Mental health conditions,individual and job characteristics and sleep disturbances among firefighters

Abstract

This study aimed to assess the associations between mental health conditions, individual and job characteristics and sleep disturbances among firefighters. Of 303 participants, 51.2% reported sleep disturbances. Psychological distress and psychosomatic disturbances were significantly associated with sleep disturbances. Suicidal ideation, unhealthy alcohol use and time as a firefighter were also associated with sleep disturbances but at a borderline level of significance (0.05 < p < .085). These findings may be related to the psychological and physical hazards of firefighting and indicate the importance of research on associated professions.

Keywords

firefighters individual differences mental health conditions occupational hazards occupational health sleep disturbances unhealthy alcohol use work-related variables

Introduction

Over 40% of police officers in the United States and Canada have sleep complaints; 34% screen positive for obstructive sleep apnea, 7% for moderate to severe insomnia and 5% for shift work disorder; 29% report excessive sleepiness and 26% report falling asleep while driving at least one time a month. These disorders have negative consequences in various social and individual spheres and interfere with quality of life, social relationships and workplace productivity; they can lead to injuries, fatigue, cognitive impairment and difficulty concentrating, accidents, and increases in medical and mental health care utilization (Rajaratnam et al., 2011).

Sleep is a complex neurophysiologic state involving many neurotransmitters, particularly serotonin (5-HT) and noradrenaline (NA), and electrophysiologic states and transitions (Graeff and Guimarães, 2005). From a clinical perspective, sleep disorders are divided into primary or secondary disorders. Primary sleep disorders may be classified as dyssomnias (alteration of the quality and quantity of sleep) or parasomnias, that are characterized by physiological altered sleep events (APA, 2000). Secondary disorders are often caused by mental disorders, medical conditions or drugs (including alcohol and medications).

Sleep disorders can have various etiologies and factors that influence their onset and maintenance. Individual characterisitcs like age, gender (Marquié and Foret, 1999; Neylan et al., 2002), domestic factors (Portela et al., 2005), pattern of alcohol use and smoking habits (Neumann et al., 2008; Noal et al., 2008) can influence sleep and sleep disorders.

Sleep can also be influenced by factors related to work such as high psychological demands, exposure to critical and traumatic incidents, work environment stressors (Neylan et al., 2002), mentally or physically stressful work and shiftwork (Rajaratnam et al., 2011).

The prevalence of sleep problems reported by workers is very high, ranging from 15% to 37%, and many factors are related to these problems. Psychosocial work stressors predict insomnia and sleep debt. These stressors include high work demands, stress regarding the impossibility of finishing them and high professional commitment (Kivistö et al., 2008).

Neylan et al. (2002), in a study aimed at comparing subjective sleep quality in police officers with that in control participants not involved in police or emergency services, found that police officers reported less average sleep time. The majority of the police officers reported disturbances in subjective sleep quality. Although the life-threatening aspects of police work are related to nightmares, the routine stressors of police service seemed to most affect global sleep quality in these participants. Global sleep disturbances, nightmares and dream-related awakenings were strongly associated with exposure to traumatic events, posttraumatic stress disorder, routine stressors of police service and psychiatric symptoms.

Witteveen et al. (2010), in a study with police officers and firefighters, found that the workers who had been exposed often to negative life events had more symptoms of post-traumatic stress disorder, depression and fatigue. Disaster-exposed subjects with more sleep disturbance symptoms had higher cortisol levels. Similar to these results, Sing and Wong (2011) found that insomnia had a mediating effect on the optimism associated with depression, suggesting that people who suffer with insomnia, have less optimism related to their depression, leading to a worse prognosis of depression.

Emergency service workers such as police officers and firefighters are constantly exposed to and conscious of imminent danger that pushes them to their psychological limits and may place them at increased risk for sleep disorders (Guidotti and Clough, 1992; Sheppard et al., 1986). They are exposed to physical, thermal, ergonomic, chemical and psychological risks (Guidotti and Clough, 1992). The unpredictability of the situations they deal with, direct contact with danger, the suffering of people they help, and risks to their own personal safety contribute to the risk for psychological consequences among firefighters. Emergency service work has high physical, psychological and psychosocial demands (including large and visible social responsibility); it involves shiftwork; and it exposes workers to stressful, risky and traumatic events.

