Research Article

Published: 2024-09-13
https://doi.org/10.20935/MHealthWellB7335

Physical activity and mortality in the general population

with and without major depression
Vagner Deuel de O. Tavares1, Jeanne V. A. Williams2, Vandad Sharifi3, Andrew Bulloch2, Gina Dimitropoulos2,
Nicole Leite Galvão-Coelho1, Scott Patten2,*

Academic Editor: Jair de Jesus Mari

Abstract
Background and aims: It is not yet well understood whether physical activity (PA) may protect against elevated mortality in major
depression. Answering this question has implications for policy and practice. The aim of this study was to estimate the association of
PA with mortality in major depressive disorder (MDD). Methods: We used data from the Canadian Community Health Survey—Mental
Health and Well-Being (CCHS 2002), conducted in 2002 (n = 31,200), and the Canadian Community Health Survey—Mental Health
(CCHS-MH 2012), conducted in 2012 (n = 20,935), for this analysis. Recreational PA was measured using metabolic equivalent of task
(MET) values, classifying participants as active, moderately active, and physically inactive in the 2002 survey and as very active, active,
moderately active, and physically inactive in the 2012 survey. Major depression was assessed using a fully structured diagnostic
instrument, the Composite International Diagnostic Interview. Mortality data was obtained through data linkage. The analysis used
Cox proportional hazard models to explore the associations. Results: PA in the 2002 survey was protective: HR 0.69 (95% CI 0.64–
0.74) and HR 0.77 (95% CI 0.72–0.83) for the active and moderately active groups, respectively. In the 2012 survey, the HRs were 0.55
(95% CI 0.43–0.70), 0.57 (95% CI 0.43–0.75), and 0.59 (95% CI 0.46–0.76) for the very active, active, and moderately active groups,
respectively. No interactions were significant. Conclusions: These results indicate that PA protects against elevated mortality to the
same extent in those with MDD as in those without. These findings, which arise from a representative sample, support continued efforts
to integrate PA into the management of depressive disorders at the population level.

Keywords: depression, mortality, physical activity, protection

Citation: Tavares VDO, Williams JVA, Sharifi V, Bulloch A, Dimitropoulos G, Galvão-Coelho NL, Patten S. Physical activity and
mortality in the general population with and without major depression. Academia Mental Health and Well-Being 2024;1.
https://doi.org/10.20935/MHealthWellB7335

1. Introduction
Major depressive disorder (MDD) affects more than 300 million An unhealthy lifestyle behavior, including physical inactivity,
people worldwide, inducing a significant impact on social which is a common feature in depressed patients [16], may also
relationships and individual functionality. It is one of the leading contribute to elevated mortality [17–19]. On the other hand,
causes of disability, incapacity, and elevated health-care cost [1, physical activity (PA) improves physical health and may be
2]. Moreover, many studies have reported elevated mortality protective against all-cause mortality [18, 20, 21].
rates in people experiencing mental disorders, including MDD
Recent epidemiological studies have also emphasized the
[3–5].
protective or beneficial effects of PA on incident depression [22,
It has been shown that patients with MDD have alterations in 23]. Pearce et al. [23] conducted a meta-analysis of the relation-
stress systems, including hypothalamic–pituitary–adrenal (HPA) ship between levels of PA and incident depression, concluding
axis function and increased levels of inflammatory biomarkers that 11% of depression cases could have been avoided if the
(CRP, IL-1, TNF-α, and IL-6) [6–9]. These factors may vary population had been active at the level of current health recom-
according to severity, duration of the disease, age, physical health mendations. Another meta-analysis showed that the protective
conditions, and social factors [10, 11]. These changes are also effects of PA against depression are observed regardless of age
associated with mortality [12–15]. and sex and are significant across all geographical regions [24].

1Laboratory of Hormone Measurement, Department of Physiology, Biosciences Center, Federal University of Rio Grande do Norte,
Natal, Rio Grande do Norte 59078-970, Brazil.
2Mathison Centre for Mental Health Research & Education, Department of Psychiatry, Cumming School of Medicine, University of

Calgary, Calgary, AB T2N-4Z6, Canada.

3Department of Psychiatry, Roozbeh Hospital, Tehran University of Medical Sciences, Tehran, Tehran Province 13337-95914, Iran.

*email: [email protected]

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 1 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

