Abstract

Background

Remote work allows one to work free from workplace policy, but it may negatively affect health-related behaviors such as tobacco product use and alcohol drinking. Our study aimed to investigate the association of remote work with tobacco dependence and hazardous alcohol use.

Methods

This nationwide Internet-based survey was conducted in Japan in February 2023. Using the data of 8,292 workers aged 18 鈥� 64, frequency of remote work (0,鈥�<鈥�1,鈥夆墺鈥�1听day per week), tobacco dependence based on the scores of the Tobacco Dependence Screener鈥夆墺鈥�5, hazardous alcohol use based on the scores of the Alcohol Use Disorders Identification Test鈥夆墺鈥�8 were assessed. Then, the associations of frequency of remote work with tobacco dependence and hazardous alcohol use were examined using a multivariable Poisson regression model.

Results

The percentages of remote work 0,鈥�<鈥�1,鈥夆墺鈥�1听day per week were 56.6%, 9.9%, and 33.5%, respectively, whereas those of tobacco dependency and hazardous alcohol use were 11.3% and 17.9%, respectively. Remote work鈥�<鈥�1听day per week (adjusted prevalence ratio [aPR] 1.30, 95%CI 1.07 鈥� 1.57), and鈥夆墺鈥�1 per week (aPR 1.29, 95%CI 1.13 鈥� 1.47) were associated with tobacco dependence. Further, remote work鈥�<鈥�1听day per week (aPR 1.43, 95%CI 1.25 鈥� 1.65), and鈥夆墺鈥�1 per week (aPR 1.40, 95%CI 1.27 鈥� 1.55) were associated with hazardous alcohol use.

Conclusions

Remote workers showed higher prevalence of tobacco dependence and hazardous alcohol use. The health behavior of remote workers should be carefully monitored.

Introduction

Tobacco and alcohol are known to negatively affect their health conditions even with a small amount of consumption. Tobacco smoking can increase individual risks of developing stroke, coronary artery disease, chronic obstructive pulmonary disease, and erectile dysfunction [1]. It is estimated that the total global cost of smoking exceeded USD 1,436 billion, equivalent to 1.8% of the world鈥檚 annual gross domestic product (GDP) [2]. Even habitual alcohol intake is associated with increased odds of hypertension and cardiovascular diseases [3, 4]. Economic loss related to alcohol drinking is also unignorable. For example, in the United States of America, the cost of excessive alcohol drinking is estimated to lead to approximately $250 billion USD, including 179 billion USD loss related to workplace productivity. In Japan, the economic costs of alcohol consumption were about 1,000 billion yen (6.8 billion USD) in direct healthcare costs and up to 5,300 billion yen (36 billion USD) in indirect costs [5]. In Japan, 16.7% of the general population regularly smokes tobacco, and 20.5% have habitual drinking of alcohol consumption (3 or more times weekly and 20听g or more of alcohol each occasion), according to a 2019 report from the Ministry of Health, Labour, and Welfare of Japan [6]. In society, everyone encounters daily stress and employs various coping strategies to manage it, such as physical activity, music, sleep, and meditation. Tobacco and alcohol are commonly used as coping tools to alleviate stress and negative emotions as well. However, unlike other coping strategies, the use of tobacco and alcohol carries a significant risk of physical and mental dependence, which can lead to functional impairment over time. This issue was particularly evident during the COVID-19 lockdown when heightened daily stress contributed to an increased prevalence of tobacco dependence and hazardous alcohol use.

During the pandemic, remote work has become widely available, and the trend still remains high globally [7]. Three years after the COVID-19 pandemic, many employees work from home. In fact, full-time employees reported to work from home 0.4 to 1.7听days per week across the globe [8]. Workplace policies for tobacco use and flexibility of their work through occasional remote work may contribute to behavioral changes in alcohol and tobacco use. A previous study showed that working from home weekly or more was associated with alcohol use after adjusting age, gender, leadership position, and education level [9]. A previous study in Japan showed that the smoking rate increased among those who work from home in the first year of the COVID-19 pandemic [10]. Another study in the U.S. has demonstrated the positive association between working from home and alcohol consumption at the beginning of the pandemic [11]. However, little is known about the association of remote work with tobacco dependence and hazardous alcohol use in the fourth year of the COVID-19 pandemic when the influence of the pandemic on daily work was mitigated. Our study aimed to investigate the association of remote work with tobacco dependence and hazardous alcohol use in 2023.

