The health condition of workers is a critical issue. The International Labour Organization (ILO) [1] stated that “Occupational accidents and diseases lead to devastating impacts on workers, enterprises, and entire communities and economies.” The ILO [2] estimated that “The losses in terms of compensation, lost work days, interrupted production, training, and reconversion, as well as healthcare expenditure, represent around 3.94 percent of the world’s annual GDP.” The World Health Organization (WHO) [3] mentioned that “Work-related health problems result in an economic loss of 4% - 6% of GDP in most countries… Research has demonstrated that workplace health initiatives can help reduce sick leave absenteeism and healthcare costs for companies by 27% and 26%, respectively.” The National Center for Chronic Disease Prevention and Health Promotion of the Centers for Disease Control and Prevention (CDC) [4] in the United States described “Preventable chronic conditions as major contributors to the costs of health insurance premiums and employee medical claims. These costs are at an all-time high and continue to increase in the United States.” In Japan, it is reported that the number of work-related diseases requiring work absences of four days and more was 7844 in 2017 [5] .

Evidently, health is not only essential for individuals’ well-being [6] but is also a very important social goal and benefit for individual workers, firms, and society as a whole [7] . In other words, maintaining and promoting workers’ health will not only improve their well-being but will also lead to medical expense reduction and increased productivity [8] . Health risks are a serious burden to employers owing to the loss of productivity, and various studies have been conducted on this topic [9] - [22] .

The monetary costs have been also estimated. Nagata et al. [23] estimated that the cost of absenteeism, presenteeism (reduction of productivity due to health conditions in the workplace), and medical/pharmaceutical expenses were $520, $3055, and $1165 per person per year, respectively, in the fiscal year 2014 in Japan. Ramasamy et al. [24] reported that direct costs for class I (body mass index (BMI) 30.0 to 34.9, BMI = weight (kg)/(height(m))²), class II (BMI 35.0 to 39.9 kg/m²), and class III (BMI ≥ 40.0 kg/m²) obesity were $1775, $3468, and $11,481 per-patient-per-year (PPPY), respectively, higher than those of the reference; and absenteeism/disability costs were higher than those of the reference by $617, $541, and $1707 PPPY for classes I, II, and III obesity, respectively. Gennep et al. [25] reported that the mean annual costs of absenteeism, presenteeism, and overall work productivity for inflammatory bowel disease patients were ?738, ?478, and ?597, respectively. Ademi et al. [26] estimated AUD 5.23 billion in lost GDP, with an average of AUD 101,366 lost per person with heterozygous familial hypercholesterolemia over their working lifetime using 2017 GDP data. Cawley et al. [27] estimated that annual productivity loss due to obesity was from $271 to $542 per employee in 2016. Kocakulah et al. [28] described that “Some sources, including Statistics Canada, cite that absenteeism approximates 15 - 20 percent of payroll (direct and indirect) costs.” They also discussed several possible solutions for the companies. Gianino [29] studied economic losses in terms of employee absenteeism due to influenza and reported that the average work loss due to influenza was ?27/person.

Another question is whether investments in improving workers’ health conditions are worthwhile. Baicker [30] estimated that medical and absenteeism costs fell by approximately $3.27 and $2.73 for every dollar spent, respectively. However, Verelst et al. [31] mentioned that while workplace influenza vaccination is relatively inexpensive and convenient, the return on investment is volatile for employers. From a managerial perspective, it is important to identify the risk factors that affect productivity and invest in improving these factors. Lawrance et al. [32] proposed a decision support system to improve health and well-being in the workplace and identify groups of employees at risk of sickness absence aiming to reduce or prevent absences.

Various studies have been conducted on overweight and obese individuals. Howard and Potter [33] evaluated the relationship between obesity, obesity-related chronic health conditions, and worker absenteeism using logistic regression. They found that obesity was related to higher rates of worker illness absence after controlling for demographic, socioeconomic, occupational, health-related, and behavioral variables. Keramat et al. [34] studied the relationship between absenteeism and obesity in Australian workers. They concluded that workplace absenteeism was significantly associated with being overweight and obese. Hashiguchi [35] evaluated the effect of BMI on working health risks determined by heart rate reserve (HRR) using a logistic regression model with a 40% HRR criterion. However, BMI was not a significant risk factor.

Presenteeism is a complex problem [36] and difficult to measure properly. Moreover, Marmot et al. [37] pointed out that worker absence is a good proxy for workers’ health. Therefore, the relationships between health factors and absenteeism are analyzed using health and working record data of 1136 employees in this study.

Table 3 shows the results of the covariate selection using the procedure described in the previous section. The bold values indicate the largest value of L R j i , satisfying over 1 in step j, where j covariates are included.

In the first step, HDL was selected. Female, B_Sugar, Height, and GOT were selected using the following steps. All values of L R 6 i were less than 1, and the procedure was stopped at this stage. The final model, which contains five covariates, is given by:

P [ y = 1 ] = Λ ( β 0 + β 1 H D L + β 2 F e m a l e + β 3 B _ S u g a r + β 4 H i g h t + β 5 G O T ) (2)

Table 4 presents the estimation results of this model. The estimate of Female is positive and significant at the 1% level, suggesting that that the female was absent from work more frequently than the male [58] . Concerning health factors, the estimates of HDL and B_Sugar were negative and positive, respectively, and were significant at the 5% level. This means that a higher level of HDL reduces long-term absenteeism but a higher level of B_Sugar significantly increases long-term absenteeism and therefore, they are considered important health factors.

