Singer, 1993). Each model controlled for the adolescent’s age at interview, gender, psychiatric disorders prior to substance involvement and person-year. Psychiatric disorder and substance involvement are time varying variables whereas
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91 Table 1 Lifetime prevalence of substance involvement.
Alcohol opportunity Alcohol use amongst those w/opportunity Alcohol abuse/dependence amongst those w/use Drug opportunity Drug use amongst those w/opportunity Drug abuse/dependence amongst those w/use a
n (unweighted)
% (weighted)
SEa
2361 1693
79.9 73.4
0.6 0.9
88
6.4
0.8
784 134
28.7 18.1
0.5 1.1
41
32.1
5.1
Standard error.
adversities are assumed to predate substance involvement. The models for substance use represent the association of adversity with substance use only among those with opportunities for use. The models for substance abuse or dependence represent the association of adversity with substance abuse/dependence only among those with substance use. We initially tested the interaction of each individual adversity with gender on alcohol and drug stages, but only 7 of the 72 gender-byadversity interactions were significant at the p < 0.05 level and were inconsistent with regards to which gender had greater odds. Therefore the final models, while controlling for gender did not include adversity-by-gender interaction terms. The final models also included each of the 12 predictors for type of adversity and additional predictors for number of family dysfunction adversities (i.e., parental mental illness, parental substance problems, parental criminal behavior, witnessing family violence, physical abuse, sexual abuse and neglect) and number of other adversities (parental death, parental divorce, other parental loss, physical illness and economic adversity) so that the odds ratio for each adversity may be considered an estimation of the “pure effect” of that adversity controlling for other adversities and number of adversities (Kessler et al., 2010). To estimate the effects of number of adversities, the models used a series of dummy predictor variables for number of adversities (e.g., one such variable for respondents who experienced exactly one adversity, another for respondents who experienced two or three adversities and another for those who experienced four or more adversities) without information about the types of adversities experienced. Additional models with gender by number-of-adversity terms were also performed, but because only one model had significant interactions these are not presented on the tables. The survival coefficients and their standard errors were exponentiated and are reported in the form of odds-ratios (OR) and 95% confidence intervals (95% CI). As a result of this complex sample design and weighting, estimates of standard errors for proportions were obtained by the Taylor series linearization method using the SUDAAN Software (Research Triangle Institute, 2002).
3. Results The prevalence of lifetime substance involvement in this representative sample of Mexican adolescents is provided in Table 1. Most adolescents (79.9%) have had alcohol opportunities, of those, a majority (73.4%) have used alcohol and few (6.4%) of those who have used, have developed alcohol abuse or dependence. With regards to illicit drugs, less than a third have had drug use opportunities. Of those with drug opportunities, 18.1% have tried drugs and of those, 32.1% have developed abuse or dependence. Table 2 presents the multivariate models for the association of childhood adversity and alcohol opportunities, alcohol use given the opportunity and alcohol abuse/dependence given use. When controlling for other adversities, number of adversities, and psychiatric disorders, parental criminal behavior, family violence and sexual abuse show significantly elevated odds (OR = 1.5, 95% CI = 1.1–1.9; OR = 1.6, 95% CI = 1.2–1.9 and OR = 1.9, 95% CI = 1.1–3.1, respectively) for alcohol opportunities. Given the opportunity to use, parent criminal behavior and sexual abuse have 1.4 (95% CI = 1.1–1.9) and 2.2 (95% CI = 1.5–3.2) times the odds of alcohol use, respectively. Given alcohol use, parental criminal behavior has four times the odds (95% CI = 1.9–9.2) and other parental loss twice the odds (95% CI = 1.2–3.9) of transitioning to alcohol abuse or dependence. Multivariate models for the association of childhood adversity and drug opportunities, use given the opportunity and drug abuse
87