Possible arguments for the protective effects of PA in MDD Canadian household residents aged 15 years and over living in
include its potential to improve social behavior and well-being any of the 10 provinces. Excluded were people living on “Indian”
[25, 26]; contribute to better regulation of the HPA axis [27]; reserves and other aboriginal settlements, full-time members of
decrease inflammation [27, 28]; contribute to the management the Canadian Armed Forces, and residents of institutions. For the
of body mass index (BMI), including body fat loss and muscle CCHS 2002, the information was collected between May 2002
mass gain [29]; and induce positive neuroplastic changes in brain and December 2002, with a sample size of approximately (n =
structures related to mood [30]. 31,200). For the CCHS 2012, data collection occurred between
January 2, 2012, and December 31, 2012. A total of (n = 20,935).
While studies have established the protective effect of PA against
Please note that these numbers are rounded to the nearest five in
depression, it is not known whether PA protects against elevated
keeping with Statistics Canada’s data-release requirements for
mortality in MDD. If so, this would provide an additional
count data. Such rounding was not required for other types of
motivation for the implementation of practices and policies to
estimates, such as frequencies and hazard ratios (HAs). In-
increase PA in patients with depression. Such practices and
terviews were conducted using computer-assisted personal
policies would likely need to account for features of depression,
interviewing by trained and experienced interviewers. Statistics
such as diminished motivation and enjoyment of activities.
Canada calculates sampling weights that help ensure valid
Only two cohort studies have investigated the relationship inference to the target (household) population, including adjust-
between PA and mortality in depression [31, 32]. Perez-Lasierra ments for nonresponse. Exclusions from the sampling frame
et al. evaluated mortality in National Health and Nutrition were similar to those in the 2002 survey: individuals living on
Examination Survey (NHANES) participants scoring 10 or above “Indian” reserves and crown lands, health-care institution
on the Patient Health Questionnaire – Brief (PHQ-9), a depres- residents, full-time members of the Canadian Armed Forces, and
sive symptom rating scale. However, they were unable to analyze residents of certain remote regions. In total, these exclusions
the prevalence of MDD, as the PHQ-9 is a screening instrument amount to less than 3% of the population. Analyses of CCHS data
rather than a diagnostic measure. Furthermore, they measured do not require Ethics Review Board approval. Additional
only two levels of PA and focused on adults aged 50 and older, information on these surveys is available from Statistics Canada
whereas major depression is common in all age groups. The study (https://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurv
by Liu and Jia [31] showed similar limitations, using a self- ey&Id=119789).
reported clinical diagnosis and PHQ-2 data from the UK
Biobank. This study also covered a limited age range (37–73 2.2. Physical activity measures
years). They found a significant risk-elevating effect of lifetime The PA module of the CCHS 2002 included items to assess
MDD and a protective effect of PA. Multiplicative interactions recreational PA in the three months preceding the survey inter-
between MDD and PA were not observed in their Cox propor- view. The activities assessed by these items were “walking for
tional hazards model [31], but greater-than-additive interactions exercise”, “gardening or yard work”, “swimming”, “bicycling”,
were observed. “popular or social dance”, “home exercises”, “ice hockey”, “ice
For a better understanding of the effects of PA on mortality in skating”, “in-line skating or rollerblading”, “jogging or running”,
major depression episode, it is necessary to have strong measures, “golfing”, “exercise class or aerobics”, “downhill skiing or snow-
a representative sample, and the ability to control for possible boarding”, “bowling”, “baseball or softball”, “tennis”, “weight
confounding variables. A particularly important confounder is training”, “fishing”, “volleyball”, “basketball”, and “other activi-
age, as MDD often has its onset in late adolescence and early ties.” Each activity is associated with an estimate of metabolic
young adulthood [33], whereas mortality is strongly influenced energy expenditure per hour of participation. While some of the
by advancing age. listed activities do not represent vigorous PA, their associated low
metabolic energy expenditures result in these activities making a
In this study, we used historical cross-sectional survey data that small but accurate contribution to overall estimates of PA. The
included administration of a validated, fully structured psychiat- CCHS respondents are classified as active, moderately active, or
ric diagnostic interview and an assessment of PA in a representa- inactive based on a PA index that estimates average daily PA over
tive sample. These surveys were conducted in 2002 and 2012 and the past three months. The index is calculated as the sum of the
were later linked to a national mortality database. We hypothe- average daily energy expenditure of all activities. Respondents are
sized that PA would reduce the effect of major depression on classified as follows: 3.0 kcal/kg/day or more = physically active;
elevated all-cause mortality. 1.5–2.9 kcal/kg/day = moderately active; less than 1.5 kcal/day =
inactive [34]. We consider this classification to be consistent with
current recommendations for 30–45 min of activity three to five
2. Materials and methods times a week [35–38]. Activities that lead to an increase in daily
2.1. Data source: Canadian Community Health Survey energy expenditure of approximately 200–300 kcal/session can
2002 and Canadian Community Health Survey 2012 add up to 1,000–1,500 kcal/week (e.g., 3 kcal/kg/day in a 70 kg
The Canadian Community Health Survey—Mental Health and person leads to an expenditure of approximately 210 kcal/day).
Well-Being (CCHS 2002) and the Canadian Community Health Physical activity with a frequency of five times a week can lead to a
Survey—Mental Health (CCHS 2012) were cross-sectional men- caloric expenditure equal to or greater than 1,050 kcal. This level
tal health surveys based on nationally representative samples of caloric expenditure is associated with substantial health
conducted by Canada’s national statistical agency, Statistics benefits. Therefore, we decided to classify as “physically active”
Canada. These surveys collected data on demographics, health those who achieved more than 150 min/week, as well as those who
status, health determinants, and health-care utilization, with a achieved 30 min or more per day for more than three days per
strong focus on mental health. The target population was week. Based on these assumptions, we classified as “moderately
active” those who have less than 150 min/week, expending

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 2 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

between 106 and 200 kcal/day. Caloric expenditures below 105 2.5.3. Type of drinker
kcal were considered inactive, e.g., walking no more than 15 min Type of drinker was assessed using field-tested items about
everyday [39]. drinking habits in both surveys, leading to the classification of
respondents as regular, occasional, former, or never drinkers in
The PA module of the CCHS-MH 2012 also included items to
2012. In the CCHS-MH 2012, each respondent was classified as
assess recreational PA, indicating the average amount of time in
consuming alcohol on a regular, occasional, or never basis.
hours that respondents reported doing moderate or vigorous PA
in the past seven days. Response categories (reported in ranges)
2.5.4. Chronic conditions
were converted to proportions of minutes, using average dura-
Both surveys asked respondents whether or not they had any
tion in hour units. We classified participants into four categories:
long-term (>6 months) physical or mental health condition that
very active physically (>300 min/week), active physically (150–
had been diagnosed by a health professional. The chronic
300 min/week), moderately active (<150 min/week), and inac-
conditions included in this analysis were asthma, arthritis, back
tive physically (<15 min/week).
problems, high blood pressure, migraine, bronchitis, lung em-
physema, diabetes, epilepsy, heart disease, cancer, stomach
2.3. Major depressive disorder
problems (e.g., ulcer), stroke, chronic disease, Alzheimer’s dis-
The CCHS 2002 identified MDD in the past 12 months using a ease, cataracts, glaucoma, and thyroid conditions.
Canadian adaptation of the World Mental Health—Composite
International Diagnostic Interview Instrument (WMH-CIDI) 2.5.5. Mortality
[40]. The WMH-CIDI is a lay-administered psychiatric interview Data from the CCHS surveys was linked by Statistics Canada to
based on Diagnostic and Statistical Manual of Mental Disorders, the Canadian Vital Statistics Deaths database. Linkage proce-
4th Edition (DSM-IV). The CCHS-MH 2012 used the same dures have been previously described and evaluated [42]. The
interview. The experienced lay interviewers employed in the linkage data from the CCHS 2002 (n = 5,635) was constructed as
surveys were fully trained in administration of the WMH-CIDI. of June 2002, providing up to 15 years of follow-up time. The
After administering the interview, a computerized algorithm was linkage data from the CCHS-MH 2012 (n = 1,175) is similarly
used to classify respondents as: (a) meeting the criteria for a available, providing up to five years of follow-up time. For
lifetime major depressive episode (one or more episodes of major respondents identified in the mortality database, the date of
depression in their past life) and (b) meeting criteria for 12- death was recorded, and time to death was calculated as the
month prevalence of MDD, based on having a major depressive number of days between their survey interview date and the date
episode in the 12 months prior to the interview. Due to concerns of their death. Respondents who did not die were censored at the
about the validity of lifetime prevalence ratings [41], past record linkage date.
12-month episodes were used in this analysis.
2.6. Statistical analysis
2.4. Demographic data (covariates)
Cox proportional hazard models were used in the analysis. We
Demographic data collected by both surveys included age, sex, operationalized our hypothesis that PA would reduce the impact
marital status, and education. These variables were assessed of major depression on all-cause mortality using interaction
using field-tested items developed by Statistics Canada. For terms. The significance of the major depression by PA interaction
marital status, we used four categories: married or common law, term, assessed using a Wald test, served as a statistical test of
widowed, separated or divorced, and single (never married). whether PA modified the effect of major depression. Since these
Married or common law was used as the referent group in the interactions were examined using a Cox proportional hazard
analysis. Education referred to the highest level of education model (HR), which is a log-transformed model, the interactions
achieved by the respondent and was categorized into less than are assessed on the multiplicative scale.
secondary school, secondary school graduate, and postsecond-
ary, with postsecondary serving as the referent category. Categorical variables were estimated as frequencies, using per-
centages (%) in descriptive tables. Continuous data is reported as
2.5. Health outcomes (covariates) means with 95% confidence intervals (CI). Covariates selected for
this analysis, as listed above, have been previously reported to be
2.5.1. Body mass index determinants of mortality and may therefore act as confounders
The BMI is a quick and accurate method to determine health risk of the association between MDD and PA. To explore the relation-
as it relates to body weight and height. BMI was measured in the ship between PA and major depression on subsequent mortality,
2002 and 2012 surveys, but in the 2002 survey, age restrictions we initially examined interaction terms between these variables.
resulted in the BMI being calculated only for a minority of the Next, the age- and sex-adjusted main effects were estimated. In
sample (n < 10,000). Exploratory analyses determined that BMI the survey dataset, we adjusted for other covariates one at a time
did not act as an effect modifying or confounding variable, so it is and in a multivariable model to assess their potential confound-
not reported in the analyses that follow. ing and/or joint confounding role. All analyses were conducted
in the Prairie Regional Data Centre at the University of Calgary
2.5.2. Smoking Campus. All analyses reported, except the relative excess risk due
A potentially important confounding variable is smoking, which to interaction (RERI) estimates, used a recommended bootstrap
was not directly assessed in the CCHS 2002. However, the survey weighting procedure incorporating a set of 1,000 replicate
interview included an item asking whether respondents smoked bootstrap weights, ensuring accurate estimates of parameters,
more when they were stressed. We classified all respondents who standard errors, and 95% CI while accounting for design effects
answered affirmatively to this item as smokers. The CCHS-MH (clustering and unequal selection probabilities). Tables 1 and 2
2012 included a survey module to classify smoking status as present all of the variables included in the analysis and specify
never, former, or current.

