The effects of aging and driving experience on reaction times of professional drivers

1 Introduction

Road traffic injuries are among the leading 10 causes of death all over the world thus traffic safety is of great importance [1]. The concept of public health concern, is the obligation and responsibility for the improvement of traffic safety, people, natural and material goods [2]. Professional drivers are active working population, mainly male. The requirements for fitness to drive are almost harmonized to the full extent in the European Union [3–6]. Operating a motor vehicle requires complex skills, quick adaptation and reacting on changes [7, 8]. The number of stimuli increases with higher driving speed, traffic congestion and changes in the road environment [9]. Driver’s reaction is determined by sensory, cognitive and motor skills [10–12] and driving depends on the skills, knowledge and experience of the driver. Expectations are, especially for professional drivers, emotional stability, calm, responsibility, healthy habits, good motor skills and specific knowledge [13–15]. The study of reaction time (RT) is a very important trait of drivers, for safely operating the motor vehicle, and plays a significant role [16–22].

Time of reaction is the most widely used measure of behavioural response expressed in time units (milliseconds), from the presentation of a specific task to its completion [23–26]. It is the time interval between the presentation of stimuli and the ending of movement [11, 26] and can be decomposed: a) Mental processing time: perceiving, identifying, analysing the stimulus and decision on the corresponding motor response. b) Movement time: performing the movement of the selected response [24, 27].

There are two kinds: simple and complex or choice RT, independent of each other, due to different physiological processes as a basis for them [26–28]. Low level of cognitive information processing, e.g. only one stimulus and possible answer: simple RT (SRT) [11], e.g. respond to a visual stimulus (an average of 190–220 ms) [29–31]. High level, choice of particular stimulus, responding with the appropriate movement to a given stimulus: choice or complex RT (CRT) (average of 420–630 ms) [12, 31]. As an important human trait, with its long history, in persons for whom driving is the main occupation, studying of RT has great practical significance e.g. if it’s slower than usual, the consequences can be fatal [32]. RT tests are generally considered relevant to driving performance. There are different methodologies and RT tests [33]. Many factors affect the speed of the reaction time, as age [34–38], gender [39, 40], physical fitness [41, 42], experience [43–46], distraction [42, 48], fatigue [14, 49]. Health condition changes which could affect perception or inappropriate decision of driver, for the vehicle control, can have serious negative consequences [9, 50]. To increase human compatibility and to design better systems, various technological solutions need to understand the cause of the variability to determine whether future investments should be directed toward technology or training resources. The aim of this study was to examine the effects of age and experience on reaction times of professional drivers.

2 Materials and methods

3 Results

The sample consisted of 278 male professional drivers. Age mean (42.36±10.44), range 22 to 65 years, driving experience (12.44±9.57), range 1–43 years. Workplace structure: truck 149 (53.59%), bus 26 (9.35%), delivery van 41(14.75%) and taxi drivers 62 (22.30%). Statistical data for dependent variables related to reaction time are shown in Table 1.

Pearson coefficients of linear correlation between age and driving experience with dependent variables are shown in Table 2.

Table 2 shows that age and driving experience are positively correlated to all dependent variables, particularly with total CRT, CRT of correct responses, total SRT and SRT of correct responses.

3.1 Age and dependent reaction time variables

Initial analyses showed that the variability of RT was age dependent. In order to examine whether there exists an effect of age on dependent variables, we chose 10-year interval and categorise professional drivers into four age groups: 22–32 years with 58 (20.86%), 33–42 years with 88 (31.65%), 43–52 years with 78 (28.06%) and 53–64 years with 54 (19.42%) drivers. Table 3 shows mean values and standard deviations for dependent variables in age groups.

