Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Sleep problems are a strong predictor of stress-related metabolic changes in police officers. A prospective study

  • Sergio Garbarino ,

    Roles Conceptualization, Data curation, Visualization, Writing – review & editing

    ‡ These authors are co-first authors on this work.

    Affiliations Post-graduate School of Occupational Health, Università Cattolica del Sacro Cuore, Rome, Italy, State Police Health Service Department, Ministry of the Interior, Rome, Italy, Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics and Maternal-Infantile Sciences (DINOGMI), Genoa, Italy

    ⨯
  • Nicola Magnavita

    Roles Formal analysis, Methodology, Writing – original draft, Writing – review & editing

    nicolamagnavita@gmail.com

    ‡ These authors are co-first authors on this work.

    Affiliations Post-graduate School of Occupational Health, Università Cattolica del Sacro Cuore, Rome, Italy, Department of Woman/Child & Public Health, Fondazione Policlinico Gemelli IRCCS, Rome, Italy

    ⨯

Abstract

Objective

Previous studies have shown that workers chronically exposed to occupational stress have an increased risk of metabolic syndrome (MetS) and sleep problems (SPs). The purpose of this study was to verify whether SPs mediate the relationship between stress and MetS.

Method

A 5-year prospective cohort study included 242 police officers from a rapid response unit engaged exclusively in maintaining law and order. Perceived stress levels were measured repeatedly with the demand-control-support and the effort-reward-imbalance questionnaires; insomnia symptoms were assessed with the Pittsburgh Sleep Quality Index; excessive daytime sleepiness was measured using the Epworth Sleepiness Scale. MetS and its components were evaluated at baseline and at follow-up.

Results

During 5-year follow-up period, 26 new cases of MetS were identified. Both occupational stress and SPs were significantly related to incident cases of MetS. Insomnia symptoms showed a highly significant association with MetS (aOR 11.038; CI95% 2.867–42.493). Mediation analysis confirmed that SPs mediate the relationship between stress and MetS.

A reciprocal relationship was found between job stress and SPs. Work-related stress was a significant predictor of insomnia symptoms, short sleep duration, sleep dissatisfaction, and sleepiness. Compared to the reference group, police officers with SPs at baseline had significantly higher odds of reporting high stress at follow-up.

Conclusion

SPs play a mediating role in the relationship between occupational stress and MetS. Prevention of MetS must include the control of stress factors and an increase in the resilience of workers, but correct sleep hygiene is also an essential factor.

Citation: Garbarino S, Magnavita N (2019) Sleep problems are a strong predictor of stress-related metabolic changes in police officers. A prospective study. PLoS ONE 14(10): e0224259. https://doi.org/10.1371/journal.pone.0224259

Editor: Juan Manuel Marquez-Romero, Instituto Mexicano del Seguro Social (IMSS) HGZ 2, MEXICO

Received: April 16, 2019; Accepted: September 25, 2019; Published: October 22, 2019

Copyright: © 2019 Garbarino, Magnavita. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Data have been deposited on Zenodo, DOI: 10.5281/zenodo.3376509.

Funding: The Università Cattolica del Sacro Cuore of Rome contributed to the funding of this research project and its publication. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Metabolic syndrome (MetS) is a cluster of metabolically related cardiovascular risk factors (obesity, insulin resistance, dyslipidaemia and hypertension) [1] that also predict a high risk of developing diabetes [2], cancer [3], and cognitive impairment [4]. MetS is highly prevalent and increasing in most parts of the world [5].

Chronic occupational stress is considered to be a risk factor for MetS [6–16]. A recent meta-analysis of longitudinal studies showed that work-related stress is associated with a significant increase in the pooled risk of metabolic syndrome (RR = 1.47; CI95% = 1.22–1.78) [17]. In a previous longitudinal study on a rapid response police unit we observed that police officers with high job strain had a higher adjusted risk of developing MetS (aOR = 2.68; CI95% = 1.08–6.70) and hypertriglyceridemia (aOR = 7.86; CI95% = 1.29–48.04) than their colleagues classified in the lowest quartile of work-related distress [18].

It is well known that sleep disorders, such as obstructive sleep apnea (OSAS), are associated with MetS [19, 20]. It would be interesting to determine whether alterations in the quantity and quality of sleep that can occur in working environments, regardless of the existence of morbid conditions such as OSAS, are able, if continued over time, to induce MetS. Studies that consider only sleep duration have produced inconsistent results. For example, short sleep duration proved to be an independent risk factor for MetS in longitudinal studies conducted in Korea [21], China [22], Canada [23] and Japan [24], while in an Iranian cohort study, long sleep duration increased the risk of MetS [25]. In a Swedish cohort, both short and long sleep durations, and problems with falling and staying asleep were linked to a higher prevalence of MetS [26]. In a sample of police officers, poor sleeping conditions, i.e. lack of sound sleep and shift sleep disorder, were associated with a high prevalence of MetS [27]. A recent review of existing cross-sectional studies [28] provided some evidence of a positive association between sleep quality, or some sleep complaints (difficulty in falling asleep, difficulty in maintaining sleep, and sleep inefficiency) and MetS. However, further studies based on a longitudinal design are needed to explore the causal relationship between sleep quality and MetS.

Work is known to disturb sleep, due to occupational exposure to many stressors [29]. Some aspects of work organization, such as lack of control over work times [30], insufficient rest between shifts [31], psychosocial burden and night work [32] or shift work [33] are predictors of sleep problems. Exposure to workplace violence is also associated with a significant increase in the pooled risk of sleep problems [34]. More generally, there is evidence that psychosocial stress is associated with sleep problems [35], and that both sleep deprivation [36, 37] and low sleep quality [38–40] are associated with stress response. Due to considerable disparity in the results of existing studies, the degree to which the relationship between job stress and sleep symptoms is unidirectional or reciprocal is unclear.

The questions to be answered in this study are: can the presence of sleep symptoms be a predictor of MetS, and in what way are sleep, stress and metabolic syndrome related? To answer these questions, we used the same cohort of workers whose surveillance revealed the existence of an association between stress and metabolic syndrome. In this cohort we studied the evolution of sleep symptoms. The main aim of the research was to assess the extent to which sleep problems were associated with an increased risk of MetS at the end of a five–year period of observation. The collateral objective was to study the relationships between occupational stress and sleep problems.

