Available via license: CC BY-ND 4.0
Content may be subject to copyright.
Title:
Turning the head while biking makes older people lose cycling
direction and balance
Authors:
● Maarten Afschrift1,2,3
● Anouck Matthijs 3,4
● Theresa De Ryck 3
● Friedl De Groote 3, *
● Jean-Jacques Orban De Xivry 3,4,*
1 Department of Mechanical Engineering, Robotics Core Lab of Flanders Make, KU
Leuven, Leuven, Belgium,
2 Department of Human Movement Sciences, Vrije Universiteit Amsterdam,
Amsterdam, The Netherlands
3 Department of Movement Sciences, KU Leuven, Leuven, Belgium
4 KU Leuven Brain Institute, KU Leuven, Leuven, Belgium
* These two authors equally contributed to the work.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
Abstract
Background: Bike safety in older subjects is a major concern in multiple countries given the
strong increase in the number of cyclists that are admitted to the hospital due to single-sided
bicycle accidents in older populations. The increase in single-sided accident rate in older
subjects suggests that older adults have more difficulty controlling balance while bicycling in
traffic, which has been related to various age-related changes in the sensorimotor system.
Yet, the impact of these age-related sensorimotor deficits on the ability of older people to
remain stable on their bike during traffic situations such as turning their upper body to check
on potential road users, remains unknown.
Methods: We instructed a group of 40 young (22.86 ± 1.53 years) and 41 older participants
(62.73 ± 1.57 years) to bike in a straight line while performing a shoulder check movement in
order to identify the color of an object presented behind them. We recorded the task-errors
(lost balance, errors in identifying the color, cycling outside the lane) and computed the
steering angle and rotation of the frame, pelvis and torso from IMU recordings.
Results: We observed that older adults made task-errors (one third of the older participants
failed at the task by making errors such as cone-identification errors, loss of balance and
cycling outside the lane). In the successful trials, we observed an increase in steering angle
and rotation of the pelvis with respect to the frame in the older subjects compared to the young
subjects.
Significance: Older cyclists lose balance while turning their head, which might contribute to
the increased number of single sided bike accidents in older subjects. The use of simple
devices, such as using a mirror, should be encouraged in older subjects to replace the
shoulder check maneuver.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
1. Introduction
Bicycling is a common mode of transport and is also a popular recreational activity in multiple
countries (e.g. Belgium, the Netherland, Denmark). In Flanders (Belgium), 83% of the adult
population uses a bicycle (VAB fietsbarometer 2018) and one out of three uses their bicycle
daily (Fietsersbond & GRACQ 2019). Similar trends are observed throughout the world. For
instance, in Denmark or Norway, bicycle use has increased by 10% over the last 20 years
[1,2]. Similarly, the number of cyclists has increased in many European countries since 1990
[3]. Despite the health benefits related to this physical activity [4–6], bike safety is a major
concern in these countries. In 2019, 10379 cyclists were injured in an accident, which is the
highest number ever reported in Flanders.
Causes of traffic accidents during cycling are related to the interaction between (1) the human
on the bicycle, (2) the mechanics of the bicycle, and (3) the environment (including other road
users) [7–9]. Accident statistics suggest that these interactions differ in different
subpopulations. In young or inexperienced cyclists, most cycling accidents occur in interaction
with motorized vehicles, whereas in older adults, there is a strong increase in the number of
cyclists that are admitted to the hospital due to single-sided bicycle accidents (i.e., accidents
involving the cyclist only) [10]. Between 60% and 95% of cyclists admitted to hospitals are the
result of single-sided bicycle accidents [11]. There is a 37% increase in single-sided accidents
in older adults compared to the average of all cyclists and a 24% increase in collisions with
another obstacle than a road user [12].
