Attention-deficit/hyperactivity disorder medication and risk of suicide attempt: A meta-analysis of observational studies
Liu WJ, Mao HJ, Hu LL, Song MF, Jiang HY, Zhang L.
Pharmacoepidemiol Drug Saf. 2020 Sep 1.
doi: 10.1002/pds.5110.
Commentary* by Dr. Margaret Weiss: If suicide risk can increase with antidepressants, and decrease with stimulants – one wonders what it is that is driving that risk?
ABSTRACT
Purpose: Epidemiologic findings are inconsistent regarding the association between attention-deficit/hyperactivity disorder (ADHD) medication exposure and suicide attempt in individuals with ADHD.
Methods: A systematic literature search of PubMed, Embase and Cochrane Library up to February 2020 was performed. A meta-analysis was conducted for outcomes in which a summary risk ratio (RR) was calculated when taking heterogeneity into account.
Results: Both population-level and within-individual analyzes showed that ADHD medication was associated with lower odds of suicide attempts (RR = 0.76, 95% confidence interval [CI], 0.58-1.00; P = .049 and RR = 0.69; 95% CI, 0.49-0.97; P = .049, respectively). However, the association only existed for participants who were treated with stimulants (RR = 0.72; 95% CI, 0.53-0.99; P = .042 on population-level analysis and RR = 0.75; 95% CI, 0.66-0.84; P < .001 on within-individual analysis). Furthermore, a lower risk of suicide attempts was not observed in subjects who took ADHD medication for 1 to 90 days (RR = 0.91; 95% CI, 0.74-1.13; P = .416 on within-individual analysis).
Conclusion: The results indicate that non-stimulant treatment is not associated with a higher risk of suicide attempt, but stimulant treatment is associated with a lower risk of suicide attempt.
* Les résumés scientifiques (abstracts) sont sélectionnés pour leur pertinence clinique par Dre. Margaret Weiss, Directrice de la recherche clinique, pédopsychiatrie, Cambridge Health Alliance, Université Harvard. Ses commentaires reflètent sa propre opinion. Ils ne sont ni approuvés par la CADDRA, ni nécessairement représentatifs de celle-ci.