Health Psychology

2013, Vol. 32, No. 6, 609 615

Copyright 2012 the American Psychological Association

0278-6133/13/$12.00 DOI: 10.1037/a0029276

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Physical Activity and Reduced Risk of Depression:

Results of a Longitudinal Study of Mexican Adults
Katia Gallegos-Carrillo

Yvonne N. Flores

Instituto Mexicano del Seguro Social, Cuernavaca, Mxico

Instituto Mexicano del Seguro Social, Cuernavaca, Mxico, and

Jonsson Comprehensive Cancer Center, University of California
at Los Angeles

Edgar Denova-Gutirrez

Pablo Mndez-Hernndez

Instituto Mexicano del Seguro Social, Cuernavaca, Mxico, and

Universidad Autnoma del Estado de Mxico

Universidad Autnoma de Tlaxcala and Secretara de Salud de

Tlaxcala, Tlaxcala, Mxico

Libia D. Dosamantes-Carrasco

Santiago Henao-Morn

Universit Lyon 1 and Lyon 2

Instituto Mexicano del Seguro Social, Cuernavaca, Mxico

Guilherme Borges

Elizabeth Halley-Castillo

Instituto Nacional de Psiquiatra Juan Ramn de la Fuente,

Secretara de Salud, Mxico, D.F., and Universidad Autnoma
Metropolitana, Unidad Xochimilco, Mxico, D.F.

Universidad Autnoma del Estado de Mxico and Hospital de

Gineco-Obstetricia Dr. Luis Castelazo Ayala, Instituto
Mexicano del Seguro Social, Mxico, D.F.

Nayeli Macias

Jorge Salmern

Instituto Nacional de Salud Pblica, Cuernavaca, Mxico

Instituto Mexicano del Seguro Social, Cuernavaca, Mxico, and

Instituto Nacional de Salud Pblica, Cuernavaca, Mxico

Objective: To evaluate the effect of physical activity (PA) on the risk of depression among Mexican adults over a
6-year follow-up period. Method: We evaluated longitudinal data from the Health Worker Cohort Study, which
follows employees of the Mexican Institute for Social Security in Morelos State, Mexico, over time. Depressive
symptoms and PA were assessed at baseline and at a follow-up measurement 6 years later. The study population
was free of depressive symptomatology at baseline, as assessed by the Center for Epidemiological StudiesDepression Scale (CES-D). After 6 years, the CES-D was completed once again by the participants to estimate their
risk of depression based on the different PA patterns they reported during the follow-up period. PA was estimated
using a questionnaire that has been applied in similar longitudinal studies and has been validated in Spanish, with
metabolic equivalents (METs) as the unit of measurement. We identified three PA patterns: highly active,
moderately active, and inactive. The relative risk of depression (CES-D score 16 points) was estimated using
multivariate logistical regression analysis according to the PA patterns at a follow-up measurement 6 years later.
Results: The incidence of depression after 6 years was higher among inactive participants (16.5%) than among those
with an active PA pattern (10.6%). We found that more active PA patterns have an important protective effect
against depression. The odds ratio (OR) for the more active PA patterns was 0.46, 95% confidence interval (CI)
[0.25, 0.87], and for individuals with a moderately active PA pattern, the OR was 0.57, 95% CI [0.34, 0.93]. These
ORs were obtained after adjusting for confounding variables and baseline PA levels. Conclusion: Our results

This article was published Online First September 3, 2012.

Katia Gallegos-Carrillo, Unidad de Investigacin Epidemiolgica y en
Servicios de Salud, Instituto Mexicano del Seguro Social, Cuernavaca,
Mxico; Yvonne N. Flores, Unidad de Investigacin Epidemiolgica y en
Servicios de Salud, Instituto Mexicano del Seguro Social, Cuernavaca,
Mxico, and Division of Cancer Prevention & Control Research, School of
Public Health, Jonsson Comprehensive Cancer Center, University of California at Los Angeles; Edgar Denova-Gutirrez, Unidad de Investigacin
Epidemiolgica y en Servicios de Salud, Instituto Mexicano del Seguro
Social, Cuernavaca, Mxico, and Centro de Investigacin en Ciencias
Mdicas, Universidad Autnoma del Estado de Mxico, Toluca, Mxico;
Pablo Mndez-Hernndez, Facultad de Ciencias de la Salud, Universidad
Autnoma de Tlaxcala, Tlaxcala, Mxico, and Jefatura de Investigacin,
Secretara de Salud de Tlaxcala, Tlaxcala, Mxico; Libia D. DosamantesCarrasco, Equipe MA2D, Laboratoire ERIC, Universit Lyon 1 and Lyon
2, Lyon, France; Santiago Henao-Morn, Unidad de Investigacin Epide-