or dependence are shown in Table 3. Parental substance problems, parental criminal behavior, and physical abuse all have significantly greater odds of drug opportunities (OR = 1.6, 95% CI = 1.0–2.5; OR = 1.7, 95% CI = 1.0–2.7 and OR = 1.6, 95% CI = 1.2–2.1, respectively). For drug use given the opportunity, these same adversities, plus family violence and neglect, are associated with larger odds ratios from 2.1 (95% CI = 1.0–4.2) for physical abuse to 3.5 (95% CI = 1.1–10.7) for neglect. Amongst drug users, parental criminal behavior, physical abuse and parental death are associated with elevated odds of transitioning to drug abuse or dependence (OR = 17.3, 95% CI = 2.7–110.8; OR = 9.1, 95% CI = 1.2–66.8 and OR = 6.7, 95% CI = 1.4–31.2, respectively). Table 4 shows the association of number of adversities with alcohol opportunity, alcohol use given the opportunity and alcohol abuse/dependence given use. A greater number of adversities is associated with increasingly greater odds of alcohol use given the opportunity, but is not associated with other stages of alcohol involvement. Table 5 shows incremental odds of drug opportunity and transition to drug use with increasing number of adversities. Odds ratios for drug opportunity range from 1.4 (95% CI = 1.2–1.8) for experiencing two or three adversities to 2.7 (95% CI = 1.8–4.0) for reporting four or more adversities. Odds ratios for drug use given the opportunity range from 3.9 for one adversity (95% CI = 1.9–7.8) to 5.1 (95% CI = 2.2–11.5) for four or more adversities. Number of adversities is not associated to drug abuse/dependence overall, but there is a significant interaction with sex (data not shown but available upon request), such that the number of adversities is associated with the transition to drug abuse/dependence for females only (p < 0.001).
4. Discussion In the Mexico City Metropolitan Area, alcohol use among adolescents is normative and accepted, as evidenced by the majority of adolescents having had exposure to alcohol opportunity and having used alcohol given the opportunity; drugs are less available and drug use is less tolerated than alcohol, as suggested by the fact that less than a third have had drug opportunities and less than a fifth of those with the opportunity have tried drugs. All models estimating the association of adversity with substance involvement stages were significant, however, not all individual adversities were independently associated to substance involvement. One particular adversity, parental criminal behavior, stands out for its consistent association in all models, from a 40% increased odds of alcohol use given the opportunity, to 17 times the odds for drug abuse or dependence given use. One explanation might be that this association reflects a genetic predisposition, such as risk taking, impulsivity, novelty seeking, or antisocial personality (Kreek et al., 2005), but cannot be explained by a genetic predisposition toward addiction as these models control for parental substance problems. Another explanation might be that parental criminal behavior involves a greater degree of chronic stress. It may also be that those willing to disclose parental criminal behavior are also more willing to disclose their own substance involvement. In general, family dysfunction adversities, such as witnessing family violence, physical and sexual abuse and neglect, have odds ratios above one, though not all are statistically significant in all models. Adversities were not only related to substance use and the development of abuse or dependence, but were also associated to substance opportunities: parental criminal behavior, witnessing family violence and sexual abuse, in the case of alcohol opportunities, and parental criminal behavior, parental substance problems and physical abuse, in the case of drug opportunities. Opportunity is usually considered an indirect measure of availability in a given context, but may also reflect, to some extent, those who search for
88
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
Table 2 Estimated effects (odds ratios) of chronic adversity on risk for alcohol opportunity, alcohol use given the opportunity and alcohol abuse/dependence given use in multivariate survival models. Alcohol opportunity
Alcohol use given opportunity
Alcohol abuse/dependence given use
Multivariate modelb
Multivariate modelb
Multivariate modelb
n
OR
OR
OR
352 162 188 582 413 50 173 169 413 413 205 738
