Physical Activity Guidelines Advisory Committee Report
Part G. Section 8: Mental Health

• Introduction
• Review of the Science
  • Overview of Questions Asked
  • Data Sources and Process Used to Answer Questions
  • Question 1: Is There an Association Between Physical Activity and Depression?
    • Introduction
    • Does Physical Activity Protect Against the Onset of Depression Disorders or Depression Symptoms?
    • Does Physical Activity Reduce Symptoms of Depression?
    • Do the Effects of Physical Activity on Depression Symptoms Differ According to Age, Sex, Race/Ethnicity, or Medical Condition?
    • Do the Effects of Physical Activity Vary According to Features of Physical Activity, Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?
  • Question 2. Is There an Association Between Physical Activity and Anxiety?
    • Introduction
    • Does Physical Activity Protect Against the Onset of Anxiety Disorders or Anxiety Symptoms?
    • Does Physical Activity Reduce Anxiety Symptoms?
    • Do the Effects of Physical Activity on Anxiety Symptoms Differ According to Age, Sex, Race/Ethnicity, or Medical Condition?
    • Do the Effects of Physical Activity Vary According to Features of Physical Activity Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?
  • Question 3: Is There an Association Between Physical Activity and Psychological Distress and Well-Being?
    • Introduction
    • Does Physical Activity Protect Against the Onset of Feelings of Distress or Enhance Well-Being?
    • Does Physical Activity Reduce Feelings of Distress or Enhance Feelings of Well-Being?
    • Do the Effects of Physical Activity on Distress or Well-Being Differ According to Race/Ethnicity, Sex, Age, or Medical Condition?
    • Do the Effects of Physical Activity Vary According to Features of Physical Activity Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?
  • Question 4: Is There an Association Between Physical Activity and Cognitive Function and Dementia?
    • Introduction
    • Does Physical Activity Protect Against the Onset of Age-Related Decline in Cognitive Function or Dementia?
    • Does Physical Activity Reduce Symptoms Associated With Alzheimer’s Disease or Other Dementias?
    • Do the Effects of Physical Activity on Cognitive Function Differ According to Genetics, Age, Sex, Race/Ethnicity, or Medical Condition?
    • Do the Effects of Physical Activity Vary According to Features of Physical Activity, Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?
  • Question 5: Is There an Association Between Physical Activity and Sleep?
    • Introduction
    • Does Physical Activity Protect Against the Onset of Insomnia or Other Sleep Problems?
    • Does Physical Activity Reduce Sleep Problems and Enhance Sleep Quality?
    • Do the Effects of Physical Activity on Sleep Differ According to Age, Sex, Race/Ethnicity, or Medical Condition?
    • Do the Effects of Physical Activity Vary According to Features of Physical Activity Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?
  • Question 6: Is There an Association Between Physical Activity and Other Aspects of Mental Health?
    • Introduction
    • Self-Esteem
    • Chronic Fatigue
  • Question 7: Is There an Association Between Physical Activity and Adverse Psychological Events?
  • Question 8: What Mechanisms Can Plausibly Explain the Association Between Physical Activity and Mental Health?
• Overall Summary and Conclusions
• Research Needs
• Reference List

List of Figures

• Figure G8.1a Depression Symptoms: Prospective Cohort Studies, 1995 Through 2007: Crude Odds
• Figure G8.1b Depression Symptoms: Prospective Cohort Studies, 1995 Through 2007: Adjusted Odds
• Figure G8.2 Depression Symptoms: Randomized Controlled Trials, 1995 Through 2007
• Figure G8.3 Depression Symptoms: Prospective Cohort Studies 1995 Through 2007: Dose Response
• Figure G8.4a Anxiety Symptoms: Randomized Controlled Trials of Healthy Adults 1995 Through 2007
• Figure G8.4b. Anxiety Symptoms: Randomized Controlled Trials of Medical Patients 1995 Through 2007
• Figure G8.5 Feelings of Distress/Well-Being: Prospective Cohort Studies, 1995 Through 2007: Crude and Adjusted Odds
• Figure G8.6 Feelings of Distress/Well-Being: Randomized Controlled Trials 1995 Through 2007
• Figure G8.7 Feelings of Distress/Well-Being: Prospective Cohort Studies 1995 Through 2007: Dose Response
• Figure G8.8 Incident Total Dementia or Alzheimer’s Disease: Prospective Cohort Studies, 1995 Through 2007: Crude and Adjusted Hazard
• Figure G8.9 Physical Activity and Symptoms of Disrupted or Insufficient Sleep: 13 Cross-Sectional Studies (Total n=84,904)

Introduction

Poor mental health, including diseases of the central nervous system (CNS), reduces the quality of life and adds a burden on public health. People with anxiety or depression disorders are more likely to have chronic physical conditions (1), and depression and dementia were among the 10 leading risk factors of disability-adjusted life expectancy in high-income nations worldwide during 2001 (2). They are projected to rank first and third by the year 2030 (3). In the United States, dementia and other CNS disorders are a leading cause of death, and mental disorders are estimated to account for more than 40% of years lost to disability (4).

The scientific evidence from prospective cohort studies and randomized controlled trials (RCTs) supports the overall conclusion that regular participation in moderate-to-vigorous physical activity is associated with improved aspects of mental well-being and reduced symptoms of several mental health disorders.

Review of the Science

Overview of Questions Asked

This chapter addresses 7 general questions about physical activity and mental health conditions. Each question includes subsections that focus on whether physical activity can protect against the onset of, or reduce, symptoms; whether the effects of physical activity on the symptoms differ by age, sex, race/ethnicity, or medical condition; and whether the effects of physical activity differ by the type, intensity, or timing of the physical activity. The general questions are:

1. Is there an association between physical activity and depression?
2. Is there an association between physical activity and anxiety?
3. Is there an association between physical activity and distress and well-being?
4. Is there an association between physical activity and cognitive function and dementia?
5. Is there an association between physical activity and sleep?
6. Is there an association between physical activity and other aspects of mental health?
7. Is there an association between physical activity and adverse psychological events?

The chapter also considers a final question dealing with mechanisms that could plausibly explain the association between physical activity and mental health.

