Abstract
In this study, the cross-sectional association of paid work hours with episodes of work absence was examined in a cohort of police officers. Study subjects were participants from the Buffalo Cardio-Metabolic Occupational Police Stress (BCOPS) study examined between 2004 and 2009. Among 395 study participants with complete data, day-by-day work history records during the one-year period prior to date of examination were used to determine episodes of one-day and three day work absence. The Negative binomial regression analysis was used to examine rate ratios (RR) of work absence. Analyses were also by gender. A one-hour increase in total work hours was associated with 5% reduction in rate of stratified one-day work absence (RR = 0.95, 95% CI: 0.92 – 0.98) and with 8% reduction in rate of three-day work absence (RR = 0.92, 95% CI: 0.89 – 0.95). The association of total work hours with episodes of one-day work absence was significant only in men while the association with episodes of three-day work absence was evident in men and women. In conclusion, in this cohort of police officers, work hours were negatively associated with both durations of work absence (one-day, ≥ 3 consecutive days).
Keywords: police officers, work hours, sickness absence, work history
Long work hours are common in the United States (Caruso, 2006) especially among first responders such as police officers (Vila, 2006). Numerous studies suggest positive association between long work hours and adverse health outcomes including injury, obesity, diabetes, hypertension, cardiovascular diseases, fatigue, stress, depression, and general health complaints (Gu et al., 2012; Charles et al., 2012; Caruso, 2006; Dembe et al., 2005; van der Hulst, 2003; Liu et al., 2002; Lipscomb et al., 2002; Nylén et al., 2001; Park et al., 2001; Shields, 1999). Long work hours have been hypothesized to negatively impact health by impairing recovery time (mentally and physiologically) from work, prolonging exposure to adverse work conditions, and modifying lifestyle behaviors including physical inactivity, reduced sleep, limited involvement in family functions, and increasing the tendency for alcohol and tobacco use (Ala-Mursula et al., 2006; Caruso et al., 2006). In addition, these adverse influences on health are thought to be more profound in women than in men (Artazcoz & Gutiérrez Vera, 2012; Ala-Mursula et al., 2006; Ala-Mursula et al., 2002).
Although there is a substantial body of research linking long work hours to negative health outcomes, comparatively few studies have examined the association of work hours with absenteeism (stress related or sickness absence). In addition, these studies were frequently based on self-reported work hours and/or absenteeism. In the present cross-sectional study, we utilized objective payroll work history records to estimate both work hours and absenteeism and examine their association among a cohort of urban police officers who generally are at high risk for a variety of chronic diseases and are exposed to long and erratic work hours. Gender stratified analyses were also conducted to assess differences in the association of interest between male and female police officers.
METHODS
Design and Study participants
The Buffalo Cardio-Metabolic Occupational Police Stress (BCOPS) study was a cross-sectional study aimed at investigating the associations of occupational stressors with psychological and physiological health of police officers. A total of 710 police officers who worked in the Buffalo Police Department in New York were invited to participate, of which 65.4% (n = 464) officers agreed and were examined between June 4, 2004 to October 2, 2009. The study was approved by the Institutional Review Board of the State University of New York at Buffalo, and the National Institute for Occupational Safety and Health Institutional Review Board (NIOSH IRB). Further details about the study design are reported elsewhere (Hartley et al., 2011, Violanti et al., 2012).
Work hours and episodes of work absence
For this study, work hours per week served as the exposure variable. The outcome of interest consisted of two types of numeric variables representing the count of episodes of one-day and ≥ 3 consecutive days of work absence. Both the exposure and outcome variables were determined using objective work history records of the participants during the 1-year period prior to date of examination. The work history data obtained from the payroll department contained a day-by-day account of activities for each officer including the start time of work, the type of activity (regular work, court work, overtime work), the type of leave (weekend, sickness, work-related injury, vacation), and the number of hours worked on each activity.
The daily hours worked by each participant during the one-year period were partitioned into regular time hours and overtime hours. Regular time hours refer to the hours worked during the regularly scheduled shift as each officer was assigned to 10-hour non-rotating shifts. On the other hand overtime hours, as defined here, may not have the usual meaning of > 40 hours per week. Examination of the payroll work history data reveals that most participants tend to have overtime work hours just prior to or after their regularly scheduled work period (i.e., coming 30-minutes or an hour earlier than the regular start time or leaving 30-minutes or an hour later than the end of regularly scheduled time). In addition, overtime also included hours at court as well as other hours worked outside the regular schedule (e.g., working when scheduled off). Weekly estimates (hours per week) of regular hours, overtime hours, and total (regular + overtime) hours were then derived.
