15.3: Exercise and Physical Activity in the Prevention of Depression

The potential for exercise and physical activity to prevent depression has been studied across a number of cross-sectional and prospective epidemiological studies. One of the earliest epidemiological studies to examine the relationship between physical activity levels and depression was conducted when Farmer and colleagues (1988) analyzed data from the Epidemiologic Follow-Up Study (1982–1984). Farmer et al. (1988) looked at the association between depressive symptomatology, as measured by the CES-D, and self-reported recreational and nonrecreational physical activity levels. Among 1,900 healthy adults between the ages of 25 and 77 included in analyses, increased depressive symptoms were related to lower levels of recreational physical activity. Interestingly, individuals engaging in little-to-no recreational physical activity were at almost a two-fold elevation in odds of experiencing increased depressive symptoms compared to those engaging in regular recreational physical activity. To determine whether recreational physical activity levels were a unique risk factor for increased depressive symptoms, the authors examined the influence of age, race, education, employment status, self-reported chronic health conditions, and household income on the relationship. After accounting for these variables, lower recreational physical activity levels still significantly related to increased depressive symptoms.

Goodwin (2003) extended this work by examining the relationship between regular physical activity levels and rates of MDD diagnoses among 5,877 Americans (~57% female) between the ages of 15 and 54 years who participated in the National Comorbidity Survey (1990–1992). Physical activity levels were assessed using a self-report item that asked participants, “How often do you get physical exercise—either in your job or in a recreational activity?”; MDD diagnoses were confirmed using a structured interview (i.e., the World Health Organization Composite International Diagnostic Interview).Goodwin found that regularly active individuals had a 25–38% reduced likelihood of having a MDD diagnosis compared to those who engaged in lower levels of physical activity. Importantly, this relationship remained significant even after accounting for important sociodemographic variables, such as age, gender, race, marital status, education, and income, suggesting that lower physical activity levels independently confer risk for MDD. There was also a significant dose-response³ relationship between frequency of physical activity and current prevalence of MDD in the sample, such that MDD was least prevalent among individuals who regularly engaged in physical activity (8.2%) compared to those who occasionally engaged in physical activity (11.6%), those who rarely engaged in physical activity (15.6%), and those who never engaged in physical activity (16.8%).

Gudmundsson and colleagues (2015) examined the relationship between physical activity and depression in a large-scale, prospective study of 676 Swedish women (M_age at baseline in 1974 = 53.4 years; SD = 0.2) followed over 32 years (1974–2005). Depression was measured using semi-structured interviews and the Montgomery-Åsberg Depression Rating Scale (Montgomery & Åsberg, 1979), which assesses the presence of symptoms and signs of depression over the past month. Physical activity levels were self-reported and measured using the Saltin-Grimby Physical Activity Level Scale (Saltin & Grimby, 1968), which indexes an individual’s physical activity frequency and intensity. Four separate assessments occurred across the 32-year study in 1974, 1992, 2000, and 2005, which allowed the authors to test several hypotheses. First, the authors examined cross-sectional associations and showed that lower physical activity levels were associated with increased depressive symptoms at the baseline assessment in 1974. Next, the authors were interested in determining whether changes in physical activity levels coincided with changes in depressive symptoms over time. There was a significant relationship between physical activity levels and depressive symptoms, such that decreasing physical activity levels were associated with increased depressive symptoms over time. Lastly, the authors tested whether the relationship between physical activity and depressive symptoms was bidirectional and wanted to determine whether initial depressive symptoms could predict subsequent physical activity levels. Women reporting increased depressive symptoms at the 1974 baseline assessment reported less engagement in subsequent physical activity at later time points in 1992 and 2000, but not in 2005. This study provides evidence for a potential bidirectional relationship between physical activity and depression from a large and well-characterized sample of women followed over the span of three decades and suggests that physical activity may not only be a risk factor for depression but may also be a long-term consequence of depression.

