Abstract
Study Objective:
To examine sleep disturbances as a predictor of cause-specific work disability and delayed return to work.
Design:
Prospective observational cohort study linking survey data on sleep disturbances with records of work disability (≥ 90 days sickness absence, disability pension, or death) obtained from national registers.
Setting:
Public sector employees in Finland.
Participants:
56,732 participants (mean age 44.4 years, 80% female), who were at work and free of work disability at the study inception.
Measurements and Results:
During a mean follow-up of 3.3 years, incident diagnosis-specific work disability was observed in 4,028 (7%) employees. Of those, 2,347 (60%) returned to work. Sleep disturbances 5-7 nights per week predicted work disability due to mental disorders (hazard ratio [HR] 1.6, 95% confidence interval [CI] 1.3-1.9) and diseases of the circulatory system (HR = 1.6, 95% CI 1.2-2.1), musculoskeletal system (HR = 1.6, 95% CI 1.4-1.8) and nervous system (HR = 1.5, 95% CI 1.0-2.2), and injuries and poisonings (HR = 1.6, 95% CI 1.2-2.1) after controlling for baseline age, sex, socioeconomic status, night/shift work, health behaviors (e.g., smoking, exercise), diagnosed somatic diseases, use of pain killers, depression, and anxiety. In addition, sleep disturbances prior to disability were associated with higher likelihood of not returning to work after work disability from musculoskeletal diseases (HR = 1.2, 95% CI 1.1-1.7) and, in men, after work disability due to mental disorders (HR = 4.4, 95% CI 1.7-11.1).
Conclusions:
Sleep disturbances are associated with increased risk for subsequent disabling mental disorders and various physical illnesses. They also predict the outcome of work disability due to musculoskeletal disorders.
Citation:
Salo P; Oksanen T; Sivertsen B; Hall M; Pentti J; Virtanen M; Vahtera J; Kivimäki M. Sleep disturbances as a predictor of cause-specific work disability and delayed return to work. SLEEP 2010;33(10):1323-1331.
Keywords: Cohort studies, retirement, sick leave, sleep initiation and maintenance disorders
SLEEP DISTURBANCES ARE COMMON IN THE WORKING POPULATION: ABOUT 30% TO 50% EXPERIENCE INSOMNIA SYMPTOMS OCCASIONALLY,1,2 AND UP TO 10% meet the criteria for clinical insomnia diagnosis.1 Direct and indirect costs of sleep disturbances are substantial as people with sleep disturbances consult health care professionals more often,3 report taking more prescription and over-the-counter medicine,4 and experience lower work productivity than good sleepers.4,5
Recent studies suggest that sleep disturbances may predict subsequent occupational disability, as indicated by sickness absence and temporary or permanent disability pension.6–14 For instance, the risk of sickness absence was found to be 1.5 times as high in individuals with sleep disturbances as in good sleepers after adjusting for important demographic and health-related confounders.6,9,12,14 Some studies reported a nearly 2-fold increase in disability pension risk among those experiencing sleep disturbances.7,13 Less understood is whether specific diseases or disorders contribute to the increased risk of work disability in individuals with disturbed sleep.6,7 Also unclear is whether sleep disturbances affect work disability prognosis, that is, whether the individual returns to work or remains permanently disabled.15,16
In this prospective observational cohort study, we examined whether sleep disturbances increased the risk of diagnosis-specific work disability and delayed return to work in a large population of public sector employees.
METHODS
Study Population and Design
This study is part of the Finnish Public Sector Study of local government employees in 10 towns and 21 public hospitals in Finland. We included employees who responded to a survey either in 2000-2002 (N = 48,598) or in 2004 (N = 48,076), for a total sample of 66,418 employees (74% of the eligible employees responded at least once, the first response was considered). This sample did not differ substantially from the eligible population (N = 90,211) in terms of mean age (43.5 years in the sample, 43.1 years in the eligible population) or the proportion of women (80.0% vs. 76.4%), manual workers (18.5% vs. 20.0%), and cases with incident work disability (7.7% vs. 8.0%).
