^aServei d'Avaluació i Mètodes d'Intervenció, Agència de Salut Pública de Barcelona, España.
^bUnitat Docent de Medicina Preventiva i Salut Pública IMAS-UPF-ASPB, Barcelona, España.

Correspondence

ABSTRACT

Objective: To determine the factors associated with the onset of cannabis use through a systematic review of cohort studies.
Methods: An internet-based search was performed using several keywords and their combinations. Original studies with longitudinal design and the onset of cannabis use as dependent variable, as well as review studies were included, published between January 1980 and May 2004. Methodology quality of the studies was assessed independently by two reviewers, according to pre-established criteria, in order to classify studies in high, mid or low quality. Agreement between reviewers was assessed through kappa coefficient.
Results:A total of 32 relevant studies were identified, of which 13 were of higher quality. Selection bias for the inclusion of consumers at the baseline measurement and lack or insufficient adjustment for confounders were the causes of exclusion. The factors of great evidence related to the onset of cannabis use were masculine sex, consumption of tobacco or alcohol, having a problematic relationship with parents, and cannabis consumption by friends.
Conclusion: Results highlight the importance of different individual, family and environmental factors on the onset of cannabis use. These must be considered to properly arrange intervention programs focusing on primary prevention among teenagers.

Key words: Cannabis. Systematic review. Longitudinal studies.

RESUMEN

Objetivo: Determinar los factores asociados al inicio del consumo de cannabis a partir de una revisión sistemática de estudios de cohortes.
Métodos: Se realizó una búsqueda bibliográfica informatizada utilizando diversas palabras clave y sus combinaciones. Entre los estudios identificados se seleccionaron los estudios originales de diseño longitudinal que utilizaran como variable dependiente el inicio de consumo de cannabis, así como los estudios de revisión, publicados entre enero de 1980 y mayo de 2004. Se evaluó la calidad metodológica de los estudios mediante 2 revisores, de manera independiente y a partir de unos criterios prestablecidos, clasificando los artículos en 3 categorías: alta, intermedia o baja calidad. Se midió el nivel de concordancia de los revisores a partir del coeficiente kappa.
Resultados: Se identificaron 32 estudios que cumplían los criterios de selección, de los cuales 13 fueron determinados de alta calidad. Las causas de exclusión fueron el sesgo de selección, por la inclusión de consumidores al inicio de la cohorte estudiada, y la falta de ajuste por los potenciales confusores. Los factores que se relacionaron con una mayor evidencia con el inicio del consumo de cannabis fueron el sexo masculino, el consumo de tabaco y alcohol, tener una relación problemática con los padres y el consumo de cannabis por parte de losamigos.
Conclusiones: Los resultados señalan la importancia de diversos factores individuales, familiares y del entorno en el inicio del consumo de cannabis, que deberían considerarse de forma conjunta en el abordaje preventivo entre los adolescentes.

Palabras clave: Cannabis. Revisión sistemática. Estudios longitudinales.

Introduction

The consumption of cannabis has various consequences for human health at the physiological, social and psychological levels^1-4. Although relatively few studies have systematically assessed the impact of cannabis on human health, the use of cannabis has been associated with cases of bronchitis, chronic cough and pre-cancerous alterations in the pulmonary tissue^1-3. In addition several studies have shown that the use of cannabis interferes with the normal workings of complex functions such as memory, concentration and learning^1-4, can cause a reduction in the educational achievement of teenagers^5,6, exacerbates psychosis^7,8 and increases the risk of traffic accidents^1-4.

Cannabis is the most frequently consumed illicit drug in most developed countries^4,9. In Europe as a whole, it is estimated that at least one in five adults has consumed cannabis at some time in their life⁹. Furthermore, in recent years there has been a noticeable increase in cannabis consumption amongst teenagers and young adults⁹. In 2003, 29.0% of the Spanish population aged between 15 and 64 admitted to having consumed cannabis at some time in their life¹⁰, while in young people aged between 15 and 29 this percentage rose to 35.9%¹¹.