Data Sources and Process Used to Answer Questions

To provide evidence-based answers to the above questions, the Mental Health subcommittee obtained data from a search of the Physical Activity Guidelines for Americans Scientific Database (see Part F: Scientific Literature Search Methodology, for a full description of the Database), which contains studies published in 1995 and later. Conclusions regarding the evidence-based review were restricted to results of RCTs and observational studies that used a prospective cohort design. Findings from selected cross-sectional, observational studies are presented when they provided additional information for questions that had a limited number of RCTs and prospective cohort studies (e.g., anxiety disorders and sleep). Studies of physical activity and mental health published since 1995 were evaluated using meta-analytic or otherwise systematic reviews when they were available. When such reviews were not available or were not sufficiently current, a systematic quantitative synthesis of results from published studies was conducted using a random effects model that weights results by the size of each study and variation in the effects across studies. Odds ratios in observational studies and standardized effect sizes in RCTs were retrieved from the published papers, or they were otherwise estimated from data and test statistics reported in the papers (5).

Question 1: Is There an Association Between Physical Activity and Depression?

Introduction

The American Psychiatric Association (6) recognizes 4 types of mood disorders: (1) depression, (2) bipolar or manic-depressive disorder, (3) mood disorders due to a medical condition, and (4) substance-induced mood disorders. Depression has an annual prevalence of about 8% among women and 4% among men worldwide and in the United States. The annual cost of depression in the United States is estimated at $83 billion per year (7). This condition includes a mild chronic form, dysthymia, and a more severe form, major depressive disorder. The rate of major depression has increased steadily during the past 50 years, with a lifetime prevalence of about 16%; the rate is higher among Hispanics than whites, and lowest, though still substantial, among African Americans. The lifetime rate of depression among adults aged 30 to 60 years is about twice the rate among people older than age 60 years (8).

People have a major depressive episode when they have depressed mood or lose interest or pleasure in normal activities most of the time for at least 2 weeks. Other symptoms include abnormalities in appetite, libido, sleep, energy levels, concentration and, often, suicidal thoughts. In some cases, anxiety and motor agitation can be more prominent symptoms than depressed mood. Also, mood disturbance can be less apparent than other features such as irritability, abuse of alcohol, and worsening of comorbid phobias, obsessions, or preoccupation with physical symptoms. Depression is not considered a major depressive episode if it is caused by grief (less than 2 months), drug abuse or medication, or a medical condition such as hyperthyroidism, heart disease, diabetes, multiple sclerosis, hepatitis, or rheumatoid arthritis. Many older patients with symptoms of depression do not meet the full criteria for major depressive disorder. If they have similar, but fewer, symptoms they may have minor depression, a sub-syndromal form of depression.

Does Physical Activity Protect Against the Onset of Depression Disorders or Depression Symptoms?

Conclusions

Population-based, prospective cohort studies provide substantial evidence that regular physical activity protects against the onset of depression symptoms and major depressive disorder. Evidence is insufficient to draw conclusions about bipolar disorder and other mood disorders.

Rationale

An association between physical activity and reduced symptoms of depression among adults has been generally supported in more than 100 population-based observational studies published since 1995, including nationally representative samples of nearly 190,000 Americans (9-15). Most of the studies looked at cross-sectional associations, which indicated that active people on average had nearly 45% lower odds of depression symptoms than did inactive people. In the national samples of Americans, active people had approximately 30% lower odds of depression.

Twenty-eight of the studies used a prospective cohort design, which reduces the likelihood that the association is explainable by people becoming less active after they experience depression symptoms. The median follow-up was about 4 years, and the range was 9 months to 37 years. The studies came from 11 nations (Australia, Canada, China, England, Finland, Germany, Israel, Italy, Netherlands, Japan), including 13 studies from the United States. In the cohort studies, the average odds of elevated symptoms were about 25 to 40% lower among active compared with inactive people, without adjustments for depression risk factors that might have differed between the active and inactive groups (OR = 0.67, 95% CI = 0.59 to 0.77). After adjustment for risk factors, such as age, sex, race, education, income, smoking, alcohol use, chronic health conditions, and other social and psychological variables, the odds remained nearly 15 to 25 percent lower among active people (OR = 0.82, 95% CI = 0.78 to 0.86).

Nearly all the comparisons (66 of 67) in the cohort studies showed less depression among physically active than inactive adults, but half the results did not reach statistical significance, often because the sample sizes were not big enough. The average number of people for each comparison was about 1,100 people, but a fourth of the comparisons included 500 people or fewer. Studies of 8 cohorts (16-23) used a clinical diagnosis to measure depression symptoms, reporting a reduction in the odds of incident cases of depression that averaged 30 percent (OR = 0.71, 95% CI: 0.61 to 0.77) among active people compared to inactive or low active people. In those studies, 18 of 19 comparisons favored active people, and nearly half the comparisons were statistically significant. Thus, the protective effect of physical activity is not limited to self-rated symptoms measured by questionnaires. Figures G8.1a and G8.1b illustrate crude and adjusted odds ratios and 95% confidence intervals from the 28 prospective cohort studies of physical activity and depression in nearly 40,000 adults from 11 countries, including 11 studies from the United States.

Does Physical Activity Reduce Symptoms of Depression?

Conclusions

The results of RCTs indicate that participation in physical activity programs reduces depression symptoms in people diagnosed as depressed, healthy adults, and medical patients without psychiatric disorders.

Rationale: Depressed Patients

A meta-analysis of 14 RCTs of chronic exercise among people diagnosed with depression (24) reported a cumulative, mean reduction in depression symptoms of 1.1 SD (95% CI: 1.5 to 0.60). However, the studies had scientific weaknesses that made it hard to conclude that the reduced depression symptoms were the independent result of exercise, and only 2 of the studies had been published after 1995 (25;26). Since that review, at least 11 RCTs have used exercise training to reduce depression symptoms in about 500 depressed patients. In 3 studies (27-29), depression was identified using cut-point scores on symptom questionnaires that have good predictive validity as screening tests to detect depression. In another 8 studies (30-37), patients were identified by diagnostic interview as having major depressive disorder, minor depression, or dysthymia.