We excluded from the analysis participants who were retired or on a long sick leave (≥ 90 days) during the preceding 12 months or at the beginning of follow-up or had incomplete baseline data (n = 9,686, 15%). Thus, the analytic sample included 56,732 employees (11,507 men; 45,225 women) who were working at baseline. We linked their responses to data on sickness absence, disability pension, and death obtained from the national registers using personal identification numbers. The register data were available up to 31 December 2005 and the linkage was successful for all participants. The study was approved by the ethics committee of the Finnish Institute of Occupational Health.
Sleep Disturbances
Sleep disturbances were quantified using the Jenkins Sleep Problem Scale.17 Individual items correspond with insomnia symptoms specified by the Diagnostic and Statistic Manual of Mental Disorders, Fourth Edition (DSM-IV), i.e., difficulty falling asleep or maintaining sleep during the night, too early morning awakenings, and non-restorative sleep with a duration ≥ 1 month. Participants reported how frequently they had experienced each of the 4 symptoms during the past 4 weeks (response scale from 1 = never to 6 = every night). For participants who reported more than one symptom, their most frequent symptom was used to assess the degree of sleep disturbance experienced. Sleep disturbances were categorized as one of the following: ≤ 1 night/week (no sleep disturbance), 2-4 nights/week (moderate sleep disturbance), or 5-7 nights/week (severe sleep disturbance).
Work Disability Outcomes
Work disability outcomes were quantified using data from national registers which cover reliably all reimbursed sickness absences and retirements. In addition, to reduce survivor bias, deaths were included in the work disability outcomes, as they represent the most severe form of illness. The work disability registers require physician diagnoses of disability and include the beginning and end dates of all work-related disability, including long-term sickness absence (≥ 90 days; Social Insurance Institution of Finland) and temporary and permanent disability pension (Finnish Centre for Pensions). Diagnosis-specific data on mortality were obtained from the Statistics Finland. Incident work disability was defined as the first long-term sickness absence, granted disability pension, or death during the follow-up. Diagnoses for disabilities were coded according to the International Classification of Diseases, 10th Revision (ICD-10). The most common disease categories were malignant tumors (ICD-10 codes C00-C99), mental disorders (F00-F99, X60-X84), diseases of the nervous system (G00-G99), circulatory system (I00-I99) and musculoskeletal system (M00-M99), and external causes, i.e., injuries and poisonings (S00-T98). A more detailed list of diagnoses is provided in Appendix 1, which is available online only at www.journalsleep.org.
Baseline Confounders
Sociodemographic characteristics including age, sex, and socioeconomic status were derived from the employers' records. Socioeconomic status was categorized according to the occupational-title classification of the Statistics Finland to upper-grade non-manual workers (e.g., teachers, physicians), lower-grade non-manual workers (e.g., registered nurses, technicians), and manual workers (e.g., cleaners, maintenance workers).18 Age and socioeconomic status were treated as confounders and sex was considered an effect modifier. Survey data regarding night or shift work was dichotomized: 0 = day work, 1 = night or shift work.
Health behaviors were assessed by questionnaire at baseline and included current smoking (no/yes), alcohol consumption, obesity, and physical activity. Participants reported their average weekly consumption of beer, wine, and spirits in portions. Those data were transformed into grams of pure alcohol. Body mass index was based on self-reported height and weight (BMI; kg/m2). Participants rated their physical activity according to whether it was equal to walking, fast walking, jogging, or running. Physical activity was measured as metabolic equivalent task (MET) hours per day.19
Baseline health.
