Skip to main content

Table 1 Characteristics of cross-sectional and longitudinal studies grouped by age groups

From: Light-intensity physical activity and mental ill health: a systematic review of observational studies in the general population

Author Year [ref]
Design
Country Participants sample size follow-up time Measurement and operationalization of LIPA Outcomes Statistical analysis and effect size [95%CI] Adjustment Main findings Quality score (%)
Loprinzi 2013 [29]
CS
US Older adults (42.8% women; mean age = 73.5 years) n = 708 Accelerometry
100–2019 counts/min
Depression (Patient Health Questionnaire-9)
[Score of ≥ 9 indicating depression]
Logistic regression
Per 60 min increase OR = 0.80 [0.67, 0.95] p = 0.01
Age, gender, race-ethnicity, BMI, marital status, education, comorbidity index, physical functioning LIPA was significantly associated with lower odds for depression. 90.9
Varma et al. 2014 [30]
CS
US Older adults (76.5% women; mean age = 66.8 years) n = 187 Step Watch Activity Monitor
 < 100 steps/min
Depressive symptoms (Geriatric Depression Scale) Linear regression
Per 1000 steps increase:
β = -0.09 [-0.17, -0.01]
Per 10 min increase:
β = -0.02 [-0.05, -0.00]
Number of bouts of 10 min activity): β = -0.04 [-0.07, -0.00]
Age, gender, race Greater amount, frequency, and duration of LIPA were significantly associated with fewer depressive symptoms. 68.2
Yasunaga et al. 2018 [31]
CS
Japan Older adults (38.0% women; mean age = 74.4 years) n = 276 Accelerometry
METs > 1.5 to < 3.0
Depressive symptoms (Japanese version of the 15-item Geriatric Depression Scale) Linear regression
Per 30 min increase β = -0.030 (-0.184, 0.124)
Isotemporal substitution model
β = -0.131 [-0.260, -0.002]
Gender, age, BMI, physical function, marital status, educational attainment, MVPA, sedentary behavior LIPA (min/day) was not significantly associated with depression score
Replacing 30 min per day of SB with the same amount of LIPA was significantly associated with lower depression score.
95.5
Ku et al. 2018 [45]
LG
Taiwan Older adults (54.4%; mean age = 74.5 years) n = 274
2 years
Accelerometry
100–1951 counts/min
Depressive symptoms (15-item Geriatric Depression Scale) Linear regression
RR = 0.67 [0.51, 0.88], p = 0.004
Sex, age, income source, drinking, number of diseases, insomnia, ADL difficulty, cognitive impairment, accelerometer wearing time, baseline depressive symptoms, MVPA Participants who spent more time in LIPA had significantly fewer depressive symptoms at follow-up, independently of MVPA. 95.5
Uemura et al. 2017 [46]
LG
Japan Older adults (49.1% female; mean age = 71.5 years) n = 3,106
15 months
Self-reported unvalidated questionnaire (dichotomous variable) Depression (15-item Geriatric Depression Scale)
[Score of ≥ 6 indicating presence of depressive symptoms]
Logistic regressions
Light exercise OR = 0.74 [0.56, 0.98]
Walking habits
OR = 0.76 [0.57, 1.01]
Age, gender, education, current smoking status, alcohol consumption, living status, self‐rated health, Mini‐Mental State Examination, Short Physical Performance Battery, medications, Geriatric Depression Scale at baseline Engagement in light physical exercise but not walking was associated with significantly lower risk of depressive symptoms. 90.9
Ribeiro et al. 2017 [47]
LG
US Older African Americans (65.1% female; mean age = 66.1 years) n = 582
9 years
Self-reported
Yale Physical Activity Survey
Depression (11-item version of the Center for Epidemiological Studies- Depression) Logistic regression (Results not reported) PA components (vigorous activity, leisure walking, moving, standing, and sitting), vegetable and fruit intake, age, gender, perceived income adequacy, years of formal education Moving was not significantly associated with depression. 84.4
Rethorst et al. 2017 [32]
CS
US Hispanic/Latino community (female 52.1%; mean age = 41.1 years) n = 11,116 Accelerometry
100–1534 counts/min
Depressive symptoms (Center for Epidemiological Studies Depression Scale 10) Linear regression
β = 0.117 [-0.003, 0.237] p = 0.055
Isotemporal substitution model
β = 0.003 [-0.011, 0.113] p = 0.955
Age, sex, Hispanic background group, BMI, household income level, education, recruitment site, physical health, general familial social support, acculturation, other PA intensities LIPA (min/day) was not significantly associated with less depressive symptoms. Substitution of 1 h of SB with LIPA did not result in a significant decrease in depressive symptoms. 95.5
Asztalos et al. 2010 [33]
CS
Belgium Healthy adults (50.5% women; age 25–64 years) n = 6,803 Self-reported