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 3 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

the choice of reference group for each variable. We also expected to manifest as a greater-than-additive interaction,
calculated the interaction on an additive scale to measure RERI, quantified by the RERI, because the causal mechanisms that
assuming that the Cox proportional hazard ratio minus one require both exposures mean that the joint exposure is associated
approximated relative excess risks. The RERI allows assessment with a higher risk than the sum of the individual exposures.
of greater-than-additive interactions between two exposures. Statistical analysis was performed in Stata v18 (StataCorp 2019),
This parameter is used as an index for causal inference in epi- and statistical significance was assigned using p < 0.05.
demiology. If two exposures interact on a biological level, this is

Table 1 • Description of the CCHS 2002 sample

Overall No major depression Major depression
% (95% CI) % (95% CI) % (95% CI)
n = 31,120a n = 29,350a n = 1,625a
Depression 95.3 (94.9–95.6) 4.7 (4.4–5.1)

Sex

Male 50.8 (50.5–51.2) 96.4 (95.9–96.8) 3.6 (3.1–4.0)

Female 49.2 (48.8–49.5) 94.2 (93.6–94.7) 5.8 (5.2–6.3)

Age

15–29 23.8 (23.1–24.5) 94.0 (93.3–94.8) 6.0 (5.2–6.7)

30–39 19.3 (18.6–19.9) 94.6 (93.9–95.4) 5.4 (4.6–6.1)

40–49 20.9 (20.1–21.6) 94.6 (93.6–95.6) 5.4 (4.4–6.4)

50–59 15.4 (14.8–16.1) 95.7 (94.9–96.4) 4.3 (3.6–5.1)

60 and above 20.6 (20.0–21.2) 97.6 (97.1–98.2) 2.4 (1.8–2.9)

Marital status

Married 62.0 (61.3–62.8) 96.7 (96.3–97.1) 3.3 (2.9–3.7)

Divorced 12.8 (12.4–13.3) 92.0 (91.0–93.0) 8.0 (7.0–9.0)

Single 25.1 (24.4–25.8) 93.4 (92.6–94.2) 6.6 (5.8–7.4)

Education

Less secondary 25.7 (25.0–26.5) 95.1 (94.4–95.7) 4.9 (4.3–5.5)

Second school 18.8 (18.1–19.5) 95.2 (94.5–95.9) 4.8 (4.1–5.5)

Some postsecondary 8.7 (8.2–9.1) 94.3 (93.2–95.5) 5.7 (4.5–6.8)

Postsecondary 46.8 (46.0–47.6) 95.6 (95.0–96.1) 4.4 (3.9–5.0)

Chronic condition

None 29.2 (28.4–30.0) 97.5 (97.1–98.0) 2.5 (2.0–2.9)

One or more 70.8 (70.0–71.6) 94.3 (93.9–94.8) 5.7 (5.2–6.1)

Type of drinker

Regular 58.1 (57.3–58.9) 95.5 (95.0–95.9) 4.5 (4.1–5.0)

Occasional 19.3 (18.7–19.9) 94.8 (94.0–95.6) 5.2 (4.4–6.0)

Former 13.0 (12.5–13.6) 94.0 (93.1–94.9) 6.0 (5.1–6.9)

Never 9.6 (9.1–10.1) 96.6 (95.5–97.7) 3.4 (2.3–4.5)

Physical activity

Inactive 47.1 (46.2–47.9) 94.8 (94.3–95.4) 5.2 (4.6–5.7)

Moderate 26.8 (26.0–27.5) 95.5 (94.9–96.2) 4.5 (3.8–5.1)

Active 26.2 (25.4–26.9) 95.8 (95.2–96.3) 4.2 (3.7–4.8)

Smoker

No 50.8 (50.5–51.2) 96.6 (96.2–96.9) 3.4 (3.1–3.8)

Yes 49.2 (48.8–49.5) 92.3 (91.5–93.1) 7.7 (6.9–8.5)

Mortality within 17 years of follow-up – 13.0% 10.1%

aThese counts are rounded to the nearest five in keeping with Statistics Canada data-release guidelines.

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 4 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

Table 2 • Description of the CCHS-MH 2012 sample

Overall Major depression (no) (n = 19,705a) Major depression (yes)
(n = 20,935) (n = 1,125 a)
Depression 95.3 (94.9–95.8) 4.7 (4.2–5.1)

Sex

Male 50.7 (49.5–51.9) 96.5 (95.8–97.2) 3.5 (3.0–4.1)

Female 49.3 (48.1–50.5) 94.2 (93.5–94.9) 5.8 (5.2–6.3)

Age (years)

15–29 23.9 (22.9–24.9) 94.0 (93.0–95.0) 6.0 (5.0–7.0)

30–39 16.2 (15.3–17.0) 94.5 (93.0–96.0) 5.5 (4.0–7.0)

40–49 17.0 (15.9–18.0) 93.2 (91.8–94.6) 6.8 (5.4–8.2)

50–59 18.4 (17.4–19.5) 96.2 (95.5–96.9) 3.8 (3.1–4.5)

60 and above 24.6 (23.7–25.5) 97.9 (97.4–98.4) 2.1 (1.6–2.6)

Marital status

Married 60.9 (59.7–62.0) 96.6 (96.1–97.1) 3.4 (2.9–3.9)

Divorced 12.5 (11.8–13.2) 93.2 (91.5–95.0) 6.8 (5.0–8.5)