Presence of any chronic conditions (no/yes) was identified based on the national health records. Data on medication for prevalent hypertension, cardiac failure, ischemic heart disease, diabetes, asthma or other chronic obstructive lung disease, rheumatoid arthritis, and severe mental disorders (i.e., psychosis) were obtained from the Drug Reimbursement Register kept by the Social Insurance Institution of Finland. Information about malignant tumors diagnosed during the preceding five years was obtained from the Finnish Cancer Register covering all diagnosed cancer cases in Finland.
Use of pain killers at baseline.
Drug Prescription Register data regarding purchases of analgesics (N02) or nonsteroidal anti-inflammatory and antirheumatic products (M01A) according to the World Health Organization's Anatomical Therapeutic Chemical (ATC) classification code20 at baseline were used to categorize participants according to use of pain killers (no/yes). The cut-off for pain killer use was 30 defined daily doses (DDD), corresponding one month's use.
Baseline mental health.
Participants were coded as positive for depression based on survey responses regarding physician diagnosed lifetime depression and/or purchases of antidepressants (ATC code N06A; > 30 DDDs) at the baseline year.20 The 6-item Trait Anxiety Inventory21 was used to quantify subjective symptoms of anxiety (continuous outcome); purchases of anxiolytics (ATC code N05B) obtained from the Drug Prescription Register were considered to be another indicator of anxiety.
Statistical Analysis
All analyses were performed using the SAS statistical software, version 9.1.3. The associations between baseline covariates and sleep disturbances (no, moderate, or severe) were analyzed using the χ2 test (categorical covariates) or analysis of variance (continuous covariates: age, alcohol consumption, BMI, physical activity, and anxiety). Outcomes were defined as the first cause-specific work disability period and return to work after such a period. Follow-up for the risk of cause-specific work disability began 1 January after the year of survey response and ended at the first of the following: long-term sickness absence, disability pension, old-age pension, death, or end of study (31 December 2005), whichever came first. Among work disability cases, the length of the work disability period was determined from the beginning of the disability to the date when an employee returned to work, was granted an old-age pension, died, or end of study (31 December 2005), of which the last 3 were categorized as not returning to work.
Cox proportional hazard models were used to study the associations between baseline sleep disturbances and work disability outcomes. Analyses were conducted separately for all-cause and cause-specific work disability. Hazard ratios (HR) and 95% confidence intervals (CI) were adjusted for age, sex, socioeconomic status, night/shift work, health behaviors, baseline health, use of pain killers, depression, and anxiety. As the associations of alcohol consumption22 and BMI23 with health may be U-shaped, these covariates were included in the fully adjusted model both as linear and squared terms. To examine sex differences, the interaction term “sex × sleep disturbance” was included in models containing main effects. Linear trends were calculated for the fully adjusted model.
Sensitivity Analyses
To examine the possibility of reverse causation, analyses were repeated after excluding cases of work disability (n = 1976) that occurred within the first 2 years of follow-up, leading to a sample size of 54,756. To further test whether the effect of sleep disturbances on work disability is independent of baseline health, this sensitivity analysis was repeated in a subcohort after excluding all participants meeting any of the following comorbidities (n = 42,595): a chronic condition or depression at the baseline (see Baseline covariates), or diagnosed sleep apnea during 10 years before beginning of the follow-up. Sleep apnea cases (codes 3472A in ICD-9 and G47.3 in ICD-10) were identified from the national Hospital Discharge Register kept by the National Institute for Health and Welfare. The register contains data on all inpatient hospital admissions in Finland. As the primary diagnosis of sleep apnea at discharge requires a positive finding in sleep registration and typical symptoms of sleep apnea, those 298 (0.5%) cases identified represented severe forms of disease.
To examine the possibility of subjectivity bias in a further sensitivity analysis, the main analysis was repeated using purchases of prescribed hypnotics or sedatives (ATC code N05C) instead of subjective sleep disturbances. Data on sleeping pill purchases were obtained from the Drug Prescription Register of the Social Insurance Institution of Finland. There were 1,161 (2%) participants who had at least one purchase at the baseline year and were coded as using sleep medication.
