Introduction

Physical activity (PA) and exercise (EX) have been investigated in the context of prevention and treatment of mental disorders over the past few decades. Epidemiological and population-based studies demonstrate significant negative relationships between PA levels and the onset of mental disorders in cross-sectional and prospective designs [1, 2]. Furthermore, a cross-sectional study [3] confirmed that psychiatric patients who regularly exercise perceive a higher quality of life (QoL).

In patients suffering from mental disorders, PA and EX are proposed to be effective for several reasons (e.g., [4–6]):

First, a wide range of biochemical and physiological changes and adaptations are related to acute bouts of EX and regular EX training. Some of these mechanisms affect mood (e.g., via serotonin, endorphins), and others have neuroprotective functions (e.g., normalization of brain-derived neurotrophic factor (BDNF) release), act as anxiolytic (e.g., release of atrial natriuretic peptide (ANP)), or alter stress reactivity (hypothalamus–pituitary–adrenal (HPA) axis).

On a psychological level, several mechanisms have been proposed, such as changes in body scheme and health attitudes/behaviors, learning and extinction, social reinforcement, experience of mastery, shift of external to more internal locus of control, or improved coping skills.

In the following sections, evidence is summed up for the acute and chronic effects of EX and PA as a sole or combined treatment in different mental disorders.

Anxiety disorders

A recent meta-analysis concluded that both aerobic and anaerobic EX are similarly effective as cognitive/behavioral therapy concerning the acute and chronic decrease of anxiety, and more effective than most other anxiety-reducing activities [7].

Panic disorder (PD)

EX is thought to act as a repeated interoceptive exposure therapy, confronting patients with bodily sensations they usually try to avoid. The earliest study compared a jogging to a walking program in neurotic patients, finding equally reduced symptoms in both interventions after 8 weeks and negative correlations between fitness increase and anxiety scores [8]. A more recent study [9] found that 10 weeks of endurance training decreased anxious and depressive symptoms in panic patients to a similar extent as clomipramine, whereas in a third study [10], no differences were found between EX and relaxation enhancement of paroxetine treatment.

In panic patients as well as in healthy controls, acute EX impedes the provocation of panic attacks via CO2 [11] and CCK-4 [12].

Post-traumatic stress disorder (PTSD)

Three pilot studies reported positive effects of aerobic EX and moderate walking on PTSD symptom severity and associated depressive and anxious symptoms in children [13], adolescents [14], and adults [15]. Since all of these studies have severe methodological limitations such as very small sample sizes, inclusion of participants without clinical PTSD diagnosis, and lacking control groups, adequately powered RCTs are needed to determine positive effects and possible risks when using EX as complemental treatment in PTSD.

Other anxiety disorders/anxiety sensitivity

To our knowledge, no studies have been published explicitly targeting EX interventions in social phobia, specific phobia, or generalized anxiety disorder. However, two studies on mixed samples were published. One found that EX training increased the amount of PA and functional capacity in patients with different anxiety disorders [16], and the other one reported larger reductions of anxiety, depression, and perceived stress in a combined CBT+walking treatment, compared to CBT alone, especially in patients with social phobia [17].

Several other studies found that aerobic EX diminished state anxiety as well as anxiety sensitivity, a personality trait related to the development and course of anxiety disorders (e.g.,[18, 19]).

Obsessive-compulsive disorder (OCD)

Two recent pilot studies investigated the effects of EX on OCD and concurrent anxiety and depression. A six-week walking augmentation of selective serotonin-reuptake-inhibitor (SSRI) treatment was reported to reduce OCD symptoms and depression at post-treatment and one-month follow-up (FU), and temporarily reduce anxiety scores [20]. Combining behavioral therapy or pharmacotherapy with a 12-week moderate aerobic EX program, another study found that EX enhancement reduced OCD symptom severity at the end of the treatment and three, six, and 26 weeks later [21]. In the same sample, anxiety, negative mood, and OCD symptoms were acutely reduced after each 20- to 40-min training session relative to baseline, especially at the beginning of the 12-week intervention [22].

However, due to the lack of control groups and very small sample sizes, these results can only serve as preliminary indicators for the efficacy of EX in the treatment of OCD patients.

