Cross-sectional association between active commuting and perceived commuting stress in Austrian adults: Results from the HOTway study

Abstract

Objective
Little is known about the acute psychological stress responses caused by commuting. Evidence for the benefits of active commuting (e.g., walking, cycling) is usually based on studies without measurements in free-living environments and without consideration of daily variations in stress. This study investigated the association between commuting mode (active, passive) and perceived commuting stress, assessed on multiple days immediately after commuting.

Methods
Adults participating in the cross-sectional ‘Healthy On The way’ (HOTway) study between 2016 and 2017 in Graz, Austria, were included. Participants completed an online survey and responded to statements about perceived stress (demands, tension) on three days before commuting (baseline stress) and after arrival (commuting stress), respectively. Active commuting was defined as cycling and/or walking (passive: car, motorbike, public transport).

Results
Of 188 participants (93 women, mean age: 28.0 ± 10.0 years) included, 124 were active and 64 were passive commuters. Active commuting was associated with less perceived commuting stress compared to passive commuting (bi = −2.95, 95% CI: −4.97 to −0.92, p = .005), even after controlling for subjective well-being, physical activity, commuting time and other confounding variables.

Conclusion
Active commuting is related to a small reduction in perceived commuting stress. The results of this study support the promotion of active commuting for population (mental) health but future studies on the causal mechanisms and the role of active commuting in the recovery from previous stressors are needed.

Full text

Mental Health and Physical Activity 19 (2020) 100356
Available online 12 September 2020
1755-2966/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Cross-sectional association between active commuting and perceived
commuting stress in Austrian adults: Results from the HOTway study
Matteo C. Sattler
a
,
*
, Tanja F¨
arber
a
,
b
, Katharina Traußnig
a
, Gottfried K¨
oberl
a
, Christoph Paier
a
,
Pavel Dietz
a
,
c
, Mireille N.M. van Poppel
a
a
Institute of Human Movement Science, Sport and Health, University of Graz, Graz, Austria
b
Institute of Psychology, Otto-Friedrich-University, Bamberg, Germany
c
Institute of Occupational, Social and Environmental Medicine, University Medical Centre of the University of Mainz, Mainz, Germany
ARTICLE INFO
Keywords:
Physical activity
Travel
Commuting
Stress
Work
ABSTRACT
Objective: Little is known about the acute psychological stress responses caused by commuting. Evidence for the
benets of active commuting (e.g., walking, cycling) is usually based on studies without measurements in free-
living environments and without consideration of daily variations in stress. This study investigated the associ-
ation between commuting mode (active, passive) and perceived commuting stress, assessed on multiple days
immediately after commuting.
Methods: Adults participating in the cross-sectional ‘Healthy On The way’ (HOTway) study between 2016 and
2017 in Graz, Austria, were included. Participants completed an online survey and responded to statements
about perceived stress (demands, tension) on three days before commuting (baseline stress) and after arrival
(commuting stress), respectively. Active commuting was dened as cycling and/or walking (passive: car,
motorbike, public transport).
Results: Of 188 participants (93 women, mean age: 28.0 ±10.0 years) included, 124 were active and 64 were
passive commuters. Active commuting was associated with less perceived commuting stress compared to passive
commuting (b
i
= − 2.95, 95% CI: −4.97 to −0.92, p =.005), even after controlling for subjective well-being,
physical activity, commuting time and other confounding variables.
Conclusion: Active commuting is related to a small reduction in perceived commuting stress. The results of this
study support the promotion of active commuting for population (mental) health but future studies on the causal
mechanisms and the role of active commuting in the recovery from previous stressors are needed.
1. Introduction
Physical activity (PA) plays a key role in promoting and maintaining
health (Lee et al., 2012). Extensive research has shown that higher levels
of moderate-to-vigorous physical activity are linked to lower risks for
cardiovascular disease and dementia, and improvements in quality of
life and mental well-being (Physical Activity Guidelines Advisory
Committee, 2018; Lee et al., 2012). Current guidelines for the adult
population recommend at least 150 min of moderate-intensity, or 75
min of vigorous-intensity aerobic PA (or an equivalent combination)
together with muscle-strengthening activities on two or more days per
week (US Department of Health and Human Services, 2018). In addition,
evidence for the health benets of short-term and lower-intensity ac-
tivities is increasing, indicating that some activity is better than none
(US Department of Health and Human Services, 2018).
Walking and cycling are activities of everyday life and are among the
best investments to increase PA in the population. They are effective in
terms of providing signicant health benets and can often be integrated
into the daily routine. For instance, many daily journeys (e.g., to the
supermarket) are short and can be covered by walking and cycling
(GAPA, 2012; World Health Organization, 2002). Walking and cycling
can also be used for daily commuting (e.g., to work or university), which
was shown to be helpful in meeting recommended levels of PA
(Sahlqvist, Song, & Ogilvie, 2012). Such ‘active commuting’ does not
only provide various environmental benets such as reducing noise and
greenhouse gas emissions (Nazelle et al., 2011), but is also benecial for
ones’ physical health. For instance, it was estimated that walking around
1.9 km in 22 min twice per day, on ve days per week, is already
* Corresponding author. Institute of Sport Science, University of Graz, Mozartgasse 14, Graz, Austria.
E-mail address: matteo.sattler@uni-graz.at (M.C. Sattler).
Contents lists available at ScienceDirect
Mental Health and Physical Activity
journal homepage: www.elsevier.com/locate/menpa
https://doi.org/10.1016/j.mhpa.2020.100356
Received 25 June 2020; Received in revised form 26 July 2020; Accepted 25 August 2020

Mental Health and Physical Activity 19 (2020) 100356
2
associated with a reduction in all-cause mortality (Shephard, 2008).
Moreover, active commuting is related to lower risks for major
non-communicable diseases such as cardiovascular disease and diabetes
(Dinu, Pagliai, Macchi, & So, 2019). Previous research also demon-
strated a positive link to mental health. For example, perceived stress in
the past month was lower in bicycle compared to motorized commuters
(Avila-Palencia et al., 2017). Based on longitudinal data from the British
Household Panel Survey, greater well-being was observed in walking
and bicycle commuters compared to car commuters (Martin, Goryakin,
& Suhrcke, 2014).
