Genetic contribution

Substance misuse is clearly linked with environmental adversities such as family and community breakdown. However, young people who misuse substances commonly exhibit a complex inherited predisposition, of which early use of drugs, alcohol or cigarettes may be a behavioural marker (Reference Costello, Erkanli and CopelandCostello 2010; Reference Green, McLaughlin and BerglundGreen 2010). For instance, data from the Minnesota Twin Family study (Reference McGue and IaconoMcGue 2008) confirm the long-held view that use of alcohol before the age of 15 is associated with increased risk of later alcohol misuse. Moreover, in a comparison with boys without early alcohol use, early drinking was linked to a range of adverse outcomes: the odds ratio for antisocial personality disorder was 5.8, for drug dependence 3.2, nicotine dependence 1.7 (and major depression 1.3). With the exception of a much lower risk for antisocial personality disorder, the pattern of associations was similar for girls. However, the risk of adverse outcomes, including alcohol dependence, associated with early-onset problem behaviour not related to alcohol misuse was sometimes even higher. Compared with young people without early problem behaviour, the risk for lifetime alcohol dependence among early smokers and early drug users was around 7 times higher. The corresponding risk for young people who had been in trouble with the police or had had early sexual intercourse was also considerably higher. McGue et al referred to this array of behaviours as ‘disinhibitory psychopathology’.

A population of adopted and non-adopted adolescents (Reference King, Keyes and MaloneKing 2009) revealed that young people living with an alcohol-dependent adoptive parent were more likely to drink than those living with a non-drinking adoptive parent, suggesting an environmental effect. However, those with alcohol-dependent biological parents displayed a range of ‘disinhibited’ behaviours, reflecting what the study’s authors argue is the inherited disinhibited predisposition. Figure 2 shows that the disinhibition is passed to the biological child of drinking parents; the adopted child of the drinking parent is ‘protected’ from the disinhibition syndrome.

FIG 2 Effect of parental alcohol dependence on standardised disinhibition factor scores in adopted and non-adopted adolescent offspring (after Reference King, Keyes and MaloneKing 2009, with permission)

These datasets support the view that what is transmitted genetically is not a specific predisposition to misuse substances, but a broad behavioural disposition. This disposition is likely to include clinical syndromes such as attention-deficit hyperactivity disorder (ADHD) and conduct disorder, each of which is independently predictive of substance misuse; these risk factors potentially act throughout adolescence (Reference Lynam, Charnigo and MoffittLynam 2009; Reference Beiderman, Petty and MonuteauxBeiderman 2010; Reference Langley, Fowler and FordLangley 2010). It appears also to include new forms of disinhibitory psychopathology such as cyberbullying (Reference Sourander, Klomek and IknoneSourander 2010).

‘Self-medication’

Self-treatment may be another contributory factor in substance misuse. Forms of psychopathology such as bipolar disorder and post-traumatic stress disorder (PTSD) are independently associated with higher risk of substance misuse (Reference Goldstein and BuksteinGoldstein 2010), perhaps especially among the most vulnerable in society (Reference Bender, Ferguson and ThompsonBender 2010). The effect of depression on risk of substance misuse appears substantially reciprocal, but depression may more often lead to misuse in females (Reference Gallerani, Garber and MartinGallerani 2010). Similarly, a study of females with PTSD that included older adolescents showed that successful treatment of PTSD predicted reduction in substance use but not vice versa (Reference Hien, Jiang and CampbellHien 2010). In a large adolescent population study, depression contributed to the uptake of smoking, and continued smoking contributed to ‘a dampening or levelling off of depression symptoms’. These findings, perhaps due to the modulating effect of nicotinic acetylcholine receptors on the release of neurotransmitters, suggest that smoking is an ‘effective’ self-therapy for depression (Reference Audrain-McGovern, Rodriguez and KasselAudrain-McGovern 2009).

Environment

Environmental factors interact with predisposition. For instance, constrained environments (e.g. those characterised by strong family bonds or those in which substances are less available) appear to suppress genetic risk, whereas unconstrained or stressful environments give it freer rein. Indeed, Reference Hicks, South and Di RagoHicks et al (2009) concluded that the greater the environmental stress, the greater the genetic effect. In disorganised environments with easy access to substances, or in which young people are left to their own devices, predispositions are more likely to be fulfilled than in environments where access to substances is limited, a potential gene–environment interaction. Social adversity may not so much ‘cause’ misuse as permit it.