miolgica y en Servicios de Salud, Instituto Mexicano del Seguro Social,

Cuernavaca, Mxico; Guilherme Borges, Instituto Nacional de Psiquiatra
Juan Ramn de la Fuente, Secretara de Salud, Mxico, D.F., and
Universidad Autnoma Metropolitana, Unidad Xochimilco, Mxico, D.F.;
Elizabeth Halley-Castillo, Centro de Investigacin en Ciencias Mdicas,
Universidad Autnoma del Estado de Mxico, Toluca, Mxico, and Hospital de Gineco-Obstetricia Dr. Luis Castelazo Ayala, Instituto Mexicano
del Seguro Social, Mxico, D.F.; Nayeli Macias, Centro de Investigacin
en Nutricin y Salud, Instituto Nacional de Salud Pblica, Cuernavaca,
Mxico; Jorge Salmern, Unidad de Investigacin Epidemiolgica y en
Servicios de Salud, Instituto Mexicano del Seguro Social, Cuernavaca,
Mxico, and Centro de Investigacin en Salud Poblacional, Instituto Nacional de Salud Pblica, Cuernavaca, Mxico.
Correspondence concerning this article should be addressed to Jorge
Salmern, Boulevard Benito Jurez 31, Centro. C.P., 62000, Cuernavaca,
Morelos, Mxico. E-mail: jsalme@prodigy.net.mx
609

GALLEGOS-CARRILLO ET AL.

610

indicate that PA may reduce risk of depression in Mexican adults. These findings have potential applications for
depression prevention programs in target populations with similar social and cultural contexts.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Keywords: depression, physical activity, adults

For several decades, the beneficial effect of physical activity

(PA) on different health outcomes has been documented (World
Health Organization, 2010). Moreover, physical inactivity is now
considered a modifiable risk factor for depression and other
chronic diseases (Warburton, Nicol, & Bredin, 2006). Regular PA
improves psychosocial welfare in the short and the long term by
reducing feelings of stress, anxiety, and depression (World Health
Organization, 2010; Warburton et al., 2006). Intervention studies,
particularly clinical trials investigating the use of PA as a treatment
for depression, have concluded that PA is an effective aspect of
depression treatment (Warburton et al., 2006; Brosse, Sheets, Lett,
& Blumenthal, 2002; Craft & Perna, 2004; Penedo & Dahn, 2005;
Saxena, Ommeren, Tang, & Armstrong, 2005).
Epidemiological evidence also supports the notion that PA may
alleviate depressive symptoms (Mead et al., 2008; Dinas, Koutedakis, & Flouris, 2010). One of the first prospective studies to
document this relationship showed that, over 8 years of follow-up,
participants who initially reported depressive symptoms and little
or no PA at baseline were more likely to have depressive symptoms at the end of the study than participants who had both
depression symptoms and higher levels of PA at baseline (Farmer
et al., 1988).
These results support those of similar studies that used different
follow-up periods (Camacho, Roberts, Lazarus, Kaplan, & Cohen,
1991), some covering up to 20 years, and included both men and
women (Paffenbarger, Lee, & Leung, 1994; Brown, Ford, Burton,
Marshall, & Dobson, 2005). These studies established the first
longitudinal evidence of the mental health benefits associated with
PA, especially for depressive symptoms. This relationship, however, is not yet fully accepted, because other longitudinal studies
have not corroborated this association (Weyerer, 1992; CooperPatrick, Ford, Mead, Chang, & Klag, 1997).
Thus, it is necessary to further examine the protective effects of
PA against the development of depressive symptoms in diverse
populations and to help reduce controversies about its beneficial
effects. To our knowledge, ours is the first study to assess the role
of PA in protecting against depressive symptoms among a sample
of Mexican adults. The present study is relevant because the
estimated prevalence of major depressive disorders among Mexican adults is 3.7% (Medina-Mora et al., 2005), making depression
an important issue for public health research and prevention program in Mxico (Murray & Lpez, 1997); and a better understanding of the cultural and social context of depression is important for
the design and implementation of effective prevention programs.

