Accepted Manuscript Title: Neurobiological effects of exercise on Major Depressive Disorder: A systematic review Author: Felipe Barreto Schuch Andrea Camaz Deslandes Brendon Stubbs Natan Pereira Gosmann Cristiano Tschiedel Belem da Silva Marcelo Pio de Almeida Fleck PII: DOI: Reference:
S0149-7634(15)30189-5 http://dx.doi.org/doi:10.1016/j.neubiorev.2015.11.012 NBR 2306
To appear in: Received date: Revised date: Accepted date:
21-9-2015 2-11-2015 23-11-2015
Please cite this article as: Schuch, F.B., Deslandes, A.C., Stubbs, B., Gosmann, N.P., Silva, C.T.B., Fleck, M.P.A.,Neurobiological effects of exercise on Major Depressive Disorder: A systematic review, Neuroscience and Biobehavioral Reviews (2015), http://dx.doi.org/10.1016/j.neubiorev.2015.11.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights: ‐ The precise neurobiological responses from exercise in people with depression are equivocal. ‐ Acutely, exercise promotes changes on ANP, BNP, Copepetin and Growth hormone. ‐ Long‐term exercise promotes changes on Copeptin, TBARS and cortical activity markers.
Ac ce pt e
d
M
an
us
cr
ip t
Page 1 of 36
Neurobiological effects of exercise on Major Depressive Disorder: A systematic review
cr
ip t
Felipe Barreto Schuch1,2*, Andrea Camaz Deslandes 3, Brendon Stubbs 4,5, Natan Pereira Gosmann2, Cristiano Tschiedel Belem da Silva1,2, Marcelo Pio de Almeida Fleck1,2
1
2
us
Programa de Pós-graduação em Ciências Médicas: Psiquiatria, Universidade Federal do Rio Grande do Sul, Porto Alegre, BR. Departamento de Psiquiatria, Hospital de Clínicas de Porto Alegre, Porto Alegre, BR.
3
an
Programa de Pós-graduação em Ciências do Exercício e do Esporte, Universidade Estadual do Rio de Janeiro, Rio de Janeiro, BR.
4
Physiotherapy Department, South London and Maudsley NHS Foundation Trust, London, UK
Ac ce pt e
d
5
M
Health Service and Population Research Department, Institute of Psychiatry, King's College London, London, UK
Running title: Neurobiology of Exercise in Major Depressive Disorder
*
Corresponding author at: Programa de Pós-Graduação em Ciências Médicas: Psiquiatria, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil Tel: +55 51 33085624; Fax: +55 51 33085624. E-mail address:
[email protected] (Felipe Barreto Schuch).
1
Page 2 of 36
Abstract Exercise displays promise as an efficacious treatment for people with depression. However, no systematic review has evaluated the neurobiological effects of exercise among people with Major Depressive Disorder (MDD). The aim of this article was to systematically review the acute and
ip t
chronic biological responses to exercise in people with MDD. Two authors conducted searches using Medline (PubMed), EMBASE and PsycINFO. From the searches, twenty studies were included within the review, representing 1,353 people with MDD. The results demonstrate that a single
cr
bout of exercise increases Atrial Natriuretic Peptide(ANP), Brain Natriuretic Peptide (BNP), Copepetin and Growth hormone among people with MDD. Exercise also potentially promotes
us
long‐term adaptations of Copeptin, Thiobarbituric Acid Reactive Species (TBARS) and Total Mean Frequency(TMF). However, there is limited evidence that exercise promotes adaptations on
an
neurogenesis, inflammation biomarkers and brain structure. Associations between depressive symptoms improvement and hippocampus volume and IL‐1B were found. Nevertheless, the paucity of studies and limitations presented within, precludes a more definitive conclusion of the
M
underlying neurobiological explanation for the antidepressant effect of exercise in people with MDD. Further trials should utilize appropriate assessments of neurobiological markers in order to
d
build upon the results of our review and further clarify the potential mechanisms associated with
Ac ce pt e
the antidepressant effects of exercise.
Key-words: Depression; Exercise; Inflammation; Hormones; Neurotrophines; Neuroplasticity; Neuronal Activity; Oxidative stress.