In the studies published since the Lawlor & Hopker review (24), the average effect of exercise for symptom reduction was 1.1 SD (95% CI: 1.55 to 0.64) compared to a no treatment control group, virtually identical to the Lawlor & Hopker results. These recent studies compared the effects of exercise to a placebo (29;32-34;36;38) or to drug therapy (30;38) or bright light therapy (35). The reductions, on average, favored exercise over placebo (0.35 SD, 95% CI: 0.49 to 0.22), and 6 of 14 comparisons were statistically significant. The average sample size was 58 people, but 25% of the studies had 25 people or fewer. Symptom reduction after exercise was similar to drug therapy and similar or superior to bright light therapy. Four of the studies reported that reductions in symptoms met criteria for a clinically meaningful decrease in symptoms (50%) or remission (32;35;37;38). Long-term reductions in symptoms were examined in a few studies and were generally favorable after supervised exercise training ended, especially when people maintained regular physical activity (28;29;36;39). The trial by Dunn and colleagues. (2005) was a placebo trial that controlled for social interaction and sunlight exposure (2 confounders common in other studies) and reported response and remission rates after exercise training that were 40 to 50% better than placebo (32).

Figure G8.1a Depression Symptoms: Prospective Cohort Studies, 1995 Through 2007: Crude Odds

Figure G8.1a Data Points

Figure G8.1b Depression Symptoms: Prospective Cohort Studies, 1995 Through 2007: Adjusted Odds

Figure G8.1b Data Points

Rationale: Healthy Adults and Adults With Medical Conditions

At least 42 RCTs since 1995, including more than 2,600 people, have examined the effects of exercise on depression symptoms in healthy adults or adults with medical conditions other than depression or disabling conditions that severely limit physical activity (i.e., spinal cord injury, multiple sclerosis, and stroke or traumatic head injury). Most of the medical patients studied had cardiovascular diseases, arthritis, chronic pain, obesity, or were cancer survivors. The mean effect of exercise compared to a control condition was 0.35 SD (95% CI: 0.24 to 0.46). The outcomes favored exercise in 90% (34 of 38) of the comparisons with control conditions, of which 60% (23 of 38) reached statistical significance. The average sample size was about 65 people, but in about a fourth of the studies it was less than 35, which was too small to detect an effect smaller than a half standard deviation (SD). When compared to a placebo (usually stretching or health education), the effect of exercise was halved to 0.15 SD (95% CI: 0.24 to 0.06). Although exercise effects exceeded placebo effects in 19 of 22 comparisons, only 4 reached statistical significance. The average sample size in the placebo studies was nearly 100 people, but samples more than 4 times that size would be needed to detect an effect that small.

Do the Effects of Physical Activity on Depression Symptoms Differ According to Age, Sex, Race/Ethnicity, or Medical Condition?

Conclusions

The available evidence supports the conclusion that regular physical activity reduces depression symptoms regardless of age, sex, race/ethnicity, or medical condition. Whether these factors modify the association between physical activity and depression has been understudied. Also, race and ethnicity have been poorly represented or not described in most studies.

Rationale: Cohort Studies

Few prospective cohort studies reported findings according to sex or age groups. Just 1 study reported results for men and women, finding that lower levels of depression symptoms were associated with more physical activity in both men and women (40). Only 11 studies had separate results for men or women. Reductions were smaller for active men (OR = 0.96, 95% CI = 0.93 to 0.99) than women (OR = 0.72, 95% CI = 0.68 to 0.77), but men were less studied. Five studies totaled fewer than 4,000 men, while 6 studies totaled more than 14,000 women. After adjusting for the number of comparisons made in each of the other cohort studies, in which men and women were represented evenly, effects for men and women did not differ. Crude odds reduction among active people were greater for people aged 55 years or older (OR = 0.58; 95% CI = 0.46 to 0.72) than people younger than age 55 year (OR = 0.72; 95% CI = 0.64 to 0.81), but physical activity was protective for all age groups, regardless of sex.

Most of the prospective cohort studies were population-based, but only a third specified the proportions of racial and ethnic groups included, and only 4 studies had evidence of good representation of African Americans and/or Hispanics/Latinos (41-44). One study reported that the odds of depression symptoms were similarly lower among white and African American adults who were active (42). Other studies had poor or no representation of other minority groups.

Rationale: Randomized Controlled Trials

In RCTs of people without depression, the average effect of exercise compared to a control condition is 0.35 SD in people with medical conditions (24 studies; 95% CI: 0.47 to 0.23) and 0.35 SD in people who have not been diagnosed with a medical condition (7 studies; 95% CI: 0.59 to 0.11). Figure G8.2 illustrates effect sizes and 95% confidence intervals comparing exercise with a control condition from the RCTs of depression in adults with or without medical conditions.

These trials reported similar reductions in depression for groups of varying ages from 28 to 83 years, but the studies were not designed to compare groups according to sex, race/ethnicity, or age. Only 1 of the trials compared men and women (45) and none compared effects between people of different ages or race/ethnicities. Seven of 11 studies of Americans described the racial/ethnic composition of their samples, but only 5 studies included African Americans (46-50); 3 studies included small numbers of Hispanics, Asian Americans or American Indians (49-51). Effects were larger in 6 studies of men (mean = 0.86 SD, 95% CI = 0.45 to 1.26), compared to 9 studies of women (mean = 0.27 SD, 95% CI = 0.04 to 0.51) and to 23 studies that combined results of men and women (mean = 0.27 SD, 95% CI = 0.16 to 0.37). Thus, it is likely that other features of the studies of men might explain the larger effects.

Reductions in depression symptoms were larger in 8 studies of heart patients (mean = 0.60 SD, 95% CI = 0.33 to 0.86) than in studies of other medical conditions (mean = 0.23 SD, 95% CI = 0.11 to 0.34) and were largest in the 3 studies of heart patients that included only men (52-54) (mean = 1.0 SD, 95% CI = 0.50 to 1.5). The small number of studies precluded meaningful comparisons among the other medical conditions.

Figure G8.2 Depression Symptoms: Randomized Controlled Trials, 1995 Through 2007

Figure G8.2. Data Points

Do the Effects of Physical Activity Vary According to Features of Physical Activity, Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?

Conclusions

The evidence from prospective cohort studies and RCTs published since 1995 suggests that moderate and high levels of physical activity similarly reduce the odds of developing depression symptoms compared to low levels of physical activity exposure, which is nonetheless more protective than inactivity or very low levels of physical activity. The minimal or optimal type or amount of exercise for reducing depression symptoms is not yet known, but it appears that an increase in physical fitness is not required.