Methods

Study design, setting, and data source

This is an internet-based cross-sectional study, a part of the Japanese 鈥淪ociety and New Tobacco Internet Survey (JASTIS),鈥� conducted between February 6 and 27, 2023. The JASTIS is a repeated cross-sectional study launched in 2015, initially aiming at annually examining the prevalence of tobacco product use, including combustible, heated tobacco, and electronic cigarettes among the general population aged 15 鈥� 79听years in Japan [12]. This internet-based survey system was operated by an internet survey agency, Rakuten Insight, which registered 2.2 million panelists as of September 2022 [13]. We targeted a sample size of 34,000 because we assumed that about 10% would be excluded due to providing irrelevant answers. A final sample size of 30,000 was deemed both feasible within our budget and sufficient to generate robust study findings across various health outcomes. As Fig.听1 shows, invitation e-mails for study participation were sent to 44,681 individuals who had been pooled as the old cohort members of JASTIS or JACSIS (i.e., a sister survey of JASTIS launched in 2020), of which 28,437 participated in the survey (response rate 63.6%). Further, 100,926 individuals aged 15鈥�79 were randomly selected from the 2,2 million registered panelists of the internet survey agency and invited to the survey as a new cohort, of which the first 5,563 respondents (5.5%) were added to secure the target sample size of 34,000, then data collection was closed.

Fig.听1

Flow chart of study participants

Data quality management

To ensure data quality, we excluded 2,964 respondents (8.7%) who provided an incorrect answer in the dummy question 鈥淧lease choose the second from the bottom out of five options (A鈥墌鈥塃)鈥� or unrealistic answers to the following three questions; responding 鈥測es鈥� in all the questions for using eight alcohol or drugs; responding 鈥測es鈥� in all the questions for having nine chronic diseases; and the number of household members was鈥夆墺鈥�15.

Eligibility criteria of study population

The eligibility criteria of the study population were individuals who were aged 18鈥�64听years old, and full-time workers. We excluded those who worked for part-time jobs, home industry, or housework, and those who were students, retired, or jobless because they may not work remotely or may experience different patterns of remote work compared to full-time workers. Additionally, we excluded those whose work required onsite communication or physical activity because their presence is typically required at the workplace. Subsequently, our final sample size for the analysis was 8,292 (Fig.听1).

Exposure variables: frequency (the number of days) of remote work per week

The exposure variable was the frequency of remote work. We collected information on the frequency of working from home and working from somewhere other than office or home (e.g., shared office) with seven categories of frequency (no; once per month; two or three times per month; once per week, two or three times per week, four or five times per week, six or seven days per week). The information was converted into the frequency per week as follows: 0, 0.25, 0.625, 1.0, 2.5, 4.5, and 6.5. Then, we combined the frequency of working from home and/or a place other than office or home, and categorized them into remote working 0,鈥�<鈥�1, or鈥夆墺鈥�1听day per week, based on the categories used in previous studies and the context of remote work in Japan in 2023 [9, 10, 14,15,16].

Outcome variables: tobacco dependence and hazardous alcohol use

We used two outcomes, including tobacco dependence, and alcohol use disorder. In this study, Tobacco product use was defined as whether the respondents used combustible tobacco (paper, or hand rolled tobacco), or heated tobacco (Ploom Tech, Ploom S, Ploom X, IQOS, glo, lil HYBRID) at least one day in the past 30听days prior to the survey. Nicotine dependence was measured using the Tobacco Dependence Screener (TDS) [6]. TDS is a standardized diagnostic tool for nicotine addiction, which has ten items with a maximum score of 10. The scores鈥夆墺鈥�5 are considered nicotine addiction. Alcohol use disorder was measured using the Alcohol Use Disorders Identification Test (AUDIT) [17]. The AUDIT comprised 10 items, with the maximum score of 40. The scores of鈥夆墺鈥�8 are considered hazardous alcohol use. The Japanese version of the two scales has been validated and is widely used in clinical practice [18, 19].