Methods

Subjects and ethics statement

The study population, which was described in detail in the previous study [18] was composed of police officers from the “VI Reparto Mobile” of Genoa. This rapid response police unit was chosen because it is exclusively engaged in maintaining law and order and is therefore homogeneously exposed to relevant factors of occupational stress. In this paper, in accordance with the literature [41–43], we used the term "stress" to describe workplace stressors that lead to strain [44], imbalance [45] or other negative responses or reactions, called “distress”. Studies on this cohort showed that occupational stress can be higher during routine, but often unpredictable activities, than during highly important public events when activities are carefully programmed in advance [46]. Although occupational exposure was homogeneous, as all the police officers performed the same tasks, personal response to stressors varied due to difference in personality [47] or individual characteristics [48]. As a result, the health consequences of stress were diverse [49]. The study population was not exposed to some of the confounding factors of MetS, such as a sedentary lifestyle, food or alcohol abuse or drug use since all the police officers regularly followed an intensive weekly program of physical exercise and a healthy diet, designed specifically to guarantee a correct nutritional intake. At the time of enrolment, the police officers were in good health and had no sleep-disordered breathing or other diseases associated with MetS.

In 2009, prior to an international G8 meeting in which they were responsible for maintaining law and order, the police officers were invited to participate in a work-related health program that included frequent monitoring. The study plan had previously been approved by the Italian Ministry of the Interior and by all police trade-union organizations. An information leaflet informed participants of the voluntary and confidential nature of the study through an information leaflet. All but two police officers accepted (292 out of 294 police officers) and gave their written informed consent by signing a form in their personal health file report. Written informed consent for access to health records and linkage was also obtained in compliance with indications in the literature [50] and the Italian Law on Privacy [51]. The study was conducted according to the ethical standards of the Declaration of Helsinki. Two female participants were excluded from subsequent analyses in order to avoid gender bias. At the end of the observation period, the cohort was composed of 234 police officers. No worker dropped out of the study, and no police officer ever submitted incomplete or non-assessable data. The reasons for abandoning the cohort were: transfer to other police units (41 persons, 73.2%); retirement (12, 21.4%), and death (3, 5.4%).

Questionnaires

Data were collected from January 2009 (T1) to January 2014 (T2) through medical examination and questionnaires. Self-assessment of job stress was performed during the 5-year period of observation by repeatedly completing the Italian version [52] of Karasek’s demand-control-support (DCS) questionnaire [44], and Siegrist’s Effort–Reward Imbalance (ERI) questionnaire [45]. In this cohort Cronbach’s α values for the sub-scales of DCS, job demand and job control, were 0.71 and 0.65, i.e. close to those found in the literature [53–57], thus indicating reasonably acceptable internal consistency [58]. The Cronbach’s α values for the sub-scales job effort and job reward of the ERI questionnaire were 0.82 and 0.89, respectively, showing high internal consistency. Blood tests and medical examinations were performed at baseline (T1) and at the follow-up (T2).

In addition to the aforementioned and already published data, in 2009 and 2014, the police officers also answered some questions from the Pittsburgh Sleep Quality Index (PSQI), [59] a self-report measure of sleep quality over a 1-month time interval, and from the SDS (sleep disorder score), an Italian questionnaire on sleep [60], in compliance with a validated protocol [61]. Evidence indicates that the PSQI is a highly reliable and valid questionnaire [62]. In this study the selected items had a Cronbach’s alpha of 0.717. The questionnaire provided a numerical variable (HS, hours of sleep, in response to the question: “During the past month, how many hours of actual sleep did you get at night?”) and three ordinal variables (DS, difficulty sleeping; SI, sleep interruptions; EA, early awakenings) from a frequency Likert scale score of 1 to 4 in response to the following questions: “During the past month, how often have you had trouble falling asleep within 30 minutes?”, “During the past month, how often have you had sleeping problems because you wake up in the middle of the night?”, “During the past month, how often have you had sleeping problems because you wake up early in the morning?”. HS were also recoded in a binary variable: police officers who slept less than 7 hours were classified as insufficient sleepers. The duration of daily sleep was then used both as an outcome (with two categories) and as a predictor variable.

The ranked T1 and T2 scales (DS, SI and EA) were also divided into no or infrequent symptoms (not during the past month, less than once a week, or once a week) and experiencing symptoms (more than once a week). The variables were used both as an outcome (with two categories, divided as described above) and as a predictor. Moreover, since insomnia is defined in the International Classification of Sleep Disorders, Third Edition [63] as difficulty in initiating or maintaining sleep, while the Australasian Sleep Edition indicates that it is a disorder characterized by difficulty with sleep initiation, sleep maintenance, early awakening, or a combination of the above difficulties [64], and the DSM-5 categorizes insomnia disorder as a sleep-wake disorder, characterized by dissatisfaction with sleep quantity or quality, associated with one (or more) of the aforementioned symptoms [65], we pooled the DS, SI and EA ranked scales in a single variable that measured insomnia symptoms. A principal component analysis of the aforementioned ordinal variables was conducted to ensure that sleep disorder variables could be incorporated in a single variable. The results suggested a one-factor solution (R2 = 44.22% of the variance with T1 variables, R2 = 50.96% of the variance with T2 variables). The insomnia symptom scale (IS), that had values ranging from 3 to 12, was used as a predictor variable.

Another ordinal variable, sleep satisfaction (SS), was obtained from a 4-point Likert scale response to the question: “During the past month, how satisfied were you with your overall sleep quantity and quality?” This variable was also used both as an outcome (with two categories) and as a predictor variable.

Sleepiness was evaluated using the Epworth Sleepiness Scale ESS [66], a self-administered questionnaire with 8 questions that rated on a 4-point scale (0–3) the chances of dozing off or falling asleep while engaged in eight different activities. A systematic review has shown that the internal consistency of the questionnaire measured by Cronbach’s alpha is good [67]. In our sample the alpha value (0.79) fell within the limits of the literature (0.73–0.86). ESS score was used as a predictor. An ESS score >10 indicated excessive daytime sleepiness (EDS) [68,69]; EDS was used as a binary outcome.

Both the demand / control (D/C) ratio and the effort / reward (E/R) ratio were used to assess occupational stress level. Karasek’s and Siegrist’s stress models are complementary and, if used together, make it possible to evaluate both the effect of the interaction between workloads and control, and the transactional effects caused by disparity between effort and the results obtained. As is known, there is no safe level: the effects of stress depend on individual resources and on the perception that each individual has of his/her own condition. Since police officers were exposed to homogeneous conditions of work stress, our aim was to determine the relative risk that each individual attributed to this exposure. For each of the five measurements made during the period, the quartiles of the two distributions were added and the final result was again subdivided into quartiles. In this way, the population was divided into four groups; the first consisted of subjects who reported the highest levels of stress, whereas the fourth was composed of resilient police officers who, although performing identical tasks, perceived the lowest levels of stress. By grouping the first two categories and the last two, a binary variable was also constructed, consisting of a high and low level of perceived stress.

