Behavioural Brain Research xxx (2005) xxx–xxx
Short communication
Automated home cage observations as a tool to measure the
effects of wheel running on cage floor locomotion
Leonie de Visser, Ruud van den Bos, Berry M. Spruijt∗
Department of Animals, Science and Society, Ethology and Animal Welfare, Faculty of Veterinary medicine,
Utrecht University, Yalelaan 2, NL-3584 CM Utrecht, The Netherlands
Received 16 September 2004; received in revised form 24 November 2004; accepted 6 December 2004
Abstract
This paper introduces automated observations in a modular home cage system as a tool to measure the effects of wheel running on the
time distribution and daily organization of cage floor locomotor activity in female C57BL/6 mice. Mice (n = 16) were placed in the home
cage system for 6 consecutive days. Fifty percent of the subjects had free access to a running wheel that was integrated in the home cage.
Overall activity levels in terms of duration of movement were increased by wheel running, while time spent inside a sheltering box was
decreased. Wheel running affected the hourly pattern of movement during the animals’ active period of the day. Mice without a running
wheel, in contrast to mice with a running wheel, showed a clear differentiation between novelty-induced and baseline levels of locomotion as
reflected by a decrease after the first day of introduction to the home cage. The results are discussed in the light of the use of running wheels
as a tool to measure general activity and as an object for environmental enrichment. Furthermore, the possibilities of using automated home
cage observations for e.g. behavioural phenotyping are discussed.
© 2004 Published by Elsevier B.V.
Keywords: Wheel running; Locomotor activity; Home cage observations; Behavioural phenotyping; Mice
Wheel running is an extensively used behavioural parameter in neurobiological studies. Roughly, three categories
can be distinguished with respect to research that is currently being done on or using wheel running behaviour.
First, a vast number of studies focus on the effects of experimental manipulations (e.g. scheduled feeding, photoperiod shifts, drug effects, housing conditions) on the circadian
rhythms using wheel running as an indicator of general activity [1,3,7,9,14,21,23,24]. A second category of research
addresses the effects of wheel running in terms of ‘physical
activity’ or ‘exercise’ on neurobiological processes, such as
brain plasticity, neurogenesis, sleep-wakefulness and aging
[5,8,12,16,19,26,31,34]. Third, some studies on the high motivation of rodents to perform wheel running have put forward
the suggestion that running wheels might improve welfare
∗
Corresponding author. Tel.: +31 30 253 2574; fax: +31 30 253 9227.
E-mail address: [email protected] (B.M. Spruijt).
0166-4328/$ – see front matter © 2004 Published by Elsevier B.V.
doi:10.1016/j.bbr.2004.12.004
when provided to laboratory animals as an extended opportunity for locomotor activity [10,22,27].
In the majority of these investigations, wheel running activity is considered as being representative for, or indicative
of, general locomotor activity or exploratory behaviour. However, recent studies suggest running wheel activity to be naturally rewarding and reinforcing [4]; this behaviour may be
described as an incentive-induced motivated behaviour similar to the intake of addictive drugs. Moreover, it is known
to affect brain organization in terms of neuronal activation of
reward systems [4,25,32]. Although it still remains unclear
what the underlying causal factors are for the high rewarding
aspects of wheel running, it is indicated that general activity
and wheel running are not similar [28]. Since wheel running
as a putative, incentive induced motivated behaviour may interact with other motivational systems, it cannot be used without any due consideration to its potential disruptive effects in
experimental set ups. Also its welfare enhancing value needs
2
L. de Visser et al. / Behavioural Brain Research xxx (2005) xxx–xxx
further attention as its display may be at the expense of other
behaviour.
Most studies on running wheel activity are performed
while recording wheel revolutions only and neglecting other
locomotor behaviours. Therefore, the present study investigated the effect of voluntary wheel running on both the time
distribution and daily organization of ‘normal’ cage floor locomotor activity by using an integrated approach in an automated home cage observation system. This system allows for
continuous, simultaneous recordings of both running wheel
activity and cage floor locomotor activity. This method has
the additional advantage of allowing differentiation between
novelty-induced and baseline behaviour. The results are discussed in the light of the use of running wheels as a tool to
measure basal activity levels and as an object for enriching
the home cage environment.
Apart from trying to acquire more knowledge on running
wheel behaviour, the second aim of this study is to introduce the home cage system as a tool to measure interactions
between different aspects of behaviour under both novelty
and baseline conditions in a more complex environment. The
possibilities of this method for e.g. behavioural phenotyping
studies are discussed at the end of this paper.