1.01 (0.77–1.32) 1.09 (0.75–1.59) 1.46 (1.12–1.90) 1.55 (1.24–1.93) 0.99 (0.69–1.42) 1.85 (1.12–3.05) 1.01 (0.71–1.46) 1.06 (0.82–1.38) 0.97 (0.77–1.22) 0.98 (0.77–1.24) 1.38 (0.99–1.91) 1.07 (0.92–1.25) 2 12 = 44.18 (p < 0.0001)
a
Parental mental illness Parental substance problems Parental criminal behavior Witnessing family violence Physical abuse Sexual abuse Neglect Parent died Parent divorce Other parent loss Physical illness Economic adversity
(95% CI)
(95% CI)
1.25 (0.97–1.61) 1.32 (0.96–1.82) 1.37 (1.02–1.85) 1.24 (0.91–1.68) 1.26 (0.90–1.75) 2.21 (1.52–3.23) 1.14 (0.73–1.77) 0.96 (0.62–1.51) 1.08 (0.82–1.41) 1.14 (0.85–1.53) 0.93 (0.58–1.49) 0.92 (0.74–1.15) 2 12 = 35.84 (p < 0.0003)
(95% CI)
1.37 (0.58–3.24) 0.86 (0.30–2.48) 4.14 (1.86–9.21) 1.84 (0.83–4.09) 1.35 (0.72–2.53) 0.42 (0.13–1.34) 1.24 (0.48–3.25) 0.21 (0.04–1.18) 0.68 (0.20–2.30) 2.16 (1.18–3.94) 1.21 (0.38–3.87) 1.54 (0.85–2.78) 2 12 = 127.22 (p < 0.0001)
a
Weighted n for each individual adversity. Discrete-time survival model with person-year the unit of analysis controlling for age, gender, person-year, psychiatric diagnosis prior to substance involvement, type of adversity, number of family dysfunction adversities (parental mental illness, parental substance problems, parental criminal behavior, witnessing family violence, physical abuse, sexual abuse and neglect), and number of other adversities (parent died, parental divorce, other parental loss, physical illness and economic adversity). b
Table 3 Estimated effects (odds ratios) of chronic adversity on risk for drug opportunity, drug use given the opportunity and drug abuse/dependence given use in multivariate survival models. Drug opportunity
Drug use given the opportunity
Drug abuse/dependence given use
Multivariate modelb
Multivariate modelb
Multivariate modelb
n
OR
OR
OR
352 162 188 582 413 50 173 169 413 413 205 738
1.14 (0.77–1.68) 1.62 (1.04–2.52) 1.65 (1.01–2.71) 1.19 (0.84–1.69) 1.56 (1.17–2.09) 1.60 (0.88–2.90) 1.19 (0.78–1.83) 1.01 (0.72–1.41) 1.10 (0.82–1.47) 1.34 (0.98–1.84) 1.19 (0.85–1.67) 0.95 (0.72–1.24) 2 12 = 29.80 (p < 0.0030)
a
Parental mental illness Parental substance problems Parental criminal behavior Witnessing family violence Physical abuse Sexual abuse Neglect Parent died Parent divorce Other parent loss Physical illness Economic adversity
(95% CI)
(95% CI)
1.50 (0.74–3.01) 2.36 (1.02–5.42) 2.61 (1.26–5.43) 3.11 (1.51–6.42) 2.07 (1.02–4.23) 1.68 (0.52–5.42) 3.45 (1.11–10.74) 1.64 (0.41–6.59) 1.54 (0.76–3.11) 1.35 (0.67–2.72) 0.64 (0.26–1.59) 1.44 (0.67–3.10) 2 12 = 38.87 (p < 0.0001)
(95% CI)
6.94 (0.76–62.94) 0.75 (0.10–5.44) 17.33 (2.71–110.81) 0.62 (0.12–3.21) 9.06 (1.23–66.82) 0.30 (0.03–2.97) 1.24 (0.11–13.85) 6.71 (1.44–31.20) 1.14 (0.19–6.67) 0.52 (0.06–4.73) 0.61 (0.19–1.95) 1.74 (0.25–12.16) 2 12 = 32.27 (p < 0.0013)
a
Weighted n for each individual adversity. Discrete-time survival model with person-year the unit of analysis controlling for age, gender, person-year, psychiatric disorder prior to substance involvement, type of adversity, number of family dysfunction adversities (parental mental illness, parental substance problems, parental criminal behavior, witnessing family violence, physical abuse, sexual abuse and neglect), and number of other adversities (parent died, parental divorce, other parental loss, physical illness and economic adversity). b
opportunities. The association of opportunity with these particular adversities (parental criminal behavior and parental substance problems) may reflect availability in the home via the parents, while the association of opportunity with physical and sexual abuse and witnessing family violence may also suggest parental (or the
perpetrator of the abuse or violence) using of substances during abuse which are available to the adolescent. Additionally this study did not find consistent gender differences for the association between adversity and substance involvement, with the exception of the finding that the cumulative
Table 4 Estimated effects (odds ratios) of number of chronic adversities on risk for alcohol opportunity, alcohol use given the opportunity and alcohol abuse/dependence given use in multivariate survival models. Alcohol opportunityb
Alcohol use given opportunityb
Alcohol abuse/dependence given useb
OR
OR
Number of adversities
na
OR
0 1 2 or 3 4+
969 1008 822 205
1.00 – 1.07 (0.89–1.28) 1.17 (0.97–1.42) 1.10 (0.82–1.48) 2 3 = 2.85 (p < 0.4150)
a
(95% CI)
(95% CI)
1.00 – 1.16 (0.96–1.40) 1.33 (1.06–1.66) 1.55 (1.11–2.17) 2 3 = 12.32 (p < 0.0064)
(95% CI)