Affective disorders

Major depression (MDD)

The efficacy of EX interventions in reducing depressive symptoms, negative affect, and sleep disturbances has widely been studied and summarized in several reviews. A recent meta-analysis [23] included 25 RCTs, finding large clinical effects of EX compared to no-treatment, with moderate effects in the long-term FU. When compared with cognitive therapy or antidepressants, EX was found to be as effective as either of these treatments.

However, when considering only trials with adequate concealment, blinded outcome assessments, and intention-to-treat analysis, this effect shrank to a moderate effect size, not reaching statistical significance. Mixed and resistance EX showed larger effect sizes (but also larger confidence intervals) than aerobic EX. In general, high EX intensities (e.g., public health dose) seemed more effective than lower doses for both EX types.

In one FU study, long-term relapse rates were lower after EX interventions than after sertraline treatment, especially when patients continued exercising on their own [24]. Neither initial treatment assignment nor antidepressant medication use, but regular PA during the FU period predicted HAM-D scores and MDD diagnosis 1 year post-treatment [25].

Surprisingly, EX compared with sertraline failed to improve neurocognitive functions in depressed middle-aged and older adults after 4 months of treatment [26], although some studies reported normalized BDNF levels in MDD patients after acute EX [27]. Acutely, EX has also been found to increase positive mood and vigor compared to quiet rest, with no significant differences in terms of distress, depression, confusion, fatigue, tension, or anger [28].

Post-natal depression (PND)

In a related field of research, a recent meta-analysis on five studies investigated the effects of PA on PND [29]. When compared with no-EX, EX considerably decreased PND symptoms. However, most studies had very small sample sizes and large CIs, and the prominent effect diminished to non-significance after exclusion of one study that combined EX with social support.

Bipolar disorder (BD)

In BD, a higher prevalence of comorbid physical health problems such as obesity and cardiovascular disorders has been reported, and patients with BD experience faster exhaustion during moderate aerobic EX than healthy controls [30]. A recent review of six studies concluded that PA interventions (both elective and prescribed) may indeed be feasible and decrease stress, depressive, and anxious symptoms in BD [31]. Acutely, PA was shown to increase self-reported well-being and dehydroepiandrosterone sulfate (DHEAS) [32].

Since all cited studies were underpowered, non-randomized trials, further research in this population is needed to determine potential benefits, as well as limitations and risks of PA like the induction of manic episodes (for detailed suggestions, see [31]). Three other reviews discussed EX-induced neurobiological changes [33], interactions between BDNF levels and allostatic load [34], and EX as a possible treatment for neurocognitive dysfunction in BD [35].

Eating disorders (ED)

In ED, the role of PA/EX is controversial. Advantages like weight loss in obese patients with binge eating disorder (BED) and prevention of bone mass loss in anorexia nervosa (AN) have as well been described as detrimental effects and poor therapy outcomes in patients practicing excessive PA with compulsive features.

Since BED patients tend not to exercise at all [36], two studies addressed the therapeutic effects of EX interventions. One study found moderately reduced weight and depression scores after 6 months of moderate EX (walking) compared to a control group [37]; the other reported significantly larger reductions in body mass index (BMI), depression scores, and binge episodes up to 12 months after treatment, when EX was added to a standard CBT treatment [38]. Interestingly, positive effects were found despite poor EX compliance and immediate return of PA levels to baseline after the end of treatment, which is in line with studies suggesting that the perceived effects of being active may be more relevant than actual fitness gains [39].

For bulimia nervosa (BN), the only published study compared EX to CBT treatment, finding that EX was as effective as CBT in reducing self-reported “Bulimia” and “Body dissatisfaction”, and even more effective than CBT in terms of “drive for thinness” and bulimic behavior up to 18 months after discharge [40].

One review [41] on EX in AN stated that although none of the six included studies satisfies RCT criteria (small sample size, lacking randomization, or quasi-experimental design), EX programs of light to moderate intensity potentially reduce obligatory attitudes and beliefs toward EX (especially in previously excessive exercisers), reduce emotional stress, protect bone mass, and enhance weight gain. However, another recent study found neither beneficial nor detrimental effects of a 12-week resistance training program in teenage AN patients [42].

