Associations of Physical and Social Workplace Characteristics with Movement Behaviors at Work

Abstract

Introduction/Purpose 

Sedentary behavior (SB) is common in desk-based work, and prolonged periods of SB are associated with negative health outcomes. This study assessed associations between workplace characteristics and setting and movement patterns during working hours.

Methods 

This secondary analysis used baseline data from the Reducing Sedentary Behavior to Decrease Blood Pressure clinical trial, which enrolled inactive, desk-based workers with elevated blood pressure (n = 271; mean age, 45.3 ± 11.6 yr; body mass index, 30.66 ± 7.1 kg·m⁻²; 59.4% women). Physical and social workplace characteristics were assessed by a study-developed questionnaire and the Office Environment and Sitting Scale. Participants also wore an activPAL activity monitor for 7 d and reported working hours in a diary to measure SB and physical activity (PA) specifically while working. Linear regression was used to analyze cross-sectional associations between workplace characteristics and SB and PA. A stratified analysis was also conducted to assess associations among home-based and in-office desk workers separately. Analyses were adjusted for age, gender, body mass index, and work wear time.

Results 

Participants spent 77% of working hours in SB. Public versus private offices, working in-office versus at home, higher local connectivity, and greater overall connectedness were associated with lower SB and/or greater PA (all P < 0.05). Higher frequency of face-to-face interactions, and greater visibility and proximity to coworkers were associated with less SB and more PA (all P < 0.05). For example, home-based workers had more total SB (+17.2 ± 8.4 min per day), more SB bouts ≥30 min (+39.1 ± 12.8 min per day), and less steps (695 ± 201 steps per day) than in-office employees. Stratification by office setting revealed differences in associations between SB and PA and workplace characteristics.

Conclusions 

More public, open spaces with more social interactions and physical walkways could improve SB and PA patterns during work. Home-based workers had more SB, less PA, and unique associations of these activities with workplace characteristics, suggesting a need for tailored interventions.

INTRODUCTION

Sedentary behavior (SB) has been linked to an increased risk of obesity, type 2 diabetes, cardiovascular disease, musculoskeletal disorders, depressive symptoms, anxiety, and stress (^1–3). SB is defined as waking activity that requires very low-energy expenditure and occurs in a seated, reclined, or lying position, often in the form of screen-based entertainment, reading, or desk-based work (⁴). Not only does the total amount of time spent in SB affect health, but emerging data suggest that a pattern of SB may also be associated with increased health risks (⁵). Accumulating SB in prolonged bouts has been more strongly associated with poor cardiometabolic health and increased risk of mortality compared with SB accumulated in shorter bouts (^6,7). Taken together, these data suggest that reducing both the total duration and patterns of prolonged SB could be important behavioral targets. SB in the workplace has become a prominent public health issue because desk-based work is a major context of excessive total and prolonged SB (⁸). Furthermore, greater workplace SB may also be associated with lower general and work-specific performance and satisfaction (^8,9).

Aspects of the physical and social work environment, such as frequency of face-to-face interactions and public versus private workspaces, are attractive intervention targets because they have been associated with greater engagement in SB and lower physical activity (PA) levels at work (^10–15). Moreover, workplace SB intervention studies suggest that interventions including environmental components (e.g., restructuring the physical environment to be more activity permissive) most consistently result in reduced workplace SB (¹⁶). However, previous observational studies have been limited by a lack of variability in workplace characteristics across participants and the absence of objective measurements of SB and activity (^17,18). Moreover, rigorous studies considering workplace characteristics and movement behaviors specifically among the growing population of home-based workers are lacking. This research gap is important given recent and sustained increases in home-based work because of the coronavirus disease 2019 (COVID-19) pandemic and resulting improvements in access to remote work capabilities (¹⁹).