The International Diabetes Federation (IDF) guide [70] and the National Cholesterol Education Program Expert Panel on Detection Evaluation and Treatment of High Cholesterol in Adults (NCEP/ATPIII) [71] were used to define metabolic syndrome components as follows: central obesity was considered to be a BMI>30 kg/m2, or a waist circumference of >85 cm in men. Hypertriglyceridemia was defined as a serum triglyceride level >150 mg/dL (1.7 mmol/L), while low high-density lipoprotein (HDL) cholesterolemia was represented by a serum HDL-cholesterol level of <40 mg/dL (1.03 mmol/L). High blood pressure was defined as a systolic blood pressure of >130 mmHg and/or a diastolic blood pressure of >85 mmHg. High fasting glucose was defined by a plasma glucose level of >100 mg/dL (5.6 mmol/L).

Statistical analyses

Firstly, the variables of interest were analyzed by assessing central tendency and variability measures. Comparison of the values of stress-related variables was made by Student’s t-test for coupled data, and the prevalence of sleep problems at baseline and at follow-up was compared by means of the chi-square test. The prevalence and incidence of metabolic disorders were also assessed.

In a further step, a study was made of the relationship between stress and sleep problems. The four groups classified according to the level of stress were compared by one-way analysis of variance (Anova) and the Bonferroni post-hoc comparison of group means to ascertain how sleep problems were distributed at the beginning and end of the observations. The distribution of sleep hours (SH); insomnia symptoms (IS) including difficulty sleeping (DS), sleep interruptions (SI) and early awakenings (EA); sleep satisfaction (SS) and daytime sleepiness (Epworth scale score, ESS) were analyzed in relation to stress levels. In this way we were able to ascertain whether, within the population, those who perceived a higher level of occupational stress also had a higher frequency of sleep problems. Hierarchical logistic regression analyses were then performed, in order to evaluate the unidirectional or bidirectional path of the association. The variables for job stress or sleep problems were entered in the first set to estimate the crude odds ratios. Subsequently, demographic variables [i.e. age (years), education level (dichotomized at 8 years of schooling), rank (officer or superior/technician), geographic origin (Northern or Southern Italy), housing (in barracks or home), marital status (single or divorced/ married or cohabiting) and presence of offspring (no/yes)] were added. The baseline variable of sleep or stress (depending on the research hypothesis) was included in the third set. The remaining covariates were added in the final set.

As a third step, multiple hierarchical logistic regression analysis was used to ascertain whether sleep problems or a combination of sleep problems and job stress were associated with metabolic syndrome incidence. Initially (model I), the average number of hours slept per day was entered as a predictor, and the incidence of metabolic syndrome as a dependent variable. In the following models, daytime sleepiness (model II), work-related stress (model III), and insomnia symptoms (model IV) were entered as predictors. Analyses were adjusted for confounding factors (age, rank, education, geographic origin, marital status, housing, offspring).

Lastly, univariate and multivariate logistic regression analyses were used once more to determine which of the sleep problems best correlated with the incidence of MetS and its components.

An assessment of the possible mediation effect of insomnia in the causal pathway linking occupational stress and metabolic syndrome was performed using the approach proposed by Hayes [72]. Firstly, calculations were made of the total (unadjusted) effect of occupational stress on metabolic syndrome as well as the regression coefficients of the links between occupational stress and insomnia (“a” pathway—by linear regression analysis) and between insomnia and metabolic syndrome (“b” pathway—by logistic regression analysis). Secondly, a multiple logistic regression model was constructed by taking metabolic syndrome as the dependent variable and both occupational stress and insomnia (i.e. the mean value of insomnia scores at baseline and at the end of follow-up) as independent variables. Lastly, the bootstrapping indirect effect of occupational stress (i.e. mediated by insomnia) was calculated and expressed by regression coefficient and bootstrapping confidence interval. The number of bootstrap samples was fixed at 5000. Analyses were performed using the Statistical Package for Social Science (IBM/ SPSS) for Windows (rel. 23.0).

Results

The characteristics of the observed sample are reported in Table 1. At baseline, the mean age was 36.0+7.4 years, and the average length of work experience was 14.6+7.8 years. According to Italian educational standards, three out of four of the participants had an average or high level of education (>8 years of education). Rank was basic (officer) for approximately half of the sample, while it was higher (special officer, superintendent, inspector, technician) for the other half, with only slight differences in salary between the various ranks, and practically no difference in the tasks performed. Half of the participants were born in Northern Italy, while the remainder came from Southern Italy. Most of the participants were quartered in barracks.

thumbnail
Download:
Table 1. Characteristics of the population studied.

https://doi.org/10.1371/journal.pone.0224259.t001

Stress levels showed some fluctuations during the observation period. A comparison of average follow-up and baseline values revealed a significant increase in control, effort, reward, and over-commitment. However, no significant changes were observed in composite distress measures, i.e. D/C and E/R weighted ratios (Table 1). During the 5-year observation period, 207 police officers (88.5%) reported a mean D/C ratio>1, which is considered a condition of job strain, and 29 (12.4%) reported a mean E/R ratio>1, considered to be indicative of effort/reward imbalance. An analysis of the different measurements performed during the 5-year period indicated that 45 officers (19.2) had high stress levels, 31 (13.2%), 82 (35%) were classified in the intermediate quartiles of distress, and 76 (32.5%) manifested low job-related stress.

The initial overall sleep hygiene picture was not optimal, with about half of the workers reporting insufficient sleep (sleeping less than 7 hours a day) and one in four claiming to be dissatisfied with the quantity or quality of sleep. At follow-up, all sleep indicators had worsened. On average, sleep time declined from 6.7 ± 1.1 to 6.4 ± 1.2 hours (t Student’s t for paired data = 6.29; p <0.001). The share of those presenting DS, SI, or EA increased from 12% to 21%; however, pre-post statistically significant differences were found only for sleep interruptions (5.1% vs. 11.1%, p<0.001), early awakening (5.1% vs. 7.3%, p<0.001), sleep dissatisfaction (23.9% vs. 27.4%, p<0.001) and daytime sleepiness (17.9% vs. 28.6%, p<0.001). The prevalence of sleep quality problems (DS, SI, EA, SS) increased from 29.5% at T1 to 36.8% at T2 (p<0.001) (Table 1).

All the components of metabolic syndrome were found to have a significantly higher prevalence at follow-up than at baseline (chi square or Fisher’s exact test p<0.001). Over the 5-year period 112 new cases of abdominal obesity (incidence:0.7), 41 of hypertension (0.20), 8 of hypertriglyceridemia (0.04), 11 of low HDL-cholesterol (0.05), and 3 of hyperglycemia (0.01) were observed leading to the diagnosis of 26 new cases of metabolic syndrome (0.12).