Method
The study population was comprised of participants from the
Health Worker Cohort Study, a longitudinal study that is investigating the relationships between lifestyle and health in
Mexico. Details of the study design, methodology, and participants characteristics have been previously reported
(Salmern-Castro,
Arillo-Santilln,
Campuzano-Rincn,

Lpez-Antuano, & Lazcano-Ponce, 2002; Mudgal, Borges,

Daz-Montiel, Flores, & Salmern, 2006). Briefly, 1,335 male
and female employees from the Mexican Institute of Social
Security (IMSS) in Morelos State were invited to take part in
the study. Baseline information was collected from 1998 2000,
and participants were assessed at a follow-up point 6 years later,
between 2004 and 2006. Information about sociodemographic,
lifestyle, and medical history factors was obtained through
self-administered questionnaires, and anthropometric measurements and clinical evaluations took place at baseline and at
Year 6.
To evaluate the effect of PA on the incidence of depressive
symptoms at the second assessment, we analyzed data from the
1,047 subjects who were free of depressive symptoms at baseline,
as determined by the cutoff point of a score of 16 on the Center
for Epidemiologic StudiesDepression Scale (CES-D; Radloff,
1977). The IMSS National Research Commission and Ethics Committee evaluated and approved all study procedures.

Data Collection
PA assessment. The PA level of participants was assessed
using a self-administered questionnaire that was applied in similar
follow-up studies (Chasan-Taber et al., 1996; Wolf et al., 1994).
The questionnaire has a validated Spanish translation (MartnezGonzlez, Lpez-Fontana, Varo, Snchez-Villegas, & Martnez,
2005), which has been adapted for use in the Mexican population,
and was previously applied to the entire IMSS cohort study population (Mndez-Hernndez et al., 2009). The questionnaire is
self-administered and estimates the minutes devoted to the practice
of different recreational physical activities during a typical week in
the last year (including walking, running, cycling, aerobics, dancing, and swimming as well as playing football, volleyball, basketball, tennis, fronton, baseball, softball, and squash, among other
activities). Each item includes time intervals that allow participants
to detail the exact number of minutes or hours they dedicate to
each form of recreational PA, as well as the intensity of each PA
(light, moderate, vigorous). The total duration of each recreational
PA was expressed in minutes per day. We calculated the number
of hours per week devoted to each activity, which were then
multiplied by the intensity of each activity, defined as multiples of
the metabolic equivalent (MET) of sitting quietly. We used the
Compendium of Physical Activities to assign METs to each activity (Ainsworth et al., 2000). We then added the average weekly
energy expenditure attributable to each activity to derive the total
MET-hours per week.
The following procedures were used to generate PA patterns.
First, we identified the total number of participants who reported
no PA or very low levels of PA ( 3 METs) (Chasan-Taber et al.,
2002). Next, we used the PA data from the rest of the participants
to generate terciles expressed as METs ( 3 to 11.01, 11.01
to 27.18, and 27.18 MET-hours per week), thus creating a PA
level variable with four categories, as a baseline measurement. In

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

PHYSICAL ACTIVITY AND REDUCED RISK OF DEPRESSION

the Year 6 assessment, we once again estimated the participants

PA levels using the same procedure ( 3 METs, 3 to 8.7,
8.7 to 21.5, and 21.5 MET-hours per week). To determine the
PA levels throughout the study, we considered the baseline values
and those reported at Year 6. This generated 16 possible combinations, which served as a basis for estimating the following three
PA patterns that the participants engaged in during the study: (a)
participants who reported being inactive at baseline and the Year
6 follow-up or participants who reported changing from another
category at baseline to being inactive at follow-up (the inactive PA
pattern); (b) participants who reported low or moderate PA level at
baseline and at Year 6 follow-up or participants who reported
changing from another category at baseline to a low or moderate
level of PA at follow-up (the moderately active PA pattern); and
(c) participants who reported high PA levels at baseline and at
Year 6 follow-up or participants who reported being inactive at
baseline and changing to a high PA level at follow-up (the highly
active PA pattern).
Depression. We used a self-administered version of the
CES-D (Radloff, 1977), which was designed to evaluate the severity of depressive symptoms in population studies, and has been
validated in several populations, including Spanish-speaking
groups (Caraveo, Medina-Mora, Villatoro, & Rascon, 1994; Salgado de Snyder & Maldonado, 1994; Benjet, Guzman, Quintanilla,
Roque, & Leon, 1999). A 20-item version was used to determine
which participants were free of depressive symptoms at baseline,
and to evaluate depressive symptoms at Year 6. The CES-D
generates a continuous scale from 0 60, and a score of 16 or more
suggests probable clinical depression (Radloff, 1977).