1.00 – 0.83 (0.41–1.71) 1.70 (0.91–3.17) 2.61 (1.00–6.84) 2 3 = 6.78 (p < 0.0794)
Weighted n of participants with specified number of adversities. Discrete-time survival model with person-year the unit of analysis estimated with dummy predictors for number of adversities and controlling for age, gender, person-year, and psychiatric disorder prior to substance involvement. b
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
89
Table 5 Estimated effects (odds ratios) of number of chronic adversities on risk for drug opportunity, drug use given the opportunity and drug abuse/dependence given use in multivariate survival models. Drug opportunityb
Drug use given opportunityb
Drug abuse/dependence given useb
OR
OR
Number of adversities
na
OR
0 1 2 or 3 4+
969 1008 822 205
1.00 – 1.10 (0.87–1.39) 1.42 (1.15–1.76) 2.66 (1.78–3.97) 23 = 31.01 (p < 0.0001)
(95% CI)
(95% CI)
1.00 – 3.85 (1.90–7.82) 3.27 (1.45–7.39) 5.05 (2.22–11.50) 23 = 21.81 (p < 0.0001)
(95% CI)
1.00 – 1.14 (0.16–8.02) 1.33 (0.23–7.86) 1.16 (0.18–7.52) 23 = 0.16 (p < 0.9835)
a
Weighted n of participants with specified number of adversities. Discrete-time survival model with person-year the unit of analysis estimated with dummy predictors for number of adversities and controlling for age, gender, person-year, and psychiatric disorder prior to substance involvement. b
number of adversities is related to greater odds of transitioning to drug abuse/dependence among females only. Prior research has also found mixed results regarding gender differences in the impact of adversity on substance involvement, some finding no differences (Greenfield et al., 2002; Lloyd and Turner, 2008) and others finding a greater impact for females (Heffner et al., 2011; Hyman et al., 2008). The inconsistency of these results may be due to differences in the substances, substance involvement stages or populations studied. False positive gender interactions could be due to differences in the prevalence of substance involvement at different stages for males and females as suggested by Lloyd and Turner (2008), whereas real gender differences in the transition to abuse/dependence could be due in part to gender differences in the frequency and types of adversity experienced. For example, females experience greater sexual abuse and research has shown that sexual abuse, in comparison to other stressors, has a particularly high risk for developing psychopathology. However, meta-analyses of studies regarding trauma and posttraumatic stress disorder have shown that sex differences in the frequency and type of adversities only partly account for the greater vulnerability of females for that particular response to trauma (Tolin and Foa, 2006). Biological explanations have also been proposed and recent evidence suggests that gonadal hormones might contribute to a greater sensitivity of females to the rewarding properties of substances (Anker and Carroll, 2011). The implications of these findings should be considered in light of the limitations of the methodology. First, the cross-sectional retrospective design does not allow for conclusions regarding causality. We have assumed, but cannot confirm, that the adversities predate substance involvement; this assumption is made given that most of the adversities considered, if endorsed, do not have discreet onsets and are likely to have been present throughout childhood (for example, economic adversity) and even events which have discreet onsets, like parental divorce, are likely to involve stress that both predate and follow the actual event for years. This assumption, therefore, should temper the interpretation of directionality; a prospective longitudinal design is necessary to address directionality and causality. Additionally, self-reports rely on the willingness of participants to disclose information on unlawful and stigmatizing behavior. Despite assurances of confidentiality, adolescents might be hesitant to disclose such information, which, undoubtedly, leads to an underestimation. Adolescent reports of parental mental illness, substance use problems and criminal behavior are also limited by the reliability with which adolescents can make such reports and are likely to be underestimated. Another factor that probably leads to underestimation of prevalence and association is that this survey only includes adolescents with a fixed residence and does not include homeless or institutionalized adolescents who are likely to have both a greater number of adversities as well as substance involvement.