International Physical Activity Questionnaire
[Dichotomized around 120 min/week]
Depression and anxiety (Symptom CheckList-90-Revised)
[Mean score of ≥ 2 indicating depression or anxiety]
Logistic regression
Men (depression)
OR = 0.863 [0.631, 1.180]
Men (anxiety)
OR = 0.918 [0.644, 1.310]
Women (depression)
OR = 0.880 [0.700, 1.106]
Women (anxiety)
OR = 0.937 [0.723, 1.214]
Age, socioeconomic status LIPA was not significantly associated with anxiety and depression in women and men. 86.4
Dillon et al. 2018 [34]
CS
Ireland Adults (53.9% female; mean age = 59.6 years) n = 397 Accelerometry dominant wrist = 191.8–281.5 counts/min
non-dominant wrist = 159.5–261.8 counts/min
-Depression (Centre for Epidemiologic Studies Depression)
-Anxiety (Hospital Anxiety and Depression Scale)
Isotemporal substitution models
Anxiety: β =  − 0.34 [-0.64, -0.04]
Depression: β = -0.63 [-1.51, 0.26]
Age, gender, season, marital status, smoking status, alcohol consumption, BMI status, education Substituting 30 min of SB for LIPA per day was associated with a significant decrease in levels of anxiety but not depression. 86.4
Poole et al. 2011 [35]
CS
UK Staff and student females (100% women; mean age = 28.7 years) n = 40 Accelerometry
191–573 counts/min
Depressive symptoms (Center for Epidemiologic Studies Depression Scale) Pearson correlations
r = -0.35 p < 0.05
None Accelerometry-measured LIPA (min/day) was significantly correlated with fewer depressive symptoms. 54.5
Mourady et al. 2017 [36]
CS
Lebanon Healthy pregnant women (100% women; mean age = 30.5 years) n = 141 Self-reported Pregnancy Physical Activity Questionnaire -Depression (Zung Self-Rating Depression Scale) Spearman correlations
Depression: r = -0.182. p = 0.031
None Light PA (MET-hrs/week) has a significant inverse correlation with depression. 68.2
von Känel R et al. 2017 [37]
CS
South Africa Adults (52.2%women; mean age = 49.8 years) n = 203 Actiheart
 > 1.5- < 3METs
Psychological distress (28-item General Health Questionnaire) Partial correlations (Results not reported) Age, gamma glutamyl transferase Partial correlations did not show any associations between PA and psychological distress. 63.6
Hamer M et al. 2014 [38]
CS
UK Adults (48.1–61.8% women; mean age = 48.3–50 years) n = 1,947 Accelerometry
200–2019 counts/min
in tertiles
Psychological distress (12-item version General Health Questionnaire)
[Score of ≥ 4 indicating psychological distress]
Logistic regression
Highest tertile vs lowest
OR = 0.73 [0.48, 1.12]
Middle vs lowest tertile
OR = 0.56 [0.37, 0.84]
Age, sex, accelerometry wear time, smoking, alcohol, education, BMI, social occupational group employment long-standing illness (non-mental), MVPA LIPA was significantly associated with lower odds for psychological distress, independently of MVPA (although the association was not linear). 90.9
Sheikh et al. 2018 [48]
LG
Norway Adults (54% female, mean age = 47.0 years) n = 10,325
13 years
Self-reported. Unvalidated questionnaire
[Categorical scale, none; less than 1; 1–2; 3 or more hrs/week]
Psychological distress (Hopkins symptom checklist, HSCL-10)
[Score of ≥ 18.5 indicating psychological distress]
Regression analysis (ordinary least square and relative risk)
β = 0.09 [-0.03, 0.22]
RR = 1.03 [0.95, 1.11]
Age, gender, history of parental psychopathology, childhood socioeconomic status, marital status, daily smoking, number of friends, perceived social isolation, education, MVPA LIPA did not confer significant protection against psychological distress at follow-up after adjusting for MVPA. 90.9
Bernard et al. 2018 [39]
CS
Canada Adults (50.3% women; mean age = 44 years) n = 8,150 Accelerometry
100 to 1534 counts/min
Overall mental health Single item question self-rated mental health Generalized additive models
LIPA = Estimate 7.2. f = 3. p = 0.003
LIPA + MVPA = Estimate 24.4. f = 3.6. p < 0.001
LIPA + SB = Estimate 21.6. f = 2.5. p < 0.001
Age, sex, daily smoking, household income, education level, accelerometer wear time, season of accelerometer assessment, BMI, MVPA A curvilinear relationship between daily mins of LIPA and mental health was found, with better mental health found in 400–550 average mins of daily LPA. 90.9
Felez-Nobrega et al. 2020 [40]
CS
Spain College students (44% women; mean age = 20.8 years) n = 360 self-reported; n = 121 activPAL Self-reported
International Physical Activity Questionnaire
[Tertile groups: T1 ≤ 3 h/week; T2 > 3 to ≤ 7 h/week; T3 > 7 h/week]
Accelerometry
%LIPAhrs/day based on time not spent standing, in MVPA or sedentary
-Perceived stress (Perceived Stress Scale)
-Anxiety (State-Trait Anxiety Inventory)