Sixteen female mice were used of the C57BL/6OlaHsd
inbred strain. All animals were obtained from the breeding stock at the laboratory animal facility of the Department
of Animals, Science and Society, Utrecht University, The
Netherlands. Subjects were wild-type animals derived from
a mu-opioid receptor knockout strain (see ref. [15] for details
on this strain). The mice were 10–12 weeks of age at the onset of the experiment, and weighed 20–25 g during the experimental period. All subjects were group housed (two–three
animals per cage) and maintained under a reversed light–dark
cycle (white light: 19:00–07:00 h, red light: 07:00–19:00 h)
with food and water available ad libitum. Per cage, animals were provided with a shelter, tissues (Kleenex® , SmithKline) and paper shreds (Enviro Dri® , TecniLab, The Netherlands) as environmental enrichment. Humidity was kept at
a constant level and ambient temperature was maintained
at 21.0 ± 2.0 ◦ C. The Animal Ethical Committee of Utrecht
University approved the experiment.
Home cage behaviour and running wheel activity was automatically recorded by videotracking in specially designed
cages (prototype of the PhenoTyper® , Noldus Information
Technology, Wageningen, The Netherlands). On top of each
cage a unit containing the hardware needed for videotracking
was placed. The unit consisted of a built-in digital infraredsensitive video camera and infrared lighting sources. The infrared sources provided a constant and even illumination of
the ground floor of the cage. Normal room lighting did not
influence these lighting conditions because of an infrared filter that was placed in front of the camera lens. This method
allowed for continuous behavioural recordings in both dark
and light periods. EthoVision 3.0 (Noldus Information Technology, Wageningen, The Netherlands) was used as videotracking software.
Cages (30 cm × 30 cm × 35 cm) were made of Perspex®
walls and an aluminium floor. They contained a fixed feeding
station and water bottle. The walls were connected in such a
way that one could easily switch between walls and change
the design of the cage depending on the specific research
question.
Four home cage test (HCT) cages were connected to a
single PC. The video images of all four cages were converted
into a single video image by a Quad unit.
The mice were assigned to either of two groups, each containing eight animals: a running wheel group (RW) and a
control group with no running wheel (no RW). For this experiment, 50% of the cages were provided with a Perspex®
wall with a running wheel attached. In the other cages, a plain
wall was placed in the same position as the wall with running
wheel. One group was placed in the HCT-cages with free
access to a running wheel. The others were placed in similar HCT-cages but were not provided with a running wheel.
The running wheels had a perimeter of 38 cm and a circular
running surface consisting of steel rods. The running wheels
were connected to an Event Logging Interface (Noldus Information Technology) that converted the sensor signal to
keyboard input. In EthoVision, the keyboard input was used
to automatically count the number of revolutions per cage
during the observation period.
Each cage contained a single animal. All cages were
provided with bedding material (sawdust) and a tissue
(Kleenex® , Kimberly-Clark) and paper shreds (Enviro Dri® ,
TecniLab) as environmental enrichment. Furthermore, all animals had access to a shelter (height: 10 cm, ø 9 cm) of nontransparent material, in which they could sleep during the day
and hide in case of unexpected environmental events such as
laboratory noise. In EthoVision, the shelter could be defined
as a “hidden zone”; the program could distinguish between
the behaviours “in shelter” and “on shelter”. This allowed
for detailed information on the duration and frequency with
which the animals visited their shelter. Observations of home
cage behaviour and running wheel activity were recorded for
6 consecutive days (144 h).
Running wheel activity was measured by automatically
counting the number of revolutions during the experiment.
Afterwards, frequencies of revolutions were converted into
distance travelled using the running wheel perimeter.
To study the effect of running wheel activity on home
cage behaviour, several parameters with regard to locomotor activity were computed using the EthoVision 3.0 analysis
module. Parameters used were: ‘sheltertime’ (percentage of
time the animals spends inside its shelter), ‘cage floor movement’ (percentage of time the animal spends moving on the
cage floor), ‘running wheel activity’ (percentage of time the
animal spends in the running wheel), ‘total movement’ (cage
floor movement and running wheel activity taken together),
‘distance moved’ (distance travelled by the animal when outside the shelter in cm/h) and ‘velocity’ (the speed of moving
during cage floor movement in cm/s). The choice of parameters used for analysis was based on prior experience with
L. de Visser et al. / Behavioural Brain Research xxx (2005) xxx–xxx
3
Fig. 1. Time distribution of locomotor activities during the dark period for mice without a running wheel (no RW; n = 5) and for mice with a running wheel
(RW; n = 6). The duration of parameters sheltertime, cage floor movement and running wheel activity are expressed as a percentage of total time. Mean and
S.E.M. of 12-h bins are used for 6 consecutive days (1–6). Significance levels for comparisons between days were set at p ≤ 0.01 (paired t-test), see text for
details.
the software package and pilot studies (unpublished data).
