Journal of Affective Disorders 200 (2016) 218221

Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research paper

Sweating away depression? The impact of intensive exercise on

depression
Ross Balchin a, Jani Linde b, Dee Blackhurst c, HG Laurie Rauch d, Georg Schnbchler e,n
a

Department of Psychology, University of Cape Town, South Africa

Division of Exercise Science and Sports Medicine, University of Cape Town, South Africa
c
Division of Chemical Pathology, Faculty of Health Sciences, University of Cape Town, South Africa
d
Division of Exercise Science and Sports Medicine, Department of Human Biology, Faculty of Health Sciences, University of Cape Town, South Africa
e
Neuropsychology Unit, Department of Neurology, University Hospital Zurich, Frauenklinikstrasse 26, CH-8091 Zurich, Switzerland
b

art ic l e i nf o

a b s t r a c t

Article history:
Received 18 January 2016
Accepted 16 April 2016
Available online 20 April 2016

Background: In periods of prolonged stress and pain the body produces endorphins to help endure pain.
The PANIC system is built on the same pathways as the pain system and is characterized by behaviour
that looks like depression. The term 'mental pain in the context of feelings of loss is arguably justied in
light of this relationship between the physical pain and social loss systems. It is reasonable to expect that
endorphin release ameliorates depression.
Methods: Moderately depressed males (n 30) were randomly assigned to one of three groups of varying
exercise intensity. Each underwent a six-week exercise programme for three days per week, one hour per
day. The HAM-D, MADRS, and ANPS were administered weekly and -endorphin levels measured.
Results: Moderate- and high-intensity exercise improved depression levels, while very-low intensity
exercise did not have as benecial an effect. -endorphin results were inconclusive. Participants showed
a slight decrease in PANIC and FEAR, and increased SEEKING.
Limitations: The potential insensitivity of the assays that were utilized, and the known problems with
measuring -endorphins, may have contributed to the ndings. The lack of a state measure of the basic
emotion systems is problematic, as a trait measure has to be relied upon, and this likely affected the
ability to accurately detect changes over time.
Conclusions: The demonstrated improvements in depressive symptoms have important implications for
the clinical treatment of patients despite the hypothesis that the PANIC system is involved in the genesis
and maintenance of depression not having been conclusively conrmed.
& 2016 Elsevier B.V. All rights reserved.

Keywords:
Depression
Exercise
Endorphin
PANIC/Separation-distress system

Evidence exists that exercise is benecial for alleviating depression. Whilst key research is required to investigate the mechanisms responsible for this relationship, the fact remains that
physical activity is good for physical and mental health (Blake,
2012). Despite studies suggesting that exercise can alleviate depressive symptoms, it remains unclear how much, how frequent,
and how intense exercise should be, and what form is best. Few
clinical-control trials have looked at exercise and its effects on
depression. Most studies nd that exercise is an effective treatment and is comparable to antidepressant therapy. Few studies
have examined different intensities of exercise, and the relationship between exercise and endorphins has been inadequately
investigated.
Based on the pioneering work of Jaak Panksepp, there are
n

Corresponding author.
E-mail address: georg.schoenbaechler@access.uzh.ch (G. Schnbchler).

http://dx.doi.org/10.1016/j.jad.2016.04.030
0165-0327/& 2016 Elsevier B.V. All rights reserved.

considered to be four 'basic-emotion command systems in the

brain: the SEEKING-, FEAR-, RAGE- and PANIC/separation-distress/
sadness system (Panksepp, 1998). It is possible that the PANIC
system is the mechanism underlying depression. This system is
built over the physical pain pathway, with the anterior cingulate
gyrus as its core (Solms and Turnbull, 2002). Stimulation of the
PANIC system can produce a full clinical depression (Solms and
Turnbull, 2002). Psychological pain when separated from an attachment gure results from a withdrawal of endogenous opioids.
Separation distress results in two distinct behaviours: (i) animals
call out to be found, known as the protest phase; (ii) if the
caregiver does not return, then after a while the animal withdraws
and waits to be found, known as the despair phase (Panksepp,
1998). The despair phase is characterized by withdrawal and hibernation-like behaviour, which looks exactly like depression
(Solms and Turnbull, 2002).
The body produces endorphins to help endure pain. These are
only released when pain intensity is at least moderately high. The

