ORIGINAL RESEARCH
Journal of Physical Activity and Health, 2011, 8,1025-1033
©2011 Human Kinetics, Inc,
Cycling to School and Cardiovascular Risk Factors:
A Longitudinal Study
Lars Bo Andersen, Niels Wedderkopp, Peter Kristensen, Niels Christian Möller,
Karsten Froberg, and Ashley R. Cooper
Background: Cycling to school may potentially increase physical activity level in sedentary children. Transport
to school occur twice a day and could improve cardiovascular health in children. Commuter cycling is associ-
ated vkiith lower mortality and cardiovascular disease rate in adults, but limited evidence exists in children.
Methods: Participants were 334 children (age 9.7 ± 0,5 years) who were followed up 6 years later. Mode of
travel to school was investigated by questionnaire. Cardiovascular (CVD) risk factors were compared by mode
of travel to school both at baseline and at follow up and for subjects who changed mode of transportation. No
difference was found between walkers and passive travelers, and these groups were merged in the analysis.
Results: A consistent pattern of better CVD risk factor profile in commuter cyclists compared with children
using other means of transport was found. Participants, who did not cycle to school at baseline, and who had
changed to cycling at follow up, were fitter, had better cholesterol/HDL ratio, better glucose metabolism, and a
lower composite CVD risk factor score than those who did not cycle at either time point. Conclusion: Cycling
to school may contribute to a better cardiovascular risk factor profile in young people.
Keywords: accelerometry, community-based research, fitness, youth, cardiovascular health
A decrease in habitual physical activity is suggested
to be a contributor to rising levels of childhood over-
weight and obesity, although there are limited direct data
to describe how children's physical activity has changed
over recent decades,' Indirect evidence for a decline
in overall physical activity comes from transportation
surveys, which have recorded a reduction in the propor-
tion of journeys taken by foot and an increase in car
travel,^ '^ These trends are refiected in the decline in active
travel to school reported in many countries. In the U,S,
children's active commuting to school declined by 37%
between 1977 and 1995, and current estimates suggest
that approximately 5 to 10% of children aged 5 to 15
years walk to school with fewer than 2 to 4% cycling,'''^
This decline has not happened in countries with a strong
tradition of active commuting. In Denmark, cycling to
school has not changed since 1983, with 63% of 16- to
19-year-olds cycling to school in 1983* compared with
66% of 15-year-olds in 1997^ and 63% in 2003.8 The
potential opportunity to reverse the decline has resulted
in the journey to school receiving attention as a clear
Andersen, Wedderkopp, Kristensen, Möller, and Froberg are
with the Center for Research in Childhood Health, Institute
of Sports Science and Clinical Biomechanics, University of
Southern Denmark, Odense, Denmark, Cooper is with the
Dept of Exercise, Nutrition, and Health Sciences, University
of Bristol, Bristol, United Kingdom,
target for intervention to increase young people's daily
physical activity,
A number of observational studies in the U.S.,''
Europe''" and elsewhere'''^ have used objective mea-
surement of physical activity with accelerometers or
pedometers to investigate the association between how
children travel to school and total physical activity. The
majority (11 of 13) of these studies have shown that
children who walk or cycle to school engage in more
physical activity than those who travel by other means.'^
As objectively measured physical activity has been shown
to be inversely associated with fatness''* and positively
associated with a better profile of cardiovascular (CVD)
risk factors in children and adolescents," a number of
studies have investigated whether active commuting to
school may be associated with reduced levels of over-
weight or obesity. To date, no consistent association with
body mass index (BMI) or other measures of adiposity
have been reported.'^'^ In addition, no studies investigat-
ing associations between active travel and other direct
health parameters that might be influenced by physical
activity such as metabolic risk factors for cardiovascular
disease have been reported in children.
In contrast, active commuting has been associated
with CVD risk factors in adults. In a large sample of
Chinese adults, daily walking or cycling to and from
work was inversely associated with serum total choles-
terol, low-density lipoprotein cholesterol and triglycéride
concentrations among men, and positively associated
1025

1026 Andersen et al
with high-density lipoprotein cholesterol concentrations
among women when compared with individuals com-
muting by bus." The Coronary Artery Risk in Young
Adults (CARDIA) study in the US explored associations
between CVD risk factors and active commuting in 2364
participants.'* Men who reported active commuting were
fitter than those who did not, and active commuting was
associated with a reduced likelihood of obesity (OR 0.5)
and improved CVD risk factor profile though these asso-
ciations disappeared after adjustment for BMI. In women,
despite higher fitness in active commuters, associations
with CVD risk factors were not seen. The separate effects
of walking and cycling were not reported in either study,
but it is likely that the physiological effects of walking
and cycling differ. In adults, commuter cycling elicits a
higher mean heart rate than walking, is associated with
greater increases in fitness" and is associated with a sub-
stantial reduction in risk of all-cause mortality.^° In young
people, those who cycle to school are fitter than those who
travel by foot or by motorized transport, and in observa-
tional studies a change in travel mode from noncycling
at baseline to cycling 6 years later is associated with an
improvement in fitness.^-^ Since higher cardiorespiratory
fitness is associated with a better CVD risk factor profile
in young people^' it is possible that the higher fitness seen
in commuter cyclists may be also associated with better
levels of CVD risk factors.