To address these gaps, we used baseline data from desk workers enrolled in the Reducing Sedentary Behavior to Decrease Blood Pressure (RESET BP) randomized trial (²⁰). In contrast to cluster-randomized workplace intervention trials that enroll a group of participants at a limited number of worksites, the individual recruitment strategy in RESET BP provided more opportunity to enroll participants with unique workspaces. This design and recruitment strategy resulted in more diverse physical and social work environments across participants and uniquely included both in-office and home-based (i.e., remote) desk workers. The primary purpose of this study was to assess associations between the physical and social workplace environments and the accumulation of SB and PA among desk workers during work time. We hypothesized that offices with physical characteristics that encourage PA and social environments that promote interaction would be associated with less time in SB and more PA measured objectively during working hours. The secondary purpose was to assess differences in associations between workplace characteristics and SB/PA after stratifying the sample into in-office versus home-based worker groups. We hypothesized that these associations could be different across workplace settings.

METHODS

Participants

The participants included in this study were enrolled in the RESET BP randomized-controlled trial (NCT03307343) between December 2017 and August 2022 (²⁰). The sample size for RESET BP was determined to have sufficient statistical power to address the aims of the clinical trial. Participants were eligible for the RESET BP trial if they had desk-based occupations, reported ≥20 h·wk⁻¹ of occupational sitting time at their desk, and lived within approximately 25 miles of the city of Pittsburgh, PA. Participants also had to have elevated systolic blood pressure (BP) between 120 and 159 mm Hg or elevated diastolic BP between 80 and 99 mm Hg. Participants were not eligible if they were taking antihypertensive or blood- glucose‐lowering medication, had systolic BP ≥160 mm Hg or diastolic BP ≥100 mm Hg, had preexisting cardiovascular disease (e.g., history of myocardial infarction, angina, or peripheral artery disease), could not obtain medical clearance from a healthcare provider to participate, were currently or recently pregnant, had bariatric surgery in the past year, or reported leisure-time PA that met or exceeded current aerobic guidelines (i.e., ≥150 min of moderate-intensity PA per week) (²¹). For the current analysis, participants were excluded if they did not have valid SB and PA at baseline as described hereinafter (n = 2), which resulted in a final analytic sample size of n = 271. All research procedures were approved by the University of Pittsburgh’s Human Research Protection Office (STUDY19030297), and informed consent was obtained from each participant before participating in the study.

Study Design

This study is a secondary analysis of baseline data from the RESET BP trial. The primary aim of RESET BP is to assess the effect of a SB-reduction intervention on resting BP and cardiovascular health. In the current analysis, we used questionnaire and objective activity monitoring data collected at baseline to evaluate cross-sectional associations of the workplace environment with workday SB and PA.

Measurements—Participant Characteristics

Participant age and gender were self-reported. Participants were also asked to report their job characteristics, including sector (e.g., academia) and job type (e.g., sales or service). Body mass index (BMI; in kilograms per meter squared) was calculated using height measured by a wall-mounted stadiometer (Perspective Enterprises, Portage, MI), and weight was measured by a calibrated digital scale (Tanita BWB-800, Arlington Heights, IL).

SB and PA

SB and PA were objectively measured for each participant for 7 d at baseline using a thigh-worn activPAL3 micro activity monitor (PAL Technologies, LTD, Glasgow, Scotland). SB and activity data were considered valid if at least 5 d with at least 10 h of waking wear time were collected (²²). Participants also completed a diary during the monitor wear period in which they recorded work, nonwork, sleep, and nonwear periods to assist in data scoring using recommended procedures (^23,24). Objective data were downloaded, exported as events using PAL Technologies software (version 7.2.38), and scored using a semiautomated, diary-informed process. Objective data were collected during times reported as working, durations of work wear time spent in any SB (total SB), in bouts of SB that were ≥30 min (SB30) and ≥60 min (SB60), standing, stepping, and total monitor wear time, as well as sit-to-stand transitions and step counts, and were averaged over valid workdays.

