It’s Hard to Stay In Recovery as a Teen

Treatments for adolescents with cannabis use disorder are only effective long term for a minority of those engaging in treatment. This innovative study tested whether it might help to add a continuing care intervention, – specifically for those adolescents who did not respond well to the initial intervention.


Adolescents report more cannabis use than all other illicit substance combined with 7% of US 18-year-olds meeting criteria for cannabis dependence (i.e. moderate/severe cannabis use disorder) in the past year. Adolescents with cannabis use disorder have historically achieved low rates of abstinence (< 25% at one year) despite high levels of engagement in substance use treatment designed specifically for adolescents. Extending treatment for those not initially responsive to treatment, including interventions designed to address the social group challenges more relevant to adolescents, may allow higher abstinence success rates.


This was an outpatient-based, randomized adaptive treatment study enrolling 161 adolescents ages 13-18 diagnosed with cannabis use disorder (based on DSM-IV criteria).


This was an outpatient-based, randomized adaptive treatment study enrolling 161 adolescents ages 13-18 diagnosed with cannabis use disorder (based on DSM-IV criteria). All participants received a seven-session weekly motivational enhancement (two individualized sessions) and cognitive behavioral therapy (CBT; 5 group sessions) intervention. By week seven, 50% of the participants had achieved abstinence.  The 80 participants that did not achieve abstinence by week seven were then randomized to receive 10 weeks of adaptive treatment: the participants were randomized to receive either individualized enhanced CBT or an Adolescent Community Reinforcement Approach (ACRA) intervention. Substance use was monitored by random urinalysis throughout the study, and abstinence was defined as abstinent from all substances tested (cannabis, cocaine, opiates, amphetamines).


Half of the adolescents achieved abstinence after seven weeks of outpatient substance use group therapy.  The therapy combined motivational enhancement and cognitive behavioral therapy, two commonly used interventions for substance use disorders, with 50% of participants achieving abstinence at week seven (completion of therapy).

The majority of those achieving abstinence at week seven had relapsed by week 17.  Only 30% of the participants abstinent at week seven were abstinent at week 17 (representing only 15% of the original sample).  This group did not receive additional substance use treatment beyond week seven. Those showing initial abstinence and subsequent relapse, however, had lower drug use and higher attendance in therapy than those who did not respond to the initial seven-week intervention.

Less than 40% of initial non-responders completed the subsequent treatment phase, and only 27% of these achieved abstinence by week 17 of the study.  Participants that had not achieved abstinence by week seven received an additional ten weeks of therapy intervention but only 26 of the 80 who went on to this treatment phase completed the ten sessions, and only 7 of these 26 participants achieved abstinence at week 17.

There was no significant difference between the two treatment arms, one of which was designed specifically for adolescents.  The two types of interventions used for those not obtaining abstinence by week 7 were individualized enhanced cognitive behavioral therapy (eCBT) and an Adolescent Community Reinforcement Approach (ACRA).  There was no significant difference in either retention or abstinence rates for these two treatment arms (5 out of 13 for ACRA vs 2 out of 13 for eCBT).


Adolescents suffering from cannabis use disorders have low treatment response rates in general to therapy-based outpatient treatment (including motivational enhancement and cognitive behavior strategies as well as treatment designed specifically for adolescents and their families). The vast majority of those achieving abstinence early in the study had used cannabis during or by week 17. Even extending treatment by an additional ten weeks had a small effect on increasing abstinence. These findings are generally in line with a series of large studies on adolescent cannabis use (Cannabis Youth Treatment summary).

The starkness of these conclusions is offset by the study’s focus on abstinence (from all substances). Those relapsing in this study did have decreased cannabis use although the study was not designed to analyze this finding in detail. Abstinence may be a less common goal for adolescents treated in outpatient settings (i.e., who have less severe substance-related problems) than controlled use, given the social pressures around using and exposure to those actively using substances. This study was not designed to look at outcomes other than complete abstinence, although simply reducing use may represent a significant positive outcome.

  1. The study was limited to the northeastern United States and 82% of the participants were male so the study finding may not be as applicable to females or other regions
  2. A large number of participants did not complete treatment (5 dropped out of the phase I without further engagement and 49 out of phase II, totaling 54 out of 161 or 34%). This suggests that treatment adoption rates might be low if implemented more generally.
  3. There is no control arm for the second phase of the study. Thus one cannot determine whether it was the enhanced treatment per se or the prolonged engagement in treatment that led to increased abstinence relative to the seven week treatment course.
  4. Given the high (>60%) relapse rates for the initial responders at week 17, it is likely that the treatment was under-dosed. The fact that the end point of the study was week 17, it could be that none of the three treatments resulted in significant sustained abstinence if they had included a one year follow-up.  Nonetheless, the three different theoretical approaches appear, in this study, to have similar outcomes despite focusing on different types of interventions.  Also, note that smoking cannabis is a less functional outcome than engagement in school, decreased conflict at home, and involvement in illegal activities, which may have been decreased even in the presence of some cannabis use at week 7 or week 17.
  5. The final outcome of this study was abstinence, not just of cannabis but of all substances. This primary outcome may not be in line with the adolescents’ goals and there was no survey on goals for treatment nor an analysis of the number of participants still meeting criteria for cannabis use disorder by the end of treatment.


  • For individuals & families seeking recovery: A goal of abstinence from cannabis use after outpatient substance use treatment is achievable (50%) but hard to sustain (only 30% at 4 months) after short-term treatment.  This suggests that longer-term engagement in treatment and other recovery support services may be necessary for more sustained remission.  Clarifying whether the goal of treatment is abstinence or controlled use may improve the chances of an individual’s success with treatment and allow for more efficient and effective tailoring of treatment. Note that imaging studies suggest that there are brain morphological changes resulting from adolescent cannabis use so abstinence may be recommended at least through the end of adolescence and brain development.
  • For scientists:  Studies of adolescents with cannabis use disorder should be informed by the participants’ longer-term goal of abstinence or reduced use.   Interventions focused on achieving abstinence, as is the focus of this study, show that the gains obtained after short-term treatment seem to not be sustained.  Potentially more attractive and engaging formats might be tested among adolescents such as those using mobile technology platforms (which can include group therapy) rather than adult-derived in-person weekly outpatient appointments as adherence to that form of treatment is generally low among youth.
  • For policy makers: The data from this study might suggest that long term outpatient treatment for adolescents with cannabis use disorder might not optimally effective until the barriers to engagement in treatment are better understood. However, decreasing the cost of quantitative cannabinoid metabolite measurements in urine might help the scientific field broaden outcome goals from abstinence to decreased use that may allow the effectiveness of short-term treatments to be assessed more accurately.
  • For treatment professionals and treatment systems: Designing interventions to engage adolescents in treatment over time may sustain the benefits of more intensive short term (7-17 session) treatment for cannabis use disorders.  This will likely require mobile technology platforms and a social context that is re-inforcing.  In addition, broadening objective measures beyond merely negative urine analyses drug monitoring may increase engagement in treatment.  For example, using quantitative urine drug monitoring tests (which are currently too expense for routine use) would allow participants to see evidence of progress as use is decreased and thus help sustain commitment to treatment.


Kaminer Y, Ohannessian CM, Burke RH 2017 Adolescents with cannabis use disorders: Adaptive Treatment for poor responders.  Addictive Behaviors 70: 102-16