The bold values give the largest value of LR_ji, satisfying over 1.

The cost of absenteeism among employees is significantly high for corporate employers. Additional replacement employees and their training are necessary to maintain the corporation’s services [14] [28] [31] [32] [59] . Moreover, investments for the reduction of absenteeism must be made efficiently from a cost-benefit perspective. The results in the previous section suggest that HDL and B_Sugar are two important factors directly related to absenteeism.

Figure 2 shows the odds ratios (ORs) and 95% confidence intervals (CIs) for HDL (high-density lipoprotein cholesterol) and B_Sugar (blood sugar). As these variables are numerical, the OR for variable x i is calculated by comparing x i and ( x i + one standard deviation). For HDL, the OR was 0.75 with a 95% CI of 0.59 - 0.95. This indicates that increasing HDL level by 15.0 mg/dL (one standard deviation) reduces long-term absenteeism by 25%. (Because the probability of L_Absence = 1 was small, the odds ratio was approximately equal to the probability ratio.) The CDC [60] has labeled HDL “good” cholesterol. According to the guidelines of the Japan Atherosclerosis Society, the normal HDL level is 40 mg/dL or over [61] . The percentage of long-term absenteeism of 1063 employees with normal HDL was 7.8%; however, for 73 employees with HDL < 40 mg/dL, the percentage was 15.1%, almost twice as large as that of the normal HDL group. The roles and mechanisms of HDL have been studied [62] [63] [64] . Ghobadi et al. [65] mentioned that nutritional intervention in the workplace seems to be more beneficial in improving HDL through a systematic review, and de Liz et al. [66] reported a positive impact of regular consumption of some types of juices on HDL levels.

For B_Sugar, the OR was 1.20 with a 95% CI of 1.01 - 1.42. Decreasing B_Sugar level by 18.9 mg/dL (one standard deviation) would reduce long-term absenteeism by 20%. Under the current Japanese criteria, an individual is diagnosed with diabetes if B_Sugar ≥ 126 mg, prediabetes if B_Sugar is 110 - 125 mg, and normal if B_Sugar < 110 mg [67] . For details on the blood sugar distribution in Japan, see Nawata [68] . The percentage of long-term absenteeism is 7.6% for 1059 normal employees; however, it becomes 18.2% for 77 prediabetic and diabetic employees. Blood sugar levels can be controlled through lifestyle improvements such as modifying eating [69] and exercise habits. CDC [70] mentioned that “The goal is to get at least 150 minutes per week of moderate-intensity physical activity.”

The correlation between HDL and B_Sugar is low and negative with a correlation coefficient of −0.144. This means that the corporation could reduce long-term absenteeism by as much as 40% by improving HDL and B_Sugar levels. This operational site has its own company cafeteria, which is used by most employees, therefore, nutrition intervention and improvement of eating habits could be relatively easy with little or no cost. It may be worthwhile to practice nutritional intervention, especially for those with low HDL or high B_Sugar levels.

Employees’ health is an important issue. It affects not only the well-being of employees but also their productivity through absenteeism and presenteeism. In this study, the health factors affecting absenteeism were analyzed because absenteeism can be directly measured without ambiguity and were good proxy of health conditions. The health and work record information of 1136 workers at an operational site of a large corporation was used in the analysis. The costs of absenteeism are significantly high and this is a serious issue from a managerial viewpoint of the corporation.

The relationship between long-term absences (more than three days in three mouth) and health factors was analyzed using logistic regression models. Because the number of observations is not very large, no health factor is significant at the 5% level when all covariates are included. Therefore, a procedure to determine the optimal selection of covariates was used. It is based on likelihood ratio statistics and the AIC.

In the selected model, the most significant health factors were high-density lipoprotein cholesterol (HDL-C) and blood sugar levels. Corporations can reduce long-term absenteeism significantly by improving these factors. These factors can be improved by modifying eating habits. The operational site has its own company cafeteria, and nutrition intervention is relatively easy. It may be worthwhile to implement nutritional interventions, especially for those with abnormal levels of HDL-C or high blood sugar.

The results of this study are based on one operational site of a corporation. The employees were mainly operators working inside the buildings. The results may be different for other types of jobs and working conditions, such as fieldwork. We may not be able to generalize the results of this study. Analyses of different types of jobs and working conditions are necessary. Evaluation of the exact cost of nutrition intervention and improvement of eating habits is also important. These topics will be investigated in future studies.

The author would like to thank the corporation for providing this dataset. The CEO of the corporation has been supporting the study. The author is grateful for the support. This study was approved by the Institutional Review Board of Hitotsubashi University (2022C003). The data presented in this study was obtained under a data use agreement with the corporation and Hitosbshsi University. The author would also like to thank an anonymous reviewer for his/her helpful comments and suggestions.

The author declares no conflicts of interest regarding the publication of this paper.

Nawata, K. (2023) An Analysis of Health Factors Affecting Em- ployees’ Absenteeism: Influences of HDL Cholesterol and Blood Sugar Levels. Health, 15, 397-412. https://doi.org/10.4236/health.2023.155027