University of Athens,

Driving Behavior in MCI and

2nd Department of
Neurology, “Attikon”
mild Alzheimers Dementia
University General
Hospital, Athens,
Greece
Sokratis G. Papageorgiou, MD, PhD
National Technical Associate Professor of Neurology
University of Athens,
Department of Cognitive Disorders/Dementia Unit
Transportation 2nd Department of Neurology, University
Planning and
Engineering, Athens, General Hospital “ATTIKON“
Greece Medical School, University of Athens
To drive or not to drive

PREDICTIVE FACTORS for accident

• Age = 85
• Gender = male
• MCI or mild dementia
• Previous accidents ( 2, last 1 year)
• Alcohol consumption
 Introduction
 Life loss in elderly represent the 26% of all life losses from car
accidents in the EU (Eurostat, 2014)

 Cognitive functions that contribute to a successful driving are

compromised in ~25% of the elderly population

(In people aged >65: 15% have MCI, 8% have dementia).

Diverse etiologies: mostly Degenerative or Vascular : (Alzheimer’s disease,
Vascular dementia, PD, DLB,)…

 Taking into account that the % of the elderly in society is increasing

while at the same time the level of motorization also increases (Yannis
et al, 2010), the need to investigate the impact of the above conditions

on driver performance becomes critical.

3
 Cognitive functions critical for safe driving
• Attention
o quick perception of the environment
• Visuospatial skills
o positioning of the car on the road
o manoeuvring the car in lane changes
o judging distances and speed

• Executive functions
o process multiple simultaneous
environmental cues
o predicting the development of traffic
situations
o make rapid, accurate and safe decisions

• Memory
o journey planning
o adapting behaviour
o sign recognition, memorization
4
(adapted form Reger et al. 2004)
 Alzheimer's disease and driving
• AD patients are 2.5 to 4.7 times more
likely to be involved in a car crash than age-
matched controls
(Brown and Ott 2004; Dobbs et al. 2002; Ernst et al. 2010; Withaar et al. 2000,
Brorsson, 1989; Massie & Campbell, 1993; Tuokko et al., 1995)

• But ~ 50% of patients with AD continue

driving for at least three years after their
initial diagnosis (Adler and Kuskowski 2003; Seiler et al.2012, Johansson and
Lundberg, 1997; Dubinsky et al., 1992; Rizzo et al., 2001; Charlton et al., 2004; Uc et al., 2005; Uc et al.,
2006; Ott 2008; Ernst et al. 2010)

• 87% of patients seen with dementia (MMSE=

21,8), were driving in a study from northern

Italy (Mauri et al, 2014)

J Am Geriatric Society 2015

(WHIMS-ECHO) mean age 83.7 +/- 3.5

were still driving:

MCI : 60%

Dementia: 40%
Am J Geriatr Psychiatry. 2017. Chee JN et. al.

9
2 studies
MVCollision
• RR= 4
• RR=1

7 studies
Fail Road Test
• RR=10
 Alzheimer's disease and driving
Patients with dementia at a moderate or severe stage
(CDR >1) are incapable of driving.

However, in the earlier stages of the

disease
not all patients are incapable of driving,
Up to a 76% of patients with
mild AD are still able to pass an on-road
driving test Duchek JM et al. J Gerontol Psychol
Sci 1998; Ott BR et al. Neurology
2008, Brown and Ott 2004; Ernst et
al., 2010; Withaar et al. 2000, Iverson
AD is a progressive disease and it is many times difficult DJ et al. Neurology 2010 (update of
the AAN 2000 practice parameter on
to decide: driving and dementia)
which is the proper time for dissuading a patient from
driving? 8
 Methodology used nowadays
1. Surveys and Questionnaires on Stated Behaviour
Information is asked from the drivers AND from their close
relatives.

2. On-road studies
• fitness to drive control (by an instructor)

• instrumented vehicles in real traffic or test site conditions

(naturalistic driving ).

3. Driving Simulator Experiments

• Advantages: Driving conditions are identical for all drivers
Exploration of any possible driving scenario like Unexpected incidents!!