R. Balchin et al. / Journal of Affective Disorders 200 (2016) 218221

endogenous opioid-based PANIC system is built on the same

pathways as the pain system. The term mental pain in the context
of feelings of loss seems justied in light of this relationship between the physical pain and social loss systems. It is reasonable to
expect that endorphin release positively affects depression, and
evidence exists that exercise is benecial for depression. However,
the precise physiological mechanism underlying the antidepressant effect of exercise has not been fully elucidated. We
hypothesized that it involves the PANIC system and the benet
that endorphin release has on this system as with its established analgesic effect with physical pain.
Key hypotheses for this study were: (i) the mechanism behind
improvements in depression due to exercise is the effect of endorphin release on the PANIC system; (ii) only high-intensity exercise greater than 70% of HR reserve would produce a signicant
-endorphin release (ktedalen et al., 2001), whereas moderateintensity exercise (less than 50% of HR reserve) and very low-intensity exercise would be insufcient, and would not result in any
signicant improvement in depression; (iii) those whose depression improved would show decreased PANIC and FEAR, and increased SEEKING.

219

while the CONs kept their HRs below 120 beats per minute (bpm)
by walking and/or doing very light cycling. Once weekly, a psychologist (blind to the participant's group) administered the HAMD, MADRS and ANPS. The HIGHs and MODs had blood drawn once
weekly (before and after exercise). The CON's had blood taken at
the start, before and after exercise, and upon completion of the
study, before and after exercise.
1.4. Data analysis
Principal data analysis involved descriptive analysis of the
psychological data. GraphPad Prism version 5.00 for Windows was
used for the statistical analyses, constituting a secondary level of
analysis conducted only on data from the 21 participants who
completed the study. Non-parametric one-way ANOVAs with
Kruskal-Wallis and Mann-Whitney tests were performed to calculate differences in: (i) the HAM-D and MADRS scores; (ii) circulating -endorphin levels; (iii) HRs of all groups.

2. Results
2.1. Depression data

1. Methodology
1.1. Sample
Thirty moderately depressed males aged between 18 and 42
years were recruited (mean age 25.4 years; mean BMI,
kg/m2 26.9). None were receiving any antidepressant therapy.
Level of depression was determined using an online Major Depression Inventory (MDI), and conrmed by interview with a
psychologist using the Hamilton Rating Scale of Depression (HAMD). Participants were randomly assigned to one of three exercise
groups: High-intensity (HIGH), Moderate-intensity (MOD), and a
Control group (CON). There were eleven MODs (mean: age24.2
years; BMI, kg/m2 24.6), 9 HIGHs (mean: age28 years; BMI,
kg/m2 25.3), and ten CONs (mean: age 23.9 years; BMI,
kg/m2 30.6). Twenty-one completed the study, with 12 dropping
out. An intention-to-treat analysis was adopted, so dropouts were
not excluded from the data analysis. Females were excluded due to
the menstrual cycle, as were individuals with a prior high-intensity exercise routine and those with medical conditions that
put them at risk.
1.2. Measures

2.1.1. Primary analysis

The qualitative analysis (n 30) revealed that 13 participants
ended with no depression, 13 had mild depression, and four had
moderate depression. Six HIGHs had no depression, two improved
to mild depression, and one remained moderately depressed. Six
MODs had no depression, four improved to mild depression, and
one remained moderately depressed. One CON ended with no
depression, seven improved to mild depression, and two stayed
moderately depressed.
2.1.2. Secondary statistical analysis
The HIGHs showed the greatest improvement, followed by the
MODs. The CONs improved somewhat, but not to the same extent.
The HIGH's average HAM-D scores improved from 15.5 to 4.2, the
MOD's from 16.2 to 5.7, and the CON's from 17.4 to 9.8. Fig. 1 shows
initial and nal HAM-D scores. There was no signicant difference
when comparing all three groups initial HAM-D scores
(p 0.0789) with each other, or when comparing their nal HAMD scores (p 0.0974). There was a signicant difference when
comparing the initial scores for each group to their nal HAM-D
scores (p o0.0001). When examining the HAM-D scores, there