OPEN IN VIEWER

Table 3

Mean values and standard deviations (M±SD) for dependent variables related to reaction times in age groups

Variable	Age
	22–32	33–42	43–52	53–64	Total
	n = 58	n = 88	n = 78	n = 54	N = 278
	(20.9%)	(31.6%)	(28.1%)	(19.4%)	(100%)
Total CRT	610.5±117.7	616.4±107.0	699.6±137.0	706.8±137.8	656.1±131.4
CRT of correct responses	578.2±88.6	589.0±85.0	648.1±96.5	673.1±117.7	619.7±102.7
% total incorrect responses	8.7±8.8	9.3±10.9	13.2±12.7	11.7±13.2	12.2±2.3
% red incorrect responses	7.6±13.3	8.1±14.1	9.9±14.8	9.2±13.9	8.7±14.1
% green incorrect responses	11.8±12.4	14.0±18.2	18.1±20.5	19.3±25.3	15.7±19.6
% yellow incorrect responses	8.0±16.9	8.8±21.3	13.5±21.1	6.0±13.1	9.4±19.1
Total SRT	278.9±32.5	298.3±49.9	308.8±56.1	334.7±79.7	304.3±58.5
SRT of correct responses	277.8±31.1	292.8±43.7	300.0±47.6	323.5±63.1	297.7±49.1

Note: Total CRT = total complex reaction time: CRT of correct responses = complex reaction time of correct responses: %  = percentage of different colour incorrect responses: Total SRT = total simple reaction time: SRT of correct responses = simple reaction time of correct responses.

ANOVA was used to test differences between mean values of dependent variables age groups. Statistically significant differences between different age groups for dependent variables were not found: the percentage of a total incorrect response, the percentage of incorrect responses for all three stimuli (red, green and yellow).

Statistically significant differences between different age groups for dependent variables were found: Total CRT, (F(3, 274) = 11.768, p < 0.001), (Fig. 1); CRT of correct responses (F(3, 274) = 14.484, p < 0.001), (Fig. 2); Total SRT, (F(3, 274) = 9.8326, p < 0.001), (Fig. 3); SRT of correct responses, (F(3, 274) = 9.2320, p < 0.001), (Fig. 4). For variables where ANOVA was leading to a conclusion that there was evidence that the group means differ, it had to been determined which of the means are different, Tukey multiple comparison test (Tukey HSD test) was used. The test compares the difference between each pair of means with appropriate adjustment for the multiple testing.

OPEN IN VIEWER

Fig. 1

Mean values of total complex reaction time (total CRT) in age groups. Mean values denoted by different letters are statistically different.

OPEN IN VIEWER

Fig. 2

Mean values of complex reaction time (CRT) correct responses in age groups. Mean values denoted by different letters are statistically different.

OPEN IN VIEWER

Fig. 3

Mean values of total simple reaction time (total SRT) in age groups. Mean values denoted by different letters are statistically different.

OPEN IN VIEWER

Fig. 4

Mean values of simple reaction times (SRT) correct responses in age groups. Mean values denoted by different letters are statistically different.

Mean values of total CRT for age groups 22–32 and 33–42 are statistically different from mean values of total CRT in age groups 43–52 and 53–64, (Tukey HSD test, p < 0.001). Age groups 22–32 and 33–42 are homogeneous and do not have statistically different mean values of Total CRT, as well as age groups 43–52 and 53–64 (Fig. 1).

Mean values of CRT correct responses in age groups 22–32 and 33–42 are statistically different from mean values of CRT correct responses in age groups 43–52 and 53–64, (Tukey HSD test, p < 0.001). Age groups 22–32 and 33–42 are homogeneous and do not have statistically different mean values of CRT correct responses, as well as age groups 43–52 and 53–64 (Fig. 2).

Mean values of total SRT in age group 53–64 is statistically different from the other age groups, which means that drivers older than 53 years have statistically significant slower total SRT. Age groups 43–52 and 53–64 are statistically different, too. (Tukey HSD test, p < 0.001). Age groups 22–32 and 33–42 are homogeneous and do not have statistically different mean values of total SRT, as well as age groups 33–42 and 43–52 (Fig. 3).

Mean values of SRT correct responses in age group 53–64 is statistically different from the other age groups, which means that drivers older than 53 years have statistically significant slower SRT of correct responses. Also, there is a statistically significant difference in mean values of SRT of correct responses between age groups 33–42 and 43–52, (Tukey HSD test, p < 0.001). Other age groups are homogenous and do not have statistically different mean values of SRT of correct responses (Fig. 4).