Physical and Social Workplace Environment

A workplace environment questionnaire was developed by the RESET BP investigators to measure relevant attributes of the participants’ physical and social workplace environment (e.g., access to common areas, face-to-face interactions) that have been linked with SB and PA in previous research. Physical characteristics that were measured included workstation environment, printer location, trash bin location, restroom location, description and use of stairwells and elevators, and access to common areas/break room (^12–15). Social characteristics assessed included supervisor support of breaks, face-to-face interactions, and socializing at work or outside of work (^13,25,26). The questions from the instrument used to measure these characteristics are available in Supplemental Content 1 (table, https://links.lww.com/TJACSM/A206). In-office or home-based work setting was assessed verbally and recorded by the assessment team using the worksite location evaluation (Supplemental Content 1, table, https://links.lww.com/TJACSM/A206). In addition, the 18-item Office Environment and Sitting Scale (OFFESS) questionnaire was used to further understand the participants’ workplace environment and how it related to their overall SB and PA during working hours (¹⁰). The OFFESS measures four distinct subscales: local connectivity, overall connectedness, proximity of coworkers, and visibility of coworkers. The first two subscales assess the physical environment, whereas the latter two subscales evaluate the social environment. More detailed definitions of these subscales can be found in Supplemental Content 2 (table, https://links.lww.com/TJACSM/A207). Each subscale consisted of 3–6 questions and is calculated as the average of all items within the subscale. The scoring range was from 1 to 4, with 1 meaning “strongly disagree” and 4 meaning “strongly agree.” The questionnaire has established reliability and has been associated with frequency of breaks in SB among office workers (^10,11).

Statistical Analysis

Participant characteristics and objectively measured SB and PA accumulated during work wear time were summarized using means and standard deviations or frequencies and percentages. Linear regression analyses determined cross-sectional associations of the physical and social workplace environment characteristics with SB and PA. Each variable was analyzed in categorical, Likert-scale, or yes/no formats. The categorical and yes/no variables had a reference answer that we hypothesized would have the least PA/most SB, whereas the Likert-scale variables were considered continuous. We also conducted a stratified analysis repeating all regression models for a home-based strata and office-based strata. Regression models were adjusted for work wear time, age, BMI, and gender. Associations were considered statistically significant if P ≤ 0.05. All analyses were conducted using Stata version 16 (StataCorp, College Station, TX).

RESULTS

Participant Characteristics

Descriptive characteristics of the participants are displayed in Table 1. Participants were, on average, middle-aged (45.1 ± 11.6 yr) and classified as obese (BMI = 30.6 ± 7.1 kg·m⁻²). Most of the participants were female (59.4%) and White/Caucasian (83.4%), and 97.4% of the participants had at least some college/associates degree. About half of participants (52.4%) self-reported working in academia, with the remaining participants working in industry (39.5%) and government (8.1%) occupations. Most (63.3%) had professional/managerial job types and worked primarily in an office (66.3%) rather than at home.

TABLE 1 - Participant Characteristics (n = 271).

	Mean ± SD or n (%)
Demographics
Age, yr	45.1 ± 11.6
BMI, kg·m−2	30.6 ± 7.1
Gender
Male	109 (40.2%)
Female	161 (59.4%)
Other	1 (0.4%)
Race
White/Caucasian	226 (83.4%)
Black/African American	22 (8.1%)
Other	23 (8.5%)
Education level
High school graduate or GED	3 (1.1%)
Vocational or training school after high school	4 (1.5%)
Some college or associate degree	35 (12.9%)
College graduate or baccalaureate degree	104 (38.4%)
Masters or doctoral degree (PhD, MD, JD, etc.)	125 (46.1%)
Work Variables
Job sector
Academia	142 (52.4%)
Industry	107 (39.5%)
Government	22 (8.1%)
Job type
Sales or service	18 (6.7%)
Clerical or administrative support	63 (23.3%)
Professional, managerial, or technical	171 (63.3%)
Other	18 (6.7%)
Worksite location
Home	91 (33.7%)
Office	179 (66.3%)
Activity levels during working hours
SB, min·d−1	397.5 ± 84.6
SB30, min·d−1	238.8 ± 106.8
SB60, min·d−1	124.8 ± 94.8
Standing, min·d−1	83.2 ± 58.1
Stepping, min·d−1	33.5 ± 19.0
Sit-to-stand transitions per day	25.7 ± 10.9
Total steps per day	2795.6 ± 1641.9
Total work time per day, h	8.6 ± 1.3
Workdays with monitor wear	4.7 ± 1.1

GED, General Educational Development Test; SD, standard deviation.