Several studies recognized commuting itself as a potential source of
psychological stress (Antoun, Edwards, Sweeting, & Ding, 2017). Psy-
chological stress, which occurs when environmental demands are
perceived to tax or exceed the adaptive capacity of the person (Cohen,
Kessler, & Gordon, 1995), is central in the downstream development of
disease (Cohen, Janicki-Deverts, & Miller, 2007). In fact, considering
commuting as a source of psychological stress is relevant because it is
well-known that daily stressors can threaten our health (Cohen,
Edmondson, & Kronish, 2015; McEwen, 2007).
Morris & Hirsch (2016) showed that driving at peak times was
associated with fatigue and stress, potentially because of greater
unpredictability due to congestion (Wener & Evans, 2011) and the
behaviour of other drivers (Rasmussen, Knapp, & Garner, 2000). Also,
public transport may be perceived as more emotionally taxing than
commuting by car, for instance due to the demands associated with
interchanges, such as waiting times and delays when changing between
different bus routes (Wardman, Hine, & Stradling, 2001). When
perceived stress of commuting is measured immediately after arrival,
car commuters reported less commuting stress than train commuters
(Wener & Evans, 2011), whereas bicycle commuters reported less
commuting stress than car commuters (Brutus, Javadian, & Panaccio,
2017). One study (Friman, Olsson, Ståhl, Ettema, & G¨
arling, 2017) did
not report any differences between different types of commuters with
respect to the dimension ‘relaxation-stress’ (although differences were
observed for ‘enthusiasm-boredom’). However, altogether, a recent re-
view on the relationship between commuting and well-being showed
that active commuting (walking, cycling) seems to be associated with
relatively lower levels of perceived commuting stress compared to more
physically passive commuting modes, such as using the car or public
transport (Chatterjee et al., 2019).
Despite this evidence, several limitations must be acknowledged.
First, much of the evidence is based on data from large surveys without
measurements in the free-living environment (LaJeunesse & Rodríguez,
2012; Morris & Hirsch, 2016). This rather unspecic approach reduces
the external validity of the results and does not provide insight into the
acute psychological stress responses during commuting, for instance due
to recall problems (Friman et al., 2017). Secondly, of those studies in
which perceived commuting stress was measured immediately upon
arrival at work (Brutus et al., 2017; Friman et al., 2017; Wener & Evans,
2011), only one study (Friman et al., 2017) performed baseline mea-
surements (e.g., moods before commuting). The lack of an appropriate
baseline increases the susceptibility to other inuences such as previous
work stress (Beattie & Grifn, 2014) and sleep quality (Blaxton, Ber-
geman, Whitehead, Braun, & Payne, 2017). Also, positive and negative
moods experienced early in the morning can inuence the perception of
subsequent, potentially stressful, events (Rothbard & Wilk, 2011).
Finally, the measurement of commuting stress at a single time point
(Brutus et al., 2017; Wener & Evans, 2011) disregards the strong
intra-individual variation in daily stress levels and may increase mea-
surement error (Hutcheon, Chiolero, & Hanley, 2010; Sliwinski,
Almeida, Smyth, & Stawski, 2009). Only few assessed commuting stress
over multiple days (Friman et al., 2017).
Therefore, the aim of this study was to investigate the cross-sectional
association between commuting mode (active, passive) and perceived
commuting stress in a sample of Austrian adults. Commuting stress was
measured on up to three days at arrival, including baseline
measurements of stress before commuting. We hypothesized that active
commuters (walking, cycling) perceive less commuting stress than
passive commuters (car, motorbike, public transport).
2. Methods
2.1. Study design and participants
This cross-sectional observational study was part of the ‘Healthy On
The way’ (HOTway) study which was designed to assess the association
between the commuting mode and several environmental and psycho-
logical factors in residents of Graz, Austria. Study centre was the Insti-
tute of Human Movement Science, Sport and Health of the University of
Graz. Recruitment of participants and data collection was performed
between October 2016 and July 2017. All participants received infor-
mation about the study procedures and provided informed consent. The
HOTway study received ethical approval by the local research ethics
committee (GZ. 39/1/63 ex 2016/179).
A convenience sample of adults was recruited using the following
eligibility criteria: i) aged ≥18 years; ii) employed and/or studying (e.g.,
at the university); iii) German-speaking; iv) having a persistent resi-
dential address and place of work for at least two months in (the sur-
rounding area of) Graz; and v) using the same commuting mode on at
least four days per week.
Study materials included a city map, paper-pencil questionnaires for
perceived baseline and commuting stress as well as a link to an once-
only online survey. Before (i.e., after waking up) and after (i.e., at
arrival) commuting on three days, participants were asked to respond to
statements about perceived stress. The three days could be chosen by
each participant, but Monday, Wednesday and Friday were recom-
mended. The route of commuting was recorded on the city map and
information regarding sociodemographic background, commuting
mode, PA and mental health were collected by the online survey. For the
reporting of this study, we followed the ‘Strengthening the Reporting of
Observational Studies in Epidemiology (STROBE)’ recommendations
(Vandenbroucke et al., 2014).
2.2. Measurements
2.2.1. Commuting stress and baseline stress
Since no commuting specic stress questionnaire was available in the
German language, seven statements, adopted from the validated
Perceived Stress Questionnaire (Fliege, Rose, Arck, Levenstein, & Klapp,
2001; Levenstein et al., 1993) and the Perceived Stress Scale (Cohen,
Kamarck, & Mermelstein, 1983; Klein et al., 2016), were used to obtain
information about baseline stress and commuting stress. The statements
refer to the dimensions ‘tension’ (internal stress response; items 1 to 3)
and ‘demands’ (perception of external stressors; items 4 to 7), which
have been used in previous research (Biehl, Boecking, Brueggemann,
Grosse, & Mazurek, 2019; Groarke et al., 2017). Although different
factor structures have been suggested, the obtained dimensions are
usually highly correlated and thus, also the sum score across the ob-
tained dimensions was shown to have good measurement properties
(Montero-Marin, Piva; Demarzo, Pereira, Olea, & García-Campayo,
2014; Fliege et al., 2001; Reis, Lehr, Heber, & Ebert, 2019; R¨
onnlund
et al., 2015).