Knowledge of the factors associated with the onset of cannabis consumption is essential for designing and implementing prevention programs aimed at teenagers and young adults. Although various authors have previously reviewed factors associated with the consumption of cannabis and other illicit drugs, such reviews have generally been rather unsystematic. They have been fundamentally based on cross-sectional studies and have tended not to exclude studies containing significant methodological limitations^12-17.

The aim of this work was to determine the factors associated with the onset of cannabis consumption by means of a systematic review of cohort studies.

Methods

An internet-based bibliographical search was carried out consulting Medline, Cochrane Library, ISI data bases, Documed, PsycINFO, PsycARTICLES and various thesis data bases (Teseo and the data bases of various universities). The following key words and their combinations were used: "cannabis", "marihuana", "associated factors", "related factors", "predictive factors", "determinants", "adolescents", "students", "youth", "school", "longitudinal", "follow-up", and "prospective". The selection criteria for the study established that references should be original articles published between January 1980 and May 2004, in English, Spanish, French, Italian or Portuguese, that they should be empirical, involve cohort studies and consider the onset of cannabis consumption as a dependent variable. The total number of articles was completed with the addition of other references from articles selected as a result of the internet search. Reviews were also used to help locate further original sources. All of the authors agreed with this search strategy.

Two of the authors (MG and DO) carried out a critical review of the studies, basing their decisions on the following criteria: sample size and representativity; average age at the beginning of the study and average age at the onset of cannabis use; length of follow-up; clear definition of exposure and outcome and valid method of assessment; conclusions supported by results that also considered potential sources of bias and limitations; level of adjustment for potential confounders; and with the onset of cannabis use as a dependent variable, having excluded from the analysis subjects who were consumers of cannabis at the onset of the monitoring period. These last 3 criteria were used to define the methodological quality of the studies. The reviewers independently assessed the articles and assigned them to one of three quality categories: high, medium or low. When discrepancies arose, a third reviewer (CA) analysed the article in question and made an independent decision. The degree of concordance between the reviewers was measured using the kappa coefficient.

Only studies that were judged to be of methodologically high quality were selected to determine the factors associated with the onset of cannabis use. The factors identified were grouped into a smaller number of dimensions on the basis of conceptual similarity in order to render the final analysis more operative (table 1).

Of the 13 studies finally selected as being of high quality, 8 used Odds Ratios (OR) as a measurement of association. Most of these studies showed a 95% confidence interval (CI), but in some only the p-value was provided. In one study, survival analysis was carried out and Hazard Ratios (HR) were used to measure association with a 95% confidence interval. Three of the 13 studies selected employed multivariate hierarchical models and used the coefficient of determination to estimate the level of association. All of the factors analysed in 2 or more studies that employed either OR or HR as a measure of association are presented in figure 1. In one of these studies³⁰ the results were analysed for a single sample of pupils but have been presented after stratification into different age groups. These results were treated as a single study and we selected the largest OR for each factor.

Results

We identified 33 articles that met the selection criteria and which corresponded to 32 different studies. Two articles presented results from the same study and so the one that presented the more preliminary results was discarded. Very few articles discussed the question of bias and methodological limitations. This parameter could not therefore be used to discriminate between the quality of different studies and was consequently rejected as a criterion for study quality. Of the 32 studies selected, 13 were classified as being of high methodological quality (table 2). Three of the 13 studies had oversampled populations with either low socio-economic levels or a high prevalence of the consumption of addictive substances^22,25,27. The main limitations identified in the 19 studies judged to be of medium or low methodological quality were: selection bias, as they did not exclude people who were already cannabis consumers at the onset of the cohort monitoring period; and the lack of adjustment for potential confounders presenting results obtained from bivariate analyses^29,32-48. The degree of concordance between the reviewers measured by the kappa coefficient was 0.85.

The 13 studies included in this review operatively defined the "onset of cannabis use" on the basis of lifetime cannabis consumption. Seven of these 13 studies defined this use on the basis of questions such as "Have you ever smoked marijuana?", which had two possible answers: 'never' or 'at some time'^{17,18,20,24,25,28,30}. Six of the 13 studies defined it from questions relating to frequency of use, such as "How often have you smoked marijuana?"^{19,21-23,26,27}. For the purposes of analysis, these studies classified the replies in terms of 'never' vs. 'whatever consumption frequency'. Two of these 6 studies explored past consumption, with one examining the previous year and the other focusing on the previous 6 months^19-21. We decided to include these studies because cannabis consumption by adolescents at some time in the past year and at some time in the past six months showed a close correlation with consumption at some time in the individual's life.

