key: cord-0901922-yn2jybjq
authors: Okawara, M.; Ishimaru, T.; Tateishi, S.; Hino, A.; Tsuji, M.; Ikegami, K.; Nagata, M.; Matsuda, S.; Fujino, Y.
title: Association between the physical work environment and work functioning impairment while working from home under the COVID-19 pandemic in Japanese workers
date: 2021-03-26
journal: nan
DOI: 10.1101/2021.03.23.21254207
sha: 8a71871e29694657520e3f7581b81b413da31cc5
doc_id: 901922
cord_uid: yn2jybjq

Objective: This study examined the relationship between the physical work environment and work functioning impairment while working from home in the context of rapid changes associated with the COVID-19 pandemic. Methods: This cross-sectional study of internet monitors was conducted between December 22 and 26, 2020. Of a total of 33,302 participants, 5,760 who worked from home at least 1 day a month, excluding those who met the exclusion criteria, were included in the analysis. A binary subjective assessment of the physical work environment while working from home was used as an exposure factor. We examined 9 items related to the physical work environment, including level of illuminance and use of suitable desks and chairs, traditionally recommended for health and safety management when working at a desk. The number of non-conformities to 7 items was also used as an exposure factor. The presence of severe work functioning impairment was measured using the Work Functioning impairment Scale (WFun), a self-reported outcome measure of the degree of work functioning impairment. Odds ratios of severe work functioning impairment were estimated using mixed-effects logistic regression analysis with the prefecture of residence as a random effect. Results: Multivariate analysis showed that the odds ratio of severe work functioning impairment was significantly higher among those who indicated "No" to all recommended items except for "I work at a desk/chair for office use." The highest odds ratio of work functioning impairment was associated with a "No" response to "There is enough light to do my work" (aOR: 2.02, 95%CI: 1.73-2.35, p<0.01). Our results also suggest the presence of a dose-response relationship between the number of non-conformities to recommendations for work environments while working from home and work functioning impairment. Conclusions: Our findings suggest that it is important for both companies and individual workers to create a work environment that prevents negative health outcomes and improves productivity while working from home.

significantly higher among those who indicated "No" to all recommended items except for "I work at 48 a desk/chair for office use." The highest odds ratio of work functioning impairment was associated 49

Introduction 60

The expansion of telecommuting brought about by the COVID-19 pandemic is one of the 61 most marked changes to the way people work around the world in decades. In February 2020, the 62 Japanese government issued a basic policy describing countermeasures against COVID-19, in which 63 it included a recommendation for companies to implement telework to prevent the spread of infection 64

(1). Furthermore, as the disease continued to spread throughout Japan, the government declared a state 65 of emergency in April 2020, requesting people to refrain from going out, and further promoting 66 telework (2). As a result, the number of companies adopting telecommuting and the percentage of 67 workers working from home have increased rapidly (3). Further, even after the state of emergency 68 ended, more and more companies continued to employ coronavirus countermeasures, mixing going to 69 work and working from home, based on the situation in each region (4). 70

Due to the unexpected emergence of COVID-19, neither companies nor workers were able 71 to sufficiently prepare their home environments for conducting work. While work-from-home systems 72 had been discussed as a way to counter overwork and promote work-life balance in Japan, few 73 effective efforts had been made (5). However, with the COVID-19 pandemic, companies were forced 74 to rapidly adopt telework at the request of the national and local governments. As a result, many 75 workers have had to work from home in environments that are not fully equipped with the necessary 76 systems or facilities, and that are suboptimal for work execution. In 2019, about 20% of companies in 77 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted March 26, 2021. ; Japan had adopted telework. In contrast, in March 2020, 24% of companies in Tokyo had adopted 78

telework, jumping to about 57% in January 2021 (4,6). 79

Working from home is expected to have an impact on workers' health and work productivity. 80

Working in environments that are not properly designed and working with poor posture can affect the 81 musculoskeletal system (7, 8, 9) . In addition, there are concerns that individuals are working longer due 82 to insufficient management of working hours. Further, reduced opportunities for direct communication 83

can make it difficult for individuals to receive support from superiors and colleagues, and for managers 84 to manage the situation. These factors have been linked to feelings of loneliness, isolation, and 85 depressed mood (10,11,12). 86

Workers' physical work environment while working from home is important for their health 87 and work productivity. There are several basic recommendations for the working-from-home 88 environment: have enough space to work; control the illumination of text, keyboard surfaces, and 89 displays; prevent glare from displays; provide windows and other ventilation equipment; maintain a 90 constant temperature and humidity in the room; reduce noise; and prepare necessary equipment, 91 including desks and chairs, to enable a good work posture and provide an effective workspace 92 (13,14,15). 93