RESULTS
The mean follow-up was 3.3 years (range 0.02-5.0, SD 1.7) for the risk of work disability and 2.0 years (range 0.01-5.0, SD 1.4) for return to work from a period of work disability. Altogether, 4,028 participants (7%) became disabled or died during the follow-up. Of those disabled (n = 3,918), 2,347 (60%) returned to work with mean duration of work disability of 166 days (range 90-1,457; SD 98). Moderate sleep problems were reported by 26% (n = 14,654), and severe sleep disturbances by 22% (n = 12,499) of the participants. Severe sleep disturbances were more common in manual workers, those with existing medical conditions at baseline, and those using prescribed pain killers (Table 1). Both moderate and severe sleep disturbances were more common in women and in those using anxiolytics, or with a history of depression. Sleep disturbances were also associated with higher alcohol consumption, greater BMI, lower physical activity, and higher anxiety.
Table 1.
Baseline characteristics
| All | Sleep disturbances |
P-value | |||
|---|---|---|---|---|---|
| No | Moderate | Severe | |||
| N | N (%) | N (%) | N (%) | ||
| Age (Mean [SD]) | 44.4 [9.7] | 43.7 [9.6] | 44.0 [9.7] | 46.3 [9.6] | < 0.001 |
| Sex | < 0.001 | ||||
| Men | 11,507 | 6,374 (55) | 2,744 (24) | 2,389 (21) | |
| Women | 45,225 | 23,205 (51) | 11,910 (26) | 10,110 (23) | |
| Socioeconomic status | < 0.001 | ||||
| Upper-grade non-manual | 16,928 | 8,810 (52) | 4,427 (26) | 3,691 (22) | |
| Lower-grade non-manual | 29,802 | 15,625 (52) | 7,783 (26) | 6,394 (22) | |
| Manual | 10,002 | 5,144 (51) | 2,444 (25) | 2,414 (24) | |
| Night/shift work | < 0.001 | ||||
| No | 37,517 | 19,562 (52) | 9,530 (25) | 8,425 (23) | |
| Yes | 19,215 | 10,017 (52) | 5,124 (27) | 4,074 (21) | |
| Smoking | 0.19 | ||||
| No | 46,261 | 24,186 (52) | 11,949 (26) | 10,126 (22) | |
| Yes | 10,471 | 5,393 (51) | 2,705 (26) | 2,373 (23) | |
| Alcohol intake (Mean [SD]) | 65.3 [100.7] | 59.7 [89.3] | 70.0 [104.5] | 73.0 [119.2] | < 0.001 |
| Body mass index (Mean [SD]) | 25.0 [4.12] | 24.9 [4.0] | 25.0 [4.1] | 25.5 [4.4] | < 0.001 |
| Physical activity (Mean [SD]) | 4.8 [4.3] | 5.1 [4.4] | 4.7 [4.1] | 4.3 [4.0] | < 0.001 |
| Somatic diagnoses | < 0.001 | ||||
| No | 49,793 | 26,451 (53) | 12,837 (26) | 10,505 (21) | |
| Yes | 6,939 | 3,128 (45) | 1,817 (26) | 1,994 (29) | |
| Use of pain killers | < 0.001 | ||||
| No | 51,134 | 27,218 (53) | 13,148 (26) | 10,768 (21) | |
| Yes | 5,598 | 2,361 (42) | 1,506 (27) | 1,731 (31) | |
| Depression | < 0.001 | ||||
| No | 49,607 | 27,467 (56) | 12,586 (25) | 9,554 (19) | |
| Yes | 7,125 | 2,112 (30) | 2,068 (29) | 2,945 (41) | |
| Anxiety (Mean [SD]) | 1.9 [0.6] | 1.7 [0.4] | 2.1 [0.5] | 2.2 [0.6] | < 0.001 |
| Use of anxiolytics | < 0.001 | ||||
| No | 55,816 | 29,281 (52) | 14,397 (26) | 12,138 (22) | |
| Yes | 916 | 298 (33) | 257 (28) | 361 (39) | |
| Use of sleep medication | < 0.001 | ||||
| No | 55,571 | 29,341 (53) | 14,352 (26) | 11,878 (21) | |
| Yes | 1,161 | 238 (21) | 302 (26) | 621 (53) | |
SD, Standard deviation
Risk of Work Disability
Figure 1 shows that sleep disturbances were associated with the time to first work disability period across the entire follow-up time. As shown in Table 2, sleep disturbance was a significant predictor of all-cause work disability after adjusting for age, sex, and socioeconomic status. Although further adjustment for all baseline confounders attenuated relationships among sleep disturbances and work disability by approximately 50%, these relationships remained significant: severe sleep disturbances were associated with a risk 1.54 times as great as good sleepers (95% CI 1.42-1.67) and moderate sleep disturbances with risk 1.19 times as great as good sleepers (95% CI 1.10-1.30) of work disability (linear trend P < 0.001). No sex interaction was observed.
