Skip to main content
PMC home page
Preventive Medicine Reports logoLink to Preventive Medicine Reports
. 2021 Nov 15;24:101640. doi: 10.1016/j.pmedr.2021.101640

Relationships among changes in walking and sedentary behaviors, individual attributes, changes in work situation, and anxiety during the COVID-19 pandemic in Japan

Shohei Nagata a, Hiroki M Adachi a, Tomoya Hanibuchi a, Shiho Amagasa b, Shigeru Inoue b, Tomoki Nakaya a,⁎
PMCID: PMC8684034  PMID: 34976691

Highlights

  • •

    Females, younger people, those living in the highest-density tended to be inactive.

  • •

    Inactivity was related to standby/work-from-home for higher socioeconomic people.

  • •

    Inactivity was related to decreased amount of work for lower socioeconomic people.

  • •

    Decreased physical activity was related to strong anxiety during the pandemic.

Keywords: Physical activity, Walking, Sedentary behavior, Anxiety, COVID-19, Japan

Abbreviations: ADI, Area deprivation index; CFI, Comparative fit index; DID, Densely-inhabited-district; NPIs, Non-pharmaceutical interventions; RMSEA, Root mean square error of approximation; SES, Socioeconomic status; SoE, State of emergency

Abstract

Studies from many countries, including Japan, have reported decreased physical activity during the coronavirus disease 2019 (COVID-19) pandemic. However, the individual attributes as related to changes in physical activity during the pandemic in Japan have been scarcely investigated. The present study explored the relationships among individual attributes including demographic, socioeconomic, and geographic characteristics, work situation changes, perception of anxiety, and changes in walking and sedentary behaviors, during the pandemic in Japan. To obtain data indicating individual circumstances during the first wave of the pandemic in Japan, we conducted a nationwide online survey from May 19 to May 23, 2020 (n = 1,200). To observe changes in walking behavior objectively and retrospectively, we collected data on the number of daily steps as measured by the iPhone’s Health application. Path analysis was employed to examine relationships between individual attributes and changes in walking and sedentary behaviors. Decreased physical activity, especially, decreased walking behavior among younger individuals and those living in highest-density neighborhoods were identified. There was increased sedentary behavior among females. Moreover, individuals with higher socioeconomic status (SES) tended to become inactive due to work-from-home/standby-at-home and individuals with lower SES tended to become inactive due to decreased amount of work. Decreased walking behavior and increased sedentary behavior were associated with a perception of strong anxiety related to the pandemic. Our findings would be helpful in considering measures to counteract health risks during the pandemic by taking into account individual backgrounds.

1. Introduction

The novel coronavirus disease (COVID-19) pandemic is ongoing as of August 2021. Although vaccinations have been administered in many countries, the World Health Organization reported that as of August 3, 2021, the number of new COVID-19 cases per week worldwide is still increasing (World Health Organization, 2021). Up until August 10, 2021, more than 200 million cases have been confirmed globally, including more than 4.3 million deaths (Johns Hopkins Coronavirus Resource Center, 2021). To reduce the spread of infection, non-pharmaceutical interventions (NPIs), such as social distancing measures and lockdowns, were implemented in many countries (BBC, 2021, Courtemanche et al., 2020, Lau et al., 2020). In Japan, to prevent the health system from collapsing due to rapid increase in new positive cases, the government declared a state of emergency (SoE) four times until August 10, 2021 and requested people to show self-restraining behaviors such as shortening business hours at restaurants, avoiding nonessential outings, and working from home (Office for Novel Coronavirus Disease Control, Cabinet Secretariat, Government of Japan, 2020).

Previous studies have revealed the effects of changes in human mobility related to NPIs during the COVID-19 outbreak (Amagasa et al., 2020, Badr et al., 2020, Li et al., 2020, Nagata et al., 2021, Yabe et al., 2020, Zhou et al., 2020). Reduced mobility that effectively reduces the spread of infection could cause physical inactivity, which is concerning (Hall et al., 2021, Lippi et al., 2020). A descriptive study showed that globally, within 30 days of declaration of the pandemic, there was a 27.3% reduction in the mean number of steps taken (Tison et al., 2020). Further, several studies have shown that reduced physical activity during the COVID-19 pandemic is associated with mental health issues, such as depression, loneliness, stress, anxiety, and sadness (Meyer et al., 2020, Pieh et al., 2020, Silva et al., 2020). Additionally, according to the World Health Organization, physical inactivity is a major risk factor for non-communicable diseases (World Health Organization, 2020) and is estimated to cause 6–10% of coronary heart diseases, type 2 diabetes, and breast and colon cancers worldwide (Lee et al., 2012). Therefore, it is important to analyze reduced physical activity caused by the implementation of NPIs to plan long-term measures against the current and any future pandemics.

Much has been discussed on the factors encouraging or discouraging physical activity. Many studies consistently demonstrate that individuals of higher socioeconomic status (SES) are more likely to be physically active during leisure time, while those of lower SES are more likely to engage in job-related physical activity (Beenackers et al., 2012, McNeill et al., 2006). Furthermore, neighborhood environments are also associated with the amount of physical activity. High-walkability neighborhoods, which have mixed land use or favorable esthetic qualities, promote residents’ walking behavior (Saelens et al., 2003, Saelens and Handy, 2008, Sallis et al., 2009), but those living in highly-deprived neighborhoods are more likely to be physically inactive (Cubbin et al., 2006, Hillsdon et al., 2008). During the COVID-19 pandemic, people with low incomes in the USA and the UK showed decreased physical activity (Fearnbach et al., 2021, Robinson et al., 2021). Conversely, a study from Bangladesh demonstrated that highly educated individuals with high income is more likely to be physically inactive (Rahman et al., 2020). Considering such differences reflect social structures and the measures taken against COVID-19 in each country; further evaluation based on the situations in different countries is needed.