For each work day, the occurrence of work absence due to sickness was identified when the payroll record indicated that an officer was paid but was off-duty due to sickness. We defined a particular work absence as a separate episode if it started following at least one day of full time regular work from the previous sick leave episode. Using this definition, a count of episodes of one-day and three-day sick leave was determined for each participant.
Covariates
On the examination date of the BCOPS study, anthropometric and biological measurements were assessed by trained clinic personnel. In addition, questionnaires were administered to collect demographic and lifestyle characteristics including age, education, marital status, number of children, race/ethnicity, years of police service, rank, smoking, sleep duration, physical activity, and alcohol consumption. Hours of sleep and physical activity were assessed using the Seven-Day Physical Activity Recall questionnaire developed in the Stanford Five-City Project (Sallis et al., 1985). Further details on covariate assessment were reported elsewhere (Charles et al., 2011).
Statistical Analysis
Of the 464 BCOPS study participants, 395 (289 men and 106 women) had work history data during the 1-year period and these subjects were used for the current analyses. The two outcome variables consisted of count data with over-dispersion and hence the negative binomial regression analysis was used as the preferred method to estimate rate ratios (RR) of work absence and their 95% confidence intervals that were associated with a one-hour increase in work hours. In order to examine the role of gender on associations of interest, stratified analyses by gender were conducted. Models were adjusted for potential covariates including age, gender (except when stratifying), race/ethnicity, rank, marital status, number of children, smoking, body mass index, alcohol consumption and sleep hours. The criterion for selection of covariates was relaxed in order to capture all potential confounders. Hence, a covariate was selected for adjustment if associated with either the dependent or exposure variable and this assessment was done separately by type of outcome and gender. Alternative statistical methods were also employed to examine the association of work hours with duration of sickness absence. These included comparing mean episodes of each type of work absence across tertiles of work hours and regression analyses to assess linear trend. For all tests, statistical significance was assessed at the 5% level and all analyses were conducted using the SAS system, version 9.3.
RESULTS
Demographic and lifestyle characteristics
The study sample (Table 1) consisted of 73% males and the majority was white (77%), married (73%), with a rank of patrol officer (71%), never smoked (61%) and overweight (82%). The mean age was 41.4 years (range: 27–66). Male officers had longer years of service (15.3 years), higher BMI (30.5 kg/m2) and alcohol consumption (6 drinks per week) and more children compared to their female counter parts. The work hours ascertained from the payroll work history records do not include paid vacation time, sickness absence or injury leave. Hence the estimate of total work hours from the work history data (33 hrs./wk.) was smaller compared to self-reported hours (43 hrs./wk.).
Table 1.
Demographic and life style characteristics of study participants, BCOPS Study, 2004–2009
| Characteristics | Men n=289 |
Women n=106 |
All n=395 |
|||
|---|---|---|---|---|---|---|
| N | % (±SD) | N | % (±SD) | N | % (±SD) | |
| Age, years | ||||||
| <40 | 119 | 41.2 | 42 | 39.6 | 161 | 40.8 |