During working hours, the average duration of SB was nearly 400 min, which equates to 6.6 h and 77.3% of work time. Participants spent an average of 238.8 min (4.0 h) in SB30 and 124.8 min (2.1 h) in SB60 per day during working hours. Standing only encompassed an average 16.2% of the workday and stepping accounted for an average of 6.5% of the workday. Participants accumulated 26 sit-to-stand transitions per day and took an average of 2796 steps per day during working hours.

Associations of Physical Workplace Characteristics with SB and PA

The first analysis investigated associations between physical workplace characteristics and SB and PA during working hours (Tables 2, 3). Having a public office was associated with less work time spent in SB30 (−45.5 min, P < 0.01) and SB60 (−37.5 min, P = 0.01) when compared with those with private offices. Using a shared printer was associated with 27.3 min less work time spent in SB60 when compared with those with a personal printer or no printer (P = 0.03). Interestingly, using an elevator to change floors in the workplace was associated with more work time spent stepping (+7.0 min, P = 0.03) and more total steps (+777.4, P = 0.01) compared with not changing floors at all. However, using stairs to change floors was not associated with differences in work SB or PA. Having better access to a break/common area was associated with 558.3 more total steps and 6.3 min more work time stepping (P = 0.03 and 0.04, respectively). Those who worked at an office spent less work time in SB (−17.2 min, P = 0.04), SB30 (−39.1 min, P < 0.01), and SB60 (−41.3 min, P < 0.01) when compared with those who worked at home. In addition, office-based workers accumulated more steps during work hours than those working from home (+695.4 steps, P < 0.01). With higher local connectivity from OFFESS, participants spent ~23 min less in bouts of SB while also accumulating 319.1 more steps during working hours (both P < 0.05). Greater overall connectedness from OFFESS was also associated with 21 min less work time spent in SB60 (P = 0.02), 4.0 min more work time spent stepping (P < 0.01), and 488.2 more steps (P = 0.02). Trash can location, restroom location, and stairwell location were not associated with SB and PA patterns during the workday (all P > 0.05).

TABLE 2 - Associations of Physical Workplace Characteristics and Durations of SB and PA During the Workday (n = 271).

	SB (Minutes Per Workday)	SB30 (Minutes Per Workday)	SB60 (Minutes Per Workday)	Standing Time (Minutes Per Workday)	Stepping Time (Minutes Per Workday)
Office type
Private (n = 156)	Reference	Reference	Reference	Reference	Reference
Semiprivate (n = 46)	−12.7 ± 11.0	−25.1 ± 16.8	−22.8 ± 15.5	9.0 ± 9.4	3.7 ± 3.1
Public (n = 68)	−17.4 ± 9.7*	−45.5 ± 14.8**	−37.5 ± 13.7**	15.1 ± 8.3*	2.4 ± 2.7
Printer location
Personal printer (n = 81)	Reference	Reference	Reference	Reference	Reference
Shared printer (n = 164)	−14.1 ± 9.0	−26.8 ± 13.9*	−27.3 ± 12.8**	12.3 ± 7.7	1.3 ± 2.5
Don’t use printer (n = 25)	13.9 ± 15.5	19.7 ± 23.8	11.4 ± 22.0	−9.4 ± 13.2	−4.0 ± 4.3
Trashcan location
Personal trash (n = 237)	Reference	Reference	Reference	Reference	Reference
Shared trash (n = 33)	−13.7 ± 12.7	−19.2 ± 19.6	−9.2 ± 18.1	9.9 ± 10.8	4.0 ± 3.5
Restroom location
Same floor (n = 252)	Reference	Reference	Reference	Reference	Reference
Different floor (n = 18)	−5.4 ± 16.1	−1.8 ± 24.8	10.7 ± 22.9	3.9 ± 13.7	1.7 ± 4.5
Traveling to different floor
Don’t change (n = 57)	Reference	Reference	Reference	Reference	Reference
Stairs (n = 138)	−9.9 ± 10.5	−6.0 ± 16.1	2.6 ± 14.9	8.9 ± 8.9	1.2 ± 2.9
Elevator (n = 71)	−16.6 ± 11.8	−27.5 ± 18.1	−21.5 ± 16.7	9.6 ± 10.0	7.0 ± 3.3**
Stairwell description
No stairwells (n = 31)	Reference	Reference	Reference	Reference	Reference
Attractive (n = 170)	10.4 ± 9.8	16.7 ± 14.9	13.1 ± 13.8	−7.7 ± 8.3	−2.4 ± 2.7
Unattractive (n = 62)	9.4 ± 14.5	−8.1 ± 22.1	−9.5 ± 20.5	−5.6 ± 12.3	−3.7 ± 4.0
Stairwell location
Can’t see stairs at desk (n = 105)	Reference	Reference	Reference	Reference	Reference
Can see stairs at desk (n = 137)	6.9 ± 8.6	13.4 ± 13.0	12.1 ± 12.3	−4.1 ± 7.3	−3.1 ± 2.4
Access to common/break room
No (n = 44)	Reference	Reference	Reference	Reference	Reference
Yes (n = 225)	−11.7 ± 10.9	−18.6 ± 16.8	−14.8 ± 15.5	4.9 ± 9.3	6.3 ± 3.0**
Worksite location
Home (n = 91)	Reference	Reference	Reference	Reference	Reference
Office (n = 179)	−17.2 ± 8.4**	−39.1 ± 12.8**	−41.3 ± 11.8**	13.0 ± 7.1*	3.9 ± 2.3
OFFESS subscales
Local connectivity (n = 266)	−12.2 ± 6.8*	−23.5 ± 10.4**	−23.0 ± 9.5**	9.7 ± 5.7*	2.1 ± 1.9
Overall connectedness (n = 266)	−12.4 ± 6.4*	−16.7 ± 9.8*	−21.3 ± 9.0**	8.3 ± 5.4	4.0 ± 1.8**