In Japan, decreased physical activity was observed during the COVID-19 outbreak (Hino and Asami, 2021, Makizako et al., 2021, Yamada et al., 2020). Hino and Asami (2021) suggested that proximity to large parks could effectively mitigate decreased walking among female older adults during the SoE. Hanibuchi et al. (2021) reported that reduced time spent outdoors is associated with individual attributes such as age, gender, income, or residential location and perception of anxiety related to the infection or the stigma. Moreover, Koohsari et al. (2021) revealed that the implementation of working-from-home was associated with decreased work-related physical activity and increased sitting time. However, few studies have attempted to clarify the comprehensive relationships between physical inactivity and individual attributes such as age, residential location, work situation changes, or anxiety related to the pandemic. The present study explored the relationships among individual attributes including demographic, socioeconomic, and geographic characteristics, work situation changes, perception of anxiety, and changes in physical activity, specifically decreased walking and increased sedentary behaviors, during the first wave of the COVID-19 pandemic in Japan.

2. Methods

2.1. Data collection

We conducted a nationwide online survey among registered panel members of a survey company (Cross Marketing Inc.) from May 19 to May 23, 2020. People aged 20–69 years with diverse demographic and socioeconomic backgrounds, owning iPhones, and living in Japan were recruited from 4.65 million panel members. The quota sample was designed to have the same distribution of population by age, gender, and geographical region based on the 2015 Japan population census. As for the definition of the geographical region, we classified the prefectures into metropolitan areas that consist of Tokyo, Kanagawa, Saitama, Chiba, Aichi, Gifu, Mie, Osaka, Hyogo, Kyoto and Nara and nonmetropolitan areas that consist of the rest. However, this sampling design was not applied to participants aged 60–69 years because of fewer responses from females of this age group.

2.2. Measurement of changes in walking and sedentary behaviors

To analyze changes in sedentary behavior, the participants were asked regarding change is duration of sitting since COVID-19 outbreak when compared with before the pandemic. Participants selected answers from the following options: significant reduction, slight reduction, no change, slight rise, significant rise. The answers were converted to integer values (1: significant reduction, 2: slight reduction, 3: no change, 4: slight rise, 5: significant rise) to obtain ordinal variables for statistical analyses.

Additionally, to observe changes in walking behavior objectively, from the participants, we collected screenshot images of the pre-installed “iPhone Health App” (Apple Inc.), which automatically records daily step counts. The screenshots of number of daily steps were captured for the previous 3 months by the participants. We then extracted this information through image processing and optical character recognition methods using Python 3.7.7, OpenCV 4.2.0, and Tesseract 5.0.0. Details of the image processing methods to obtain the step counts in numbers have been described previously (Adachi et al., 2021). Subsequently, we calculated the differences between the number of mean steps before and after the first SoE for each participant and used them for each period to measure the changes in walking behavior. According to a previous study, the pre-SoE period was from February 19, 2020 to March 23, 2020 and the post-SoE period was from April 16, 2020 to May 19, 2020 (Adachi et al., 2021).

2.3. Demographic, socioeconomic, and geographic variables

The demographic and socioeconomic attributes of the participants considered as variables were: gender (0: males, 1: females), age (20–29 years, 30–39 years, 40–49 years, 50–59 years, 60–69 years), chronic disease (0: no, 1: yes), educational status (junior high school/high school, junior (technical) college/vocational school, undergraduate/graduate school), occupation (white-collar job including administrators, professionals, and office clerks; gray-collar job including sales clerks and service workers; blue-collar job including security workers and production, construction, and transportation workers; and other/not working), household annual income (<3 million yen, 3–7 million yen, ≥7 million yen, and unknown), living alone (0: no, 1: yes), living with child(ren) under 18 years (0: no, 1: yes), and living with person(s) aged 65 years and older (0: no, 1: yes). The data corresponding to the categorical variables, such as age, educational status, occupation, and household income, were converted to a binary system indicating whether the participants belong to each group or not (0: no, 1: yes).

Neighborhood-level population density and deprivation were considered as geographic variables. The neighborhoods were defined by the postal code of the participants’ residential address and were categorized as follows into groups of approximately equal sample sizes: lowest density (non-densely-inhabited-district (non-DID), defined by the 2015 Japan population census), middle-low density (DID with 7,015 people/km2 or fewer), middle-high density (DID with 7,016–10,214 people/km2), and highest density (DID with 10,215 people/km2 or more). The neighborhood-level deprivation indicator was calculated using the area deprivation index (ADI) derived from the 2015 Japan population census, which has been explained previously (Nakaya et al., 2014). A higher ADI indicates that the neighborhood has more deprived conditions. We categorized the neighborhoods based on ADI quartiles as lowest ADI, middle-low ADI, middle-high ADI, and highest ADI groups.

2.4. Variables representing changes related to work situation and anxiety

We hypothesized that the work situation changes and perception of anxiety due to COVID-19 are also associated with walking and sedentary behaviors. The variables that indicate changes in work situation were: introduction of work-from-home/standby-at-home measures (0: no, 1: yes) and decreased amount of work (0: no, 1: yes). Three anxiety variables were also used: strong anxiety about getting infected (0: no, 1: yes), spreading the infection to others (0: no, 1: yes), and stigma associated with going out (0: no, 1: yes).

2.5. Statistical analysis

To examine the direct and indirect relationships among the individual attributes including demographic, socioeconomic, and geographic variables, changes in work situation, perception of anxiety due to the pandemic, and changes in physical activity, we assumed the following models (Fig. 1): model A represents that the individuals’ background affected their work situation and perception of anxiety which affected the changes in physical activity; model B assumes the inverse relationship between perception of anxiety and the changes in physical activity represented by model A to account for the possibility that the physical inactivity causes increased anxiety during the pandemic (Silva et al., 2020); model C and D assumes the direct relationship between the work situation changes and perception of anxiety in addition to the frameworks of model A and B. All models assume the direct relationships between the individuals’ background and the changes in physical activity.

Fig. 1.

Fig. 1

Open in a new tab

Conceptual frameworks of the models indicating hypotheses of the relationships among individual attributes, work situation changes, perception of anxiety, and changes in physical activity.