Despite job stress and the high risk of sleep disturbances among emergency service workers, little is known about them. Therefore, the aim of the present study was to assess the independent associations between individual characteristics (sociodemographics), mental health conditions, and job characteristics, and sleep disturbances among firefighters. To do so, we surveyed firefighters from the Fourth Military Firefighters Battalion in Juiz de Fora, a medium-sized city in the south-eastern Brazilian state of Minas Gerais. This is one of the five ‘Operational Execution Units’ in the state, responsible for assistance to 2,200,000 inhabitants, in 146 towns.

Because of the lack of clear theoretical and empirical guidance, we did not develop formal hypotheses regarding the effects of variables of interest on firefighters’ sleep disturbances; we were interested in descriptive and exploratory analysis.

Method

The study was carried out at the Fourth Military Firefighters Battalion (4MFB) in Juiz de Fora – MG in Brazil, a city of approximately 500,000 inhabitants, as part of a study called Firefighters’ Mental Health Evaluation. Psychology students or the Battalion psychologist invited up to 12 firefighters at a time to a private room where the study was explained and participation in it was offered. The firefighters were told that their participation was voluntary and that the information they provided would be kept confidential. This study was approved by the Research Ethics Commitee of the Federal University of Juiz de Fora (Opinion Document number 014/2005, consistent with the directive in Resolution CNS number 196/96).

First, to assess individual characteristics, a structured written questionnaire with a unique identification number was distributed. The questionnaire asked socio-demographics including age, sex, marital status, educational level and race.

Second, to assess mental health conditions (including sleep disturbances), six psychological tests were used:

Lipp Stress Symptom Inventory for Adults (LSSI), that evaluates self-perceived stress (Lipp, 2000).

Beck Depression Inventory (BDI), that evaluates severity of depression (Cunha, 2001).

Beck Anxiety Inventory (BAI), that evaluates severity of anxiety (Cunha, 2001).

Beck Scale for Suicidal Ideation (SSI), that evaluates the presence or absence of suicidal ideation (Cunha, 2001).

Alcohol Use Disorders Identification Test (AUDIT), that assesses unhealthy alcohol use (Babor et al., 2003).

General Health Questionnaire (GHQ) and its subscales that evaluate the presence and severity of mental health conditions. The subscales are: psychological distress (HQPD), desire for death (HQDD), diffidence (HQD), psychosomatic alteration (HQPA) and sleep disturbances (HQSD) (Goldberg, 1996).

To assess work-related factors, a questionnaire asked about job characteristics, including military status (commissioned, non-commissioned); work shift (shiftwork or regular shift – administrative services during daytime); time as a firefighter; work section (operational or internal); income and work station (main Battalion or other suburban posts). Operational work for these firefighters means rescue, firefighting, work in catastrophes and similar work responding to emergencies. Internal work includes administrative services and answering phone calls. Answering phone calls includes the responsibility for planning the emergency response route and the strategies to be used to address each call. The non-commissioned firefighters are the corporals and the sergeants, and the commissioned ones are the first and second lieutenants, lieutenants commanders, captains, majors, a lieutenant colonel and a lieutenant general.

The assessments were validated for the Brazilian population (Babor et al., 2003; Cunha, 2001; Goldberg, 1996; Lipp, 2000). We also validated the assessments in a sample of firefighters from the same Battalion as the sample for the current study. Cronbach’s alphas were: BDI (α = 0.85); BAI (α = 0.91); SSI (α = 0.78); HQPD (α = 0.84); HQDD (α = 0.93); HQD (α = 0.88); HQPA (α = 0.84); HQSD (α = 0.84). The Beck Hopelessness Scale was not included because its Cronbach’s α coefficient was low.

Participants completed the questionnaires in a private room. They were administered in the morning, during two sessions, approximately one hour each. Questionnaire completion was preceded by explanations given by the psychology students or by the psychologist, who followed the test manual procedures for the selected instruments.

Questionnaire responses were entered into a database and analyses were performed using SPSS (versions 8.0 and 15.0).We dichotomized sociodemographic variables (gender, marital status, education (college or not), age (at the median), work-related factors (military commission or not, shiftwork or not, operational versus internal work, suburban station, and income) and time as a firefighter (at the median).

Mental health condition scores were dichotomized at published cutoffs, with the exception of the GHQ subscales, for which the T-score cutoff ≥ 60 was used. The GHQ manual recommends that the T-score-cutoff ≥ 90 should be considered psychiatric cases. However, as the objective of this study was to identify sleep disturbances and to study the effects of mental health condition symptoms (not only the presence of diagnosis), we used a T-score cutoff of 60 or greater.

