2015 Volume 57 Issue 1 Pages 51-57
Objectives: Long work hours and overwork may increase the cardiovascular load of workers. But long work hours and overwork are not the same. Cardiovascular overload from working is dependent on the physical demand of the work and the worker's physical fitness, as well as the working hours. This cross-sectional study was designed to identify the association between overwork and cerebrocardiovascular disease, taking into account the physical demand of work, physical fitness, and work hours. Methods: Study data were obtained from surveillance of occupational cerebrocardiovascular disease. Questionnaire surveys including general and work-related characteristics were conducted. Maximum acceptable work time was estimated using the physical work demand and physical fitness of the subjects. The overwork index, which was the ratio of maximum acceptable work time and actual work hours of the subjects, was calculated. Results: In the workers with a moderate or high physical demand of work, the adjusted odds ratios for overwork indexes of 1.01–1.20, 1.21–1.50, and >1.50 were 2.679 (95% confidence interval (CI) 1.025–6.999), 3.124 (95% CI 1.111–8.783), and 4.331 (95% CI 1.719–10.908), respectively. Conclusions: The results indicate that the risk of cerebrocardiovascular disease might be high in the workers with long working hours, high physical demand of work, and poor physical fitness engaged in work with a moderate to high physical work demand. Work hours should be accommodated according to the worker's physical fitness and the physical demand of work, and this could lower the risk of cerebrocardiovascular disease.
(J Occup Health 2015; 57: 51–57)
Long work hours recently emerged as a workplace stressor that can have a significant, negative impact on health status1, 2). The health problems associated with long work hours include obesity3), cerebrocardiovascular diseases4, 5), depression6), anxiety5), sleep disturbances7), and occupational injuries8).
Cerebrocardiovascular disease is the second most common cause of death in Korea9), and the working hours of workers in the Republic of Korea were the longest among the countries of the Organization for Economic Cooperation and Development (OECD) until 2007; they were the second longest from 2008 to 201110). In the Republic of Korea, approval for compensation for cerebrocardiovascular disease is based mainly on a worker's working hours, which is determined by the Enforcement Rules of the Enforcement Decree of the Industrial Accident Compensation Insurance Act. According to the regulations of the Ministry of Employment and Labor, working of more than 60 hours per week has a strong association with the occurrence of cerebrocardiovascular disease.
There have been many studies about the relationship between long work hours and cerebrocardiovascular disease4, 5), but only a few studies have considered the physical demand of work and physical fitness as well as long work hours11–13). According to these studies, the risk of cerebrocardiovascular disease is related to the physical demands of work and the worker's physical fitness as well as working hours12, 13). Thus, the effects of long work hours on the cerebrocardiovascular system might vary with a worker's physical fitness and the physical demands of his work.
We assumed that long hours of working with a light physical work demand or good physical fitness might not influence the risk of cerebrocardiovascular disease, whereas even short to moderate hours working with a heavy physical work demand or poor physical fitness might influence the risk of cerebrocardiovascular disease. The conditions that could induce physical fatigue and result in an increase in the risk of cerebrocardiovascular disease were considered overwork. The present study was designed to identify the association between overwork and cerebrocardiovascular disease while considering additional factors mentioned above.
The Occupational Safety and Health Research Institute of Korea carried out an occupational cerebrocardiovascular disease surveillance (OCDS) from 2007 to 2010. The targeted reportable diseases were cerebral infarction, intracerebral haemorrhage, subarachnoid haemorrhage, and acute myocardial infarction. Patients who were 20–65 years of age who had been diagnosed with these cereberocardiovascular diseases at the emergency centers of three university hospitals were included in the OCDS registry. To collect general and work-related information, trained interviewers administered a survey questionnaire to each patient at the emergency centers or in the ward after acute treatment for their disease.
The data used in this case-control study were derived from the OCDS surveillance conducted between November 2009 and October 2010. Only data obtained at two of the three Hospitals were used because region-matched control patients were not available from one hospital. A total of 711 patients were registered in the OCDS from November 2009 to October 2010. Patients who were unemployed, with an inaccurate diagnosis, or with incomplete work-related information were excluded. The final size of the case group was 471 patients.