We examined the relationships by path analysis, a special case of structural equation modeling, and evaluated the most suitable model to explain the comprehensive relationships by comparative fit index (CFI) and root mean square error of approximation (RMSEA) which indicate how well the model fits. CFI is expressed as a value between 0 and 1, and models with a value greater than 0.95 are often interpreted as good fitting (Ullman and Bentler, 2012). RMSEA represents that the smaller the value, the better fitting the model, and values of 0.06 or less can be interpreted as good fitting (Ullman and Bentler, 2012). Furthermore, to evaluate the relationship between each variable, the relationship was considered statistically significant if the path’s p-value was<0.05. To estimate the coefficients to the dichotomous variables indicating changes in work situation and perception of anxiety and ordinal variables indicating changes in time spent in sedentary behavior, binary probit and ordered probit regression models were employed, respectively. All statistical analyses were performed using R 3.6.1, and the lavaan package, version 0.6–7, was used to run path analysis.

This study was approved by the Research Ethics Committee of the Graduate School of Engineering, Tohoku University (approval number: 20A-3). Informed consent was obtained from all participants.

3. Results

In the online survey, 1,200 panel members participated. We excluded the data of 282 participants from whom the daily step counts could not be obtained for more than 15 days both before and after the SoE because of errors such as low image resolution. We excluded the data of five participants whose educational data was unavailable and 17 participants whose postal code was missing. Finally, data of 896 participants were used for analyses (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Flow diagram of participant inclusion.

Fig. 3 shows the differences in mean step counts between late-February and mid-May 2020. The step counts decreased gradually from late-February, and there was more than 20% reduction after the SoE. Table 1 summarizes the changes in walking and sedentary behaviors during the COVID-19 outbreak based on demographic, socioeconomic, and geographic attributes. More than 60% of participants increasingly spent time in sedentary behavior. The average step count consistently decreased across all attributes, while sedentary behavior increased during the outbreak. There were significant differences in step count reductions when considering age, educational status, living with child(ren), neighborhood density, ADI, and introduction of work-from-home/standby-at-home measures. Furthermore, the changes in time spent in sedentary behavior across groups categorized according to gender, educational status, neighborhood density, ADI, introduction of work-from-home/standby-at-home, decreased amount of work, and strong anxiety about getting infected, spreading the infection to others, and the stigma associated with going out, were significant.

Fig. 3.

Fig. 3

Open in a new tab

Changes in mean step counts of the participants (7-day moving average). Vertical dash lines represent the start/end date of state of emergency (SoE) declaration (April 7, 2020, SoE declaration for Tokyo, Osaka, Kanagawa, Saitama, Chiba, Hyogo, and Fukuoka; April 16, 2020, SoE declaration for the remaining prefectures; May 14, 2020, lifting of SoE for prefectures excluding Hokkaido, Saitama, Chiba, Tokyo, Kanagawa, Kyoto, Osaka, and Hyogo).

Table 1.

Changes in walking and sedentary behaviors during the COVID-19 outbreak based on characteristics of respondents.