The increase in single-sided accident rate in older subjects suggests that older adults have
more difficulty controlling cycling direction and balance while bicycling in traffic. Bulsink et al
suggested that older adults need more effort to restore from perturbations based on the
observation that older subjects use additional strategies, such as outward knee movement, to
control balance in response to a perturbation [13]. Neuro-physiological changes in the human
body with aging [14], such as frailty, slower reaction times [15], less accurate sensory abilities
[16,17], reduced range of motion in the musculoskeletal system [18] and higher mental
workload experienced [19,20], might be associated with the observed differences in balance
control strategies [13] and the increased number of single sided accidents. For example, the
increase of accident risk in older cyclists (on an e-bike) has been related to a relatively higher
mental workload for older cyclists (65+ years) compared to a group of younger cyclists (30-45
years) [20]. Yet, the impact of these age-related sensorimotor deficits on the ability of older
people to remain stable on their bike during traffic situations such as turning their upper body
to check on potential road users, remains unknown.
To fill this gap, we instructed a group of young and a group of older participants to bike in a
straight line while performing a shoulder check movement in order to identify the color of an
object presented behind them. This task involves both the rotation of trunk and head for
perception of the environment and the concurrent control of balance and cycling direction.
Therefore, it allows us to investigate the ability of older adults to perform these two concurrent
motor tasks. This task is ecologically valid as it is relevant for multiple traffic situations: such
as identifying cars behind the cyclist when changing lane or overtaking another road user. We
evaluated if the young and older adults could perform the task without errors (e.g. wrong
identification, cyclist outside the test track, loss of balance) and if they performed steering
corrections to maintain cycling direction.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
2. Methods
2. 1 participants
A group of 40 young (22.86 ± 1.53 years) and 41 older subjects (62.73 ± 1.57 years, minimum
60 years and maximum 65 years) subjects participated in this study. Participants were
recruited through mailings to local organizations and sport clubs for seniors, through local
advertisements and by contacting the social network of the research team. Only participants
who cycled at least once a month were included. The study was approved by the local ethics
committee (SMEC G-2020-2410) and all participants provided their written informed consent.
2.2 Experimental procedure
We instructed participants to cycle on a self-designed test track that mimicked several real-
world conditions on an aluminum bike with (e-bike) or without a motor (conventional bike)
(order of the bike was randomized across participants, mass conventional bike = 19.8 kg,
details in appendix). Participants biked on the track under three different conditions: first at
self-selected speed, then twice more slowly than the self-selected speed and finally while
performing a concurrent dual-task. For this paper, we focus on the self-selected speed
condition on the conventional bike but the results are identical for the e-bike and for the slow
condition (Appendix 2). The subjects familiarized themselves with the cycling track for five
minutes (two repetitions) with both the e-bike and the normal bike and subsequently performed
one trial that was recorded and analyzed. In the present study we analyzed the part in the test
track where subjects had to identify a yellow or red cone behind the cyclist by first looking over
the left and secondly over the right shoulder (Figure 1). During the test, the experimenter wrote
down (a) whether the participants lost balance and needed to place a foot on the ground or
not, (b) whether the participants made mistakes in identifying the cone, (c) and whether they
cycled outside the narrow lane with yellow cones (Figure 1).
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
Figure 1: Overview of the kinematic analysis. Young and older subjects identified a yellow or red cone behind the
cyclist by first looking over the left and secondly over the right shoulder while cycling in a narrow track between
cones (A). The kinematics of the bike and cyclist were analyzed using IMU’s on the handlebar, frame, pelvis, torso.
Sensor orientations were computed from the IMU raw data and orientations were evaluated around the vertical
axis (B). The start and end of the rotation over the right and left shoulder was identified manually based on the
angular velocity of sensors (C). Relative changes in orientation between the sensor on the frame, pelvis and torso
were computed in the identified time-window to quantify the relative rotation of the segments (D). The range of
motion was computed from the relative rotation between the sensors (E). The steering angle was computed as the
orientation of the handlebars with respect to the frame along the axis of the stem (F), and the resulting standard
deviation in steering angle was compared between the young and older subjects (G).
2.3 Kinematic analysis + determine steering angle
The kinematics of the cyclist and bike was analyzed using inertial measurement units (IMU)
mounted on the handlebars, frame, pelvis and torso (MTw Awinda, Xsens, The Netherlands).