3.2 Driving experience and dependent variables related to reaction times

Driving experience was considered in order to test importance of professional practice. Further analysis carried out to determine how driving experience linked with dependent variables. In order to examine whether there exists an effect of driving experience on dependent variables, we chose 5-years interval and categorised drivers into seven driving experience groups: 1–5 years 82 (29.5%), 6–10 years 70 (25.2%), 11–15 years 39 (14.0%), 16–20 years 37 (13.3%), 21–25 years 18 (6.5%), 26–30 years 16 (5.8%) and over 30 years 16 (5.8%). Table 4 shows mean values and standard deviations for dependent variables in driving experience groups (in years).

OPEN IN VIEWER

Table 4

Mean values and standard deviations for dependent variables related to reaction time in driving experience groups (in years)

Variable	Driving experience
	1–5	6–10	11–15	16–20	21–25	26–30	>30	Total
	n = 82	n = 70	n = 39	n = 37	n = 18	n = 16	n = 16	N = 278
	(29.5%)	(25.2%)	(14.0%)	(13.3%)	(6.5%)	(5.8%)	(5.8%)	(100%)
Total CRT	617.9±130.7	635.0±121.2	661.1±113.0	678.1±129.9	709.1±127.1	706.8±141.7	770.5±128.2	656.1±131.4
CRT of correct responses	590.3±104.0	598.4±96.3	626.7±87.5	640.4±90.1	667.2±110.7	659.0±94.5	705.1±109.1	619.7±102.7
% total incorrect responses	10.6±11.5	9.1±9.3	9.0±8.8	12.3±12.5	10.8±12.9	15.9±15.3	14.4±17.6	12.2±2.3
% red incorrect responses	6.5±10.7	9.8±15.8	7.1±11.5	7.9±12.9	7.3±14.9	21.1±22.1	10.6±13.9	8.7±14.1
% green incorrect responses	17.6±21.3	11.6±15.3	13.4±15.8	20.2±21.8	11.8±15.0	19.8±23.2	19.5±27.7	15.7±19.6
% yellow incorrect responses	8.4±19.7	9.0±16.7	9.2±18.0	9.8±20.7	10.1±15.6	11.3±24.8	13.1±25.1	9.4±19.1
Total SRT	292.2±53.0	297.4±49.4	296.6±43.5	313.1±45.2	318.4±66.7	316.2±40.6	366.9±120.3	304.3±58.5
SRT of correct responses	284.9±42.0	294.0±45.1	289.5±38.8	308.2±40.9	312.1±58.0	312.4±42.0	343.8±89.8	297.7±49.1

Note: Total CRT = total complex reaction time: CRT of correct responses = complex reaction time of correct responses: %  = percentage of different colour incorrect responses: Total SRT = total simple reaction time: SRT of correct responses = simple reaction time of correct responses.

ANOVA was used to test differences between mean values of dependent variables in driving experience groups. Statistically significant differences between driving experience groups for dependent variables: the percentage of the total incorrect response, the percentage of incorrect responses for all three stimuli (red, green, and yellow), were not found.

Statistically significant differences between driving experience groups, were found for dependent variables: Total CRT, (F (6, 271) = 4.9306, p < 0.001), (Fig. 5); CRT of correct answers (F (6, 271) = 5.2161, p < 0.001), (Fig. 6); Total SRT, (F (6, 271) = 4.6868, p < 0.001), (Fig. 7); SRT of correct responses, (F (6, 271) = 4.6482, p < 0.001), (Fig. 8). For variables where ANOVA was leading to a conclusion that there was evidence that the group means differ, it had to been determined which of the means are different, Tukey multiple comparison test (Tukey HSD test) was used. The test compares the difference between each pair of means with appropriate adjustment for the multiple testing.

OPEN IN VIEWER

Fig. 5

Mean values of total complex reaction time (total CRT) in driving experience groups. Mean values denoted by different letters are statistically different.

OPEN IN VIEWER

Fig. 6

Mean values of complex reaction time (CRT) correct responses in driving experience groups. Mean values denoted by different letters are statistically different.

OPEN IN VIEWER

Fig. 7

Mean values of total simple reaction time (total SRT) in driving experience groups. Mean values denoted by different letters are statistically different.

OPEN IN VIEWER

Fig. 8

Mean values of simple reaction time (SRT) correct responses in different driving experience groups. Mean values denoted by different letters are statistically different.

Mean value of total CRT for group > 30 years driving experience is statistically different from the total CRT in groups 1–5 and 6–10 years driving experience, (Tukey HSD test, p < 0.001). Mean values of total CRT are not statistically significant different for other homogenous groups (Fig. 5).