Figure 1 shows the factors associated with the onset of cannabis consumption identified in 10 of the 13 studies judged to be of high methodological quality. In one of these studies, it was only possible to chart a single factor, which was seen as the most interesting with respect to the aims of that particular study, while the only information provided for the other factors was their respective degree of significance¹⁸. Of these factors, low socio-economic status and poor relations with parents were the factors that were significantly associated with the onset of cannabis use. No statistically significant association was observed with either coming from a single parent family, having moved home, having an emotionally sensitive mother, or avoiding punishment.

In the 3 studies that could not be charted, the factors identified and significantly associated with the onset of cannabis consumption were: female sex²⁷, age (great age, greater use)²⁷, low socio-economic status¹⁸, belonging to a single parent family^18,27, forming part of a problem family^26,27, suffering the effects of strong parental discipline²³, individual personality^23,26 and having a cannabis consuming father²⁷. In contrast, the factors that were identified but not significantly associated with the onset of cannabis consumption were: drug consumption by the individual him/herself³¹; cannabis consumption by friends²³; and tobacco and alcohol consumption by the individual's father and mother²⁷.

Factors that were analysed in isolated studies (not represented in figure 1) and which had a significant association with cannabis consumption include: going to church (OR = 0.73; 95% CI, 0.56-0.94)²⁴; belonging to a problem family (OR = 1.22; p<.05)¹⁹; intending to consume drugs in the future (OR = 2.8; 95% CI, 1.8-4.4)¹⁷; having a 'non-conventional' personality (OR = 1.32; 95% CI, 1.01-1.75)²⁴; suffering psychological problems (OR = 1.08; 95% CI, 1.01-1.10)²⁸; exhibiting socially conflictive behaviour and rejecting conventional norms (OR=1.75; 95% CI, 1.17-1.47)²⁰; following rules against the consumption of cannabis (HR = 0.64; 95% CI, 0.46-0.81)²⁵; being subjected to strict parental discipline (HR = 0.75; 95% CI, 0.53-0.96)²¹; and having siblings who consume alcohol (OR = 1.45; 95% CI, 1.01-2.12)²⁴ or cannabis (HR = 1.48; 95% CI, 1.36-1.60)²⁵. Three studies from the United States explored the variable 'belonging to a particular race/ethnicity' without obtaining any conclusive results^22,25,30.

Discussion

As far as we know, this is the only systematic review of cohort studies that has aimed to identify the factors associated with the onset of cannabis consumption. A review of longitudinal studies makes it possible to identify independent risk factors that can help predict the onset of cannabis consumption and which play a causal role in, or are otherwise associated with, the onset of cannabis consumption. As tends to occur in cases of behaviour exhibiting a complex and multi-factorial etiology, the studies identify a wide range of risk factors which include many different aspects. These include factors relating to social and family circumstance or educational performance and other factors such as the availability of drugs in the immediate environment, the behavioural models of parents and -above all- friends, and other psychological pathologies or sources of conflictive behaviour. However, the factors identified that seem to merit particular attention on account of their frequent appearance in a large number of studies are: male sex^{17,22,24,25,28}, the consumption of tobacco^17,21-24 and alcohol^17,21-23,25, problematic relations with parents^17,19,24,30 and cannabis consumption by friends^{17,21,24,25,30}.