Single 26.6 (25.6–27.7) 93.4 (92.5–94.3) 6.6 (5.7–7.5)

Education

Less secondary 7.1 (6.7–7.6) 94.0 (92.2–95.8) 6.0 (4.2–7.8)

Second school 10.3 (9.7–11.0) 95.2 (93.9–96.4) 4.8 (3.6–6.1)

Some postsecondary 4.5 (3.9–5.1) 91.5 (88.5–94.5) 8.5 (5.5–11.5)

Postsecondary 78.0 (77.1–79.0) 95.7 (95.1–96.2) 4.3 (3.9–4.8)

Chronic condition

None 26.7 (25.5–27.8) 97.5 (96.8–98.1) 2.5 (1.9–3.2)

One or more 73.3 (72.2–74.5) 94.5 (94.0–95.1) 5.5 (4.9–6.0)

Type of drinker

Regular 59.3 (58.1–60.5) 95.8 (95.3–96.4) 4.2 (3.6–4.7)

Occasional 18.8 (17.8–19.8) 93.5 (92.4–94.6) 6.5 (5.4–7.6)

Never 21.9 (20.9–22.9) 95.7 (94.8–96.6) 4.3 (3.4–5.2)

Physical activity

Very active 27.9 (26.8–29.1) 95.8 (95.0–96.6) 4.2 (3.4–5.0)

Active 17.6 (16.7–18.4) 96.5 (95.6–97.3) 3.5 (2.7–4.4)

Moderate active 27.9 (26.8–28.9) 94.6 (93.5–95.6) 5.4 (4.4–6.5)

Inactive 26.7 (25.6–27.7) 94.8 (94.0–95.6) 5.2 (4.4–6.0)

Smoker

Nonsmoker 40.0 (38.8–41.2) 96.6 (95.9–97.3) 3.4 (2.7–4.1)

Current smoker 20.8 (19.8–21.8) 92.1 (90.9–93.2) 7.9 (6.8–9.1)

Former smoker 39.2 (38.1–40.4) 95.8 (95.1–96.4) 4.2 (3.6–4.9)

Mortality within 17 years of follow-up 4.1% 3.5%

aThese counts are rounded to the nearest five in keeping with Statistics Canada data-release guidelines.

3. Results higher among those who were inactive (5.2%) than those who
were moderately active (4.5%), and physically active (4.2%). In
3.1. Canadian Community Health Survey—Mental addition, past-year major depressive episodes were more
Health and Well-Being (2002) prevalent in females (5.8%), younger individuals (15–29 years,
The overall sample (n = 31,200) showed a prevalence proportion 6.0%), divorced people (8.0%), participants with some postsec-
for past-year major depression of 4.7%. The prevalence was ondary education but less than a university or college degree

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 5 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

(5.7%), underweight individuals (8.9%), former alcohol users Occasional 0.91 -1.86 0.063 0.82–1.00
(6.0%), and smokers (7.7%). More details are presented in Table 1.
Former 1.10 2.06 0.039 1.00-1.20
The frequency of deaths during follow-up (17 years) among those
who were depressed was 13.0%, compared to 10.1% among those Never – – – –
without major depression. Smoker
Initially, we estimated the HR for depression, adjusting for age and Yes 1.77 17.72 <0.001 1.66–1.89
sex (HR = 1.74, p = 0.0001). When PA levels were added to the
No (ref) – – – –
model along with their interaction terms, these interactions were
not significant, indicating a lack of multiplicative interaction Physical activity
(p > 0.05). With the removal of the interaction terms, the HRs for Moderate 0.77 -7.38 <0.001 0.72–0.83
PA indicated protective effects of being active (HR = 0.64,
p < 0.0001) and being moderately active (HR = 0.71, p = 0.0001). Active 0.69 -9.61 <0.001 0.64–0.74

In this model, the association of major depression with elevated Inactive (ref) – – – –
mortality weakened slightly (HR = 1.66, p < 0.0001), as shown in aAge is treated as a continuous variable in these models.
Table 3.
Note: “ref” refers to the control group used as the reference.

Table 3 • Association of physical activity with major depres-

After analyzing age- and sex-adjusted hazard ratio models, we
sion-associated mortality: Cox proportional hazard models
examined a model that included adjustments for all variables
based on the linkage of the CCHS 2002 to vital statistics data
(sex, age, marital status, education, chronic condition, type of
CCHS 2002 drinker, and smoking frequency). Elevated mortality in MDD
HR Z p Value LCI–UCI continued to be observed (HR = 1.36, p < 0.0001), as shown in
Table 3. Thus, being physically active was protective (HR = 0.69,
Model 1
p < 0.0001), as was being moderately active (HR = 0.77,
Major depression 1.74 7.65 <0.001 1.51–2.00 p < 0.0001), with adjustment for the set of covariates available
Male sex 1.53 15.71 <0.001 1.45–1.61 in this study.

Agea 1.10 92.64 <0.001 1.10–1.10 3.2. Canadian Community Health Survey—Mental
Model 2 Health (2012)
Major depression 1.66 6.98 <0.001 1.44–1.91 There was a past-year prevalence of MDD of 4.7%, similar to that
observed in the 2002 survey. The prevalence of depression varied
Male sex 1.61 17.37 <0.001 1.52–1.70
across different demographics in a manner consistent with the
Agea 1.10 91.75 <0.001 1.10–1.10 2002 data. Notably, those who engaged in moderate PA reported
Inactive (ref) – – – –
a prevalence of 5.4%, while inactive individuals showed a slightly
lower prevalence of 5.2%. Individuals who were highly active
Moderate 0.64 -11.9 <0.001 0.59–0.69 physically demonstrated a prevalence of 4.2%, and those who
Active 0.71 -9.99 <0.001 0.66–0.76 were moderately active exhibited a prevalence of 3.5%. Further-
more, the data indicates that depression was more pronounced
Model 3
among specific subgroups. Females exhibited a higher prevalence
Major depression 1.36 4.22 <0.001 1.18–1.58 at 5.8%, and individuals within the middle age group (40–49
Male sex 1.77 19.08 <0.001 1.67–1.88
years) reported a prevalence of 6.8%. Those who were divorced
also showed a 6.8% prevalence of depression, while individuals
Agea 1.10 79.93 <0.001 1.09–1.10 with some postsecondary education but no university or college
Marital status degree recorded a prevalence of 8.5%. Other groups, such as
underweight individuals (6.5%), former alcohol users (6.0%),
Single 1.55 9.18 <0.001 1.41–1.71
and smokers (7.9%), also demonstrated a higher prevalence of
Divorced 1.30 8.44 <0.001 1.22–1.38 past-year major depression. For additional information, refer to
Married (ref) – – – – Table 2. The frequencies of deaths during follow-up (five years)
among those who were depressed were 4.1%, compared to 3.5%
Education among those without major depression.
Less secondary 1.16 4.60 <0.001 1.09–1.23
Initially, we estimated the HR for depression while adjusting for
Second school 1.08 1.81 0.071 0.99–1.18 age and sex (HR = 3.11, p < 0.0001). Subsequently, when we
incorporated PA levels into the model, along with the corre-
Some postsecondary 1.26 3.93 <0.001 1.12–1.42
sponding interaction term, the analysis revealed that these
Postsecondary (ref) – – – – interaction terms were not statistically significant, indicating a
Chronic condition lack of multiplicative interaction (p > 0.05). Upon removal of
these interaction terms, the HRs for PA emerged as significant.
One or more 1.45 9.22 <0.001 1.34–1.58
This updated model with PA levels exhibited a slight attenuation
Type of drinker in the association between major depression and elevated
Regular 0.86 -3.13 0002 0.78–0.94
mortality, resulting in an (HR = 2.96, p < 0.0001), as detailed in
Table 4. Specifically, being highly active physically showed a