| 40–49 | 128 | 44.3 | 55 | 51.9 | 183 | 46.3 |
| ≥50 | 42 | 14.5 | 9 | 8.5 | 51 | 12.9 |
| Race/Ethinicity+ | ||||||
| White | 222 | 78.4 | 76 | 71.7 | 298 | 76.6 |
| Black | 54 | 19.1 | 30 | 28.3 | 84 | 21.6 |
| Hispanic | 7 | 2.5 | 0 | 0.0 | 7 | 1.8 |
| Education+ | ||||||
| ≤High school/GED | 37 | 12.9 | 4 | 3.8 | 41 | 10.4 |
| College <4 yrs | 157 | 54.7 | 66 | 62.3 | 223 | 56.7 |
| College 4+ yrs | 93 | 32.4 | 36 | 34.0 | 129 | 32.8 |
| Marital status+ | ||||||
| Single | 25 | 8.7 | 26 | 24.5 | 51 | 12.9 |
| Married | 226 | 78.8 | 61 | 57.6 | 287 | 73.0 |
| Divorced | 36 | 12.5 | 19 | 17.9 | 55 | 14.0 |
| Smoking status+ | ||||||
| Current | 38 | 13.2 | 27 | 26.2 | 65 | 16.6 |
| Former | 58 | 20.1 | 31 | 30.1 | 89 | 22.8 |
| Never | 192 | 66.7 | 45 | 43.7 | 237 | 60.6 |
| Rank | ||||||
| Patrol officer | 195 | 67.7 | 84 | 79.3 | 279 | 70.8 |
| Sergeant/Lieutenant | 43 | 14.9 | 12 | 11.3 | 55 | 14.0 |
| Captain/Detective | 50 | 17.4 | 10 | 9.4 | 60 | 15.2 |
| Body mass index, kg/m2+ | ||||||
| Normal (<25) | 19 | 6.6 | 51 | 48.6 | 70 | 17.8 |
| Overweight (25–29) | 127 | 43.9 | 37 | 35.2 | 164 | 41.6 |
| Obese (30+) | 143 | 49.5 | 17 | 16.2 | 160 | 40.6 |
| Age, years | 289 | 41.5 ± 6.9 | 106 | 41.1 ± 5.7 | 395 | 41.4 ± 6.6 |
| Years of service, years + | 288 | 15.3 ± 7.2 | 106 | 13.6 ± 6.5 | 394 | 14.8 ± 7.0 |
| Body mass index, kg/m2 + | 289 | 30.5 ± 4.2 | 105 | 26.1 ± 4.7 | 394 | 29.3 ± 4.8 |
| No. of alcohol drinks/week+ | 286 | 6.3 ± 10.6 | 103 | 3.6 ± 6.0 | 389 | 5.6 ± 9.7 |
| Average hours of sleep/day | 288 | 6.2 ± 1.1 | 105 | 6.3 ± 1.2 | 393 | 6.2 ± 1.2 |
| Hours of physical activity/week¥ | 287 | 14.8 ± 12.9 | 105 | 17.2 ± 14.1 | 392 | 15.4 ± 13.3 |
| No. of children+ | 274 | 2.3 ± 1.5 | 104 | 1.8 ± 1.5 | 378 | 2.2 ± 1.5 |
| Self-reported hours at regular shift per week | 286 | 39.8 ± 4.7 | 106 | 39.8 ± 2.4 | 392 | 39.8 ± 4.2 |
| Self-reported overtime hours per week + | 281 | 3.7 ± 5.6 | 101 | 2.1 ± 4.1 | 382 | 3.3 ± 5.3 |
| Self-reported hours per week at second job | 289 | 5.3 ± 7.9 | 106 | 2.0 ± 5.2 | 395 | 4.4 ± 7.4 |
| Total work hours per week1, + | 289 | 32.8 ± 6.6 | 106 | 31.4 ± 4.9 | 395 | 32.4 ± 6.2 |
| Regular time hours per week1 | 289 | 29.6 ± 5.4 | 106 | 29.7 ± 4.1 | 395 | 29.6 ± 5.1 |
| Overtime hours per week1, + | 289 | 3.2 ± 3.0 | 106 | 1.8 ± 2.3 | 395 | 2.8 ± 2.9 |
Results for the continuous variables are means ± SD.
Significant differences between men and women based on χ2 tests of independence or Fisher’s exact for categorical variables and ANOVA for continuous variables (p-values <0.05).
Physical activity hours include occupational, household and leisure time activities.
Work hours derived from payroll work history data.
Episodes of work absence
The count of one-day work absence during the past year ranged from zero to 19 with women showing slightly more frequent episodes (2.7 ± 3.1) compared to men (1.8 ± 3.0). The episodes of three-day work absence (range: 0–12) were also significantly higher for women (1.8 ± 2.7) compared to men (1.2 ± 1.9). Examination of episodes of work absence in relation to other demographic and lifestyle factors (Table 2) revealed that current smokers and patrol officers had the highest mean episodes of one-day work absence while older age (r = −0.25) and longer years of service (r = −0.35) were associated with lower episodes of one-day work absence. On the other hand, episode of three-day work absence was lowest for those with ≥ 4 years of education. Alcohol consumption was positively correlated (r = 0.12) with three-day work absence.
Table 2.
Associations of demographic and life style characteristics with episodes of work absence.