611

Covariates
Height was measured using a conventional stadiometer while
participants were standing barefoot with their shoulders in a normal position. Weight was measured with a previously calibrated
electronic Tanita scale (model BC-533). Body mass index (BMI)
was obtained from standardized measurements of weight and
height, and was computed as a ratio of weight (in kilograms) to
height (in square meters). BMI values were categorized into normal weight (BMI range: 18.524.9 kg/m2), overweight (BMI
range: 2529.9 kg/m2), and obese (BMI range: 30kg/m2)
(World Health Organization, 1995).
The following sociodemographic and health data were collected
using a self-administered questionnaire: sex, age ( 40, 40 60,
60 years), education level (primary, secondary, vocational school,
and university or more), marital status (married, widowed, separated/divorced, and single), and tobacco use (never, ex-smoker,
and smoker). The following variables were considered to determine participants health status: (a) self-reported health problems
that have impeded daily tasks in the last 6 months; (b) self-reported
previous medical diagnoses of chronic conditions, such as arthritis,
asthma, bronchitis, cirrhosis, diabetes, cardiovascular disease, arterial hypertension, or chronic renal insufficiency (based on the
chronic condition information participants reported, we generated
a variable to represent the number of chronic conditions, with the
following two categories: none, and one or more chronic conditions); (c) self-reported mobility issues that impeded daily life
activities, this variable was defined as either with and without
mobility problems; and (d) self-reported health status over the

Table 1
Characteristics of the Study Population According to Physical Activity Level at Baseline (N 1,047)
Population characteristics free
of depressive symptoms
Sex
Male
Female
Age
40 years
4060 years
60 years
Marital status
Married
Widowed
Divorced
Single
Educational level
Elementary
Secondary
High/technical
Bachelor or higher
Tobacco consumption
Never
Ex-smoker
Smoker
Body mass index
Normal ( 25 kg/m2)
Overweight (2529.9 kg/m2)
Obese ( 30 kg/m2)

Inactive
PAa

PA Tertile 2c

PA Tertile
3d

%
22.5
77.5

3.0 METs
24.6
44.6

3.0 11.0 METs

19.9
19.1

11.0 27.2 METs

25.4
18.2

27.2 METs
30.1
18.0

45.2
51.6
3.25

34.1
44.7
52.9

18.2
20.0
20.6

21.8
18.0
23.5

25.8
17.2
2.9

67.3
2.9
11.8
17.9

39.4
48.4
50.4
34.2

20.9
22.6
11.4
17.6

19.8
22.6
17.9
21.4

19.8
6.4
20.3
26.7

8.6
17.4
21.3
52.7

39.8
46.9
40.3
36.5

19.3
20.9
18.5
18.2

20.4
22.0
18.5
20.0

20.4
10.2
22.7
24.3

64.8
13.0
22.2

40.0
40.6
37.2

21.2
15.6
17.9

16.6
24.2
23.8

22.1
19.5
21.1

38.2
43.2
18.6

35.7
37.2
55.9

21.5
20.6
11.8

20.2
20.8
16.9

22.5
21.5
15.4

Note. PA physical activity; METs metabolic equivalents.

n 420 (40.1%). b n 202 (19.3%). c n 208 (19.9%).

PA Tertile 1b

n 217 (20.7%).

GALLEGOS-CARRILLO ET AL.

612

previous 12 months, measured using a question that allowed participants to choose from the following options: improved, remained the same, or worsened.

Table 2
Physical Activity Patterns During the Study According to
Population Characteristics at Follow-Up (N 1,047)
Inactive Moderately
Highly
pattern active pattern active pattern
(36.5%)a
(40.2%)b
(23.3%)c

Statistical Analysis

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Population characteristics

The sociodemographic and health variables were examined by

the baseline PA levels and PA patterns (inactive, moderately
active, and highly active) of participants at Year 6. Differences in
PA patterns based on sociodemographic and health variables were
evaluated using Pearsons chi-square test (Plackett, 1983). We
then estimated the risk of developing depressive symptoms
(CES-D scores 16) in the follow-up assessment based on the
participants PA patterns.
Logistical regression analyses were performed using multivariate models to estimate the odds ratio (OR) of developing depression according to the different PA patterns at the follow-up assessment. The logistical regression models were carried out with
progressive adjustment to determine the independent contribution
of the different variables to the relative risk between depression
and PA patterns. Model 1 was adjusted by age, sex, and PA at
baseline. Model 2 was further adjusted by education level, marital
status, BMI (continuous), tobacco use, health problems in the last
6 months, number of chronic diseases, and mobility problems.
Model 3 was further adjusted by change in self-reported health
status.
All analyses were carried out using Stata, Version 10.1 (Stata
Statistical Software).