All parameters were calculated in 1-h bins and subsequently
lumped into 12-h fragments to distinguish between dark and
light periods. For circadian rhythmicity no lumping was performed, instead hourly values were used.
Statistical analyses were conducted using SPSS 10.0 for
Windows. Repeated measures ANOVAs were performed
with the factors: group (no running wheel versus running
wheel) as between-subjects factor and phase (light or dark
period of day) and day (6 consecutive days of observations)
as within-subjects factors. Post-hoc comparisons between
groups and days were performed using the appropriate ttest. Significance levels were set at p ≤ 0.05 when significant, 0.05 < p ≤ 0.10 when a trend and p > 0.10 when not
significant. For post-hoc comparisons between days, the significance level was set at p ≤ 0.01 after Bonferroni-correction
(five comparisons for each day).
Five animals were discarded from the analysis. Two animals, one of each group, were left out due to technical problems with the system’s hardware. After the experiment, it was
decided to discard three more animals because they built a
nesting place outside the shelter. For these animals, considerably less space was left on the ground floor for locomotion,
which would affect the results. Furthermore, the reason behind the decision made by the animals to sleep uncovered is
unclear. Preferably these animals should be treated as a sep-
arate group but in the present experiment numbers were too
small to produce reliable results. This ultimately leaves the
number of animals at five for the control group (no RW) and
six for the running wheel group (RW).
Significant results from three-way ANOVA (factors: group
(between-subjects), phase (within-subjects) and day (withinsubjects)) are discussed for both the time distribution of locomotor activity and the development of locomotor activity over
the course of the experiment. Post-hoc comparisons between
days (performed with paired-samples t-test) are discussed in
terms of the differences of day 2–6 compared to the first day
of the experiment, as hardly any significant differences were
found amongst day 2–6.
Fig. 1 shows the time distribution of locomotor activity and
running wheel activity for the dark periods of 6 consecutive
days. Time distribution for sheltertime, cage floor movement
and running wheel activity are represented in terms of their
relative duration. For animals with a running wheel, total
movement consists of both cage floor movement and running wheel activity. As no differences were found on these
parameters during the light period, values are not shown in
Fig. 1 but instead summed up in Table 1. Mice with access to a running wheel spent significantly less time moving on the cage floor (main-effect ANOVA, F(1,9) = 13.072;
p ≤ 0.006), but more time on total movement (cage floor
movement plus running wheel activity; main-effect ANOVA,
Table 1
Light period values for parameters sheltertime, cage floor movement and running wheel activity for mice without and with access to a running wheel
Parameter
Group
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
Mean
S.E.M.
Mean
S.E.M.
Mean
S.E.M.
Mean
S.E.M.
Mean
S.E.M.
Mean
S.E.M.
Sheltertime
No RW
RW
93.31
91.21
1.43
2.12
92.53
94.33
0.94
1.18
90.46
95.36
4.01
1.12
93.46
95.72
1.29
0.91
92.27
96.18
2.19
1.08
94.62
97.44
1.34
1.08
Cage floor movement
No RW
RW
1.69
1.23
0.13
0.39
1.90
0.78
0.34
0.19
1.02
0.55
0.14
0.18
1.03
0.47
0.22
0.17
2.96
0.40
1.31
0.21
0.94
0.29
0.18
0.12
Running wheel activity
No RW
RW
n.a.
4.17
n.a.
1.41
n.a.
1.66
n.a.
0.62
n.a.
1.17
n.a.
0.38
n.a.
1.03
n.a.
0.26
n.a.
0.86
n.a.
0.36
n.a.
0.64
n.a.
0.27
Mean and S.E.M. of 12-h bins are used for 6 consecutive days. All parameters are durations and expressed as a percentage of total time.