One-way Anova demonstrated that sleep symptoms differed significantly between groups of officers characterized by different levels of stress (p<0.001). Post-hoc comparisons using the Bonferroni test confirmed that the police officers with persistently lower levels of perceived stress had significantly better sleep conditions than their colleagues in the highest quartile of job strain. The police officers in the lowest stress quartile were significantly more satisfied with the quality and quantity of sleep, and had fewer insomnia symptoms and lower daytime sleepiness than other groups (p<0.001 for all comparisons). On average, the non-distressed police officers slept more than 7 hours a day (on average, 7.1±0.9 hours), while the policemen in the highest quartile of perceived job stress slept less than 6 hours a day (5.0±0.8; Student’s t test p<0.001) (Table 2).

thumbnail
Download:
Table 2. Stress and sleep problems.

Difference in prevalence of sleep complaints (average scores) between groups exposed to different levels of work stress at baseline and at follow-up.

https://doi.org/10.1371/journal.pone.0224259.t002

Sleep problems were significantly related to incident cases of metabolic syndrome. The average number of hours of sleep had a protective effect. Using logistic regression analysis, the association between the number of hours slept and the prevalence of MetS was non-significant at baseline (OR 0.736; CI95% 0.434–1.247). At follow-up sleep hours were significantly associated with MetS prevalence (OR 0.568; CI95% 0.414–0.780; p<0.001). In longitudinal observation the average number of hours slept during the observation period was significantly associated with the occurrence of MetS cases (crude OR 0.476; CI95% 0.310–0.730; p<0.001). The adjusted OR (corrected for age, rank level, education, geographic origin, marital status, housing, and presence of offspring) was 0.452 (CI95% = 0.289–0.710; p = 0.001), and remained significant even after correction for sleepiness, stress and sleep satisfaction (aOR 0.468; CI95% = 0.2327–0.942; p = 0.033). However, a much stronger association was observed between the quality of sleep and MetS than between the latter and the quantity of sleep: after adding insomnia symptoms to the multivariate hierarchical logistic model, hours of sleep were no longer significant, and insomnia symptoms showed a highly significant aOR (11.038; CI95% = 2.867–42.493; p<0.001). The coefficient of determination of the model improved significantly when insomnia symptoms were introduced into the analysis (Table 3) (Fig 1).

thumbnail
Download:
Table 3. Association of sleep problems with incident cases of metabolic syndrome.

Hierarchical regression analysis.

https://doi.org/10.1371/journal.pone.0224259.t003

thumbnail
Download:
Fig 1. Mediation analysis of the inter-relationships between occupational stress, insomnia and metabolic syndrome (unstandardized coefficients).

https://doi.org/10.1371/journal.pone.0224259.g001

Findings for the relationship between insomnia symptoms and MetS components revealed an unadjusted significant association with incident cases of dyslipidemia (hypertriglyceridemia OR = 62.65; CI95% 9.33–420.72; p<0.001; low HDL-cholesterol OR = 9.96; CI95% 2.28–43.54; p = 0.002). The effects of insomnia symptoms were robust as the addition of adjustment variables in three separate sets failed to produce noticeable alterations in these estimates (Table 4).

thumbnail
Download:
Table 4. Metabolic effects of insomnia.

Odds ratios and 95% confidence intervals of Insomnia symptoms (DS, SI, EA) on incident cases of metabolic syndrome and its components. Hierarchical regression analysis.

https://doi.org/10.1371/journal.pone.0224259.t004

The crude estimates showed that high stress among policemen predicted higher odds of sleep disorders at follow-up compared to the reference group. Work-related stress was a significant predictor of insomnia symptoms (Table 5), short sleep duration (Table 6), sleep dissatisfaction (Table 7), and sleepiness (Table 8). Adjusted ratios were still significant, although the associations were weaker.

thumbnail
Download:
Table 5. Effect of stress on insomnia symptoms.

Multivariate hierarchical logistic regression analysis.

https://doi.org/10.1371/journal.pone.0224259.t005

thumbnail
Download:
Table 6. Effect of stress on short sleep duration.

Multivariate hierarchical logistic regression analysis.

https://doi.org/10.1371/journal.pone.0224259.t006

thumbnail
Download:
Table 7. Effect of stress on sleep satisfaction.

Multivariate hierarchical logistic regression analysis, according to stress levels.

https://doi.org/10.1371/journal.pone.0224259.t007

thumbnail
Download:
Table 8. Effect of stress on sleepiness.

Multivariate hierarchical logistic regression analysis.

https://doi.org/10.1371/journal.pone.0224259.t008

The crude estimates also indicated that police officers with sleep problems had significantly higher odds of reporting a high level of work-related stress than the reference group (Table 9). Univariate logistic regression analysis showed that insomnia symptoms, sleep hours, sleep satisfaction and daytime sleepiness were significant determinants of distress. Insomnia symptoms (OR = 7.72 CI95% 3.84–15.51) and excessive daytime sleepiness (OR = 10.00 CI95% 5.22–19.16) were associated with high stress levels, while hours of sleep (OR = 0.17, CI95% 0.11–0.28) and sleep satisfaction (OR = 0.14, CI95% 0.08–0.22) had a reverse association. These associations remained significant after correction for demographic variables. However, the addition of sleep complaints that were present at baseline into the model reduced the association between sleepiness and stress. In the fully corrected model, only sleep duration, insomnia symptoms and sleep satisfaction were significant predictors of the perception of occupational stress.

thumbnail
Download:
Table 9. Effect of sleep problems on distress.

Multivariate hierarchical logistic regression analysis.

https://doi.org/10.1371/journal.pone.0224259.t009

Mediation analysis

The results of the mediation analysis are illustrated in Fig 1. The total (unadjusted) effect of occupational stress on MetS was statistically significant (regression coefficient: 0.382, p = 0.039) and this was also true for the links between occupational stress and insomnia (“a” pathway, regression coefficient: 0.92, p<0.001) and between insomnia and metabolic syndrome (“b” pathway, regression coefficient: 0.452, p<0.001). A multiple logistic regression model that took MetS as the dependent variable and both occupational stress and insomnia (i.e. the mean value of insomnia scores at baseline and at the end of follow-up) as independent variables, demonstrated that the relationship between occupational stress and MetS was no longer significant (regression coefficient: -0.066, p = 0.78) after data adjustment for insomnia. An assessment of the bootstrapping indirect effect of occupational stress (i.e. mediated by insomnia) revealed that the regression coefficient of such an effect was 0.438, with a bootstrapping confidence interval that did not include zero (i.e. it was statistically significant) ranging from 0.182 to 0.801.