Results
At baseline, 40.1% of all participants (N 1,047) who were free
from depressive symptoms reported no or very low levels of PA
( 3 METs). The prevalence of an inactive PA pattern was 52.9%
in those over 60 years of age; 44.6% among women; 50.4% among
separated or divorced individuals; and 55.9% among obese participants. Subjects in the high PA tercile shared the following characteristics: they were more likely to be male, to be under 40 years
of age, single, and to have a high level of education (see Table 1).
Exactly 36.5% of participants reported an inactive PA pattern
and 40.2% had a moderately active PA pattern in the assessment
made 6 years after the baseline measurements. Analysis of the
PA patterns based on the characteristics of the study population
at Year 6 indicated significant differences by sex (men tended
to have more active PA patterns than women); age (only 10% of
those over age 60 reported a high PA pattern); marital status
(15% of widowers had a highly active PA pattern); and BMI
(16.6% of obese participants had a highly active PA pattern
(p .05) (Table 2).
Incidence of depression in the follow-up assessment was higher
in participants with an inactive PA pattern (16.5%), than those who
had a highly active PA pattern throughout the study (10.6%)
(Figure 1).
When adjusted by age, sex, and PA level at baseline, our
logistical regression results demonstrated that highly active PA
patterns reduced the risk of depression in the follow-up assessment
(Model 1) by 56%. It should therefore be noted that similar results
were found when, instead of using PA patterns, we carried out the
same model using PA level at follow-up as an independent vari-

Sex
Male
Female
Age
40 years
4060 years
60 years
Marital status
Married
Widowed
Divorced
Single
Educational level
Elementary
Secondary
High/technical
Professional/postgraduate
Tobacco consumption
Never
Ex-smoker
Smoker
Body mass index
Normal (25 kg/m2)
Overweight (2529.9 kg/m2)
Obese (30 kg/m2)
Health problems in the last 6
months
No
Yes
Mobility problems
No
Yes
Number of chronic diseases
0
1 or more

27.1
39.2

43.6
39.2

29.2
21.6

39.7
35.3
37.9

37.6
39.8
51.9

22.7
24.9
10.1

38.1
44.7
23.9
38.9

38.2
39.5
51.4
39.5

23.7
15.8
24.6
21.6

42.9
43
39.8
32.7

35.7
38.6
36.4
42.4

21.4
18.4
23.7
24.9

37.5
34.5
38.6

39.9
42.1
36.2

22.6
23.4
25.1

33.1
35.1
44.5

41.8
39.6
38.9

25.1
25.3
16.6

36.2
35.6

38.9
44.9

24.9
19.5

35.9
41.1

40.5
43.2

23.6
15.8

35.1
40.4

40.9
38.2

23.9
21.4

Remained at zero or low PA or changed to low or zero. b Remained at

low or regular PA or changed to low or regular. c Remained at a high PA
level or changed from low or zero to a high level.

p .05 (Pearsons chi-square).

able, adjusting by PA level at baseline (data not shown). Figure 2

shows that PA had a protective effect against the relative risk of
depression, which increased when the models were adjusted by
controlling for variables for acute and chronic morbidity, mobility
problems that impeded daily activities, tobacco use, and BMI for
the moderately active PA pattern, OR 0.62, 95% confidence
interval (CI) [0.38, 0.99], and for the highly active PA pattern,
OR 0.47, 95% CI [0.25, 0.86] (Model 2). This significant
protective effect was maintained when the model was also adjusted
by perceived change in health status (Model 3).

Discussion
Our results supported the hypothesis that PA may reduce the
incidence of depressive symptoms. The findings suggest that individuals with higher levels of PA have a lower risk of developing
depressive symptoms as compared to those who have an inactive
or moderate PA pattern.

PHYSICAL ACTIVITY AND REDUCED RISK OF DEPRESSION

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Figure 1. Risk of depression according to different PA patterns at 6 years

of baseline measurement (N 1,047).

In the present study, a highly active PA pattern reduced the risk

of depression by about 56%. This finding is consistent with the
results of previous longitudinal studies of adults of both sexes
(Farmer et al., 1988; Camacho et al., 1991). Yet despite the high
prevalence of depression in the Mexican adult population
(Medina-Mora et al., 2005), this is the first study in Mexican adults
to establish a prospective relationship between low levels of PA
and an increased incidence of depressive symptoms. At the end of
the follow-up period, we found that the risk of developing depression was lower among subjects who had highly active PA patterns;
which indicates that PA may help prevent depressive symptoms
(Brown et al., 2005; Strawbridge, Deleger, Roberts, & Kaplan,
2002). The positive and preventive effects of PA can be explained
by the increased production of brain neurotransmitters, such as
endorphins (Goldfarb & Jamurtas, 1997) and monoamines, which
are produced during PA (Meeusen, 2005). The link may also be

Figure 2.