In most studies^{17,22,24,25,28}, it was observed that boys had a greater risk of becoming cannabis consumers than girls. Even so, some studies^19,27 reported a greater incidence of consumption among girls, especially before the age of 21¹⁹, as well as a greater degree of dependence amongst female consumers. In another study²⁷, an association was observed between cannabis consumption by mothers and the onset of consumption by their daughters. Whatever the case, beyond any specific mechanisms that may explain part of these differences, the influence of gender should be evaluated in the light of the consumption tendencies described in several recent studies⁴⁹. In our setting, recent years have seen a certain convergence in the indicators relating to the consumption of tobacco, alcohol and cannabis by teenagers. Several studies carried out in the United States show that the differences observed in the prevalence of drug consumption between the two sexes could be largely due to differences in opportunities for consumption, rather than to differences in the probability of consuming for the first time when an opportunity presents itself^50,51. This suggests that the greater cannabis consumption observed amongst boys could be explained by the fact that they tend to have more opportunities to start consuming the drug than girls: when girls have the same opportunities they exhibit the same probability of consuming cannabis.

Kandel et al⁵². have presented a stepping-stone model for drug consumption which is currently subject to discussion. This model is based on a putative progression in consumption, which starts with legal drugs such as tobacco and alcohol. According to these authors, this would facilitate the consumption of marihuana, which -in turn- would subsequently act as a gateway to the consumption of other illegal and psychoactive drugs. In this review, the consumption of alcohol and tobacco both appear as factors associated with the consumption of cannabis, which would tend to confirm the stepping-stone model^17,21-25, at least with regard to the increased risk of cannabis consumption amongst the consumers of legalised drugs. This association could be due to the form of consumption, in the case of tobacco, and to the search for immediate effects, in the case of alcohol, in addition to other common mechanisms such as the symbolic nature of transgression and socialising aspects of consumption. The stepping-stone consumption model has, however, been questioned in some studies^53-55 that showed that cannabis consumption was a factor associated with the onset of tobacco consumption. This could be explained by the fact that in environments in which the consumption of cannabis is very prevalent, it is possible to observe the onset of cannabis consumption previous to that of tobacco amongst some young adults. Other authors suggest that the consumption of tobacco, alcohol and cannabis may start at the same time, because the determining factors associated initiation are common to all three drugs^56-58. Moreover, in another study referred to in this review, the use of other illegal and psychoactive drugs was also signalled as a factor associated with initiation in the use of cannabis. This was probably due, at least in part, to the fact that the study in question focused on a population with a low socio-economic level and a high prevalence of drug abuse²². In such cases, it has been suggested that the populations in question exhibits a different pattern of use from the rest of the population and that the risk factors associated with initiation in the use of drugs are different, although any such differences could be fundamentally due to environmental conditions such as the availability of drugs in the local setting.

Another relevant factor is poor relations with parents^17,19,24,30. This factor encompasses the importance of having good and harmonious communication with parents, relates to the ability to establish such communication, and also concerns the relationship between mother and child and the interaction between children and their parents and how the former identify with the latter. The importance of the parental role model is demonstrated within the range of social influences⁵⁹. A bad child-parent relationship implies the elimination of the passive model as a protective factor. It also implies the loss of the subjective norm, which is what prompts adolescents to adjust their behaviour in accordance with the position of relevant figures within their social environment, such as -in this case- their parents. The existence of a good, close parent-child relationship tends to favour the development of self-regulatory mechanisms and more effective social abilities⁶⁰.

Cannabis use by peers is signalled as another important factor in the initiation of its use by young people^{17,21,24,25,30}. The role of friends has, however, been the subject of much argument in recent years. Several authors have suggested that it is more a matter of peer selection than of peer pressure^61-63. Whatever the case, these would no doubt be interrelated phenomena, and it should be stressed that in the present review this factor appeared amongst the most important factors for a wide age range; for young people aged between 14 and 26.

The availability of drugs in the local setting appears to be a factor of average importance. It would, however, seem more credible if it were a factor of major importance as it is an accessibility factor, and is seen as a key facilitating factor in the PRECEDE model⁶⁴. This average rating could be due to the fact that this variable was not taken into account in all of the studies.

As far as the methodological limitations of this study are concerned, it should first be stressed that the review process was subject to publication bias, as the search was only conducted using free searc h engines and we only selected articles published in English, Spanish, French, Italian or Portuguese. Another important limitation was the fact that the same factors were not analysed in all of the studies considered. As seen in table 1, there were different approximations to the same dimension in the different studies selected. In the studies that examined the use of drugs and different types of conduct, the diversity of potentially related factors made it difficult to consider all of the variables in each study. If a factor was only identified in a single study, it could mean that it was less important than others or that it had simply not been investigated in other studies. It should also be remembered that in these studies the variables were gathered by means of self-administered questionnaires or interviews and may therefore have been subject to information bias due to the limited validity of some responses.