Discussion and conclusions

This study demonstrates that low sleep quantity and quality are powerful predictors of MetS, and lends support to the hypothesis that the association between some sleep complaints (difficulty sleeping, sleep interruptions, and early awakenings) and MetS, that has been frequently observed in cross-sectional studies, also implies the existence of a causal link. Moreover, our study supports the idea that sleep loss is a causal factor of MetS, and confirms the findings of some previous studies on this topic.

Likewise, the present study demonstrates that sleep problems mediate the effect of occupational stress on MetS. Should this result be generalized in longitudinal studies on other populations, researchers would have at their disposal an indicator of MetS risk that would be much simpler to use than the questionnaires for occupational stress. In fact, the latter require a certain competence on the part of the physician who must first of all select the stress model best suited to the population and then evaluate the results correctly. Moreover, workers may find it difficult to express their distress openly. Although investigating sleep symptoms can be much easier than investigating stress, very few doctors systematically collect sleep-related symptoms.

The study also sheds new light on the relationship between stress and sleep problems. Relatively consistent findings in the literature have demonstrated that experiencing job stress enhances the risk of sleep problems. The results of the present study clearly show a bidirectional relationship between sleep and stress: workers exposed to chronic occupational stress have an increased incidence of sleep problems, and bad sleepers suffer more from occupational stress factors than good sleepers. This result reinforces similar observations from previous studies. Data from the Swedish Longitudinal Occupational Survey of Health (SLOSH) detected reverse and reciprocal relationships in addition to the commonly hypothesized causal relationships between occupational stress and sleep problems [73]. More recently, a Norwegian study confirmed the commonly hypothesized unidirectional forward associations between occupational stress and sleep problems among women, and demonstrated a reverse and reciprocal relation between work stressors and sleep problems among men [74]. Data from the Workplace Bullying and Harassment cohort in Denmark also found evidence for a reciprocal association [75]. A small reverse effect, with reduced sleep duration (but not sleep quality) predicting increased post-traumatic stress, was also observed in World Trade Center (WTC) responders [76].

The results of this research should be interpreted principally in relation to members of the police force. Poor sleep quality is very common in the law enforcement profession. According to a recent systematic review with meta-analysis, 51% of police officers report bad sleep quality [77]. Sleep problems have been associated with night and evening work schedules [78] or abnormal work hours [79]. However, very few studies have been conducted that take into consideration work-related stressors other than night shifts, or the association between stress and sleep problems. Some studies have examined the way stress can alter sleep quality, depending upon the different coping strategies adopted by police officers [80]. Reverse casualty, i.e. that sleep problems can influence the response to stress, has also been studied: a small sample of Iowa law enforcement officers with short sleep duration or poor sleep quality reported more stress, burnout, and depression symptoms than controls [81]. The present study confirms previous experiences, thus indicating that sleep problems in police officers may be connected with MetS and cardiovascular risk [82, 83].

The results of this study on policemen confirm findings reported in research conducted on the general population and other categories of workers as regards the association between stress and cardiovascular risk factors. Studies undertaken in the 1960s and conducted up to the present day have revealed that work-related stress is associated with cardiovascular risk [84]. A summary of the studies that used the ERI model to assess occupational stress as the imbalance between the effort required for work tasks and the rewards received, specifically indicates that stress increases cardiovascular risk [85]. Furthermore, a very recent systematic review of the literature reported a strong association between stress and alteration in biological indicators that is especially consistent with MetS, alteration in blood lipids and variability in heart rate [86]. Similar evidence has been found with the DCS model [87]. Surveys conducted on over 600,000 workers confirm that stress induces MetS [88–90]. Our study carries forward this research by indicating that sleep acts as a mediating factor in the relationship between occupational stress and cardiovascular risk.

The main strength of the present study lies in the fact that it benefited from a selected cohort in which exposure to relevant risk factors (stress, abnormal shifts) was homogeneous and did not include any important confounders for MetS (gender, diet, physical activity). A further strong point was the frequent assessment of stress levels by means of two complementary and standardized questionnaires thus enabling the police officers to be classified into different levels of resilience.

The results of this study need to receive experimental confirmation in other working populations. The main limitation of the study was the scant number of observations that inevitably led to a high standard error. Our findings must be interpreted with caution, in consideration of the range of the confidence intervals that makes it difficult to relate our observations to the general population. The observed sample can however be considered similar to other police units engaged in maintaining law and order. The limitation due to the meager number of workers was offset by constant participation and the negligible loss of cases. The cheap and simple method used in this research should encourage occupational health services to verify our results by conducting similar studies in other productive sectors.

The conclusions of a previous study [18] highlighted the need to avoid stress in police officers in order to prevent MetS, while the present study indicates the importance of avoiding a vicious circle involving stress and sleep problems as this would be more likely to cause MetS than stress alone. Health promotion in the police force should have the two-fold aim of increasing resilience to stressors and spreading healthy sleep patterns and proper sleep hygiene practices.

Acknowledgments

We thank Ms. Elisabeth Ann Wright for assistance with the English text of this manuscript. We thank Prof. Giovanni Tripepi, biostatistician, MSc, PhD, First Researcher from the Italian National Research Centre (CNR-IBIM), Reggio Calabria, Italy, for assistance in the revision of statistical analyses.

The Università Cattolica del Sacro Cuore of Rome contributed to the funding of this research project and its publication. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