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 6 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

protective effect (HR = 0.47, p < 0.0001), as did being physically Smoker
active (HR = 0.53, p < 0.0001) and moderately active (HR = 0.49,
Current 1.77 5.65 <0.001 1.75–3.20
p < 0.0001). Following the examination of hazard ratios adjusted
for age and sex, we proceeded to analyze a model adjusted for all Former 1.17 1.39 0.165 0.93–1.46
available covariates, similar to the CCHS 2002 (see Table 4). In Never (ref) – – – –
this more inclusive analysis, the association between major
Physical activity
depression and elevated mortality persisted (HR = 2.19,
p < 0.0001). Thus, being very active physically (HR = 0.55, Moderate 0.59 −4.13 <0.001 0.46–0.76
p < 0.0001), active (HR = 0.57, p < 0.0001), and moderately
Active 0.57 −3.86 <0.001 0.43–0.75
active (HR = 0.59, p < 0.0001) continued to be associated with
lower mortality. Very active 0.55 −4.65 <0.001 0.43–0.70

Inactive (ref) – – – –
Table 4 • Association of physical activity with major
aAge is treated as a continuous variable in these models.
depression-associated mortality: Cox proportional hazard
Note: “ref” refers to the control group used as the reference.
models based on the linkage of the CCHS-MH 2012 to vital
statistics data
The analysis of RERI was based on the age- and sex-adjusted
CCHS-MH 2012 models (Model 1 in Tables 3 and 4). The analysis yielded nega-
HR Z p Value LCI–UCI tive results. Specifically, for the CCHS 2002, we found
RERI = 0.45, p = 0.125, and for the CCHS-MH 2012, we found
Model 1
RERI = 0.61, p = 0.744.
Major depression 3.11 4.93 <0.001 1.98–4.88

Male sex 1.85 6.82 <0.001 1.55–2.22

4. Conclusions
Agea 1.10 27.03 <0.001 1.09–1.11
This is the first study to assess the protective effects of PA against
Model 2
elevated mortality in individuals with MDD, as evaluated through
Major depression 2.96 4.36 <0.001 1.81–4.82 a diagnostic interview across all ages (15+). The results from the
CCHS 2002 indicate that being physically active or moderately
Male sex 2.01 7.39 <0.001 1.67–2.42
active is associated with lower mortality, independent of MDD,
Agea 1.10 24.81 <0.001 1.09–1.11 with no interaction observed between PA and MDD on either the
Inactive (ref) – – – – multiplicative or additive scales. In the same context, regardless
of depression status, results from the CCHS-MH 2012 study show
Moderate 0.49 −5.72 <0.001 0.38–0.62
that being very active (45%), active (43%), or moderately active
Active 0.53 −3.73 <0.001 0.38–0.74 (41%) is associated with lower mortality. Following the current
recommendations for PA >150 min should be the minimum
Very active 0.47 −5.71 <0.001 0.37–0.61
target to achieve benefits [20]. However, we also highlight that
Model 3 even without meeting the full PA guidelines, there are still
Major depression 2.18 2.88 0.004 1.29–3.70 benefits in reducing mortality.

Male sex 2.30 8.01 <0.001 1.87–2.82 A robust body of evidence has shown that PA may decrease
mortality rates. Our study investigated the protective effect of
Agea 1.10 23.56 <0.001 1.09–1.11
being physically active, including three PA levels: very active,
Marital status active, and moderately active, on elevated mortality in MDD.
Single 1.71 2.83 0.005 1.18–2.50 Considering data from the CCHS 2002, our results indicate that
following current health recommendations or even being below
Divorced 1.28 2.49 0.013 1.05–1.56 recommendations but not entirely inactive may lower mortality
Married (ref) – – – – risk by 36% and 29% when controlled for sex and age, respec-
tively. PA continues to have a protective effect on mortality
Education
independent of MDD and the mortality determinants measured
Less secondary 0.99 −0.03 0.976 0.80–1.22 in this study, a statement that applies both to physically active
Second school 1.25 1.62 0.104 0.95–1.64 (31%) and moderately active (23%) participants.

Some postsecondary 1.10 0.31 0.759 0.58–2.10 This study adds to the literature by demonstrating substantial
benefits in a representative sample drawn from the general
Postsecondary (ref) – – – –
population. The 5–10% of the population that experience major
Chronic condition depressive episodes can benefit from PA, which is a substantial
Have one or more 1.96 4.48 <0.001 1.46–2.63
proportion of the population. Although the RERI results do not
provide statistical confirmation of synergy, this does not negate
Type of drinker the importance of the protective effect of PA on mortality.
Regular 0.59 −4.65 <0.001 0.47–0.73
The protective effects reported here should be considered
Occasional 0.80 −1.86 0.062 0.63–1.01 conservative estimates, since some of the effects of MDD on
mortality may be mediated by other determinants. For example,
None or former – – – –
MDD increases the incidence of chronic diseases and smoking

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 7 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