Sensor orientations were computed from the raw IMU data using the Xsens sensor fusion
algorithm in proprietary software MT manager (Xsens, The Netherlands). Sensor orientations
were exported as rotation matrices for further analysis.
The steering angle is defined as the rotation of the handlebars with respect to the frame along
the hinge of the stem. First, we estimated the axis of the stem by solving a least-squares
problem based on a calibration movement where the subject rotated the handlebars through
the full range of motion when the bike was in a static position [21]. This method relies on the
assumption that the change in orientation between the steer and the frame mainly occurs
around one axis (i.e., the axis of the stem) during the calibration movement. Second, we used
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
euler angles to compute the steering angle as the orientation of the stem with respect to the
handlebars along the axis of the stem for all recorded frames.
The orientation of the sensors and the relative angles between sensors were analyzed along
the vertical (gravity) axis by converting the rotation matrices to euler angles. The range of
motion, defined as the maximal angle between sensors, was computed for the rotation of the
pelvis with respect to the frame, the torso with respect to the frame and the torso with respect
to the pelvis (Figure 1). The start and end of the rotation over the right and left shoulder was
identified manually based on the angular velocity of the torso sensor along the vertical axis
(Figure 1c).
2.4 Excluded data
To explore if young and older subjects used different kinematic strategies during this task we
evaluated the range of motion between the IMU sensors on frame, pelvis and torso for the
subjects that performed the trial successfully. The data of subjects that lost balance (i.e. foot
on the ground, cycling outside the lane) or made mistakes in identifying the cones was
excluded from the kinematic analysis. We additionally had to exclude data for several young
and older subjects due to a faulty wireless connection between the IMU sensors and the base
station. There was mainly a data loss for the sensor on the pelvis (17 young subjects, 9 older
subjects) and to a lesser extent for sensors on frame, torso and handlebars (4 young and 4
older subjects). We therefore have a different number of participants depending if the outcome
measure included the pelvis sensor or not (See Figure 2B and 3). Note that the errors in the
task-performance were reported for all subjects (Figure 2A).
2.5 Outcomes and statistics
The assumption of a normal distribution was tested using a Shapiro-wilk test for the data of
the young and older subjects. Depending if the assumption of normal distribution was violated,
we used a two-tailed Mann-Whitney U test or an independent t-test (ttest2, Matlab 2019) with
an alpha of 0.05 to compare the relative range of motion of the sensors on the frame, pelvis
and torso and the standard deviation in steering angle between the young and older subjects.
3. Results
Older adults had more difficulties than young adults in identifying the color of a cone behind
them by looking over their shoulder. Even when older adults were successful in identifying the
color of the cone, they more often made other task-related errors and used a different
kinematic strategy than young adults.
3.1 Increased number of task-related errors in older adults
The shoulder check task was successful in 100% and 69% of the trials performed respectively
by the young and older adults (Figure 2). For the older adults, placing one or both feet on the
ground was observed as the most common error during the shoulder check task (22%),
followed by cycling outside the lines (5%), and errors in identifying the color of the cone (5%).
In addition, 15% of the healthy older adults lifted their pelvis from the saddle during the
shoulder check task while all younger subjects remained seated.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
Even when older adults performed the shoulder check successfully, they required more
steering corrections than younger adults to remain stable and to continue cycling straight
ahead while performing a shoulder check. We found that the standard deviation in steering
angle in successful trials was larger in the young compared to the older subjects (Figure 2B,
t(62) = -3.73, p<0.001, Cohen’s d=0.95).
3.2 Increased torso rotation in the older adults
Older adults rotated significantly further with the torso with respect to the frame during the task
than younger adults (Figure 3A, t(62) = -6.28, p<0.001, Cohen’s d=1.6). The increased
rotation of the torso with respect to the frame was caused by an increased rotation of the pelvis
with respect to the frame (Figure 3B, t(44) = -6.75 ,p < 0.001, Cohen’s d = -2.04) as we did
not find any evidence for an age-related increase in the rotation of the torso with respect to
the pelvis (Figure 3C, t(44) = 1.50 ,p= 0.14, Cohen’s d= -0.45). As a result, older subjects
required less rotation in the neck and/or smaller eye-head rotations than young adults to
identify the color of the cone (Figure 3 D).