Running title: Neurobiology of Exercise in Major Depressive Disorder 2
Page 3 of 36
1 Background
Major Depressive Disorder (MDD) is a relatively common condition and a leading cause of
ip t
years lived with disability across the world (Ferrari et al., 2013). Several models have been proposed to explain the etiology of MDD, with one original hypothesis being attributed to
cr
“chemical imbalance in the brain” (Schildkraut, 1965). More recently, emerging evidence has
us
demonstrated that MDD has a more complex etiology, involving other neurobiological mechanisms such as neurotrophins, oxidative stress, inflammation, and changes in brain structure
an
and activation (Belmaker and Agam, 2008; Furtado and Katzman, In Press; Rive et al., 2013). In summary, several studies have shown that Patients with MDD present with: 1)
M
decreased levels peripheral (plasma and serum) Brain‐Derived Neurothrophic Factor (BDNF) levels, a marker of neurogenesis (Brunoni et al., 2008); 2) increased levels of peripheral (plasma and
d
serum) pro‐inflammatory markers, such as Interleukin(IL)‐6 (Dowlati et al., 2010; Valkanova et al.,
Ac ce pt e
2013); 3) increased serum oxidative stress markers, such as superoxide dismutase (SOD) and decreased antioxidant enzymes, such as glutathione peroxidase (GPX) (Lopresti et al., 2014), and 4) changes in brain anatomy (e.g: decrease in the hippocampus volume) and activity of some cortical structures (E.g: abnormally reduced activity in lateral prefrontal cortices during explicit voluntary control of emotional experiences) (Hamilton et al., 2012; Soares and Mann, 1997; Steffens and Krishnan, 1998). Taken together, these results clearly suggest that the etiology of MDD is complex and multifaceted, involving numerous interlined neurobiological systems (Song and Wang, 2011).
Physical exercise has been shown to be an efficacious treatment for MDD, with effect
sizes ranging from small (‐0.4) to very large (‐1.4) (Cooney et al., 2013; Craft and Landers, 1998;
3
Page 4 of 36
Daley, 2008; Danielsson et al., 2013; Josefsson et al., 2014; Krogh et al., 2011a; Rethorst et al., 2009; Silveira et al., 2013; Stathopoulou et al., 2006). Indeed, a number of studies have previously demonstrated that exercise may offer comparable benefits to antidepressant medication in those
ip t
with depression (Blumenthal et al., 2007; Blumenthal et al., 1999). Moreover, exercise is efficacious for outpatients (Dunn et al., 2005; Park and Yu, 2015) and inpatients (Schuch et al.,
cr
2015). Despite its efficacy, the mediators or mechanisms underlying the antidepressant effects of
us
exercise in MDD are unclear, speculative and predominantly derived from animal studies or findings from studies conducted in people without MDD (Eyre and Baune, 2012b; Fuqua and
an
Rogol, 2013; Knaepen et al., 2010; Pedersen and Hoffman‐Goetz, 2000; Radak et al., 2008a; Radak et al., 2008b; Radak et al., 2001; Scheewe et al., 2013).
M
It is essential that underlying mechanisms through which exercise exerts its antidepressant effects are better understood, since this will enable more optimal targeted interventions to be developed.
d
Current hypotheses for the antidepressant effect of exercise include both acute (transient
Ac ce pt e
responses that occur during or immediately after the exercise bout) and chronic responses (adaptive changes following a training period of two or more consecutive exercise bouts) that influence several systems such as neuroendocrine, neurogenesis, oxidative stress, auto‐immune and cortical structural changes (Eyre and Baune, 2012b; Fuqua and Rogol, 2013; Knaepen et al., 2010; Pedersen and Hoffman‐Goetz, 2000; Radak et al., 2008a; Radak et al., 2008b; Radak et al., 2001; Scheewe et al., 2013). It is important that acute and chronic responses are considered separately due to the fact that the responses to exercise may be different, and even opposite directions. For example, acute exercise increases some pro‐inflammatory and oxidants, while chronic responses to exercise over several weeks appears to decrease pro‐inflammatory and oxidant markers (Pedersen and Hoffman‐Goetz, 2000; Radak et al., 2008a; Radak et al., 2008b; Radak et al., 2001). 4
Page 5 of 36
Research considering the neurobiological response to exercise among people with MDD is equivocal. Given the rising burden of MDD and the promise of exercise as an intervention, there is an urgent need to consider the plausible mechanisms underlying the antidepressant response
ip t
elicited from exercise. To our knowledge, no systematic review has addressed this gap and reviewed studies conducted in humans with MDD. Given the aforementioned, the aim of the
cr
present study was to systematically review studies that have evaluated acute and chronic
us
biomarker responses to exercise across five current biological hypotheses proposed to explain MDD etiology including neuroendocrine, neurogenesis, oxidative stress, inflammation and cortical
an
thickness and activity. The present study provides a comprehensive review of the main pathways proposed to explain the antidepressant effects of exercise in subjects with MDD.