persistence, both of which can partially mediate the MDD– Benefits may include improved mental health outcomes and
mortality association. Additionally, after an 11-year interval, data reduced all-cause mortality.
from the CCHS-MH 2012 continues to underscore the enduring
Our study has some limitations. First, the baseline data is limited
protective impact of PA on mortality, independent of its influence
by the cross-sectional nature of the study, which did not include
on depression. It is noteworthy that when a more robust
certain important populations (e.g., residents of institutions).
statistical model was applied, including potential covariates
Second, although the study included several questions related to
similar to those used in the CCHS 2002 analysis, the hazard
the intensity, frequency, and type of PA, the questionnaire was
ratios increased substantially (HR = 2.18). Despite this elevation
not validated and did not include objective measures of PA (e.g.,
in mortality risk, it is worth highlighting that PA levels exhibited
accelerometer). Additionally, there were differences in the PA
even greater protective effects after a decade. In practical terms,
questions between the 2002 and 2012 surveys. Therefore, future
adhering to PA recommendations (e.g., even slightly below or
epidemiological studies should apply direct measures to provide
above 150 min/week) may yield a substantial 40–45% reduction
more accurate quantification of PA levels in a population. Addi-
in the risk of mortality.
tionally, since inflammation may partially mediate the associa-
Furthermore, to delve deeper into our findings, we conducted an tion between MDD and mortality, future studies may benefit
analysis aimed at evaluating the relative excess risk resulting from measuring pro-inflammatory biomarkers. Given that the
from the interaction between PA levels and their association with limitations of PA measurement apply equally to both depressed
mortality in major depression. While the interaction model did and nondepressed participants, any bias arising from the
not yield statistically significant results, our exploratory analysis misclassification of PA status would likely be non-differential.
did reveal a possible greater-than-additive interaction, indicating This could lead to a dilution of the HRs, potentially under-
a heightened risk of mortality in both the CCHS 2002 (RERI = estimating the benefits of PA. Furthermore, as the survey
0.45) and CCHS-MH 2012 (RERI = 0.61) datasets. Although our datasets are cross-sectional, they cannot fully clarify temporal
results are consistent with the greater-than-additive interaction relationships. In some cases, medical conditions or injuries may
reported by Liu et al., the lack of statistical significance means have limited participants’ PA, and the treatment of depression
that our results did not replicate this finding decisively. could not be clearly delineated using the survey data.
Physical inactivity causes economic and social losses and In conclusion, regardless of MDD, following PA recommen-
increases mortality, regardless of mental health status [43, 44]. dations of >150 min/week or even engaging in less than the
Physical activity may protect against mortality through multiple recommended levels of PA may contribute to a greater protective
pathways, including the reduction of risk factors for mortality, impact against elevated mortality and reduce health-related risk
such as cardiovascular disease, diabetes, and other noncom- factors, thereby enhancing overall health and well-being.
municable diseases. In addition to prevention, regular PA may Strategies to reduce physical inactivity and promote PA should
help to manage several health conditions, preventing the pro- be encouraged. Every step is important, and engaging in some PA
gression of comorbidities associated with chronic diseases. For is better than none.
example, there is a high prevalence of hypertension, diabetes,
and other chronic diseases among individuals with MDD [45–
48]. These chronic conditions may improve when PA levels Acknowledgments
follow the World Health Organization (WHO) recommendations This work was supported by the Cuthbertson & Fischer Chair in
[20]. Regular PA stimulates mechanisms, including improve- Pediatric Mental Health held by Patten during the time period
ments in vascular endothelial function, structural vascular when this research was conducted.
adaptations, cardioprotection, and improved autonomic balance
(e.g., increases vagal tone to the heart) [49]. In addition, PA may
promote a reduction in inflammation (e.g., cytokines). If uncon- Funding
trolled, acute inflammation can become chronic, potentially
reducing the effectiveness of the anti-inflammatory response S.P. is supported by the Cuthbertson and Fischer Chair in Pediatric
[50]. During PA, muscles act on immune receptors by secreting Mental Health at the University of Calgary. The analysis was
muscle-specific cytokines or “myokines”, which may have local conducted at the Prairie Regional Research Data Centre, which is
and systemic influences on inflammation and metabolism [51]. part of the Canadian Research Data Centre Network (CRDCN). The
From a neurobiological perspective, PA may contribute to services and activities provided by the CRDCN are made possible
increased neuroplasticity involving brain-derived neurotrophic by the financial or in-kind support of the SSHRC, CIHR, CFI,
factor and increased blood flow to the brain, which may provide Statistics Canada, and participating universities, whose support is
neuronal support to prevent atrophy in brain regions associated gratefully acknowledged. The views expressed in this paper do not
with several types of mental disorder, including depression [52, necessarily represent those of the CRDCN or its partners.
53]. In this perspective, PA might regulate serotonergic and
dopaminergic levels, which are associated with improved mood
and stress response [54, 55]. Additionally, we highlight the effect
Author contributions
of PA on cognitive function, which enhances an individual’s Conceptualization, V.D.O.T., N.L.G.C. and S.P.; methodology,
ability to produce appropriate responses to the environment [56]. V.D.O.T. and N.L.G.C.; software, V.D.O.T.; validation, S.P.;
formal analysis, V.D.O.T.; resources, S.P.; writing—original draft
Symptoms of depression, such as fatigue, anhedonia, and hope-
preparation, V.D.O.T.; writing—review and editing, V.D.O.T.,
lessness, may interfere with participation in PA. Specialized
J.V.A.W., V.S., A.B. and G.D.; supervision, S.P. and N.L.G.C. All
interventions capable of overcoming these barriers should be
authors have read and agreed to the published version of the
developed and implemented in psychiatric treatment settings.
manuscript.

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 8 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

Conflict of interest beneficiaries with treatment-resistant depression. J Manag

Care Spec Pharm. 2018;24(3):226–36.
The authors declare no conflict of interest.
3. Institute for Health Metrics and Evaluation. Global health
data exchange (GHDx). Seattle (WA): Institute of Health
Data availability statement Metrics and Evaluation; 2021.
Data supporting these findings are available within the article, at 4. Momen NC, Plana-Ripoll O, Agerbo E, Christensen MK,
https://doi.org/10.20935/MHealthWellB7335, or upon request. Iburg KM, Laursen TM, et al. Mortality associated with
mental disorders and comorbid general medical conditions.
JAMA Psychiatry. 2022;79:444–53. doi: 10.1001/jamapsy-
Institutional review board statement chiatry.2022.0347
Not applicable. 5. Patten SB, Williams JVA, Bulloch AGM. Major depressive
episodes and mortality in the Canadian household popu-
lation. J Affect Disord. 2019;242:165–71. doi: 10.1016/j.
Informed consent statement jad.2018.08.030
Not applicable. 6. Bremmer MA, Deeg DJH, Beekman ATF, Penninx BWJH,
Lips P, Hoogendijk WJG. Major depression in late life is
associated with both hypo- and hypercortisolemia. Biol
Sample availability Psychiatry. 2007;62:479–86. doi: 10.1016/j.biopsych.2006.
The authors declare no physical samples were used in the study. 11.033