There may be specific aspects of depression that could explain the prediction of lower subsequent physical activity from depressive symptoms. For example, one of the core criterion symptoms of depression is the loss of interest or pleasure in almost all typically or previously enjoyable activities (i.e., anhedonia). In a study by Leventhal (2012), the author examined whether anhedonia was related to physical activity participation levels in a sample of 157 undergraduate students (M_age = 19.9 years; SD = 1.8). Increased anhedonia was associated with lower physical activity levels, providing preliminary support for the notion that anhedonia, rather than other symptoms of depression, may be associated with lower physical activity participation. It is important to be cautious in interpreting these preliminary data. Leventhal used a normative sample of undergraduate students to examine the relationship between concurrent levels of anhedonia and physical activity. The sample was relatively homogenous in terms of physical and mental health status and was more physically active compared to the general population. Further, the causative nature of the relationship cannot be established since the study employed a cross-sectional design. Therefore, these findings are only suggestive, but provide a template for examining specific symptoms of depression (e.g., anhedonia) and whether they can explain physical activity participation in clinically depressed samples.

Other prospective studies have shown significant relationships between physical activity levels and depression. Aten-year cohort study of 424 patients with depression (M_age at baseline = 39.9 years; SD = 14.1; 54% female) found that higher physical activity levels were associated with lower concurrent depressive symptoms at four separate assessment points spanning 10 years (Harris et al., 2006). This evidence corroborates previous findings demonstrating that increased physical activity is associated with a reduced incidence of depression diagnoses and symptoms.

The preventive effects of physical activity on depression (i.e., symptoms and diagnoses) have also been summarized in reviews (Mammen & Faulkner, 2013; Teychenne et al., 2008). Teychenne and colleagues (2008) performed a literature review of observational and interventional research on the relationship between physical activity and depression in adults. The authors were interested in identifying the influence of physical activity characteristics (i.e., dose, domain, and physical activity setting) on depression outcomes. There were inverse associations between physical activity and the likelihood of developing depression. Even lower doses of physical activity (i.e., 1.5 hours of moderate-intensity leisure time physical activity per week)demonstrated a protective effect against developing depression. Leisure-time physical activity was associated with a decreased likelihood of depression,while there was insufficient evidence regarding the role of other physical activity characteristics on depression.

In 2013, the Mammen and Faulkner review aimed to update the findings from Teychenne et al. by reviewing high quality research, which was determined using criteria from the Critical Appraisal Skills Programme (CASP; see https://casp-uk.net/) for prospective studies. Mammen and Faulkner aimed to address three important questions: (a) do baseline physical activity levels prevent follow-up depression?; (b) is there a specific dose of physical activity that protects against depression?; and (c) do physical activity levels over time impact risk for subsequent depression?. To address their first question, Mammen and Faulkner (2013) examined studies including assessments spanning at least two separate time intervals (range of follow-up period = 1–40 years) among nonclinical community samples of women/girls and/or men/boys between the ages of 11 and 100 years. Twenty-five of 30 reviewed studies supported the notion that baseline physical activity levels protect against the incidence or onset of subsequent depression. To address their second question, the authors found that engaging in as little as 10–29 min of daily physical activity could reduce subsequent risk for depression by 10% (e.g., Lucas et al., 2011), while other research showed that being physically active just one-to-two or more times per week was associated with a 40% reduced risk for depression (Bernaards et al., 2006; Hamer et al., 2009). To address their third question, the authors found that engaging in less physical activity over time was associated with an increased risk of developing depression compared to those who either maintained or increased their activity levels over the same time frame. These findings led Mammen and Faulkner to conclude that there was sufficient support for the preventive effects of physical activity on depression, even when physical activity is performed at low levels.