Results are reported as adjusted β ± standard error with adjustment for age, BMI, gender, and work wear time.

*0.05 ≤ P ≤ 0.10

**P ≤ 0.05.

TABLE 3 - Associations of Physical Workplace Characteristics with Steps and Sit-to-Stand Transitions (n = 271).

	Steps Per Workday	Sit-to-Stand Transitions Per Workday
Office type
Private (n = 156)	Reference	Reference
Semiprivate (n = 46)	319.4 ± 267.6	−0.04 ± 1.7
Public (n = 68)	246.9 ± 236.2	3.5 ± 1.5*
Printer location
Personal printer (n = 81)	Reference	Reference
Shared printer (n = 164)	205.5 ± 219.1	1.1 ± 1.4
Don’t use printer (n = 25)	−451.9 ± 376.5	2.2 ± 2.5
Trashcan location
Personal trash (n = 237)	Reference	Reference
Shared trash (n = 33)	261.9 ± 307.5	2.3 ± 2.0
Restroom location
Same floor (n = 252)	Reference	Reference
Different floor (n = 18)	126.5 ± 389.6	−0.8 ± 2.6
Traveling to different floor
Don’t change (n = 57)	Reference	Reference
Stairs (n = 138)	221.5 ± 251.2	−2.3 ± 1.6
Elevator (n = 71)	777.4 ± 281.8*	−0.1 ± 1.8
Stairwell description
No stairwells (n = 31)	Reference	Reference
Attractive (n = 170)	−297.3 ± 235.1	−1.5 ± 1.5
Unattractive (n = 62)	−516.6 ± 349.5	4.6 ± 2.2*
Stairwell location
Can’t see stairs at desk (n = 105)	Reference	Reference
Can see stairs at desk (n = 137)	−319.3 ± 207.2	−1.8 ± 1.2
Access to common/break room
No (n = 44)	Reference	Reference
Yes (n = 225)	558.3 ± 262.3*	1.7 ± 1.7
Worksite location
Home (n = 91)	Reference	Reference
Office (n = 179)	695.4 ± 200.5*	2.1 ± 1.3
OFFESS subscales
Local connectivity (n = 266)	319.1 ± 162.2*	1.6 ± 1.1
Overall connectedness (n = 266)	488.2 ± 151.6*	0.2 ± 1.0

Results are reported as adjusted β ± standard error with adjustment for age, BMI, gender, and work wear time.

*P ≤ 0.05.

