Pharmacological treatment reduces the risk of motor vehicle crashes among men and women with ADHD

COMMENTARY ON: Chang Z, Quinn PD, Hur K, Gibbons RD, Sjölander A, Larsson H & D’Onofrio. Association between medication use for attention-deficit/hyperactivity disorder and risk of motor vehicle crashes. JAMA Psychiatry 2017;74(6):597–603.

What is already known on this topic

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder with core symptoms of inattention, hyperactivity and impulsivity,; traits that may increase the risk of injuries and motor vehicle crashes (MVCs). Injuries in early childhood predict later ADHD,¹ a diagnosis that is associated with a twofold increased risk of premature death, with accidents being the most common cause of death.² ADHD is commonly treated with pharmacotherapy that successfully targets most of the impairing core symptoms.³ In children with ADHD, pharmacological treatment also reduces the risk of injuries by up to 43%.⁴ Similarly, a Swedish study of adults with ADHD suggests that pharmacotherapy may reduce the risk of MVCs in men.⁵ It is unclear if the same association is true for women with ADHD. Furthermore, the rates of MVCs and prescriptions of ADHD medications are higher in the USA than in most European countries.⁶

Methods of the study

This US cohort study analysed data from a national commercial health insurance database, on 2 319 450 adults with ADHD (defined as an International Classification of Diseases, Ninth Revision (ICD-9) code of 314 or filling a prescription for ADHD medications, between 1 January 2005 and 31 December 2014), who were followed from the date of ADHD diagnosis or age 18 years, whichever came later, until 31 December 2014. Disenrolment was defined as zero days of medical or drug coverage in a given month. Outcome was defined as an emergency department (ED) visit for MVCs (ICD-9 codes E810-825). Associations between concurrent medication for ADHD (the same month) and the occurrence of MVCs were estimated as ORs by logistic regression. The long-term association (2 years later) was also estimated. Both analyses were done using two different statistical approaches: between-individuals analyses (ie, at population level) and to reduce confounding by indication, within-individual analyses (ie, self-controlled).

What this paper adds

• This is the first study to establish an association between use of ADHD medication and a reduced risk of MVC, in adult women with ADHD.

• The study replicates that adults with ADHD have a significantly higher risk of MVCs, than controls (OR 1.49, 95% CI 1.46 to 1.54 in men and OR 1.44, 1.41 to 1.48 in women).

• In this first US study, between-individuals analyses showed that concurrent ADHD medication reduced the risk of MVCs, in both men (OR 0.88, 0.84 to 0.93) and women (OR 0.86, 0.82 to 0.90).

• In within-individual analyses, medication had larger protective effects, reducing the risk of MVCs by 38% in men (OR 0.62, 0.56 to 0.67) and 42% in women (OR 0.58, 0.53 to 0.62).

• This is the first study of long-term associations between medications and MVCs. In within-individual analyses, medication was associated with a 34% (OR 0.66, 0.58 to 0.76) lower risk of MVCs in men and a 27% (OR 0.73, 0.64 to 0.84) lower risk in women, 2 years later.

• Assuming that the association was causal, the results suggest that 22% of MVCs in adults with ADHD could have been avoided, if patients had been on medication during the entire follow-up period.

Limitations

• Use of ADHD medication was defined as a filled prescription. Compliance with the prescribed pharmacotherapy is unknown.

• The primary outcome was defined as ED visits due to MVC. The actual number of MVCs may be underestimated, as MVCs not requiring hospital treatment are not represented in the study.

• The sample is limited to the commercially insured population and is likely not representative o the whole US population.

What next in research

• Validation in other samples as the study focused on commercially insured US patients, who might represent a higher socioeconomical group.

• Triangulation with other research designs mixing quantitative and qualitative methods.

Do these results change your practices and why?

ADHD medication should be titrated to optimal beneficial effect without causing adverse effects. Patients with ADHD responding to medication should be encouraged not to discontinue treatment, even for short periods. Patients with ADHD holding a driving license should be informed that pharmacotherapy may reduce their risk of MVCs and thereby possibly reduce the risk of premature death associated with the disorder.