Impacts of Smoking and Alcohol Consumption on Workplace Presenteeism: A Cross-Sectional Study

Background: Presenteeism refers to the practice of going to work despite poor health, resulting in subpar performance. The problems caused by smoking and drinking alcohol while on company premises have received limited attention despite their health hazards, which are linked to sick leave, occupational injuries, and reduced productivity. This study aimed to investigate the effects of smoking and alcohol drinking on presenteeism. Methods: The study participants were 60,051 wage workers from the database of the second and third Korean Working Conditions Surveys (KWCS) in 2010 and 2011 , respectively. A total of 41,672 workers aged 19 and over who had worked for at least one hour in the previous week answered the survey questions. Chi-squared tests and multiple logistic regression analyses were conducted using SPSS 18.0 to determine the impacts of smoking and drinking alcohol on presenteeism. Results: Of the 41,672 Korean workers, 8565 (20.6%) had experienced presenteeism in the past 12 months. Presenteeism was higher among women; the elderly; people with low educational status; people with low income; people with 5-11 days absenteeism; people with more working time; people with health problems; heavy smokers; and high-risk alcohol drinkers than other participants. Based on the results of multiple regression analysis, heavy smoking (AOR = 2.10, 95% CI [1.56, 2.82]) and high-risk drinking of alcohol (AOR = 1.20, 95% CI [1.10, 1.31]) were signi�cantly related to presenteeism among workers. Conclusions: Smoking and alcohol drinking are associated with and potentially in�uence presenteeism; therefore,

2. Are smoking and drinking signi cant predictors of presenteeism, after adjusting for potentially confounding variables?

Design
This study used a cross-sectional survey research design to conduct a secondary analysis of screening survey data from the second and third KWCS databases. We investigated the effects of smoking and alcohol drinking on presenteeism among wage workers.

Setting and sample subjects
The original KWCS surveys were approved by the Korea Occupational Safety and Health Agency. The SPSS data les classi ed by year were merged by cases (Approval No. 38002).
KWCS is a state-recognized statistical survey that has been conducted since 2006 to identify risk factors by working type and conditions-including mechanical, physical, and chemical hazards in the workplace -and psychosocial factors that in uence working conditions. The population represented by the KWCS are economically active people 15 years and older residing in Korea. A two-step strati ed cluster sampling method was used for the rst and second sampling units. In contrast, sampling frameworks were strati ed according to the criteria city-province, neighborhood/township/town, and housing type (single houses/apartment homes).
The second and third KWCS enrolled a combined 60,051 participants, from which we selected the 41,703 wage workers who were over 19 years old. After excluding surveys with insu cient or incomplete data, completed surveys from 41,672 workers were included in the present analysis.

Instruments
Presenteeism Presenteeism needed to be identi ed. "Yes" responses to the question "Have you ever come to work while sick in the past 12 months?" were considered to indicate presenteeism in this research.

Smoking and drinking alcohol
We de ned smokers as people who had smoked ve packs of cigarettes or more (100 cigarettes or more) to date in their lifetime. "Non-smokers" were those who had never smoked; those who were smokers at the time of the survey were classi ed as "current smokers"; those who no longer smoked but had in the past were classi ed as "ex-smokers." Based on preceding studies [22,23], those who smoked 40 cigarettes or more a day were referred to as "heavy smokers." Drinkers were de ned as those who drank at least once a month regardless of the type of liquor. High-risk drinkers were de ned as those who drank more than twice a week, seven glasses of alcohol or more for men and ve glasses of alcohol or more for women at a time, following the standard suggested by the Korea National Health and Nutrition Examination Survey [3].

General characteristics and health problems
Based on previous studies, general characteristics of the subjects included gender, age, educational status, income, absenteeism, and smoke exposure as factors potentially associated with presenteeism [16,24].
To examine the health problems that affect presenteeism, we included "Yes" responses to the question "Have you ever had health problems over the past 12 months, such as a hearing problem, a dermatologic problem, musculoskeletal pain (back pain, shoulder/neck/arm muscular pain, or lower extremities muscular pain), headache/asthenopia, abdominal pain, respiratory distress, cardiovascular disease, injury, depression/anxiety disorder, fatigue, or insomnia/sleep disorder?"

Data analysis
Survey data obtained in this study were analyzed using SPSS 18.0, and workers' presenteeism rate was calculated in terms of frequency and percentage. Chi-square tests were used to determine differences in presenteeism according to the subjects' general characteristics, smoking, drinking, and health problems. Multiple logistic regression analysis was used to evaluate the associations among smoking, drinking, and presenteeism. Adjusted odds ratio (AOR) and 95% con dence intervals (CI) were calculated, and the level of statistical signi cance was set at p < .05.