It is also appropriate to point out that previously published reviews relating to factors associated with drug use were not systematic^12-17: they included cross-sectional studies and did not exclude studies of low methodological quality, such as studies whose results were not adjusted for potential confounders. Moreover, these reviews did not focus their attention on the onset of drug use but rather on the use of drugs themselves, without differentiating between incident and prevalent use. Furthermore, these other reviews included the use of alcohol as well as that of cannabis and other illegal drugs.

Models that analyse behaviour are not determinant causal models. Above all, this type of study helps to define risk patterns which tend to favour certain types of conduct. The predictive factors that have been identified in this review are principally social and psychological in nature. In consequence, preventative measures should focus on increasing the awareness of adolescents of the social influences that could lead them to consume cannabis and also on providing them with the tools with which to face up to such situations. It is also important to educate young people in values and beliefs related to drug use and, at the same time, to introduce preventive measures to discourage the use of tobacco and alcohol. Botvin's Life Skills Training model^65,66 has helped to reduce the consumption of addictive substances, as well as provoking a change in behaviour with respect to drugs. It therefore provides an initial sign that primary prevention of cannabis use in this age group is possible. Such programs also need to include components that are directed towards parents in order to strengthen relations between adolescents and their parents. In doing this, they should also help to prevent the consumption of drugs.

In conclusion, the factors associated with the onset of cannabis consumption tend to relate to: the individual, including such considerations as age, sex, poor academic achievement, legal drug use and antisocial conduct; the family, with factors such as the level of communication between parents and children and relations within the family, family problems, belonging to a single parent family, and strict parental discipline; and environmental factors, such as the influence of friends, the consumption of cannabis by friends and the availability of drugs in the local setting. These factors should all be taken into account when developing strategies aimed at preventing the consumption of cannabis by adolescents and young people.

Acknowledgements

The authors are grateful to Isaac Subirana for help and assessment when drawing up the figures and to María Grau and Andrea Burón for reviewing the final version of this manuscript.

Correspondence:
Dr. Manel Nebot.
Servei d'Avaluació i Mètodes d'Intervenció.
Agència de Salut Pública de Barcelona.
Pl. Lesseps, 1. 08015 Bacelona. España.
Correo electrónico: mnebot@aspb.es

Received: 8 de febrero de 2006.
Accepted: 21 de diciembre de 2006.

References

1. Ashton H. Cannabis or health? Curr Opin Psychiatry. 2002;15: 247-53.        [ Links ]

2. Bobes J, Calafat A. Monografía del cannabis. Addiciones. 2000;12 Supl 2:5-329.        [ Links ]

3. Kalant H. Adverse effects of cannabis on health: an update of the literature since 1996. Prog Neuropsychopharmacol Biol Psychiatry. 2004;28:849-63.        [ Links ]

4. Minister of Public Health of Belgium. Cannabis 2002 Report. A joint international effort at the initiative of the Ministers of Public Health of Belgium, France, Germany, The Netherlands, Switzerland. Brussels: Technical Report of the International Scientific Conference; 2002.        [ Links ]

5. Fergusson DM, Horwood LJ, Beautrais AL. Cannabis and educational achievement. Addiction. 2003;98:1681-92.        [ Links ]

6. Macleod J, Oakes R, Copello A, Crome I, Egger M, Hickman M, et al. Psychological and social sequelae of cannabis and other illicit drug use by young people: a systematic review of longitudinal, general population studies. Lancet. 2004;363: 1579-88.        [ Links ]

7. Henquet C, Krabbendam L, Spauwen J, Kaplan C, Lieb R, Wittchen HU, et al. Prospective cohort study of cannabis use, predisposition for psychosis, and psychotic symptoms in young people. BMJ. 2005;330:11.        [ Links ]