• Limitation: Simulator sickness (~25% of persons)

Alzheimer's disease and driving errors
Although, basic control abilities of the vehicle are normal

Drivers with mild AD made significantly:

• more incorrect turns

• got lost more often

• more at-fault safety errors

• less likely to use a safety belt

• Drove smaller and fewer distances, at lower traffic roads

• stayed closer at home and had a preference for daylight driving

• lower driving speed

In-vehicle technology continuous registration of driving parameters (Eby et al.,2012)

On-road assessment (Uc et al., 2004)
Alzheimer's disease and driving errors

on-road driving
assessment.

most of the dangerous

actions in patients
with dementia
occurred

while driving
straight
condition.
 Predictors of driving ability in patients with AD

Some studies have shown that

performance on various cognitive
tests is associated with driving
competence in patients with AD:

1. visuospatial
2. attentional
3. executive
4. memory

• However, other studies have found no

association of cognitive testing with
driving competence in patients with
dementia
(Brown et al., 2005; De Raedt et al., 2001; Paccalin et al.,
2005; Uc et al., 2004; Whelihan et al, 2005, Brown et Ott,
2004; Elkin-Frankston et al., 2003; Ott et al., 2003; Ott et al.,
2008; Reger et al., 2004; Szlyk et al., 2002; Uc et al., 2005;
Grace et al., 2005; Bieliauskas et al., 1998; Brown and Ott,
2004; Molnar et al., 2006; Asimakopoulos et al., 2012;
Etienne et al., 2013)
Mild Cognitive Impairment and driving
A controversial issue

Driving Ability Driving Ability

• Wadley et al., 2009 • Snellgrove et al., 2005

on-road on-road (50% of MCI
failed the on-road test)

• Devlin et al., 2012 • Kawano et al., 2012

simulator simulator

• Jeong et al., 2012 • O’ Connor et al., 2010

questionnaire questionnaire
Our data in patients with
Cognitive Disorders
 A large driving simulator experiment
 on driving behaviour including driver distraction
(fall 2013 – fall 2015)
 Neurologists, Neuropsychologists,
Transportation Engineers
National Technical
 Assessments University of Athens,
Department of
- Medical, Neurological & Ophthalmological Transportation
Planning and
evaluation (~2 hours) Engineering, Athens,
Greece
- Neuropsychological evaluation (~2,5 hours) and
Questionnaire on driving habits (~20 minutes) University of Athens,
Department of
- Driving simulation evaluation (~1,5 hour) Psychology, Athens,
Greece

 Sample size: 225 persons fully examined, University of Athens,

2nd Department of
 154 persons > 55 years old Neurology, “Attikon”
University General
Hospital, Athens,
Greece
(MCI = 59, AD= 25, PD= 25, Normal Controls= 45)
co-funded by the Greek Research Secretariat and the European
Commission http://www.nrso.ntua.gr/driverbrain/ 16
 Inclusion Criteria

• Valid driving license

• Regular driving, without frequent accidents
• CDR: 0 to 1
• No history of psychosis, or other Psychiatric or Neurological
disease
• No dizziness, nausea while driving, either as a driver or as
a passenger
• No alcoholism or drug addiction
• No visual disturbance preventing them from driving safely
“Driving at the simulator” assessment
• 1 practice drive (usually 15-20 minutes)

• 1 rural route (2,1km long, single carriageway,

3m lane width)

• 1 urban route (1,7km long, at its bigger part

dual carriageway, 3.5m lane width)

• 2 traffic scenarios for each route:

• QL: Moderate traffic conditions (Q=300
vehicles/hour)
• QH: High traffic conditions (Q=600
vehicles/hour)

• 2 unexpected incidents at each trial:

• Sudden appearance of an animal (deer or
donkey) on the roadway
• Sudden appearance of a child chasing a ball
on the roadway or of a car suddenly getting
out of a parking position.