A few years later, Schuch et al. (2018) conducted a meta-analysis of 49 prospective studies that included 266,939 individuals (median proportion of males = 47%) with physical activity and depression measurements at baseline and a follow-up assessments (mean follow-up period = 7.4 years; range of follow-up period = 2–15 years) across studies. The authors found that individuals of all ages (youths, working-age adults, elderly persons) who engaged in more physical activity had 16.3% decreased odds of developing future depression. These findings were observed across geographical regions around the world (Asia, Europe, North America, and Oceania) and across genders.

Exercise programs may also have preventive effects. In 2020, Hu and colleagues performed a systematic review of eight meta-analyses that comprised a total of 134 individual studies examining the role of exercise interventions in the prevention of depression. In their review, Hu et al. (2020) examined the impact of exercise on depression across the lifespan among the general population and found moderate depressive symptom reductions among children, adolescents, adults, and the elderly. Although exercise interventions decreased depressive symptoms among the general population, the authors were unable to examine whether the incidence of MDD could be prevented, as no research has specifically investigated this relationship. A notable finding was that low-intensity exercise was comparable to higher intensities of exercise in reducing depressive symptoms; however, the authors noted that this evidence is of low-quality and needs to be investigated further.

The above-mentioned evidence highlights the preventive effects of physical activity and exercise on depression and symptoms of depression; however, the following questions remain: (a) does physical activity causally protect against—or decrease risk for—depression; or (b) does depression causally reduce physical activity? These two questions formed the basis for an investigation by Choi and colleagues (2019). To answer these questions, the authors used bidirectional Mendelian Randomization (MR),⁴ a genetically informed method for establishing causality in large-scale observational studies by testing whether genetic variants in a potential underlying trait (or risk factor) are associated with an outcome of interest. In bidirectional MR, researchers can evaluate whether the underlying trait causes the outcome and vice versa. In the Choi et al. (2019) study, the authors performed MR analysis in one direction (i.e., physical activity to depression), and then performed the analysis in the opposing direction (i.e., depression to physical activity) using genetic variants associated with each trait.

Choi et al. (2019) extracted physical activity and depression data from the U.K. BiobankStudy (Klimentidis et al., 2018; Wray et al., 2018) to examine relationships between physical activity and depression. Physical activity was assessed using device-based (accelerometry) and subjective (self-report) assessments and MDD was assessed using structured clinical interviews to make a lifetime MDD diagnosis,clinician-administered checklists,or medical record reviews (see Wray et al., 2018 for details). When testing the causal pathway of physical activity to depression, the authors found that each 1SD unit increase in accelerometer-assessed physical activity was associated with a 26% reduction in risk for developing MDD. This 1SD increase in accelerometer-assessed physical activity was reported as being equivalent to replacing time spent engaging in sedentary behaviors (e.g., sitting) with 15 min of vigorous physical activity, 60 min of moderate physical activity, or a combination of light and more vigorous activities. Notably, this relationship was only evident for the accelerometer-assessed measurements of physical activity, not self-reported physical activity, suggesting that measurement artifacts associated with different self-report instruments (e.g., cognitive biases, memory recall bias) could influence inferences drawn from studies examining the physical activity-depression relationship. When testing the causal pathway of depression to physical activity, the authors found no associations, regardless of how physical activity was measured. Thus, there seems to be more evidence supporting the causal pathway of physical activity to depression rather than the reverse. That is, physical activity may offer protection against the development of depression and symptoms of depression.

Overall, the available evidence indicates that physical activity and exercise are effective strategies for preventing depression. Although some research suggests that the relationship may be bidirectional (e.g., Gudmundsson et al., 2015), the study by Choi and colleagues (2019) indicates that there may be more robust evidence supporting the preventive role of physical activity on depression.

---

Reference

³ Dose-response refers to the amount of exercise or physical activity (duration, intensity, and frequency) that results in graded effects.

⁴ For additional reading on MR and how it can be used in physical activity studies to strengthen causal inferences, refer to the research by Byrne et al. (2017) and Choi et al. (2019), respectively. Zheng et al. (2017) also provide an overview of the different types of MR studies that can be used to evaluate causality in observational epidemiological studies.