  1. 1. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome—a new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med. 2006;23(5):469–80. pmid:16681555
  2. 2. van der Berg JD, Stehouwer CD, Bosma H, van der Velde JH, Willems PJ, Savelberg HH, et al. Associations of total amount and patterns of sedentary behaviour with type 2 diabetes and the metabolic syndrome: The Maastricht Study. Diabetologia. 2016;59(4):709–18. pmid:26831300
  3. 3. Esposito K, Chiodini P, Colao A, Lenzi A, Giugliano D. Metabolic syndrome and risk of cancer: a systematic review and meta-analysis. Diabetes Care. 2012;35(11):2402–11. pmid:23093685
  4. 4. Coteanu C, Gusti A, Coteanu MF, Gusti S. Association of Some Components of the Metabolic Syndrome with the Subtype of Mild Cognitive Impairment. Curr Health Sci J. 2017;43(4):351–354. pmid:30595902
  5. 5. Mendrick DL, Diehl AM, Topor LS, Dietert RR, Will Y, La Merrill MA, et al. Metabolic Syndrome and Associated Diseases: From the Bench to the Clinic. Toxicol Sci. 2018;162(1):36–42. pmid:29106690
  6. 6. Magnavita N, Fileni A. Work stress and metabolic syndrome in radiologists: first evidence. Radiol Med. 2014;119(2):142–8. pmid:24297580
  7. 7. Magnavita N. Work-related psychological injury is associated with metabolic syndrome components in apparently healthy workers. PLoS One, 2015; 10(6):e0130944. pmid:26086387
  8. 8. Kang MG, Koh SB, Cha BS, Park JK, Woo JM, Chang SJ. Association between job stress on heart rate variability and metabolic syndrome in shipyard male workers. Yonsei Med J. 2004;45(5):838–46. pmid:15515194
  9. 9. Kawada T. Relationship between components of the metabolic syndrome and job strain using a brief job stress questionnaire (BJSQ). Int Arch Occup Environ Health. 2013;86(6):725–6. pmid:23584233
  10. 10. Djindjic N, Jovanovic J, Djindjic B, Jovanovic M, Jovanovic JJ. Associations between the occupational stress index and hypertension, type 2 diabetes mellitus, and lipid disorders in middle-aged men and women. Ann Occup Hyg. 2012;56(9):1051–62. pmid:22986427
  11. 11. Chandola T, Brunner E, Marmot M. Chronic stress at work and the metabolic syndrome: prospective study. BMJ. 2006;332:521–525. pmid:16428252
  12. 12. Edwards EM, Stuver SO, Heeren TC, Fredman L. Job strain and incident metabolic syndrome over 5 years of follow-up: the coronary artery risk development in young adults study. J Occup Environ Med 2012 54 1447–1452. pmid:23171915
  13. 13. Bergmann N, Gyntelberg F, Faber J. The appraisal of chronic stress and the development of the metabolic syndrome: a systematic review of prospective cohort studies. Endocr Connect. 2014;3(2):R55–80. pmid:24743684
  14. 14. Siegrist J, Li J. Work Stress and Altered Biomarkers: A Synthesis of Findings Based on the Effort-Reward Imbalance Model. Int J Environ Res Public Health. 2017;14(11). pii: E1373. pmid:29125555
  15. 15. Magnavita N, Capitanelli I, Garbarino S, Pira E. Work-related stress as a cardiovascular risk factor in police officers. A systematic review of evidence. Int Arch Occup Environ Health, pmid:29344727
  16. 16. Pedersen JM, Lund R, Andersen I, Clark AJ, Prescott E, Rod NH. Psychosocial risk factors for the metabolic syndrome: A prospective cohort study. Int J Cardiol. 2016;215:41–6. pmid:27107545
  17. 17. Watanabe K, Sakuraya A, Kawakami N, Imamura K, Ando E, Asai Y, et al. Work-related psychosocial factors and metabolic syndrome onset among workers: a systematic review and meta-analysis. Obes Rev. 2018;19(11):1557–1568. pmid:30047228
  18. 18. Garbarino S, Magnavita N. Work Stress and Metabolic Syndrome in Police Officers. A Prospective Study. PLoS ONE 2015;10(12): e0144318. pmid:26641879
  19. 19. Garbarino S, Lanteri P, Durando P, Magnavita N, Sannita WG. Co-morbidity, mortality, quality of life and the healthcare/welfare/social costs of disordered sleep. Int J Environ Res Public Health. 2016;13(8). pii: E831. pmid:27548196
  20. 20. Gagnadoux F, Priou P, Urban T, Meslier N, Trzepizur W. Sleep apneas, metabolic syndrome and cardiovascular risk: Data from the Pays de la Loire sleep cohort. Rev Pneumol Clin. 2017;73(6):299–305. pmid:29122399
  21. 21. Kim JY, Yadav D, Ahn SV, Koh SB, Park JT, Yoon J, et al. A prospective study of total sleep duration and incident metabolic syndrome: the ARIRANG study. Sleep Med. 2015;16(12):1511–5. pmid:26611949
  22. 22. Song Q, Liu X, Zhou W, Wang X, Wu S. Changes in sleep duration and risk of metabolic syndrome: the Kailuan prospective study. Sci Rep. 2016;6:36861. pmid:27857185
  23. 23. Chaput JP, McNeil J, Després JP, Bouchard C, Tremblay A. Short sleep duration as a risk factor for the development of the metabolic syndrome in adults. Prev Med. 2013;57(6):872–7. pmid:24099879
  24. 24. Itani O, Kaneita Y, Tokiya M, Jike M, Murata A, Nakagome S, et al. Short sleep duration, shift work, and actual days taken off work are predictive life-style risk factors for new-onset metabolic syndrome: a seven-year cohort study of 40,000 male workers. Sleep Med. 2017;39:87–94. pmid:29157594
  25. 25. Ostadrahimi A, Nikniaz Z, Faramarzi E, Mohammadpoorasl A, Ansarin K, Somi MH. Does long sleep duration increase risk of metabolic syndrome in Azar cohort study population? Health Promot Perspect. 2018; 27;8(4):290–295. pmid:30479983
  26. 26. Titova OE, Lindberg E, Elmståhl S, Lind L, Schiöth HB, Benedict C. Associations Between the Prevalence of Metabolic Syndrome and Sleep Parameters Vary by Age. Front Endocrinol (Lausanne). 2018;9:234. pmid:29867766
  27. 27. Shiozaki M, Miyai N, Morioka I, Utsumi M, Koike H, Arita M, et al. Assessment of the risk of ischemic heart disease and its relevant factors among Japanese police officers. Sangyo Eiseigaku Zasshi. 2013;55(4):115–24. pmid:23676527
  28. 28. Lian Y, Yuan Q, Wang G, Tang F. Association between sleep quality and metabolic syndrome: A systematic review and meta-analysis. Psychiatry Res. 2019;274:66–74. pmid:30780064
  29. 29. Magnavita N, Garbarino S. Sleep, Health and Wellness at Work: A Scoping Review. Int J Environ Res Public Health. 2017;14(11). pii: E1347. pmid:29113118
  30. 30. Salo P, Ala-Mursula L, Rod NH, Tucker P, Pentti J, Kivimäki M, et al. Work time control and sleep disturbances: prospective cohort study of Finnish public sector employees. Sleep. 2014;37(7):1217–25. pmid:25061250