613

explained by PAs ability to improve ones health status and, in

turn, his or her self-esteem (McAuley & Blissmer, 2000).
In our regression models, the protective role of PA was maintained after the inclusion of possible confounding variables due to
their association with depression and PA, such as age, educational
level, mobility problems (Geerlings, Beekman, Deeg, Twisk, &
van Tilburg, 2001; Chan et al., 2009), health status (Beekman,
Kriegsman, Deeg, & van Tilburg, 1995), chronic diseases (Hays,
Wells, Sherbourne, Rogers, & Spritzer, 1995; Gallegos-Carrillo et
al., 2009), and other variables, including BMI (Onyike, Crum, Lee,
Lyketsos, & Eaton, 2003). Therefore, after adjusting the statistical
model for those variables, the association between PA and depressive symptoms was maintained. These findings demonstrate the
consistency of our model and also corroborate the positive association between PA and depressive symptoms found in other
groups, including women (Brown et al., 2005) and elderly adults
(Strawbridge et al., 2002).
The levels of PA that participants reported in our study are
lower than the levels reported in other prospective studies conducted with both sexes (Cooper-Patrick et al., 1997). This finding
indicates that this sample of generally healthy urban Mexican
adults engages in relatively low levels of PA; nearly 60% of our
study population engaged in low or very low levels of PA. These
results are consistent with prior studies that have reported 65.5% of
participants affiliated with IMSS did not perform sufficient PA for
good health (Acosta-Czares, Aranda-Alvarez, & Reyes-Morales,
2006), and differ from the findings reported in other longitudinal
studies such as Brown et al. (2005), in which only 26% of the
study subjects engaged in low or very low levels of PA. Finally, it
is important to mention that our results can only be generalized to

Relative risk of depression according to different PA patterns (N 1,047).

GALLEGOS-CARRILLO ET AL.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

614

populations of working or retired Mexican adults in urban areas.

Further studies that include more representative population samples are needed to generalize the results to the broader population.
The long period between the baseline and follow-up data collections as well as the use of only one follow-up measurement are
limitations of the study. This design may increase the possibility of
inaccurate classification of both depressive symptoms and PA
levels. Erroneous PA level classification could have occurred
because participation in PA is likely to vary over time, so some
participants may not have engaged in consistent amounts of PA
between the baseline and the period at the end of the follow-up. A
preliminary evaluation of the personal variability in PA levels
suggests that 61.4% of the study population showed changes in
levels of activity during the follow-up period.
We recognize that depressive symptoms and episodes vary in
length and are not constant; however, the present study design does
not provide information for periods between the baseline and Year 6.
Nonetheless, this lack of information does not change the fact that
subjects who reported diminished PA levels at the Year 6 evaluation
were found to have a high risk of developing symptoms of depression.
Another study limitation is the use of a self-administered questionnaire to measure PA, which may have over- or underestimated
the amount of participants PA. However, such questionnaires are
an accepted form of measurement and are widely used in epidemiological studies (Chasan-Taber et al., 1996; Wolf et al., 1994;
Martnez-Gonzlez et al., 2005) The use of a self-administered
instrument, like the CES-D, to measure depressive symptoms also
offers the added benefit of greater privacy, which may help ensure
more accurate data collection involving socially sensitive topics
like depressive symptomatology. Finally, although use of a clinical
evaluation is the criterion standard for assessing depressive symptoms, the CES-D is a well-accepted instrument that has been
validated and widely used (Caraveo et al., 1994; Salgado de
Snyder & Maldonado, 1994; Benjet et al., 1999).
Despite these limitations, this is the first longitudinal study of an
urban, adult Mexican population to address the important public
health issue of the protective effect of PA against depression.
Prevention is crucial in populations in which mental illnesses like
depression are a widespread public health problem. Promoting
healthy lifestyles that include high levels of PA has been confirmed to be an excellent way to address this public health challenge. Public health policies must also account for the specificities
of particular social and cultural contexts. Thus, the present findings suggest that health promotion programs for Mexican workers
should encourage participants to increase their PA. This intervention could help reduce the enormous social and economic burden
of depressive disorders facing Mexico.