Participants were asked to refer to their current perceptions (‘You
feel rested’) when responding to these statements before commuting
(baseline stress) and to refer to the commute (‘On your commute, you
felt rested’) when responding to them after commuting (commuting
stress). Reponses were made on a 10-point Likert scale ranging from
‘very little’ to ‘very strong’. For each participant, the sum of all re-
sponses, averaged across all days (i.e., up to three days), was calculated
(minimum: 0, maximum: 63). The original statements can be found in
the Supplementary File (Appendix A).
M.C. Sattler et al.

Mental Health and Physical Activity 19 (2020) 100356
3
2.2.2. Commuting mode
A single question within the online survey was used to assess the
commuting mode (walking, cycling, car, motorbike, public transport)
and the average time spent in different modes. Participants who re-
ported only walking and/or cycling were considered as active com-
muters whereas those reporting any combination of the remaining
modes were considered as passive commuters.
2.2.3. Confounding variables
Information on age (years), gender (female/male), education (uni-
versity degree: yes/no), employment status (student: yes/no), monthly
income (<1000, 1000–1800, >1800
€
), marital status (relationship:
yes/no) as well as height and weight [used to calculate body mass index
(BMI; kg/m
2
)] were obtained. Furthermore, information about
commuting time [i.e., “How long (in minutes) does it take you to get to
work, on average?] was available.
Information about PA was obtained by the self-administered German
version of the International Physical Activity Questionnaire – short form
(IPAQ-SF) (Craig et al., 2003). This questionnaire showed acceptable
construct validity and reliability (Lee, Macfarlane, Lam, & Stewart,
2011) and is commonly used for assessing PA in European populations
(Loyen et al., 2016). The IPAQ-SF assesses moderate PA, vigorous PA
and walking performed in the past seven days across all domains (e.g.,
leisure time, work). Data processing was performed in accordance with
the analysis guide (IPAQ Research Committee, 2005) although we also
included values of <10 min due to the increasing evidence for the health
benets of shorter bouts of PA (Physical Activity Guidelines Advisory
Committee, 2018). The sum of all three items (moderate, vigorous,
walking), expressed in average minutes per day (min/day), was
calculated.
The self-administered German version of the World Health Organi-
zation Well-Being Index (WHO-5) was used to measure subjective well-
being (Br¨
ahler, Mühlan, Albani, & Schmidt, 2007; World Health Orga-
nization, 1998). The WHO-5 consists of ve positively phrased questions
about the well-being during the past two weeks and was shown to be a
valid screening tool for depression (Topp, Østergaard, Søndergaard, &
Bech, 2015). Responses were made on a 6-point Likert scale (0: all time,
5: at no time). The answers were summarized for each participant
(minimum: 0, maximum: 25).
Overall, the following confounding variables were considered in our
models based on previous research (Avila-Palencia et al., 2017; Wener &
Evans, 2011): Baseline stress, age, sex, education, employment status,
marital status, BMI, well-being, PA and commuting time. As employ-
ment status and income were highly correlated (r <−.50) and infor-
mation about income was missing for a substantial proportion of the
sample, only employment status was included as a confounding
variable.
2.3. Sample size
No criterion for a minimum sample size was applied in this study.
However, a sample size of approximately 150 participants was targeted
which allowed us to detect a moderate effect (f
2
=.15) with at least 80%
statistical power in the nal model, considering exposure, outcome and
all confounding variables (Faul, Erdfelder, Lang, & Buchner, 2007).
2.4. Statistical analysis
Statistical analyses were performed using SPSS Data Analysis version
26 (IBM Corp, Armonk, NY, USA). Descriptive statistics for all variables
were calculated, including mean and standard deviation (SD) for
continuous variables with normal distribution, median and interquartile
range for continuous variables without normal distribution, and count
and percentage for categorical variables. Normal distribution was
assessed by Shapiro-Wilk, Q-Q-Plots and Histogram. Skewed variables
were log-transformed prior to analysis. Bivariate Pearson correlation
coefcients (r) were calculated to investigate the association between
commuting stress, commuting mode and all confounding variables.
Ordinary least squares regression was used to assess the association
between commuting mode and commuting stress. We performed mini-
mally and fully adjusted models based on a complete-case analysis.
Differences between active and passive commuters as well as between
included and non-included participants were evaluated using indepen-
dent t-test, Mann-Whitney U test or Chi Square test, depending on the
distribution of the data. Signicance was considered as p <0.05.
Several sensitivity analyses were performed. First, we excluded BMI
because it may have been on the causal pathway between PA and mental
health (Kandola, Lewis, Osborn, Stubbs, & Hayes, 2020). This was done
to reduce overadjustment bias (Schisterman, Cole, & Platt, 2009). Sec-
ondly, we excluded employment status due to the moderate overlap with
education in our sample. This was done to reduce multi-collinearity.
Thirdly, the two dimensions of commuting stress (tension, demands)
may be considered separately, depending on the observed factor struc-
ture (Fliege, Rose, Arck, Levenstein, & Klapp, 2001). Therefore, we
repeated our main analysis for the outcomes ‘tension’ and ‘demands’.
Lastly, to evaluate the impact of missing data on the results, multiple
imputation was used based on the missing at random (MAR) assumption
(i.e., MAR depending on outcome and covariates). Using a fully condi-
tional specication method [Markov Chain Monte Carlo (MCMC)], 50
imputed datasets were created to account for the uncertainty about the
missing values. Our imputation model included outcome, exposure and
all confounding variables of the fully adjusted model. In addition, a
model was calculated in which employment status was replaced by in-
come. In order to reduce the risk of misspecication because of miss-
ingness in many variables for the same individual (Seaman & White,
2013), we imputed missing values for all participants who did not
change their mode during the measurement period and (partially)
completed the online survey. After repeating the analysis, the results
were combined using Rubin’s rules (Rubin, 1987).