• 3 distraction conditions
“Driving at the simulator” assessment
• 1 practice drive (usually 15-20 minutes)

• 1 rural route (2,1km long, single carriageway,

3m lane width)

• 1 urban route (1,7km long, at its bigger part

dual carriageway, 3.5m lane width)

• 2 traffic scenarios for each route:

• QL: Moderate traffic conditions (Q=300
vehicles/hour)
• QH: High traffic conditions (Q=600
vehicles/hour)

• 2 unexpected incidents at each trial:

• Sudden appearance of an animal (deer or
donkey) on the roadway
• Sudden appearance of a child chasing a ball
on the roadway or of a car suddenly getting
out of a parking position.

• 3 distraction conditions
tested for the 1st time
 3 distraction conditions
• Undistracted condition

• Conversation with a
passenger

• Conversation on the
Mobile phone
True?
 Driving Simulator - Quantitative Measures
1. average driving speed (km/h)
2. speed variation (variation of average speed)
3. Average wheel position
4. wheel position variation (variation of wheel steering angle in degrees)
5. lateral position (average vehicle distance from the central road axis in meters)
6. lateral position variation (the standard deviation of lateral position)
7. average headway (average time to cover the distance from other vehicles in meters)
8. headway variation (the standard deviation of headway)
9. Sudden brakes
10. Engine Stops
11. Speed limit violations
12. Hits of side bars
13. number of crashes
14. reaction time in unexpected incidents (in milliseconds)

•Urban Driving:
•Rural Driving:

With and Without DISTRACTION

 Driving behaviour of drivers with Mild Cognitive Impairment and Alzheimer’s Disease: A Driving Simulator Study
Dimosthenis Pavlou, Eleonora Papadimitriou, Constantinos Antoniou, Panagiotis Papantoniou, George Yannis, John Golias, Sokratis G. Papageorgiou

Proceedings of Transportation Research Board (TRB) 2015, Washington DC

Rural Road Urban Road

Parameter Estimates Low Traffic Volume High Traffic Volume Low Traffic Volume High Traffic Volume
Dependent Std. Std. Std. Std.
B t Sig. B t Sig. t Sig. B t Sig.
Variable Error Error Error Error
47,9
Intercept 1,2 39,7 ,000** 45,3 1,0 45,6 ,000** 33,7 0,9 37,5 ,000** 30,4 0,7 43,9 ,000**

Mean speed MCI -6,1 2,0 -3,1 ,003** -6,2 1,6 -3,9 ,000** -4,9 1,7 -2,8 ,007** -3,7 1,3 -2,8 ,007**
(km/h) AD -14,0 2,3 -6,0 ,000** -13,4 1,9 -7,0 ,000** -4,4 2,4 -1,8 ,079* -4,6 1,9 -2,5 ,017**
Control 0 0 0 0
Intercept 46,6 4,8 9,8 ,000** 22,4 4,7 4,7 ,000** 48,6 5,2 9,4 ,000** 23,8 2,3 10,3 ,000**

Mean MCI 12,4 7,8 1,6 ,120 12,0 7,7 1,6 ,123 7,5 9,9 0,8 ,460 12,0 4,5 2,7 ,009**

headway AD 40,4 9,2 4,4 ,000**

51,3 9,1 5,6 ,000**
4,3 13,9 0,3 ,760 7,3 6,3 1,2 ,250
(sec)

Control 0 0 0 0
923,1
Intercept 154,0 6,0 ,000** 996,3 159,1 6,3 ,000** 1294,5 66,6 19,4 ,000** 1284,2 63,0 20,4 ,000**
Reaction MCI 481,9 250,7 1,9 ,059* 532,6 259,1 2,1 ,043** 198,1 116,0 1,7 ,092* 139,1 121,4 1,2 ,260
time (millisec) AD 580,3 296,7 2,0 ,054* 446,4 266,7 1,7 ,097* 296,2 165,7 1,8 ,078* 209,7 170,5 1,2 ,220
Control 0 0 0 0
Results-Repeated measures GLM - Longitudinal control measures
• AD and MCI patients drive at significantly lower mean speed compared to healthy drivers, both at low and high traffic volumes, in both rural and
urban areas
• AD drivers’ speed is significantly lower than MCI, in both traffic volumes in rural road, whereas in urban road the two groups have almost the same
speed
• AD patients have significantly longer mean headway compared to healthy drivers (and compared to MCI group) at both traffic volumes in rural
area
• MCI patients seem to have significantly longer mean headway compared to healthy drivers only at high traffic volume in urban road
 Which are the effects of driver distraction and brain pathologies on reaction time and accident risk?
Dimosthenis Pavlou, Panagiotis Papantoniou, Eleonora Papadimitriou, Sophia Vardaki, George Yannis, John Golias, Sokratis G. P apageorgiou