Rationale

Epidemiologic studies typically use a variety of criteria and methods to classify people into 2 or more activity groups. This limits the evaluation of the dose-response relation across the full range of reported physical activity and can misclassify people who overestimate or underestimate their activity. Only 7 prospective cohort studies of depression symptoms included the 3 or more levels of physical activity necessary to determine whether the association of physical activity with lower odds of depression has a dose-gradient with increased levels of exposure (16;18;21;22;55-57). After adjustment for age, sex, and other risk factors, the reduction of odds was smaller for the lowest level of physical activity (OR = 0.86, 95% CI = 0.79 to 0.94) compared to the next 2 levels of physical activity, which did not differ (OR = 0.77, 95% CI = 0.72 to 0.82). Thus, the highest levels of participation did not confer more protective benefits than did more moderate levels. Each was more protective than the lowest levels in the studies.

Figure G8.3 shows the odds ratios and 95% CI in prospective cohort studies that examined the dose-response association between levels of physical activity and depression symptoms.

Those studies used different measures and criteria for defining levels of physical activity, so it is not possible to convert their findings to a standard estimate (e.g., MET-hours or kilocalories per kilogram) of the amount of physical activity at each level. However, about half the prospective cohort studies provided enough information to determine whether active people were meeting existing public health recommendations (58;59) for participation in moderate or vigorous physical activity (i.e., moderate-intensity aerobic [endurance] physical activity for a minimum of 30 minutes on 5 days per week, or vigorous-intensity aerobic physical activity for a minimum of 20 minutes on 3 days per week). Odds reduction favored people who met or exceeded recommendations for moderate or vigorous participation (OR = 0.64, 95% CI = 0.57 to 0.73) compared to active people who did not meet either recommended level (OR = 0.70, 95% CI = 0.58 to 0.82). Odds were not different between vigorous and moderate participation, but it was possible to distinguish moderate from vigorous participation in only a few studies. After adjustment for other risk factors, a similar protective benefit favored regular participation in moderate or vigorous physical activity (OR = 0.77, 95% CI = 0.72 to 0.82) compared to participation at levels less than recommended (OR = 0.84, 95% CI = 0.78 to 0.90).

Figure G8.3 Depression Symptoms: Prospective Cohort Studies 1995 Through 2007: Dose Response

Figure G8.3 Data Points

A few RCTs of people without depressive disorders have manipulated type (e.g., aerobic versus resistance or walking versus aquatic exercise or Qigong) or timing (i.e., continuous versus intermittent [the intermittent studies often used resistance, circuit, interval or mixed-mode exercise]), of exercise to examine whether those features modify the effects of exercise on symptoms of depression (60-64). However, these studies did not include a control group who did not exercise. Only a few studies have evaluated Eastern health practices that include exercise (65). About 40 percent of the studies used an aerobic exercise intervention such as walking, jogging, cycling, or aquatic exercise, and another third of the studies combined aerobic activity with resistance exercise. Only 3 studies used resistance exercise alone, and none compared aerobic versus resistance exercise. On average, reductions in depression have been similar regardless of the mode of activity used. About two-thirds of the studies used continuous exercise. Regardless of mode, a larger reduction in depression tends to occur after continuous exercise (0.45 SD, 95% CI = 0.63 to 0.27) than intermittent exercise (0.18 SD, 95% CI = 0.30 to 0.06), but differing features of the studies other than timing might explain this finding.

Three-fourths of the RCTs of healthy adults and non-psychiatric medical patients used a moderate-to-vigorous exercise intensity of 60-80% of people’s aerobic capacity or maximum strength that occurred 3 days per week. Intensity was lower or the frequency was 2 days per week in the other studies. The average duration of each session was about 35 minutes, but it was less than 30 minutes in a fourth of the studies and more than 1 hour in another fourth of the studies. However, fewer than half the studies were clear about how the time was partitioned into warm-up, exercise, and cool down. Nonetheless, reductions in depression symptoms did not differ across these varying features of exercise. Studies lasted an average of 6 months, and a fourth of the studies lasted less than 3 months. Length of the exercise program also was unrelated to symptom outcome. About half the studies measured cardiorespiratory fitness, which was increased significantly in 16 of 24 comparisons. Only 3 studies measured strength. In each study, fitness increases and symptom reduction were not associated when changes were defined by statistical significance, which depends on both the size of change and the sample size. After adjusting for sample sizes, the magnitude of depression reduction was moderately correlated with the magnitude of fitness increase (r = 0.40). However, that relation was not independent of increases in primary outcomes other than fitness in the studies of medical patients.

Perhaps the clearest experimental evidence for a dose-dependent effect of exercise on symptom reduction comes from the dose study (32). Adults aged 20 to 45 years diagnosed with mild to moderate major depressive disorder expended either 7.0 or 17.5 kilocalories per kilogram per week at a frequency of 3 or 5 days per week or engaged in 3 days per week of stretching exercises as a placebo control. Physician-rated symptoms after 12 weeks were reduced 47% from baseline for the higher dose, compared with 30% for the lower dose and 29% for control, regardless of whether exercise frequency was 3 days or 5 days each week.

Question 2. Is There an Association Between Physical Activity and Anxiety?

Introduction

Anxiety is characterized by apprehensive or worrisome thoughts and is typically accompanied by agitation, feelings of tension, and activation of the autonomic nervous system. A distinction is made between transient anxiety symptoms, termed state anxiety, persistent symptoms, termed trait anxiety, and a group of disabling conditions characterized by excessive, chronic anxiety that are known as anxiety disorders. The anxiety disorders, listed from most to least common, are:

• Specific phobia — an intense fear of an object, place, or situation that poses little or no actual danger.
• Social phobia — an overwhelming fear of scrutiny and embarrassment in social situations, leading to avoidance of potentially enjoyable activities.
• Generalized anxiety disorder — recurrent or persistent excessive worry about everyday, routine life events and activities, lasting at least 6 months.
• Panic disorder — repeated episodes of intense fear and physical symptoms that strike without warning and without an obvious source, often producing fear of being alone or going into public places (agoraphobia) and persistent fear of an attack.
• Obsessive-compulsive disorder — repeated, unwanted thoughts or compulsive behaviors that seem impossible to stop, typified by repetitive acts or rituals to relieve anxiety.
• Post–traumatic stress disorder — a delayed or prolonged response (including flashbacks, dreams, insomnia, hypervigilance) to a stressful event or situation (either short- or long-lasting) that was especially threatening or catastrophic (6).