Figure 1.
The associations of sleep disturbances with work disability and return to work. Crude cumulative hazard rates for risk of work disability (Panel A) and probability of returning to work (Panel B) for participants with no, moderate, or severe sleep disturbance, as estimated by Kaplan-Meier survival analysis.
Table 2.
The association of sleep disturbances and the risk of work disability
| Sleep disturbances | Risk of work disability |
|||||||
|---|---|---|---|---|---|---|---|---|
| Any disease (n = 4028) HR (95% CI) | Malignancy (n = 398) HR (95% CI) | Mental disorders (n = 841) HR (95% CI) | Diseases of the nervous system (n = 174) HR (95% CI) | Diseases of the circulatory system (n = 252) HR (95% CI) | Diseases of the musculo-skeletal system (n = 1522) HR (95% CI) | Injuries and poisonings* |
||
| Men (n = 79) HR (95% CI) | Women (n = 284) HR (95% CI) | |||||||
| Model 1 | ||||||||
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) |
| Moderate | 1.36 (1.26-1.47) | 1.14 (0.90-1.45) | 1.90 (1.59-2.27) | 1.44 (1.00-2.09) | 1.12 (0.81-1.54) | 1.26 (1.10-1.43) | 0.92 (0.53-1.60) | 1.41 (1.05-1.89) |
| Severe | 2.02 (1.88-2.18) | 1.29 (1.02-1.64) | 3.35 (2.85-3.95) | 1.89 (1.33-2.69) | 1.71 (1.29-2.27) | 1.89 (1.68-2.13) | 0.95 (0.54-1.67) | 2.08 (1.58-2.74) |
| Model 2 | ||||||||
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) |
| Moderate | 1.19 (1.10-1.30) | 1.11 (0.86-1.42) | 1.32 (1.10-1.59) | 1.32 (0.90-1.93) | 1.06 (0.76-1.48) | 1.16 (1.02-1.33) | 0.85 (0.48-1.51) | 1.35 (0.99-1.82) |
| Severe | 1.54 (1.42-1.67) | 1.23 (0.95-1.59) | 1.61 (1.34-1.94) | 1.51 (1.02-2.22) | 1.57 (1.15-2.14) | 1.60 (1.41-1.81) | 0.82 (0.44-1.53) | 1.90 (1.41-2.57) |
| P for linear trend | < 0.001 | 0.29 | < 0.001 | 0.10 | 0.01 | < 0.001 | 0.77 | < 0.001 |
P-values for linear trends are calculated for model 2. Hazard ratios and their 95% confidence limits were derived from Cox proportional hazard models.
Significant sex interactions: injuries and poisonings P = 0.04, other diagnosis groups 0.13-0.88.
Model 1 adjusted for age, sex, and socioeconomic status. Model 2 adjusted for age, sex, socioeconomic status, night/shift work, smoking, alcohol intake, body mass index, physical activity, diagnosed somatic disease, use of pain killers, depression, anxiety, and use of anxiolytics.