We then stratified the sample into those working from home and those working in an office and repeated the same models relating physical workplace characteristics to workday SB and PA. Participant characteristics are reported by strata in Supplemental Content 3 (table, https://links.lww.com/TJACSM/A208). In both strata (Supplemental Contents 4 and 5, tables, https://links.lww.com/TJACSM/A209 and https://links.lww.com/TJACSM/A210), and similar to the main results, those who perceived their workspace to be more public (e.g., cubicle, open space) spent less time in SB30 and SB60, although this was only statistically significant in the office-based workers’ strata. Using a shared versus personal printer became significantly associated with less SB and more steps in the home-based strata, which may reflect sharing with family members versus having a personal printer in a home office, but there were no significant associations in the office-based strata. There were more steps accumulated with access to common areas/break rooms among the office-based workers only. Local connectivity and overall connectedness were not associated with SB and PA in either of the stratified analyses, although they were generally in the same direction as the main analysis. However, overall connectedness was significantly associated with greater steps in the office-based strata only, similar to the main analysis.

Associations of Social Workplace Characteristics with SB and PA

We used regression models to evaluate the association between social workplace characteristics and SB and PA (Tables 4, 5). Higher frequency of face-to-face interactions with coworkers was associated with less work time spent in total SB (−6.2 min), SB30 (−14.8 min), and SB60 (−13.7 min) and higher standing time (+4.9 min), steps (+143.5 per day), and sit-to-stand transitions (+1.5 per day). Higher visibility of coworkers was associated with 15.8 min less spent in SB60, whereas closer proximity to coworkers was associated with less SB30 (−16.2 min) and SB60 (−16.7 min) and 6.8 min more standing during work time (all P < 0.05). Social workplace characteristics that were not associated included supervisor support for breaks, socializing with coworkers at work, and socializing with coworkers outside of work (all P > 0.05).

TABLE 4 - Associations of Social Workplace Characteristics and Durations of SB and PA During the Workday (n = 270).

	SB (Minutes Per Workday)	SB30 (Minutes Per Workday)	SB60 (Minutes Per Workday)	Standing Time (Minutes Per Workday)	Stepping Time (Minutes Per Workday)
Breaks support from supervisor (n = 270)	2.0 ± 4.8	7.7 ± 7.4	7.2 ± 6.8	−1.0 ± 4.1	−1.0 ± 1.3
Face-to-face interactions (n = 270)	−6.2 ± 2.6*	−14.8 ± 4.0*	−13.7 ± 3.7*	4.9 ± 2.2*	1.2 ± 0.7
Socializing at work (n = 270)	−0.6 ± 3.2	−4.4 ± 4.9	−6.5 ± 4.5	1.2 ± 2.7	−0.4 ± 0.9
Socializing outside of work (n = 270)	2.3 ± 2.8	−2.6 ± 4.3	−3.5 ± 4.0	−2.3 ± 2.4	0.1 ± 0.8
OFFESS subscales
Visibility of coworkers (n = 267)	−4.4 ± 4.9	−12.7 ± 7.4**	−15.8 ± 6.8*	3.5 ± 4.1	0.8 ± 1.3
Proximity to coworkers (n = 267)	−7.7 ± 4.0**	−16.2 ± 6.1*	−16.7 ± 5.6*	6.8 ± 3.4*	0.9 ± 1.1

Results are reported as adjusted β ± standard error with adjustment for age, BMI, gender, and work wear time.

*P ≤ 0.05.

**0.05 ≤ P ≤ 0.10.

TABLE 5 - Associations of Social Workplace Characteristics with Steps and Sit-to-Stand Transitions (n = 270).

	Steps Per Workday	Sit-to-Stand Transitions Per Workday
Breaks support from supervisor (n = 270)	−111.9 ± 115.7	−1.1 ± 0.8
Face-to-face interactions (n = 270)	143.5 ± 64.1*	1.5 ± 0.4*
Socializing at work (n = 270)	−2.7 ± 76.5	0.7 ± 0.5
Socializing outside of work (n = 270)	47.6 ± 68.2	0.7 ± 0.4
OFFESS subscales
Visibility of coworkers (n = 267)	164.4 ± 116.8	1.1 ± 0.8
Proximity to coworkers (n = 267)	164.0 ± 96.9**	0.8 ± 0.6

Results are reported as adjusted β ± standard error with adjustments for age, BMI, gender, and work wear time.