Anxiety disorders are common, affecting more than 16 million people in the United States each year (roughly 4% of women and 2% of men). More than 80 million people in the United States at some point in their lives suffer from an anxiety disorder (8). Anxiety disorders begin at a median age of 15 years, often persist throughout life and are associated with numerous physical and mental co-morbidities, especially depression (66). People aged 15 to 24 years experience episodes of anxiety about 40% more often than people aged 25 to 54 years, regardless of race. Although less than 30% of those who suffer from anxiety disorders seek treatment, they strain the health care system because of direct psychiatric and nonpsychiatric treatment costs. Additional indirect costs of anxiety disorders are incurred from reduced work productivity. The total annual costs in 1990 of all anxiety disorders were estimated to be $42 billion to $47 billion dollars (67;68), and they likely cost double that amount today.

Does Physical Activity Protect Against the Onset of Anxiety Disorders or Anxiety Symptoms?

Conclusions

The weight of evidence from a small number of nationally representative and population-based cross-sectional and prospective cohort studies supports that regular physical activity protects against the onset of anxiety disorders and anxiety symptoms.

Rationale

At least 4 population-based cross sectional studies published since 1995, including data from nationally representative samples of nearly 121,000 Americans, show that regular physical activity is associated with lower odds of anxiety symptoms (13;69-71). Results of the US National Co-Morbidity Study found that regular physical activity reduced the odds of a diagnosed anxiety disorder (i.e., specific phobia, social phobia, generalized anxiety, panic, and agoraphobia) by an average of 43% (10). After controlling for sociodemographic and illness variables, regular physical activity reduced the odds of an anxiety disorder by an average of 28% (10).

At least 2 population-based studies used a prospective cohort design. The odds of developing any anxiety disorder were reduced by an average of 53% among Australians who reported more than 3 hours per week of vigorous physical activity compared to those reporting no activity (16). The effect was not statistically significant, in part due to the small number of participants who developed an anxiety disorder (n=67). The second study, which adjusted for age and sex and had more participants who developed an anxiety disorder (n=228), found that regularly active young adults (representative of Munich, Germany) on average had a statistically significant, 48% lower odds of developing any anxiety disorder compared to those reporting no activity (22).

Does Physical Activity Reduce Anxiety Symptoms?

Conclusions

The results of RCTs conducted with medical patients and healthy adults indicate that participation in physical activity programs reduces anxiety symptoms.

Rationale

Before 1995, no reports of RCTs of anxiety disorder patients had been reported. Since then, 2 RCTs have been conducted with anxiety disorder patients and reported statistically large effects. One trial conducted with 46 panic disorder patients found that those who completed a 10-week (3 times per week) walking or jogging program reported, on average, a large reduction in anxiety symptoms (1.1 SD) compared to those who took daily placebo capsules (72). The other trial involved 74 patients with social phobia, generalized anxiety, or panic disorder. The addition of a moderate intensity home-based exercise program to 8 to 10 weeks of group cognitive-behavioral therapy (GCBT) resulted in a large reduction in anxiety symptoms (1.36 SD) compared to the control condition (GCBT + nutrition education) after statistically controlling for potential confounding variables (73).

Before 1995, at least 40 quasi-experimental and experimental exercise training studies of varying quality had reported a cumulative effect on reducing anxiety symptoms of approximately 0.40 SD (74). Since 1995, at least 46 RCTs involving more than 3,550 people have examined the effects of chronic exercise on anxiety symptoms among inactive adults who were healthy or had a medical condition other than anxiety disorders or disabling CNS disorders including multiple sclerosis, traumatic head injury, stroke or spinal cord injury. The effect of exercise compared to control conditions in reducing anxiety symptoms was 0.38 SD (95% CI: 0.30 to 0.46). Outcomes favored exercise in 84% (67 of 80) of the comparisons with control conditions, and 28% (22 of 80) reached statistical significance. In 6 studies that compared moderate-to-vigorous exercise to a placebo-type condition (usually low intensity exercise such as stretching) the effect of moderate-to-vigorous exercise on reducing anxiety symptoms was 0.19 SD (95% CI: 0.05 to 0.33). Exercise effects were favorable in 90% (9 of 10) of the placebo-type comparisons, and 30% reached statistical significance. The average study included about 60 people, but a fourth of the studies had less than 40 participants, too few to detect small but possibly meaningful effects.

Do the Effects of Physical Activity on Anxiety Symptoms Differ According to Age, Sex, Race/Ethnicity, or Medical Condition?

Conclusions

The weight of the evidence supports the conclusion that regular physical activity reduces anxiety symptoms regardless of age, sex or medical condition. No published data address whether race or ethnicity modifies the effects of physical activity on anxiety symptoms.

Rationale

In the 46 RCTs of healthy adults and medical patients, the mean age of the samples was weakly related to anxiety reductions after exercise training. Two of the trials compared men and women and found no sex-related difference in the effect of exercise on anxiety symptoms (45;75). The average effect of exercise training on reduced symptoms of anxiety was larger in those studies that involved only men (mean = 0.62 SD, 95% CI: 0.34 to 0.89, n=7) compared to those that included only women (mean = 0.33 SD, 95% CI: 0.16 to 0.50, n=12) and those that included similar proportions of men and women (mean = 0.42 SD, 95% CI: 0.28 to 0.57, n=19). It is likely that other features of the studies of men might explain the larger effects. Regardless, exercise benefited both men and women.

None of the available investigations was designed to determine whether the effect of chronic exercise was moderated by race or ethnicity. The racial or ethnic composition of the samples was described in 17% of the reviewed studies (25;51;76-81). Of the 3,550 people who participated in the RCTs, approximately 89% were whites and approximately 11% were African Americans or Hispanics. Statistical analyses by race or ethnic category were not presented in any of the studies.

Of the RCTs, about 40% (19 of 46) were conducted with healthy adults, and 60% (27 of 46) involved patients with various medical conditions. The most frequently studied conditions were cancer and cardiovascular diseases. The average effect of exercise compared to a control condition was 0.40 SD in healthy adults (40 comparisons; 95% CI: 0.27 to 0.53) and 0.36 in people with medical conditions (40 comparisons; 95% CI: 0.26 to 0.47). The reduction in anxiety symptoms was somewhat larger than average in studies of cardiovascular conditions (mean = 0.53 SD, 95% CI: 0.15 to 0.92, n=6), but too few studies are available to conclude that the effect of exercise training on anxiety symptoms is modified by medical condition. Figures G8.4a and G8.4b summarize the findings from the RCTs of physical activity and anxiety symptoms in healthy adults and medical patients.