3. Results
3.1. Participants and descriptive data
Of 253 who consented to participate in the study, data on baseline
and commuting stress were available for 247 participants (232 provided
all three days and 15 provided two days). Of those, 213 (84.2%)
answered the online survey and provided information on commuting
mode and confounding variables. However, 12 participants were
excluded because they changed their commuting mode throughout the
measurement period (based on participants’ records on the paper-pencil
questionnaires). For further 13 participants information about
commuting mode was missing because of reporting an implausible
combination of modes (with respect to the total commuting time) in the
online survey. Therefore, the nal sample of the present analysis con-
sisted of 188 participants (74.3%; overall, missing data ranged from
2.3% for baseline and commuting stress to 39.5% for income).
Of the nal sample, two participants were removed from all analyses
including the IPAQ-SF [one had an implausible high value (daily PA >
960 min) and one did not report on minutes of PA per day]. Finally, one
participant reported 120 min for the commuting time (one-way). This
extreme value was replaced by 60 min, the maximum value observed in
the remaining sample.
The sample included 124 active and 64 passive commuters (Table 1).
Participants were between 18 and 64 years old (M =28.0 ±10.0).
Active and passive commuters showed similar levels of baseline stress
(M
active
=16.9 ±8.0, M
passive
=16.9 ±8.0, p =.988) but active com-
muters reported less commuting stress (M
active
=13.9 ±7.1, M
passive
=
17.3 ±9.2, p =.007, Fig. 1). The results for the comparison between
included (n =186) and non-included (n =27) participants with respect
to the confounding variables of the main model are shown in Supple-
mentary Table S1. Compared to included participants, non-included
M.C. Sattler et al.

Mental Health and Physical Activity 19 (2020) 100356
4
participants reported greater commuting time. Bivariate inter-
correlations among all variables included in the main model are
shown in Supplementary Table S2.
3.2. Commuting mode and commuting stress
In both minimally and fully adjusted models, active commuting was
negatively associated with commuting stress (Supplementary Table S3).
Baseline stress was positively associated with commuting stress. The
fully adjusted model revealed a decrease of 3 points [b
i
= − 2.95, 95%
condence interval (CI): −4.97 to −0.92] in commuting stress for active
compared to passive commuters (minimally adjusted model: b
i
= − 3.55,
95% CI: −5.34 to −1.76). Excluding BMI or employment status had no
considerable inuence on the coefcients in the model (Supplementary
Table S4). Likewise, similar results were obtained when considering the
two dimensions of commuting stress (tension, demands) as separate
outcomes (Supplementary Table S5). The combined results from mul-
tiple imputed datasets did not change the main results concerning the
association between commuting mode and commuting stress (Supple-
mentary Table S6 and S7).
4. Discussion
The present study investigated the association between commuting
mode (active, passive) and psychological commuting stress in a sample
of Austrian adults. Commuting stress was measured on up to three days
at arrival. As hypothesized the results showed that active commuting
(walking, cycling) was associated with lower perceived commuting
stress compared to passive commuting (car, motorbike, public trans-
port). The results were also robust to several sensitivity analyses.
Overall, our results support the promotion of active commuting for
population (mental) health even though we only observed a small effect
of active commuting.
In contrast to previous research on commuting stress (Brutus et al.,
2017; LaJeunesse & Rodríguez, 2012; Morris & Hirsch, 2016; Wener &
Evans, 2011), we performed eld-based measurements of commuting
stress on multiple days and considered daily differences in stress before
commuting. Using this methodology, our results are in line with those
from previous studies. For instance, Avila-Palencia et al. (2017) showed
that bicycle commuters had lower risks of being stressed than
non-bicycle commuters. Employees in Montreal who cycled to work
reported less commuting stress than those who commuted by car (Brutus
et al., 2017). Also, Gatersleben & Uzzell (2007) observed lower levels of
commuting stress for active compared to passive commuters among
British university employees. A recent review concluded that commuters
who walk or cycle are generally more satised with their commute
compared to car and public transport commuters (Chatterjee et al.,
2019).
The observed differences in commuting stress in the current study
are consistent with research showing that PA can have acute psycho-
logical effects. For example, exercise sessions in moderate intensity of
around 20 min (e.g., cycling, yoga, resistance training) can reduce state
anxiety in healthy adults (Asmundson et al., 2013) while similar bouts of
exercise can increase mood and decrease rumination in psychiatric pa-
tients (Brand et al., 2018). Commuting can also provide an opportunity
for relaxation and detachment and thereby, may help to recover from
previous (work) stress (Gatersleben & Uzzell, 2007; van Hooff, 2015).
However, the role of active commuting in the recovery from previous
stressful experiences needs to be explored in future studies.
When commuting by car or public transport several factors, such as
unexpected delays, congestion and the behaviour of other travellers, can
be stressful for the individuum (Morris & Hirsch, 2016; Rüger, Pfaff,
Weishaar, & Wiernik, 2017). For instance, exposure to congestion
resulted in acute increases in systolic and diastolic blood pressure in a
recent study in Lebanon drivers (Bou Samra et al., 2017). While de-
mands can also occur during active commuting, for example when
cycling close to trafc (Caviedes & Figliozzi, 2018), passive commuting
may be associated with somewhat higher levels of demanding events.
Moreover, passive commuting may reduce the extent to which a person
can exercise behavioural control (e.g., due to congestion) or predict
outcomes of the commute (e.g., time of arrival) (Evans, Wener, &
Phillips, 2002; Sposato, R¨
oderer, & Cervinka, 2012; Wener & Evans,
2011). Although the potential discrepancy in demanding events,
perceived control and predictability may help in explaining the differ-
ences in commuting stress between active and passive commuters, the
causal mechanisms remain unclear from this cross-sectional investiga-
tion. For instance, also several other factors such as weather or
Table 1
Sample characteristics of the 188 Austrian commuters.