n % Average change in step counts a)b) Number of participants by change levels of sedentary behavior c)
Significant reduction Slight reduction No change Slight rise Significant rise
Total 896 −918.4 (SD = 2276.2) 1 (0.1%) 9 (1.0%) 321 (35.8%) 293 (32.7%) 272 (30.4%)
Gender p = 0.263 p = 0.002
 Males 459 51.2 −1001.5 (SD = 2318.7) 1 (0.2%) 5 (1.1%) 178 (38.8%) 160 (34.9%) 115 (25.1%)
 Females 437 48.8 −831.2 (SD = 2230.0) 0 (0.0%) 4 (0.9%) 143 (32.7%) 133 (30.4%) 157 (35.9%)
Age p < 0.001 p = 0.066
 20–29 years 134 15.0 −1991.5 (SD = 2604.2) 0 (0.0%) 3 (2.2%) 40 (29.9%) 34 (25.4%) 57 (42.5%)
 30–39 years 180 20.1 −771.0 (SD = 2586.5) 0 (0.0%) 2 (1.1%) 60 (33.3%) 64 (35.6%) 54 (30.0%)
 40–49 years 214 23.9 −907.6 (SD = 1903.6) 0 (0.0%) 1 (0.5%) 78 (36.4%) 71 (33.2%) 64 (29.9%)
 50–59 years 171 19.1 −756.1 (SD = 1921.4) 1 (0.6%) 1 (0.6%) 73 (42.7%) 52 (30.4%) 44 (25.7%)
 60–69 years 197 22.0 −475.9 (SD = 2186.2) 0 (0.0%) 2 (1.0%) 70 (35.5%) 72 (36.5%) 53 (26.9%)
Chronic disease p = 0.388 p = 0.983
 No 682 76.1 −952.9 (SD = 2338.1) 1 (0.1%) 7 (1.0%) 243 (35.6%) 225 (33.0%) 206 (30.2%)
 Yes 214 23.9 −808.5 (SD = 2067.8) 0 (0.0%) 2 (0.9%) 78 (36.4%) 68 (31.8%) 66 (30.8%)
Educational status p < 0.001 p = 0.025
 Junior high school/high school 177 19.8 −493.9 (SD = 1849.1) 1 (0.6%) 1 (0.6%) 77 (43.5%) 52 (29.4%) 46 (26.0%)
 Junior (technical) college/vocational school 219 24.4 −668.6 (SD = 2506.6) 0 (0.0%) 3 (1.4%) 84 (38.4%) 69 (31.5%) 63 (28.8%)
 Undergraduate/graduate school 500 55.8 −1178.1 (SD = 2276.6) 0 (0.0%) 5 (1.0%) 160 (32.0%) 172 (34.4%) 163 (32.6%)
Occupation p = 0.788 p = 0.601
 White-collar job 407 45.4 −901.1 (SD = 1876.6) 0 (0.0%) 3 (0.7%) 152 (37.3%) 126 (31.0%) 126 (31.0%)
 Gray-collar job 152 17.0 −1067.1 (SD = 2910.4) 0 (0.0%) 1 (0.7%) 57 (37.5%) 45 (29.6%) 49 (32.2%)
 Blue-collar job 75 8.4 −913.1 (SD = 2575.5) 0 (0.0%) 0 (0.0%) 31 (41.3%) 27 (36.0%) 17 (22.7%)
 Other/not working 262 29.2 −860.6 (SD = 2343.2) 1 (0.4%) 5 (1.9%) 81 (30.9%) 95 (36.3%) 80 (30.5%)
Household annual income p=0.347 p=0.085
 Less than 3 million yen 122 13.6 −808.0 (SD=3125.0) 0 (0.0%) 3 (2.5%) 50 (41.0%) 35 (28.7%) 34 (27.9%)
 3–7 million yen 377 42.1 −886.4 (SD=2082.5) 0 (0.0%) 2 (0.5%) 128 (34.0%) 121 (32.1%) 126 (33.4%)
 7 million yen or more 298 33.3 −950.5 (SD=1965.7) 1 (0.3%) 2 (0.7%) 100 (33.6%) 108 (36.2%) 87 (29.2%)
 Unknown 99 11.0 −1079.9 (SD=2619.3) 0 (0.0%) 2 (2.0%) 43 (43.4%) 29 (29.3%) 25 (25.3%)
Living alone p=0.134 p=0.363
 No 736 82.1 −856.5 (SD=2159.7) 1 (0.1%) 6 (0.8%) 267 (36.3%) 246 (33.4%) 216 (29.3%)
 Yes 160 17.9 −1203.5 (SD=2739.1) 0 (0.0%) 3 (1.9%) 54 (33.8%) 47 (29.4%) 56 (35.0%)
Living with child(ren) under 18 years p<0.001 p=0.226
 No 650 72.5 −1084.2 (SD=2295.6) 1 (0.2%) 7 (1.1%) 225 (34.6%) 213 (32.8%) 204 (31.4%)
 Yes 246 27.5 −480.3 (SD=2168.3) 0 (0.0%) 2 (0.8%) 96 (39.0%) 80 (32.5%) 68 (27.6%)
Living with person(s) aged 65 years and older p=0.423 p=0.091
 No 762 85.0 −942.0 (SD=2314.5) 1 (0.1%) 6 (0.8%) 268 (35.2%) 248 (32.5%) 239 (31.4%)
 Yes 134 15.0 −784.7 (SD=2047.5) 0 (0.0%) 3 (2.2%) 53 (39.6%) 45 (33.6%) 33 (24.6%)
Neighborhood density p<0.001 p<0.001
 Lowest density 218 24.3 −360.0 (SD=2117.8) 1 (0.5%) 1 (0.5%) 99 (45.4%) 66 (30.3%) 51 (23.4%)
 Middle-low density 232 25.9 −506.8 (SD=2046.2) 0 (0.0%) 0 (0.0%) 87 (37.5%) 80 (34.5%) 65 (28.0%)
 Middle-high density 229 25.6 −1029.4 (SD=2268.8) 0 (0.0%) 3 (1.3%) 81 (35.4%) 80 (34.9%) 65 (28.4%)
 Highest density 217 24.2 −1802.4 (SD=2399.2) 0 (0.0%) 5 (2.3%) 54 (24.9%) 67 (30.9%) 91 (41.9%)
Areal deprivation index (ADI) p=0.002 p=0.019
 Lowest ADI 224 25.0 −1118.2 (SD=2699.2) 1 (0.4%) 3 (1.3%) 75 (33.5%) 66 (29.5%) 79 (35.3%)
 Middle-low ADI 224 25.0 −1278.3 (SD=2359.6) 0 (0.0%) 0 (0.0%) 66 (29.5%) 84 (37.5%) 74 (33.0%)
 Middle-high ADI 224 25.0 −661.9 (SD=2055.4) 0 (0.0%) 2 (0.9%) 93 (41.5%) 71 (31.7%) 58 (25.9%)
 Highest ADI 224 25.0 −615.3 (SD=1840.5) 0 (0.0%) 4 (1.8%) 87 (38.8%) 72 (32.1%) 61 (27.2%)
Introduction of work-from-home/standby-at-home p<0.001 p<0.001
 No 595 66.4 −638.5 (SD=1996.7) 1 (0.2%) 5 (0.8%) 249 (41.8%) 193 (32.4%) 147 (24.7%)
 Yes 301 33.6 −1471.8 (SD=2664.6) 0 (0.0%) 4 (1.3%) 72 (23.9%) 100 (33.2%) 125 (41.5%)
Decreased amount of work p=0.080 p<0.001
 No 724 80.8 −849.8 (SD=2234.5) 1 (0.1%) 7 (1.0%) 284 (39.2%) 233 (32.2%) 199 (27.5%)
 Yes 172 19.2 −1207.2 (SD=2429.5) 0 (0.0%) 2 (1.2%) 37 (21.5%) 60 (34.9%) 73 (42.4%)
Strong anxiety about getting infected p=0.806 p=0.002
 No 613 68.4 −932.1 (SD=2112.0) 1 (0.2%) 6 (1.0%) 233 (38.0%) 209 (34.1%) 164 (26.8%)
 Yes 283 31.6 −888.8 (SD=2600.3) 0 (0.0%) 3 (1.1%) 88 (31.1%) 84 (29.7%) 108 (38.2%)
Strong anxiety about spreading the infection to others p=0.499 p=0.002
 No 668 74.6 −883.5 (SD=2063.6) 1 (0.1%) 6 (0.9%) 253 (37.9%) 226 (33.8%) 182 (27.2%)
 Yes 228 25.4 −1020.8 (SD=2810.7) 0 (0.0%) 3 (1.3%) 68 (29.8%) 67 (29.4%) 90 (39.5%)
Strong anxiety about stigma associated with going out p=0.942 p<0.001
 No 745 83.1 −915.2 (SD=2070.3) 1 (0.1%) 7 (0.9%) 280 (37.6%) 251 (33.7%) 206 (27.7%)
 Yes 151 16.9 −934.4 (SD=3107.0) 0 (0.0%) 2 (1.3%) 41 (27.2%) 42 (27.8%) 66 (43.7%)
Open in a new tab

a) Changes in step counts were calculated by the differences in the mean step counts between the pre-SoE period (from February 19, 2020 to March 23, 2020) the post-SoE period (from April 16, 2020 to May 19, 2020).

b) p for ANOVA or t-test

c) p for Wilcoxon rank sum test or Kruskal-Wallis test

Table 2 summarizes the fit indices of each model based on path analysis. Judging from CFI and RMSEA, model B best explained the relationships among the demographic, socioeconomic, and geographic variables, changes in work situation, perception of anxiety, and changes in walking and sedentary behaviors (CFI = 0.996, RMSEA = 0.057).