Advances in Transportation Studies an international Journal 2016 Special Issue, Vol. 1

,0001 ,0001

• In rural area AD group had the worst reaction times (more than 40% worse reaction
times than the control group)
• Mobile phone use seemed to have a significant effect on reaction time.
• Conversing with passenger didn’t seem to have an important effect on reaction
time in all examined groups
• A statistical analysis was carried out by means of mixed generalized linear modelling and the results
indicated significant differences between the driving performance of healthy drivers and patients.
 Which are the effects of driver distraction and brain pathologies on reaction time and accident risk?
Dimosthenis Pavlou, Panagiotis Papantoniou, Eleonora Papadimitriou, Sophia Vardaki, George Yannis, John Golias, Sokratis G. Papageorgiou

Advances in Transportation Studies an international Journal 2016 Special Issue, Vol. 1

,0001

,0001
,0001
,0001

• AD drivers had in all conditions the higher accident probability, and especially
when conversing on the mobile phone (more than 60%)
• Conversation with passenger didn’t increase the possibility of causing an
accident
• In urban area the differences between the groups were approximately the same
with the rural area
• A statistical analysis was carried out by means of mixed generalized linear modelling and the results indicated significant
differences between the driving performance of healthy drivers and patients.
 Driving behavior in AD and MCI

Compensatory behaviour Cognitive

Driving deficits
deficits
Pathology Speed Headway Lateral position LP variability
Reaction Reaction
time timerisk
- Accident
Healthy
MCI
AD

Road and traffic environment complexity

AD predictors
• Severity of Neuropsychiatric symptoms (and particularly of Depression
and Apathy) is correlated with driving behavior in mild AD.
Association between motor ability tasks and driving parameters in healthy, patients with
mild cognitive impairment (MCI) and mild Alzheimer’s disease (AD)
Andronas et. al. Athens-Springfield 2016
Variables Controls MCI AD ANOVA Post-Hoc
M SD M SD M SD F p
Age 61.7 8.4 69.9 9.6 74.5 7.6 17.5 .000 HC<MCI**, HC<AD**
Education 15.5 3.2 13.1 3.2 16.8 26.3 5.5 .006
Driving Exp. 35.2 9.2 42.8 9.3 42.6 9.7 49.9 .000 HC<MCI*, HC<AD*
MMSE 29.3 .8 27.6 1.9 22.7 4.2 .69 .50 HC>MCI>AD**

25
MCI predictors
 Number of Accidents in MCI
Predictors:
(1st level) general cognitive functioning (MMSE)
(2nd level) visuospatial memory (BVMT_Recognition, β=-.40, p=.056)
and speed of attention (UFV_1, β=.48, p=.027)

The model explained 77.3% of the variance in number of accidents

R²=.773, F(3,10)=11.35, p=.001

• In the cognitively intact group the same regression model did not
contribute to the prediction of number of crashes

R²=.279, F(3,10)=1.29, p=.330

Normal Group (Mean=.43, SD=.65) vs MCI Group(Mean=.56, SD=.81)

t(28)=.49, p=.25
Papageorgiou et al, EFNS- EAN Congress, 2014
• Depression and Irritability are correlated
with number of errors in driving

• In Urban condition, Depression, Anxiety

and Sleep disorders are related to Engine
stops.