Characteristic
Total (N =
188)
Active
commuters (n
=124)
Passive
commuters (n
=64)
p
Age (years) 28.0 ±10.0 25.4 ±6.5 33.1 ±13.3 <.001
Sex (female) 93 (49.5) 61 (49.2) 32 (50.0) .917
Education
(university)
78 (41.5) 51 (41.1) 27 (42.2) .889
Employment
status
(student)
81 (43.1) 68 (54.8) 13 (20.3) <.001
Monthly
income
a
<.001
<1000
€
80 (42.6) 62 (50.0) 18 (28.1)
1000–1800
€
44 (23.4) 27 (21.8) 17 (26.6)
>1800
€
29 (15.4) 10 (8.1) 19 (29.7)
Marital status
(relationship)
92 (48.9) 56 (45.2) 36 (56.3) .150
BMI (kg/m
2
) 22.7 ±2.5 22.3 ±2.3 23.4 ±2.7 .002
Well-being
(WHO-5, sum
score)
15.1 ±4.0 15.0 ±4.0 15.3 ±4.1 .689
PA (IPAQ-SF,
min/day)
111.4
(66.3–162.5)
b
114.3
(74.8–167.9)
b
105.0
(43.2–150.5)
.103
Commuting time
(min)
15.0
(10.0–20.0)
11.0 (8.3–18.0) 20.0
(15.0–25.0)
<.001
Note. BMI =body mass index, IPAQ-SF =International Physical Activity
Questionnaire – short form, M =mean, min =minutes, PA =physical activity,
Q
1/3
=quartile 1 or 3, SD =standard deviation, WHO-5 =World Health Or-
ganization (Five) Well-being Index. Values are shown as M ±SD, Median
(Q
1
–Q
3
) or n (%) depending on the distribution of the variable. Active com-
muters: walking, cycling; passive commuters: car, motorbike, public transport.
PA and commuting time were log-transformed prior to signicance testing.
a
A total of 35 participants selected ‘no answer’ (25 active and 10 passive
commuters).
b
Values from two participants (active commuters) were removed.
Fig. 1. Perceived baseline and commuting stress for the 188 Austrian com-
muters (active commuters: n =124, passive commuters: n =64). Mean values
for perceived stress, derived from up to three days of commuting, are shown
together with 95% condence intervals (CI). Active commuters: walking,
cycling. Passive commuters: car, motorbike, public transport. Asterisks (*)
indicate signicant differences (p <.05).
M.C. Sattler et al.

Mental Health and Physical Activity 19 (2020) 100356
5
greenness along the commute may shape the relationship between
commuting mode and stress. In summary, this study showed a positive,
yet weak, association between active commuting and psychological
commuting stress.
4.1. Strengths and limitations
This study has several strengths: i) commuting stress was measured
on three days in the free-living environment to reduce measurement
error and increase external validity; ii) commuting stress was measured
immediately at arrival which allows a better assessment of acute psy-
chological stress responses; and iii) baseline measurements of perceived
stress, namely before commuting, were obtained to control for stressful
events before commuting.
However, a convenience sample of participants was included which
increases selection bias and limits the generalizability of the results.
Secondly, we were unable to control for the exact time when partici-
pants reported on baseline and commuting stress. For instance, it is
possible that some participants responded to the statements not imme-
diately after arrival at work but somewhat later. Furthermore, the design
of the study was cross-sectional which makes conclusions about cau-
sality impossible. One may also consider a bi-directional relationship
between commuting mode and commuting stress (i.e., the stress caused
by a specic commute potentially affects the choice of the mode). This
study compared two types of commuters, which does not allow to draw
conclusions about commuting stress for different sub-types such as
electric bike or motorbike commuters, and may lack in addressing all
relevant variables involved in the relationship between commuting
mode and commuting stress (e.g., job satisfaction, weather). For
example, it was shown that temperature, precipitation and wind can
affect the emotional state of active commuters (B¨
ocker, Dijst, & Faber,
2016). Likewise, we did not control for PA prior commuting (e.g., ex-
ercise sessions early in the morning), the (relative) intensity of PA
involved during active commuting (e.g., brisk walking in relation to
aerobic capacity or age) or whether a person was a driver or co-driver.
Finally, one may argue to consider public transport as a form of active
commuting due to the active parts of the journey (Rissel, Curac,
Greenaway, & Bauman, 2012). However, differences in physical activity
energy expenditure (PAEE) illustrate a distinction between them.
Walking [4.6 metabolic equivalents (METs)] and cycling (6.4 METs) for
commuting require more energy than car (1.3 METs) and bus use (1.7
METs) (Costa et al., 2015). Considering walking and cycling as the
primarily active modes is also in line with previous studies (Celis-Mor-
ales et al., 2017).
4.2. Future studies
Future studies should use devices (e.g., accelerometers) to measure
the level of PA involved during active commuting. This can help to
identify aspects of the dose-response relationship between active
commuting and commuting stress such as the optimum ‘dose’ (e.g.,
duration, intensity) (Shephard, 2008). Future studies may consider a
more sophisticated approach by measuring perceived stress also when-
commuting home after work, by including environmental aspects such
greenness and congestion as well as by implementing longitudinal and
experimental studies to better understand the causal mechanism be-
tween commuting and stress and the potential of active commuting to
facilitate coping with previous stressors. Lastly, we observed a small
difference in commuting stress between active and passive commuters.
Because we are unaware about clinically relevant differences in daily
commuting stress, future studies should also investigate the long-term
consequences of different commuting modes, including associations
with other (mental) health outcomes.
5. Conclusion
Walking and cycling are everyday activities and important for health
promotion in the population. The results of this study expand the evi-
dence for the mental health benets of active commuting and showed
that active commuters perceive less commuting stress than passive
commuters. Although the results support the promotion of active
commuting, we only observed a small effect. More longitudinal and
experimental studies are needed to identify the underlying causal
mechanism.
Funding
No funding was received.
Author contributions
Matteo C. Sattler: Conceptualization, Methodology, Investigation,
Project administration, Formal analysis, Validation, Visualization,
Writing - Original Draft. Tanja F¨
arber: Software, Data Curation, Writing
– Review and Editing. Katharina Trauβnig: Software, Data Curation,
Writing – Review and Editing. Gottfried K¨
oberl: Conceptualization,
Investigation, Software, Writing – Review and Editing. Christoph Paier:
Software, Data Curation, Validation, Writing – Review and Editing.