Epidemiologic studies have not yet examined race/ethnicity or medical condition as potential modifiers of the effect of exercise on anxiety disorders or symptoms. One cross-sectional study of 41,914 participants in the 2001 Behavioral Risk Factor Surveillance System found low anxiety symptoms among physically active people compared to inactive people across young (age 18 to 40 years), middle-aged (aged 41 to 60 years) and older adult (aged 61 years and older) categories, but the inactive younger adults were approximately 20% more likely to experience anxiety symptoms than inactive middle-aged and older adults (13). In a population-based cross sectional study of 19,288 twins and their families, potential interactions of exercise and age, exercise and sex and exercise, age and sex on anxiety symptoms were statistically non-significant (69).

Do the Effects of Physical Activity Vary According to Features of Physical Activity Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?

Conclusions

Limited cross-sectional, observational evidence suggests that the odds of an anxiety disorder may be reduced by higher weekly frequency of exercise bouts. However, there is an absence of evidence from prospective cohort studies or RCTs that examine whether anxiety symptoms vary according to features of physical activity exposure.

Rationale

One large population-based study found a dose-response relation between cross-sectional measures of physical activity frequency and lower prevalence of anxiety disorders (i.e., specific phobia, social phobia, generalized anxiety, panic, and agoraphobia) (10). The percentage of adults with these anxiety disorders was highest among those who reported no physical activity and was reduced in a step-wise fashion among those who reported rare, occasional, and regular physical activity (10).

Figure G8.4a Anxiety Symptoms: Randomized Controlled Trials of Healthy Adults 1995 Through 2007

Figure G8.4a Data Points

Figure G8.4b. Anxiety Symptoms: Randomized Controlled Trials of Medical Patients 1995 Through 2007

Figure G8.4b Data Points

Prospective cohort studies show mixed results as to whether participation in sports, activities that often involve high-intensity exercise, has a protective effect against anxiety symptoms. Some show a protective effect (23) and others show no protection (82). One nationally representative cross-sectional study found that a group of regularly physically active (i.e., met or exceeded the Healthy People 2010 physical activity guidelines) people reported a higher frequency of anxiety symptoms compared to those who were physically active but not active enough to meet the Healthy People 2010 recommendations (13).

Few RCTs of healthy adults or people with medical conditions have manipulated any feature of an exercise intervention to learn whether those features causally modify anxiety symptoms. No exercise training studies have manipulated exercise program length or exercise type for the purpose of examining anxiety symptom outcomes.

About 43% (20 of 46) of the studies used a single aerobic exercise intervention such as walking, jogging, or cycling, 13% (6 of 46) used resistance exercise alone, and 9% (4 of 46) combined aerobic with resistance exercise. The magnitude of anxiety reduction has been similar across these 3 categories of studies. Regardless of mode, anxiety reductions also were similar after continuous (0.36 SD, 95% CI: 0.27 to 0.45) and intermittent (0.39 SD, 95% CI: 0.16 to 0.63) exercise (about 60% of studies used continuous exercise).

Three investigations manipulated exercise session duration and found no significant differences in anxiety symptom outcomes among exercise training studies that used bouts of varying session durations (80;83;84). Across all trials, the average duration of the exercise sessions was approximately 40 minutes. In a fourth of the studies, the duration was less than 25 minutes, and in another fourth the duration was more than 60 minutes. The effects of exercise training on anxiety symptoms were similar across studies of all durations. Imprecise descriptions of exercise duration contributed to the difficulty in determining whether session duration has a true effect. For example, less than half of the studies were specific regarding how time was partitioned into warm-up, exercise, and cool down.

Only 2 exercise training experiments manipulated exercise intensity to examine whether it modifies the effects of exercise on anxiety symptoms. Statistically non-significant differences in anxiety symptoms were found between moderate- and high-intensity resistance exercise conditions on the anxiety symptom outcomes (81;85). About 55% (44 of 80) of the comparisons from all the trials used moderate-to-vigorous exercise intensity (i.e., 60 – 80% of aerobic capacity or maximum strength) with a weekly frequency of 3 or more days per week. Reductions in anxiety symptoms were similar across variations in exercise intensity.

About 63% (29 of 46) of the trials measured fitness, 7 studies measured strength, and 22 studies measured cardiorespiratory fitness. Cardiorespiratory fitness was increased significantly in 36% (13 of 36) of the comparisons and strength was increased significantly in 25% (2 of 8) of the comparisons. Those studies that defined changes by statistical significance showed no association between fitness increases and anxiety symptom reduction. After adjusting for sample sizes, the magnitude of anxiety reduction was weakly correlated with the magnitude of fitness increase (r = .24). This relation was independent of increases in primary outcomes other than fitness in the studies.

Question 3: Is There an Association Between Physical Activity and Psychological Distress and Well-Being?

Introduction

Psychological distress is a risk factor for psychiatric disorders (86;87) and coronary heart disease (88), and it is negatively associated with quality of life. Conversely, a feeling of well-being can reduce psychiatric risk and is an important feature of high life quality and health (89). People frequently experience feelings of distress during the normal course of living and during challenging life events, including chronic medical conditions. Thus, it is important to understand the association between physical activity and feelings of distress or well-being because they bear not only on disease risk but also on overall mental health. Measures in this area are not uniform, but most studies have used a scale that assessed the presence of distress (e.g., combined symptoms of anxiety and depression or perceived stress) or the absence of distress (e.g., well-being or positive mental health). Findings of physical activity studies have not differed when measures of distress or well-being were used, so the following results apply regardless of the direction of odds (i.e., decrease in distress or increase in well-being).

Does Physical Activity Protect Against the Onset of Feelings of Distress or Enhance Well-Being?

Conclusions

The available evidence from prospective cohort studies indicates a small-to-moderate association that favors people who are physically active.

Rationale

The association between physical activity and reduced feelings of distress or enhanced well-being among adults was virtually unstudied in large groups of people before 1995 (90). Since then, more than 30 population-based observational studies have been published, including nationally representative samples of more than 175,000 Americans (91-95). Most of the studies looked at cross-sectional associations, which indicated that active people on average had more than a 30% lower odds of feeling of distress or 30% higher odds of enhanced well-being than did inactive people. In the national samples of Americans, the odds favored active people by approximately 25%.

