Nationwide population study finds high relative risk of traffic crashes among the elderly with ADHD, but with very low frequency, muddling interpretation of the results

Researchers from the Swedish Department of Global Public Health, the Swedish Transport Agency, and the Swedish National Road and Transport Research Institute collaborated in a nationwide population study of motor vehicle crashes among the elderly, defined as 65 and older.

They availed themselves of the country’s all-encompassing national registers to identify the anonymized records of all such drivers from 2011 through 2016. That enabled them to compare crash records of those with known driving-impairing conditions with matched drivers who had no record of such conditions.

They looked only at road traffic crashes that resulted in injury to the driver or a passenger. For anyone with multiple crash records, they only looked at the first.

This was a case-control study, with two controls matched to each case wherever possible. For every case of a 65 or older driver involved in an injurious crash, the team randomly matched two individual controls by sex, birth year, municipality of residence, and other medical conditions. Place of residence was used to distinguish residents of large cities, who would tend to drive less frequently and in denser traffic, from those in small towns and rural areas. To minimize controls that never drive, only those with a driver's license and car were considered.

Of the thirteen medical conditions examined, elderly drivers with “ADHD, autism spectrum disorder, and similar conditions” had by far the highest odds of being in crashes that resulted in injury – at almost three times the rate of those without those conditions.

But note carefully the serious limitations in the data:

·        ADHD was bundled in with autism spectrum disorder and “similar conditions,” making an unalloyed evaluation impossible.

·        Out of a total of 13,701 crashes, only 26 involved any of these conditions.

·        Because of the small number, the two-for-one matching broke down completely. Only 17 matched controls could be found, less than a third of the target of 52.

·        That means that despite a nationwide sample involving over 40,000 cases and controls, the sample size for “ADHD, autism spectrum disorder and similar conditions” was only 43.

The authors noted that while the high odds ratio “for ADHD, autism spectrum disorder, and similar conditions, are in line with previous studies on young adult drivers and adult drivers …in this recent cohort of older, Swedish adults, such conditions are very uncommon compared to younger adults, suggesting likely under-diagnosis. Hence, the results should be interpreted with caution.”

Marie Skydiving, Åsa Forsman, Tania DukicWillstrand, LucieLaflamme, JetteMöller, “Medical impairment and road traffic crashes among older drivers in Sweden – A national, population-based, case-control study,” AccidentAnalysis and Prevention 163 (2021) 106434, published online,