| Characteristics | N | One-day work absence
|
≥3 day work absence
|
||
|---|---|---|---|---|---|
| Mean ±SD | p-value | Mean ±SD | p-value | ||
| Gender | |||||
| Women | 106 | 2.7 ± 3.1 | 0.008 | 1.8 ± 2.7 | 0.007 |
| Men | 289 | 1.8 ± 3.0 | 1.2 ± 1.9 | ||
| Race/Ethinicity+ | |||||
| White | 298 | 2.3 ± 3.1 | 0.056 | 1.3 ± 2.0 | 0.409 |
| Black | 84 | 1.5 ± 2.7 | 1.5 ± 2.5 | ||
| Hispanic | 7 | 0.9 ± 1.2 | 2.1 ± 3.3 | ||
| Education+ | |||||
| ≤High school/GED | 41 | 1.7 ± 2.7 | 0.408 | 1.1 ± 1.4 | 0.004 |
| College <4 yrs | 223 | 2.3 ± 3.2 | 1.6 ± 2.4 | ||
| College 4+ yrs | 129 | 1.9 ± 2.8 | 1.0 ± 1.7 | ||
| Marital status+ | |||||
| Single | 51 | 1.9 ± 2.4 | 0.855 | 1.6 ± 2.6 | 0.448 |
| Married | 287 | 2.1 ± 3.1 | 1.2 ± 2.1 | ||
| Divorced | 55 | 2.1 ± 3.6 | 1.5 ± 2.1 | ||
| Smoking status+ | |||||
| Current | 65 | 2.8 ± 3.5 | 0.035 | 1.8 ± 2.2 | 0.160 |
| Former | 89 | 1.6 ± 2.8 | 1.3 ± 2.1 | ||
| Never | 237 | 2.1 ± 3.0 | 1.2 ± 2.1 | ||
| Rank | |||||
| Patrol officer | 279 | 2.6 ± 3.3 | 0.001 | 1.3 ± 1.9 | 0.547 |
| Sergeant/Lieutenant | 55 | 0.8 ± 1.7 | 1.6 ± 2.7 | ||
| Captain/Detective | 60 | 0.8 ± 1.7 | 1.2 ± 2.6 | ||
| Age, years | 395 | r = −0.25 | 0.001 | r = 0.08 | 0.105 |
| Years of service, years | 394 | r = −0.33 | 0.001 | r = 0.09 | 0.075 |
| Body mass index, kg/m2 | 394 | r = −0.07 | 0.179 | r = −0.10 | 0.056 |
| No. of alcohol drinks/week | 389 | r = 0.04 | 0.461 | r = 0.12 | 0.020 |
| Average hours of sleep/day | 393 | r = 0.08 | 0.096 | r = 0.06 | 0.201 |
| Hours of physical activity/week | 392 | r = 0.01 | 0.801 | r = −0.03 | 0.588 |
| No. of children+ | 378 | r = −0.07 | 0.178 | r = 0.09 | 0.075 |
Results for the categorical covariates are means ± SD. Results for the continuous covariates are correlation coefficients (r).
Work hours and episode of one-day work absence
The rate ratios (RR) of one-day work absence for men and women combined (Table 3) shows that, after accounting for potential covariates, a one-hour increase in overtime hours (RR = 0.92, 95% CI: 0.87 – 0.98), and total hours (RR = 0.95, 95% CI: 0.92 – 0.98) was associated with 8% and 5% reduction in rate of one-day work absence, respectively. Results from stratified analyses by gender (Tables 4 and 5) indicate longer total work hours were significantly associated with lower rate of one-day work absence in men (RR = 0.95, 95% CI: 0.92 – 0.99) but not in women (RR = 0.97, 95% CI: 0.92 – 1.01).
Table 3.
Rate ratios (RR) of episodes of work absence (n=395) in relation to work hours.
| Characteristics | Unadjusted
|
Age-adjusted
|
Multivariable-adjusted
|
|||
|---|---|---|---|---|---|---|
| RR | 95% CI | RR | 95% CI | RR | 95% CI | |
| Episodes of work absence (one-day)1
|
||||||
| Overtime hours per week | 0.91 | 0.86 – 0.96 | 0.90 | 0.86 – 0.95 | 0.92 | 0.87 – 0.98 |
| Total work hours per week | 0.96 | 0.92 – 0.99 | 0.94 | 0.91 – 0.98 | 0.95 | 0.92 – 0.98 |
| Episodes of work absence (≥3 consecutive days)2
|
||||||
| Overtime hours per week | 0.93 | 0.87 – 0.98 | 0.92 | 0.87 – 0.98 | 0.94 | 0.88 – 0.99 |
| Total work hours per week | 0.91 | 0.88 – 0.93 | 0.91 | 0.88 – 0.93 | 0.92 | 0.89 – 0.95 |
For the multivariable-adjusted model, adjustment was made for age, gender, race, smoking status, rank, and sleep hours.
Adjustment was made for age, gender, rank, education, BMI, alcohol consumption and number of children.
Table 4.