7. Galvão ACM, de Almeida RN, Silva EAS, Freire FAM,

Additional information Palhano-Fontes F, Onias H, et al. Cortisol modulation by
ayahuasca in patients with treatment resistant depression
Received: 2024-06-13 and healthy controls. Front Psychiatry. 2018;9:185. doi: 10.
Accepted: 2024-08-26 3389/fpsyt.2018.00185
Published: 2024-09-13 8. Galvão-Coelho NL, de Menezes Galvão AC, de Almeida RN,
Palhano-Fontes F, Campos Braga I, Lobão Soares B, et al.
Academia Mental Health and Well-Being papers should be cited
Changes in inflammatory biomarkers are related to the
as Academia Mental Health and Well-Being 2024, ISSN 2997-
antidepressant effects of Ayahuasca. J Psychopharmacol.
9196, https://doi.org/10.20935/MHealthWellB7335. The journal’s
2020;34:1125–33. doi: 10.1177/0269881120936486
official abbreviation is Acad. Ment. Health WellB.
9. Varela YM, de Almeida RN, Galvão ACM, de Sousa GM, de
Lima ACL, da Silva NG, et al. Psychophysiological responses
Publisher’s note to group cognitive-behavioral therapy in depressive patients.
Academia.edu Journals stays neutral with regard to jurisdictional Curr Psychol. 2021;42:592–601. doi: 10.1007/s12144-020-
claims in published maps and institutional affiliations. All claims 01324-9
expressed in this article are solely those of the authors and do not 10. De Menezes Galvão AC, Almeida RN, De Sousa Junior GM,
necessarily represent those of their affiliated organizations, or Leocadio-Miguel MA, Palhano-Fontes F, De Araujo DB, et al.
those of the publisher, the editors, and the reviewers. Any Potential biomarkers of major depression diagnosis and
product that may be evaluated in this article, or claim that may chronicity. PLoS One. 2021;16:0257251. doi: 10.1371/journal.
be made by its manufacturer, is not guaranteed or endorsed by pone.0257251
the publisher.
11. Galvão ACM, Almeida RN, de Sousa Júnior GM, Leocadio-
Miguel MA, Palhano-Fontes F, de Araujo DB, et al. Potential
Copyright biomarkers of major depression diagnosis and chronicity.
PLoS One. 2021;16(9):e0257251.
© 2024 copyright by the authors. This article is an open access
article distributed under the terms and conditions of the 12. Furman D, Campisi J, Verdin E, Carrera-Bastos P, Targ S,
Creative Commons Attribution (CC BY) license (https:// Franceschi C, et al. Chronic inflammation in the etiology of
creativecommons.org/licenses/by/4.0/). disease across the life span. Nat Med. 2019;25:1822–32. doi:
10.1038/s41591-019-0675-0

13. Gilman SE, Sucha E, Kingsbury M, Horton NJ, Murphy JM,

References Colman I. Depression and mortality in a longitudinal study:
1. GBD 2019 Mental Disorders Collaborators. Global, regional, 1952–2011. CMAJ. 2017;189:E1304–10. doi: 10.1503/cmaj.
and national burden of 12 mental disorders in 204 countries 170125
and territories, 1990–2019: a systematic analysis for the
14. Holt-Lunstad J, Smith TB, Layton JB. Social relationships
Global Burden of Disease Study 2019. Lancet Psychiatry.
and mortality risk: a meta-analytic review. PLoS Med.
2022;9:137–50. doi: 10.1016/S2215-0366(21)00395-3
2010;7:e1000316. doi: 10.1371/journal.pmed.1000316
2. Olfson M, Amos TB, Benson C, Mcrae J, Marcus SC.
15. Oude Voshaar RC, Aprahamian I, Borges MK, van den Brink
Prospective service use and health care costs of medicaid
RHS, Marijnissen RM, Hoogendijk EO, et al. Excess

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 9 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

mortality in depressive and anxiety disorders: the lifelines 27. Ross RE, VanDerwerker CJ, Saladin ME, Gregory CM. The
cohort study. Eur Psychiatry. 2021;64(1):e54. doi: 10.1192/ role of exercise in the treatment of depression: biological
j.eurpsy.2021.2229 underpinnings and clinical outcomes. Mol Psychiatry. 2023.
28(1):298–328. doi: 10.1038/s41380-022-01819-w
16. Schuch F, Vancampfort D, Firth J, Rosenbaum S, Ward P,
Reichert T, et al. Physical activity and sedentary behavior in 28. Lavebratt C, Herring MP, Liu JJ, Wei YB, Bossoli D,
people with major depressive disorder: a systematic review Hallgren M, et al. Interleukin-6 and depressive symptom
and meta-analysis. J Affect Disord. 2017;210:139–50. doi: severity in response to physical exercise. Psychiatry Res.
10.1016/j.jad.2016.10.050 2017;252:270–6. doi: 10.1016/j.psychres.2017.03.012
17. Katzmarzyk PT, Powell KE, Jakicic JM, Troiano RP, Piercy 29. Swift DL, McGee JE, Earnest CP, Carlisle E, Nygard M,
K, Tennant B. Sedentary behavior and health: update from Johannsen NM. The effects of exercise and physical activity
the 2018 physical activity guidelines advisory committee. on weight loss and maintenance. Prog Cardiovasc Dis.
Med Sci Sports Exerc. 2019;51:1227–41. doi: 10.1249/ 2018;61:206–13. doi: 10.1016/j.pcad.2018.07.014
MSS.0000000000001935
30. Firth J, Stubbs B, Vancampfort D, Schuch F, Lagopoulos J,
18. Lear SA, Hu W, Rangarajan S, Gasevic D, Leong D, Iqbal R, Rosenbaum S, et al. Effect of aerobic exercise on hippo-
et al. The effect of physical activity on mortality and campal volume in humans: a systematic review and meta-
cardiovascular disease in 130000 people from 17 high- analysis. Neuroimage. 2018;166:230–8. doi: 10.1016/j.
income, middle-income, and low-income countries: the neuroimage.2017.11.007
PURE study. Lancet. 2017;390:2643–54. doi: 10.1016/
31. Liu BP, Jia CX. The associations of physical activity and
S0140-6736(17)31634-3
lifetime depression with all-cause and cause-specific mortal-
19. Xu C, Furuya-Kanamori L, Liu Y, Færch K, Aadahl M, Seguin ity: evidence from a prospective cohort study. Psychiatry
AR, et al. Sedentary behavior, physical activity, and all-cause Res. 2023;324:115206. doi: 10.1016/j.psychres.2023.115206
mortality: dose-response and intensity weighted time-use
32. Perez-Lasierra JL, Moreno-Franco B, González-Agüero A,
meta-analysis. J Am Med Dir Assoc. 2019;20:1206–12.e3.
Lobo E, Casajus JA. A cross-sectional analysis of the
doi: 10.1016/j.jamda.2019.05.001
association between physical activity, depression, and all-
20. Bull FC, Al-Ansari SS, Biddle S, Borodulin K, Buman MP, cause mortality in Americans over 50 years old. Sci Rep.
Cardon G, et al. World Health Organization 2020 guidelines 2022;12(1):2264. doi: 10.1038/s41598-022-05563-7
on physical activity and sedentary behaviour. Br J Sports
33. Solmi M, Radua J, Olivola M, Croce E, Soardo L, Salazar de
Med. 2020;54(24):102955. doi: 10.1136/bjsports-2020-
Pablo G, et al. Age at onset of mental disorders worldwide:
102955
large-scale meta-analysis of 192 epidemiological studies.
21. Kraus WE, Powell KE, Haskell WL, Janz KF, Campbell WW, Mol Psychiatry. 2022;27:281–95. doi: 10.1038/s41380-021-
Jakicic JM, et al. Physical activity, all-cause and cardio- 01161-7
vascular mortality, and cardiovascular disease. Med Sci
34. Lix L, Finlayson G, Yogendran M, Bond R, Bodnarchuk J,
Sports Exerc. 2019;51:1270–81. doi: 10.1249/MSS. 00000
Soodeen R. Primary prevention: an examination of data
00000001939
capabilities in Manitoba. Winnipeg (MB): University of
22. Hallgren M, Nguyen TTD, Lundin A, Vancampfort D, Stubbs Manitoba; 2005.
B, Schuch F, et al. Prospective associations between physical
35. Ainsworth BE, Haskell WL, Leon AS, Jacobs DRJ, Montoye
activity and clinician diagnosed major depressive disorder in
HJ, Sallis JF, et al. Compendium of physical activities:
adults: a 13-year cohort study. Prev Med (Baltim). 2019;118:
classification of energy costs of human physical activities.
38–43. doi: 10.1016/j.ypmed.2018.10.009
Med Sci Sports Exerc. 1993;25:71–80.
23. Pearce M, Garcia L, Abbas A, Strain T, Schuch FB, Golubic
36. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz
R, et al. Association between physical activity and risk of
AM, Strath SJ, et al. Compendium of physical activities: an
depression: a systematic review and meta-analysis. JAMA
update of activity codes and MET intensities. Med Sci Sports
Psychiatry. 2022;79(6):550–9. doi: 10.1001/jamapsychiatry.
Exerc. 2000;32:S498–504.
2022.0609
37. Katzmarzyk PT, Tremblay MS. Limitations of Canada’s
24. Schuch FB, Vancampfort D, Firth J, Rosenbaum S, Ward PB,
physical activity data: implications for monitoring trends.
Silva ES, et al. Physical activity and incident depression: a
Appl Physiol Nutr Metab. 2007;32:S2E. doi: 10.1139/H07-
meta-analysis of prospective cohort studies. Am J Psychi-
113
atry. 2018;175:631–48. doi: 10.1176/appi.ajp.2018.17111194
38. Manley AF. Physical activity and health: a report of the
25. Lee Y, Park S. Understanding of physical activity in social
surgeon general. Collingdale (PA): DIANE Publishing; 1996.
ecological perspective: application of multilevel model.
Front Psychol. 2021;12:622929. doi: 10.3389/fpsyg.2021. 39. Canadian Fitness and Lifestyle Research Institute. Increas-
622929 ing physical activity: trends for planning effective commu-
nication. Ottawa (ON): The Institute; 2004.
26. Zhao G, Lu Z, Sun Y, Kang Z, Feng X, Liao Y, et al. Dissecting
the causal association between social or physical inactivity 40. Kessler RC, Üstün TB. The World Mental Health (WMH)
and depression: a bidirectional two-sample Mendelian Survey Initiative version of the World Health Organization
randomization study. Transl Psychiatry. 2023;13:94. doi: (WHO) Composite International Diagnostic Interview
10.1038/s41398-023-02492-5