4
L. de Visser et al. / Behavioural Brain Research xxx (2005) xxx–xxx
F(1,9) = 12.261; p ≤ 0.007) and less time inside the shelter
(main-effect ANOVA, F(1,9) = 13.554; p ≤ 0.005) during the
dark period of the day compared to mice without a running wheel. Mice with running wheels ran approximately
5.0 km/24 h (data not shown) and almost all running was performed during the dark period.
The parameter cage floor movement is further split up
in terms of the parameters distance moved and velocity
of moving, for both dark and light period of the day, in
Fig. 2(a)–(d). The distance travelled while moving on the
cage floor (Fig. 2(a) and (b)) was significantly higher in mice
without running wheel during both light and dark period
(main-effect ANOVA, F(1,9) = 11.208; p ≤ 0.009) compared
to mice with a running wheel. However, during the dark period, this difference was only significant on the first 4 days
of the experiment. No main group effect was seen in velocity
(Fig. 2(c) and (d)) during either dark or light period (maineffect ANOVA, F(1,9) = 0.318; p ≤ 0.587).
Group differences were observed in the development of
different aspects of locomotor activity over the course of the
experiment. This was reflected by significant group × day
interactions from a three-way ANOVA for the parameters total movement (F(1,9) = 2.800; p ≤ 0.028), distance
moved (F(1,9) = 8.837; p < 0.001) and velocity (F(1,9) = 3.370;
p ≤ 0.011) and a trend for the parameter sheltertime
(F(1,9) = 2.103; p ≤ 0.082).
In general, animals without a running wheel in their cage
showed elevated activity levels during the dark period in
terms of duration of cage floor movement (Fig. 1) and distance moved (Fig. 2) on the first day. This was followed by a
significant decrease in duration of cage floor movement from
day 2 until day 6 and in distance moved from day 3 until
day 6. Sheltertime increased significantly after the first day
until the end of the experiment, but only during the dark period. All parameters remained at a stable level after the first 1
or 2 days. During the light period, no significant differences
between days were noted.
Animals with access to a running wheel also showed an
initial decrease in total movement, distance moved and an
increase in sheltertime on day 2 during the dark period. But
in this group, levels of total movement started to rise again
after 4 days (reflected by non-significant differences on day
5 and 6 compared to day 1). This increase in total movement
is caused by an increase in time spent in the running wheel,
while levels of cage floor movement remain stable after the
first day. Similarly, after the initial increase on day 2, sheltertime decreased from day 5 on (reflected by non-significant
differences compared to day 1) during the dark period.
The parameter velocity behaved differently compared to
the other parameters. An increase in velocity for mice without a running wheel was observed on day 3, but not on the
days thereafter. The increase in velocity for the group with
running wheels was significant from day 3 until the end of
the experiment. During the light period, no differences were
found between days for mice without a running wheel. However, mice with a running wheel did tend to show a decrease
Fig. 2. Cage floor movement is further split up in (a) and (b) distance moved in cm/h, and (c) and (d) velocity in cm/s for mice without a running wheel (no
RW) and mice with access to a running wheel (RW). Mean and S.E.M. of 12-h bins for both dark and light period for 6 consecutive days. * p ≤ 0.01 when
compared to day 1 (paired t-test).
L. de Visser et al. / Behavioural Brain Research xxx (2005) xxx–xxx
in velocity during the light period on day 5 and 6. This resulted in an increase in the difference between dark and light
period on the last 2 days of the experiment.
Fig. 3(a) and (b) show duration of movement and running wheel revolutions for 1-h intervals during both dark and
light period of 24 h. Values shown are means for day 2–6.
Day 1 was excluded from this figure due to very high activity levels in the first hours after introduction in the cage
(data not shown). Fig. 3(a) and (b) differ with respect to the
parameter that was used for analysis. In Fig. 3(a), the parameter cage floor movement is set against time, whether in
Fig. 3(b) the parameter total movement is used, thus including the time spent in the running wheel. Obviously, for mice
without running wheels the total movement is equal to the
cage floor movement. These data show that activity patterns
during dark phase differ in mice with and without running
wheel. The pattern of movement in mice without a running
wheel consists of two peaks and a decrease in activity around
3 h before lights on. In mice with running wheels, this pattern is eliminated when looking at movement on the cage
floor (Fig. 3(a)). When running wheel activity is included
(Fig. 3(b)), a pattern emerges that is different from the activity pattern of mice without a running wheel.