Rate ratios (RR) of episodes of work absence in relation to work hours among men (n=289).
| Characteristics | Unadjusted
|
Age-adjusted
|
Multivariable-adjusted
|
|||
|---|---|---|---|---|---|---|
| RR | 95% CI | RR | 95% CI | RR | 95% CI | |
| Episodes of work absence (one-day)1
|
||||||
| Overtime hours per week | 0.92 | 0.86 – 0.98 | 0.90 | 0.85 – 0.96 | 0.94 | 0.88 – 1.00 |
| Total work hours per week | 0.95 | 0.91 – 0.99 | 0.94 | 0.90 – 0.98 | 0.95 | 0.92 – 0.99 |
| Episodes of work absence (≥3 consecutive days)2
|
||||||
| Overtime hours per week | 0.93 | 0.86 – 0.99 | 0.93 | 0.86 – 0.99 | 0.94 | 0.88 – 1.00 |
| Total work hours per week | 0.91 | 0.88 – 0.95 | 0.92 | 0.88 – 0.95 | 0.95 | 0.92 – 0.99 |
For the multivariable-adjusted model, adjustment was made for age, race, rank, and alcohol consumption.
Adjustment was made for age, race, rank, education, alcohol consumption, sleep hours, and number of children.
Table 5.
Rate ratios (RR) of episodes of work absence in relation to work hours among women (n=106).
| Characteristics | Unadjusted
|
Age-adjusted
|
Multivariable-adjusted
|
|||
|---|---|---|---|---|---|---|
| RR | 95% CI | RR | 95% CI | RR | 95% CI | |
| Episodes of work absence (one-day)1
|
||||||
| Overtime hours per week | 0.94 | 0.84 – 1.05 | 0.94 | 0.85 – 1.05 | 0.96 | 0.86 – 1.07 |
| Total work hours per week | 0.98 | 0.93 – 1.03 | 0.97 | 0.93 – 1.02 | 0.97 | 0.92 – 1.01 |
| Episodes of work absence (≥3 consecutive days)2
|
||||||
| Overtime hours per week | 0.98 | 0.87 – 1.09 | 0.96 | 0.86 – 1.08 | 0.96 | 0.85 – 1.09 |
| Total work hours per week | 0.89 | 0.84 – 0.94 | 0.89 | 0.85 – 0.94 | 0.89 | 0.85 – 0.94 |
For the multivariable-adjusted model, adjustment was made for age, race, rank, marital status, BMI, and sleep hours.
Adjustment was made for age, race, rank, and sleep hours.
Work hours and episodes three-day work absence
After adjusting for covariates, a one-hour increase in overtime hours (RR = 0.94, 95% CI: 0.88 – 0.99) and total hours (RR = 0.92, 95% CI: 0.89 – 0.95) was associated with 6% and 8% reduction in rate of three-day work absence, respectively. Stratified analyses by gender (Tables 4 and 5) show that total work hours were inversely associated with three-day work absence in both men (RR = 0.95, 95% CI: 0.92 – 0.99) and women (RR = 0.89, 95% CI: 0.85 – 0.94).
Results from the alternative statistical analyses (analyses of covariance) are displayed in Figure 1 which shows covariate adjusted mean episodes of work absence by tertiles of total work hours. The data indicates that among men, the average episode of one-day work absence was largest for those in the lowest tertile and that all pairwise mean comparisons across the three categories were significant in men but not in women. The decreasing linear trends in one-day work absence with increasing work hours were not significant at the 5% level (trend p-value = 0.174 for men, 0.282 for women). A significant reduction in episodes of three-day work absence across increasing tertiles of total work hours in both men (trend p-value <0.0001) and women (trend p-value <0.0001) are shown.
Figure 1.
Mean episodes of one-day and three-day work absence by tertiles of total work hours for male and female officers. The numbers of subjects in the low, medium and high categories were 96, 97, 96 respectively for men and 35, 36, and 35 for women.
DISCUSSION
The study examined the association of paid work hours and work absenteeism (one-day, ≥ 3 consecutive days) among police officers and assessed the impact of gender on the association. Both variables of interest (work hours and absenteeism) were derived based on objective day-to-day work-history records of the officers for a one-year period. Although the work history data do not indicate the type of sickness or its severity, it is an administrative policy to provide certification from a physician when the work absence lasts three consecutive days or longer. Hence absence of three or more consecutive days was treated as proxy measure of physician-certified sickness absence. It is possible that one-day work absences included both stress-related work absence and short term absence due to physical sickness. On average, women had approximately 3-episodes of one-day work absence and 2-episodes of three-day work absence per year while the estimates for men were 2 and 1-episodes respectively. During the 1-year period, over half of the study participants experienced one or more episodes of one-day or three-day work absences.