• In High Traffic Depression and Anxiety are

related to increased Reaction Time in
unexpected incidents.
There was a unique contribution of DEPRESSIVE SYMPTOMS (PHQ-9) on
predicting various indexes of driving performance only in the MCI group
even after controlling for the role of neuropsychological measures and sleep disturbances

Outcome measure PHQ-9 contribution

β t p R2 Overall
Model
Lateral Position .60 2.89 .011* .635
Variation
Average Speed .62 2.52 .023* .490
Average Headway -.61 2.43 .028* .468
Headway Variation -.59 2.34 .034* .463
No. of accidents .70 2.84 .012* .485
Hits of Side Bars .39 2.11 .052 .705
Speed Limit Violation .61 2.84 .012* .613
Average Wheel Position -.59 2.50 .025* .524
 Mild Cognitive Impairment and driving: Does in-vehicle distraction affect driving performance?
Ion N. Beratis, Dimosthenis Pavlou, Eleonora Papadimitriou, Nikolaos Andronas, Dionysia Kontaxopoulou, Stella Fragkiadaki, George Yannis, S. G.
Papageorgiou

Average Speed Submitted, (2016) • The mixed ANOVA revealed a

greater effect of distraction on
Accident Probability
MCI patients.

• use of mobile phone

induced a more pronounced
impact on reaction time and
accident probability in the
group of patients, as
compared to healthy controls.

Reaction Time • Also, a greater negative effect

Average Speed of “conversing with
Variability passenger” was observed in
the group of drivers with MCI,
but of a lesser extent than in
the case of the mobile-phone
condition.
• despite the effort of the
drivers with MCI to apply a
compensatory strategy by
reducing significantly their
speed.
 sleep disturbances were correlated with driving
variables, only in MCI patients

*
*

* *

sleepiness - insomnia

Beratis et al, 1st EAN Congress, Berlin, 2015

2/15/2019
 Conclusions
A. While 30-50% of patients with mild AD may be still able to drive,
their driving abilities are compromised
B. Driving abilities are also compromised even in MCI patients.
C. Distraction during driving plays a critical role in driving behavior
in patients with MCI and AD.
D. Useful predictors of driving performance-behavior in AD and
MCI, may be provided from:
1. cognitive testing (visuospatial memory, attention, executive
functions)
2. neurological examination (motor coordination, balance)
3. neuropsychiatric evaluation (behavioral symptoms,
depression)
4. sleep quality evaluation
Department of Transportation Planning and Engineering,
National Technical University of Athens,
2nd Department of Neurology, “Attikon” University General Hospital,
Department of Psychology University of Athens
Traffic engineers
Pr. G. Yannis
Pr. J. Golias
Pr. K. Antoniou
Dr. S. Vardaki
Dr. E. Papadimitriou
Mr. D. Pavlou
Dr. P. Papantoniou
Neurologists
Pr. S. Papageorgiou
Dr. N. Andronas
Psychiatrist
Dr. J. Papatriantafyllou
Neuropsychologists
Pr. A. Economou
Pr. M. Kosmidis
Pr. A. Papanicolaou
Dr. I. Beratis
Mrs. D. Kontaxopoulou
Mrs. S. Frangiadaki
AD cases
 Participant 1
Diagnosis: Alzheimer’s Disease Neuropsychological findings:
• MMSE: 22/30
• A 70 year old right handed man
• MoCA: 14/30
• 10 years of education
• sales manager (retired) • CDT Free: 6/7
• FAB: 10/18
Neuropsychiatric • Verbal Fluency: 9/1 min
Symptoms • TMT A: 203 sec

NPI=0 • TMT B: >5 min

GDS=1
 Participant 2

Diagnosis: Alzheimer’s Disease Neuropsychological findings:

• MMSE: 27/30
• A 80 year old right handed man
• MoCA: 22/30
• 18 years of education
• doctor (retired) • CDT Free: 6/7
• FAB: 12/18
Neuropsychiatric • Verbal Fluency: 9/1 min
Symptoms
• TMT A: 160 sec
NPI=0 • TMT B: >5 min
GDS=1
 Clinical
Diagnosis: FTD
Case, E.V.
Neuropsychological
• A 49 year old right handed
findings
woman MMSE: 25/30 ( -4 calculation, -1
language)
• 9 years of education
• Housewife • MoCA: 22/30
• CDT Free: 6/7
• Mild Speech disorders • FAB: 12/18
(logopenia)
• Verbal Fluency: 11/1 min
• Apathy, Loss of Insight • TMT A: 33 sec
• TMT B: >5 min
• Very mild Frontal atrophy