First we performed descriptive analyses (frequencies, medians) for all variables. A chi-square goodness of fit test was used to test the association between mental health condition scores (dichotomized) and the presence or absence of sleep disturbances. Then we fit a conditional logistic regression analysis using the backwards stepwise procedure (p < .10 for entry, p < .05 for stay or retention in the model). Three separate regression models were fit to assess the associations between: (1) socio-demographic variables; (2) mental health conditions; and (3) job characteristics and sleep disturbances (GHQ sleep disturbances subscale, T-score cutoff ≥ 60). The final model included variables that were retained in the three separate models. From these models we report odds ratios and 95% confidence intervals.

Results

The Battalion was composed of 542 firefighters; of those, 303 (55.9%) agreed to take part in the study. The characteristics of the 303 participating firefighters are in Table 1.

Table 1.

Characteristics of the firefighters (N = 303)

Characteristics	Category	n	%
Sex	Male	277	91
	Female	26	9
Educational level	Completed high school	217	72
	More than high school	80	26
Race	White	160	53
	Non-white	140	46
Marital status	Married/living with a partner	210	69
	Single or other	93	31
Age (years)	Median	33	–
Work station	Main Battalion	257	85
	Suburban	46	15
Work shifts	Shiftwork	167	55
	Regular shifts	136	45
Military status	Non-commissioned	281	93
	Commissioned officer	22	7
Salary (US$ per month)	Median	651.16	–
Work section	Operational activities	141	47
	Administrative and phone activities	151	50
Time as firefighter (months)	Median	126	–

Note: Missing for age, n = 11; race, n = 3; educational level, n = 6; salary, n = 7; work section, n = 11; time as firefighter, n = 1.

The descriptive analysis of the mental health conditions appears in Table 2.

Table 2.

Mental health conditions in 303 firefighters

Psychological tests	Negative		Positive		χ² (d.f.=1)	p-value
	n	%	n	%
Stress (LSSI)	215	71	88	29	53.231	< .001**
Depression (BDI)	272	90	30	10	193.921	< .001**
Anxiety (BAI)	276	91	27	9	204.624	< .001**
Suicidal ideation (SSI)	255	85	46	15	145.120	< .001**
Unhealthy alcohol use (AUDIT)	208	69	94	31	43.033	< .001**
Psychological distress (HQPD)*	171	56	132	44	5.020	.025**
Desire for death (HQDD)*	231	76	72	24	83.436	< .001**
Diffidence (HQD)*	179	59	124	41	9.983	< .002**
Psychosomatic alteration (HQPA)*	171	56	132	44	5.020	.025**
Sleep disturbance (HQSD)*	148	49	155	51	0.162	.688

Note: LSSI – Lipp Stress Symptom Inventory for Adults; BDI – Beck Depression Inventory; BAI – Beck Anxiety Inventory; SSI – Scale for Suicidal Ideation; AUDIT – Alcohol Use Disorders Identification Test; GHQ – General Health Questionnaire and the following subscales: HQPD – psychological distress; HQDD – desire for death; HQD – diffidence; HQPA – psychosomatic alteration; HQSD – sleep disturbance.

T-score-cutoff ≥ 60; **p < ,05.

Logistic regression analysis

In the logistic regression analysis the following variables were not retained as they were not statistically significant (p was > .05): all of the socio-demographic variables, all of the work-related variables (except time as a firefighter), stress (LSSI), desire for death (HQDD), diffidence (HQD), depression (BDI) and anxiety (BAI). The significant variables (p < .05) from the three separate initial models were entered into the final model. In the final model, psychological distress (HQPD) and psychosomatic alteration (HQPA) were significantly associated with sleep disturbances at the p < .05 level and suicidal ideation (SSI), unhealthy alcohol use (AUDIT) and time as a firefighter were of borderline statistical significance (p > .05 but < .085) (see Table 3). All variables were associated with a greater odds of sleep disturbances except longer time working as a firefighter, which was protective for the presence of sleep disturbances. Overall, the model correctly classified those with and without sleep alterations 71% of the time.

Table 3.