The present study introduced automated home cage observations as a tool to measure the effect of wheel running on
different aspects of locomotor activity in mice. The method
allows for simultaneous recordings of both running wheel
activity and other locomotor activities during both the dark
and light period for consecutive days. Wheel running seems
to induce both an increase in overall activity levels and a
change in activity pattern during the animal’s active period
of day.
The results show that free access to a running wheel in
a home cage environment is competing with time spent on
other locomotor activities, such as cage floor movement (in
terms of duration, distance travelled and speed of moving)
and sheltering behaviour. This was reflected by an increase
in levels of total movement, thus including the time spent in
the running wheel, and a decrease in cage floor movement and
5
sheltertime for wheel running animals. Notably, these effects
were only evident during the dark period, i.e. the animal’s active period of the day. These results are in concordance with
earlier findings by Harri et al. [13] and Koteja et al. [18]. The
latter group selectively bred a strain of mice that ran 2.2 times
as many revolutions per day (12.1 km/day) than their normal
counterparts (5.5 km/day). In the selection experiment, mice
from both lines had access to either a free wheel or a locked
wheel in an adjacent cage. Running wheel activity in control mice with access to a free wheel reached similar levels
as found in the present study (5.0 km/day). In this control
line, mice with free wheels spent less time sleeping and had
increased total activity levels. Cage locomotion (scored by
human observers in blocks of 2 h, twice a day) was decreased
at the expense of wheel time. These results are in agreement
with findings in the present study, although the experimental
set-up differed in the sense that all mice in the selection experiment were provided with a running wheel. For the present
study it was decided not to provide the control group with a
running wheel because when locked, the running wheel may
be an object used for climbing which could affect home cage
activity in a different, and unknown way, than when it is used
for running. The physical presence of the running wheel in
terms of a slight decrease in ground floor area that could be
used for locomotion did not affect the results (a correction
showed no differences on either parameter).
To our knowledge, velocity or speed of moving is not used
as a parameter in earlier studies. There was a clear increase in
velocity over the course of the experiment for the mice with
a running wheel that was absent in the mice without running
wheels. Interpretation of this finding should be made with
caution until it is supported through replication of the present
data. However, one might argue that when wheel running is
competing with time to be spent on cage floor movement,
increasing speed of moving is an efficient way to decrease
the time needed to reach feeding and drinking places and for
monitoring the environment. Another explanation for the increased velocity in wheel running animals in this experiment
might be that wheel running brings the animal in a state of
Fig. 3. (a) Twenty-four-hour distribution of duration of cage floor movement for mice without (n = 5) and with (n = 6) access to a running wheel expressed
as a percentage of total time. Mean and S.E.M. of day 2–6 are used in 1-h bins. Black bar represents the 12-h dark period of the day, whereas the white bar
represents the 12-h light period of the day. (b) Twenty-four-hour distribution of duration of total movement (cage floor movement and running wheel activity
added up) for mice without (n = 5) and with (n = 6) access to a running wheel expressed as a percentage of total time. Mean and S.E.M. of day 2–6 are used in
1-h bins. Black bar represents the 12-h dark period of the day, whereas the white bar represents the 12-h light period of the day.
6
L. de Visser et al. / Behavioural Brain Research xxx (2005) xxx–xxx
high arousal or hyperactivity that is not only affecting general
activity levels but also speed of moving.
Running wheel availability had a profound influence on
the hourly distribution of movement during the dark period of
the day. In terms of cage floor movement, the distinct pattern
found in mice without a running wheel was similar to home
cage activity patterns found in other studies [17,29]. This pattern consisted of a sharp increase in movement following the
onset of darkness, a clear decrease in movement around twothird of the dark period followed by a small increase until the
end of the dark period. In mice with a running wheel, this pattern was absent when considering cage floor movement only
and was replaced by a different pattern that was almost completely made up by running wheel movement. The pattern of
cage floor movement in these mice was a mere “flat line”,
while the pattern of total movement reached its peak only
halfway the dark period followed by a decrease lasting until
the end of the dark period. The same hourly pattern was found
for sheltertime (data not shown), supporting the suggestion
that running wheel availability is affecting the organization
of locomotor activity, but only during the animal’s active period. Increases in running wheel activity and velocity of cage
floor movement at the end of the experiment, indicate that the
time period of 6 days might be too short to speak of baseline
activity in the wheel running animals. Given the present data
it is not possible to determine if and when the increase in
activity, which can be seen at the end of 6 days, will level off.