Mean value of CRT correct responses in group > 30 years driving experience is statistically different from mean values of CRT correct responses for groups 1–5 and 6–10 years driving experience. Driving experience groups 1–5 and 21–25 are statistically different, too. (Tukey HSD test, p < 0.001). Mean values of CRT correct responses are not statistically significant different for other homogenous groups (Fig. 6).

Mean value of total SRT for group > 30 years driving experience is statistically different from total SRT in groups 1–5, 6–10, 11–15 and 16–20 years driving experience, (Tukey HSD test, p < 0.001). Mean values of total SRT are not statistically significant different for other homogenous groups (Fig. 7).

Mean value of SRT correct responses for group > 30 years driving experience is statistically different from SRT correct responses for groups 1–5, 6–10 and 11–15 years driving experience, (Tukey HSD test, p < 0.001). Mean values of SRT correct responses are not statistically significant different for other homogenous groups (Fig. 8).

This study assesses part of driving efficiency of professional drivers by measuring SRT (one stimulus) and CRT (three stimuli) in the domains (cognitive, perceptual and motor skills) that are relevant to driving safety. Driving requires an executive function for attention, multi-tasking and fast responses. The response time is an important subjective characteristic of the driver from which safety and efficiency of driving significantly depend. This is an individual characteristic, which depends on driver traits and of a large number of objective circumstances. Different drivers have varying response times, but even one driver can show different response times, depending on his psycho-physical condition and objective circumstances.

Measurement methodologies of driver reaction time vary. What is the best approach, it is hard to give answer because there are a lot of studies differ with respect to variables that affect reaction time, such as composition of the sample, testing procedures and analysing of data. Some reaction time studies use responses with a finger so the time of movement always influences the measurement of the reaction time [52–56]. There are studies that use foot response [22, 57–64]. In all of these studies total reaction times measured in drivers were in the range 500–850 ms. In our study participants followed stimuli on screen and responds to appearance of stimuli at a predetermined location. However, the cognitive load was high because we recorded incorrect responses. Mean total reaction time i.e. CRT in our study was 656 ms (Table 1). Green M. [33] in his review concluded it is possible to estimate driver’s reaction times under specific conditions. He showed that the most important variable affecting RTs is expectation. When drivers are fully aware of the time and location of stimuli, they can detect the stimulus and react by movement in about 700 to 750 ms [33]. Some studies recorded values of reaction time in drivers up to 45 years old, in various road conditions in the range 465–730 ms [18, 65–68]. Average age of our professional drivers was 42,36 years.

To evaluate the effectiveness of driving appropriately, it is necessary to take into account both the speed and accuracy. Also, in our study we examined effects of age and driving experience. In further analysis, Pearson’s correlation was used to test the linear association between each dependent variable related to reaction time with age and driving experience (Table 2). Age and driving experience were correlated positively with all variables, especially to total CRT, CRT correct responses, total SRT and SRT correct responses. This indicates that the older drivers, regardless of driving experience, have slower reaction times. However, there was not found positive correlation between driving experience with the percentage of total incorrect answers and percentage of incorrect answers for each of three colours (red, green, and yellow), which can be interpreted as a positive effect of practice and experience, on attention, accuracy, caution or executive performance while driving. This could to some extent decrease the negative effect of ageing on the speed of reaction time.

In order to drive safely, the driver needs to maintain a large number of complex perceptual and cognitive skills. Included in this skill set is the ability to accurately identify hazards in the driving environment. Impairments of cognitive functions resulting from normal ageing hamper attention, memory and executive functions [53, 69–71]. The speed of reaction response reaches a peak in the third and fourth decade of human life, and later begin to weaken, but on the other side, the working memory, information processing, concentration, perception, attention, accuracy, do not mandatory, so dependent on age [26, 72]. Basic reaction time studies generally find correlation between decreased reaction time and the natural process of ageing [62, 74]. The slowing of motor and sensory conduction velocity with ageing correlates with degeneration of horn cells in the spinal cord and neuromuscular junctions [31, 75]. These neurostructural changes may particularly account for increased SRT, where central processing is limited. Changes in driving performance are progressive throughout the driver age range. A few previous studies showed slowing down of age-related SRT [53, 76]. Our study showed similar results. SRT in the oldest group of drivers (53–64) - 334.7 ms was significantly higher compared to SRT in the other groups of drivers (Fig. 3). Examining driving performance need caring out throughout the whole age range. Also, in our study, total CRTs in the older groups of drivers were significantly higher than in the younger groups (Fig. 1). CRT deteriorates during ageing too.