8. Semple DM, McIntosh AM, Lawrie SM. Cannabis as a risk factor for psychosis: systematic review. J Psychopharmacol. 2005;19:187-94.        [ Links ]

9. European Monitoring Centre for Drugs and Drug Addiction. Annual Report 2004: The state of the drugs problem in the European Union and Norway. Office for Official Publications of the European Communities, Luxembourg, pp. 28-30 [citado 12 Jun 2006]. Disponible en: http://annualreport.emcdda.eu.int/download/ar2004-en.pdf        [ Links ]

10. Observatorio Español sobre Drogas. Informe 2004. Ministerio del Interior Delegación del Gobierno para el Plan Nacional sobre Drogas, España, pp. 22-37 [citado 31 Dic 2004]. Disponible en: http://www.pnsd.msc.es/Categoria2/publica/pdf/oed-2004.pdf        [ Links ]

11. Observatorio Español sobre Drogas. Informe n.º 5, julio 2002. Ministerio del Interior Delegación del Gobierno para el Plan Nacional sobre Drogas, España, p. 25-26 [citado 31 Dic 2004]. Disponible en: http://www.pnsd.msc.es/Categoria2/publica/pdf/oed-5.pdf        [ Links ]

12. Cabrera S. Drug use among hispanic youth: examinig comon and unique contributing factors. Hisp J Behav Sci. 1999;21:89-103.        [ Links ]

13. Glynn T. From family to peer: a review of transitions of influence among drug using youth. J Youth Adolesc. 1981;10:363-83.        [ Links ]

14. Hawkins JD, Catalano RF, Miller JY. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: implications for substance abuse prevention. Psychol Bull. 1992;112:64-105.        [ Links ]

15. Kandel DB. Drug and drinking behavior among youth. Ann Rev Sociol. 1980;6:235-85.        [ Links ]

16. Kandel DB. The parental and peer contexts od adolescent deviance: an algebra of interpersonal influences. J Drug Issues. 1996;26:289-315.        [ Links ]

17. Von Sydow K, Lieb R, Pfister H, Hofler M, Wittchen HU. What predicts incident use of cannabis and progression to abuse and dependence? A 4-year prospective examination of risk factors in a community sample of adolescents and young adults. Drug Alcohol Depend. 2002;68:49-64.        [ Links ]

18. Fergusson DM, Lynskey MT, Horwood LJ. Conduct problems and attention deficit behaviour in middle childhood and cannabis use by age 15. Aust NZ J Psychiatry. 1993;27:673-82.        [ Links ]

19. McGee R, Williams S, Poulton R, Moffitt T. A longitudinal study of cannabis use and mental health from adolescence to early adulthood. Addiction. 2000;95:491-503.        [ Links ]

20. Pedersen W. Adolescents initiating cannabis use: cultural opposition or poor mental health? J Adolesc. 1990;13:327-39.        [ Links ]

21. Coffey C, Lynskey M, Wolfe R, Patton GC. Initiation and progression of cannabis use in a population-based Australian adolescent longitudinal study. Addiction. 2000;95:1679-90.        [ Links ]

22. Aitken SS, DeSantis J, Harford TC, Cases MF. Marijuana use among adults. A longitudinal study of current and former users. J Subst Abuse. 2000;12:213-26.        [ Links ]

23. Morojele NK, Brook JS. Adolescent precursors of intensity of marijuana and other illicit drug use among adult initiators. J Genet Psychol. 2001;162:430-450.        [ Links ]

24. Brook JS, Kessler RC, Cohen P. The onset of marijuana use from preadolescence and early adolescence to young adulthood. Dev Psychopathol. 1999;11:901-14.        [ Links ]

25. Kosterman R, Hawkins JD, Guo J, Catalano RF, Abbott RD. The dynamics of alcohol and marijuana initiation: patterns and predictors of first use in adolescence. Am J Public Health. 2000; 90:360-6.        [ Links ]

26. Siqueira LM, Brook JS. Tobacco use as a predictor of illicit drug use and drug-related problems in Colombian youth. J Adolesc Health. 2003;32:50-7.        [ Links ]