Pre-existing mental illness

Reference Swendsen, Conway and DegenhardtSwendsen et al (2010) examined risks posed by a range of mental disorders and the degree to which substance use, misuse and dependence might be eliminated through their treatment. Analysing follow-up data from the US National Comorbidity Survey, and consistent with other longitudinal studies (e.g. Reference Copeland, Shanahan and CostelloCopeland 2009), they argue that risk is linked to affective disorders but that pre-existing behavioural disorders, including ADHD and conduct disorder, are the most powerful antecedents of substance misuse and that their successful early treatment could reduce illicit drug use and dependence by more than 70%. Indeed, early treatment of ADHD appears to reduce risk of later misuse: in a longitudinal case–control study of girls with ADHD, those treated with stimulants were 73% less likely to develop a substance use disorder (Reference Wilens, Adamson and MonuteauxWilens 2008).

Prevention programmes

Reference Riggs, Chou and PentzRiggs et al (2009) presented a long-term follow-up (from 11 to 28 years of age) of a universal intervention in the form of an early-adolescent substance misuse prevention programme. This comprised 10–13 approximately monthly teacher-led sessions of ‘resistance skills’ and ‘social-normative change’, a five-session booster 12 months later (age 13), and a parent programme including interactive sessions about homework and parent–child communication. Significant programme effects emerged at age 15 (2 years after the end of the programme) and appeared to increase until age 17, with differences between the intervention and control groups persisting thereafter. A European study has reported reduced drunkenness and less frequent cannabis use following a teacher-led classroom intervention (Reference Faggiano, Vigna-Taglianti and BurkhartFaggiano 2010). A parent-only intervention was significant at 12-month outcome but a pupil-only intervention had no effect on any outcomes (Reference Koning, Vollebergh and SmitKoning 2009). However, two sessions of a school-based targeted intervention involving group work with at-risk young people compared with no intervention yielded reduced illicit use over the 2-year follow-up (Reference Conrod, Castellanos-Ryan and StrangConrod 2010). Perhaps by helping mothers develop a sense of mastery, nurse home visiting of vulnerable mothers during their pregnancy and the first 2 years of their children’s lives was linked to reduced affective symptoms and substance misuse in the children 12 years later (Reference Kitzman, Olds and ColeKitzman 2010).

These data suggest that altering school curricula to attend to wider aspects of personal development can be effective in enhancing life skills, thus reducing later problem behaviours, and that combined and booster interventions may be important for detecting sustained effects in universal interventions. However, targeted interventions have the advantage of excluding those at low risk and, because of their smaller scale, are likely to be cheaper. What the exact balance of universal and targeted programmes should be remains unclear and is likely to depend on local social and political priorities as much as on science. Prolonged follow-up can detect important delayed or sleeper effects, perhaps reflecting altered developmental trajectories. Any interventions should be evaluated in different cultures. If successful, they may be cost-effective and may benefit whole societies, although reasonable fidelity to the original concept or technology is likely to be important.

Treatment

Opportunistic brief interventions in healthcare settings have shown positive effects on adult drinking. Consistent with this, focusing on adolescents exposed to aggression and substance use, a half-hour therapy in an emergency department reduced reported exposure to violence and aggression at 3-month and alcohol consequences at 6-month follow-up compared with an information brochure control intervention (Reference Walton, Chermack and ShopeWalton 2010). The precise components of a successful brief intervention for young people are not known, but such a finding shows that even taking a substance history and briefly discussing its findings may be helpful, opening to a wide range of clinicians the possibility of intervening.

In an RCT targeting substance misusing adolescents with ADHD, individual psychotherapy was linked with significant reduction in substance misuse (Reference Riggs, Winhusen and DaviesRiggs 2011). Those additionally treated with long-acting osmotic-release methylphenidate (OROS MPH) over the course of the 4-month trial showed additional benefits. Diversion of the methylphenidate (e.g. selling it on) and interaction with drugs of misuse were not reported, allowing the authors to state that OROS MPH is safe to use in this population. Focusing on opiate-dependent adolescents, Reference Woody, Poole and SubramaniamWoody et al (2008) reported the greater efficacy (as measured by abstinence) of 12 weeks as opposed to 2 weeks of buprenorphine treatment and concluded that a case exists for longer-term maintenance for this group, supplemented by psychosocial interventions.

Adolescent development occurs in ‘an ecology of nested systems’, such as peers, family, schools, recreation and juvenile justice. Within these systems, the relative influence of attachments to family and school versus, for instance, to deviant peers may shape behaviour (Reference Liddle, Rowe and DakofLiddle 2009). Multidimensional family therapy (MDFT) focuses on engagement of the young person and family, targeting adolescent, parental and family interactions as well as extra-familial domains of functioning (Reference Henderson, Rowe and DakofHenderson 2009).

Reference Henderson, Dakof and GreenbaumHenderson et al (2010) conducted a secondary analysis of two RCTs comparing MDFT with either cognitive–behavioural therapy (CBT) or enhanced (through support with engagement and transport) treatment as usual. This demonstrated that, for milder forms of misuse, the more intensive MDFT treatment had no greater impact than less intensive CBT (Fig. 3). However, statistically significant differences favouring MDFT emerged among those with indicators of more severe substance misuse. Consistent with clinical experience, this important interaction seems to confirm that treatment should be tailored to the severity and complexity of the presenting problems.

