Driving Fitness in Different Forms of Dementia: An Update

Max Toepper, PhD; Michael Falkenstein, MD


J Am Geriatr Soc. 2019;67(10):2186-2192. 

In This Article


Our database search resulted in 4745 titles and abstracts in total. After removing duplicates, 1286 articles remained. We screened possibly eligible and relevant abstracts and full texts of the remaining articles, with a focus on the most relevant and recent publications. The vast majority of articles was excluded because these studies had nothing to do with driving, were not designed to assess driving fitness, or made only indirect assumptions. Eventually, we included 53 articles of the database search results (Supplementary Table S1). In addition, we included 19 references from other sources involving diagnostic and clinical features of the different dementia syndromes (n = 10), driving skills in healthy older people (n = 3), negative consequences of driving cessation (n = 4), and methodological guidelines (n = 2).

Driving Fitness and Dementia Severity

Results show that dementia severity, cognitive and functional impairments, as well as greater increases in dementia severity and cognitive and functional impairments over time are significant predictors of driving cessation.[14] A tool that is commonly used for assessing the severity of dementia is the CDR.[15] While higher CDR levels 2 and 3 (moderate and severe dementia) are incompatible with driving, subjects with very mild (CDR score = 0.5) and mild (CDR score = 1) dementia are candidates whose fitness to drive needs to be studied in more detail, preferably with an on-road test.

For drivers with MCI, our search results suggest lower safety ratings compared to healthy older drivers in on-road tests[16] and twice as many driving errors in simulated driving.[17] MCI patients show suboptimal lane control, have difficulties with left-sided turns under complex traffic conditions, and seem to be more distracted by smartphones while driving than healthy drivers.[17,18] Noteworthy, this particularly applies to drivers with multidomain MCI.[17] Moreover, depressive symptoms, which are frequent in older adults, appear to have a greater negative impact on driving performance in drivers with MCI than in cognitively healthy drivers.[19] Driving difficulties in MCI may be reflected by increased avoidance behavior[20] and by driving cessation within 3 years after diagnosis.[21] In a study by Fuermaier and colleagues, a third of the patients failed an on-road assessment (no baseline fail rate in healthy older drivers reported).[22]

Taken together, there is consensus that people with severe or moderate dementia are no longer fit to drive.[23,24] People with MCI or mild dementia may still be able to drive under certain circumstances, which may be at least partly dependent on the type of the respective dementia syndrome.

Driving Fitness and Type of Dementia

Alzheimer Disease Dementia. Driving fitness is scientifically best investigated in ADD.[25] Particularly in complex traffic situations, drivers with ADD show more driving self-restriction due to impaired attention and executive functioning.[26] Search results show that ADD patients make significantly more driving errors than healthy older drivers.[27,28] In particular, they have difficulties in remembering routes or traffic rules[28,29] and in identifying landmarks or traffic signs.[30] They misjudge the distance to other vehicles, make judgment errors at intersections, or act/react too slowly.[27] Rizzo and colleagues examined 39 older drivers in a driving simulator, 21 of them with ADD.[31] Of the patients, 29% were involved in crashes compared to none in the group of healthy controls. Moreover, ADD patients experienced near accidents more than twice as likely as healthy drivers (74% vs 35%). Thereby, the best predictors of accident risk were impairments in visual-spatial attention and spatial thinking. A survey of relatives of ADD patients revealed changes in driving behavior in 58% of the patients, with orientation deficits being the most striking dysfunctions.[32] Nearly 20% were responsible for at least one accident since disease onset. Overall, accident risk in Alzheimer patients is more than four times higher than in healthy older people[33] and increases with the time of driving since disease onset.[34] In a heterogeneous sample including drivers with probable and suspected ADD, more than half of the patients did not pass a practical driving test (55.5%).[35]

Noteworthy, this may not necessarily apply to prodromal, very mild, or mild stages of the disease. The prodromal stage of ADD is defined as amnestic MCI (aMCI), which is typically characterized by an isolated episodic memory dysfunction.[12] Results suggest that drivers with aMCI rather show suboptimal performances than a definite impairment of driving fitness.[27,36]