Thirteen studies of adults in Australia, Canada, Denmark, England, Netherlands, Scotland, Wales, and 3 studies of Americans (18;23;94) used a prospective cohort design. In those studies, the average odds of reduced feelings of distress or of enhanced well-being favored active people by about 30% compared with inactive people, without adjustments for risk factors (OR = 0.69, 95% CI = 0.61 to 0.78). After adjustment for risk factors, such as age, sex, race, education, social class, occupation, income, smoking, alcohol use, substance abuse, chronic health conditions, disability, marital status, life events, job stress, and social support, the odds still favored active people by nearly 20% (OR = 0.82, 95% CI = 0.77 to 0.86). About 80% of the comparisons (58 of 70) favored active adults, but half the results did not reach statistical significance, often because of too small sample sizes. The average number of people for each comparison was approximately 1,300 people, but a fourth of the comparisons had 700 people or less. Figure G8.5 illustrates 18 crude and/or adjusted odds ratios and 95% confidence intervals from the 13 prospective cohort studies of physical activity and distress or well-being in more than 100,000 adults from 8 countries, including 67,000 American women (23;94).

Does Physical Activity Reduce Feelings of Distress or Enhance Feelings of Well-Being?

Conclusions

The effects of RCTs indicate small benefits of physical activity that often do not exceed the effects of placebo control conditions, such as health education or stretching.

Rationale

Since 1995, at least 26 RCTs, including nearly 3,000 people, have examined the effects of exercise on feelings of distress or well-being in healthy adults or adults with medical conditions other than psychiatric disorders or disabling conditions that severely limit physical activity (i.e., spinal cord injury, multiple sclerosis, and stroke or severe head trauma). The average effect of exercise compared to a control condition was 0.27 SD (95% CI = 0.16 to 0.38). The outcomes were favorable after exercise in nearly 80 percent of the comparisons (26 of 33) with control conditions, but only 13 of 33 comparisons reached statistical significance. The average sample size was about 60 people, but a fourth of the studies had less than 45, a much smaller number than the 200 or so people needed to detect an effect as small as one-third SD. When compared to a placebo (usually stretching or health education), the effect of exercise was reduced to 0.10 SD (95% CI: −0.12 to 0.32) and was significant in just 2 of 9 comparisons.

Figure G8.5 Feelings of Distress/Well-Being: Prospective Cohort Studies, 1995 Through 2007: Crude and Adjusted Odds

Figure G8.5 Data Points

Do the Effects of Physical Activity on Distress or Well-Being Differ According to Race/Ethnicity, Sex, Age, or Medical Condition?

Conclusions

The available evidence supports that regular physical activity is associated with reduced feelings of distress and enhanced feelings of well-being regardless of age, sex, race/ethnicity, or medical condition. However, whether these factors modify those associations has not been studied. Race and ethnicity have been poorly represented or not described in most studies.

Rationale

Cross-sectional samples of Americans in the National Health Interview Survey (92), the Behavioral Risk Factor Surveillance Survey (93;95), and the National Physical Activity and Weight Loss Survey (91) included African American, Hispanics, and small numbers of Asian Americans and American Indians, but results were not compared between groups. Most of the prospective cohort studies were conducted in Europe or Australia and did not describe minority representation of the sample. The US Nurses’ Health Study includes minority women, but their representation was not described in the study of physical activity and well-being in the cohort (94).

Similarly, it is unclear whether sex and age modify the protective effect of physical activity on distress/well-being. Only 4 cohort studies reported findings separately for males and females, with mixed results (96-99), and none compared age groups. Across cohort studies, findings were similar for men and women. However, age was inversely related to reduced odds of distress or increased odds of enhanced well-being regardless of sex. After adjustment for other risk factors, odds were 0.65 (95% CI = 0.61 to 0.72) for men and women younger than age 55 years and 0.90 (95% CI = 0.86 to 0.97) for men and women aged 55 years and older. In RCTs, the average effect of exercise compared to a control condition is similar in both healthy adults (0.28 SD, 95% CI = 0.16 to 0.41) and people with non-psychiatric medical conditions (0.26, 95% CI: 0.10 to 0.41). Figure G8.6 illustrates effect sizes and 95% confidence intervals feelings of distress or well-being in studies comparing exercise with a control condition from the RCTs of adults with or without medical conditions.

These trials reported similar effects of exercise in both men and women and in groups varying in mean aged 31 to 74 years. However, the studies were not designed to compare groups according sex, race/ethnicity, or age. Only 1 of the trials compared men and women (45), and results were mixed across the 3 studies of men (0.49 SD, 95% CI = −0.11 to 1.08). Ten studies reported results for women (0.21 SD, 95% CI = 0.02 to 0.40), and 15 studies combined results for similar proportions of men and women (0.26 SD, 95% CI = 0.12 to 0.41). None compared effects between people of different ages or race/ethnicities. Seven of 16 studies of Americans described the racial/ethnic composition of their samples, but only 2 studies included a substantial portion of African Americans (46;100). Three studies included small numbers of Hispanics and/or Asian Americans (51;78;100). The small number of studies precluded meaningful comparisons among the medical conditions.

Figure G8.6 Feelings of Distress/Well-Being: Randomized Controlled Trials 1995 Through 2007

Figure G8.6 Data Points

Do the Effects of Physical Activity Vary According to Features of Physical Activity Including Type, Intensity, or Timing (i.e., Session Duration, Weekly Frequency, and Length of Participation)?

Conclusions

Population-based studies indicate that participation in either moderate or high levels of physical activity is associated with reduced feelings of distress or enhanced well-being, when compared with inactivity or very low physical activity exposure. The minimal or optimal type or amount of exercise for reducing feelings of distress or enhancing feelings of well-being are not yet known, but it appears that an increase in physical fitness is not required.

Rationale

Seven prospective cohort studies of feelings of distress or well-being included 3 or more levels of physical activity, using inactivity or low activity as the reference group (18;55;56;94;96;97;101). Independently of age and sex, and with or without adjustment for other risk factors, a linear reduction in odds of about 10% occurred for each level of physical activity compared to people who were inactive or had very low activity. Similarly, odds favored people who met or exceeded recommendations for moderate or vigorous participation by about 5% compared to active people who did not meet either recommended level. No studies permitted a direct comparison of moderate and vigorous recommended levels, but unadjusted odds were lower for participation in moderate-to-vigorous physical activity (OR = 0.77, 95% CI = 0.70 to 0.84) compared to participation at levels less than recommended (OR = 0.84, 95% CI = 0.78 to 0.91). Figure G8.7 shows the odds ratios and 95% CI in prospective cohort studies that examined the dose-response association between levels of physical activity and feelings of distress or well-being.