Materials included the MDI, HAM-D, Montgomery-sberg Depression Rating Scale (MADRS) and the Affective Neuroscience Personality Scale 2.4 (ANPS), along with a Suunto t6d watch with chest
strap, a Lode Excalibur Sport Ergometer, the BORG RPE Scale, and the
Peak Sustained Power Output test (PSPO).
Blood was collected in gold Vacutaner tubes containing Z Serum
Sep Clot Activator. The rst -endorphin kit used was a CUSABIO
ELISA KIT (catalogue number CSB-E06821h; human -endorphin,
-EP ELISA kit) ordered from Biocom Biotech. After several attempts, this kit was deemed to lack sufcient sensitivity and a
second kit from MD Bioproducts (catalogue number M056011) was
instead used.
1.3. Design
This three-armed prospective randomized control pilot study
involved all groups participating in a six-week programme comprising exercise three days per week, for one hour per day. The
HIGHs exercised at 7075% of HR reserve, the MODs at 4550%,

Fig. 1. Box and whiskers indicating the initial and nal HAM-D scores for all
groups.

220

R. Balchin et al. / Journal of Affective Disorders 200 (2016) 218221

was a signicant difference (p 0.0048) for the HIGHs when

comparing initial and nal scores. The MODs showed a signicant
difference between their initial and nal scores (p 0.0002). The
same trend was apparent for the CONs, although not as signicant
(p 0.0160).
There were no signicant differences when comparing the
three groups initial MADRS scores (p 0.2600), or when comparing their nal scores (p 0.1596). There was a signicant difference when comparing the initial scores within each group to
the nal MADRS scores (p 0.0182). The HIGHs (p 0.0450) and
MODs (p 0.0340) showed a signicant difference between their
initial and nal scores, but the CONs did not (p 0.3443). When
comparing nal MADRS scores between the HIGHs and MODs,
there was no signicant difference (p 0.4139), nor was there
when comparing nal scores between the HIGHs and CONs
(p 0.0988).
2.2. ANPS
Twenty-four participants results were analyzed six dropouts
completed fewer than six sessions and did not have a second
questionnaire for comparison purposes. Across groups, the trend is
one of a decrease in PANIC and FEAR, and an increase in SEEKING.
None of the groups trends were of statistical signicance.
2.3.

-endorphins

A signicant difference (p 0.0222) occurred in the post-exercise -endorphin concentrations amongst the three groups.
When comparing post-exercise values of the CONs to the MODs, a
signicant difference was found (p 0.0173). When comparing the
post-exercise values of the CONs with the HIGHs, there was no
signicant difference (p 0.0856). Since -endorphin levels in the
CONS decreased throughout, unlike the other groups whose endorphins did not decrease, it can be concluded that low-intensity exercise does not yield an endorphin release. The CONs
showed a signicant decrease in -endorphins following exercise,
whereas there was an almost equal number of increased and decreased values for the MODs. The HIGHs showed a slightly greater
number of decreased values. The means after exercise for the
MODS and HIGHs were not signicantly different to baseline values. Differences between the baseline and post-exercise measurements were not signicantly different for the MODs
(p 0.953) and HIGHs (p 0.992).

3. Discussion
All participants exercised within the assigned HR ranges and
the average HRs of the three groups were signicantly different
from each other. All but four experienced at least some depression
reduction. Thirteen improved to mild depression, and 13 ended
with no depression. Both the HIGHs and MODs contained six
participants who ended with no depression. These ndings are
inconsistent with the hypothesis that only the high-intensity exercise would yield improvements. Nonetheless, the HIGHs still
derived the greatest benet. It is interesting that the MODs improved so much, suggesting that a lower intensity of exercise is
therapeutic. The HAM-D scores after six weeks reduced 77% from
baseline for the HIGHs, signicantly more than the MODs (65%)
and CONs (44%). The HIGH's MADRS scores reduced 58%, the
MOD's improved 43% and the CON's 24%. The HIGHs showed the
greatest improvement, with a reduction of 77%, signicantly better
than the CONs.
The hypothesis that -endorphin release would only occur
during high-intensity exercise may have been disproven, as there