Cause-specific analyses show that sleep disturbances predicted increased risk of incident work disability in almost all disease categories evaluated (Table 2). Participants with severe sleep disturbances had a hazard of work disability due to mental disorders 3.35 times as high as good sleepers. They also had a hazard 1.71 to 1.89 times as high as good sleepers for work disability due to diseases of the nervous system, circulatory system, and musculoskeletal system, and injuries and poisonings. After full adjustments, hazard ratios were still 1.51 to 1.61 times higher in each of these disease categories. For injuries and poisonings, a statistically significant sex interaction was observed (P = 0.04): after full adjustments, severe sleep disturbances were associated with a 1.90 fold (95% CI 1.41-2.57) increased risk of work disability in women; but no statistically significant association was observed in men. A linear trend was found for work disability due to mental disorders (both sexes P < 0.001), diseases of the circulatory system (P = 0.01) or musculoskeletal system (P < 0.001), and injuries and poisonings (women P < 0.01). Sleep disturbances were not associated with work disability due to malignancies.
Return to Work
Figure 1 shows that the likelihood of returning to work from any disability period was lower among individuals with sleep disturbances compared to those with no sleep disturbances across the entire follow-up period. However, those relationships varied by sex (P = 0.03) (Table 3). Compared to men with no sleep disturbances, the probability of not returning to work was higher among men with moderate (HR 1.39, 95% CI 1.05-1.82) and severe baseline sleep disturbances (HR 1.67, 95% CI 1.25-2.22) in the fully adjusted model (linear trend P < 0.01). In women, sleep disturbances were associated with a higher likelihood of not returning to work in the partially adjusted model only.
Table 3.
The association of sleep disturbances and probability of not returning to work after work disability (number of returnees of all absentees)
| Sleep disturbances | Probability of not returning to work |
||||||||
|---|---|---|---|---|---|---|---|---|---|
| Any disease* |
Malignancy |
Mental disorders* |
Diseases of the nervous system |
Diseases of the circulatory system |
Diseases of the musculo-skeletal system |
Injuries and poisonings |
|||
| Men (n = 372 of 703) HR (95% CI) | Women (n = 1975 of 3215) HR (95% CI) | (n = 220 of 379) HR (95% CI) | Men (n = 57 of 112) HR (95% CI) | Women (n = 423 of 710) HR (95% CI) | (n = 76 of 172) HR (95% CI) | (n = 118 of 216) HR (95% CI) | (n = 884 of 1522) HR (95% CI) | (n = 289 of 363) HR (95% CI) | |
| Model 1 | |||||||||
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) |
| Moderate | 1.43 (1.10-1.85) | 1.12 (1.01-1.25) | 1.25 (0.90-1.75) | 1.47 (0.75-2.86) | 1.04 (0.81-1.33) | 1.69 (0.89-3.23) | 0.94 (0.58-1.54) | 1.18 (0.99-1.39) | 1.18 (0.88-1.59) |
| Severe | 1.79 (1.39-2.27) | 1.15 (1.04-1.28) | 1.16 (0.85-1.59) | 2.56 (1.33-4.76) | 1.20 (0.96-1.52) | 0.54 (0.32-0.90) | 1.05 (0.68-1.61) | 1.35 (1.16-1.59) | 1.09 (0.82-1.45) |
| Model 2 | |||||||||
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) |
| Moderate | 1.39 (1.05-1.82) | 1.10 (0.98-1.23) | 1.28 (0.90-1.79) | 2.38 (1.03-5.56) | 0.96 (0.74-1.23) | 1.37 (0.69-2.70) | 0.93 (0.56-1.56) | 1.18 (0.99-1.41) | 1.11 (0.82-1.49) |
| Severe | 1.67 (1.25-2.22) | 1.08 (0.96-1.20) | 1.20 (0.84-1.75) | 4.35 (1.72-11.11) | 1.04 (0.80-1.35) | 0.37 (0.20-0.69) | 0.98 (0.61-1.56) | 1.32 (1.11-1.56) | 1.01 (0.74-1.39) |
| P for linear trend | < 0.01 | 0.21 | 0.34 | < 0.01 | 0.45 | < 0.001 | 0.97 | < 0.01 | 0.80 |
P-values for linear trends are calculated for model 2. Hazard ratios and their 95% confidence limits were derived from Cox proportional hazard models.