Figure 2: A. 100% and 69% of the trials performed respectively by the young and older adults were successful (no
error). B. The standard deviation in steering angle was lower in the young compared to the older subjects (B).
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
Figure 3: A. Older subjects rotated more with the torso with respect to the frame than young adult during the
shoulder check. B. Older adults rotated more with the pelvis with respect to the frame than young adults during
the shoulder check. C. The range of motion of the trunk with respect to the pelvis was not different in young and
older adults. D. The average differences in kinematics between young and older adults is illustrated in pane D.
Differences in torso-frame rotations imply that young adults relied more on rotation in the neck and/or eye-head
rotation to identify the color of the cones than older subjects. Note that we excluded data of several young on older
subjects due to (1) an error in the task or (2) a faulty wireless connection between the IMU sensors on the
handlebars, frame, pelvis and torso during the task (see details in method section).
4. Discussion
Older cyclists were worse than young cyclists in performing a shoulder check. We asked
participants to identify the color of objects behind them while controlling cycling direction. Even
though our group of older cyclists was by no means representative for the oldest cyclists on
the road as they were aged up to 65 years only, one third of them failed the task by either
making cone-identification errors, interrupting cycling to prevent loss of balance, or cycling
outside the lane while all young adults performed the task successfully (Fig. 2A). In addition,
older subjects who performed the task successfully used more steering actions and rotated
their torsos more with respect to the frame than young adults (Fig. 3). The observed
differences in kinematic strategies between the group of young and older subjects show that
even when older subjects could perform the task successfully, they had more difficulties in
performing the concurrent tasks of object identification and controlling balance and cycling
direction. The increase in variation in steering angle in the older subjects shows that they had
to make more corrections to control the cycling direction or to control balance.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
A first explanation for the increase in steering corrections is the different kinematic strategy
used by the older subjects. Older subjects rotated more with the torso with respect to the
frame, possibly to compensate for a decreased axial rotation range of motion in the neck and
in eye-head rotation [22]. This is in agreement with research in car drivers where a reduced
range of motion in the neck, and functional neck range of motion during the task, in older
adults was associated with an increase in target detection errors during a shoulder check [23].
Yet, in contrast to our results in biking, older participants exhibited less trunk rotation to check
the blind spot than younger participants when driving a car [23], which directly impacted their
ability to detect a target. Reduced neck range of motion has also been identified as a risk
factor for older car drivers [24] but this can be improved by training neck, shoulder and trunk
flexibility [25,26].
Alternatively, given the mechanical coupling between the trunk and the arms, the increased
trunk rotation in the older cyclists might cause undesired arm motion, which results in
undesired steering. Additional steering corrections are most likely needed to compensate for
the co-rotation of the trunk and arms to control balance and cycling direction. Note that the
neck and head-eye rotations strategy used by the young subjects is mechanically not directly
linked to arm motion and therefore undesired steering angles. Hence, the difference in
kinematic strategies (increase in torso rotation) during the shoulder check task might explain
the increase in steering angle variance, and potentially also the increase in balance loss during
the task in the older subjects. Several studies found that inter-joint coordination was not
impaired in older people, suggesting that older people can compensate for the passive
dynamics due to their own movements as well as younger people [27].
A third potential explanation for the increase in steering corrections is a reduced ability in older
subjects to estimate the position of the arms in the absence of vision. When turning the head,
the arm and the steer are located outside of the field of view. Given the age-related decrease
in proprioceptive acuity [16,28–30,but see 31–33], older adults rely much more on vision than
on proprioception to control movements such as bimanual upper limb movements [34],
reaching or grasping [35,36] or walking [37]. The preponderance of visual inputs to maintain
a stable steer control would explain why older adults exhibit more steer variability when the
gaze is away. The plausibility of this alternative is reinforced by the fact that, during car driving,
a multi-tasking condition that took the gaze away from the road led the older adults to exhibit
deviations from their trajectories similar to those observed in the current study [41].