27. Andrews JA, Hops H, Ary D, Tildesley E, Harris J. Parental influence on early adolescent substance use: Specific and nonspecific effects. J Early Adolesc. 1993;13:285-310.        [ Links ]

28. Hammer T, Vaglum P. Users and nonusers within a high risk milieu of cannabis use. A general population study. Int J Addict. 1991;26:595-604.        [ Links ]

29. Fergusson DM, Horwood LJ. Early onset cannabis use and psychosocial adjustment in young adults. Addiction. 1997;92: 279-96.        [ Links ]

30. Bailey SL, Hubbard RL. Developmental variation in the context of marijuana initiation among adolescents. J Health Soc Behav. 1990;31:58-70.        [ Links ]

31. Brook JS, Lukoff IF, Whiteman M. Initiation into adolescent marijuana use. J Genet Psychol. 1980;137:133-42.        [ Links ]

32. Andrews JA, Tildesley E, Hops H, Li F. The influence of peers on young adult substance use. Health Psychol. 2002;21:349-57.        [ Links ]

33. Aseltine RH Jr. A reconsideration of parental and peer influences on adolescent deviance. J Health Soc Behav. 1995; 36:103-21.        [ Links ]

34. Brook DW, Brook JS, Rosen Z, Montoya I. Correlates of marijuana use in Colombian adolescents: a focus on the impact of the ecological/cultural domain. J Adolesc Health. 2002;31: 286-98.        [ Links ]

35. Brook JS, Brook DW, rencibia-Mireles O, Richter L, Whiteman M. Risk factors for adolescent marijuana use across cultures and across time. J Genet Psychol. 2001;162:357-74.        [ Links ]

36. Bryant AL, Schulenberg JE, O'Malley PM, Bachaman JG, Johnston LD. How academic achievement, attitudes, and behaviors relate to the course of substance use during adolescence: a 6-year, multiwave national longitudinal study. J Res Adol. 2003;13:361-97.        [ Links ]

37. Fleming JP, Kellam SG, Brown CH. Early predictors of age at first use of alcohol, marijuana, and cigarettes. Drug Alcohol Depend. 1982;9:285-303.        [ Links ]

38. Griffin KW, Botvin GJ, Scheier LM, Nichols TR. Factors associated with regular marijuana use among high school students: a long-term follow-up study. Subst Use Misuse. 2002; 37:225-38.        [ Links ]

39. Hofler M, Lieb R, Perkonigg A, Schuster P, Sonntag H, Wittchen HU. Covariates of cannabis use progression in a representative population sample of adolescents: a prospective examination of vulnerability and risk factors. Addiction. 1999; 94:1679-94.        [ Links ]

40. Hops H, Duncan TE, Duncan SC. Parent substance use as a predictor of adolescent use: a six-year lagged analysis. Ann Behav Med. 1996;18: 57-64.        [ Links ]

41. Kandel DB. On processes of peer influences in adolescent drug use: a developmental perspective. Adv Alcohol Subst Abuse. 1985;4:139-63.        [ Links ]

42. Molina BS, Pelham WE Jr. Childhood predictors of adolescent substance use in a longitudinal study of children with ADHD. J Abnorm Psychol. 2003;12:497-507.        [ Links ]

43. Newcomb MD, Felix-Ortiz M. Multiple protective and risk factors for drug use and abuse: cross-sectional and prospective findings. J Pers Soc Psychol. 1992;63:280-96.        [ Links ]

44. Poikolainen K, Tuulio-Henriksson A, Alto-Setala T, Marttunen M, Anttila T, Lonnqvist J. Correlates of initiation to cannabis use: a 5-year follow-up of 15-19-year-old adolescents. Drug Alcohol Depend. 2001;62:175-80.        [ Links ]

45. Sieber MF, Angst J. Alcohol, tobacco and cannabis: 12-year longitudinal associations with antecedent social context and personality. Drug Alcohol Depend. 1990;25:281-92.        [ Links ]

46. Stacy AW, Bentler PM, Flay BR. Attitudes and health behavior in diverse populations: drunk driving. Alcohol use, binge eating, marijuana use, and cigarette use. Health Psychol. 1994; 13:73-85.        [ Links ]