Substance use disorders (SUD)

For EX in smoking cessation, a large number of RCTs have been conducted, indicating supplementary benefits of EX treatment, when combined with nicotine replacement and/or CBT. A review [43] summarizing 13 studies stated that in order to successfully support patients, EX programs should begin prior to smoking cessation, have rather high intensities and a minimum duration of 10 weeks, and promote EX as a coping strategy to prevent relapse by regulating mood and reducing craving.

In contrast, evidence is weaker for the efficacy of EX in alcohol and drug treatment. Although since the 1970s, nine studies in alcoholic patients and eight studies in patients abusing drugs have been published which mostly point to favorable effects of EX, most studies did not employ adequate control groups (alcohol [44–47], drugs [48–52]), had too small sample sizes (alcohol [47, 53, 54], drugs [49–52]), non-generalizable populations [55, 56], partly without clinical diagnoses [57], or no intention-to-treat-analyses to correct for the high number of dropouts (alcohol [58, 59], drugs [50, 60]).

However, there is preliminary evidence that EX can improve abstinence, as well as comorbid symptoms of depression and anxiety. RCTs with sufficient sample sizes are needed to confirm or disprove these findings. Besides effects specific for EX, the impact of structured social events, general lifestyle modifications, and a non-substance-related social environment are mechanisms that should be further evaluated in the context of SUD.

Schizophrenia (SZ)/psychosis

A recent review [61] included three RCTs (described below) and listed several other studies awaiting methodological assessment. Out of two RCTs comparing EX to standard care, one [62] found stronger (but non-significant) reductions in body fat, BMI, and the Positive and Negative Symptoms Scale (PANSS) after 16 weeks of aerobic EX, whereas the other [63] reported improvements in the Mental Health Inventory score after 12 weeks of combined aerobic and strength training, which were correlated with increased functional capacity. The third RCT found that 4 months of light EX were less effective than yoga therapy in reducing negative symptoms and depression and increasing social and occupational functioning and QoL [64].

Using a quasi-experimental design, an additional study [65] found significant reductions in both positive and negative symptoms after 10 weeks of moderate aerobic EX compared to standard therapy.

One possible mechanism of action in SZ is EX-induced neurogenesis, which has been demonstrated in a study finding not only EX-induced decreases in positive and negative symptoms, but also increases in hippocampal volumes after 3 months of aerobic EX, which were correlated with fitness increases [66].

To our knowledge, there is presently no evidence about the use of EX treatment in schizotypal, delusional, or schizoaffective disorder.

Summary

Evidence is weak for almost all groups of disorders with the exception of MDD. Studies using contact control groups generally yield smaller effects than studies comparing EX with no intervention. This leads to the conclusion that unspecific effects such as therapeutic contact, social support, and distraction may drive some of the effects, particularly of low-intensity EX.

Cost-efficacy cannot be estimated for any groups of disorders yet. Precise description of study conditions, standardized interventions, validated assessment strategies, adequate randomization strategies and control conditions, and power estimations are essential to obtain meaningful results. Future studies also should consider risks and adverse effects, besides the benefits of EX.

Implications for research/open questions

  • Full blinding is not possible (need for adequate placebo interventions as control conditions)

  • Patients know that EX is supposed to make them feel better (Rosenthal effect)

  • Dose–response relationship remains unclear (except for MDD and some aspects of anxiety)

  • Costs, efficacy, risks, adverse events, contraindications of EX interventions need to be specified

  • Which is the most effective type of EX for which mental disorder?

  • Are there differences concerning supervised versus unsupervised, indoor versus outdoor, group versus individual EX?

  • Does symptom improvement require changes in fitness? [39]

  • How can patients’ motivation be sustained during and after EX programs?

Implications for practice

  • Use individually adjusted training plans to maintain compliance, yet ensure an intensity level that is high enough to induce the intended changes (follow public health recommendations)

  • Provide professional supervision and training management, especially in the beginning, and integrate PA and EX into psychoeducation and psychotherapy (e.g., using training and mood diaries)

  • Respect disorder-specific conditions of patients (personality traits, frustration tolerance, cognitions, attitudes, apprehensions)

  • Consider differential acute effects depending on the patient’s training history and actual fitness: trained subjects experience greater improvements in vigor, positive affect, and fatigue than non-trained subjects (e.g., [67])

  • Make use of unspecific effects such as social support, time structure, therapeutic contact, and positive reinforcement

  • Ensure patients’ compliance during the EX intervention and continuation after the end of treatment