For very mild ADD, meta-analytical evidence suggests on-road fail rates of about 13% compared to 1.6% in healthy controls,[27] which might be partly due to inappropriate behavior at intersections (ie, stopping at intersections).[27] In the case of mild ADD, already 33% of the patients fail an on-road test,[27] and single studies reported even higher fail rates of 50.6% and 58% in patients with very mild and mild ADD, compared to 4.4% and 11%, respectively, in healthy controls.[28,37] Drivers with very mild and mild ADD show impaired on-road performances on the operational, tactical, and visual level.[37] They make more turning errors than healthy drivers, lose orientation, and endanger road safety with their maneuvers.[29] Carr and colleagues reported that drivers with very mild and mild ADD show tendentially more at-fault crashes than healthy drivers and more accidents resulting in personal injuries.[38]

Driving Fitness in Non-Alzheimer Dementias. In non-Alzheimer dementias, there is much less empirical evidence regarding fitness to drive than in ADD. Across different non-Alzheimer etiologies, however, recent on-road driving studies reported fail rates of 59%[39] and 48%[37] for very mild to mild disease stages, compared to 11% in healthy controls.

Vascular dementia: The few studies that investigated driving skills in patients with VaD suggest that drivers with VaD exhibit severe driving difficulties. A study by Fitten and colleagues revealed that patients with multi-infarct dementia show poorer on-road driving skills than older people with diabetes, healthy older people, or healthy young drivers.[40] Thereby, driving scores were inversely associated with cognitive skills, number of collisions, and violations per 1000 miles driven. Moreover, our literature review resulted in one case report highlighting the difficulties of a family in the interaction with a driving family member with multi-infarct dementia.[41] Two more recent studies revealed that about 70% of drivers with very mild and mild VaD fail an on-road driving test,[37,39] compared to 11% in healthy controls. Patients with a single stroke also show poor driving performances;[42] meta-analytical evidence suggests that 46% of these patients do not pass an on-road test.[43]

Frontotemporal dementia: In patients with FTD,[44] fitness to drive is also severely impaired. Between 50%[37] and 60%[39] of these patients do not pass an on-road test in very mild to mild disease stages, compared to 11% in healthy controls. Next to cognitive deficits in executive functions, patients with FTD show early behavioral symptoms, such as disinhibition, impulsivity, antisocial behavior, or increased risk tolerance.[44] All of these factors can severely impair driving fitness as being reflected by hit-and-run crashes, ignoring red lights, speeding, and overlooking pedestrians at intersections.[45,46] In a driving simulator study, drivers with FTD received more speeding tickets, ran more stop signs, and were involved in more off-road crashes and collisions compared to their healthy peers.[47] In two surveys of relatives of FTD patients on potential changes in the patients' driving behavior, changes were reported for about 90% of the patients.[32,48] In particular, an aggressive, risk-taking driving style was described, including frequent violations of traffic regulations, speeding, inappropriate driving behavior and driving against the will of the family.[32] Moreover, FTD patients were reported to frequently misjudge distances, be easily distracted, and ignore road signs or red traffic lights.[48] A large number of patients caused at least one accident since onset of the disease,[32,48] mostly because traffic rules were ignored. Some of these accidents resulted in personal injuries;[32] nevertheless, the majority of patients saw no reason to give up driving.

Noteworthy, most driving studies included different FTD subtypes and did not differentiate between these subtypes, so that the above results may apply to all FTD variants, such as the behavioral variant as well as the semantic and nonfluent variants of primary progressive aphasia (PPA). Drivers with the semantic FTD variant, however, showed poor traffic sign knowledge in a neuropsychological study and even worse performances than drivers with ADD.[49]

Dementia with Lewy bodies: In DLB, driving fitness may also be impaired already in early disease stages, although the number of studies identified by our search criteria is small. Yamin and colleagues reported that patients with mild DLB showed poorer performances than healthy controls in all parameters of a driving simulator ride, including an increased number of collisions.[50] Moreover, on-road fail rates of 40%[39] and 35%[37] are reported in very mild to mild disease stages, compared to 11% in healthy controls.

Parkinson disease dementia: Empirical studies on driving fitness in PDD are also sparse. Driving impairments in Parkinson patients without dementia, however, are frequent and depend on disease duration, motor symptoms, and cognitive dysfunctions.[51–53] A progression to dementia in these patients may further impair driving skills. Singh and colleagues explored driving problems in 154 Parkinson patients by a combination of clinical tests, reaction time measures, and an on-road driving test.[54] Results show that only 18% of patients with PDD were suited to drive (no baseline rates for healthy drivers reported).