Finally, the increase in steering variability and decrease in task performance could also stem
from the decrease in dual-tasking abilities with aging [19,38,39]. The shoulder check task
requires participants to simultaneously bike straight and look behind. During car driving, dual-
tasking or multitasking made older participants drive in a more variable way than their younger
counterparts [40,41]. Dual tasking during driving causes an increase reaction time to brake,
sloppy steer control and worse distance estimation, especially in older adults [41–43]
We demonstrated that the ability to perform a shoulder check task already drastically
decreased in adults aged 60 to 65 as compared to young adults. While we need to further
understand the causes, this study could have direct policy implications. As shoulder checks
are a common task while cycling, e.g. to safely interact with other road users, the reduced
ability to perform a shoulder task while maintaining heading direction might contribute to the
age-related increase in single-sided accidents [11]. Indeed, when cycling lanes are narrow,
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
deviations from the heading direction might cause cyclists to encounter changes in terrain
(e.g. curbs, soft borders), which might cause them to lose balance. Our results raise
awareness for balance threads imposed by shoulder checks in older cyclists, which might
prompt policy makers to improve cycling infrastructure, such as wider cycling lanes that allow
larger deviations in heading direction, individual cyclists to use assistive devices such as
mirrors, or designers to develop new assistive devices that overcome the limitations of existing
devices such as the limited field of view of bicycle-mounted mirrors.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
5. References
[1] L.B. Andersen, A. Riiser, H. Rutter, S. Goenka, S. Nordengen, A.K. Solbraa, Trends in
cycling and cycle related injuries and a calculation of prevented morbidity and mortality,
J. Transp. Health. 9 (2018) 217–225. https://doi.org/10.1016/j.jth.2018.02.009.
[2] S. Nordengen, L.B. Andersen, A. Riiser, A.K. Solbraa, National Trends in Cycling in
Light of the Norwegian Bike Traffic Index, Int. J. Environ. Res. Public. Health. 18 (2021)
6198. https://doi.org/10.3390/ijerph18126198.
[3] P. Schepers, M. Helbich, M. Hagenzieker, B. de Geus, M. Dozza, N. Agerholm, A.
Niska, N. Airaksinen, F. Papon, R. Gerike, T. Bjørnskau, R. Aldred, The development of
cycling in European countries since 1990, Eur. J. Transp. Infrastruct. Res. 21 (2021)
41–70. https://doi.org/10.18757/ejtir.2021.21.2.5411.
[4] J. Pucher, J. Dill, S. Handy, Infrastructure, programs, and policies to increase bicycling:
An international review, Prev. Med. 50 (2010) S106–S125.
https://doi.org/10.1016/j.ypmed.2009.07.028.
[5] S. Nordengen, L.B. Andersen, A.K. Solbraa, A. Riiser, Cycling is associated with a
lower incidence of cardiovascular diseases and death: Part 1 - systematic review of
cohort studies with meta-analysis, Br. J. Sports Med. 53 (2019) 870–878.
https://doi.org/10.1136/bjsports-2018-099099.
[6] S. Nordengen, L.B. Andersen, A.K. Solbraa, A. Riiser, Cycling and cardiovascular
disease risk factors including body composition, blood lipids and cardiorespiratory
fitness analysed as continuous variables: Part 2—systematic review with meta-
analysis, Br. J. Sports Med. 53 (2019) 879–885. https://doi.org/10.1136/bjsports-2018-
099778.
[7] M. Møller, K.H. Janstrup, N. Pilegaard, Bicycle accidents in Denmark – the contribution
of cyclist behavior, the vehicle and the road, (2018).
https://doi.org/10.5281/zenodo.1456413.
[8] R. Buehler, J. Pucher, eds., Cycling for Sustainable Cities, MIT Press, Cambridge, MA,
USA, 2021.
[9] M.S. Myhrmann, K.H. Janstrup, M. Møller, S.E. Mabit, Factors influencing the injury
severity of single-bicycle crashes, Accid. Anal. Prev. 149 (2021) 105875.
https://doi.org/10.1016/j.aap.2020.105875.