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 10 of 11

https://www.academia.edu/journals/academia-mental-health-and-well-being/about https://doi.org/10.20935/MHealthWellB7335

(CIDI). Int J Methods Psychiatr Res. 2004;13:93–121. doi: 49. Fiuza-Luces C, Santos-Lozano A, Joyner M, Carrera-Bastos
10.1002/mpr.168 P, Picazo O, Zugaza JL, et al. Exercise benefits in cardi-
ovascular disease: beyond attenuation of traditional risk
41. Streiner DL, Patten SB, Anthony JC, Cairney J. Has lifetime
factors. Nat Rev Cardiol. 2018;15:731–43. doi: 10.1038/
prevalence’ reached the end of its life? An examination of the
s41569-018-0065-1
concept. Int J Methods Psychiatr Res. 2009;18:221–8. doi:
10.1002/mpr.296 50. Slavin MB, Khemraj P, Hood DA. Exercise, mitochondrial
dysfunction and inflammasomes in skeletal muscle. Biomed
42. Sanmartin C, Decady Y, Trudeau R, Dasylva A, Tjepkema M,
J. 2023;47(1):100636. doi: 10.1016/j.bj.2023.100636
Finès P, et al. Health reports linking the canadian
community health survey and the Canadian mortality 51. Nieman DC, Wentz LM. The compelling link between
database: an enhanced data source for the study of mortality. physical activity and the body’s defense system. J Sport
Health Rep. 2016;27:10–8. Health Sci. 2019;8(3):201–17. doi: 10.1016/j.jshs.2018.09.
009
43. Santos AC, Willumsen J, Meheus F, Ilbawi A, Bull FC. The
cost of inaction on physical inactivity to public health-care 52. Phillips C. Brain-derived neurotrophic factor, depression,
systems: a population-attributable fraction analysis. Lancet and physical activity: making the neuroplastic connection.
Glob Health. 2023;11:e32–9. doi: 10.1016/S2214-109X(22) Neural Plast. 2017;2017:7260130. doi: 10.1155/2017/726
00464-8 0130

44. WHO. Global status report on physical activity 2022. 53. Schwert C, Stohrer M, Aschenbrenner S, Weisbrod M,
Geneva; 2022. Schröder A. Neurocognitive profile of outpatients with
unipolar depressive disorders. J Clin Exp Neuropsychol.
45. Anwar N, Kuppili PP, Balhara YPS. Depression and physical
2019;41:913–24. doi: 10.1080/13803395.2019.1634180
noncommunicable diseases: the need for an integrated
approach. WHO South East Asia J Public Health. 2018; 54. Marques A, Marconcin P, Werneck AO, Ferrari G, Gouveia
6:12–7. ÉR, Kliegel M, et al. Review bidirectional association
between physical activity and dopamine across adulthood-a
46. Cohen BE, Edmondson D, Kronish IM. State of the art
systematic review. Brain Sci. 2021;11(7):829. doi: 10.3390/
review: depression, stress, anxiety, and cardiovascular
brainsci
disease. Am J Hypertens. 2015;28:1295–302. doi: 10.1093/
ajh/hpv047 55. Zimmer P, Stritt C, Bloch W, Schmidt FP, Hübner ST,
Binnebößel S, et al. The effects of different aerobic exercise
47. Scott KM, Lim C, Al-Hamzawi A, Alonso J, Bruffaerts R,
intensities on serum serotonin concentrations and their
Caldas-de-Almeida JM, et al. Association of mental
association with Stroop task performance: a randomized
disorders with subsequent chronic physical conditions:
controlled trial. Eur J Appl Physiol. 2016;116:2025–34. doi:
world mental health surveys from 17 countries. JAMA
10.1007/s00421-016-3456-1
Psychiatry. 2016;73:150–8. doi: 10.1001/jamapsychiatry.
2015.2688 56. Tavares VDO, Rossell SL, Schuch FB, Herring M, Menezes
G, Galvão-Coelho NL, et al. Effects of exercise on cognitive
48. Vancampfort D, Correll CU, Galling B, Probst M, De Hert M,
functioning in adults with serious mental illness: a meta
Ward PB, et al. Diabetes mellitus in people with schizo-
analytic review. Psychiatry Res. 2023;321:115081. doi: 10.
phrenia, bipolar disorder and major depressive disorder: a
1016/j.psychres.2023.115081
systematic review and large scale meta-analysis. World
Psychiatry. 2016;15:166–74. doi: 10.1002/wps.20309

ACADEMIA MENTAL HEALTH AND WELL-BEING 2024, 1 11 of 11