FIG 3 Treatment effect differences for milder and more severe substance misuse. MDFT, multidimensional family therapy; CBT, cognitive–behavioural therapy (after Reference Henderson, Dakof and GreenbaumHenderson 2010, with permission).

Prognosis may be related to aftercare. In an 8-year follow-up of adolescents who had received in-patient treatment for alcohol or drug dependence, Reference Kelly, Brown and AbrantesKelly et al (2008) showed a relationship between abstinence and attendance at Alcoholics or Narcotics Anonymous meetings. One meeting per week independently predicted abstinence, and most attending three meetings were abstinent. Attendance during the first 6 months after discharge from hospital was associated with abstinence but this effect diminished over the period. Reference Kaminer, Burleson and BurkeKaminer et al (2008) demonstrated reduced relapse rates among young people with alcohol use disorders initially treated as in-patients who were offered aftercare comprising face-to-face contact but not those offered only telephone contact, underlining the importance for many of a continued relationship with a helping other or others.

Borderline personality disorder

Many young substance misusers display impulsive aggression, self-harm, interpersonal difficulties, affective distress and mistrust. These are core features of borderline personality disorder, and adults exhibiting them have shown benefits from sustained structured interventions such as dialectical behaviour therapy (DBT) and mentalisation-based treatment (MBT).Footnote †

In a trial involving adults with borderline personality disorder, 59% of whom reported lifetime comorbid substance use disorders, Reference McMain, Links and GnamMcMain et al (2009) compared DBT with expert general psychiatric management (case management, psychodynamic psychotherapy and medication). The core strategies used in the DBT involved balancing validation of patients’ experience with behavioural change to help them develop more effective coping strategies, thereby eliminating behavioural dyscontrol. Over a 12-month period, 25 therapists delivered 30+ sessions to 183 patients, 62% of whom completed treatment. Both treatments were associated with equivalent symptomatic improvement.

In another trial, Reference Bateman and FonagyBateman & Fonagy (2009) compared MBT with structured clinical management (crisis contact and crisis plans, pharmacotherapy, general psychiatric review, and written information about treatment). The aim of MBT, a psychodynamic treatment, was to strengthen patients’ capacity to understand their own and others’ mental states in attachment contexts, thus enabling them to address their difficulties with affect regulation, impulse control and interpersonal functioning. Of the 134 adult participants, 54% reported a substance use disorder. All were offered 18 months or 140 sessions of treatment, delivered by 11 therapists, and 75% completed at least 70 sessions in the first year. The authors concluded that focusing on psychological functions related to symptoms of borderline personality disorder (e.g. difficulty in reflecting on the mental states of oneself and others) could improve a structured programme providing generic psychological support.

Reference OldhamOldham (2009) commented that ‘often the biggest hurdle is to engage the patient in treatment – to establish a partnership that can endure so that any one of the many types of effective treatments can take hold and lead to lasting change’. He speculates that the overarching goal of therapy may be to enable rational control of emotion, to ‘teach the cortex to control the amygdala’.

Overall, these studies suggest that a coherent package of interventions, sustained over at least 6 months, involving relatively intensive contact can usefully affect symptoms often considered intractable. Taken with the MDFT data, they may also suggest a dose–response effect, so that complex psychopathology requires more elaborate and sustained interventions.

Antisocial personality disorder

Adult antisocial behaviour has close links with child and adolescent conduct disorder, which are closely entwined with substance misuse. As treatment studies of adolescent conduct disorder are rare (Reference Riggs, Winhusen and DaviesRiggs 2011), it is instructive to consider lessons from adult treatment studies.

Frank antisocial behaviour can appear intractable. However, even among those with antisocial personality disorder, observer ratings of change in therapeutic alliance predict behaviour change (Reference Polaschek and RossPolaschek 2010). A Cochrane review of psychological interventions for antisocial or dissocial personality disorder (Reference Gibbon, Duggan and StoffersGibbon 2010) found 6 months of treatment that included contingency management to have a positive effect on attendance at counselling sessions as well as on social functioning and cocaine use (assessed by an addiction severity measure). Twenty-four weeks of CBT was associated with a significant reduction in cocaine-positive urine samples. However, CBT was not associated with change in verbal or physical aggression, compared with treatment as usual (TAU). As both interventions were associated with reduced aggression, high-quality TAU could be an effective intervention. Indeed, as the participants came from mental health and forensic services within the National Health Service (Reference Davidson, Tyrer and TataDavidson 2009), it is likely that TAU was a reasonably sustained intervention emphasising engagement, general supportive interventions and problem-solving.