Independent variables associated with sleep disturbances in the multivariable logistic regression analysis

Variables	Odds ratio (95% CI)
Suicidal ideation (SSI)	2.50 (0.91–6.83)
Unhealthy alcohol use (AUDIT)>7	1.77 (0.96–3.25)
Psychological distress (HQPD)*	2.21 (1.14–4.28)
Psychosomatic alteration (HQPA)*	3.42 (1.81–6.47)
Time as a firefighter	0.61 (0.35–1.06)

Note: SSI – Scale for Suicidal Ideation; AUDIT – Alcohol Use Disorders Identification Test; GHQ – General Health Questionnaire and the following subscales: HQPD – psychological distress; HQPA – psychosomatic alteration.

T-score-cutoff ≥ 60.

Discussion

The prevalence of sleep disturbances in these firefighters was high (51%). This prevalence is higher than that reported for those in other occupations and in the general population (15%–37%) (Jansson-Fröjmark et al., 2007; Jansson and Linton, 2006; Kivistö et al., 2008). This high prevalence could be related to the physical and psychologically stressful nature of firefighters’ jobs. The high prevalence also suggests it is important to study firefighter health, in particular, sleep disturbances (Guidotti and Clough, 1992; Parkes, 1999).

The presence of psychological distress and psychosomatic alteration was significantly related to the presence of sleep disturbances among the firefighters in this study. Parkes (1999) considered sleep disturbances as a type of psychosomatic alteration that could be influenced by psychological factors. In Parkes’ study there was an association between individual personality characteristics and sleep, and shiftwork was significantly related to sleep disturbances, gastric problems, psychological distress (assessed by the GHQ) and injuries. Sleep disturbances and psychosomatic problems were also significantly related to the physical work environment. The results of Parkes’ study were similar to those in the present study except for the associations with shiftwork, that we did not detect. This discrepancy may be due to the lack of adjustment for potential confounders not measured in the current study. Alternatively, shiftwork may not have been associated with sleep disturbances in the current study because military firefighters who participated rotate across work station and work section, which decreases their exposure to each type of work and shiftwork.

Psychological distress may affect sleep in a number of ways. According to Vgontzas and Kales (1999) some people do not have sufficient coping mechanisms to confront stressful life events and deal with stress, internalizing conflicts and emotions, generating a combination of emotional arousal and physiological activation. This mental and physiological hyperarousal leads to difficulties with sleep. Then fear of sleeplessness intensifies the emotional arousal. Behavioral conditioning, with these other factors, leads to a vicious circle that perpetuates insomnia. These same authors suggest that when parasomnias happen in late childhood, adolescence, or resurface in adulthood, they are often associated with psychological disturbances and tend to worsen under stress.

The presence of psychological and emotional stress predicts lower efficiency of sleep, a higher percent of wake time during the night, longer latency to initiate sleep (Åkerstedt et al., 2007) and lower subjective (self-reported) and objective (measured through physiological exams) duration of sleep. Stress can also intensify the negative effect that other factors like obesity and sleep complaints have on the subjective duration of sleep (Caldwell and Redeker, 2009; Vgontzas et al., 2008).

The findings of the present study are consistent with previous studies and highlight the importance of the assessing for psychological distress among those at risk for sleep disturbances. This implication may be of particular importance among military firefighters who deal with a great number of psychological stressors in their work, in addition to the stress of everyday life. Accordingly to Guidotti and Clough (1992), besides being concerned with the safety of other people, firefighters need to worry about their own security and they are often witnesses to pain, injury and strong emotion. In addition, firefighters are routinely exposed to the sound of an alarm, that is followed by apprehension and anxiety because of the unpredictability of the situation they are about to encounter. Some firefighters are stressed to their limits during the execution of their job duties. The fact that this increased stress does not result in increased mortality, may reflect a strong ‘healthy worker effect’ and a decrease in effort on the job that comes with more time as a firefighter and with promotion.

The presence of suicidal ideation and unhealthy alcohol use was associated with sleep disturbance at a borderline level of significance, suggesting they may increase the risk. Problems with duration and quality of sleep are related to unhealthy alcohol use (Kinney and Coyle, 2006; Neumann et al., 2008) and alcohol use can lead to a perception of sleep as less satisfying, less relaxing and less refreshing. Drinking can also lead to fatigue on the day after alcohol use (Kinney and Coyle, 2006). Even lower risk (‘moderate’) alcohol use can lead to an increase in the frequency and the severity of obstructive sleep apnea (Noal et al., 2008; Scanlan et al., 2000).