However, already in the relative short time of 6 days, running
wheel activity is affecting the animal on a behavioural level
by competing with other locomotor activities and probably
even by rearranging the organisation of locomotor activity.
To further investigate the suggestion that wheel running
is altering the organisation of behaviour, experiments may
be performed in which the running wheel is removed after
6 days. When wheel running is competing with other behaviours rather than reorganising behavioural patterns, the
effects found in the present experiment should vanish soon
after the removal of the wheels. Studies on the effect of wheel
running on brain activity, found a change in neuronal activity
in the core of the nucleus accumbens, after 90 min of voluntary wheel running when compared to novel environment
exploration [32]. Increased neuronal activity in this brain
region is also found as a response to repeated exposure to
drugs of abuse (e.g. cocaine, nicotine and amphetamine; ref.
[6]). This indicates that wheel running has potential addictive properties that are likely to affect the organisation of
behaviour and that this effect may endure for some period
after the removal of the running wheels from the animal’s
cage.
Although it can be a highly automated and time saving
tool, one should be cautious when using the running wheel
to measure the effects of experimental manipulations on activity. Running wheel activity is not representative of general
locomotor activity because the measuring instrument is reinforcing the measured behaviour. Furthermore, running wheel
activity has its own dynamics and follows a different hourly
pattern when compared to cage floor locomotion. Also, the
clear effects of wheel running on activity levels should be
taken into consideration when using the running wheel as
environmental enrichment. Wheel running cannot be considered solely as an extended opportunity for performing locomotion, but it is actually affecting the daily organization of
behaviour. This might be especially important in behavioural
studies where amounts and patterns of baseline locomotor
activity are likely to influence the outcomes of behavioural
tests [2,17,29].
The second objective of the present study was to introduce a newly developed home cage system as a tool for
e.g. behavioural phenotyping. With respect to the identification of genes underlying complex and interacting behavioural
traits, refinement of current phenotyping methods is required
[11,30,33]. Recently, much attention is being paid to the development of multi-dimensional tests as an alternative to the
standard, one-dimensional behavioural tests such as the open
field and elevated plus maze (e.g. [20]). The home cage system used in the present study proved to be a valuable tool to
measure interactions between different aspects of locomotor
activity in an integrated set-up. Furthermore, the fully automated nature of the system is timesaving, minimizes the
need for human intervention and contributes to standardisation. Currently, we are exploring the possibilities of this system for behavioural phenotyping studies. This also includes
finding solutions in the field of mathematical modelling for
the analysis of the large amount of data that are generated
with this system.
References
[1] Abe H, Kida M, et al. Feeding cycles entrain circadian rhythms of
locomotor activity in cs mice but not in c57bl/6j mice. Phys Behav
1989;45(2):397–401.
[2] Ammassari-Teule M, Caprioli A. Spatial learning and memory, maze
running strategies and cholinergic mechanisms in two inbred strains
of mice. Behav Brain Res 1985;17(1):9–16.
[3] Bacon Y, Ooi A, et al. Screening for novel enu-induced
rhythm, entrainment and activity mutants. Genes Brain Behav
2004;3(4):196–205.
[4] Belke W. Investigating the reinforcing properties of running: or, running is its own reward. In: Epling WF, Pierce WD, editors. Activity
anorexia: theory, research, and treatment. Mahwah, NJ: Erlbaum;
1996. p. 45–55.
[5] Brown J, Cooper-Kuhn CM, et al. Enriched environment and physical activity stimulate hippocampal but not olfactory bulb neurogenesis. Eur J Neurosci 2003;17(10):2042–6.
[6] Cadoni S, Solinas M, et al. Selective psychostimulant sensitization
by food restriction: differential changes in accumbens shell and core
dopamine. Eur J Neurosci 2003;18(8):2326–34.
[7] Drickamer LC, Evans TR. Chemosignals and activity of wild stock
house mice, with a note on the use of running wheels to assess
activity in rodents. Behav Proc 1996;36(1):51–66.
[8] Edgar DM, Kilduff TS, et al. Influence of running wheel activity
on free-running sleep/wake and drinking circadian rhythms in mice.
Phys Behav 1991;50(2):373–8.
[9] Gamble KL, Novak CM, et al. Neuropeptide y and n-methyl–aspartic
acid interact within the suprachiasmatic nuclei to alter circadian
phase. Neuroscience 2004;126(3):559–65.