However, because working from home has rapidly become the new way of working in Japan during 94 the COVID-19 pandemic, it is not known how the physical work environment while working from 95 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in home affects worker's health and productivity. In fact, many companies have no control over workers' 96 work environment while they are working from home. A previous survey reported that, of the 1,256 97 companies in Japan that had introduced telework, 75.8% indicated that they did not check any 98 components of workers' physical work environment, such as air conditioning, lighting, and desk and 99 chair condition, while they were working from home (16). Working from home in an environment that 100

is not properly maintained can cause work functioning impairment through worker's health problems. 101

The purpose of this study was to investigate the relationship between the physical work 102 environment while working from home and work functioning impairment in the context of rapid 103 changes associated with the COVID-19 pandemic. 104 105 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in 

The survey was conducted between December 22 and 26, 2020. A total of 33,302 workers 113 between the ages of 20 and 65 years at the time of the survey were included, and sampling was 114 designed such that participants' sex and occupation (office workers and non-office workers) were 115 balanced by region of residence, based on the cumulative infection rate of COVID-19. Excluding those 116 who were deemed to have provided inappropriate responses, the total number of participants in this 117 study was 27,036. Details of the exclusion criteria are described in the protocol (17). Briefly, we 118 excluded those who had extremely short response times, were shorter than 140 cm tall, weighed less 119 than 30 kg, or provided inconsistent responses to multiple identical questions. In this study, workers 120 who were working from home at least 1 day a month were included in the analysis. 121 122 Assessment of physical work environment while working from home 123 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.23.21254207 doi: medRxiv preprint was used as an exposure factor. The following nine items were examined: "There is a place/room 125 where I can concentrate," "There is enough light to do my work," "There is enough space on my desk 126 to work," "There is enough space at my feet," "The temperature and humidity in the room are 127 comfortable," "The environment is quiet," "I work at a desk/chair for office use," "I work at a 128 desk/chair not for office use," "I work at a coffee table or kotatsu." We included the final item "I work 129 at a coffee table or kotatsu" because sitting on the floor at a low table such as a coffee table or kotatsu 130 (a low table equipped with a heater) to eat or do light work is a common way of life in Japan. 131

Additionally, we examined the number of non-conformities to seven of the nine 132 recommended items for work environments; we excluded "I work at a desk/chair for office use" and 133 combined responses to the items "I work at a desk/chair not for office use" and "I work at a 134 table/kotatsu." Thus, non-conformity was indicated by "No" responses to the following items: "There 135 is a place/room where I can concentrate," "There is enough light to do my work," "There is enough 136 space on my desk to work," "There is enough space at my feet," "The temperature and humidity in the 137 room are comfortable," "The environment is quiet," and "I work at a desk/chair not for office use" or 138 "I work at a coffee table or kotatsu." The number of non-conformities was stratified into five categories 139 (0 items, 1-2 items, 3-4 items, 5-6 items, and 7 items) and used as an exposure factor. 140

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The presence or absence of severe work functioning impairment was measured using the 142 Work Functioning impairment Scale (WFun) and used as the primary outcome. WFun is a self-reported 143 outcome measure of the degree of work functioning impairment developed based on the Rasch model 144 and validated according to Consensus-based Standards for the selection of health Measurement 145 Instruments (COSMIN) (18). Subjects provide responses to a total of seven items, including "I haven't 146 been able to behave socially" and "I have felt that my work isn't going well," on a five-point scale, 147 and the total score indicates the degree of work functioning impairment. The total score ranges from 148 7 to 35, and a score of 21 or higher is defined as severe work functioning impairment based on the 149 results of interviews with occupational health nurses in a previous study (19) . While the original WFun 150 contains seven items, we used a six-item version, from which the scores can be equivalently converted 151 to those of the original version based on the Rasch model. 152

For socioeconomic factors, we examined the following items: sex, age, education (junior 153 high school; high school; vocational school, junior college, or technical college; university or graduate 154 school), marital status (married; divorced/bereaved; never married), job type (mainly desk work; Jobs 155 mainly involving interpersonal communication; mainly physical work), equivalent income (household 156 income divided by the square root of household size), company size (total number of employees in the 157 company where the respondent mainly works [self-employed answered one]). 158

For work-related factors, we examined the following items: frequency of working from 159 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.23.21254207 doi: medRxiv preprint home (at least 4 days a week; at least 2 days a week; at least 1 day a week; at least 1 day a month; 160 hardly ever), change in working hours compared to pre-2019 (increased; no change; decreased), one-161 way commute time. 162