References
Acosta-Czares, B., Aranda-Alvarez, J. G., & Reyes-Morales, H. (2006).
ENCOPREVENIMSS 2004. Patrones de actividad fsica de la mujer y
del hombre. Revista Mdica del Instituto Mexicano del Seguro Social,
44, S79 S86.
Ainsworth, B. E., Haskell, W. L., Whitt, M. C., Irwin, M. L., Swartz,
A. M., Strath, S. J., Leon, A. S. (2000). Compendium of physical
activities: An update of activity codes and MET intensities. Medicine
and Science in Sports & Exercise, 32, S498 S516.
Beekman, A. T., Kriegsman, D. M., Deeg, D. J., & van Tilburg, W. (1995).
The association of physical health and depressive symptoms in the older

population: Age and sex differences. Social Psychiatry and Psychiatry

Epidemiology, 30, 3238.
Benjet, C., Guzman, L., Quintanilla, G., Roque, A., & Leon, R. (1999).
Validez y confiabilidad de la Escala CES-D en peri pberes. Revista
Mexicana de Psicologa, 16, 175185.
Brosse, A. L., Sheets, E. S., Lett, H. S., & Blumenthal, J. A. (2002).
Exercise and the treatment of clinical depression in adults. Sports Medicine, 32, 741760.
Brown, W. J., Ford, J., Burton, N. W., Marshall, A. L., & Dobson, A.
(2005). Prospective study of physical activity and depressive symptoms
in mid age women. American Journal of Preventive Medicine, 29,
265272.
Camacho, T. C., Roberts, R. E., Lazarus, N. B., Kaplan, G. A., & Cohen,
R. D. (1991). Physical activity and depression: Evidence from the
Alameda County Study. American Journal of Epidemiology, 134, 220
231.
Caraveo, J., Medina-Mora, M. E., Villatoro, J., & Rascon, M. L. (1994). La
depresin en el adulto mayor como factor de riesgo en la salud mental
de los nios. Salud Mental, 17, 56 60.
Chan, S. W., Chiu, H. F., Chien, W. T., Goggins, W., Thompson, D., &
Hong, B. (2009). Predictors of change in health-related quality of life
among older people with depression: A longitudinal study. International
Psychogeriatrics, 21, 11711179.
Chasan-Taber, L., Erickson, J. B., McBride, J. W., Nasca, P. C., ChasanTaber, S., & Freedson, P. S. (2002). Reproducibility of a selfadministered lifetime physical activity questionnaire among female college alumnae. American Journal of Epidemiology, 155, 282289.
Chasan-Taber, S., Rimm, E. B., Stampfer, M. J., Spiegelman, D., Colditz,
G. A., Giovannucci, E., . . . Willett, W. C. (1996). Reproducibility and
validity of a self-administered physical activity questionnaire for male
health professionals. Epidemiology, 7, 81 86.
Cooper-Patrick, L., Ford, D. E., Mead, L. A., Chang, P. P., & Klag, M. J.
(1997). Exercise and depression in midlife: A prospective study. American Journal of Public Health, 87, 670 673.
Craft, L. L., & Perna, F. M. (2004). The benefits of exercise for the
clinically depressed. Journal of Clinical Psychiatry, 6, 104 113.
Dinas, P. C., Koutedakis, Y., & Flouris, A. D. (2010). Effects of exercise
and physical activity on depression. Irish Journal of Medical Sciences,
180, 319325.
Farmer, M. E., Locke, B. Z., Moscicki, E. K., Dannenberg, A. L., Larson,
D. B., & Radloff, L. S. (1988). Physical activity and depressive symptoms: The NHANES I Epidemiologic Follow-up Study. American Journal of Epidemiology, 128, 1340 1351.
Gallegos-Carrillo, K., Garca-Pea, C., Mudgal, J., Romero, X., DurnArenas, L., & Salmern, J. (2009). Role of depressive symptoms and
comorbid chronic disease on health-related quality of life among community dwelling older adults. Journal of Psychosomatic Research, 66,
127135.
Geerlings, S. W., Beekman, A. T., Deeg, D. J., Twisk, J. W., & van
Tilburg, W. (2001). The longitudinal effect of depression on functional
limitations and disability in older adults: An eight-wave prospective
community-based study. Psychological Medicine, 31, 13611371.
Goldfarb, A. H., & Jamurtas, A. Z. (1997). Beta-endorphin response to
exercise. An update. Sports Medicine, 24, 8 16.
Hays, R. D., Wells, K. B., Sherbourne, C. D., Rogers, W., & Spritzer, K.
(1995). Functioning and well-being outcomes of patients with depression compared with chronic general medical illnesses. Archives of General Psychiatry, 52, 1119.
Martnez-Gonzlez, M. A., Lpez-Fontana, C., Varo, J. J., SnchezVillegas, A., & Martinez, J. A. (2005). Validation of the Spanish version
of the physical activity questionnaire used in the Nurses Health Study
and the Health Professionals Follow-up Study. Public Health Nutrition,
8, 920 927.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

PHYSICAL ACTIVITY AND REDUCED RISK OF DEPRESSION

McAuley, E., & Blissmer, B. (2000). Self-efficacy determinants and consequences of physical activity. Exercise and Sport Science Reviews, 28,
85 88.
Mead, G. E., Morley, W., Campbell, P., Greig, C. A., McMurdo, M., &
Lawlor, D. A. (2008). Exercise for depression. Cochrane Database of
Systematic Reviews, 4, CD004366.
Medina-Mora, M. E., Borges, G., Lara, C., Benjet, C., Blanco, J., Fleiz, C.,
. . . Zambrano, J. (2005). Prevalence, service use, and demographic
correlates of 12-month DSMIV psychiatric disorders in Mexico: Results
from the Mexican National Comorbidity Survey. Psychological Medicine, 35, 17731783.
Meeusen, R. (2005). Exercise and the brain: Insight in new therapeutic
modalities. Annals of Transplantation, 10, 49 51.
Mndez-Hernndez, P., Flores, Y., Siani, C., Lamure, M., DosamantesCarrasco, L. D., Halley-Castillo, E., & Salmern, J. (2009). Physical
activity and risk of metabolic syndrome in an urban Mexican cohort.
BMC Public Health, 9, 276.
Mudgal, J., Borges, G., Daz-Montiel, J. C., Flores, Y., & Salmern, J.
(2006). Depression among health workers: The role of social characteristics, work stress, and chronic diseases. Salud Mental, 29, 1 8.
Murray, C., & Lpez, A. (1997). Alternative projections of mortality and
disability by cause, 19902020. Global Burden of Disease Study. Lancet, 349, 1498 1504.
Onyike, C., Crum, R., Lee, H., Lyketsos, C., & Eaton, W. (2003). Is obesity
associated with major depression? Results from the Third National
Health and Nutrition Examination Survey. American Journal of Epidemiology, 158, 1139 1147.
Paffenbarger , R. S., Jr. Lee, I. M., & Leung, R. (1994). Physical activity
and personal characteristics associated with depression and suicide in
American College men. Acta Psychiatrica Scandinavica, 377, 16 22.
Penedo, F. J., & Dahn, J. R. (2005). Exercise and well-being: A review of
mental and physical benefits associated with physical activity. Current
Opinion in Psychiatry, 18, 189 193.
Plackett, R. L. (1983). Karl Pearson and the chi-squared test. International
Statistical Review, 51, 59 72.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for
research in the general population. Applied Psychological Measurement,
1, 385 401.

615

Salgado de Snyder, V. N., & Maldonado, M. (1994). Caractersticas

psicomtricas de la Escala de Depresin del Centro de Estudios Epidemiolgicos en mujeres mexicanas adultas de reas rurales. Salud Pblica
de Mexico, 36, 200 209.
Salmern-Castro, J., Arillo-Santilln, E., Campuzano-Rincn, J. C., LpezAntuano, F. J., & Lazcano-Ponce, E. C. (2002). Smoking among health
professionals of the Mexican Social Security Institute. Salud Pblica de
Mexico, 44, S67S75.
Saxena, S., Ommeren, M. V., Tang, K. C., & Armstrong, T. P. (2005).
Mental health benefits of physical activity. Journal of Mental Health, 15,
445 451.
Stata (Version 10) [Computer software]. College Station, TX: StataCorp.
Strawbridge, W. J., Deleger, S., Roberts, R. E., & Kaplan, G. A. (2002).
Physical activity reduces the risk of subsequent depression for older
adults. American Journal of Epidemiology, 156, 328 334.
Warburton, D. E., Nicol, C. W., & Bredin, S. S. (2006). Health benefits of
physical activity: The evidence. Canadian Medical Association Journal,
174, 801 809.
Weyerer, S. (1992). Physical inactivity and depression in the community:
Evidence from the Upper Bavarian Field Study. International Journal of
Sports Medicine, 13, 492 496.
Wolf, A. M., Hunter, D. J., Colditz, G. A., Manson, J. E., Stampfer, M. J.,
Corsano, K. A., . . . Willett, W. C. (1994). Reproducibility and validity
of a self-administered physical activity questionnaire. International
Journal of Epidemiology, 23, 991999.
World Health Organization. (1995). Physical status: The use and interpretation of anthropometry. Report of a WHO Expert Committee (Technical
Report No. Series 854): Geneva, Switzerland: Author.
World Health Organization. (2010). Global recommendations on physical activity
for health. Geneva, Switzerland: World Health Organization. Retrieved from
http://whqlibdoc.who.int/publications/2010/9789241599979_eng.pdf

Received July 5, 2011

Revision received February 29, 2012
Accepted May 14, 2012

E-Mail Notification of Your Latest Issue Online!

Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!