While we were able to evaluate the association of adversity with alcohol and drug involvement separately, our sample size did not allow for evaluating individual drugs separately. Most drug use in this sample was with marijuana (3.2%) and to a lesser extent tranquilizers and stimulants used without a medical prescription (2%) and cocaine (1.6%; Benjet et al., 2007b). Even evaluating alcohol and drug involvement separately does not allow for teasing apart these substances entirely, as some youth who use alcohol also use drugs and many youth who use drugs also use alcohol. Alcohol is considered a gateway substance for other drugs. For example, in Mexico only 21.5% of those having used cannabis had not already used either alcohol or tobacco before cannabis use (Degenhardt et al., 2010). Further research should address whether the association between adversity and illicit drug involvement can be explained by first use of alcohol or whether adversity increases risk for trying other types of drugs beyond the increased risk for alcohol involvement. Because of the low prevalence of abuse or dependence, the confidence intervals for these models are large, limiting the precision of these estimations and their interpretation. Finally, because many of these adolescents are still experiencing adversity, or little time has passed for the development of a substance disorder, these data may represent only the acute effect of adversity upon substance involvement. Only longitudinal data can address conclusively the persistence of substance involvement or the transition to further stages of substance involvement later in life. Age is likely to play an important role in the transition to abuse or dependence, as is length of time since first use. More complex understanding of the association between adversity and the transition between substance involvement stages would benefit from future research into how age might moderate this association. Despite these limitations, this study has several important novel strengths. First, while most surveys of adolescents are school-based samples, this is a general population study, which includes youth who no longer attend school (18.8% of these youth); a group that has greater substance involvement than school attending youth (Benjet et al., 2012). Secondly, this survey, which was designed to evaluate overall mental health in adolescents, allowed us to control in our models for psychiatric disorders such as mood, anxiety and behavioral disorders, which are highly related to both substance involvement and adversity, and thus possible confounding factors. And finally, while many studies have reported an association of adversity with substance abuse and or dependence (Enoch et al., 2010; Turner and Lloyd, 2003; Lloyd and Turner, 2008; Veijola et al., 2008), this study is novel in evaluating the impact of adversity on the transition to different stages of substance involvement. In conclusion, we addressed the question of whether the association between chronic childhood adversity and substance use disorders found in prior research is due to those with childhood adversity having greater substance use opportunities, being more likely to use given the opportunity, being more likely to develop substance use problems given use, or all of the above. These
90
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91
findings suggest that the answer is all of the above. Chronic adversity is associated with both alcohol and drug involvement at all stages including opportunities for use of substances, not just use given the opportunity or the development of abuse or dependence given use. Attending to the needs of youth living in adversity, particularly adversities related to parental dysfunction and child abuse should be a part of strategies to prevent substance involvement and addiction. Role of Funding Source Funding for this study was provided by the National Council on Science and Technology (CONACyT) Grant SSEDF-2003-CO1-22 and Grant CB-2006-01-60688 with complementary funds from Fundación Miguel Alemán AC. Neither funding source had any further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. Contributors Corina Benjet was responsible for obtaining funding, implementation of the