As the community-level factor, we used the cumulative incidence of COVID-19 infection 163 in the prefecture of residence one month before the survey. Data were collected from the websites of 164 public institutions. 165 166

Age and WFun scores are presented as continuous variables using median and interquartile 168 range (IQR). Other variables are presented as categorical variables using percentages. We conducted 169 mixed-effects logistic regression analysis using the presence or absence of severe work functioning 170 impairment as the dependent variable and subjective evaluation of the physical work environment or 171 the number of non-conformities to the recommendations for work environments while working at 172 home as independent variables, with the prefecture of residence as the random effect. To adjust for 173 potential confounders, we used sex, age, education, job type, equivalent income, company size, 174 frequency of working from home, change in working hours compared with pre-2019, one-way 175 commute time, and cumulative infection rate by prefecture as covariates. 176

All statistical analyses were conducted using Stata (Stata Statistical Software: Release 14.2; 177 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The demographic and sociological characteristics of the analyzed population are shown in 188 Table 1 . A total 3,361 (58%) were male and the median age was 50 years (IQR: 42-57). Of the total 189 population, 4,052 (70%) were desk workers and 2,790 (48%) telecommuted four or more days per 190

week. The overall median WFun score was 12 (IQR: 7-20) and severe work functioning impairment 191 (WFun≥21) was observed in 1,309 workers (23%). There were no missing data because the survey 192 was designed such that all responses were mandatory. 193

[Insert Table 1 . here] 194

The associations of severe work functioning impairment with workers' physical work 195 environment while working from home are shown in Table 2 . There was no significant association of 196 severe work functioning impairment with the item "I work at a desk/chair for office use" (adjusted 197 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in Table 3 shows the relationship between severe work functioning impairment and the number 205 of non-conformities (out of seven) to recommendations for work environments. From the nine items 206

in Table 2 , we excluded "I work at a desk/chair for office use" and combined responses to the items "I 207 work at a desk/chair not for office use" and "I work at a table/kotatsu." When we used those who 208 indicated no non-conformities to recommendations for work environments as the reference group, the 209 OR of severe work functioning impairment was significantly higher for any non-conformity to the 210 seven items examined. Moreover, an increase in the number of non-conformities was correlated with 211 an increase in the OR of severe work functioning impairment, with non-conformities to all seven items 212 associated with the highest risk (aOR: 4.51, 95% CI: 2.86-7.11, p<0.01). The results of Model 1, 213 which was adjusted only for sex and age, and Model 2, which was adjusted for other potential 214 confounders, were in a similar direction. 215 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in To our knowledge, this is the first study to identify an association between the physical work 219 environment while working from home and work functioning impairment. We found that non-220 conformity to any of the recommendations for work environments examined in this study, except for 221 "I work at a desk/chair for office use," was associated with work functioning impairment. Additionally, 222

there was a dose-response relationship between the number of non-conformities to recommendations 223 for work environments and risk of work functioning impairment. 224

We found that non-conformity to any of the recommendations for work environments while 225 working from home examined in this study, except for "I work at a desk/chair for office use," was 226 associated with work functioning impairment. The work environment items examined in this study are 227 traditionally recommended for health and safety management when working at a desk (13,14,15) . 228

These included items related to the quality of the surrounds, such as temperature, humidity, and 229 quietness of the workplace; those linked to back pain, stiff shoulders, and musculoskeletal strain; and 230 those for concentration and performance. Comfortable temperature and humidity are expected to 231 improve workers' concentration and performance. In addition, items related to work posture, such as 232 the choice of a desk and chair suitable for work and space on the desk and at the feet, are thought to 233 reduce musculoskeletal strain during work. In addition, managing proper work posture, temperature, 234 humidity, and illumination will help prevent visual display terminals syndrome. Setting up these 235 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in Furthermore, our findings suggest that there is a dose-response relationship between the 238 number of non-conformities to recommendations for work environments while working from home 239 and work functioning impairment. In other words, no one specific item among the seven items 240 examined in this study had a positive impact on productivity; rather, adapting as much of the work 241 environment as possible to recommendations for work environments while working from home had 242 an increasingly positive impact on productivity. Interestingly, only 20% of workers indicated that they 243 conformed to all seven items examined in this study, suggesting the difficulty of creating an optimal 244 work environment while working from home. For example, purchasing and installing new desks and 245 chairs suitable for working from home can be a financial burden and a housing space problem. Further, 246 adequate management of room temperature and humidity is linked to issues such as insulation of the 247 dwelling and the purchase and electricity costs of air conditioning. Based on the results of a previous 248 survey and our present findings, we can infer that the methods and systems for checking and improving 249 the working environment while work from home are still lacking and that they are not being managed 250 within companies (16). To increase productivity when working from home, both companies and 251 workers need to account for each of these individual items when creating a work environment to ensure 252 that as many items conform to the recommendations as possible. 253 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in There is currently no consistent evidence on the impact of working from home on the health 254 and productivity of workers, as it varies depending on the worker's situation (20). While working from 255 home has its advantages, including improved work-life balance and reduced burden of commuting, it 256 also has disadvantages, including the psychological burden of a lack of separation between living and 257 working spaces and time, and the lack of communication with colleagues. Furthermore, as shown in 258 this study, the impact of various aspects of the physical work environment while working from home 259 can be significant. In fact, the number of conformities to recommendations for work environments 260 examined in this study varied among workers. Therefore, to understand the impact of working from 261 home on the health and productivity of workers, it is necessary to continue to study differences in the 262 physical work environment among individuals conducting work from home. 263