[10] Ormel, W., Hertog, P. den, Klein Wolt, K., Enkelvoudige fietsongevallen; Een LIS-
vervolgonderzoek., (2009).
https://puc.overheid.nl/rijkswaterstaat/doc/PUC_133429_31/.
[11] P. Schepers, N. Agerholm, E. Amoros, R. Benington, T. Bjørnskau, S. Dhondt, B. de
Geus, C. Hagemeister, B.P.Y. Loo, A. Niska, An international review of the frequency of
single-bicycle crashes (SBCs) and their relation to bicycle modal share, Inj. Prev. 21
(2015) e138–e143. https://doi.org/10.1136/injuryprev-2013-040964.
[12] D.V. Nathan, Oudere fietsers, (2021).
[13] V.E. Bulsink, H. Kiewiet, D. van de Belt, G.M. Bonnema, B. Koopman, Cycling
strategies of young and older cyclists, Hum. Mov. Sci. 46 (2016) 184–195.
https://doi.org/10.1016/j.humov.2016.01.005.
[14] S.W. Shaffer, A.L. Harrison, Aging of the Somatosensory System: A Translational
Perspective, Phys. Ther. 87 (2007) 193–207. https://doi.org/10.2522/ptj.20060083.
[15] R.M. Hardwick, A.D. Forrence, M.G. Costello, K. Zackowski, A.M. Haith, Age-related
increases in reaction time result from slower preparation, not delayed initiation, (2021)
2021.06.12.448183. https://doi.org/10.1101/2021.06.12.448183.
[16] G.E. Stelmach, A. Sirica, Aging and proprioception, AGE. 9 (1986) 99–103.
https://doi.org/10.1007/BF02432281.
[17] R. Goodman, G.A. Manson, L. Tremblay, Age-related Differences in Sensorimotor
Transformations for Visual and/or Somatosensory Targets: Planning or Execution?,
Exp. Aging Res. 46 (2020) 128–138. https://doi.org/10.1080/0361073X.2020.1716153.
[18] P. Intolo, S. Milosavljevic, D.G. Baxter, A.B. Carman, P. Pal, J. Munn, The effect of age
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
on lumbar range of motion: a systematic review, Man. Ther. 14 (2009) 596–604.
https://doi.org/10.1016/j.math.2009.08.006.
[19] M.P. Boisgontier, I.A.M. Beets, J. Duysens, A. Nieuwboer, R.T. Krampe, S.P. Swinnen,
Age-related differences in attentional cost associated with postural dual tasks:
Increased recruitment of generic cognitive resources in older adults, Neurosci.
Biobehav. Rev. 37 (2013) 1824–1837. https://doi.org/10.1016/j.neubiorev.2013.07.014.
[20] W.P. Vlakveld, D. Twisk, M. Christoph, M. Boele, R. Sikkema, R. Remy, A.L. Schwab,
Speed choice and mental workload of elderly cyclists on e-bikes in simple and complex
traffic situations: A field experiment, Accid. Anal. Prev. 74 (2015) 97–106.
https://doi.org/10.1016/j.aap.2014.10.018.
[21] R.M. Ehrig, W.R. Taylor, G.N. Duda, M.O. Heller, A survey of formal methods for
determining functional joint axes, J. Biomech. 40 (2007) 2150–2157.
https://doi.org/10.1016/j.jbiomech.2006.10.026.
[22] R.B. Isler, B.S. Parsonson, G.J. Hansson, Age related effects of restricted head
movements on the useful field of view of drivers, Accid. Anal. Prev. 29 (1997) 793–801.
https://doi.org/10.1016/S0001-4575(97)00048-1.
[23] K.B. Chen, X. Xu, J.-H. Lin, R.G. Radwin, Evaluation of older driver head functional
range of motion using portable immersive virtual reality, Exp. Gerontol. 70 (2015) 150–
156. https://doi.org/10.1016/j.exger.2015.08.010.