  31. 31. Vedaa Ø, Harris A, Bjorvatn B, Waage S, Sivertsen B, Tucker P, et al. Systematic review of the relationship between quick returns in rotating shift work and health-related outcomes. Ergonomics. 2016;59(1):1–14. pmid:26072668
  32. 32. Chazelle E, Chastang JF, Niedhammer I. Psychosocial work factors and sleep problems: findings from the French national SIP survey. Int Arch Occup Environ Health. 2016;89(3):485–95. pmid:26376909
  33. 33. Burman D. Sleep Disorders: Circadian Rhythm Sleep-Wake Disorders. FP Essent. 2017;460:33–36. pmid:28845960
  34. 34. Magnavita N, Di Stasio E, Capitanelli I, Lops EA, Chirico F, Garbarino S. Sleep problems and workplace violence: A systematic review and meta-analysis. Frontiers in Neuroscience, 2019 (in press)
  35. 35. Akerstedt T. Psychosocial stress and impaired sleep. Scand J Work Environ Health. 2006;32(6):493–501 pmid:17173205
  36. 36. Schwarz J, Gerhardsson A, van Leeuwen W, Lekander M, Ericson M, Fischer H, et al. Does sleep deprivation increase the vulnerability to acute psychosocial stress in young and older adults? Psychoneuroendocrinology. 2018;96:155–165. pmid:29982098
  37. 37. Ahmed AE, Albalawi AN, Alshehri AA, AlBlaihed RM, Alsalamah MA. Stress and its predictors in pregnant women: a study in Saudi Arabia. Psychol Res Behav Manag. 2017;10:97–102. pmid:28435340
  38. 38. Bassett SM, Lupis SB, Gianferante D, Rohleder N, Wolf JM. Sleep quality but not sleep quantity effects on cortisol responses to acute psychosocial stress. Stress. 2015;18(6):638–44. pmid:26414625
  39. 39. Goodin BR, Smith MT, Quinn NB, King CD, McGuire L. Poor sleep quality and exaggerated salivary cortisol reactivity to the cold pressor task predict greater acute pain severity in a non-clinical sample. Biol Psychol. 2012;91(1):36–41. pmid:22445783
  40. 40. Solomonova E, Lee YEA, Robins S, King L, Feeley N, Gold I, et al. Sleep quality is associated with vasopressin methylation in pregnant and postpartum women with a history of psychosocial stress. Psychoneuroendocrinology. 2019;107:160–168. pmid:31132568
  41. 41. Glazer S, Liu C. Work, Stress, Coping, and Stress Management. In: Oxford Research Encyclopedia of Psychology. Published online April 2017 | e-ISBN: 9780190236557 | https://dx.doi.org/10.1093/acrefore/9780190236557.013.30
    • 42. Burr H, Formazin M, Pohrt A. Methodological and conceptual issues regarding occupational psychosocial coronary heart disease epidemiology. Scand J Work Environ Health. 2016;42(3):251–5. pmid:26960179
    • 43. Chirico F. Job stress models for predicting burnout syndrome: a review. Ann Ist Super Sanita. 2016;52(3):443–456. pmid:27698304
    • 44. Karasek RA. Job demands. job decision latitude. and mental strain. Implication for job redesign. Adm Sci Q 1979; 24: 285–308
    • 45. Siegrist J. Adverse health effects of high-effort/low-reward conditions. J Occup Health Psychol 1996; 1:27–41 pmid:9547031
    • 46. Garbarino S, Magnavita N, Elovainio M, Heponiemi T, Ciprani F, Cuomo G, et al. Police job strain during routine activities and a major event. Occup Med (Lond). 2011; 61(6):395–9
    • 47. Garbarino S, Chiorri C, Magnavita N. Personality traits of the Five-Factor Model are associated with work-related stress in special force police officers. Int Arch Occup Environ Health 2014;87(3):295–306. pmid:23463443
    • 48. Chiorri C, Garbarino S, Bracco F, Magnavita N. Personality Traits Moderate the Effect of Workload Sources on Perceived Workload in Flying Column Police Officers. Front. Psychol. 2015;6:1835. pmid:26640456
    • 49. Garbarino S, Cuomo G, Chiorri C, Magnavita N. Association of work-related stress with mental health problems in a special police force. BMJOpen 2013;3(7). pii: e002791. pmid:23872288
    • 50. da Silva ME, Coeli CM, Ventura M, Palacios M, Magnanini MM, Camargo TM, et al. Informed consent for record linkage: a systematic review. J Med Ethics. 2012;38(10):639–42. pmid:22403083
    • 51. Legislative Decree 196, 30/6/2003 “Codice in materia di protezione dei dati personali”. http://garanteprivacy.it/web/guest/home/docweb/-/docweb-display/docweb/1311248
      • 52. Magnavita N. Two tools for health surveillance of job stress: the Karasek Job Content Questionnaire and the Siegrist Effort Reward Imbalance Questionnaire. G Ital Med Lav Ergon 2007; 29 (3): 667–670
      • 53. Sale JE, Kerr MS. The psychometric properties of Karasek’s demand and control scales within a single sector: data from a large teaching hospital. Int Arch Occup Environ Health. 2002; 75(3):145–52. pmid:11954981
      • 54. Niedhammer I. Psychometric properties of the French version of the Karasek Job Content Questionnaire: a study of the scales of decision latitude, psychological demands, social support, and physical demands in the GAZEL cohort. Int Arch Occup Environ Health. 2002; 75(3):129–44. pmid:11954980
      • 55. Griep RH, Rotenberg L, Vasconcellos AG, Landsbergis P, Comaru CM, Alves MG. The psychometric properties of demand-control and effort-reward imbalance scales among Brazilian nurses. Int Arch Occup Environ Health. 2009; 82(10):1163–72. pmid:19756699
      • 56. Sanne B, Torp S, Mykletun A, Dahl AA. The Swedish Demand-Control-Support Questionnaire (DCSQ): factor structure, item analyses, and internal consistency in a large population. Scand J Public Health. 2005; 33(3):166–74. pmid:16040456
      • 57. Fujishiro K, Landsbergis PA, Diez-Roux AV, Stukovsky KH, Shrager S, Baron S. Factorial invariance, scale reliability, and construct validity of the job control and job demands scales for immigrant workers: the multi-ethnic study of atherosclerosis. J Immigr Minor Health. 2011; 13(3):533–40. pmid:20582720
      • 58. Pelfrene E, Vlerick P, Mak R, Smets Pd, Kornitzer M, Backer Gd. Scale reliability and validity of the Karasek ‘Job Demand–Contol–Support’ model in the Belstress study. Work Stress 2001; 15:297–313.
      • 59. Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index (PSQI): A new instrument for psychiatric research and practice. Psychiatry Res 1989; 28(2), 193–213 pmid:2748771
      • 60. Garbarino S, De Carli F, Nobili L, Mascialino B, Squarcia S, Penco MA, et al. Sleepiness and sleep disorders in shift workers: a study on a group of Italian police officers. Sleep. 2002;25(6):648–53. pmid:12224843