Working from home has been suggested to be associated with sickness presenteeism, a 264 condition in which individuals work while experiencing health problems. For example, the use of 265

properly designed desks and chairs can affect the health of workers with musculoskeletal disorders 266 and chronic pain, important problems that cause sickness presenteeism (21,22). Further, a study has 267 suggested that removing barriers to work, such as commuting and going to the office, is also associated 268 with the occurrence of sickness presenteeism (23). Improving workers' work environment while 269 working from home is important for reducing the negative health outcomes associated with working 270 from home, improving presenteeism, and increasing productivity. Many companies and workers who 271 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in (24). Given that our study protocol found that 21% of the entire study population has severe work 282 functioning impairment (17), we concluded that our present study population was relatively unbiased. 283

Second, we relied on respondents' self-assessment of their physical environment while 284 working from home, but did not examine the actual physical environments. Therefore, there may be 285 discrepancies with objective evaluation. However, because we inquired about the physical 286 environment, the possibility of erroneous answers is low. 287

Third, since this study was a cross-sectional study, it is impossible to determine the causal 288 relationship between the exposure factors and outcome. However, we think it is unlikely that 289 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in In conclusion, individuals who work in poor work environments while working from home 297 may exhibit work functioning impairment during a period of rapid change owing to the COVID-19 298 pandemic. As work environments become more diverse, it is important for both companies and 299 individual workers to create a work environment that prevents negative health outcomes and improves 300 productivity while working from home. 301 302 303 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in Figure 1 . Flow chart of participants in the study *Respondents who had extremely short response times, were shorter than 140 cm tall, weighed less than 30 kg, or provided inconsistent responses to multiple identical questions. **Respondents who answered "hardly ever" to the question inquiring about the frequency of working from home *** Respondents who work from home at least 1 day a month perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in 

Decisions by the Headquarters for Novel Coronavirus Disease Control. Basic Policies for Novel 305

Decisions by the Headquarters for Novel Coronavirus Disease Control. Basic Policies for Novel 308

Coronavirus Disease Control (Revised on

Decisions by the Council for the Realization of Work Style Reform. Action Plan for the Realization 317 of Work Style Reform

Psychosocial and physical risk factors for musculoskeletal disorders of 323 the neck, shoulders, and lower back in salespeople

Musculoskeletal symptoms in pharmaceutical sales representatives

Determination of pain in 328 musculoskeletal system reported by office workers and the pain risk factors

The psychological impact of teleworking: stress, emotions and health

Telework and occupational health: a Quebec empirical study and regulatory 333 implications

Psychological stress and disease

Ministry of Health, Labour and Welfare. [Guidelines for the Appropriate Introduction and 337

Implementation of Work Outside the Workplace Using Information and Communication Technology

Ministry of Health, Labour and Welfare

Ministry of Health, Labour and Welfare

Research Project on Labor Management for Teleworkers as a Ministry of Health, Labour and Welfare 348 commissioned project: Fact-finding Survey on Labor Management of Teleworkers

Internet-based health survey in workers during the COVID-19 pandemic in 2020. medRxiv

Development and validity 355 of a work functioning impairment scale based on the Rasch model among Japanese workers

Functioning Impairment Scale (WFun): A Method to Detect Workers Who Have Health Problems 359

Affecting their Work and to Evaluate Fitness for Work

A rapid review of mental and physical health effects 361 of working at home: how do we optimise health

Total Health-Related Costs Due 365 to Absenteeism, Presenteeism, and Medical and Pharmaceutical Expenses in Japanese Employers

Grit PhD Home-Based 368

Association between the course 371 of hypnotics treatment for insomnia and work functioning impairment in Japanese workers