Table 1
Depression scores.
Group n

HIGH
MOD
CON

HAM-D initial HAM-D nal

score Mean
score Mean
(SD)
(SD)

9 15.89 (1.76)
11 16.36 (1.43)
10 17.1 (1.20)

5.67 (5.85)
6.64 (5.07)
11.8 (3.94)

MADRS initial
score Mean
(SD)

MADRS nal
score Mean
(SD)

12.67 (3.97)
14.36 (4.27)
18.8 (6.41)

7 (6.69)
9 (6.75)
15 (5.19)

Note: HAM-D: 0-7 Normal, 8-13 Mild, 14-18 Moderate, 19-22 Severe
MADRS: 0-6 Normal, 7-19Mild, 20-34 Moderate, 434 Severe

was no signicant difference between the baseline and post-exercise serum levels for the HIGHs (p 0.992) or MODs (p 0.953).
There were an almost equal number of increased and decreased
values for the MODs. Surprisingly, the HIGHs showed more decreased values over all their visits. Mean values following exercise
for the MODs and HIGHs were not signicantly different to baseline. The signicant difference in post-exercise concentrations
amongst the CONs, MODs and HIGHs is probably not meaningful
in the context of the sample size. Interestingly, the CONs showed a
signicant decrease following exercise. The CON's values of absolute change were greater than those of the MOD's and HIGH's.
These unexpected results may relate to shortcomings of the assay,
especially with respect to sensitivity and correlations between
circulating and central nervous system levels (Schwarz and
Kindermann, 1992; Meyer et al., 2000). A more accurate means
would be to measure levels in the brain, yet this is not possible.
The decreased PANIC and FEAR across all groups, and the increased SEEKING, is pleasing, suggesting that the PANIC system
may be the mechanism underlying depression. Part of this hypothesis is the prediction that an increase in SEEKING corresponds
to a lifting of the despair phase of the separation distress response this trend was observed, yet it is clear from the descriptive statistical analysis that these results are not statistically
signicant. A likely explanation is that the ANPS is designed as a
trait measure of the basic emotion systems. This study sought to
establish whether the ANPS could effectively serve as a state
questionnaire, yet it remains unclear whether it is accurate/sensitive enough to detect possible uctuations in the basic emotions
over short periods.(Table 1).

4. Conclusion
High- and moderate intensity exercise had a positive impact on
moderate depression. -endorphin values did not increase signicantly during exercise. The PANIC system was not conclusively
rmed as the mechanism underlying depression.

Acknowledgements
The authors would like to sincerely thank the Institute for the Study of Affective
Neuroscience (ISAN) for their generous grant which supported this research.

Appendix A. Supporting information

Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.jad.2016.04.030.

References
Blake, H., 2012. Physical activity and exercise in the treatment of depression. Front
Psychiatry 3, 106.
Meyer, T., Schwarz, L., Kindermann, W., 2000. Exercise and endogenous opiates. In:

R. Balchin et al. / Journal of Affective Disorders 200 (2016) 218221

Warren, O., Constantini, N.W. (Eds.), Contemporary Endocrinology: Sports Endocrinology. Humana Press Inc., Totowa, NJ, pp. 3142.
ktedalen, O., Solberg, E.E., Haugen, A.H., Opstad, P.K., 2001. The inuence of
physical and mental training on plasma beta-endorphin level and pain perception after intensive physical exercise. Stress Health 127, 121127.
Panksepp, J., 1998. Affective Neuroscience: The Foundations of Human and Animal

221

Emotions. Oxford University Press,, Oxford.

Schwarz, L., Kindermann, W., 1992. Changes in beta-endorphin levels in response to
aerobic and anaerobic exercise. Sports Med. 13, 2536.
Solms, M., Turnbull, O., 2002. The Brain and the Inner World: An Introduction to the
Neuroscience of Subjective Experience. Other Press,, New York.