Suicidal ideation and suicide attempts among adults have been associated with short duration of sleep (Goodwin and Marusic, 2008), and sleep disturbances are associated with suicide completion among adolescents (Goldstein et al., 2008). Suicidal ideation can remit or at least be reduced when the quality of sleep is improved with treatment (Krahn et al., 2008). Goodwin and Marusic (2008) also found that alcohol dependence increases suicidal ideation among individuals with a short sleep duration. Those observations reinforce the results of the present study and highlight the necessity of greater attention to sleep problems.

In summary, psychological distress, psychosomatic alteration, suicidal ideation and unhealthy alcohol use were risk factors for sleep disturbances among firefighters. Time on the job was a protective factor, consistent with prior studies, possibly indicative of a ‘healthy worker effect’. This effect suggests that while the worker gets adapted to job conditions, he or she creates coping mechanisms to confront the adverse conditions and avoid the most harmful exposures, reducing the negative consequences of the job (Guidotti and Clough, 1992; Marquié and Foret, 1999). In addition, the workers that do not adapt well to the job, generally abandon their jobs: ‘Thus, self-selection produces an older “survivor” group, better able to cope with environmental and work-related demands than those who quit’ (Parkes, 1999: 266).

From a practical perspective, the current findings suggest a way to reduce sleep disturbances among firefighters, by reducing stressful job factors and addressing other mental health conditions. The findings also suggest that reduced alcohol use and improvements in coping strategies (to cope with their job and mental health conditions) might reduce sleep disturbances. By educating firefighters regarding the beneficial effects of healthy habits like sleep hygiene and a schedule of activities including both personal and job actitivities, sleep disturbances might also be reduced (Yang et al., 2010). Such reductions could decrease health care costs related to sleep disturbances and their negative consequences.

This study has some limitations. First, the results may not be generalizable to firefighters beyond this city in Brazil. Second, the use of only one scale from a general mental health test to identify sleep disturbances could be considered a limitation. However, we have little reason to suspect that firefighters who participated in the study would be very different from those elsewehere, at least in Brazil, and we used scales validated in the Brazilian population (Goldberg, 1996). In addition, the tests were self-report and therefore the evaluation of the sleep disturbances was intentionally subjective. Subjective assessment can detect what laboratory testing may not: Dru et al. (2007) found that some people with normal sleep patterns according to electroencephalography and polysomnography exams still report having sleep disturbances. In addition, negative subjective perception of sleep may have influence in diverse life spheres, including depression, irritability and cognitive function (Dru et al., 2007). Creti et al. (2010) compared the objective and subjective ways to measure sleep complaints and found that objectively measured nocturnal sleep time effectively approximated subjective experience but that subjective self-report measures of daytime sleepiness, fatigue and non-refreshing sleep were not reflected by corresponding objective sleep-related measures (polysomnography and actigraphy).

Future research using physiological exams like electroencephalography and polysomnography are needed to better understand sleep disturbances among firefighters. Furthermore, additional longitudinal research is needed to better understand firefighter health given the high health risks and the great social importance of this occupation.

Footnotes

Acknowledgements

We express our appreciation to the National Council of Technological and Scientific Development (CNPq), to the Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) and the Research Pro-Rectory of the Federal University of Juiz de Fora (Propesq-UFJF) for financial support, to the Fourth Military Firefighters Battalion in Juiz de Fora for infrastructure and to Andréa Batista and Paula Almeida de Oliveira for staff support.

Competing Interests

None declared.

References

Åkerstedt

Kecklund

Axelsson

(2007) Impaired sleep after bedtime stress and worries. Biolological Psychology 76(3): 170–173.

American Psychiatric Association (APA) (2000) Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: APA.

Babor

Higgins-Biddle

Saunders

Monteiro

(2003) AUDIT: Teste para identificação de problemas de álcool – roteiro para uso em atenção primária. PAI-PAD: Ribeirão Preto, São Paulo.

Caldwell

Redeker

(2009) Sleep patterns and psychological distress in women living in an inner city. Research in Nursing & Health 32(2): 177–190.

Creti

Libman

Baltzan

Rizzo

Bailes

Fichten

(2010) Impaired sleep in chronic fatigue syndrome: How is it best measured? Journal of Health Psychology 15(4): 596–607.

Cunha

(ed.) (2001) Manual de versão em português das escalas Beck. São Paulo: Casa do Psicólogo.

Dru

Bruge

Benoit

. (2007) Overnight duty impairs behaviour, awake activity and sleep in medical doctors. European Journal of Emergency Medicine 14(4): 199–203.