Results
General characteristics of participants Participants' average age was 40.7 years; 58.8% were male and 42.2% female. There was a high proportion of college graduates, and for most participants, household income was less than 3 million Korean won per month. Of all respondents, 90.7% had no absences while 4.1% reported missing 1-4 days within the previous 12 months. Musculoskeletal pain was the most common symptom of health problems, and 1.9% had been exposed to tobacco smoke for over three-fourths of their working hours ( Presenteeism and general characteristics Table 3 shows the relationships among smoking, drinking, general characteristics, and presenteeism. Heavy smokers had more presenteeism than non-smokers, ex-smokers, or current smokers (p<.001). Highrisk drinkers had more experience of presenteeism than non-drinkers or current drinkers (p<.001). Females had more experience of presenteeism than males (p<.001). Presenteeism was also more common among the elderly (p<.001), workers who had lower educational status (p<.001), and workers who had lower income (p<.001).Workers who were absent from 5 to 11 days had more presenteeism than those who were never absent, were absent from 1 to 4 days, or were absent more than 12 days (p<.001). Workers who were exposed to tobacco smoke for at least three-quarters of their work time had more presenteeism than those exposed to tobacco smoke less than half of their work time (p<.001). Presenteeism was higher among subjects with health problems than subjects without. Associations among smoking, drinking, and presenteeism

Discussion
Our results indicate differences in presenteeism according to general characteristics and health problems, including smoking and drinking. Several other studies have also shown female workers to have more presenteeism than male workers [8, 10]. Many employees had health issues that signi cantly affected presenteeism [10][11][12]25], including psychological distress and emotional ill-health [8, 11,12,25] and, in recent studies, drug and alcohol problems [10,12,25] and smoking [10,25]. The results of our study support these ndings of earlier studies.
Businesses and governments are concerned about employee health because it affects employers' medical costs and productivity. Hence, recent studies have investigated relations between workers' health problems and presenteeism as related to productivity. Zakrzewska [26] reported that presenteeism costs more than absenteeism and disability, so it is reasonable to link workers' health problems to presenteeism. However, most studies only link presenteeism to physical health problems [8, 25,27], mental stress [8, 10,11,12], and work environment stressors [28]. Some have addressed links with health behaviors such as smoking [12,16], alcohol use [12], obesity, and physical activity [10,16,21].
Though all employees can occasionally be unproductive, research suggests that smoking status alone negatively impacts productivity beyond the work time lost due to smoking breaks [16]. Some people might work intensely despite poor health conditions; however, previous studies have demonstrated smoking and drinking among workers affects their productivity [10,11,12], and loss of productivity may lead to absenteeism and presenteeism [13]. In recent studies, binge drinking was associated with presenteeism [29], and higher levels of alcohol consumption were associated with higher levels of impaired work performance (i.e., presenteeism) [19]. It is reasonable to assume that presenteeism might occur when people work through a hangover or smoke at work and, therefore, excessive drinking and heavy smoking can impact presenteeism.
In this study, alcohol drinking and smoking were found to be related to presenteeism. These ndings are consistent with previous studies [14,16,19,30]. The multiple logistic regression analysis from our study revealed that heavy smoking and high-risk alcohol drinking signi cantly affected presenteeism. After adjusting for general characteristics and the variable "health problems," we found that the in uence of heavy smoking on presenteeism increased. The longer employees were exposed to tobacco smoking in the workplace, or the more health problems they had, the greater the odds of presenteeism. There are several possible explanations for such results. Health status has a critical in uence on work ability [7], and health risks are related to reduced productivity [8,9]. Furthermore, smoking among workers affects productivity [10][11][12], may lead to presenteeism [13], and causes health problems related to presenteeism such as cancer [31] and cardiovascular diseases [32][33][34][35]. Similarly, alcohol drinking causes chronic liver diseases and cancer [36].
This study's results have several implications. Smoking or drinking increases presenteeism, even after adjusting for physical illnesses, which implies costs to employers. Other recent studies have also found that presenteeism was associated with alcohol consumption [19,29] and smoking [20], and several studies suggest that heavy drinking increases absenteeism and leads to declines in productivity, which can be more harmful to productivity than absenteeism [14]. Further, research suggests that high costs are incurred due to smokers' absenteeism, presenteeism, smoking breaks, healthcare costs, and pension bene ts [16].
Changes are needed to improve the health of workers who smoke and drink. The WHO states that one of the most effective ways to reduce tobacco consumption is to raise tobacco prices, but South Korea has the highest smoking rate and the lowest tobacco prices among the 41 OECD countries [3,37]. Our society should pay more attention to heavy smoking and anti-smoking intervention programs, and based on our study, it would be wise to target heavy smokers in the workplace. Smoke-free workplace programs encourage smokers to quit and improve their overall health and productivity [24]. A recent study showed that when heavy smokers were part of a smoking cessation program, their success rates became higher than non-heavy smokers [30]. There is little information on whether changes in work structure can reduce the harmful effects of alcohol in the workplace [38]; however, mandatory screening programs are effective for industries such as transportation. Furthermore, investments in social welfare policies outside the workplace can decrease alcohol-related absences and increase productivity [14].
Though our study yielded important results, it had some limitations. First, this cross-sectional study could only con rm the associations between factors; in other words, it could not prove that smoking and drinking are direct causes of presenteeism. Second, errors regarding health problems may have occurred from using participants' responses instead of doctors' diagnoses. Nonetheless, this research is signi cant in being the rst to investigate the effects of both smoking and drinking on presenteeism. The results of this study suggest that smoking and drinking, which have mainly been treated as health risk factors, may also be related to national productivity. In addition, this research, based on a national survey, can serve as a basic guide for future researchers about the impacts of heavy smoking and high-risk alcohol use in the workplace and the necessity of implementing relevant programs.

Conclusions
Smoking and alcohol consumption are associated with chronic health diseases and high socio-economic costs worldwide. The present findings further highlight the economic costs of smoking and drinking, as these are associated with greater odds of presenteeism among Korean workers. These findings demonstrate a critical need for social and behavioral interventions to alter employees' smoking and drinking habits. Further research is needed to identify efficacious treatments as well as the impacts of reduced smoking and drinking on presenteeism. Availability of data and materials: All data generated or analyzed during this study are included in this published article.