A newly published meta-analysis of 57 studies encompassing almost a third of a million participants has uncovered a very strong association between ADHD and suicide, a strong association with suicidal ideation, and a small-to-medium association with suicide attempts.

The population examined included children, adolescents, and adults. Only persons formally diagnosed were considered to have ADHD. Studies that included self-injuries without suicidal intent were excluded. Most of the studies focused on European and American populations, with one in six from other locations, mostly Asian.

The most striking result was for actual suicides. The odds ratio (OR) for four datasets encompassing roughly one hundred forty thousand participants was 6.69 (95% CI 3.24 to 17.39, p < .0001). As a frame of reference, an OR of 1.5 is a small effect size, 2.5 a medium one, and 4.3 a large one. That means the effect size in this case is very large.

For suicidal ideation, 23 datasets with a combined total of just over 73,000 participants produced a medium-to-large OR of 3.5 (95% CI 2.94 to 4.25, p < .0001). In three datasets with more than nine thousand participants that adjusted for confounders, the adjusted OR was 4.5 (95% CI 1.72 to 11.63, p < .0001), indicating a large effect size.

For suicide attempts, 44 datasets encompassing over 228,000 participants produced an OR of 2.4 (95% CI 1.64 to 3.43, p < .0001). In six datasets with over 65,000 participants that adjusted for confounders, the adjusted OR dropped to 2.1 (95% CI 1.27 to 3.47, p = .005).

There was no evidence of publication bias for studies on suicides or suicidal ideation, but significant evidence of bias for studies on suicide attempts (Egger’s p = .03). This means that studies with positive findings were more likely to be published than negative studies.

There was, however, strong statistical evidence for differences between studies in the size of their ORs. This indicates that the pooled OR cannot summarize results from all datasets and more work is needed to clarify why the ORs differ among studies.

The authors appropriately caution that their meta-analysis is “not informative on cause-effect relationships,” but offer as a hypothesis that ADHD contributes to suicidal spectrum behaviors (SSBs) through “Impulsivity, a core symptom of ADHD, along with impaired decision-making and risk taking, that characterize a number of individuals with ADHD … Additionally, a sizeable portion of individuals with ADHD present with deficits in executive functions. As executive functions are implicated in the regulation of impulse control and emotions, executive dysfunctions may contribute to SSBs.”

In view of the large to very large effect sizes for suicide and suicidal ideation, the authors advise: “Awareness of this association should prompt practitioners to systematically screen for SSBs in patients with ADHD at the first assessment and at each follow-up, which in turns should contribute to decrease the risk of SSBs. This is particularly noteworthy considering that questionnaires/scales commonly used to screen/assess ADHD symptoms generally do not include suicide related items.”

REFERENCES

Septier M, Stordeur C, Zhang J, Delorme R, Cortese S, Association between suicidal spectrum behaviors and Attention-Deficit/Hyperactivity Disorder: A systematic review and meta-analysis, Neuroscience and Biobehavioral Reviews (2019), https://doi.org/10.1016/j.neubiorev.2019.05.022.