Pavel Dietz: Conceptualization, Methodology, Investigation, Writing –
Review and Editing. Mireille N. M. van Poppel: Conceptualization,
Methodology, Resources, Investigation, Project administration, Super-
vision, Writing – Review and Editing.
Data availability
The dataset used and/or analyzed during this current study is
available from the corresponding author on reasonable request.
Declaration of competing interest
None.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.mhpa.2020.100356.
References
Antoun, M., Edwards, K. M., Sweeting, J., & Ding, D. (2017). The acute physiological
stress response to driving: A systematic review. PloS One, 12(10), Article e0185517.
https://doi.org/10.1371/journal.pone.0185517.
Asmundson, G. J. G., Fetzner, M. G., Deboer, L. B., Powers, M. B., Otto, M. W., &
Smits, J. A. J. (2013). Let’s get physical: A contemporary review of the anxiolytic
effects of exercise for anxiety and its disorders. Depression and Anxiety, 30(4),
362–373. https://doi.org/10.1002/da.22043.
Avila-Palencia, I., Nazelle, A. de, Cole-Hunter, T., Donaire-Gonzalez, D., Jerrett, M.,
Rodriguez, D. A., et al. (2017). The relationship between bicycle commuting and
perceived stress: A cross-sectional study. BMJ Open, 7(6), Article e013542. https://
doi.org/10.1136/bmjopen-2016-013542.
Beattie, L., & Grifn, B. (2014). Day-level uctuations in stress and engagement in
response to workplace incivility: A diary study. Work & Stress, 16(1), 1–19. https://
doi.org/10.1080/02678373.2014.898712.
Biehl, R., Boecking, B., Brueggemann, P., Grosse, R., & Mazurek, B. (2019). Personality
traits, perceived stress, and tinnitus-related distress in patients with chronic tinnitus:
Support for a vulnerability-stress model. Frontiers in Psychology, 10, 3093. https://
doi.org/10.3389/fpsyg.2019.03093.
Blaxton, J. M., Bergeman, C. S., Whitehead, B. R., Braun, M. E., & Payne, J. D. (2017).
Relationships among nightly sleep quality, daily stress, and daily affect. Journals of
Gerontology Series B: Psychological Sciences and Social Sciences, 72(3), 363–372.
https://doi.org/10.1093/geronb/gbv060.
B¨
ocker, L., Dijst, M., & Faber, J. (2016). Weather, transport mode choices and emotional
travel experiences. Transportation Research Part A: Policy and Practice, 94, 360–373.
https://doi.org/10.1016/j.tra.2016.09.021.
Bou Samra, P., El Tomb, P., Hosni, M., Kassem, A., Rizk, R., Shayya, S., et al. (2017).
Trafc congestion and blood pressure elevation: A comparative cross-sectional study
M.C. Sattler et al.

Mental Health and Physical Activity 19 (2020) 100356
6
in Lebanon. Journal of Clinical Hypertension, 19(12), 1366–1371. https://doi.org/
10.1111/jch.13102.
Br¨
ahler, E., Mühlan, H., Albani, C., & Schmidt, S. (2007). Teststatistische Prüfung und
Normierung der deutschen Versionen des EUROHIS-QOL Lebensqualit¨
at-Index und
des WHO-5 Wohlbendens-Index. Diagnostica, 53(2), 83–96. https://doi.org/
10.1026/0012-1924.53.2.83.
Brand, S., Colledge, F., Ludyga, S., Emmenegger, R., Kalak, N., Sadeghi Bahmani, D., …
Gerber, M. (2018). Acute bouts of exercising improved mood, rumination and social
interaction in inpatients with mental disorders. Frontiers in Psychology, 9, 249.
https://doi.org/10.3389/fpsyg.2018.00249.
Brutus, S., Javadian, R., & Panaccio, A. J. (2017). Cycling, car, or public transit: A study
of stress and mood upon arrival at work. International Journal of Workplace Health
Management, 10(1), 13–24. https://doi.org/10.1108/IJWHM-10-2015-0059.
Caviedes, A., & Figliozzi, M. (2018). Modeling the impact of trafc conditions and
bicycle facilities on cyclists’ on-road stress levels. Transportation Research Part F:
Trafc Psychology and Behaviour, 58, 488–499. https://doi.org/10.1016/j.
trf.2018.06.032.
Celis-Morales, C. A., Lyall, D. M., Welsh, P., Anderson, J., Steell, L., Guo, Y., …
Gill, J. M. R. (2017). Association between active commuting and incident
cardiovascular disease, cancer, and mortality: Prospective cohort study. BMJ, 357,
j1456. https://doi.org/10.1136/bmj.j1456.
Chatterjee, K., Chng, S., Clark, B., Davis, A., Vos, J. de, Ettema, D., … Reardon, L. (2019).
Commuting and wellbeing: A critical overview of the literature with implications for
policy and future research. Transport Reviews, 40(1), 5–34. https://doi.org/10.1080/
01441647.2019.1649317.
Cohen, B. E., Edmondson, D., & Kronish, I. M. (2015). State of the art review: Depression,
stress, anxiety, and cardiovascular disease. American Journal of Hypertension, 28(11),
1295–1302. https://doi.org/10.1093/ajh/hpv047.
Cohen, S., Janicki-Deverts, D., & Miller, G. E. (2007). Psychological stress and disease.
Journal of the American Medical Association, 298(14), 1685–1687. https://doi.org/
10.1001/jama.298.14.1685.
Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress.
Journal of Health and Social Behavior, 24(4), 385–396.
Cohen, S., Kessler, R. C., & Gordon, L. U. (1995). Strategies for measuring stress in studies
of psychiatric and physical disorders. In S. Cohen, R. C. Kessler, & L. U. Gordon
(Eds.), Measuring stress: A guide for health and social scientists (pp. 3–26). New York:
Oxford University Press.
Costa, S., Ogilvie, D., Dalton, A., Westgate, K., Brage, S., & Panter, J. (2015). Quantifying
the physical activity energy expenditure of commuters using a combination of global
positioning system and combined heart rate and movement sensors. Preventive
Medicine, 81, 339–344. https://doi.org/10.1016/j.ypmed.2015.09.022.