Adjustment variables

We included ten adjustment variables in the analysis: sex (men or women), age (10听s-20听s, 30听s, 40听s, 50听s, or 60听s), educational attainment (high school / lower, college, or university / higher), marital status (married, single, or widowed/divorced), equivalent household income (very low [<鈥�2.25 million JPY (Japanese yen)], low [鈮モ��2.25,鈥�<鈥�3.25 million JPY], high [鈮モ��3.25,鈥�<鈥�4.75 million JPY], very high [鈮モ��4.75 million JPY], or no answer), an average working hour per day (鈮も��7听h, 8 鈥� 9听h,鈥夆墺鈥�10听h), self-rated health (excellent, very good, good, fair, or poor), having at least one of the following chronic diseases (i.e., hypertension, diabetes, dyslipidemia, asthma, angina pectoris/myocardial infarction, stroke, chronic obstructive pulmonary disease, chronic kidney disease, chronic hepatitis/liver cirrhosis, immune disorder, or cancer) that is required for regular medical checkup (yes, or no), and having mental illness (i.e., depression or psychiatric diseases) (yes, or no), and region of residence in Japan (Hokkaido/Tohoku, Kanto, Hokuriku/Koshinetsu, Tokai, Kansai, Chugoku, Shikoku, or Kyusyu/Okinawa region).

Statistical analysis

Descriptive analysis was performed to describe the distributions of participants鈥� socio-demographic, economic and health-related characteristics, tobacco product use status, nicotine dependence, and hazardous alcohol use according to the frequency of remote work, and reported with p-values for a chi-square test. Further, the associations of remote work frequency with tobacco dependency and hazardous alcohol use were examined using a multivariable Poisson regression model with a robust variance estimator because the prevalence of the two outcomes was greater than 10% [20]. In the model, we adjusted for sex, age group, educational attainment, marital status, equivalence household income, working hours per day, self-rated health, chronic disease(s), mental illness, and region of residence (main model). Tobacco product use status was not included in the model because it is a precondition of tobacco dependence. Additionally, we performed a further analysis by incorporating an interaction term of frequency of remote work and sex into the main model (interaction model). We also conducted the stratified analysis by sex (stratification model). The results of interaction and stratification models were presented in Tables S1 and S2 (supporting information). Furthermore, we conducted a sensitivity analysis that included individuals without chronic diseases only and reported in Table S3.

Ethical considerations

This study was conducted according to the ethical standards of the Declaration of Helsinki. This study was approved by the Institutional Review Board of Osaka International Cancer Institute (Number: 20084) and Ethics Committee Tohoku University Graduate School of Medicine (Number: 2024鈥�1鈥�231). We obtained electronic informed consent from all participants for the use of their data for research purposes. The confidential data of the participants were removed by the internet survey agency before the researchers received the dataset. A credit point called 鈥楨points鈥�, which could be used for online shopping, was offered to the respondents as an honorarium for completing the questionnaire.

Results

Table 1 shows the distribution of participants according to their socio-demographic and economic characteristics. The percentages of remote working per 0,鈥�<鈥�1, and鈥夆墺鈥�1听day per week were 56.6%, 9.9%, and 33.5%, respectively. Tobacco dependence and hazardous alcohol use were observed in 11.3% and 17.9%, respectively. The distributions of all socio-demographic, economic, health-related characteristics, and the study outcomes were significantly different.

Table 1 Basic characteristics of study participants per frequency of remote work

Table 2 shows the association between remote work frequency and tobacco dependence (scores of TDS鈥夆墺鈥�5). Compared to those not working remotely, those who worked remotely鈥�<鈥�1听day per week (adjusted prevalence ratio [aPR] 1.30, 95% confidence interval [CI] 1.07 鈥�1.57), and鈥夆墺鈥�1听day per week (aPR 1.32, 95%CI 1.15 鈥� 1.51) were more likely to be tobacco-dependent. Additionally, the stratified analysis revealed that the association remained significant among men and female (though the association between remote work鈥�<鈥�1听day per week and TDS鈥夆墺鈥�5 was not significant for women). There was no significant interaction between remote work and sex in relation to tobacco dependence (Table S1).

Table 2 The association between remote work frequency and tobacco dependence (n鈥�=鈥�8292)

Table 3 shows the association between remote work frequency and hazardous alcohol use (scores of AUDIT鈥夆墺鈥�8). Those who worked remotely鈥�<鈥�1听day (aPR 1.41, 95%CI 1.23 鈥� 1.62), and鈥夆墺鈥�1听day per week (aPR1.35, 95%CI 1.22 鈥� 1.50) were more likely to be hazardous alcohol drinkers, compared to those who did not work remotely. Additionally, the stratified analysis showed that this association remained significant among both men and women, although there was no significant interaction between remote work and sex in relation to hazardous alcohol consumption (Table S2).