[24] R.A. Marottoli, E.D. Richardson, M.H. Stowe, E.G. Miller, L.M. Brass, L.M. Cooney,
M.E. Tinetti, Development of a test battery to identify older drivers at risk for self-
reported adverse driving events, J. Am. Geriatr. Soc. 46 (1998) 562–568.
https://doi.org/10.1111/j.1532-5415.1998.tb01071.x.
[25] A.C. Ostrow, P. Shaffron, K. McPherson, The effects of a joint range-of-motion physical
fitness training program on the automobile driving skills of older adults, J. Safety Res.
23 (1992) 207–219. https://doi.org/10.1016/0022-4375(92)90003-R.
[26] R.D. Ashman, R.R. Bishu, B.G. Foster, P.T. McCoy, Countermeasures to improve the
driving performance of older drivers, Educ. Gerontol. 20 (1994) 567–577.
https://doi.org/10.1080/0360127940200603.
[27] G. Lee, L. Fradet, C.J. Ketcham, N. Dounskaia, Efficient control of arm movements in
advanced age, Exp. Brain Res. 177 (2007) 78–94. https://doi.org/10.1007/s00221-006-
0648-7.
[28] D.J. Goble, J.P. Coxon, N. Wenderoth, A. Van Impe, S.P. Swinnen, Proprioceptive
sensibility in the elderly: Degeneration, functional consequences and plastic-adaptive
processes, Neurosci. Biobehav. Rev. 33 (2009) 271–278.
https://doi.org/10.1016/j.neubiorev.2008.08.012.
[29] M. Henry, S. Baudry, Age-related changes in leg proprioception: implications for
postural control, J. Neurophysiol. 122 (2019) 525–538.
https://doi.org/10.1152/jn.00067.2019.
[30] F. Ribeiro, J. Oliveira, Aging effects on joint proprioception: the role of physical activity
in proprioception preservation, Eur. Rev. Aging Phys. Act. 4 (2007) 71–76.
https://doi.org/10.1007/s11556-007-0026-x.
[31] N.M. Kitchen, R.C. Miall, Adaptation of reach action to a novel force-field is not
predicted by acuity of dynamic proprioception in either older or younger adults, BioRxiv.
(2020) 2020.07.13.200733. https://doi.org/10.1101/2020.07.13.200733.
[32] K. Vandevoorde, J.-J. Orban de Xivry, Does proprioceptive acuity influence the extent
of implicit sensorimotor adaptation in young and older adults?, J. Neurophysiol. (2021).
https://doi.org/10.1152/jn.00636.2020.
[33] Z.J. Djajadikarta, S.C. Gandevia, J.L. Taylor, Age has no effect on ankle proprioception
when movement history is controlled, J. Appl. Physiol. 128 (2020) 1365–1372.
https://doi.org/10.1152/japplphysiol.00741.2019.
[34] D.J. Serrien, N. Teasdale, C. Bard, M. Fleury, Age-Related Differences in the
Integration of Sensory Information During the Execution of a Bimanual Coordination
Task, J. Mot. Behav. 28 (1996) 337–347.
https://doi.org/10.1080/00222895.1996.10544603.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
[35] S. Chaput, L. Proteau, Aging and Motor Control, J. Gerontol. Ser. B. 51B (1996) P346–
P355. https://doi.org/10.1093/geronb/51B.6.P346.
[36] R.O. Coats, J.P. Wann, The reliance on visual feedback control by older adults is
highlighted in tasks requiring precise endpoint placement and precision grip, Exp. Brain
Res. 214 (2011) 139. https://doi.org/10.1007/s00221-011-2813-x.
[37] J.R. Franz, C.A. Francis, M.S. Allen, S.M. O’Connor, D.G. Thelen, Advanced age
brings a greater reliance on visual feedback to maintain balance during walking, Hum.
Mov. Sci. 40 (2015) 381–392. https://doi.org/10.1016/j.humov.2015.01.012.
[38] P. Verhaeghen, D.W. Steitz, M.J. Sliwinski, J. Cerella, Aging and dual-task
performance: A meta-analysis., Psychol. Aging. 18 (2003) 443–460.
https://doi.org/10.1037/0882-7974.18.3.443.