Table 2.

Fit indices of each model.

Model CFIa) RMSEAb)
A 0.064 0.753
B 0.996 0.057
C 0.913 0.364
D 0.995 0.124
Open in a new tab

a) Comparative Fit Index

b) Root Mean Square Error of Approximation

Fig. 4 shows the significant paths in model B. Details of all the estimated coefficients of the models are provided in the appendix. Considering individual attributes and the changes in physical activity, respondents aged 20–29 years, aged 40–49 years, and living in the highest-density neighborhoods were more likely to experience reduced step counts. However, white-collar workers and living with child(ren) under 18 years were positively associated with differences in step counts and such individuals were less likely to experience reduced walking behavior. Moreover, females and workers in jobs other than white, gray, and blue-collar jobs or non-working participants were more likely to show increased sedentary behavior, while participants with household income less than 3 million yen were less likely to show increased sedentary behavior.

Fig. 4.

Fig. 4

Open in a new tab

Coefficients of model B estimated by path analysis. Only significant paths and their coefficients among demographic, socioeconomic, and geographic variables, changes in work situation and perception of anxiety, and changes in walking and sedentary behaviors are shown.*** statistical significance at 0.1%; ** statistical significance at 1%; * statistical significance at 5%. CFI = 0.996, RMSEA = 0.057. Changes in step counts were determined by calculating the difference in the mean step counts between the pre-state of emergency (SoE) period (from February 19, 2020 to March 23, 2020) and the post-SoE period (from April 16, 2020 to May 19, 2020). The ordered categories of change in time spent in sedentary behavior were defined as follows: 1: significant reduction, 2: slight reduction, 3: no change, 4: slight rise, 5: significant rise.

The work-from-home/standby-at-home group was positively associated with undergraduate/graduate school, and white-collar jobs, along with middle-low density, middle-high density, and highest density neighborhoods, while it was negatively associated with female respondents and the other jobs/not working group. Decreased amount of work was positively associated with an income of<3-million-yen group and negatively associated with the other jobs/not working group. The work-from-home/standby-at-home group and the decreased work group were associated with decreased step counts and increased time spent in sedentary behavior.

Female participants, age 30–39 years, and incomes<3 million yen were positively associated with strong anxiety about getting infected and about spreading the infection to others. Further, participants aged 20–29 years, aged 30–39 years, living with child(ren) under 18 years were positively associated with strong anxiety about the stigma of going out. Living in middle-low ADI were negatively associated with strong anxiety about spreading the infection to others and living in the highest density neighborhoods were negatively associated with the stigma of going out. Decreased step counts were associated with strong anxiety about spreading the infection to others or about the stigma of going out while increased time spent in sedentary behavior was associated with all anxiety variables.

4. Discussion

Several studies have documented decreased physical activity during the COVID-19 outbreak in Japan (Adachi et al., 2021, Hino and Asami, 2021, Makizako et al., 2021, Yamada et al., 2020). However, to the best of our knowledge, this is the first study to examine the association of changes in physical activity during the COVID-19 outbreak in Japan with the demographic, socioeconomic, and geographical attributes, and changes in work situation and perception of anxiety, simultaneously.

We found that overall, people became inactive during the first wave of the outbreak; based on the direct relationships estimated by path analysis, especially younger individuals and those living in high-density neighborhoods were more likely have decreased walking behavior. Previous studies have shown a similar trend in outing and walking behaviors (Adachi et al., 2021, Hanibuchi et al., 2021, Hino and Asami, 2021), and this could be attributed to more walks for daily activities before the pandemic among younger individuals and residents in urban area, thereby resulting in a significant decrease in the number of steps. Additionally, sedentary behavior clearly increased among females, similar to previous reports that showed that females became inactive during the outbreak in Japan (Hanibuchi et al., 2021, Hino and Asami, 2021, Makizako et al., 2021). Containment measures such as school closure or self-isolation may have increased the burden of housework on females (Hanibuchi et al., 2021, Power, 2020), making them more inactive.

As for the indirect relationships between individual attributes and the changes in physical activity via changes in work situation, individuals of higher SES, such as undergraduate/graduate school graduates and white-collar job workers, were more likely to implement preventive measures like work-from-home or standby-at-home, which were associated with decreased step counts and increased sedentary behavior. This could be because of decreased walking needed for commuting. Moreover, those living in high-density neighborhoods were more likely to implement work-from-home or standby-at-home strategies. This is expected as companies not requiring on-site work, such as information technology companies, are normally located in urban areas. Moreover, it is suggested that those with incomes<3 million yen were more likely to experience decreased work causing decreased step counts and increased sedentary behavior. Therefore, individuals with lower SES were more likely to experience economic problems and decreased physical activity simultaneously. Previous studies have revealed that unemployment is often associated with deterioration of mental health (Paul and Moser, 2009) and increased smoking and drinking behaviors (Montgomery et al., 1998). Individuals with lower SES with decreased work would have particularly high health risks during the pandemic.

Our results showed that changes in physical activity were associated with perception of anxiety, similar to several previous studies (Pieh et al., 2020, Silva et al., 2020). The model assuming that decreased physical activity causes increased anxiety better explained the comprehensive relationships compared to the model assuming that anxiety causes physical inactivity or the models assuming direct relationships between the work situation changes and anxiety. Increased anxiety during the pandemic is a crucial concern as a previous study demonstrated that COVID-19 related anxiety was strongly associated with functional impairment, alcohol or drug coping, negative religious coping, extreme hopelessness, and passive suicidal ideation (Lee, 2020). Our path model showed that females, younger individuals, and those living in high-density neighborhoods were more likely to experience decreased physical activity associated with anxiety. During the pandemic in Japan, several studies reported an increase in the number of females committing suicide (Nomura et al., 2021, Tanaka and Okamoto, 2021) and high urbanization to be associated with severe psychological distress and new-onset suicidal ideation (Okubo et al., 2021). These indicated that decreased physical activity could be one of the factors affecting such serious mental health problems during the pandemic in Japan.