Linear regression
Self-reported LIPA:
State anxiety: T2 vs T1 B = -1.70 [-5, 1.97]; T3 vs T1 B = -2.24 [-5.52, 0,86]
Trait anxiety: T2 vs T1 B = -1,49 [-3.66, 0.90]; T3 vs T1 B = -1.47 [-3.51, 0.82]
Perceived stress: T2 vs T1 B = -1.59 [-3.62, 0.46]; T3 vs T1 B = -2.07 [-3.90, -0.06]
Partial correlation
activPAL LIPA weekday; weekend day:
State anxiety: r = 0.10, p = 0.69; r = -0.07, p = 0.51
Trait anxiety: r = -0.04, p = 0.64; r = -0.06, p = 0.54
Perceived stress: r = 0.04, p = 0.69; r = 0.07, p = 0.43
Age, gender Self-reported LIPA was significantly associated with lower perceived stress but not with state-trait anxiety. No significant associations were found for device-based measures of LIPA (activPAL) with any of the mental health outcomes. 81.8
Tao et al. 2019 [41]
CS
China College students (52.3% women; mean age = 20.3 years) n = 220 Accelerometry
Steps rate (steps/min) 20–99
Anxiety and depression (24-item Patient-Reported Outcomes Measurement Information System) Pearson correlations
Anxiety r = -0.023 (N.S)
Depression r = 0.011 (N.S)
None LIPA (min/day) was not significantly correlated with anxiety or depression. 72.7
Lee et al. 2013 [42]
CS
Hong Kong Non exercising healthy adolescents and adults (59.7% women; mean age = 46.2 years) n = 2,417 Accelerometry
101–1951 counts/min
Depression (Patient Health Questionnaire 9)
[Score of ≥ 5 indicating depression (mild)]
Difference in mean z scores among those with and without depression
-0.10 [-0.20, 0.01]
Age, sex LIPA (min/day) was not significantly associated with depression. 95.5
Costigan et al. 2019 [43]
CS
Australia Healthy adolescents (44.9% girls; mean age = 12.9 years) n = 1,223 Accelerometry
101–2295 counts/min
Negative affect (the Positive and Negative Affect Scale for children) Quantile regressions
Standardized β = 0.010, p = 0.837
Sex, BMI, ethnicity, wear time, other physical activity intensities No significant associations between LIPA and negative affect were found. 95.5
Kandola et al. 2020 [49]
LG
UK Healthy adolescents (56.1% girls; mean age 12 years) n = 4,257
6 years
Accelerometry
200–3599 counts/min
Depression
Computerized Clinical Interview Schedule-Revised for depression at 18 years of age
-The Short Moods and Feelings Questionnaire for depression at age 12, 14, and 16
Negative binomial regression models and by group-based trajectory modelling
Increasing LIPA at age 12 (IRR = 0.904, [0.850, 0.961], p = 0.0012). Similar estimates were found for the other age groups
Sex, ethnicity, maternal social class, baseline depression, IQ, parental psychiatric history, parental education, total accelerometer wear time At all timepoints, each 60 min/day increase in LIPA was significantly associated with a lower depression score at 18 years of age: 9.6% for LIPA at 12 years, 7.8% at 14 years, and 11.1% at 16 years of age.
Lower depression scores were identified in participants with persistently high levels of LIPA.
100
Parfitt et al. 2009 [44]
CS
UK Healthy primary school children (59.6% girls; age 9–10 years) n = 57 Accelerometry
Very LIPA = 100–470.1 counts/min
LIPA = 470.1–976.8 counts/min
-Depression (Child Depression Inventory)
-Anxiety (State-trait inventory for children)
Pearson correlations
Very LIPA and anxiety r = 0.331. p < 0.05
Very LIPA and depression r = 0.282. p < 0.05
LIPA and anxiety r = 0.173. p = N.S
LIPA and depression r = 0.202. p = N.S
Very light PA adjusted for percent body fat
Correlations for LIPA not adjusted
Very LIPA (min/day) was significantly correlated with higher measures of anxiety and depression. No significant correlations were found for LIPA. 81.8
Ahn et al. 2018 [50]
LG
UK Children (61.3% female, mean age 11 years) n = 6,153
2 years
Accelerometry
100–2240 counts/min
Emotional symptoms (Strengths and Difficulties Questionnaire) Linear regression
Boys:
β = -0.039 [-0.106, 0.028]
Girls:
β = -0.026 [-0.094, 0.041]
Age, season, total difficulties at age 7, limiting illness, special education needs, weight status, self-esteem, ethnicity, income, siblings, family structure, maternal education, maternal depression, maternal employment, British Ability Scale pattern construction, British Ability Scale word reading LIPA was not significantly related to emotional symptoms at follow-up. 95.5
  1. Only fully confounder-adjusted estimates are shown in tables
  2. Abbreviations: LIPA Light intensity physical activity, MVPA Moderate-to-vigorous intensity physical activity, PA Physical activity, min Minutes, BMI Body mass index, IQ Intelligence quotient, ADL Activities of daily living, SB Sedentary behavior, CS Cross-sectional, LG Longitudinal