L. de Visser et al. / Behavioural Brain Research xxx (2005) xxx–xxx
[10] Gattermann R, Weinandy R, et al. Running-wheel activity and body
composition in golden hamsters (Mesocricetus auratus). Phys Behav
2004;82(2-3):541–4.
[11] Gerlai R. Phenomics: fiction or the future? Trends Neurosci
2002;25(10):506–9.
[12] Hanagasioglu M, Borbely AA. Effect of voluntary locomotor activity
on sleep in the rat. Behav Brain Res 1982;4(4):359–68.
[13] Harri M, Lindblom J, et al. Effect of access to a running wheel on
behavior of c57bl/6j mice. Lab Anim Sci 1999;49(4):401–5.
[14] Hofstetter JR, Grahame NJ, et al. Circadian activity rhythms in
high-alcohol-preferring and low-alcohol-preferring mice. Alcohol
2003;30(1):81–5.
[15] Kas MJH, van den Bos R, et al. Mu-opioid receptor knockout
mice show diminished food-anticipatory activity. Eur J Neurosci
2004;20(6):1624–32.
[16] Kitamura T, Mishina M, et al. Enhancement of neurogenesis by
running wheel exercises is suppressed in mice lacking nmda receptor
[epsiv]1 subunit. Neurosci Res 2003;47(1):55–63.
[17] Kopp C. Locomotor activity rhythm in inbred strains of mice: implications for behavioural studies. Behav Brain Res 2001;125(12):93–6.
[18] Koteja P, Garland Jr T, et al. Behaviour of house mice artificially
selected for high levels of voluntary wheel running. Anim Behav
1999;58(6):1307–18.
[19] Leszczynski R, van Dellen A. Environmental stimulation, including
running wheel usage, delays onset and progresssion of Huntington’s
disease in r6/1 mice. Soc Neurosci 2001 (abstract).
[20] Ohl F. Testing for anxiety. Clin Neurosci Res 2003;3(4-5):233–8.
[21] Oliverio A, Malorni W. Wheel running and sleep in two strains of
mice: plasticity and rigidity in the expression of circadian rhythmicity. Brain Res 1979;163(1):121–33.
[22] Pietropaolo S, Branchi I, et al. Utilisation of a physically-enriched
environment by laboratory mice: age and gender differences. Appl
Anim Behav Sci 2004;88(1-2):149–62.
7
[23] Possidente B, McEldowney S, et al. Aging lengthens circadian
period for wheel-running activity in c57bl mice. Phys Behav
1995;57(3):575–9.
[24] Refinetti R. Daily activity patterns of a nocturnal and a diurnal rodent in a seminatural environment. Phys Behav 2004;82(2-3):285–
94.
[25] Rhodes JS, Garland Jr T, et al. Patterns of brain activity associated
with variation in voluntary wheel-running behavior. Behav Neurosci
2003;117(6):1243–56.
[26] Rhodes JS, van Praag H, et al. Exercise increases hippocampal neurogenesis to high levels but does not improve spatial learning in
mice bred for increased voluntary wheel running. Behav Neurosci
2003;117(5):1006–16.
[27] Sherwin CM. The use and perceived importance of three resources
which provide caged laboratory mice the opportunity for extended
locomotion. Appl Anim Behav Science 1998;55(3-4):353–67.
[28] Sherwin CM. Voluntary wheel running: a review and novel interpretation. Anim Behav 1998;56(1):11–27.
[29] Tang X, Orchard SM, et al. Home cage activity and behavioral performance in inbred and hybrid mice. Behav Brain Res
2002;136(2):555–69.
[30] Tecott LH, Nestler EJ. Neurobehavioral assessment in the information age. Nat Neurosci 2004;7(5):462–6.
[31] Van Praag H, Kempermann G, et al. Running increases cell proliferation and neurogenesis in the adult mouse dentate gyrus. Nat
Neurosci 1999;2(3):266–70.
[32] Vargas-Perez H, Mena-Segovia J, et al. Induction of c-fos in nucleus
accumbens in naive male balb/c mice after wheel running. Neurosci
Lett 2003;352(2):81–4.
[33] Wahlsten D, Rustay NR, et al. In search of a better mouse test.
Trends Neurosci 2003;26(3):132–6.
[34] Welsh DK, Richardson GS, et al. Effect of running wheel availability
on circadian patterns of sleep and wakefulness in mice. Phys Behav
1988;43(6):771–7.