Attributes of the built environment of the workplace neighborhood and sitting position at work and for transportation among Japanese office workers


Result variables

Sedentary behaviors

A validated Japanese questionnaire14, which seeks to assess sedentary time in six specific behaviors in three domains (related to work, transport and leisure) separately for working days and non-working days, was used (see Supplementary Table 1 for the full questionnaire ). Participants were asked to report their average daily sedentary time for each behavior during the previous week. We used three specific sedentary behaviors on work days as outcome measures: time sitting at work; time sitting in cars; and sitting time in public transport. These behaviors were considered to occur in or near participants’ work quarters, as workers performed them on work days. This questionnaire showed moderate to high test-retest reliability (intraclass correlation coefficient[ICC]= 0.83) for the work area with a recall period of one week14. The validity criterion of sedentary time in all areas for working days (rho = 0.57, p 14.

Exposure variables

Built environment of the perceived work district

The Japanese version of the Japanese version of the Short Neighborhood Scale for the Pedestrian Environment (ANEWS-J) was used to measure environmental perceptions in the workplace neighborhood. The workplace neighborhood was defined as being within a 10 to 15 minute walk of the workplace. In total, six subscales were assessed: diversity of land use (16 items), access to land use (6 items), street connectivity (3 items), availability and quality of walking infrastructure / bicycle (4 elements), aesthetics (4 elements) and criminal security (5 elements). The de Cronbach, an indicator of internal consistency, for the diversity of land use, access to land use, street connectivity, availability and quality of walking / cycling infrastructure, aesthetics and criminal security were 0.91, 0.65, 0.64, 0.72, 0.73 and 0.56, respectively. We did not include the residential density subscales, which were not applicable to the study, and road safety due to poor internal consistency (α = 0.26)15. Details of the modified ANEWS-J used in this study were provided in Supplementary Table 2. All items of the subscale were rated on a four-point scale except those intended to assess the land use diversity (six point scale). The scoring of the subscales followed ANEWS-J procedures published online ( Higher scores indicate greater walking ability. ANEWS-J was found to have acceptable test-retest reliability (ICC = 0.76-0.96) for residential areas16. We examined the test-retest reliability of ANEWS-J for the work district in a subsample of participants (n = 200). Participants reported their perceptions of their work environment in their neighborhood twice in two weeks. The test-retest reliability of ANEWS-J was moderate to high for all subscales (ICC = 0.57-0.87) (Supplementary Table 3).

Pedestrian power of the neighborhood at work measured objectively

The level of walkability in work areas was estimated using the Walk Score®. This is a measure of access to local destinations, using a function of decreasing the distance to destinations such as grocery stores, restaurants, banks, parks and schools, with a two-measure adjustment. street connectivity: density of intersections and length of blocks17. The Walk Score® can be assigned to locations (for example, postal codes or addresses) and is normalized between 0 and 100. A higher Walk Score® indicates that there are more destinations within walking distance. Walk Score® uses open source data such as Google, and Open Street Map as source data to identify relevant destinations17. The Walk Score® has been confirmed as a valid measure to assess the pedestrian potential of neighborhoods in Japan18. About 60% of participants provided their seven-digit workplace postal codes (n = 1360), with 777 unable to provide their complete workplace postal codes. Each workplace postal code was manually entered on the Walk Score® website ( to obtain the score in July-August 2020. The Walk Score® was available for 1163 participants. The website did not generate a Walk Score® for 197 participants who provided a postcode at their workplace due to limited data for spatial details for Japan. Because the Walk Score® was negatively skewed (median score = 82, 25th percentile = 63, 75th percentile = 94), we used the Walk Score® as a categorical measure. We classified participants into three groups according to the Walk Score®: car dependent (0-69); fairly practicable (70-89); and very practicable (90-100).


Individual-level covariates included sex, age group (20-29, 30-39, 40-49, or 50-59), marital status (not married or married), level of education ( higher or lower than higher education), individual annual income (19 assess the amount of physical activity in three areas (work, transportation and recreation). GPAQ data has been verified for valid responses following standard procedures provided by the World Health Organization20. The total amount of physical activity for these domains was used as a covariate. Four other participants were excluded due to missing data on total physical activity. Information on the possession of a driving license (yes / no) was also collected for sedentary behavior linked to transport. Hours worked were assessed using the question “How many hours did you work in the last 7 days?” The covariate at the workplace level was workplace size, which was measured by the self-reported number of workers in the participant’s workplace (

statistical analyzes

Differences in characteristics between categories of sub-samples were examined using Pearson’s chi-square tests for categorical variables and independent t-tests for continuous variables. Spearman’s correlation was used to examine correlations between the perceived attributes of the work area’s built environment and the Walk Score®, as the latter was skewed.

We used linear regression models to investigate associations between work neighborhood attributes and sedentary behavior at work. Unstandardized regression coefficients (β) and 95% confidence intervals (CI), corresponding to one standard deviation (SD) increment of the perceived environmental attributes, were estimated for the associations. We also calculated β and 95% CI for the Walk Score® category, using the intermediate category (rather accessible on foot) as a reference. Each attribute of the workplace neighborhood was examined individually in the models. All regression analyzes were performed using Stata 15 (Stata Corp, College Station, Texas, USA), and the level of significance was set at p


Comments are closed.