The control group was recruited at the occupational health institutes of four university hospitals from November 2009 to October 2010. The controls were matched for gender, age, type of work (manual and nonmanual), and region with 2 matched controls per case. Control subjects with a history of cerebrocardiovascular disease or incomplete work-related information were excluded. The control group included 836 participants. Trained interviewers administered the same survey questionnaire to controls that was used in the case group. The present study was supported by a grant from the Occupational Safety and Health Research Institute, Korean Occupational Safety and Health Agency. The Institutional Review Board of Seoul St. Mary's Hospital approved the present study (approval ID: KC13RISI0647).
The survey questionnaire collected information regarding general and work-related characteristics of the cases and controls. The general characteristics included age, gender, height, weight, education, smoking, alcohol consumption, exercise, sleep hours per day, and past medical history such as hypertension, diabetes, dyslipidemia and other diseases that could increase the risk of a cerebrocardiovascular disease. Work-related characteristics included occupation, work schedule, and work hours.
The metabolic equivalent of task (MET) was used to represent the physical demand of work, and the maximal oxygen uptake of the subjects represented their physical fitness. We compared the jobs of the subjects with the job categories of the Department of Labor in the United States14) and the study of Ainsworth et al.15). When the identical or most similar job was identified, the MET value of the identified job was used as the subject's physical work demand. Bruce et al.16) suggested the equation for maximal oxygen uptake without exercise testing. The maximal oxygen uptake of the subjects was calculated with this equation using the variables including age, gender, body weight, and exercise. The maximum acceptable work time (MAWT) was defined as the maximum amount of time for which the worker could work without fatigue17). Wu and Wang17) suggested the equation for MAWT using physical work demand and maximal oxygen uptake. We estimated MAWT with this equation using the MET values and the subjects' maximal oxygen uptake. The weekly MAWT was calculated as 5 times the MAWT; because Koreans usually work five days a week. The weekly MAWT was divided by the actual weekly work hours of the subject, and that value was defined as the overwork index.
Case and control group data were compared using the McNemar and paired t-tests. Odds ratios of the overwork index were estimated, with 95% confidence intervals (CI) by conditional logistic regression analysis. Cereberocardiovascular disease was the dependent variable; independent variables included general and work-related characteristics of the case and control participants. Alcohol consumption was categorized as “none or social” or “moderate to heavy”. Moderate to heavy drinking was defined as drinking more than once each week and consuming more than one bottle of soju. Work schedules were categorized as day or shift work, and occupations were categorized as manual or nonmanual work. Blue-collar work and white-collar work were considered manual and nonmanual work, respectively. The hours worked per week were organized into five ranges: <32, 33–40, 41–48, 49–60, and >60 hours. The physical demand of work was expressed in METs and organized into four categories described by the United States Department of Labor14): sedentary work (1.5–2.1), light work (2.2–3.5), moderate work (3.6–4.5), and heavy or very heavy work (≥4.5). Weekly MAWT was grouped into five intervals: ≤32, 33–40, 41–48, 49–60, and >60 hours. The overwork index was expressed in five ranges: ≤0.80, 0.81–1.00 (reference), 1.01–1.20, 1.21–1.50, and >1.50. In the conditional logistic regression analyses for the subjects with <3.6 METs for physical work demand, the overwork index was given as ≤0.8, 0.81–1.00 (reference), 1.01–1.20, and >1.20, because the number of subjects with an overwork index >1.50 was too small. Model 1 was not adjusted for any variables, and Model 2 was adjusted for past medical history, smoking, alcohol consumption, education, hours of sleep per day, work schedule, and work hours per day. Statistical analyses were conducted with SAS windows version 9.2 (SAS Institute Inc, Cary, NC, USA). The statistical significance level was set at p<0.05.
Table 1 shows the general and work-related characteristics of the subjects. The age and gender distributions of the control and case groups were similar. The prevalence of obesity (BMI ≥25 kg/m2) in the case group (34.0%) was higher than in the control group (30.4%), but the difference was not statistically different. The prevalence of current smoking in the case group (52.0%) was significantly higher than in the control group (36.8%). There were more shift and nonmanual workers in the case group compared with the control group (p<0.05). In the case group, the proportion of the workers with long working hours (>60 hours per week) were significantly higher than in the control group (p<0.05).
| Variables | Case (n=471) | Control (n=836) |
|---|---|---|
| Age (years) | ||
| <30 | 8 (1.7) | 18 (2.2) |
| 30–39 | 34 (7.2) | 61 (7.3) |
| 40–49 | 168 (35.7) | 301 (36.0) |
| 50–59 | 184 (39.1) | 339 (40.5) |
| ≥60 | 77 (16.3) | 117 (14.0) |
| Gender | ||
| Male | 389 (17.4) | 692 (17.2) |
| Female | 82 (82.6) | 144 (82.8) |
| Body mass index (m/kg2) | ||
| <25 | 311 (66.0) | 582 (69.6) |
| ≥25 | 160 (34.0) | 254 (30.4) |
| Education* | ||
| Less than high school | 6 (1.3) | 27 (3.3) |
| High school | 311 (70.2) | 520 (63.1) |
| University | 116 (26.2) | 238 (28.9) |
| Graduate school | 10 (2.3) | 39 (4.7) |
| Smoking* | ||
| Never or ex-smoker | 225 (48.0) | 523 (63.2) |
| Current smoker | 244 (52.0) | 304 (36.8) |
| Alcohol consumption | ||
| None or social | 315 (66.9) | 558 (66.6) |
| Moderate or heavy | 156 (33.1) | 279 (33.4) |
| Sleep hours per day* | ||
| <5 | 24 (5.4) | 119 (14.4) |
| 5–6 | 230 (52.2) | 577 (69.9) |
| 7–8 | 136 (30.8) | 98 (11.9) |
| >8 | 51 (11.6) | 31 (3.8) |
| Past medical history* | ||
| No | 224 (47.6) | 553 (66.2) |
| Yes | 247 (52.4) | 283 (33.8) |
| Work schedule* | ||
| Day work | 41 (10.0) | 191 (23.1) |
| Shift work | 368 (90.0) | 636 (76.9) |
| Occupation* | ||
| Manual | 335 (28.9) | 468 (44.0) |
| Nonmanual | 136 (71.1) | 368 (56.0) |
| Work hours per week* | ||
| ≤32 | 121 (25.7) | 101 (12.1) |
| 33–40 | 61 (12.9) | 196 (23.4) |
| 41–48 | 69 (14.7) | 194 (23.2) |
| 49–60 | 109 (23.1) | 221 (26.4) |
| >60 | 111 (23.6) | 124 (14.8) |
Data are shown as numbers and frequencies (%).
Table 2 shows the estimated values for physical demands of jobs, maximal oxygen consumption, MAWT, and overwork index. Most subjects in the control group (46.0%) were engaged in moderate work (3.6–4.5 METs of physical demand). In the case group, 38.8% were employed in light work (2.2–3.5 METs of physical demand), and 32.3% were employed in moderate work (3.6–4.5 METs of physical demand). Maximal oxygen consumption in the case group was significantly lower than that in the control group (p<0.05). The proportions of case subjects with short (≤40.0 hours, 29.9%) and long (>60.0 hours 50.7%) MAWTs per week were higher than in the control group (27.9% and 42.1%, respectively).
| Variables | Case (n=471) | Control (n=836) |
|---|---|---|
| Physical demand of work (METs)*† | ||
| 1.5–2.1 | 82 (17.4) | 203 (24.3) |
| 2.2–3.5 | 183 (38.8) | 106 (12.7) |
| 3.6–4.5 | 152 (32.3) | 385 (46.0) |
| ≥4.5 | 54 (11.5) | 142 (17.0) |
| Physical fitness (ml/kg2)‡ | ||
| Men* | 37.4 ± 3.6 | 39.0 ± 3.6 |
| Women* | 25.0 ± 3.9 | 26.1 ± 3.9 |
| MAWT per week (hours)*§ | ||
| ≤32.0 | 91 (19.3) | 183 (21.9) |
| 32.1–40.0 | 50 (10.6) | 50 (6.0) |
| 40.1–48.0 | 49 (10.4) | 89 (10.6) |
| 48.1–60.0 | 42 (8.9) | 162 (19.4) |
| >60.0 | 239 (50.7) | 352 (42.1) |
| Overwork index*‖ | ||
| ≤0.80 | 296 (62.9) | 470 (56.2) |
| 0.81–1.00 | 38 (8.1) | 90 (10.8) |
| 1.01–1.20 | 30 (6.4) | 71 (8.5) |
| 1.21–1.50 | 33 (7.0) | 45 (5.4) |
| >1.50 | 74 (15.7) | 160 (19.1) |
Data are numbers (%) or means ± standard deviations. MAWT: maximum acceptable work time.
Table 3 shows the results of multiple conditional logistic regression analyses with cerebrocardiovascular disease as the dependent variable. The odds ratios for an overwork index of >1.50 was 2.877 (95% CI, 1.283–6.452) in model 2. Table 4 shows the results of multiple conditional logistic regression analyses with cerebrocardiovascular disease as the dependent variable stratified by physical demand of work. In the subjects with <3.6 METs of physical work demand, the odds ratios were less than 1.0; and were not significant. In subjects with ≥3.6 METs of physical work demand, the odds ratios tended to increase with overwork index after adjusting for possible confounding variables (Model 2). The odds ratios for overwork indexes of 1.01–1.00, 1.21–1.50 and >1.50 were 2.679 (95% CI, 1.025–6.999), 3.124 (95% CI, 1.111–8.783) and 4.331 (95% CI, 1.719–10.908), respectively.
| Overwork index | Model 1 | Model 2 |
|---|---|---|
| ≤0.80 | 0.796 (0.455–1.392) | 1.092 (0.536–2.227) |
| 0.81–1.00 | 1.000 | 1.000 |
| 1.01–1.20 | 1.809 (0.995–3.288) | 1.917 (0.844–4.358) |
| 1.21–1.50 | 1.933 (0.984–3.797) | 2.097 (0.885–4.971) |
| >1.50 | 3.271 (1.731–6.179) | 2.877 (1.283–6.452) |
Data are shown as odds ratios (95% CI). Model 1 was not adjusted with any variables, and Model 2 was adjusted with past medical history, smoking, alcohol consumption, education, sleep hours per day, and work schedule.
| Overwork index | Model 1 | Model 2 |
|---|---|---|
| Physical demand <3.6 METs | ||
| ≤0.80 | 0.516 (0.196–1.359) | 2.326 (0.505–10.714) |
| 0.81–1.00 | 1.000 | 1.000 |
| 1.01–1.20 | 0.779 (0.217–2.791) | 0.843 (0.135–5.256) |
| >1.20 | 0.300 (0.070–1.289) | 0.968 (0.087–10.764) |
| Physical demand ≥3.6 METs | ||
| ≤0.80 | 0.802 (0.405–1.589) | 1.097 (0.433–2.781) |
| 0.81–1.00 | 1.000 | 1.000 |
| 1.01–1.20 | 2.284 (1.149–4.540) | 2.679 (1.025–6.999) |
| 1.21–1.50 | 2.642 (1.258–5.548) | 3.124 (1.111–8.783) |
| >1.50 | 3.708 (1.873–7.338) | 4.331 (1.719–10.908) |
Data are shown as odds ratio (95% CI). Model 1 was not adjusted with any variable, and Model 2 was adjusted with past medical history, smoking, alcohol consumption, education, sleep hours per day, and work schedule.
The variables of age, gender, and body mass index were used in the estimation of physical fitness and the overwork index. For this reason, the variables of age, gender, and body mass index were not adjusted in the multiple conditional logistic regression analyses. When these variables were adjusted in the multiple conditional logistic regression analysis, the odds ratios were overestimated 3.594 rather than 3.124, overwork index 1.21–1.50; 6.354 rather than 4.331, overwork index >1.50, and they were not presented in any tables.
In the present study, overwork increased the risk of cerebrocardiovascular disease, in the subjects engaged in moderate or heavy work, whereas the association was not observed in the subjects engaged in sedentary or light work.
Many studies have reported that long work hours were associated with cerebrocardiovascular disease. Virtanen et al.18) reported that 3–4 hours of overtime work per day was associated with cerebrocardiovascular disease (hazard ratio 1.56, 95% CI, 1.11–2.19). Kivimäki et al.1) reported that working more than 11 hours per day was associated with cerebrocardiovascular disease (hazard ratio 1.67, 95% CI, 1.10–2.55). Other studies reported that the risk of cardiovascular disease varied with physical work demands and the physical fitness of workers11, 12, 19). According to those studies, the risk of cerebrocardiovascular disease from long work hours was increased in workers with low physical fitness and high occupational activity, i.e., physical work demand.
There are several ways to explain the association between overwork and cerebrocardiovascular disease. First, fatigue due to long working hours could increase sympathetic nervous system activity, resulting in and increased in blood pressure and heart rate, which may induce cardiovascular abnormalities and dysfunction20, 21). Second, incomplete recovery from work-induced changes could increase the risk of cerebrocardiovascular disease and mortality22, 23). Third, insufficient sleep might influence the onset of cerebrocardiovascular disease24, 25). Fourth, undesirable lifestyle behaviors could account for a part of the increased risk of cerebrocardiovascular disease. People working long hours tend to have undesirable lifestyle behaviors, such as reduced leisure-time physical activity and unhealthy eating habits that could increase the risk of cerebrocardiovascular disease26). The fifth mechanism is psychosocial stress related to long working hours. Long working hours may produce high job stress that could increase the risk of cerebrocardiovascular disease.22, 27, 28)
Many previous studies of the association between overwork and cerebrocardiovascular disease have adjusted for other risk factors, including age, gender, past medical history, body mass index, behavioral habits, socioeconomic status, and work-related factors4, 29). In the present study, after adjusting with age, gender, body mass index, past medical history, behavioral habits, sleep hours, socioeconomic status, and work-related factors, an overwork index above 1.2 was significantly associated with cerebrocardiovascular disease in subjects engaged in moderate or heavy work. Therefore, insufficient sleep duration and undesirable lifestyle behaviors24, 25), which were the third and fourth mechanisms described above, could not account for the results of the present study.
We have, then, focused on the effects of fatigue and incomplete recovery time, which were the first and second mechanisms. Exhaustion from overwork might lead to increased sympathetic nervous system activity and increased serum norepinephrine levels, blood pressure, and heart rate, which may not have returned to normal upon return to work. Cerebrocardiovascular disease could occur if this vicious pattern continues for a period of time. These mechanisms could account for the results of the present study showing that overwork increased the risk of cerebrocardiovascular disease in the subjects engaged in moderate or heavy work.
The present study found that the risk of cerebrocardiovascular disease increased with the overwork index in the subjects engaged in work with a moderate or high physical demand. The overwork index represents the degree of overwork and was the ratio of the actual working hours and MAWT of the subjects. The MAWT was estimated using the physical fitness and physical work demand of the subjects, and it showed a positive relationship with physical fitness and a negative relationship with physical work demand17). Therefore, the overwork index had a positive relationship with work hours and physical work demand; and a negative relationship with physical fitness. This means that the risk of cerebrocardiovascular disease was high in workers with long work hours, high physical demand of work, and poor physical fitness. In other words, workers engaged in work beyond the capabilities of his or her physical fitness had an increased risk of cerebrocardiovascular disease. This result is consistent with previous studies reporting that the risk of cerebrocardiovascular disease from long working hours was increased in workers with low physical fitness and high physical demand of work11, 12, 19). According to the results of the present study and the above previous studies, the risk of cerebrocardiovascular disease should be decreased in workers given working hours or physical demand of work appropriate to their levels of physical fitness.
In the present study, the risk of cerebrocardiovascular disease increased with the overwork index in subjects with moderate or high physical demand of work, but this association was not observed in subjects engaged in sedentary or light work. Several factors might account for this finding. First, overwork might be infrequent in those engaged in sedentary or light work because the physical demands do not generate fatigue and exhaustion. Second, psychosocial factors might have a greater impact on cerebrocardiovascular disease in the workers engaged in sedentary or light work.
This study has some limitations. First, a causal association between the risk of cerebrocardiovascular disease and overwork could not be determined because of the case-control study design. Second, the values of physical demand of work and maximal oxygen uptake used to determine the overwork index, were estimated, so the overwork index was also an estimated value. Third, selection bias may have occurred because the study subjects were patients who visited a university hospital.
The risk of cerebrocardiovascular disease may be increased in workers with long working hours, high physical demand of work, and poor physical fitness, especially in those with a moderate or high physical demand of work. Therefore, adjusting work hours to accommodate physical fitness levels and the physical demand of work could lower the risk of cerebrocardiovascular disease.
Conflicts of interests: The authors declare that they have no conflicts of interest.