survey, and wrote the first draft of the manuscript. Guilherme Borges was responsible for quality control during fieldwork and made a significant intellectual contribution to the study design, interpretation of data, and drafting of the manuscript. María Elena Medina-Mora contributed to the study protocol, design and implementation. Enrique Méndez undertook the statistical analysis. All authors contributed to and have approved the final manuscript. Conflict of Interest No conflict declared. Acknowledgements The survey was carried out in conjunction with the World Health Organization World Mental Health (WMH) Survey Initiative. We thank the WMH staff for assistance with instrumentation and fieldwork. References Anker, J.J., Carroll, 2011. Females are more vulnerable to drug abuse than males: evidence from preclinical studies and the role of ovarian hormones. Curr. Top. Behav. Neurosci. 8, 73–96. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, (DSM-IV), 4th ed. American Psychiatric Association, Washington, DC. Becker, H.C., Lopez, M.F., Doremus-Fitzwater, T.L., 2011. Effects of stress on alcohol drinking: a review of animal studies. Psychopharmacology (Berl.) 218, 131–156. Benjet, C., Borges, G., Medina-Mora, M.E., 2010. Chronic childhood adversity and onset of psychopathology during three life stages: childhood, adolescence and adulthood. J. Psychiatr. Res. 44, 732–740. Benjet, C., Borges, G., Medina-Mora, M.E., Blanco, J., Zambrano, J., Orozco, R., Fleiz, C., Rojas, E., 2007a. Drug use opportunities and the transition to drug use among adolescents in the Mexico Metropolitan Area. Drug Alcohol Depend. 90, 128–134. Benjet, C., Borges, G., Medina-Mora, M.E., Fleiz, C., Blanco, J., Zambrano, J., Rojas, E., Ramirez, M., 2007b. Prevalence and socio-demographic correlates of drug use among adolescents: results from the Mexican Adolescent Mental Health Survey. Addiction 102, 1261–1268. Benjet, C., Borges, G., Medina-Mora, M.E., Zambrano, J., Cruz, C., Méndez, E., 2009b. Descriptive epidemiology of chronic childhood adversity in Mexican adolescents. J. Adolesc. Health 45, 483–489. Benjet, C., Hernandez-Montoya, D., Borges, G., Mendez, E., Medina-Mora, M.E., Aguilar-Gaxiola, S., 2012. Youth who neither study nor work: mental health, education and employment. Salud Publica Mex. 54, 410–417. Benjet, C., Medina-Mora, M.E., Borges, G., Zambrano, J., Aguilar-Gaxiola, S., 2009a. Youth mental health in a populous city of the developing world: results from the Mexican Adolescent Mental Health Survey. J. Child Psychol. Psychiatry 50, 386–395.
Bruffaerts, R., Demyttenaere, K., Borges, G., Haro, J.M., Chiu, W.T., Hwang, I., Karam, E.G., Kessler, R.C., Sampson, N., Alsonso, J., Andrade, L.H., Angermeyer, M., Benjet, C., Bromet, E., de Girolamo, G., de Graaf, R., Florescu, S., Gureje, O., Horiguchi, I., Hu, C., Kovess, V., Levinson, D., Posada-Villa, J., Sagar, R., Scott, K., Tsang, A., Vassilev, S.M., Williams, D.R., Nock, M.K., 2010. Childhood adversities as risk factors for onset and persistence of suicidal behavior. Br. J. Psychiatry 197, 20–27. Chen, C.Y., Dormitzer, C.M., Bejarano, J., Anthony, J.C., 2004. Religiosity and the earliest stages of adolescent drug involvement in seven countries of Latin America. Am. J. Epidemiol. 159, 1180–1188. Courtney, M.E., Piliavin, I., Grogan-Kaylor, A., Nesmith, A., 1998. Foster youth transitions to adulthood: a longitudinal view of youth leaving care. Child Welfare 80, 685–717. Degenhardt, L., Dierker, L., Chiu, W.T., Medina-Mora, M.E., Neumark, Y., Sampson, N., Alonso, J., Angermeyer, M., Anthony, J.C., Bruffaerts, R., de Girolamo, G., de Graaf, R., Gureje, O., Karam, A.N., Kostyuchenko, S., Lee, S., Lépine, J.P., Levinson, D., Nakamura, Y., Posada-Villa, J., Stein, D., Wells, J.E., 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 Alcohol Depend. 108, 84–97. Delva, J., Van Etten, M.L., Gonzalez, G.B., Cedeno, M.A., Penna, M., Caris, L.H., Anthony, J.C., 1999. First opportunities to try drugs and the transition to first drug use: evidence from a national school survey in Panama. Subst. Use Misuse 34, 1451–1467. Efron, B., 1988. Logistic regression, survival analysis and the Kaplan–Meier curve. J. Am. Stat. Assoc. 