The aim of this study was to use longitudinal data
from the European Youth Heart Study to determine the
CVD risk factor profile in children and adolescents who
cycled to school compared with noncyclists, and to
investigate whether the levels of risk factors change in
participants who changed travel mode to school between
baseline and follow-up.
Methods
Participants
This paper describes the participants from the Danish
arm of the EYHS, who were 9 years old in 1997-98
and 15 years old when recontacted in 2003-04. For the
original sample, all schools (n = 35) in the region of
Odense, Denmark, were stratified according to location
(urban, suburban, rural), and socioeconomic profile of
uptake area (high, middle, low). A proportional, two-stage
cluster sample of children was taken from each stratum.
Twenty-five of 28 sampled schools agreed to participate.
In all, 771 children were invited to participate in EYHS-I
in 1997 to 98, of whom 589 (76.4%; 310 girls and 279
boys) consented. Six years later, 384 of these children
(212 girls and 172 boys) were reexamined in EYHS-II.
Twelve participants provided nonfasting blood samples
at one of the occasions and were excluded, and another
38 participants did not reach a maximal heart rate of 185
bpm during the fitness test. There were no differences
in BMI between participants excluded for an invalid
fitness test and subjects with valid fitness tests in 9- or
15-year-olds. A tendency was seen for fewer cyclists
among subjects excluded for invalid fitness test (P = . 1 ).
The sample for the current study was 187 girls and 147
boys with complete data. Written informed consent was
obtained from the parent/guardian of each participant,
who also gave verbal assent to taking part.
Measurements
All measurements were identical for EYHS-I and II.
Completion of measurements took 2 to 3 hours. Height
was measured to the nearest 1mm in bare or stockinged
feet with a transportable Harpenden stadiometer and
weight was measured to the nearest 0.1kg using a cali-
brated beam scale. BMI was calculated as weight (kg)
/ height^ (m). The sum of the thickness of 4 skinfolds
(biceps, triceps, subscapular, and suprailiac) was mea-
sured using a Harpenden caliper, with the mean of 3 mea-
surements used at each site. Pubertal stage was assessed
according to Tanner stages using a 5-point scale of pic-
tures—girls according to breast development and pubic
hair growth and boys according to genital development
and pubic hair growth. Assessments were carried out by
a trained researcher of the same gender as the participant
and privacy was maintained at all times.
To investigate travel mode, participants completed
a computerized questionnaire including the questions:
"How do you usually travel to school?" (response options:
by car or motorcycle, by bus or train, by bicycle, by foot)
with a similar question for travel home, and "How long
does it usually take you to travel to school from your
home?" (response options: less than 5 minutes, 5 to 15
minutes, 16 to 30 minutes, 31 minutes to I hour, more
than 1 hour). In the current study the responses "by car
or motorcycle, by bus or train" were combined into the
group "passive transport," and travel mode was defined
by the way the participant usually traveled to school.
Blood pressure was measured using a Dinamap
pediatric/adult neonatal vital signs monitor (model XL,
Critikron, Inc., Tampa, FL). Five measurements were
taken at two-minute intervals with the mean of the final
3 measurements used in all analyses.
Blood samples were obtained after an overnight fast
and stored at -80°C before analysis. All samples were
analyzed for blood lipids, glucose, and insulin at clinical
pathology accredited laboratories (Bristol and Cam-
bridge, UK).^ 2 Insulin resistance was estimated according
to the homoeostasis model assessment (HOMA) as the
product of fasting glucose (mmol/L) and insulin (|iU/
mL) divided by the constant 22.5.^^
Cardio-respiratory fitness, defined as maximal power
output per kilogram (W^ax-kg"'), was determined using
a cycle ergometer test with progressively increasing
workload until exhaustion, on an electronically braked
cycle ergometer (Monark 839 Ergomedic). This test has
been validated in both children and adolescents with a
low test-retest coefficient of variation (2.5% to 4.8%) and
is highly correlated (r= .90 in boys and r = .95 in girls)
with directly measured V02max.^' The cycle ergometer