Craig, C. L., Marshall, A. L., Sj¨
ostr¨
om, M., Bauman, A. E., Booth, M. L., Ainsworth, B. E.,
… Oja, P. (2003). International physical activity questionnaire: 12-country
reliability and validity. Medicine & Science in Sports & Exercise, 35(8), 1381–1395.
https://doi.org/10.1249/01.MSS.0000078924.61453.FB.
Dinu, M., Pagliai, G., Macchi, C., & So, F. (2019). Active commuting and multiple health
outcomes: A systematic review and meta-analysis. Sports Medicine, 49(3), 437–452.
https://doi.org/10.1007/s40279-018-1023-0.
Evans, G. W., Wener, R. E., & Phillips, D. (2002). The morning rush hour: Predictability
and commuter stress. Environment and Behavior, 34(4), 521–530. https://doi.org/
10.1177/00116502034004007.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A exible statistical
power analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39(2), 175–191. https://doi.org/10.3758/BF03193146.
Fliege, H., Rose, M., Arck, P., Levenstein, S., & Klapp, B. F. (2001). Validierung des
“Perceived Stress Questionnaire“ (PSQ) an einer deutschen Stichprobe. Diagnostica,
47(3), 142–152. https://doi.org/10.1026//0012-1924.47.3.142.
Friman, M., Olsson, L. E., Ståhl, M., Ettema, D., & G¨
arling, T. (2017). Travel and residual
emotional well-being. Transportation Research Part F: Trafc Psychology and
Behaviour, 49, 159–176. https://doi.org/10.1016/j.trf.2017.06.015.
Gatersleben, B., & Uzzell, D. (2007). Affective appraisals of the daily commute.
Environment and Behavior, 39(3), 416–431. https://doi.org/10.1177/
0013916506294032.
Global Advocacy for Physical Activity (GAPA) the Advocacy Council of the International
Society for Physical Activity and Health. (2012). Ncd prevention: Investments that
work for physical activity. British Journal of Sports Medicine, 46(10), 709–712.
https://doi.org/10.1136/bjsm.2012.091485.
Groarke, A., Curtis, R., Groarke, J. M., Hogan, M. J., Gibbons, A., & Kerin, M. (2017).
Post-traumatic growth in breast cancer: How and when do distress and stress
contribute? Psycho-Oncology, 26(7), 967–974. https://doi.org/10.1002/pon.4243.
Hutcheon, J. A., Chiolero, A., & Hanley, J. A. (2010). Random measurement error and
regression dilution bias. BMJ, 340, c2289. https://doi.org/10.1136/bmj.c2289.
IPAQ Research Committee. (2005). Guidelines for data processing and analysis of the
International Physical Activity Questionnaire (IPAQ)-short and long forms. IPAQ Group.
Kandola, A., Lewis, G., Osborn, D. P. J., Stubbs, B., & Hayes, J. F. (2020). Depressive
symptoms and objectively measured physical activity and sedentary behaviour
throughout adolescence: A prospective cohort study. The Lancet Psychiatry, 7(3),
262–271. https://doi.org/10.1016/S2215-0366(20)30034-1.
Klein, E. M., Br¨
ahler, E., Dreier, M., Reinecke, L., Müller, K. W., Schmutzer, G., …
Beutel, M. E. (2016). The German version of the Perceived Stress Scale -
psychometric characteristics in a representative German community sample. BMC
Psychiatry, 16, 159. https://doi.org/10.1186/s12888-016-0875-9.
LaJeunesse, S., & Rodríguez, D. A. (2012). Mindfulness, time afuence, and journey-
based affect: Exploring relationships. Transportation Research Part F: Trafc
Psychology and Behaviour, 15(2), 196–205. https://doi.org/10.1016/j.
trf.2011.12.010.
Lee, P. H., Macfarlane, D. J., Lam, T. H., & Stewart, S. M. (2011). Validity of the
international physical activity questionnaire short form (IPAQ-SF): A systematic
review. International Journal of Behavioral Nutrition and Physical Activity, 8, 115.
https://doi.org/10.1186/1479-5868-8-115.
Lee, I. M., Shiroma, E. J., Lobelo, F., Puska, P., Blair, S. N., & Katzmarzyk, P. T. (2012).
Effect of physical inactivity on major non-communicable diseases worldwide: An
analysis of burden of disease and life expectancy. The Lancet, 380(9838), 219–229.
https://doi.org/10.1016/S0140-6736(12)61031-9.
Levenstein, S., Prantera, C., Varvo, V., Scribano, M. L., Berto, E., Luzi, C., et al. (1993).
Development of the perceived stress questionnaire: A new tool for psychosomatic
research. Journal of Psychosomatic Research, 37(1), 19–32. https://doi.org/10.1016/
0022-3999(93)90120-5.
Loyen, A., van Hecke, L., Verloigne, M., Hendriksen, I., Lakerveld, J., Steene-
Johannessen, J., … van der Ploeg, H. P. (2016). Variation in population levels of
physical activity in European adults according to cross-European studies: A
systematic literature review within DEDIPAC. International Journal of Behavioral
Nutrition and Physical Activity, 13, 72. https://doi.org/10.1186/s12966-016-0398-2.
Martin, A., Goryakin, Y., & Suhrcke, M. (2014). Does active commuting improve
psychological wellbeing? Longitudinal evidence from eighteen waves of the British
Household Panel survey. Preventive Medicine, 69, 296–303. https://doi.org/10.1016/
j.ypmed.2014.08.023.
McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role
of the brain. Physiological Reviews, 87(3), 873–904. https://doi.org/10.1152/
physrev.00041.2006.
Montero-Marin, J., Piva Demarzo, M. M., Pereira, J. P., Olea, M., & García-Campayo, J.
(2014). Reassessment of the psychometric characteristics and factor structure of the
’Perceived Stress Questionnaire’ (PSQ): Analysis in a sample of dental students. PloS
One, 9(1), Article e87071. https://doi.org/10.1371/journal.pone.0087071.
Morris, E. A., & Hirsch, J. A. (2016). Does rush hour see a rush of emotions? Driver mood
in conditions likely to exhibit congestion. Travel Behaviour & Society, 5, 5–13.
https://doi.org/10.1016/j.tbs.2015.07.002.