Furthermore, the variables of anxiety regarding spreading the infection to others and stigma associated with going out were related to decreased walking behavior while the variable of anxiety about getting infected was not. This may be because in Japanese society with tight social norms (Gelfand et al., 2021), people are more anxious about disrupting social harmony by spreading the infection or going out than about getting infected. Although Gelfand et al. (2021) suggested that tightening social norms may mitigate the COVID-19 outbreak, in strict societies, social isolation due to decrease in opportunities to go out would increase people’s concerns about disturbing others or being criticized by others which may make people more inactive. Appropriate social norms should be considered based on overall health risks including infection, physical inactivity, and mental health.

This study has several limitations. First, the method of sample selection was not random, and the questionnaire has not been validated using external data; therefore, our findings are limited in their generalization. Second, a bias may have occurred due to the method of observation of walking behavior. Although we observed only the steps walked while the participants carried their iPhones, its frequency may vary according to personal attributes such as gender and age. We employed a simple method for quantitatively counting steps before the survey during the emergency; however, measurements by wearing an accelerometer at all times would have ideally reflected the changes in physical activity more accurately. Third, we attempted to explore the relationships between the changes in physical activity and individual attributes, the work situation changes, and perception of anxiety during the COVID-19 pandemic by comparing the path models based on several hypotheses; however, this cross-sectional study could not conclude the causality.

5. Conclusion

Physical inactivity during the COVID-19 pandemic is a serious concern causing various health problems. By examining the relationship of the changes in step count and time spent in sedentary behavior during the first wave of the outbreak in Japan to individual circumstances, the present study revealed that younger individuals, those living in high-density neighborhoods, and females were clearly associated with decreased walking behavior or increased sedentary behavior, and the changes in physical activity were associated with strong anxiety related to the pandemic. Further, while individuals with high SES were more likely to implement preventions such as work-from-home or standby-at-home, lower SES leads to decreased amounts of work, and both of those changes in work situation were related to decreased walking behavior and increased sedentary behavior. The health of people with low SES facing economic burden and females, younger individuals, and those living in urban areas who experience decreased physical activity should be continuously observed. Finally, considering that the pandemic is a changing, evolving situation, further analyses of changes in physical activity are warranted.

CRediT authorship contribution statement

Shohei Nagata: Conceptualization, Formal analysis, Methodology, Software, Writing – original draft. Hiroki M. Adachi: Methodology, Software, Writing – review & editing. Tomoya Hanibuchi: Conceptualization, Data curation, Writing – review & editing, Funding acquisition. Shiho Amagasa: Conceptualization, Writing – review & editing. Shigeru Inoue: Conceptualization, Writing – review & editing, Funding acquisition. Tomoki Nakaya: Conceptualization, Writing – review & editing, Funding acquisition, Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by JSPS KAKENHI, Japan (Grant Numbers 17H00947, 19H03910, and 20H00040) and JST-Mirai Program, Japan (Grant Number JPMJMI20B2).

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pmedr.2021.101640.

Contributor Information

Shohei Nagata, Email: shohei.nagata.r7@dc.tohoku.ac.jp.

Hiroki M. Adachi, Email: adachi.hiroki.p2@dc.tohoku.ac.jp.

Shiho Amagasa, Email: amagasa@tokyo-med.ac.jp.

Shigeru Inoue, Email: inoue@tokyo-med.ac.jp.

Tomoki Nakaya, Email: tomoki.nakaya.c8@tohoku.ac.jp.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary data 1
mmc1.docx (182.4KB, docx)