A few RCTs manipulated type (walking versus aquatic exercise or Qigong) or timing (i.e., intermittent versus continuous) of exercise to examine whether those feature modify the effects of exercise on feelings of distress or well-being (60-62;64), but the studies did not include a control group who did not exercise. About 45 percent of the controlled studies used an aerobic exercise intervention such as walking, jogging, cycling, or aquatic exercise, and another 30 percent of the studies combined aerobic activity with resistance exercise. Only 3 studies used resistance exercise alone, and none compared aerobic with resistance exercise. About half the studies used continuous exercise, and a third used intermittent exercise. One study of home-based aerobic dance reported positive effects when daily exercise was a continuous 30-minute session but not when it was 2 sessions of 15 minutes each separated by 4 hours (80). Timing of the exercise could not be determined in the other studies. Effects of exercise did not differ according to the mode or timing of exercise.

Figure G8.7 Feelings of Distress/Well-Being: Prospective Cohort Studies 1995 Through 2007: Dose Response

Figure G8.7 Data Points

Three-fourths of the studies used a moderate-to-vigorous exercise intensity of 60% to 80% of people’s aerobic capacity or maximum strength. The average session lasted 45 minutes and occurred 3 days per week, but studies had a wide range of session duration (10 to 85 minutes) and weekly frequency (1 to 7 days). Nonetheless, these features of exercise did not modify effects independently of each other. Studies lasted an average of 6 months, and a fourth of the studies lasted less than 3 months. Length of the exercise program was also unrelated to symptom outcome.

Eleven of 13 studies that measured cardiorespiratory fitness reported a significant increase that averaged 1 SD, or about 20%. Strength was significantly increased in each of 3 studies that measured strength. However, there was no association between fitness increases and changes in feelings of distress or well-being.

Question 4: Is There an Association Between Physical Activity and Cognitive Function and Dementia?

Introduction

Cognition can be conceptualized as processes involved in selecting, manipulating, and storing information derived from experiences and how these processes guide behavior. Cognitive abilities are functional properties of the individual that are not directly observed but are inferred from behavior. Researchers in the disciplines of psychometrics, cognitive psychology, and neuropsychology have developed more than 400 tests designed to assess specific types of mental processing (102). They range from those designed to evaluate specific processes (e.g., working memory, information-processing speed, inhibition) to those that assess global mental functioning involving multiple processes. Assessment methods include those designed specifically to evaluate the effects of injury and degenerative disease on cognitive function and those designed to evaluate individual differences in healthy individuals. The multidimensionality of cognitive function and the diversity of assessment methods present a special challenge for the interpretation of the evidence about the effect of physical activity and exercise on cognitive function.

Does Physical Activity Protect Against the Onset of Age-Related Decline in Cognitive Function or Dementia?

Conclusions

The weight of the available evidence from prospective cohort studies supports the conclusion that physical activity delays the incidence of dementia and the onset cognitive decline associated with aging.

Rationale: Prospective Cohort Studies

At least 17 prospective population-based cohort studies have been published since 1995 that have assessed the association of individuals’ level of physical activity with the onset of age-related decline in cognitive functioning among healthy adults or with incident cases of dementia. Studies that confounded the measure of physical activity with other leisure or mental activities were not considered. Four studies showed protective effects against cognitive decline in healthy aging adults (103-106) and two studies did not find protective effects (107;108).

Of the 11 studies of dementia, 7 reported a protective effect of physical activity. Nine of 16 comparisons were statistically significant (mean OR = 0.63 95% CI = 0.50 to 0.80). Though limited in number, results were stronger for Alzheimer’s disease than for other dementias, including vascular dementia. Figure G8.8 describes results from prospective cohort studies of physical activity and incident dementia or Alzheimer’s disease.

The studies varied considerably in sample size, methods used to assess participant’s physical activity level and mental function, and the duration between baseline and follow-up measurements. In general, most studies with sample sizes greater than 1,000 individuals report that physical activity delays the onset of cognitive decline or dementia. The results of studies conducted with smaller numbers of participants are inconsistent. The lack of agreement among the studies reviewed may be explained, at least in part, in terms of statistical power. An alternative explanation for the inconsistencies among the results of these studies is the confounding influence of cognitive stimulation derived while engaged in physical activities (108). In studies that report a relation between physical activity and delayed onset of dementia, the effects were detected using both clinical assessments and standardized measures of cognitive function. Further, early-life (103), mid-life (109), and current levels of physical activity all appear to postpone symptoms of dementia. Thus, although exercise does not prevent dementia, it may be associated with a delay in its onset, perhaps by maintaining a higher level of cognitive function for physically active adults than less physically active individuals as they age (110).

Rationale: Randomized Controlled Trials

A meta-analysis of 18 RCTs (111) indicated that aerobic exercise training produced an effect size (SD) of 0.48 for improving performance on all cognitive tasks, with the greatest effect size (0.68) for executive processing tasks that measure goal-oriented decision-making behavior, compared to 0.46 for controlled-processing tasks that assess attentional effort, 0.42 for visuospatial tasks that evaluate the perceptual organization, and 0.27 for speeded tasks that focus on rapid responses and movements. The results of 2 recently conducted experiments provide support for the facilitative effects of resistance exercise training on information-processing functions of cognitively healthy older adults (85;112), though one experiment found no significant effects of either yoga training or a walking program on older adults’ memory or information-processing speed (113).

Figure G8.8 Incident Total Dementia or Alzheimer’s Disease: Prospective Cohort Studies, 1995 Through 2007: Crude and Adjusted Hazard

Figure G8.8 Data Points

Does Physical Activity Reduce Symptoms Associated With Alzheimer’s Disease or Other Dementias?

Conclusions

Evidence from RCTs of healthy older adults and people with Alzheimer’s disease or other dementias support that regular participation in physical activity improves aspects of cognitive function or reduces symptoms of dementia.

Rationale

A meta-analytic review of 10 randomize