Significant sex interactions: Any disease P = 0.03, mental disorders P < 0.05, other diagnosis groups 0.24-1.00.
Model 1 adjusted for age, sex, and socioeconomic status. Model 2 adjusted for age, sex, socioeconomic status, night/shift work, smoking, alcohol intake, body mass index, physical activity, diagnosed somatic disease, use of pain killers, depression, anxiety, and use of anxiolytics.
Compared to men without disturbed sleep, men with sleep disturbances had a higher probability of not returning to work following a disability caused by mental disorder, even after adjusting for all potential confounders (moderate sleep disturbance: HR 2.38, 95% CI 1.03-5.56; severe sleep disturbance: HR 4.35, 95% CI 1.72-11.11, linear trend P < 0.01). No such association was found in women (P-value for sex interaction < 0.05). Probability of not returning to work was higher also after disease of the musculoskeletal system among participants with severe sleep disturbances (fully adjusted HR 1.32, 95% CI 1.11-1.56, linear trend P < 0.01; Table 3). Unexpectedly, severe sleep disturbances at baseline were associated with higher probability of returning to work among individuals diagnosed with diseases of the nervous system (fully adjusted HR 0.37, 95% CI 0.20-0.69). Sleep disturbances were not associated with return to work after disability due to malignancies, diseases of the circulatory system, or injuries and poisonings.
Sensitivity Analyses
In a sensitivity analysis excluding cases of work disability within the first two years of follow-up, severe sleep disturbances were associated with 1.37 times greater risk of all-cause work disability and 1.36-1.83 times greater risk of disability due to mental disorders, and diseases of the circulatory and musculoskeletal system, and injuries and poisonings (Table 4). All these associations were statistically significant. Corresponding lagged sensitivity analyses in a subsample additionally excluding those with chronic conditions, depression, or sleep apnea at baseline (n = 42,595) replicated those findings. In the fully adjusted model, severe sleep disturbances were associated with an increased risk of work disability caused by mental disorders (HR 1.55, 95% CI 1.09-2.22), diseases of the circulatory system (HR 2.34, 95% CI 1.35-4.06), diseases of the musculoskeletal system (HR 1.43, 95% CI 1.14-1.81), and injuries and poisonings (HR 1.78, 95% CI 1.13-2.82). A statistically significant linear trend for sleep disturbances was observed for all these associations (all P < 0.05).
Table 4.