References

  1. Adachi H.M., Hanibuchi T., Nagata S., Amagasa S., Inoue S., Nakaya T. Developing a new method to measure daily steps through iPhone Health App and an Internet survey: a case study of temporal changes of step counts before and after the declaration of a state of emergency against the epidemic of COVID-19 in Japan. Res. Exerc Epidemiol. 2021 In press. [Google Scholar]
  2. Amagasa S., Kojin H., Inoue S. Mobility Trends Reports Revealed “Self-isolation Fatigue” in Japan: Use of Mobility Data for Coronavirus Disease Control. JMA J. 2020;3:272–273. doi: 10.31662/jmaj.2020-0031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Badr H.S., Du H., Marshall M., Dong E., Squire M.M., Gardner L.M. Association between mobility patterns and COVID-19 transmission in the USA : a mathematical modelling study. Lancet Infect. Dis. 2020;20:1247–1254. doi: 10.1016/S1473-3099(20)30553-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. BBC Covid: How is Europe lifting lockdown restrictions? https://www.bbc.com/news/explainers-53640249 2021 accessed August 10, 2021.
  5. Beenackers M.A., Kamphuis C.BM., Giskes K., Brug J., Kunst A.E., Burdorf A., van Lenthe F.J. Socioeconomic inequalities in occupational, leisure-time, and transport related physical activity among European adults: A systematic review. Int. J. Behav. Nutr. Phys. Act. 2012;9(1):116. doi: 10.1186/1479-5868-9-116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Courtemanche C., Garuccio J., Le A., Pinkston J., Yelowitz A. Strong Social Distancing Measures In The United States Reduced The COVID-19 Growth Rate. Health Aff. 2020;39(7):1237–1246. doi: 10.1377/hlthaff.2020.00608. [DOI] [PubMed] [Google Scholar]
  7. Cubbin C., Sundquist K., Ahlén H., Johansson S.-E., Winkleby M.A., Sundquist J. Neighborhood deprivation and cardiovascular disease risk factors: Protective and harmful effects. Scand. J. Public Health. 2006;34(3):228–237. doi: 10.1080/14034940500327935. [DOI] [PubMed] [Google Scholar]
  8. Fearnbach S.N., Flanagan E.W., Höchsmann C., Beyl R.A., Altazan A.D., Martin C.K., Redman L.M. Factors Protecting against a Decline in Physical Activity during the COVID-19 Pandemic. Med. Sci. Sport. Exerc. 2021;53:1391–1399. doi: 10.1249/MSS.0000000000002602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gelfand M.J., Jackson J.C., Pan X., Nau D., Pieper D., Denison E., Dagher M., Van Lange P.A.M., Chiu C.-Y., Wang M.o. The relationship between cultural tightness–looseness and COVID-19 cases and deaths: a global analysis. Lancet Planet. Heal. 2021;5(3):e135–e144. doi: 10.1016/S2542-5196(20)30301-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hall G., Laddu D.R., Phillips S.A., Lavie C.J., Arena R. A tale of two pandemics: How will COVID-19 and global trends in physical inactivity and sedentary behavior affect one another? Prog. Cardiovasc. Dis. 2021;64:108–110. doi: 10.1016/j.pcad.2020.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hanibuchi T., Yabe N., Nakaya T. Who is staying home and who is not? Demographic, socioeconomic, and geographic differences in time spent outside the home during the COVID-19 outbreak in Japan. Prev. Med. Reports. 2021;21:101306. doi: 10.1016/j.pmedr.2020.101306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hillsdon M., Lawlor D.A., Ebrahim S., Morris J.N. Physical activity in older women: associations with area deprivation and with socioeconomic position over the life course: observations in the British Women’s Heart and Health Study. J. Epidemiol. Community Health. 2008;62(4):344–350. doi: 10.1136/jech.2006.058610. [DOI] [PubMed] [Google Scholar]
  13. Hino K., Asami Y. Change in walking steps and association with built environments during the COVID-19 state of emergency: A longitudinal comparison with the first half of 2019 in Yokohama. Japan. Health Place. 2021;69:102544. doi: 10.1016/j.healthplace:2021.102544. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Johns Hopkins Coronavirus Resource Center COVID-19 Dashboard https://coronavirus.jhu.edu/map.html 2021 accessed August 10, 2021.
  15. Koohsari M.J., Nakaya T., Shibata A., Ishii K., Oka K. Working from Home After the COVID-19 Pandemic: Do Company Employees Sit More and Move Less? Sustainability. 2021;13:939. doi: 10.3390/su13020939. [DOI] [Google Scholar]
  16. Lau H., Khosrawipour V., Kocbach P., Mikolajczyk A., Schubert J., Bania J., Khosrawipour T. The positive impact of lockdown in Wuhan on containing the COVID-19 outbreak in China. J. Travel Med. 2020;27:taaa037. doi: 10.1093/jtm/taaa037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Lee I.-M., Shiroma E.J., Lobelo F., Puska P., Blair S.N., Katzmarzyk P.T. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380(9838):219–229. doi: 10.1016/S0140-6736(12)61031-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lee S.A. Coronavirus Anxiety Scale: A brief mental health screener for COVID-19 related anxiety. Death Stud. 2020;44(7):393–401. doi: 10.1080/07481187.2020.1748481. [DOI] [PubMed] [Google Scholar]
  19. Li X., Rudolph A.E., Mennis J. Association Between Population Mobility Reductions and New COVID-19 Diagnoses in the United States Along the Urban-Rural Gradient, February-April, 2020. Prev. Chronic Dis. 2020;17:E118. doi: 10.5888/pcd17.200241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lippi G., Henry B.M., Sanchis-Gomar F. Physical inactivity and cardiovascular disease at the time of coronavirus disease 2019 (COVID-19) Eur. J. Prev. Cardiol. 2020;27(9):906–908. doi: 10.1177/2047487320916823. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Makizako H., Akaida S., Shono S., Shiiba R., Taniguchi Y., Shiratsuchi D., Nakai Y. Physical Activity and Perceived Physical Fitness during the COVID-19 Epidemic: A Population of 40- to 69-Year-Olds in Japan. Int. J. Environ. Res. Public Health. 2021;18:4832. doi: 10.3390/ijerph18094832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. McNeill L.H., Kreuter M.W., Subramanian S.V. Social Environment and Physical activity: A review of concepts and evidence. Soc. Sci. Med. 2006;63(4):1011–1022. doi: 10.1016/j.socscimed.2006.03.012. [DOI] [PubMed] [Google Scholar]
  23. Meyer J., McDowell C., Lansing J., Brower C., Smith L., Tully M., Herring M. Changes in Physical Activity and Sedentary Behavior in Response to COVID-19 and Their Associations with Mental Health in 3052 US Adults. Int. J. Environ. Res. Public Health. 2020;17:6469. doi: 10.3390/ijerph17186469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Montgomery S.M., Cook D.G., Bartley M.J., Wadsworth M.E.J. Unemployment, cigarette smoking, alcohol consumption and body weight in young British men. Eur. J. Public Health. 1998;8:21–27. doi: 10.1093/eurpub/8.1.21. [DOI] [Google Scholar]