Nazelle, A. de, Nieuwenhuijsen, M. J., Ant´
o, J. M., Brauer, M., Briggs, D., Braun-
Fahrlander, C., … Lebret, E. (2011). Improving health through policies that promote
active travel: A review of evidence to support integrated health impact assessment.
Environment International, 37(4), 766–777. https://doi.org/10.1016/j.
envint.2011.02.003.
Physical Activity Guidelines Advisory Committee. (2018). 2018 physical activity guidelines
advisory committee scientic report. Washington, DC: US Department of Health and
Human Services.
US Department of Health and Human Services. (2018). Physical activity guidelines for
Americans (2nd ed.). Wasington, DC: US Department of Health and Human Services.
Rasmussen, C., Knapp, T. J., & Garner, L. (2000). Driving-induced stress in urban college
students. Perceptual & Motor Skills, 90(2), 437–443. https://doi.org/10.2466/
pms.2000.90.2.437.
Reis, D., Lehr, D., Heber, E., & Ebert, D. D. (2019). The German version of the perceived
stress scale (PSS-10): Evaluation of dimensionality, validity, and measurement
invariance with exploratory and conrmatory bifactor modeling. Assessment, 26(7),
1246–1259. https://doi.org/10.1177/1073191117715731.
Rissel, C., Curac, N., Greenaway, M., & Bauman, A. (2012). Physical activity associated
with public transport use–a review and modelling of potential benets. International
Journal of Environmental Research and Public Health, 9(7), 2454–2478. https://doi.
org/10.3390/ijerph9072454.
R¨
onnlund, M., Vestergren, P., Stenling, A., Nilsson, L.-G., Bergdahl, M., & Bergdahl, J.
(2015). Dimensionality of stress experiences: Factorial structure of the Perceived
Stress Questionnaire (PSQ) in a population-based Swedish sample. Scandinavian
Journal of Psychology, 56(5), 592–598. https://doi.org/10.1111/sjop.12214.
Rothbard, N. P., & Wilk, S. L. (2011). Waking up on the right or wrong side of the bed:
Start-of-Workday mood, work events, employee affect, and performance. Academy of
Management Journal, 54(5), 959–980. https://doi.org/10.5465/amj.2007.0056.
Rubin, D. B. (1987). Multiple Imputation for nonresponse in surveys. Wiley series in
probability and mathematical statistics Applied probability and statistics. New York, NY:
Wiley. https://doi.org/10.1002/9780470316696.
Rüger, H., Pfaff, S., Weishaar, H., & Wiernik, B. M. (2017). Does perceived stress mediate
the relationship between commuting and health-related quality of life?
Transportation Research Part F: Trafc Psychology and Behaviour, 50, 100–108.
https://doi.org/10.1016/j.trf.2017.07.005.
Sahlqvist, S., Song, Y., & Ogilvie, D. (2012). Is active travel associated with greater
physical activity? The contribution of commuting and non-commuting active travel
to total physical activity in adults. Preventive Medicine, 55(3), 206–211. https://doi.
org/10.1016/j.ypmed.2012.06.028.
Schisterman, E. F., Cole, S. R., & Platt, R. W. (2009). Overadjustment bias and
unnecessary adjustment in epidemiologic studies. Epidemiology, 20(4), 488–495.
https://doi.org/10.1097/EDE.0b013e3181a819a1.
Seaman, S. R., & White, I. R. (2013). Review of inverse probability weighting for dealing
with missing data. Statistical Methods in Medical Research, 22(3), 278–295. https://
doi.org/10.1177/0962280210395740.
Shephard, R. J. (2008). Is active commuting the answer to population health? Sports
Medicine, 38(9), 751–758. https://doi.org/10.2165/00007256-200838090-00004.
Sliwinski, M. J., Almeida, D. M., Smyth, J., & Stawski, R. S. (2009). Intraindividual
change and variability in daily stress processes: Findings from two measurement-
burst diary studies. Psychology and Aging, 24(4), 828–840. https://doi.org/10.1037/
a0017925.
Sposato, R. G., R¨
oderer, K., & Cervinka, R. (2012). The inuence of control and related
variables on commuting stress. Transportation Research Part F: Trafc Psychology and
Behaviour, 15(5), 581–587. https://doi.org/10.1016/j.trf.2012.05.003.
M.C. Sattler et al.

Mental Health and Physical Activity 19 (2020) 100356
7
Topp, C. W., Østergaard, S. D., Søndergaard, S., & Bech, P. (2015). The WHO-5 well-
being index: A systematic review of the literature. Psychotherapy and Psychosomatics,
84(3), 167–176. https://doi.org/10.1159/000376585.
Van Hooff, M. L. M. (2015). The daily commute from work to home: Examining
employees’ experiences in relation to their recovery status. Stress and Health : Journal
of the International Society for the Investigation of Stress, 31(2), 124–137. https://doi.
org/10.1002/smi.2534.
Vandenbroucke, J. P., Elm, E. von, Altman, D. G., Gøtzsche, P. C., Mulrow, C. D.,
Pocock, S. J., … Egger, M. (2014). Strengthening the reporting of observational
studies in Epidemiology (STROBE): Explanation and elaboration. International
Journal of Surgery, 12(12), 1500–1524. https://doi.org/10.1016/j.ijsu.2014.07.014.
Wardman, M., Hine, J., & Stradling, S. G. (2001). Interchange and travel choice, Volumes 1,
2 and executive summary. Edinburgh: Scottish Executive Central Research Unit.
Wener, R. E., & Evans, G. W. (2011). Comparing stress of car and train commuters.
Transportation Research Part F: Trafc Psychology and Behaviour, 14(2), 111–116.
https://doi.org/10.1016/j.trf.2010.11.008.
World Health Organization. (1998). Wellbeing measures in primary health care/the
DEPCARE project. Report on a WHO meeting. Copenhagen: WHO Regional Ofce for
Europe.
World Health Organization. (2002). A physically active life through everyday transport.
Copenhagen: WHO Regional Ofce for Europe.
M.C. Sattler et al.