It seems that the ageing could be the main cause of the slower reaction response of professional drivers. Driving leads to automated sensorimotor associations between perception and action. Fortunately, driving events (e.g. dangerous manoeuvres) are not so frequent, and weaknesses in driving skills may be evident only when the driver is in a challenging situation that requires him to rapidly evaluate and respond [56, 77]. These are situations that older drivers avoid [78]. However, it was found that older drivers are slower in making driving decisions than the younger one, but if they are given sufficient time the quality of their decisions does not decline [56, 79]. One study found that in the absence of disease or functional impairment there is no evidence of adverse effects of ageing upon driving ability [80]. Obviously, many different types of vehicles on the road have different acceleration, cornering and braking capabilities. That to a large extent explains different reactions of drivers [81]. The actual differences in the driving behaviour of middle-aged drivers, compared with elderly drivers, are rather small in the majority of traffic situations [82].

We have not found correlation between the driving experience and speed of response, which indicates that driving experience does not affect the speed of reaction times. Our results showed that total CRT and SRT in group over 30 years of driving experience were significantly slower compared to the other groups with lees driving experience. However, there was not found statistically significant difference between other driving experience groups (Fig. 5 and 7). CRT and SRT of correct responses in group over 30 years driving experience were significantly different from the other groups (Fig. 6 and 8). Between the other groups, only statistically difference between group 1–5 and group 21–25 years of driving experience was found (Fig. 6).

Driving requires multitasking. In many operational situations, drivers use planning activities in order to achieve a fast response with high accuracy. One factor commonly cited in favour of elderly drivers is their experience [77]. Elderly drivers showed a tendency to slow down their response time to improve accuracy [83–85]. Furthermore, it is important to note that many, but not all, of the professional drivers, retain the benefits of long years of driving experience. Other studies found different results. In the simple task tests, the elderly driver’s group was as accurate as the younger group, even though elderly drivers needed more time to respond. In the other more complex tests, the elderly made more errors but exhibited only a slightly higher rate of no-response than the younger group. This means that they probably needed more time to process the information since they more frequently responded outside of the required time-window or did not respond at all [56, 62].

A definite answer cannot be given because there are a lot of differences in studies concerning several variables that could affect RT: composition of the samples, testing procedure, and handling of data. There are limitations of our study. Our research considered health requirements according to regulations related to driver health. According to the legislation, the health requirements for professional drivers are stricter than for amateur drivers. The basic directive governing the ability to drive vehicles in EU countries is Directive 2006/126/EC (52). Annex III to this Directive, based on the results of major world studies, defines a minimum list of diseases that are contraindicated for driving. Diseases that can effect major of sensory functions (visual, hearing or balance disorders), on the general state of consciousness (consequences of head injuries, cardiovascular, respiratory, endocrine, neurological and psychiatric diseases), which affect reflexes and muscle strength (e.g. operations) [4, 79] and reaction time can be assessed for various types of physical illnesses and to predict who can and who cannot drive. However, we did not consider other health disorders which can effect safe driving of amateurs. We tested the reaction time of professional drivers in comfortable conditions. The speed of reaction time can certainly be different in the real environment, and under other circumstances. Nevertheless, we are certain that the results of this study can serve as a good base for future studies.

5 Conclusions

Our results showed a positive correlation between ageing and speed of response. This suggests that the natural ageing process clearly slows down reaction times. So, the response times of professional drivers are more dependent on age than on driving experience. Although, increased reaction times of elderly drivers, mandatory not have negative implications for some driving performance and efficiency, e.g. attention and accuracy which could to some extent decrease the effect of ageing on reaction time speed. We did not find statistically significant differences in the number of incorrect answers, regardless of age. Practice at a certain extent could compensate effects of ageing on the driving efficiency and safety of professional drivers. These observations suggest that slower reaction time is not necessarily a factor for safe driving criteria.

Conflict of interest

None to report.