      • 61. Garbarino S, Magnavita N. Obstructive sleep apnea syndrome (OSAS), metabolic syndrome and mental health in small enterprise workers. Feasibility of an action for health. PLoS One. 2014; 9(5): e97188. pmid:24810290
      • 62. Mollayeva T, Thurairajah P, Burton K, Mollayeva S, Shapiro CM, Colantonio A. The Pittsburgh sleep quality index as a screening tool for sleep dysfunction in clinical and non-clinical samples: A systematic review and meta-analysis. Sleep Med Rev. 2016;25:52–73. pmid:26163057
      • 63. American Academy of Sleep Medicine. International Classification of Sleep Disorders, 3rd ed. American Academy of Sleep Medicine, Darien, IL, 2014.
        • 64. Australasian Sleep Association. Insomnia. https://www.sleep.org.au/documents/item/355
          • 65. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th ed.; DSM-5). Washington, DC: 2013.
            • 66. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14(6):540–5. pmid:1798888
            • 67. Kendzerska TB, Smith PM, Brignardello-Petersen R, Leung RS, Tomlinson GA. Evaluation of the measurement properties of the Epworth sleepiness scale: a systematic review. Sleep Med Rev. 2014;18(4):321–31. pmid:24135493
            • 68. Johns MW. Sleepiness in different situations measured by the Epworth Sleepiness Scale. Sleep. 1994;17(8):703–10. pmid:7701181
            • 69. Johns MW. Epworth Sleepiness Scale, official website. http://epworthsleepinessscale.com/about-the-ess/ Accessed July 7, 2018.
              • 70. International Diabetes Foundation IDF. 2006. The IDF consensus worldwide definition of the metabolic syndrome. http://www.idf.org/webdata/docs/IDF_Meta_def_final.pdf
                • 71. National Cholesterol Education Program: Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001; 285: 2486–2497 pmid:11368702
                • 72. Hayes AF. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York, NY: The Guilford Press; 2013.
                  • 73. Hanson LL, Åkerstedt T, Näswall K, Leineweber C, Theorell T, Westerlund H. Cross-lagged relationships between workplace demands, control, support, and sleep problems. Sleep. 2011;34(10):1403–10. pmid:21966072
                  • 74. Johannessen HA, Sterud T. Psychosocial factors at work and sleep problems: a longitudinal study of the general working population in Norway. Int Arch Occup Environ Health. 2017;90(7):597–608. pmid:28429107
                  • 75. Hansen ÅM, Grynderup MB, Rugulies R, Conway PM, Garde AH, Török E, et al. A cohort study on self-reported role stressors at work and poor sleep: does sense of coherence moderate or mediate the associations? Int Arch Occup Environ Health. 2018;91(4):445–456. pmid:29383438
                  • 76. Dietch JR, Ruggero CJ, Schuler K, Taylor DJ, Luft BJ, Kotov R. Posttraumatic stress disorder symptoms and sleep in the daily lives of World Trade Center responders. J Occup Health Psychol. 2019 Jun 17. pmid:31204820
                  • 77. Garbarino S, Guglielmi O, Puntoni M, Bragazzi NL, Magnavita N. Sleep Quality among Police Officers: Implications and Insights from a Systematic Review and Meta-Analysis of the Literature. Int J Environ Res Public Health. 2019;16(5). pii: E885. pmid:30862044
                  • 78. Fekedulegn D, Burchfiel CM, Charles LE, Hartley TA, Andrew ME, Violanti JM. Shift Work and Sleep Quality Among Urban Police Officers: The BCOPS Study. J Occup Environ Med. 2016;58(3):e66–71. pmid:26949891
                  • 79. Violanti JM, Burchfiel CM, Hartley TA, Mnatsakanova A, Fekedulegn D, Andrew ME, et al. Atypical work hours and metabolic syndrome among police officers. Arch Environ Occup Health. 2009;64(3):194–201. pmid:19864222
                  • 80. Hartley TA, Violanti JM, Sarkisian K, Fekedulegn D, Mnatsakanova A, Andrew ME, et al. Association Between Police-Specific Stressors and Sleep Quality: Influence of Coping and Depressive Symptoms. J Law Enforc Leadersh Ethics. 2014;1(1):31–48. pmid:28480317
                  • 81. Yoo H, Franke WD. Sleep habits, mental health, and the metabolic syndrome in law enforcement officers. J Occup Environ Med. 2013;55(1):99–103. pmid:23207742
                  • 82. McCanlies EC, Slaven JE, Smith LM, Andrew ME, Charles LE, Burchfiel CM, et al. Metabolic syndrome and sleep duration in police officers. Work. 2012;43(2):133–9 pmid:22927620
                  • 83. Hartley TA, Burchfiel CM, Fekedulegn D, Andrew ME, Violanti JM. Health disparities in police officers: comparisons to the U.S. general population. Int J Emerg Ment Health. 2011;13(4):211–20. pmid:22900455
                  • 84. Theorell T. A long-term perspective on cardiovascular job stress research. J Occup Health. 2019;61(1):3–9. pmid:30698340
                  • 85. Eddy P, Wertheim EH, Kingsley M, Wright BJ. Associations between the effort-reward imbalance model of workplace stress and indices of cardiovascular health: A systematic review and meta-analysis. Neurosci Biobehav Rev. 2017;83:252–266. pmid:29111334
                  • 86. Siegrist J, Li J. Work Stress and Altered Biomarkers: A Synthesis of Findings Based on the Effort-Reward Imbalance Model. Int J Environ Res Public Health. 2017;14(11). pii: E1373. pmid:29125555
                  • 87. Karasek R, Collins S, Clays E, Bortkiewicz A, Ferrario M. Description of a large-scale study design to assess work-stress-disease associations for cardiovascular disease. Int J Occup Med Environ Health. 2010;23(3):293–312. pmid:21306975
                  • 88. Kivimäki M, Jokela M, Nyberg ST, Singh-Manoux A, Fransson EI, Alfredsson L, et al. Long working hours and risk of coronary heart disease and stroke: a systematic review and meta-analysis of published and unpublished data for 603,838 individuals. Lancet. 2015;386(10005):1739–46. pmid:26298822
                  • 89. Kivimäki M, Kawachi I. Work Stress as a Risk Factor for Cardiovascular Disease. Curr Cardiol Rep. 2015;17(9):630. pmid:26238744
                  • 90. Fishta A, Backé EM. Psychosocial stress at work and cardiovascular diseases: an overview of systematic reviews. Int Arch Occup Environ Health. 2015;88(8):997–1014. pmid:25687981