Sensitivity analysis on the association of sleep disturbances and the risk of work disability
| Sleep disturbances | Work disability |
|||||||
|---|---|---|---|---|---|---|---|---|
| Any disease (n = 2053) HR (95% CI) | Malignancy (n = 209) HR (95% CI) | Mental disorders (n = 448) HR (95% CI) | Diseases of the nervous system (n = 97) HR (95% CI) | Diseases of the circulatory system (n = 141) HR (95% CI) | Diseases of the musculo-skeletal system (n = 760) HR (95% CI) | Injuries and poisonings (n = 167) HR (95% CI) | ||
| Total sample | (n = 2053) | (n = 209) | (n = 448) | (n = 97) | (n = 141) | (n = 760) | (n = 167) | |
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | |
| Moderate | 1.12 (1.00-1.25) | 1.03 (0.73-1.46) | 1.22 (0.95-1.56) | 1.19 (0.73-1.94) | 1.24 (0.80-1.93) | 1.10 (0.92-1.32) | 0.89 (0.59-1.34) | |
| Severe | 1.37 (1.22-1.53) | 1.31 (0.92-1.86) | 1.45 (1.13-1.88) | 0.94 (0.55-1.62) | 1.83 (1.20-2.77) | 1.36 (1.13-1.63) | 1.48 (1.01-2.17) | |
| P for linear trend | < 0.0001 | 0.27 | 0.01 | 0.66 | 0.02 | < 0.01 | 0.04 | |
| Healthy subsample | (n = 1251) | (n = 153) | (n = 201) | (n = 52) | (n = 80) | (n = 491) | (n = 114) | |
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | |
| Moderate | 1.16 (1.01-1.34) | 1.15 (0.78-1.70) | 1.14 (0.80-1.61) | 1.04 (0.54-1.98) | 1.58 (0.90-2.76) | 1.20 (0.96-1.50) | 0.91 (0.56-1.50) | |
| Severe | 1.40 (1.21-1.62) | 1.17 (0.76-1.79) | 1.55 (1.09-2.22) | 0.63 (0.28-1.43) | 2.34 (1.35-4.06) | 1.43 (1.14-1.81) | 1.78 (1.13-2.82) | |
| P for linear trend | < 0.0001 | 0.70 | < 0.05 | 0.48 | 0.01 | < 0.01 | 0.02 | |
Cases of work disability within the first two years of follow-up have been excluded and the analyses have been carried out for the whole sample (n = 54,756) and for a healthy subsample without chronic condition, depression, or sleep apnea at baseline (n = 42,595). The fully adjusted model is reported. Hazard ratios and their 95% confidence limits were derived from Cox proportional hazard models. P-values for sex interaction 0.10-0.89.
Model adjusted for age, sex, socioeconomic status, night/shift work, smoking, alcohol intake, body mass index, physical activity, diagnosed somatic disease, use of pain killers, depression, anxiety, and use of anxiolytics.
Sensitivity analyses based on the use of sleep medication replicated the majority of the findings. Use of sleep medication predicted all-cause (Figure 2) and cause-specific work disability, except for disability due to malignant tumors, after adjusting for demographics (Table 5). After full adjustments, use of sleep medication was associated with all cause work disability (HR 1.63, 95% CI 1.42-1.88) and work disability due to mental disorders (HR 2.02, 95% CI 1.58-2.58) and diseases of the musculoskeletal system (HR 1.56, 95% CI 1.23-1.98). After adjusting for demographics, the probability of not returning to work for those on sleep medication compared to non-users was 1.32 (95% CI 1.09-1.59) after disability caused by any disease, and 1.39 (95% CI 0.99-1.96) after disabling mental disorders. These associations were nonsignificant in the fully adjusted models. Sleep medication use was not associated with delayed return to work after disease of the musculoskeletal system. We did not analyze probability of not returning to work after work disability due to malignant tumors, diseases of the nervous or circulatory system, or injuries and poisonings because of insufficient statistical power (< 10 cases among those using sleep medication). No interaction between sex and sleep medication use was found either for risk of work disability or probability of not returning to work.
Figure 2.
The association between use of sleep medication and work disability. Crude cumulative hazard rates for risk of work disability for participants with or without sleep medication use, as estimated by Kaplan-Meier survival analysis.
Table 5.
The association between use of sleep medication and the risk of work disability
| Sleep medication | Work disability |
||||||
|---|---|---|---|---|---|---|---|
| Any disease (n = 4028) HR (95% CI) | Malignancy (n = 398) HR (95% CI) | Mental disorders (n = 841) HR (95% CI) | Diseases of the nervous system (n = 174) HR (95% CI) | Diseases of the circulatory system (n = 252) HR (95% CI) | Diseases of the musculoskeletal system (n = 1522) HR (95% CI) | Injuries and poisonings (n = 363) HR (95% CI) | |
| