Experimental Gerontology 112 (2018) 68–75

Contents lists available at ScienceDirect

Experimental Gerontology
journal homepage: www.elsevier.com/locate/expgero

Associations of physical activity and depression: Results from the Irish T

Longitudinal Study on Ageing
⁎
C.P. McDowella, , R.K. Dishmanb, M. Hallgrenc, C. MacDonnchaa,d, M.P. Herringa,d
a
Department of Physical Education and Sport Sciences, University of Limerick, Limerick, Ireland
b
Department of Kinesiology, The University of Georgia, Athens, GA, United States
c
Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden
d
Health Research Institute, University of Limerick, Limerick, Ireland

A R T I C LE I N FO A B S T R A C T

Keywords: Physical activity (PA) can protect against depression, but few studies have assessed whether meeting PA
Prospective guidelines is sufficient, or if benefits can be derived from greater volumes of PA. The present study examines
Cross-sectional cross-sectional and prospective associations between different volumes of moderate-to-vigorous PA (MVPA) and
Elderly walking, and depressive symptoms and status. Participants (n = 4556; 56.7% female) aged ≥ 50 years com-
Depression
pleted the International PA Questionnaire (IPAQ) at baseline and the Center for Epidemiological Studies
Physical activity
Depression Scale at baseline and two years later. Prevalence and incidence of depression were 9.0% (n = 410)
and 5.0% (n = 207), respectively. After full adjustment, odds of prevalent depression were: 40% (OR = 0.60,
95%CI: 0.48–0.76) lower among those meeting PA guidelines; 23% (OR = 0.77, 0.49–1.21) and 43%
(OR = 0.57, 0.45–0.73) lower among those in moderate and high categories, respectively; and, 22% (OR = 0.78,
0.61–1.01) and 44.0% (OR = 0.56, 0.42–0.74) lower among those in moderate and high walking tertiles, re-
spectively. Odds of incident depression were: 23% (OR = 0.77, 0.58–1.04) lower among those meeting PA
guidelines; 37% (OR = 0.63, 0.32–1.22) and 20.0% (OR = 0.80, 0.59–1.09) lower among those in moderate and
high categories, respectively; and, 21% (OR = 0.79, 0.56–1.12) and 25% (OR = 0.75, 0.52–1.07) lower among
those in moderate and high walking tertiles, respectively. Moderate and high volumes of MVPA were sig-
nificantly associated with lower odds of concurrent depression, and significantly and non-significantly associated
with reduced odds of incident depression, respectively. Meeting recommended levels of MVPA and walking were
associated with significantly lower odds of concurrent depression, and non-significantly reduced odds of the
development of depression over two years.

1. Introduction 2006). As health services continue to adapt and respond to the im-
plications of an ageing population within the European Union, research
Depression and depressive symptoms are recognized by the World into health promotion and disease prevention in older populations is
Health Organization (WHO) as one of the greatest contributors to required (Rechel et al., 2013).
overall global disease burden, affecting over an estimated 300 million Despite this, evidence has shown that many people with mood
people worldwide (WHO, 2017). The 12-month prevalence of major disorders do not seek treatment (Wang et al., 2005b). Studies show that
depressive disorder (MDD) is highest among those aged 18–64 years, among those that do, delays can range from six to eight years (Wang
but approximately 2.3% of people aged ≥ 65 years experience MDD in et al., 2005a), many do not have their needs met by evidence-based
a 12-month period (Kessler et al., 2012). Among older adults depression treatment (Forsell, 2006), and depression can persist in approximately
is associated with increased risk of frailty (Soysal et al., 2017), mor- 67% after first-line treatment (Trivedi et al., 2006) and at least 30% can
bidity, and suicide (Chapman and Perry, 2008), and decreased physical, remain depressed even after four rounds of distinct treatments (Rush
social, and cognitive functioning (Blazer, 2003). Moreover, the fi- et al., 2006). A recent umbrella review examined potential environ-
nancial burden associated with depression continues to grow, ac- mental risk factors for depression, finding that, despite the large
counting for at least 1% of the total European economy (Sobocki et al., number of putative risk factors investigated in the literature, few

⁎
Corresponding author at: PESS 1039, Department of Physical Education and Sport Sciences, University of Limerick, Limerick, Ireland.
E-mail addresses: cillian.mcdowell@ul.ie (C.P. McDowell), rdishman@uga.edu (R.K. Dishman), mats.hallgren@ki.se (M. Hallgren),
Ciaran.MacDonncha@ul.ie (C. MacDonncha), Matthew.Herring@ul.ie (M.P. Herring).

https://doi.org/10.1016/j.exger.2018.09.004
Received 2 April 2018; Received in revised form 13 August 2018; Accepted 10 September 2018
Available online 12 September 2018
0531-5565/ © 2018 Elsevier Inc. All rights reserved.
C.P. McDowell et al. Experimental Gerontology 112 (2018) 68–75

associations are supported by robust evidence (Köhler et al., 2018). The were excluded from longitudinal analyses, leaving a sample of 4146.
factors with the most convincing evidence were largely stressors, such
as children exposed to physical abuse, military personnel exposed to 2.2. Study measures
combat, recently widowed adults, obese individuals, and people with
4–5 five metabolic risk factors. Thus, there is a critical need to identify 2.2.1. Physical activity
both risk factors for depression and the healthful behaviours that may PA was measured using the short-form International Physical
attenuate those risk factors to better inform interventions for preven- Activity Questionnaire (IPAQ-SF) (Craig et al., 2003). Respondents who
tion and treatment. reported walking or MVPA greater than a combined 16 h/day were
A recent meta-analysis reviewed twenty-two prospective cohort excluded (n = 11). The remaining respondents were classified as
studies in older adults that largely support the protective effects of PA meeting PA guidelines (i.e., reporting ≥150 min weekly of MVPA,
on depression, with all except for one (García-Peña et al., 2013) re- ≥75 min weekly of vigorous PA, ≥600 MET.minutes weekly). Three
porting reductions in incident depression (Schuch et al., 2018). How- dose categories, Low (0 to < 600 MET.minutes weekly), Moderate (600
ever, just six studies used validated instruments to measure PA dose to < 1200 MET.minutes weekly), or High (≥1200 MET.minutes
(i.e., a product of PA duration and intensity; Groffen et al., 2013; Joshi weekly) (WHO, 2010). Weekly minutes of walking were divided into
et al., 2016; Park et al., 2015; Pasco et al., 2011; Rius-Ottenheim et al., tertiles (i.e., 0–110 min, 120–400 min, 420+ minutes).
2013; Tsutsumimoto et al., 2017). Available evidence suggests that
meeting WHO recommended PA levels (WHO, 2010) may protect 2.2.2. Depression
against other mental health outcomes such as generalized anxiety dis- At Waves 1 and 2, depression symptoms were assessed using the
order which is highly comorbid with depression (McDowell et al., Center for Epidemiological Studies Depression Scale (CES-D) (Radloff,
2018b) and convey additional benefit for depressive symptoms. For 1977). Reliability coefficients of the CES-D are high (0.85–0.91) among
example, depressive symptom reductions following exercise training older adults (Radloff and Teri, 1986). We used a score of ≥16 to define
were significantly larger among chronically-ill patients who were caseness of depression (Radloff, 1977). Respondents above this cut-off
meeting recommended PA levels (Herring et al., 2012) and recent score are referred to as depressed throughout the current manuscript;
evidence showed significantly larger improvements in depressive however, they are at increased risk of depression but do not have
symptoms following internet-based cognitive behavioural therapy and clinically diagnosed depression. This cut-score demonstrates 100%
usual care among depressed patients engaging in PA levels that corre- sensitivity and 87.6% specificity in older populations (Beekman et al.,
sponded to consensus recommendations for maximizing general health 1997). Positive predictive values can be low (13.2%) and negative
(Hallgren et al., 2016; Herring et al., 2012). Further, people with major predictive values can be high (99–100%) (Beekman et al., 1997).
depression are less likely to meet recommended levels of PA
(Vancampfort et al., 2017) which is a particular worry given that de- 2.2.3. Covariates
pression is associated with increased risk of cardiovascular disease Age was divided into four categories (i.e., 50–59 years.,
(CVD) and death from CVD (Correll et al., 2017) and the well-estab- 60–69 years., 70–79 years., and 80+ years). Waist circumference was
lished role of PA in protecting against CVD. Despite this, of the six classified as low- or increased-risk according to WHO guidelines (i.e.,
prospective-cohort studies that measured PA dose, none examined the Males: > 94 cm; Females: > 80 cm) (WHO, 2000; Grundy et al., 2005).
benefits of meeting WHO guidelines. This information is important for Social class was defined according to the European Socioeconomic
clinical practice and public health recommendations. Classification (ESeC) scheme (Rose and Harrison, 2007). The ESeC
Therefore, the present study used data from two waves of The Irish classifies people into seven categories according to their positions
Longitudinal Study of Ageing (TILDA) to examine associations between: within labour markets and with special attention to their employment
1) meeting PA guidelines, 2) different volumes of MVPA, and 3) weekly relations. In order to improve sample coverage, those who were not in
minutes spent walking, and prevalent and incident depressive symp- paid employment were allocated to a “Not Working” group (n = 1254).
toms and status, both within age and sex categories, and across the total Smoking status was assessed by self-reported current or never/previous
population. smoker. Participants reported whether a doctor had ever told them that
they have angina, asthma, cancer/a malignant tumour, diabetes/high
2. Methods blood sugar, hypertension, osteoporosis, or a stroke. The number of
these comorbidities was summed.
This study used STROBE recommendations to guide reporting (Von
Elm et al., 2008). 2.3. Statistical analysis

2.1. Study population Statistical analyses were conducted using SPSS Version 22.0
(Armonk, NY: IBM Corp.). Chi-square tests examined differences in
TILDA is an ongoing cohort study that contains a nationally re- meeting PA guidelines, depression, sex, age, waist circumference, social
presentative sample of community dwelling adults aged ≥ 50 years, class, and smoking categories between those included and excluded
and their partners of any age, resident in the Republic of Ireland. An from analyses (i.e., those with complete IPAQ-SF (Wave 1) and CES-D
initial multi-stage probability sample of addresses was chosen by means (Waves 1 and 2) data but missing covariate data). Chi-square tests ex-
of the RANSAM sampling procedure (Whelan, 1979), with District amined differences in sex, age, waist circumference, social class, and
Electoral Divisions selected at the first stage and household addresses smoking categories between those with and without depression. For
selected at the second stage. The response rate was 62.0%. Participants significant Chi-square tests, Z tests were calculated for column pro-
gave full informed consent to participate in the study and ethical ap- portions for each row in the Chi-square contingency table and adjusted
proval was obtained from the Trinity College Dublin Faculty of Health using a Bonferroni correction (Sharpe, 2015).
Sciences Research Ethics Committee. For the present analyses we report
data from Wave 1 (2009–2010) (Barrett et al., 2011) and Wave 2 2.3.1. Cross-sectional and longitudinal walking, PA guidelines, and dose
(2012−2013) (Nolan et al., 2014). PA and depression data at Waves 1 categories analyses
and 2 were available for 6977 respondents. Following exclusion of One-way ANOVAs and Fisher's LSD planned contrasts quantified
participants with incomplete covariate data (n = 2416), the final differences in Waves 1 and 2 depressive symptoms between those
sample size for the current study consisted of 4556 individuals (56.7% meeting and not meeting PA guidelines, dose categories, and walking
female). Respondents who reported depression at baseline (n = 415) tertiles. Hedges' g effect sizes and associated 95% confidence intervals

69
C.P. McDowell et al. Experimental Gerontology 112 (2018) 68–75

(95%CI) were calculated to quantify the magnitude of differences in was 9.0% (n = 410), and the incidence rate was 5.0% (n = 207).
depressive scores (Cumming, 2014).
Binomial logistic regression quantified associations (i.e., odds ratios 3.2. Cross-sectional associations between meeting PA guidelines and
(ORs)) between meeting PA guidelines, dose categories, and walking depression
tertiles, and depression status at Waves 1 and 2 within age and sex
categories, and the total population. Covariates in adjusted models Depressive symptoms were significantly higher among people not
were age and sex (Model 1) and age, sex, waist circumference, social meeting PA guidelines than those meeting PA guidelines
class, and smoking status (Model 2). The Hosmer-Lemeshow test was (F(1,4554) = 74.83, p < 0.001; g = 0.26; 95%CI: 0.20 to 0.31).
performed and Nagelkerke R2 calculated to assess the goodness-of-fit of In Model 2, meeting PA guidelines was significantly associated with
the model. Likelihood ratio tests examined covariate significance. 39.7% (95%CI: 24.3 to 52.1; p < 0.001) reduced odds of depression
Supplementary tables 1 and 2 provide the proportion of people with (Table 2). The model fit the data well (χ2(8) = 11.78, p = 0.16;
depression by levels of PA and walking and significant covariates at R2 = 0.16). Age, sex, social class, smoking status, and number of co-
Waves 1 and 2 respectively. morbidities were significant covariates (all p < 0.03).

3.3. Cross-sectional associations between dose categories and depression

3. Results
Planned contrasts showed significantly lower depressive symptoms
3.1. Participant characteristics
for the High category compared to Low (p < 0.001; g = 0.27: 0.21 to
0.33) and Moderate (p = 0.042; g = 0.14: 0.02 to 0.26), and sig-
Compared to the analytic sample, a significantly greater proportion
nificantly lower depressive symptoms for the Moderate category com-
of excluded respondents were aged 80+ years (7.5% vs 5.6%), male
pared to Low (p = 0.013; g = 0.14: 0.12 to 0.26).
(46.9% vs 43.3%), and current smokers (20.7% vs 15.3%; all
In Model 2, the Moderate category was non-significantly, and High
p < 0.05). There were no significant differences in waist cir-
category significantly, associated with 33.1% (−20.7% to 51.0%;
cumference, depression, or PA. Wave 1 participant characteristics are
p = 0.25) and 42.6% (26.7% to 45.0%; p < 0.001) lower odds of de-
presented in Table 1. Sex (p < 0.01), age (p < 0.001), social class
pression, respectively (Table 2). The model fit the data well
(p < 0.001), and smoking status (p < 0.001) significantly differed
(χ2(8) = 11.12, p = 0.20; R2 = 0.16). Age, sex, social class, smoking
according to depression, and waist circumference did not (p > 0.64).
status, and number of comorbidities were significant covariates (all
Briefly, a greater proportion of respondents with depression were aged
p < 0.02).
80+ years (5.9% vs 2.9%), female (70.2% vs 55.4%) and current
smokers (27.8% vs 14.1%; all p < 0.05). The prevalence of depression
3.4. Cross-sectional associations between walking and depression
Table 1
Planned contrasts showed significantly lower depression for the
Baseline participant characteristics.
high tertile compared to the low (p < 0.001; g = 0.24: 0.17 to 0.32)
No Depression Probable Depression and moderate (p = 0.01; g = 0.10: 0.03 to 0.16) tertiles, and sig-
(n = 4146) (n = 410)
nificantly lower depressive symptoms for the moderate compared to the
(n (%) or mean ± SD) (n (%) or mean ± SD)
low (p < 0.001; g = 0.15: 0.08 to 0.22) tertile.
CES-D score 3.97 ± 3.96 23.23 ± 7.29 In Model 2, moderate walking was non-significantly, and high
Walking (hrs wk−1) 6.32 ± 7.48 4.78 ± 6.93 walking significantly, associated with 21.8% (−0.6% to 39.2%;
MVPA (hrs wk−1) 8.13 ± 13.20 9.52 ± 4.44
p = 0.056) and 44.0% (26.1% to 57.6%; p < 0.001) lower odds of
Meeting PA guidelines
Yes 2063b (50.2) 134a (32.7) depression, respectively (Table 2). The model fit the data well
No 2083b (49.8) 276a (67.3) (χ2(8) = 11.30, p = 0.19; R2 = 0.15). Age, sex, social class, smoking
Sex status, and number of comorbidities were significant covariates (all
Female 1851b (44.6) 288a (70.2) p < 0.03).
Male 2295b (55.4) 122a (29.8)
Age category (yrs)
50–59 1640b (39.6) 194a (47.3) 3.5. Longitudinal associations between meeting PA guidelines and
60–69 1438a (34.7) 133a (32.4) depression
70–79 825a (19.9) 71a (17.3)
80+ 243b (5.9) 12a (2.9)
Depressive symptoms were significantly higher among people not
Waist circumference
Normal 1993 (48.1) 202 (49.3) meeting PA guidelines than those meeting PA guidelines
At risk 2153 (51.9) 208 (50.7) (F(1,4144) = 17.46, p < 0.001; g = 0.13: 0.07 to 0.19).
Social class In Model 2, meeting PA guidelines was non-significantly associated
Professional workers 176b (4.2) 6a (4.0) with 22.6% (−3.9% to 42.4%; p = 0.09) reduced odds of depression
Managerial and 1079b (26.0) 55a (13.4)
technical
(Table 3). The model fit the data well (χ2(8) = 2.31, p = 0.97;
Non-manual 629b (15.2) 43a (14.7) R2 = 0.06). Sex, social class, smoking, and number of comorbidities
Skilled manual 412a (9.9) 33a (9.8) were significant covariates (all p < 0.03).
Semi-skilled 409a (9.9) 44a (9.9)
Unskilled 116a (2.8) 18a (2.9)
3.6. Longitudinal associations between dose categories and depression
Farmers 278b (6.7) 8a (6.3)
Not working 1047b (25.3) 203a (27.4)
Smoking status Planned contrasts showed significantly lower depressive symptoms
Current 585b (14.1) 296a (72.2) for the High category compared to Low (p < 0.001; g = 0.13: 0.07 to
Past/never 3561b (85.9) 114a (27.8) 0.19) but not Moderate (p = 0.85; g = 0.12: −0.01 to 0.24), and non-
Number of comorbidities 0.47 ± 0.73 1.06 ± 1.07
significantly lower depressive symptoms for Moderate compared to Low
Different subscript letters indicate a subset of each category whose column (p = 0.058; g = 0.01: −0.11 to 0.14).
proportions differ statistically significantly at the 0.05 level. In Model 2, the Moderate and High categories were non-sig-
Abbreviations: hrs: hours; mins: minutes; MVPA: moderate-to-vigorous physical nificantly associated with 37.3% (−21.8% to 67.7%; p = 0.17) and
activity; wk: week; yrs: years. 20.0% (−8.8% to 41.2%; p = 0.16) reduced odds of depression,

70
 C.P. McDowell et al.

Table 2
Odds ratios (OR) and 95% confidence intervals (CI) derived from binominal logistic regression analyses as indicators of cross-sectional associations between physical activity (PA) and depression within age and sex
categories and the total population.
Model 1 Model 2

Age Sex Total population Age Sex Total population

50–59 60–69 70–79 80+ Male Female 50–59 60–69 70–79 80+ Male Female
Meeting PA
Guidelines
No REF REF REF REF REF REF REF REF REF REF REF REF REF REF
Yes 0.68 (0.50 0.37 (0.25 0.34 (0.18 to 0.22 (0.03 0.31 (0.21 0.63 (0.48 to 0.50 (0.40 to 0.86 (0.63 0.45 (0.30 0.35 (0.18 0.27 (0.03 0.39 (0.26 0.75 (0.57 to 0.60 (0.48 to
to 0.93)a to 0.55)c 0.673)c to 1.70) to 0.46)c 0.81)c 0.62)c to 1.19) to 0.68)c to 0.68)b to 2.18) to 0.59)c 0.98)a 0.76)c
Dose Categories
Low REF REF REF REF REF REF REF REF REF REF REF REF REF REF
Moderate 0.77 (0.42 0.48 (0.22 0.66 (0.23 to – 0.46 (0.19 0.69 (0.41 to 0.61 (0.40 to 0.95 (0.50 0.57 (0.25 0.84 (0.28 – 0.55 (0.23 0.87 (0.51 to 0.77 (0.49 to

71
to 1.42) to 1.06) 1.89) to 1.07) 1.15)c 0.95)a to 1.81) to 1.30) to 2.51) to 1.34) 1.47) 1.21)
High 0.67 (0.49 0.36 (0.23 0.28 (0.14 to 0.29 (0.04 0.29 (0.19 0.61 (0.46 to 0.48 (0.38 to 0.85 (0.60 0.43 (0.28 0.28 (0.13 0.36 (0.04 0.37 (0.24 0.72 (0.54 to 0.57 (0.45 to
to 0.92)a to 0.54)c 0.57)c to 2.28) to 0.45)c 0.81)c 0.61)c to 1.19) to 0.67)c to 0.59)c to 3.00) to 0.57)c 0.97)a 0.73)c
Walking
Low REF REF REF REF REF REF REF REF REF REF REF REF REF REF
Moderate 0.75 (0.53 0.65 (0.43 0.57 (0.32 to 0.52 (0.13 0.71 (0.46 0.66 (0.50 to 0.68 (0.53 to 0.83 (0.57 0.78 (0.50 0.64 (0.35 0.60 (0.14 0.80 (0.51 0.75 (0.56 to 0.78 (0.61 to
to 1.08) to 0.98)a 1.04) to 2.02) to 1.09) 0.88)b 0.86)b to 1.21) to 1.20) to 1.21) to 2.56) to 1.27) 1.02) 1.01)
High 0.57 (0.39 0.41 (0.26 0.59 (0.32 to 0.27 (0.03 0.42 (0.26 0.55 (0.40 to 0.50 (0.39 to 0.62 (0.41 0.46 (0.28 0.60 (0.32 0.31 (0.04 0.47 (0.29 0.60 (0.43 to 0.56 (0.42 to
to 0.84)b to 0.66)c 1.07) to 2.19) to 0.67)c 0.76)c 0.66)c to 0.93)a to 0.76)b to 1.14) to 2.69) to 0.78)b 0.84)b 0.74)c

Model 1 adjusted for age and/or sex.

Model 2 adjusted for Model 1 and waist circumference, social class, smoking status, and number of comorbidities.
Abbreviations: REF: reference category.
a
p < 0.05.
b
p < 0.01.
c
p < 0.001.
Experimental Gerontology 112 (2018) 68–75
 C.P. McDowell et al.

Table 3
Odds ratios (OR) and 95% confidence intervals (CI) derived from binominal logistic regression analyses as indicators of prospective associations between physical activity (PA) and depression within age and sex
categories and the total population.
Model 1 Model 2

Age (years) Sex Total population Age (years) Sex Total population

50–59 60–69 70–79 80+ Male Female 50–59 60–69 70–79 80+ Male Female
Meeting PA
Guidelines
No REF REF REF REF REF REF REF REF REF REF REF REF REF REF
Yes 0.80 (0.53 0.55 (0.32 0.83 (0.40 0.56 (0.12 0.61 (0.38 0.78 (0.54 to 0.71 (0.53 to 0.90 (0.59 0.61 (0.35 0.79 (0.38 0.45 (0.09 0.67 (0.41 0.84 (0.58 to 0.77 (0.58 to
to 1.21) to 0.95)a to 1.70) to 2.72) to 0.98)a 1.12) 0.95)a to 1.37) to 1.05) to 1.65) to 2.26) to 1.10) 1.22) 1.04)
Dose Categories
Low REF REF REF REF REF REF REF REF REF REF REF REF REF REF
Moderate 0.60 (0.23 0.53 (0.16 0.85 (0.19 0.51 (0.15 0.62 (0.28 to 0.57 (0.30 to 0.68 (0.26 0.56 (0.17 0.93 (0.21 0.53 (0.16 0.69 (0.31 to 0.63 (0.32 to

72
– –
to 1.52) to 1.77) to 3.72) to 1.67) 1.37) 1.10) to 1.76) to 1.89) to 4.15) to 1.79) 1.54) 1.22)
High 0.84 (0.55 0.55 (0.31 0.82 (0.38 0.73 (0.15 0.62 (0.38 0.81 (0.56 to 0.73 (0.54 to 0.94 (0.61 0.61 (0.34 0.77 (0.35 0.55 (0.11 0.69 (0.42 0.87 (0.59 to 0.80 (0.59 to
to 1.27) to 0.98)a to 1.77) to 3.56) to 1.02) 1.19) 0.99)a to 1.44) to 1.09) to 1.68) to 2.88) to 1.79) 1.28) 1.09)
Walking
Low REF REF REF REF REF REF REF REF REF REF REF REF REF REF
Moderate 0.92 (0.55 0.64 (0.35 0.59 (0.26 0.40 (0.08 0.80 (0.45 0.68 (0.45 to 0.72 (0.51 to 0.96 (0.57 0.72 (0.39 0.63 (0.27 0.33 (0.06 0.88 (0.49 0.74 (0.48 to 0.79 (0.56 to
to 1.54) to 1.15) to 1.35) to 2.02) to 1.41) 1.04) 1.01) to 1.63) to 1.34) to 1.47) to 1.79) to 1.59) 1.14) 1.12)
High 1.05 (0.63 0.38 (0.19 0.71 (0.32 0.66 (0.13 0.57 (0.31 0.81 (0.53 to 0.70 (0.50 to 1.10 (0.65 0.41 (0.20 0.75 (0.32 0.50 (0.09 0.63 (0.34 0.84 (0.54 to 0.75 (0.52 to
to 1.76) to 0.76)b to 1.60) to 3.40) to 1.03) 1.25) 1.00)a to 1.86) to 0.84)a to 1.73) to 2.84) to 1.16) 1.30) 1.07)

Model 1 adjusted for age and/or sex.

Model 2 adjusted for Model 1 and waist circumference, social class, smoking status, and number of comorbidities.
Abbreviations: REF: reference category.
a
p < 0.05.
b
p < 0.01.
Experimental Gerontology 112 (2018) 68–75
C.P. McDowell et al. Experimental Gerontology 112 (2018) 68–75

respectively (Table 3). The model fit the data well (χ2(8) = 3.46, hours of MVPA rather than MVPA volume. Collectively, these findings
p = 0.90; R2 = 0.06). Sex, social class, smoking, and number of co- appear to support a potential lower threshold volume for similar ben-
morbidities were significant covariates (all p < 0.02). efits among males, and a larger protective effect for moderate volumes
of MVPA among females.
3.7. Longitudinal associations between walking and depression A non-significant protective effect of walking on the development of
subsequent depression was observed in the present study. Among the
Planned contrasts showed significantly lower depressive symptoms total population there appeared to be a threshold effect, with similar
for the high (p < 0.001; g = 0.15: 0.07 to 0.23) and moderate effects for those in the moderate (21%) and high (25%) tertiles com-
(p = 0.03; g = 0.08: 0.01 to 0.16) tertiles compared to the low, and pared to the low tertile. However, when examined within sexes, the
non-significantly lower depressive symptoms for the high compared to high tertile (37%) exhibited greater effects than the moderate (12%)
the moderate (p = 0.08; g = 0.06: −0.01 to 0.14) tertile. among males, while the opposite was observed among females (16% vs
In Model 2, moderate and high walking were non-significantly as- 26%). Similarly, those in the oldest age category appear to derive the
sociated with 20.7% (−12.2% to 43.9%; p = 0.19) and 25.1% (−7.0% greatest benefits from moderate amounts of walking. The protective
to 47.6%; p = 0.11) lower odds of depression, respectively (Table 3). effect of walking on the development of incident depression has pre-
The model fit the data well (χ2(8) = 6.25, p = 0.62; R2 = 0.06). Sex, viously been reported in The Honolulu-Asia Aging Study (Smith et al.,
social class, smoking status, and number of comorbidities were sig- 2010). However, in this all-male sample, a threshold effect was ob-
nificant covariates (all p < 0.02). served, with only slight differences between those in intermediate- and
high-walking-distance groups. Harvey et al. (2017) also reported that
4. Discussion any level of PA, even low levels, can be protective against depression,
and light PA has been shown to be associated with lower depression
In the current large, nationally-representative cohort of older adults, levels (Loprinzi, 2013) and protective of future cognitive ability in older
meeting PA guidelines was significantly associated with 40% lower adults, irrespective of MVPA levels (Stubbs et al., 2017). If future
odds of prevalent depression, and non-significantly associated with prospective cohort studies or randomized controlled trials corroborate
approximately 23% reduced odds of incident depression. Results did these and the current findings, then promotion of lighter-intensity and
not support a dose-response for MVPA and depression. Rather, a 37% lower volumes of PA among older adults may be more practical, as their
reduction in odds of incident depression was observed for moderate ability and desire to engage in higher intensity PA can be limited by the
volumes of MVPA, and a reduction of 20% was observed for high vo- ageing process.
lumes of MVPA. Furthermore, walking was significantly associated with Although the examination of plausible mechanisms of the associa-
lower odds of prevalent depression, with a larger effect observed with tion between PA and depression was not the purpose of the current
more time spent walking. Associations of walking with prevalent de- investigation, the available evidence has provided some support for
pression were non-significant but indicate a potential 25% reduced several psychological and biological mechanisms that may underlie
odds of incident depression across both levels of walking. To the au- these associations. Briefly, exercise training has brain monoaminergic
thors' knowledge, this is the first prospective cohort study to examine (Dishman, 1997; Greenwood and Fleshner, 2011) and neurotrophic
the protective effect of meeting MVPA guidelines, and different vo- (Dishman et al., 2006a; Firth et al., 2018) effects, and may reduce in-
lumes of walking, on depression among a sample of older adults. flammatory and oxidant markers (Schuch et al., 2016). Further, phy-
The present findings demonstrate the potential benefits of meeting sical activity may be associated with depression through psychological
PA guidelines for reduced risk of depression in addition to previously factors such as self-esteem (Dishman et al., 2006b; Motl et al., 2005).
established reduced risk of mortality (Hupin et al., 2015) and increased
healthy ageing (Daskalopoulou et al., 2017). Older adults who met PA 4.1. Limitations
guidelines were 40% less likely to report concurrent depression and
23% less likely to develop depression two years later. However. despite The main limitation of the current study is the use of self-reported
these known benefits, approximately 50% of older adults in Europe do PA and depressive symptoms, predisposing the results to over-reporting
not meet PA guidelines (Hallal et al., 2012). This is particularly con- or under-reporting. Future studies would benefit from objective PA
cerning given the ageing population within Europe, and the consider- measurement which can indicate both activity duration and intensity,
able burden of depressive illness. The present findings reiterate the such as GPS or accelerometery-based measurements, and more direct
necessity to promote meeting, and where appropriate exceeding, re- measures of depression, such as physician diagnosis. Secondly, despite
commended PA levels at the population-level. the prospective design, causality cannot be inferred from these results.
A dose-response between MVPA and prevalent depression was ob- Undoubtedly, depression can lead to less PA and vice versa.
served, but not between MVPA and incident depression. Compared to Nonetheless, the present study included a large sample size, a PA
inactive older adults, those who engaged in Moderate and High activity measure that measured important components of PA, and direct ex-
were 23% and 43% less likely to report concurrent depression, re- amination of the protective effect of meeting MVPA guidelines, dif-
spectively, and 37% and 20% less likely to develop new depression two fering volumes of MVPA, and walking on depression.
years later, respectively. This difference in prospective associations may
be driven by sex-related differences as, among females, Moderate vo- 4.2. Future recommendations
lumes of MVPA exhibited greater protective effects than High (31% vs
13%), while there was a relatively smaller difference between volumes Given the potential increased practicality of walking among older
among males (47% vs 31%). This supports previous findings in a adults, future prospective cohort studies or randomized controlled trials
smaller sample of elderly Koreans, which found larger effects for mild should corroborate the present findings on the protective effect of
PA than above-moderate, in addition to larger effects overall (Park walking. Secondly, recent evidence has shown that people with MDD
et al., 2015). Similarly, exercise performed at either light, moderate, or engage in higher levels of sedentary behaviour (Schuch et al., 2017)
vigorous intensities have been shown to be equally effective at reducing and that cross-sectionally sedentary behaviour is positively associated
depressive symptoms among people with mild-to-moderate depression with depression (Stubbs et al., 2018). The influence of sedentary be-
(Helgadóttir et al., 2016). Dose-responses have previously been ob- haviour on mental health remains relatively understudied, and research
served among older females from the Nurses' Health Study (Chang to further explore these associations is warranted. Finally, as older
et al., 2016), and adolescents from The Youth and Mental Health Study adults with pain are less physically active than asymptomatic controls
(Sund et al., 2011), although these analyses were based on weekly (Stubbs et al., 2013), and pain is associated with poorer mental health

73
C.P. McDowell et al. Experimental Gerontology 112 (2018) 68–75

(McDowell et al., 2018a; Tian et al., 2018), the potential moderating/ 2008. The Strengthening the Reporting of Observational Studies in Epidemiology
mediating role of pain in the PA-depression relationship warrants ex- [STROBE] statement: guidelines for reporting observational studies. Gac. Sanit. 22,
144–150.
ploration. Firth, J., Stubbs, B., Vancampfort, D., Schuch, F., Lagopoulos, J., Rosenbaum, S., Ward,
P.B., 2018. Effect of aerobic exercise on hippocampal volume in humans: a systematic
4.3. Conclusion review and meta-analysis. NeuroImage 166, 230–238.
Forsell, Y., 2006. The pathway to meeting need for mental health services in Sweden.
Psychiatr. Serv. 57, 114–119.
The present study supports previous research on the protective García-Peña, C., Wagner, F.A., Sánchez-García, S., Espinel-Bermúdez, C., Juárez-Cedillo,
benefits of PA for depression among older adults, and suggests that T., Pérez-Zepeda, M., Arango-Lopera, V., Franco-Marina, F., Ramírez-Aldana, R.,
Gallo, J.J., 2013. Late-life depressive symptoms: prediction models of change. J.
meeting WHO recommended PA levels is could be recommended in Affect. Disord. 150, 886–894.
order to prevent the onset of depression. Although there is no apparent Greenwood, B.N., Fleshner, M., 2011. Exercise, stress resistance, and central serotonergic
additional benefit for depression to exceeding this dose in the present systems 39 (3), 140–149.
Groffen, D.A., Koster, A., Bosma, H., Van Den Akker, M., Kempen, G.I., Van Eijk, J.T., Van
study, additional health benefits to exceeding this dose have previously
Gool, C.H., Penninx, B.W., Harris, T.B., Rubin, S.M., Pahor, M., 2013. Unhealthy
been identified. Future research should focus on associations of de- lifestyles do not mediate the relationship between socioeconomic status and incident
pression and volumes of PA less than PA guidelines, light intensity PA, depressive symptoms: the Health ABC study. Am. J. Geriatr. Psychiatry 21 (7),
and sedentary behaviour. 664–674.
Grundy, S.M., Cleeman, J.I., Daniels, S.R., Donato, K.A., Eckel, R.H., Franklin, B.A.,
Gordon, D.J., Krauss, R.M., Savage, P.J., Smith, S.C., 2005. Diagnosis and manage-
Acknowledgements ment of the metabolic syndrome: an American Heart Association/National Heart,
Lung, and Blood Institute scientific statement. Circulation 112, 2735–2752.
Hallal, P.C., Andersen, L.B., Bull, F.C., Guthold, R., Haskell, W., Ekelund, U., for the
The authors have no financial disclosures. The authors disclose no Lancet Physical Activity Series Working Group, 2012. Global physical activity levels:
conflicts of interest. surveillance progress, pitfalls, and prospects. Lancet 380, 247–257.
This research received no specific grant from any funding agency, Hallgren, M., Nakitanda, O.A., Ekblom, Ö., Herring, M.P., Owen, N., Dunstan, D.,
Helgadottir, B., Forsell, Y., 2016. Habitual physical activity levels predict treatment
commercial or not-for-profit sectors. The TILDA study was supported by outcomes in depressed adults: a prospective cohort study. Prev. Med. 88, 53–58.
the Irish Government, the Atlantic Philanthropies, and Irish Life PLC. Harvey, S.B., Øverland, S., Hatch, S.L., Wessely, S., Mykletun, A., Hotopf, M., 2017.
This work was also supported by the Health Research Board of Ireland Exercise and the prevention of depression: Results of the HUNT cohort study. Am. J.
Psychiatry 175, 28–36 (appi. ajp. 2017.16111223).
(Grant reference: HRA_PHS/2012/30). The sponsors played no role in Helgadóttir, B., Hallgren, M., Ekblom, Ö., Forsell, Y., 2016. Training fast or slow? Exercise
study design, methods, subject recruitment, data collection, analysis or for depression: a randomized controlled trial. Prev. Med. 91, 123–131.
preparation of the paper. Herring, M.P., Puetz, T.W., O'Connor, P.J., Dishman, R.K., 2012. Effect of exercise
training on depressive symptoms among patients with a chronic illness: a systematic
review and meta-analysis of randomized controlled trials. Arch. Intern. Med. 172,
Availability of data and materials 101–111.
Hupin, D., Roche, F., Gremeaux, V., Chatard, J.C., Oriol, M., Gaspoz, J.M., Barthélémy,
Dataset files are publicly accessible and provided by The Irish J.C., Edouard, P., 2015. Even a low-dose of moderate-to-vigorous physical activity
reduces mortality by 22% in adults aged ≥ 60 years: a systematic review and meta-
Longitudinal Study on Ageing. Information on how to access these is analysis. Br. J. Sports Med. 0, 1–8.
available at www.tilda.tcd.ie. Joshi, S., Mooney, S.J., Kennedy, G.J., Benjamin, E.O., Ompad, D., Rundle, A.G., Beard,
J.R., Cerdá, M., 2016. Beyond METs: types of physical activity and depression among
older adults. Age Ageing 45 (1), 103–109.
References Kessler, R.C., Petukhova, M., Sampson, N.A., Zaslavsky, A.M., Wittchen, H.U., 2012.
Twelve-month and lifetime prevalence and lifetime morbid risk of anxiety and mood
Barrett, A., Burke, H., Cronin, H., Hickey, A., Kamiya, Y., Kenny, R.A., Layte, R., Maty, S., disorders in the United States. Int. J. Methods Psychiatr. Res. 21, 169–184.
McGee, H., Morgan, K., 2011. Fifty Plus in Ireland 2011: First Results from The Irish Köhler, C.A., Evangelou, E., Stubbs, B., Solmi, M., Veronese, N., Belbasis, L., Bortolato, B.,
Longitudinal Study on Ageing (TILDA). Available at: https://epubs.rcsi.ie/ Melo, M.C., Coelho, C.A., Fernandes, B.S., Olfson, M., 2018. Mapping risk factors for
psycholrep/45/. depression across the lifespan: an umbrella review of evidence from meta-analyses
Beekman, A., Deeg, D., Van Limbeek, J., Braam, A., De Vries, M., Van Tilburg, W., 1997. and Mendelian randomization studies. J. Psychiatr. Res. https://doi.org/10.1016/j.
Criterion validity of the Center for Epidemiologic Studies Depression scale (CES-D): jpsychires.2018.05.020.
results from a community-based sample of older subjects in the Netherlands. Psychol. Loprinzi, P.D., 2013. Objectively measured light and moderate-to-vigorous physical ac-
Med. 27, 231–235. tivity is associated with lower depression levels among older US adults. Aging Ment.
Blazer, D.G., 2003. Depression in late life: review and commentary. J. Gerontol. A Biol. Health 17, 801–805.
Sci. Med. Sci. 58, M249–M265. McDowell, C., Gordon, B., Andrews, K., MacDonncha, C., Herring, M., 2018a.
Chang, S.C., Pan, A., Kawachi, I., Okereke, O.I., 2016. Risk factors for late-life depression: Associations of physical activity with anxiety symptoms and status: results from The
a prospective cohort study among older women. Prev. Med. 91, 144–151. Irish Longitudinal Study on Ageing. Epidemiol. Psychiatr. Sci. 1–10.
Chapman, D.P., Perry, G.S., 2008. Depression as a major component of public health for McDowell, C.P., Dishman, R.K., Vancampfort, D., Hallgren, M., Stubbs, B., MacDonncha,
older adults. Prev. Chronic Dis. 5, A22. C., Herring, M.P., 2018b. Physical activity and generalized anxiety disorder: results
Correll, C.U., Solmi, M., Veronese, N., Bortolato, B., Rosson, S., Santonastaso, P., Thapa- from The Irish Longitudinal Study on Ageing (TILDA). Int. J. Epidemiol. https://doi.
Chhetri, N., Fornaro, M., Gallicchio, D., Collantoni, E., Pigato, G., 2017. Prevalence, org/10.1093/ije/dyy141.
incidence and mortality from cardiovascular disease in patients with pooled and Motl, R.W., Konopack, J.F., McAuley, E., Elavsky, S., Jerome, G.J., Marquez, D.X., 2005.
specific severe mental illness: a large-scale meta-analysis of 3,211,768 patients and Depressive symptoms among older adults: long-term reduction after a physical ac-
113,383,368 controls. World Psychiatry 16 (2), 163–180. tivity intervention. J. Behav. Med. 1 (28) ((4):385).
Craig, C.L., Marshall, A.L., Sjorstrom, M., Bauman, A.E., Booth, M.L., Ainsworth, B.E., Nolan, A., O'Regan, C., Dooley, C., Wallace, D., Hever, A., Cronin, H., Hudson, E., Kenny,
Pratt, M., Ekelund, U., Yngve, A., Sallis, J.F., 2003. International physical activity R., 2014. The over 50 s in a changing Ireland: economic circumstances, health and
questionnaire: 12-country reliability and validity. Med. Sci. Sports Exerc. 35, well-being. In: Dublin: Trinity College: The Irish Longitudinal Study on Ageing,
1381–1395. Available at: https://www.esri.ie/publications/the-over-50s-in-a-changing-ireland-
Cumming, G., 2014. The new statistics: why and how. Psychol. Sci. 25, 7–29. economic-circumstances-health-and-well-being/.
Daskalopoulou, C., Stubbs, B., Kralj, C., Koukounari, A., Prince, M., Prina, A., 2017. Park, J.E., Lee, J.Y., Kim, B.S., Kim, K.W., Chae, S.H., Cho, M.J., 2015. Above-moderate
Physical activity and healthy ageing: a systematic review and meta-analysis of physical activity reduces both incident and persistent late-life depression in rural
longitudinal cohort studies. Ageing Res. Rev. 38, 6–17. Koreans. Int. J. Geriatr. Psychiatry 30, 766–775.
Dishman, R.K., 1997. Brain monoamines, exercise, and behavioral stress: animal models. Pasco, J.A., Williams, L.J., Jacka, F.N., Henry, M.J., Coulson, C.E., Brennan, S.L., Leslie,
Med. Sci. Sports Exerc. 29 (1), 63–74 (Exerc Sport Sci Rev). E., Nicholson, G.C., Kotowicz, M.A., Berk, M., 2011. Habitual physical activity and
Dishman, R.K., Berthoud, H.R., Booth, F.W., Cotman, C.W., Edgerton, V.R., Fleshner, the risk for depressive and anxiety disorders among older men and women. Int.
M.R., Gandevia, S.C., Gomez-Pinilla, F., Greenwood, B.N., Hillman, C.H., Kramer, Psychogeriatr. 23 (2), 292–298.
A.F., Levin, B.E., Moran, T.H., Russo-Neustadt, A.A., Salamone, J.D., Van Hoomissen, Radloff, L.S., 1977. The CES-D scale: a self-report depression scale for research in the
J.D., Wade, C.E., York, D.A., Zigmond, M.J., 2006a. Neurobiology of exercise. Obesity general population. Appl. Psychol. Meas. 1, 385–401.
(Silver Spring) 14 (3), 345–356. Radloff, L., Teri, L., 1986. Use of the center for epidemiological studies-depression scale
Dishman, R.K., Hales, D.P., Pfeiffer, K.A., Felton, G.A., Saunders, R., Ward, D.S., Dowda, with older adults. Clin. Gerontol. 5, 119–136.
M., Pate, R.R., 2006b. Physical self-concept and self-esteem mediate cross-sectional Rechel, B., Grundy, E., Robine, J.M., Cylus, J., Mackenbach, J.P., Knai, C., McKee, M.,
relations of physical activity and sport participation with depression symptoms 2013. Ageing in the European union. Lancet 381, 1312–1322.
among adolescent girls. Health Psychol. 25 (3), 396. Rius-Ottenheim, N., Geleijnse, J.M., Kromhout, D., van der Mast, R.C., Zitman, F.G.,
Von Elm, E., Altman, D.G., Egger, M., Pocock, S.J., Gøtzsche, P.C., Vandenbroucke, J.P., Giltay, E.J., 2013. Physical activity after myocardial infarction: is it related to mental

74
C.P. McDowell et al. Experimental Gerontology 112 (2018) 68–75

health? Eur. J. Prev. Cardiol. 20 (3), 399–408. 42,469 people in low-and middle-income countries. J. Affect. Disord. 229, 231–238.
Rose, D., Harrison, E., 2007. The European socio-economic classification: a new social Sund, A.M., Larsson, B., Wichstrøm, L., 2011. Role of physical and sedentary activities in
class schema for comparative European research. Eur. Soc. 9, 459–490. the development of depressive symptoms in early adolescence. Soc. Psychiatry
Rush, A.J., Trivedi, M.H., Wisniewski, S.R., Nierenberg, A.A., Stewart, J.W., Warden, D., Psychiatr. Epidemiol. 46, 431–441.
Niederehe, G., Thase, M.E., Lavori, P.W., Lebowitz, B.D., 2006. Acute and longer- Tian, X., Wang, C., Qiao, X., Liu, N., Dong, L., Butler, M., Si, H., Jin, Y., 2018. Association
term outcomes in depressed outpatients requiring one or several treatment steps: a between pain and frailty among Chinese community-dwelling older adults: depres-
STAR* D report. Am. J. Psychiatry 163, 1905–1917. sion as a mediator and its interaction with pain. Pain 159, 306–313.
Schuch, F.B., Deslandes, A.C., Stubbs, B., Gosmann, N.P., da Silva, C.T., de Almeida Fleck, Trivedi, M.H., Rush, A.J., Wisniewski, S.R., Nierenberg, A.A., Warden, D., Ritz, L.,
M.P., 2016. Neurobiological effects of exercise on major depressive disorder: a sys- Norquist, G., Howland, R.H., Lebowitz, B., Mcgrath, P.J., 2006. Evaluation of out-
tematic review. Neurosci. Biobehav. Rev. 1 (61) (1-1). comes with citalopram for depression using measurement-based care in STAR* D:
Schuch, F., Vancampfort, D., Firth, J., Rosenbaum, S., Ward, P., Reichert, T., Bagatini, implications for clinical practice. Am. J. Psychiatry 163, 28–40.
N.C., Bgeginski, R., Stubbs, B., 2017. Physical activity and sedentary behavior in Tsutsumimoto, K., Makizako, H., Doi, T., Hotta, R., Nakakubo, S., Shimada, H., Suzuki, T.,
people with major depressive disorder: a systematic review and meta-analysis. J. 2017. Prospective associations between sedentary behaviour and incident depressive
Affect. Disord. 210, 139–150. symptoms in older people: a 15-month longitudinal cohort study. Int. J. Geriatr.
Schuch, F.B., Vancampfort, D., Firth, J., Rosenbaum, S., Ward, P.B., Silva, E.S., Hallgren, Psychiatry 32 (2), 193–200.
M., Ponce De Leon, A., Dunn, A.L., Deslandes, A.C., Fleck, M.P., 2018. Physical ac- Vancampfort, D., Firth, J., Schuch, F.B., Rosenbaum, S., Mugisha, J., Hallgren, M., Probst,
tivity and incident depression: a meta-analysis of prospective cohort studies. Am. J. M., Ward, P.B., Gaughran, F., De Hert, M., Carvalho, A.F., 2017. Sedentary behavior
Psychiatr. https://doi.org/10.1176/appi.ajp.2018.17111194. and physical activity levels in people with schizophrenia, bipolar disorder and major
Sharpe, D., 2015. Your chi-square test is statistically significant: now what? Pract. Assess. depressive disorder: a global systematic review and meta-analysis. World Psychiatry
Res. Eval. 20, 2–10. 16 (3), 308–315.
Smith, T.L., Masaki, K.H., Fong, K., Abbott, R.D., Ross, G.W., Petrovitch, H., Blanchette, Wang, P.S., Berglund, P., Olfson, M., Pincus, H.A., Wells, K.B., Kessler, R.C., 2005a.
P.L., White, L.R., 2010. Effect of walking distance on 8-year incident depressive Failure and delay in initial treatment contact after first onset of mental disorders in
symptoms in elderly men with and without chronic disease: the Honolulu-Asia Aging the National Comorbidity Survey Replication. Arch. Gen. Psychiatry 62, 603–613.
Study. J. Am. Geriatr. Soc. 58, 1447–1452. Wang, P.S., Lane, M., Olfson, M., Pincus, H.A., Wells, K.B., Kessler, R.C., 2005b. Twelve-
Sobocki, P., Jönsson, B., Angst, J., Rehnberg, C., 2006. Cost of depression in Europe. J. month use of mental health services in the United States: results from the National
Ment. Health Policy Econ. 9, 87–98. Comorbidity Survey Replication. Arch. Gen. Psychiatry 62, 629–640.
Soysal, P., Veronese, N., Thompson, T., Kahl, K.G., Fernandes, B.S., Prina, A.M., Solmi, Whelan, B.J., 1979. Ransam a random sample design for Ireland. Econ.Soc. Rev. 10, 169.
M., Schofield, P., Koyanagi, A., Tseng, P.T., 2017. Relationship between depression World Health Organization, 2000. Obesity: preventing and managing the global epi-
and frailty in older adults: a systematic review and meta-analysis. Ageing Res. Rev. demic. In: Geneva: WHO Technical Report Series 894, p. 252. Available at. http://
36, 78–87. www.who.int/nutrition/publications/obesity/WHO_TRS_894/en/, Accessed date: 10
Stubbs, B., Binnekade, T.T., Soundy, A., Schofield, P., Huijnen, I.P., Eggermont, L.H., October 2017.
2013. Are older adults with chronic musculoskeletal pain less active than older adults World Health Organization, 2010. Global Recommendations on Physical Activity for
without pain? A systematic review and meta-analysis. Pain Med. 14, 1316–1331. Health. World Health Organization, Geneva Available at. http://www.who.int/
Stubbs, B., Chen, L.J., Chang, C.Y., Sun, W.J., Ku, P.W., 2017. Accelerometer-assessed dietphysicalactivity/factsheet_recommendations/en/, Accessed date: 22 February
light physical activity is protective of future cognitive ability: a longitudinal study 2018.
among community dwelling older adults. Exp. Gerontol. 91, 104–109. World Health Organization, 2017. Depression: A Global Public Health Concern. World
Stubbs, B., Vancampfort, D., Firth, J., Schuch, F.B., Hallgren, M., Smith, L., Gardner, B., Health Organization, Geneva Available at. http://www.who.int/mediacentre/
Kahl, K.G., Veronese, N., Solmi, M., 2018. Relationship between sedentary behavior factsheets/fs369/en/, Accessed date: 22 February 2018.
and depression: a mediation analysis of influential factors across the lifespan among

75