This review looked at treatments for ADHD in teens aged 12 to 18. The authors gathered 16 randomized trials and 1 meta-analysis with 2,668 participants. Six trials and the meta-analysis tested long-acting FDA-approved medicines in 1,752 patients. Ten trials tested psychosocial treatments in 916 participants. Studies of teens with other diagnosed mental health or developmental conditions were left out.
Both stimulant and nonstimulant medicines led to clinically significant drops in ADHD symptom scores. In the trials, scores fell further on medicine than on placebo. Evidence was stronger for stimulants and atomoxetine than for guanfacine (1 study) or clonidine (no teen studies).
Psychosocial programs had inconsistent effects on ADHD symptoms, with small to medium gains where parents rated symptoms. They helped more with school and organizing skills, such as planning tasks, using a planner and finishing homework. No study found a clear effect on teacher-rated ADHD symptoms.
Few studies were high quality. Medication trials lasted under 13 weeks and left out many kinds of patients. Psychosocial studies got a lower evidence rating, partly due to small samples and unclear methods. No study tested both treatment types together.
The authors suggest long-acting methylphenidate and amphetamine as first choices for teens, followed by atomoxetine and long-acting guanfacine, plus multicomponent psychosocial programs for school skills.
Talk to a clinician about your own situation.
The numbers Stats and what the terms mean
The statistics behind the plain-language claims above.
- Reviewed studies (cited in Chan, 2016): drop in ADHD Rating Scale scores (0 is least and 54 is most symptomatic) with stimulant and nonstimulant medicines, 14.93 to 24.60 absolute points.
- One included study (cited in Chan, 2016): methylphenidate osmotic-release oral system versus placebo, mean reduction of 47% (mean [SD], −14.93 points [10.72 points]) vs 31% (−9.58 points [9.73 points]) in the placebo group (P = .001).
- One included study (cited in Chan, 2016): methylphenidate transdermal system versus a placebo transdermal system, mean −9.96 points (95% CI, −13.39 points to −6.53 points).
- One included study (cited in Chan, 2016): extended-release mixed amphetamine salts versus placebo, −17.8 points vs −9.4 points in the placebo group (P < .001).
- One included study (cited in Chan, 2016): lisdexamfetamine versus placebo, −18.3 to −21.1 points across dose groups vs −12.8 points for placebo (P < .006).
- One included meta-analysis (cited in Chan, 2016): atomoxetine versus placebo in a subgroup aged 12 to 17 years (n = 176), −13.99 points [12.97 points] vs −6.95 points [10.07 points] for placebo (P < .001).
- One included study (cited in Chan, 2016): extended-release guanfacine versus placebo, −24.6 points vs −18.5 points for placebo (P < .001).
- Reviewed studies (cited in Chan, 2016): psychosocial treatments on parent-rated ADHD symptoms, co-occurring emotional or behavioral symptoms and interpersonal functioning, Cohen d range 0.30 to 0.69.
- CHP afterschool, STAND and HOPS studies (cited in Chan, 2016): task planning, organized actions and planner use compared with control, Cohen d range 0.51 to 5.15.
- CHP afterschool, HOPS and Survey, Question, Read, Write, Recite studies (cited in Chan, 2016): parent-reported homework problems or homework completion compared with control, between-group Cohen d range 0.42 to 0.85.
- Chan, 2016: overall evidence level, 1B for medications and 2B for psychosocial treatments (Oxford Centre for Evidence-Based Medicine criteria). What the terms mean:
- Cohen d is a standard way to size a difference between groups or over time. Roughly, 0.2 is small, 0.5 is moderate and 0.8 is large. A minus sign shows the direction of the change, not its size.
- SD (standard deviation) shows how spread out people's scores were around the average.
- 95% CI is the range the true result most likely falls in.
- P is the chance of seeing a result this large if the treatment had no real effect; smaller means less likely to be down to chance.
- n is the number of people in the group.
- Evidence level is a grade for how strong the study designs were.
Source: Chan E, Fogler J, Hammerness P. Treatment of Attention-Deficit/Hyperactivity Disorder in Adolescents. JAMA (2016). doi:10.1001/jama.2016.5453