M
Ac ce pt e
d
5
Page 6 of 36
2 Methods
The present systematic review was conducted according to the PRISMA (Moher et al.,
ip t
2009) statement.
cr
2.1 Literature search
Two independent authors conducted searches of Medline (PubMed), EMBASE and
us
PsycINFO from inception till January 2015. The following search strategy was used (((("exercise" OR "physical activity")) AND "depress*") AND ("cortical activity" OR "nerve growth factors" OR
an
"oxidant" OR "anti‐oxidant" OR "oxidative" OR "endocrin*" OR "neurogenesis" OR "Immune*" OR "immunol*" OR "inflammat*" OR "hormones" OR "hormona*" OR "oxidative stress" OR
M
"Electroencephalography" OR "neurotrophin" OR "VGF" OR "BDNF" OR "alpha asymmetry" OR "brain‐derived neurotrophic factor" OR "cytokines" OR "5HT2" OR "serotonin" OR "cortisol" OR
d
"Interleukin" OR "TBARS" OR "EEG" OR "GH" OR "FMRi" OR "magnetic functional resonance" OR
Ac ce pt e
"growth‐hormone" OR "hpa axis" OR "endocannabinoid" OR "endorphin" OR "IGF‐1" OR "ANP" OR "atrial natriuretic" OR "Hypothalamic‐pituitary‐adrenal axis" OR "prolactin" OR "adrenaline" OR "dopamine")) NOT ("mice" OR "rat" OR "hamster" OR "arthritis" OR "stroke" OR "fibromyalgia" OR "diabetes" OR "cancer" OR "kidney" OR "coronary artery disease" OR "asthma" OR "parkins*" OR "HIV" OR "multiple sclerosis" OR "chronic heart failure" OR "Spine*" OR "Spinal" OR "epilepsy" OR "COPD" OR "bowel" OR "bipolar"). Reference lists of all included studies were also reviewed for potentially eligible articles.
2.2 Study selection Two independent reviewers (FBS) and (NPG) selected the articles deemed potentially
eligible at the title and abstract level. The inclusion criteria were: 1) published in English; 2)
6
Page 7 of 36
present original data; 3) involve clinically depressed patients (individuals with a diagnosis of MDD according the Diagnostic and Statistical Manual for Mental Disorders (DSM), International Classification of Diseases (ICD) or Research Diagnostic Criteria (RDC) criteria, assessed by
ip t
psychiatrists and/or through the use of standardized instruments such as Mini International Neuropsychiatric Interview (MINI), Composite International Diagnostic interview (CIDI), Structured
cr
Clinical Interview (SCID)); 4) evaluate the effects of an exercise‐based intervention (a single bout or
us
a training intervention) on biological outcomes as hormones, neurotrophines, inflammation biomarkers, oxidative stress and cortical plasticity and activity.
an
Articles were excluded that: 1) included patients with clinical comorbidities (for instance stroke); or 2) included patients other psychiatric diagnosis (e.g. bipolar disorder) or if the study
M
included participants with subthreshold depressive symptoms. If we encountered studies presenting data from the same sample but reported different biomarkers, both studies were
2.3 Extraction of data
Ac ce pt e
d
included.
Two independent reviewers (FBS) and (NPG) extracted the data on a predetermined
database. The data extracted from each study included: study design, participant characteristics, exercise intervention details and biomarker response results. Studies presenting both acute and chronic effects of exercise were analyzed as different studies. Whenever a disagreement arose, a third reviewer was available for mediation. 2.4 Outcomes measures
Our primary outcome measure was the acute or chronic response from exercise within
neurogenesis, biomarkers of neuroendocrine responses, inflammation and oxidative stress, neuroimaging and electroencephalogram (EEG). Any technique of neuroimaging (functional, structural and molecular) or EEG were considered. We defined acute as a single bout of exercise
7
Page 8 of 36
(Acute studies) and chronic as studies that evaluated the adaptations in longer interventions of two or more sessions of exercise. 2.5 Analysis
The Standardized Mean Difference (SMD) test was used to calculate the effect sizes of
ip t
each mediator (Hedges and Olkin, 2014). For studies evaluating the acute effects of exercise, the
cr
effect size was calculated through the change from pre to post‐test. The baseline or rest measure
us
prior to exercise value was used as the pre‐test value. For post‐test, the measure acquired immediately after exercise was used. The chronic effects were calculated through the mean
an
difference between exercise and control groups at the end of the intervention. When two or more studies evaluated the same outcome, the data was pooled using the random‐effects analysis. If we
M
encountered a study presents in which the authors included two exercise groups, both groups were included separately in our analyses. Effect sizes were interpreted according Cohen`s criteria
d
(Cohen, 1988) including : small (SMD=0.2), medium (SMD=0.5), and large effects (SMD=0.8). All
Ac ce pt e
the analyses were performed using Review Manager (Version 5.3).