[39] V. de Rond, D. Orcioli-Silva, B.W. Dijkstra, J.-J. Orban de Xivry, A. Pantall, A.
Nieuwboer, Compromised Brain Activity With Age During a Game-Like Dynamic
Balance Task: Single- vs. Dual-Task Performance, Front. Aging Neurosci. 13 (2021)
345. https://doi.org/10.3389/fnagi.2021.657308.
[40] T.T. Kubose, K. Bock, G.S. Dell, S.M. Garnsey, A.F. Kramer, J. Mayhugh, The effects
of speech production and speech comprehension on simulated driving performance,
Appl. Cogn. Psychol. 20 (2006) 43–63. https://doi.org/10.1002/acp.1164.
[41] K. Wechsler, U. Drescher, C. Janouch, M. Haeger, C. Voelcker-Rehage, O. Bock,
Multitasking During Simulated Car Driving: A Comparison of Young and Older Persons,
Front. Psychol. 0 (2018). https://doi.org/10.3389/fpsyg.2018.00910.
[42] D. Lamble, T. Kauranen, M. Laakso, H. Summala, Cognitive load and detection
thresholds in car following situations: safety implications for using mobile (cellular)
telephones while driving, Accid. Anal. Prev. 31 (1999) 617–623.
https://doi.org/10.1016/S0001-4575(99)00018-4.
[43] A. Chaparro, J.M. Wood, T. Carberry, Effects of age and auditory and visual dual tasks
on closed-road driving performance., Optom. Vis. Sci. Off. Publ. Am. Acad. Optom. 82
(2005) 747–754. https://doi.org/10.1097/01.opx.0000174724.74957.45.
[44] C. Engbers, R. Dubbeldam, J.H. Buurke, N. Kamphuis, S.H.H.M.D.H. Buijssen, F.
Westerhuis, D.D. Waard, J.S. Rietman, A front- and rear-view assistant for older
cyclists: evaluations on technical performance, user experience and behaviour, Int. J.
Hum. Factors Ergon. 5 (2018) 257. https://doi.org/10.1504/IJHFE.2018.096099.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
Appendix 1: details on conventional bike and ebike and sensor placement
The kinematics of the cyclist and bike was analyzed using inertial measurement units (IMU)
mounted on the handlebars, frame, pelvis and torso (Figure Appendix 1). Cycling data of all
subjects was collected while riding the same conventional and electrical bicycles, provided by
the KU Leuven. Both bicycles were similar models with step-through frames. Only a small
difference in frame height was present, with the frame of the electrical bicycle being slightly
higher. The weight of the electrical bicycle was 27.7 kg, which was 7.9 kg heavier than the
conventional bicycle. The electrical bicycle had a front wheel hub motor, and the battery was
located above the rear wheel. Power support could be set to five levels. In the present study
the third level was chosen, which can be described as normal support‘. Each bicycle had six
gears, which were set on the fourth level. Participants were asked to not change these settings
during the whole experiment. Participants could change the saddle height to a comfortable
height. Wearing a bicycle helmet was mandatory.
Figure appendix 1: IMU placement on the frame and handlebars of the bike (A) and on the
pelvis and torso of the subject (B)
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
Appendix 2: Kinematic analysis for e-bike and conventional bike
We observed the difference between the young and older subjects in steering angle and ROM
of the sensor was similar on the conventional bike and ebike during cycling in self-selected
and slow speed conditions (Figure appendix 2).
Figure appendix 2: Standard deviation in steering angle (A) and range of motion between
Frame and torso (B) frame and pelvis (C ) and torso and pelvis (D) during the shoulder
check task when cycling with a conventional bike (N-Bike) at preferred or slow (N-Bike Slow)
speed or with an e-bike (E-Bike) at preferred or slow speed (E-Bike Slow). We observed that
the effect of age on steering angle and ROM was similar for the conventional bike and e-bike
at preferred and slower speed.
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
.CC-BY-ND 4.0 International licensemade available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this preprintthis version posted March 2, 2022. ; https://doi.org/10.1101/2022.03.01.481993doi: bioRxiv preprint