83, 414–425. Endicott, J., Andreasen, N., Spitzer, R.L., 1978. Family History Research Diagnostic Criteria. New York State Psychiatric Institute, New York, NY. Enoch, M.A., Hodgkinson, C.A., Yuan, Q., Shen, P.H., Goldman, D., Roy, A., 2010. The influence of GABRA2, childhood trauma, and their interaction on alcohol, heroin, and cocaine dependence. Biol. Psychiatry 67, 20–27. Garland, E.L., Pettus-Davis, C., Howard, M.O., 2012. Self-medication among traumatized youth: structure equation modeling of pathways between trauma history, substance misuse, and psychological distress. J. Behav. Med. (Epub ahead of print). Goeders, N.E., Guerin, G.F., 1996. Effects of surgical and pharmacological adrenalectomy on the initiation and maintenance of intravenous cocaine selfadministration in rats. Brain Res. 722, 145–152. Goeders, N.E., 2002. Stress and cocaine addiction. J. Pharmacol. Exp. Ther. 30, 785–789. Greenfield, S.F., Kolodziej, M.E., Sugarman, D.E., Muenz, L.R., Vagge, L.M., He, D.Y., Weiss, R.D., 2002. History of abuse and drinking outcomes following inpatient alcohol treatment: a prospective study. Drug Alcohol Depend. 67, 227–234. Heffner, L.J., Blom, T.J., Anthenelli, R.M., 2011. Gender differences in trauma history and symptoms as predictors of relapse to alcohol and drug use. Am. J. Addict. 20, 307–311. Higley, A.E., Crane, N.A., Spadoni, A.D., Quello, S.B., Goodell, V., Mason, B.J., 2011. Craving in response to stress induction in a human laboratory paradigm predicts treatment outcome in alcohol-dependent individuals. Psychopharmacology (Berl.) 218, 121–129. Hyman, S.M., Paliwal, P., Chaplin, T.M., Mazure, C.M., Rounsaville, B., Sinha, R., 2008. Severity of childhood trauma is predictive of cocaine relapse outcomes in women but not men. Drug Alcohol Depend. 92, 208–216. Kessler, R.C., Avenevoli, S., Green, J., Gruber, M.J., Guyer, M., He, Y., Jin, R., Kaufman, J., Sampson, N.A., Zaslavsky, A.M., 2009. National Comorbidity Survey Replication Adolescent Supplement: III. Concordance of DSM.IV/CIDI diagnoses with clinical reassessments. J Am. Acad. Child Adolesc. Psychiatry 48, 386–399. Kessler, R.C., McLaughlin, K.A., Green, J.G., Gruber, M.J., Sampson, N.A., Zaslavsky, A.M., Aguilar-Gaxiola, S., Alhamzawi, A.O., Alonso, J., Angermeyer, M., Benjet, C., Bromet, E., Chatterji, S., de Girolamo, G., Demyttenaere, K., Fayyad, J., Florescu, S., Gal, G., Gureje, O., Haro, J.M., Hu, C., Karam, E.G., Kawakami, N., Lee, S., Lépine, J.P., Ormel, J., Posada-Villa, J., Sagar, R., Tsang, A., Üstün, B., Vassilev, S., Viana, M.C., Williams, D.R., 2010. Childhood adversities and adult psychopathology in the WHO World Mental Health Surveys. Br. J. Psychiatry 197, 378–385. Khantzian, E.J., 1985. The self-medication hypothesis of addictive disorders: focus on heroin and cocaine dependence. Am. J. Psychiatry 142, 1259–1264. Knauper, B., Cannell, C.F., Schwarz, N., Bruce, M.L., Kessler, R.C., 1999. Improving the accuracy of major depression age of onset reports in the US National Comorbidity Survey. Int. J. Methods Psychiatr. Res. 8, 39–48. Koob, G., Kreek, M.J., 2007. Stress, dysregulation of drug reward pathways and the transition to drug dependence. Am. J. Psychiatry 164, 1149–1159. Kreek, M.J., Nielson, D.A., Butelman, E.R., LaForge, K.S., 2005. Genetic influences on impulsivity, risk taking, stress responsivity and vulnerability to drug abuse and addiction. Nat. Neurosci. 8, 1450–1457. Liang, W., Lenton, S., Allsop, S., Chikritzhs, R., 2011. Does availability of illicit drugs mediate the association between mental illness and substance use? Subst. Use Misuse 46, 1304–1308. Lloyd, D.A., Turner, R.J., 2008. Cumulative life-time adversities and alcohol dependence in adolescence and young adulthood. Drug Alcohol Depend. 93, 217–226. Lopez, M.F., Doremus-Fitzwater, T.L., Becker, H.C., 2011. Chronic social isolation and chronic variable stress during early development induce later elevated ethanol intake in adult C57BL/6J mice. Alcohol 45, 355–364. Lynch, W.J., Kushner, M.G., Rawleigh, J.M., Fisdon, J., Caroll, M.E., 1999. The effects of restraint stress on voluntary ethanol consumption in rats. Exp. Clin. Psychopharmacol. 7, 318–323. McEwen, B.S., 2000. Allostatis and allostatic load: implications for neuropsychopharmacology. Neuropsychopharmacology 22, 1083–1124.