3 Results
3.1 Characteristics of Included Studies and Samples
Initially, 4174 studies were localized as potentially relevant. Following the application of
the eligibility criteria, twenty studies were included for review (Boettger et al., 2010; Deslandes et al., 2010; Filser et al., 1988; Foley et al., 2008; Gustafsson et al., 2009; Hallberg et al., 2010; Hennings et al., 2013; Kiive et al., 2004; Krogh et al., 2014a; Krogh et al., 2013; Krogh et al., 2010; Krogh et al., 2014b; Krogh et al., 2011b; Lechin et al., 1995; Rethorst et al., 2013; Salehi et al., 2014; Schuch et al., 2014; Silveira et al., 2010; Toups et al., 2011; Wisén et al., 2011). Full details of the search results, including reasons for exclusion are summarized in figure 1. 8
Page 9 of 36
Figure 1 here. 3.2 Summary of included studies
ip t
Overall, 1,353 people with MDD were represented across the included studies. Eight studies evaluated acute response to exercise (Boettger et al., 2010; Filser et al., 1988; Gustafsson
cr
et al., 2009; Hallberg et al., 2010; Kiive et al., 2004; Krogh et al., 2011b; Lechin et al., 1995; Wisén et al., 2011), ten evaluated chronic response (Deslandes et al., 2010; Foley et al., 2008; Hennings
us
et al., 2013; Krogh et al., 2014a; Krogh et al., 2014b; Rethorst et al., 2013; Salehi et al., 2014; Schuch et al., 2014; Silveira et al., 2010; Toups et al., 2011) and two evaluated acute and chronic
an
response to exercise (Krogh et al., 2013; Krogh et al., 2010). The most commonly considered biomarker was hormones (n=9), followed by neurogenesis biomarkers (n=5), inflammatory
M
biomarkers (n=5), cortical activity (n=2), structural neuroimaging (n=1) and oxidative stress (n=1). The distribution of the articles per decade is shown in Figure 2. Further details of the included
Ac ce pt e
d
studies are presented in table 1 and 2 for acute and chronic responses respectively. Across the 20 included studies, 4 did not provide sufficient data for analysis and the authors could not provide supplementary data upon request (Boettger et al., 2010; Filser et al., 1988; Gustafsson et al., 2009; Lechin et al., 1995).
Figure 2 here
3.3 Acute response to exercise
The studies considering acute response to exercise predominantly included adults with
mild to severe MDD episodes, as judged by symptom severity checklists. The acute studies primarily focused on hormonal, inflammatory, and neurogenesis biomarker responses to exercise. No study evaluated the acute responses to exercise on oxidative stress or cortical activity. Data
9
Page 10 of 36
were not evaluable for serotonin, dopamine, epinephrine and BDNF effect size estimation. A detailed description of the studies is provided in Table 1. The effect sizes can be seen in the Table 2. The acute response to exercise will now briefly be explored across the five main neurobiological
Insert Table 1 here
cr
ip t
systems.
us
Insert Table 2 here
3.3.1.1 Atrial Natriuretic Peptide (ANP)
an
3.3.1 Neuroendocrine hypothesis
M
Two studies evaluated acute responses of exercise in ANP (Krogh et al., 2011b; Wisén et al., 2011). The pooled analysis revealed that exercise results in a large reduction in ANP (SMD=‐
d
1.22,p=0.0002).
Ac ce pt e
3.3.1.2 Bain Natriuretic Peptide (BNP)
A large and significant reduction in BNP levels among people with MDD was found (SMD=‐
0.88, 95% CI [‐1.58, ‐0.17] p=0.02). 3.3.1.3 Copeptin
The Copeptin acute response to exercise was evaluated in one study across two groups
(aerobic x anaerobic) before and after 16 weeks of exercise training (Krogh et al., 2013). The pooled data demonstrated a significant increase in copeptin with medium effect size (SMD=0.56, p