  25. Nagata S., Nakaya T., Adachi Y.u., Inamori T., Nakamura K., Arima D., Nishiura H. Mobility Change and COVID-19 in Japan: Mobile Data Analysis of Locations of Infection. J. Epidemiol. 2021;31(6):387–391. doi: 10.2188/jea.JE20200625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Nakaya T., Honjo K., Hanibuchi T., Ikeda A.i., Iso H., Inoue M., Sawada N., Tsugane S., Dalby A.R. Associations of All-Cause Mortality with Census-Based Neighbourhood Deprivation and Population Density in Japan: A Multilevel Survival Analysis. PLoS One. 2014;9(6):e97802. doi: 10.1371/journal.pone.0097802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Nomura S., Kawashima T., Yoneoka D., Tanoue Y., Eguchi A., Gilmour S., Kawamura Y., Harada N., Hashizume M. Trends in suicide in Japan by gender during the COVID-19 pandemic, up to September 2020. Psychiatry Res. 2021;295:113622. doi: 10.1016/j.psychres.2020.113622. [DOI] [PubMed] [Google Scholar]
  28. Office for Novel Coronavirus Disease Control, Cabinet Secretariat, Government of Japan COVID-19 Information and Resources: Declaration of the State of Emergency. 2020. https://corona.go.jp/emergency/ accessed August 10, 2021.
  29. Okubo R., Yoshioka T., Nakaya T., Hanibuchi T., Okano H., Ikezawa S., Tsuno K., Murayama H., Tabuchi T. Urbanization level and neighborhood deprivation, not COVID-19 case numbers by residence area, are associated with severe psychological distress and new-onset suicidal ideation during the COVID-19 pandemic. J. Affect. Disord. 2021;287:89–95. doi: 10.1016/j.jad.2021.03.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Paul K.I., Moser K. Unemployment impairs mental health: Meta-analyses. J. Vocat. Behav. 2009;74(3):264–282. doi: 10.1016/j.jvb.2009.01.001. [DOI] [Google Scholar]
  31. Pieh C., Budimir S., Probst T. The effect of age, gender, income, work, and physical activity on mental health during coronavirus disease (COVID-19) lockdown in Austria. J. Psychosom. Res. 2020;136:110186. doi: 10.1016/j.jpsychores.2020.110186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Power K. The COVID-19 pandemic has increased the care burden of women and families. Sustain. Sci. Pract. Policy. 2020;16(1):67–73. doi: 10.1080/15487733.2020.1776561. [DOI] [Google Scholar]
  33. Rahman M.E., Islam M.S., Bishwas M.S., Moonajilin M.S., Gozal D. Physical inactivity and sedentary behaviors in the Bangladeshi population during the COVID-19 pandemic: An online cross-sectional survey. Heliyon. 2020;6(10):e05392. doi: 10.1016/j.heliyon.2020.e05392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Robinson E., Boyland E., Chisholm A., Harrold J., Maloney N.G., Marty L., Mead B.R., Noonan R., Hardman C.A. Obesity, eating behavior and physical activity during COVID-19 lockdown: A study of UK adults. Appetite. 2021;156:104853. doi: 10.1016/j.appet.2020.104853. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Saelens B.E., Handy S.L. Built environment correlates of walking: a review. Med. Sci. Sports Exerc. 2008;40:S550–S566. doi: 10.1249/MSS.0b013e31817c67a4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Saelens B.E., Sallis J.F., Black J.B., Chen D. Neighborhood-Based Differences in Physical Activity: An Environment Scale Evaluation. Am. J. Public Health. 2003;93(9):1552–1558. doi: 10.2105/AJPH.93.9.1552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sallis J.F., Bowles H.R., Bauman A., Ainsworth B.E., Bull F.C., Craig C.L., Sjöström M., De Bourdeaudhuij I., Lefevre J., Matsudo V., Matsudo S., Macfarlane D.J., Gomez L.F., Inoue S., Murase N., Volbekiene V., McLean G., Carr H., Heggebo L.K., Tomten H., Bergman P. Neighborhood Environments and Physical Activity Among Adults in 11 Countries. Am. J. Prev. Med. 2009;36(6):484–490. doi: 10.1016/j.amepre.2009.01.031. [DOI] [PubMed] [Google Scholar]
  38. Silva L.R.B., Seguro C.S., de Oliveira C.G.A., Santos P.O.S., de Oliveira J.C.M., de Souza Filho L.F.M., de Paula Júnior C.A., Gentil P., Rebelo A.C.S. Physical Inactivity Is Associated With Increased Levels of Anxiety, Depression, and Stress in Brazilians During the COVID-19 Pandemic: A Cross-Sectional Study. Front. psychiatry. 2020;11 doi: 10.3389/fpsyt.2020.565291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tanaka T., Okamoto S. Increase in suicide following an initial decline during the COVID-19 pandemic in Japan. Nat. Hum. Behav. 2021;5(2):229–238. doi: 10.1038/s41562-020-01042-z. [DOI] [PubMed] [Google Scholar]
  40. Tison G.H., Avram R., Kuhar P., Abreau S., Marcus G.M., Pletcher M.J., Olgin J.E. Worldwide Effect of COVID-19 on Physical Activity: A Descriptive Study. Ann. Intern. Med. 2020;173(9):767–770. doi: 10.7326/M20-2665. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Ullman, J.B., Bentler, P.M., 2012. Structural Equation Modeling. Handb. Psychol. Second Ed., Major Reference Works. https://doi.org/10.1002/9781118133880.hop202023.
  42. World Health Organization Weekly epidemiological update on COVID-19 - 3 August 2021 https://www.who.int/publications/m/item/weekly-epidemiological-update-on-covid-19–-3-august-2021 2021 accessed August 10, 2021.
  43. World Health Organization Physical activity https://www.who.int/news-room/fact-sheets/detail/physical-activity 2020 accessed August 10, 2021.
  44. Yabe T., Tsubouchi K., Fujiwara N., Wada T., Sekimoto Y., Ukkusuri S.V. Non-compulsory measures sufficiently reduced human mobility in Tokyo during the COVID-19 epidemic. Sci. Rep. 2020;10:18053. doi: 10.1038/s41598-020-75033-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Yamada M., Kimura Y., Ishiyama D., Otobe Y., Suzuki M., Koyama S., Kikuchi T., Kusumi H., Arai H. Effect of the COVID-19 Epidemic on Physical Activity in Community-Dwelling Older Adults in Japan: A Cross-Sectional Online Survey. J. Nutr. Health Aging. 2020;24(9):948–950. doi: 10.1007/s12603-020-1501-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhou Y., Xu R., Hu D., Yue Y., Li Q., Xia J. Effects of human mobility restrictions on the spread of COVID-19 in Shenzhen, China: a modelling study using mobile phone data. Lancet Digit. Heal. 2020;2(8):e417–e424. doi: 10.1016/S2589-7500(20)30165-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary data 1
mmc1.docx (182.4KB, docx)

Articles from Preventive Medicine Reports are provided here courtesy of Elsevier

RESOURCES