Goldberg

(1996) Questionário de saúde geral de Goldberg: Manual técnico QSG: Adaptação brasileira. São Paulo: Casa do Psicólogo.

Goldstein

Bridge

Brent

(2008) Sleep disturbance preceding completed suicide in adolescents. Journal of Consulting and Clinical Psychology 76(1): 84–91.

10.

Goodwin

Marusic

(2008) Association between short sleep and suicidal ideation and suicide attempt among adults in the general population. Sleep 31(8): 1097–1101.

11.

Graeff

Guimarães

(2005) Fundamentos de Psicofarmacologia. São Paulo: Atheneu.

12.

Guidotti

Clough

(1992) Occupational health concerns of firefighting. Annual Review of Public Health 13: 151–171.

13.

Jansson

Linton

(2006) Psychosocial work stressors in the development and maintenance of insomnia: A prospective study. Journal of Occupational Health Psychology 11(3): 241–248.

14.

Jansson-Fröjmark

Lundqvist

Linton

(2007) Psychosocial work stressors for insomnia: A prospective study on 50–60-year-old adults in the working population. International Journal of Behavioral Medicine 14(4): 222–228.

15.

Kinney

Coyle

(2006) Alcohol hangover effects on measures of affect the morning after a normal night’s drinking. Alcohol & Alcoholism 41(1): 54–60.

16.

Kivistö

Härmä

Sallinen

Kalimo

(2008) Work-related factors, sleep debt and insomnia in IT professionals. Occupational Medicine 58: 138–140.

17.

Krahn

Miller

Bergstrom

(2008) Rapid resolution of intense suicidal ideation after treatment of severe obstructive sleep apnea. Journal of Clinical Sleep Medicine 4(1): 64–65.

18.

Lipp

(2000) Manual do inventário de sintomas de estresse para adultos de Lipp. São Paulo: Casa do Psicólogo.

19.

Marquié

Foret

(1999) Sleep, age, and shiftwork experience. Journal of Sleep Research 8: 297–304.

20.

Neumann

Neuner

Weiß-Gerlach

Spies

(2008) Complaints about sleep in trauma patients in an emergency depatment in respect to alcohol use. Alcohol & Alcoholism 43(3): 305–313.

21.

Neylan

Metzler

Best

. (2002) Critical incident exposure and sleep quality in police officers. Psychosomatic Medicine 64: 345–352.

22.

Noal

Menezes

AMB

Canani

Siqueira

(2008) Ronco Habitual e Apnéia Obstrutiva Observada em Adultos: Estudo de Base Populacional, Pelotas, RS. Revista de Saúde Pública 42(2): 224–233.

23.

Parkes

(1999) Shiftwork, job type, and the work environment as joint predictors of health-related outcomes. Journal of Occupational Health Psychology 4: 256–268.

24.

Portela

Rotenberg

Waissmann

(2005) Health, sleep and lack of time: Relations to domestic and paid work in nurses. Revista de Saúde Pública 39(5): 802–808.

25.

Rajaratnam

SMW

Barger

Lockley

. (2011) Sleep disorders, health, and safety in police officers. JAMA 306(23): 2567–2578.

26.

Scanlan

Roebuck

Little

Redman

Naughton

(2000) Effect of moderate alcohol upon obstructive sleep apnoea. European Respiratory Journal 16: 909–913.

27.

Sheppard

Distefano

Morse

Becker

(1986) Acute effects of routine firefighting on lung function. American Journal of Industrial Medicine 9: 333–340

28.

Sing

Wong

(2011) The effect of optimism on depression: The mediating and moderating role of insomnia. Journal of Health Psychology 16(8): 1251–1258.

29.

Vgontzas

Kales

(1999) Sleep and its disorders. Annual Review of Medicine 50: 387–400.

30.

Vgontzas

Lin

Papaliaga

. (2008) Short sleep duration and obesity: The role of emotional stress and sleep disturbances. International Journal of Obesity 32(5): 801–809.

31.

Witteveen

Huizink

Slottje

Bramsem

Smid

Ploeg

HMVD

(2010) Associations of cortisol with posttraumatic stress symptoms and negative life events: A study of police officers and firefighters. Psychoneuroendocrinology 35: 1113–1118.

32.

Yang

Lin

Hsu

Cheng

(2010) Maladaptive sleep hygiene practices in good sleepers and patients with insomnia. Journal of Health Psychology 15(1): 147–155.