*P ≤ 0.05.

**0.05 ≤ P ≤ 0.10.

In the stratified analyses, the associations of more face-to-face interactions with lower total and prolonged SB, as found in the main analysis, were only observed among home-based workers and were not significant in office-based workers (all P ≤ 0.04; Supplemental Contents 6 and 7, tables, https://links.lww.com/TJACSM/A211 and https://links.lww.com/TJACSM/A212). A similar pattern, albeit nonsignificant, was observed for the visibility and proximity to coworkers subscales, where higher proximity/visibility was inversely associated with SB only among home-based workers.

DISCUSSION

This cross-sectional analysis provides further, and some novel, evidence that the physical and social characteristics of a workplace are associated with SB and PA during working hours. In relation to physical workplace characteristics, those who had public office spaces and higher perceived connectedness in the workplace spent significantly less time in prolonged bouts of SB than those who worked in more private or less connected offices. We also found that higher perceived connectedness was associated with less time in prolonged SB and increased steps/stepping time, whereas access to common/break rooms was associated with more steps/stepping time during working hours. Interestingly, only using an elevator (and not using stairs) was associated with more stepping time and total steps during working hours when compared with not changing floors. We speculate that we might have captured a phenomenon where individuals who use elevators are in larger buildings that require more movement overall during the workday. Lastly, those who worked in an office spent significantly less time in prolonged SB and accumulated more steps during working hours when compared with those who worked from home. Social constructs within the workplace also had a significant association with SB and PA. Higher frequency of face-to-face interactions was associated with less SB and more PA, whereas greater visibility and closer proximity to coworkers were related to reduced bouts of SB during working hours. However, it should be noted that face-to-face interactions may differ for home workers given they could characterize these interactions as video calls (e.g., Zoom calls). These results were consistent with our overall hypothesis that physical and social workplace characteristics that promote and permit movement and interaction would be associated with increased PA and reduced SB during working hours. Furthermore, we offer some new evidence that SB and PA are less favorable among desk workers that work from home versus in office.

There are few other studies that comprehensively evaluate associations between both physical and social office characteristics and workday SB and PA. Specifically, with respect to physical office characteristics, a 2020 systematic review by Zhu et al. (¹⁸) of 87 studies identified workplace characteristics commonly associated with SB and PA. Although the office characteristics we studied sometimes differed, our findings were generally consistent with the findings from observational studies included in the review. Zhu et al. (¹⁸) concluded that office spaces that were open and more public, included shared spaces and equipment, and provided opportunities for interaction were typically associated with less time spent in SB and higher PA. In another recent systematic review by Sugiyama et al. (¹⁷) that included 20 studies, some studies found that public office spaces were associated with less time spent in total and prolonged SB when compared with closed or private office spaces, which is consistent with our findings. Intervention studies included in the review by Zhu et al. (¹⁸) found that modifying physical workplace characteristics, such as adding communal spaces or relocating to an activity-promoting building, typically resulted in favorable changes in activity patterns. It is important to note that the review consistently identified that activity-encouraging workstations (e.g., sit–stand desks) commonly reduced SB and increased standing in both observational and intervention studies (¹⁸). We could not investigate this factor because the participants in RESET BP were not allowed to be using a sit–stand workstation upon enrollment in the intervention study, which included the provision of a sit–stand desk. Despite this, the current analysis was able to control for important confounding factors including age, gender, and BMI, which addresses a limitation of previous literature identified by the review (¹⁸). Furthermore, the strength of our conclusion that a more public, connected, and open physical work environment seems to promote a healthier activity pattern is bolstered by the evaluation of physical characteristics across many workspaces of individually recruited participants and by the objectively measured PA and SB specifically during working hours.