47. Sullivan TN, Farrell AD. Identification and impact of risk and protective factors for drug use among urban African American adolescents. J Clin Child Psychol. 1999;28:122-36.        [ Links ]

48. Yamaguchi K, Kandel DB. Patterns of drug use from adolescence to young adulthood (III). Predictors of progression. Am J Public Health. 1984;74:673-81.        [ Links ]

49. Nebot M, Giménez E, Ariza C, Tomás Z. Tendencias en el consumo de tabaco, alcohol y cannabis en los adolescentes de Barcelona entre 1987 y 2004. Med Clin (Barc). 2006;126: 159.        [ Links ]

50. Van Etten ML, Anthony JC. Comparative epidemiology of initial drug opportunities and transitions to first use: marijuana, cocaine, hallucinogens and heroin. Drug Alcohol Depend. 1999;54:117-25.        [ Links ]

51. Van Etten ML, Anthony JC. Male-female differences in transitions from first drug opportunity to first use: searching for subgroup variation by age, race, region, and urban status. J Womens Health Gend Based Med. 2001;10:797-804.        [ Links ]

52. Kandel DB, Yamaguchi K, Chen K. Stages of progression in drug involvement from adolescence to adulthood: further evidence for the gateway theory. J Stud Alcohol. 1992;53:447-57.        [ Links ]

53. Patton GC, Coffey C, Carlin JB, Sawyer SM, Lynskey M. Reverse gateways? Frequent cannabis use as a predictor of tobacco initiation and nicotine dependence. Addiction. 2005;100: 1518-25.        [ Links ]

54. Amos A, Wiltshire S, Bostock Y, Haw S, McNeill A. «You can't go without a fag... you need it for your hash»: a qualitative exploration of smoking, cannabis and young people. Addiction. 2004;99:77-81.        [ Links ]

55. Tullis LM, Dupont R, Frost-Pineda K, Gold MS. Marijuana and tobacco: a major connection? J Addict Dis. 2003;22:51-62.        [ Links ]

56. Morral AR, McCaffrey DF, Paddock SM. Reassessing the marijuana gateway effect. Addiction. 2002;97:1493-504.        [ Links ]

57. Anthony JC. Death of the «stepping-stone» hypothesis and the «gateway» model? Comments on Morral et al. Addiction. 2002;97:1505-7.        [ Links ]

58. Lynskey MT, Fergusson DM, Horwood LJ. The origins of the correlations between tobacco, alcohol, and cannabis use during adolescence. J Child Psychol Psychiatry. 1998;39:995-1005.        [ Links ]

59. DeVries H, Mudde AN, Dijkstra A, Willemsen MC. Differential beliefs, perceived social influences, and self-efficacy expectations among smokers in various motivational phases. Prev Med. 1998;27:681-9.        [ Links ]

60. Rothbart MK, Ahadi SA. Temperament and the development of personality. J Abnorm Psychol. 1994;103:55-66.        [ Links ]

61. Wang MQ, Eddy JM, Fitzhugh EC. Smoking acquisition: peer influence and self-selection. Psychol Rep. 2000;86:1241-6.        [ Links ]

62. Wills TA, Cleary SD. Peer and adolescent substance use among 6th-9th graders: latent growth analyses of influence versus selection mechanisms. Health Psychol. 1999;18:453-63.        [ Links ]

63. Oetting ER, Donnermeyer JF. Primary socialization theory: the etiology of drug use and deviance (I). Subst Use Misuse. 1998;33:995-1026.        [ Links ]

64. Green L, Kreuter MW. Health promotion planning: An educational and environmental approach. 2001. 2nd ed. Palo Alto: Mayfield Publishing Company; 1991.        [ Links ]

65. Botvin GJ, Baker E, Dusenbury L, Botvin EM, Diaz T. Long-term follow-up results of a randomized drug abuse prevention trial in a white middle-class population. JAMA. 1995;273: 1106-12.        [ Links ]

66. Botvin GJ, Griffin KW. Life skills training as a primary prevention approach for adolescent drug abuse and other problem behaviors. Int J Emerg Ment Health. 2002;4:41-7.        [ Links ]