Table 3 The association between remote work frequency and hazardous alcohol use

A sensitivity analysis that included individuals without chronic diseases showed similar results to the main model. However, the prevalence ratios of frequency of remote work for the study outcomes among those without chronic diseases tended to be smaller than among all participants.

Discussion

To our best knowledge, this is the first study to investigate the association of remote work with tobacco dependence and hazardous alcohol use three years after the COVID-19 pandemic. A large-scale, nationwide survey in Japan showed that those who work remotely even less one day per week were more likely to be tobacco-dependent and hazardous alcohol drinkers compared with those who do not work remotely. Our study utilized the TDS and AUDIT to examine tobacco dependence and hazardous alcohol use, respectively, offering a more nuanced understanding of substance use behaviors instead of binary indicators.

A 2020 study in Japan examining reviewing smoking behavior change in the COVID-19 state of emergency found 32.1% of smokers increased the number of cigarettes smoked per day [10]. Furthermore, smokers who work remotely were likely to increase their smoking [10]. Another study in Norway demonstrated that remote work more than 15听h per week was associated with higher alcohol consumption after adjustment of all variables such as age, gender, leadership position, and educational level [9]. These findings were consistent with our study findings, showing the association of remote work with increased risk for smoking and alcohol behaviors. Of note, while the previous study examined remote work 15听h or more per week, equivalent to two full days per week [9], our study showed that even working remotely once a week or less can contribute to the increased risk of tobacco dependence and hazardous alcohol use. This may be a unique health problem in the fourth year of COVID-19 pandemic.

There are two main potential reasons for the association of remote work with tobacco dependence and hazardous alcohol use. First, remote work, even less than once a week, could have increased access to tobacco and alcohol by being at home more often. In Japan, workplace bans on smoking during working hours are common, limiting access to cigarettes at workplace [21]. Greater freedom to smoke cigarettes during their remote works can contribute to more tobacco dependence. The habits formed at home through remote work might have persisted even after remote work ends. Second, remote work reduces face-to-face interaction with their colleagues, limiting work-related social support which plays a critical role in enhancing the well-being [22]. Due to the reduced social support, people might have utilized dysfunctional coping strategies, such as tobacco and alcohol use, to alleviate their stress [9].

In our study, those who have lower self-rated health conditions, chronic disease, or mental illness were more likely to have tobacco dependence and hazardous alcohol behaviors. While individuals aged 50听s to 60听s were more likely to have tobacco dependence, they were less likely to be hazardous alcohol drinkers. Similarly, previous studies reported that risk factors of smoking behavior included male gender, older age group (55听years old or older), and living alone [10], and those for increased alcohol use included male gender, higher age, leadership position, and higher educational level [9].

There are several limitations in our study. First, our study could show only associations rather than causation, given the cross-sectional study design. Second, while we surveyed residents living in diverse regions, our studied pool may have been biased during the selection process. Third, surveyed individuals might not correctly report or underreport their behavior of tobacco and alcohol use due to their recall or social desirability bias. In fact, some studies have shown that individuals tend to underreport their behaviors towards tobacco and alcohol [23,24,25]. Fourth, information on chronic diseases and mental illness was collected on a self-reported base, and may not necessarily match data from medical records. Despite these limitations, this study has established that those who work remotely even less than once per week are more likely to have tobacco dependence and hazardous alcohol use.

To the best of our knowledge, it is the first study to investigate the association of remote work with tobacco dependence and hazardous alcohol use three years after the beginning of the pandemic when its impact on daily work was mitigated. This finding can aid clinicians in identifying individuals at higher prevalence for developing tobacco use disorder and alcohol use disorder at an early phase. Based on the study鈥檚 findings, we recommend that employees who work remotely even once a week should be carefully reviewed for their tobacco and alcohol use behaviors. Finally, as this was a cross-sectional study, future research is needed to conduct a causal analysis and examine whether people are more likely to develop tobacco or alcohol dependence after transitioning to remote work from onsite work.