After adjusting for a number of demographic and life style factors, results showed that work hours were inversely and significantly associated with episodes of one-day work absence (in men) and episodes of three-day work absence in both genders. For the whole cohort, a one-hour increase in total work hours was associated with 5% reduction in rate of one-day work absence (RR = 0.95, 95% CI: 0.92 – 0.98) and with 8% reduction in rate of three-day work absence (RR = 0.92, 95% CI: 0.89 – 0.95). The magnitude of the inverse association between work hours and episodes of one-day work absence were similar regardless of gender (interaction p-value > 0.680) but the significance of the association did vary by gender. On the other hand, the effect of work hours on three-day work absence although evident in both genders appears to be more pronounced among women than men (interaction p-value = 0.276); a one-hour increase in total work hours were associated with 11% reduction in rate of three-day work absence in women and with 5% reduction in men.
Literature on work hours and absenteeism is scarce but the inverse association from the current analyses is consistent with one previous report. A prospective study of public sector employees in Finland (Ala-Mursula et al., 2006) reported that long paid work hours were associated with lower rates of self-certified sickness absence (≤ 3 days) among both men and women. Among women the rate of absenteeism for those who worked 40 to 50 hours per week was 8% lower (RR = 0.92, 95% CI: 0.88 – 0.96) compared to those who worked less than 40 hours per week, while among men the rate of work absence was 21% smaller (RR = 0.79, 95% CI: 0.73 – 0.85). The associations of paid work hours with medically certified sickness absence (>3 days) was not significant in both men (RR = 0.97, 95% CI: 0.93 – 1.02) and women (RR = 0.90, 95% CI: 0.82 – 1.00). However, the study reported that total work hours (paid + domestic + commuting) were associated with higher rates of medically certified sickness absence. Other studies (Van der Hulst, 2003; Spurgeon et al., 1997, Sparks et al., 1997) also point out that the effects of work hours depend on the type of health outcome measured; positive association with health complaints, CVD, diabetes, etc. but inverse association with sickness absenteeism.
Work absenteeism is not a simple function of ill health or other individual factors such as job dissatisfaction (Kristensen T., 1991, Eriksen et al., 2003; Ala-Mursula et al., 2006, Duijts et al., 2007, Foss et al., 2011, Fekedulegn et al., 2013). In fact, the causes of work absenteeism could be multifactorial and complex in nature as it may involve a number of interacting physical and psychosocial conditions at individual, organizational, and household levels. These additional factors may include perceived control over working hours, job demand and control, coping method, social support, burnout, shift schedule, psychosomatic and psychological problems, marital status, and medication use.
The results from the current analyses give rise to a number of questions. For example, how should these findings be interpreted and how should they be related to existing knowledge about the links between long work hours and health or absenteeism? The answers to such questions are of importance from a scientific viewpoint. Some possible explanations for the inverse association of work hours with absenteeism could include (1) those who work long hours could be committed employees and hence may work while ill or take sick leave only when absolutely necessary (Ala-Mursula et al., 2006), (2) those employees engaged in long work hours may be the healthiest (healthy worker effect) and hence may contribute to the inverse association between long work hours and low rates of absence, and (3) long hours may result in overtime pay and hence officers may use such opportunity to gain additional income. Our data however do not appear to support the healthy worker hypothesis. The proportion of officers who self-reported their general health as excellent or very good did not differ (p=0.376) by tertiles of work hours (90.1%, 94.7%, and 92.4% for low, moderate, and high work hours respectively). The prevalence of metabolic syndrome, a clustering of CVD risk factors, was not significantly different (p=0.159) across tertiles of work hours (23.4%, 25.2%, and 33.3% for low, moderate, and high work hours respectively). The proportion of officers with BMI <25kg/m2 was significantly lower (p=0.031) among those in the highest tertile of work hours (9.9%) compared to those in the middle (24.4%) and lowest (19.1%) tertiles. In addition, we conducted analyses of work absence and regular time work hours (excluding overtime hours) and the result showed significant associations with one-day (RR = 0.96, 95% CI: 0.92 – 0.99) and three-day (RR = 0.89, 95% CI: 0.86 – 0.93) work absence. This may suggest that the inverse association may not be due to overtime pay (income).
The study has limitations and strengths. Shortcomings include the cross-sectional study design which limits casual inference. The type of sickness was not available. The study sample consisted of only police officers which limits external validity of the results. Due to lack of variability in self-reported work hours (95% of the officers self-reported working exactly 40 hrs/wk), we were not able to examine the association of self-reported hours and work absence. One of the strengths of this study is the use of routinely collected objective data detailing day-to-day work history records from which work hours and sickness absence were ascertained. The concept of one-day sickness absence includes absences due to stress or fatigue (the so called “mental health” day off from work) which usually lasts for a shorter duration whereas absences lasting three or more consecutive days are more likely related to sickness absence with physical symptoms.
In summary, this study showed that officers working longer hours were less likely to have one-day and three-day work absences. The findings are in contrast with prior studies that indicate adverse health outcomes associated with long work hours. This study adds to the body of knowledge regarding the association of work hours with absenteeism among high stress occupations. Determination of the underlying reasons for the inverse association requires further studies with larger samples and a prospective design that would enable estimation of the true incidence rates of absenteeism associated with long work hours and also identification of potential mediators and effect modifiers of the association. Future studies ought to consider commuting and domestic work hours in addition to paid work hours when assessing effects on absenteeism.
Acknowledgments
Sources of Funding: This work was supported by the National Institute for Occupational Safety and Health contract number 200-2003-01580.
Footnotes
Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.
Contributor Information
Desta Fekedulegn, National Institute for Occupational Safety and Health, Ctrs. for Disease Control & Prevention, Morgantown, WV.
Cecil M. Burchfiel, National Institute for Occupational Safety and Health, Ctrs. for Disease Control & Prevention, Morgantown, WV
Tara A. Hartley, National Institute for Occupational Safety and Health, Ctrs. for Disease Control & Prevention, Morgantown, WV
Penelope Baughman, National Institute for Occupational Safety and Health, Ctrs. for Disease Control & Prevention, Morgantown, WV.
Luenda E. Charles, National Institute for Occupational Safety and Health, Ctrs. for Disease Control & Prevention, Morgantown, WV
Michael E. Andrew, National Institute for Occupational Safety and Health, Ctrs. for Disease Control & Prevention, Morgantown, WV
John M. Violanti, University at Buffalo, Buffalo, NY The State University of New York, Buffalo, NY
References
- Ala-Mursula L, Vahtera J, Kivimäki M, Kevin MV, Pentti J. Employee control over working times: Associations with subjective health and sickness absences. Journal of Epidemiology and Community Health. 2002;4:272–278. doi: 10.1136/jech.56.4.272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ala-Mursula L, Vahtera J, Kouvonen A, Väänänen A, Linna A, Pentti J, et al. Long hours in paid and domestic work and subsequent sickness absence: Does control over daily working hours matter? Occupational and Environmental Medicine. 2006;9:608–616. doi: 10.1136/oem.2005.023937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Artazcoz L, Gutiérrez Vera A. Gender differences in the relationship between long working hours and health status in Catalonia. Arch Prev Riesgos Labor. 2012;3:129–135. doi: 10.12961/aprl.2012.15.3.03. [DOI] [PubMed] [Google Scholar]
- Caruso CC. Possible broad impacts of long work hours. Industrial Health. 2006;4:531–536. doi: 10.2486/indhealth.44.531. [DOI] [PubMed] [Google Scholar]
- Caruso CC, Bushnell T, Eggerth D, Heitmann A, Kojola B, Newman K, et al. Long working hours, safety, and health: Toward a National Research Agenda. American Journal of Industrial Medicine. 2006;11:930–942. doi: 10.1002/ajim.20373. [DOI] [PubMed] [Google Scholar]
- Charles LE, Gu JK, Andrew ME, Violanti JM, Fekedulegn D, Burchfiel CM. Sleep duration and biomarkers of metabolic function among police officers. Journal of Occupational and Environmental Medicine. 2011;8:831–837. doi: 10.1097/JOM.0b013e31821f5ece. [DOI] [PubMed] [Google Scholar]
- Charles LE, Fekedulegn D, Burchfiel CM, Fujishiro K, Landsbergis P, Diez Roux AV, et al. Associations of work hours with carotid intima-media thickness and ankle-brachial index: The Multi-Ethnic Study of Atherosclerosis (MESA) Occupational and Environmental Medicine. 2012;10:713–720. doi: 10.1136/oemed-2012-100665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dembe AE, Erickson JB, Delbos RG, Banks SM. The impact of overtime and long work hours on occupational injuries and illnesses: New evidence from the United States. Occupational and Environmental Medicine. 2005;9:588–597. doi: 10.1136/oem.2004.016667. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Duijts SF, Kant I, Swaen GM, van den Brandt PA, Zeegers MP. A meta-analysis of observational studies identifies predictors of sickness absence. Journal of Clinical Epidemiology. 2007;60:1105–1115. doi: 10.1016/j.jclinepi.2007.04.008. [DOI] [PubMed] [Google Scholar]
- Eriksen W, Bruusgaard D, Knardahl S. Work factors as predictors of sickness absence: A three month prospective study of nurses’ aides. Occupational and Environmental Medicine. 2003;60:271–278. doi: 10.1136/oem.60.4.271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fekedulegn D, Burchfiel CM, Hartley TA, Andrew ME, Charles LE, Tinney-Zara CA, et al. Shiftwork and sickness absence among police officers: the BCOPS study. Chronobiology International. 2013;7:930–941. doi: 10.3109/07420528.2013.790043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Foss L, Gravseth HM, Kristensen P, Claussen B, Mehlum IS, Knardahl S, Skyberg K. The impact of workplace risk factors on long-term musculoskeletal sickness absence: A registry-based 5-year follow-up from the Oslo health study. Journal of Occupational and Environmental Medicine. 2011;53:1478–1482. doi: 10.1097/JOM.0b013e3182398dec. [DOI] [PubMed] [Google Scholar]
- Gu JK, Charles LE, Burchfiel CM, Fekedulegn D, Sarkisian K, Andrew ME, et al. Long work hours and adiposity among police officers in a US north-east city. Journal of Occupational and Environmental Medicine. 2012;11:1374–1381. doi: 10.1097/JOM.0b013e31825f2bea. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hartley TA, Burchfiel CM, Fekedulegn D, Andrew ME, Knox SS, Violanti JM. Associations between police officer stress and the metabolic syndrome. International Journal of Emergency Mental Health. 2011;4:243–256. [PMC free article] [PubMed] [Google Scholar]
- Kristensen TS. Sickness absence and work strain among Danish slaughterhouse workers: An analysis of absence from work regarded as coping behavior. Social Science and Medicine. 1991;1:15–27. doi: 10.1016/0277-9536(91)90122-s. [DOI] [PubMed] [Google Scholar]
- Lipscomb JA, Trinkoff AM, Geiger-Brown J, Brady B. Work-schedule characteristics and reported musculoskeletal disorders of registered nurses. Scandinavian Journal of Work, Environment and Health. 2002;28:394–401. doi: 10.5271/sjweh.691. [DOI] [PubMed] [Google Scholar]
- Liu Y, Tanaka H The Fukuoka Heart Study Group. Overtime work, insufficient sleep, and risk of non-fatal acute myocardial infarction in Japanese men. Occupational and Environmental Medicine. 2002;59:447–451. doi: 10.1136/oem.59.7.447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nylen L, Voss M, Floderus B. Mortality among women and men relative to unemployment, part time work, overtime work: A study based on data from the Swedish twin registry. Occupational and Environmental Medicine. 2001;58:52–57. doi: 10.1136/oem.58.1.52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park J, Kim Y, Chung H, Hisanaga N. Long working hours and subjective fatigue symptoms. Industrial Health. 2001;39:250–254. doi: 10.2486/indhealth.39.250. [DOI] [PubMed] [Google Scholar]
- Sallis JF, Haskell WL, Wood PD, Fortmann SP, Rogers T, Blair SN, et al. Physical activity assessment methodology in the Five-City Project. American Journal of Epidemiology. 1985;121:91–106. doi: 10.1093/oxfordjournals.aje.a113987. [DOI] [PubMed] [Google Scholar]
- Shields M. Long working hours and health. Health Reports. 1999;11:33–48. [PubMed] [Google Scholar]
- Sparks K, Cooper C, Fried Y, Shirom A. The effects of hours of work on health: A meta-analytic review. Journal of Occupational and Organizational Psychology. 1997;70:391–408. [Google Scholar]
- Spurgeon A, Harrington JM, Cooper CL. Health and safety problems associated with long working hours: A review of the current position. Occupational and Environ Medicine. 1997;54:367–375. doi: 10.1136/oem.54.6.367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- van der Hulst M. Long work hours and health. Scandinavian Journal of Work, Environment and Health. 2003;3:171–188. doi: 10.5271/sjweh.720. [DOI] [PubMed] [Google Scholar]
- Vila B. Impact of long work hours on police officers and the communities they serve. American Journal of Industrial Medicine. 2006;11:972–980. doi: 10.1002/ajim.20333. [DOI] [PubMed] [Google Scholar]
- Violanti JM, Fekedulegn D, Andrew ME, Charles LE, Hartley TA, Vila B, et al. Shift work and the incidence of injury among police officers. American Journal of Industrial Medicine. 2012;3:217–227. doi: 10.1002/ajim.22007. [DOI] [PubMed] [Google Scholar]