Little research has evaluated associations between social workplace environments and activity patterns. In our study, more face-to-face interactions was the social workplace characteristic most consistently associated with more favorable activity patterns. Few previous studies directly assess this relationship, especially with SB. Sugiyama et al. (²⁷) conducted a cross-sectional analysis in 2017 of movement behaviors and face-to-face interactions during working hours. The authors detected a significant association where the frequency of face-to-face interactions was 20% lower with each 1-h increase in sitting time (²⁷). Similarly, Mullane et al. (¹³) assessed associations of social workplace factors and SB by activPAL in a study of 24 workplaces and 478 participants. That study found that greater face-to-face interactions were associated with significantly less SB during working hours (¹³). A possible limitation of the current study is that participants’ definitions of face-to-face interactions and visibility/proximity to coworkers may have been defined differently in home-based versus in-office environments. Having video conference calls with a coworker may be seen as face-to-face interactions, although the question intends to assess physical closeness. Future studies should delineate between video and in-person contacts, which may have opposing effects on activity patterns, and the results of this study should be considered with this caveat.

Our study contributes to a novel understanding of how remote/home-based work may influence SB and PA patterns during working hours. The current analysis, which has a sizable home-based worker sample because of the recruitment timing that spanned the COVID-19 pandemic (December 2019 through August 2022), supports the hypothesis that those who work in office spaces spend significantly less time in objectively measured prolonged and total SB and accumulate significantly more steps than those who work from home. A few studies during the early months of the COVID-19 pandemic found that home-based workers self-reported increased SB and decreased PA during lockdown periods (^28,29). The current study provides more generalizable conclusions compared with the lockdown studies because our participants were not specifically recruited or measured during this unique event. Given that the COVID-19 pandemic has significantly changed workplace practices such that many desk-based employees can either choose to or are required to work remotely, the policy implications of this study’s findings are that workplaces may need programs to support PA and reduce SB for home-based employees (^30,31). Given our results from the stratified analyses, we can also infer that intervention approaches will have to be tailored to address home-based workers. For example, using a shared versus personal printer was associated with less SB and more steps in home-based workers but not in-office workers. It is possible that modifying this factor by moving the printer to another room in the house may be a more effective strategy at home than in-office. Given that some differences in associations were found in the office-based versus home-based group, formative research to improve understanding of the unique barriers and facilitators of workday SB and PA for remote workers is needed before developing interventions. Future research should also consider that questionnaires developed to measure workplace characteristics among in-office workers may need to be validated or modified for use in home-based workers.

Strengths and Limitations

This study benefits from having a large sample of desk-based workers, each with a unique work situation, and from the measurement of detailed information on physical and social workspace characteristics. The sample was enriched from having continued recruitment and assessment throughout the COVID-19 pandemic, which resulted in the inclusion of a significant number of home-based workers. In addition, objective SB and PA data using best-practice assessment methodology via a thigh-worn accelerometer along with detailed time-use diaries facilitated more precise analyses of working-time movement patterns.

This study is limited by the cross-sectional nature of the analysis. As temporality is not established, causality cannot be inferred. Also, generalizability is limited because of the fact that the participants in the sample all had elevated BP, were inactive by self-report, and were not using a sit–stand desk because of the inclusion criteria for the RESET BP clinical trial. It should be noted that the questionnaires used for assessing work environment characteristics were not specifically developed with home-based work in mind; therefore, the validity may be limited when addressing the characteristics of home-based workers. Also, caution is advised as this analysis may be susceptible to type I error because of the many hypothesis tests conducted; to reduce this possibility, each tested association was hypothesis driven and based on previous literature. Thus, the conclusions from this study may not generalize to a broader sample of office workers.

There are several indications from the findings of this study. Businesses interested in decreasing SB and encouraging PA among their workforce could seek to create more connected, public, and activity-permissive workspaces. In addition, we found that home-based workers had more SB and less PA during working hours than those in office spaces. To combat this with effective interventions, we must further understand the differences in the workplace environment and the most common determinants of SB and PA in a home setting. Future studies could also expand to consider 24-hour behaviors because these may be a more relevant measure for home-based workers who may have less structured workdays, more free time (secondary to commuting less), and different leisure time movement and sleep behaviors.

Conclusions

We found that SB during working hours was significantly associated with physical and social workplace characteristics. An overarching theme of these cross-sectional findings is that having more public, open, and activity-permissive workspaces that encourage social interaction seems to contribute to improved SB and PA patterns during working hours. Our data suggest a less favorable SB and PA pattern during the workday for home-based compared with in-office work settings. Future research studies should focus on identifying influential components of the working environment when working from home to design and test targeted interventions.