C. Benjet et al. / Drug and Alcohol Dependence 131 (2013) 85–91 Merikangas, K.R., Avenevolli, S., Costello, E.J., Koretz, D., Kessler, R.C., 2009. National Comorbidity Survey Replication adolescent supplement: I. Background and measures. J. Am. Acad. Child Adolesc. Psychiatry 48, 367–379. Research Triangle Institute, 2002. Sudaan Release 8.0.1. Research Triangle Institute, Research Triangle Park, NC. Robinson, J., Sareen, J., Cox, B.J., Bolton, J.M., 2011. Role of self-medication in the development of comorbid anxiety and substance use disorders: a longitudinal investigation. Arch. Gen. Psychiatry 68, 800–807. Sinha, R., Fox, H.C., Hong, K.I., Hansen, J., Tuit, K., Kreek, M.J., 2011. Effects of adrenal sensitivity, stress- and cue-induced craving, and anxiety on subsequent alcohol relapse and treatment outcomes. Arch. Gen. Psychiatry 68, 942–952. Sinha, R., Garcia, M., Paliwal, P., Kreek, M.J., Rounsaville, B.J., 2006. Stress-induced cocaine craving and hypothalamic-pituitary-adrenal responses are predictive of cocaine relapse outcomes. Arch. Gen. Psychiatry 63, 324–331. Storr, C.L., Wagner, F.A., Chen, C.Y., Anthony, J.C., 2011. Childhood predictors of first chance to use and use of cannabis by young adulthood. Drug Alcohol Depend. 117, 7–15. Strauss, M.A., 1979. Measuring intrafamily conflict and violence: the conflict tactics (CT) scales. J. Marriage Fam. 41, 75–88. Tolin, D.F., Foa, E.B., 2006. Sex differences in trauma and posttraumatic stress disorder: a quantitative review of 25 years of research. Psychol. Bull. 132, 959–992.
91
Turner, R.J., Lloyd, D.A., 2003. Cumulative adversity and drug dependence in young adults: racial/ethnic contrasts. Addiction 98, 305–315. Uhart, M., Wand, G.S., 2009. Stress, alcohol and drug interaction: an update of human research. Addict. Biol. 14, 43–64. Van Etten, M.L., Anthony, J.C., 1999. Comparative epidemiology of initial drug opportunities and transitions to first use: marijuana, cocaine, hallucinogens, and heroin. Drug Alcohol Depend. 54, 117–125. Van Etten, M.L., Neumark, Y.D., Anthony, J.C., 1999. Male–female differences in the earliest stages of drug involvement. Addiction 94, 1413–1419. Veijola, J., Laara, E., Joukamaa, M., Isohanni, M., Hakko, H., Haapea, M., Pirkola, S., Maki, P., 2008. Temporary parental separation at birth and substance use disorder in adulthood. A long-term follow-up of the Finnish Christmas Seal Home Children. Soc. Psychiatry Psychiatr. Epidemiol. 43, 11–17. Vengeliene, V., Siegmund, S., Singer, M.V., Sinclair, J.D., Li, T.K., Spanagel, R., 2003. A comparative study on alcohol-preferring rat lines: effects of deprivation and stress phases on voluntary alcohol intake. Alcohol. Clin. Exp. Res. 27, 1048–1054. Willett, J.B., Singer, J.D., 1993. Investigating onset, cessation, relapse, and recovery: why you should, and how you can, use discrete-time survival analysis to examine event occurrence. J. Consult. Clin. Psychol. 61, 952–965.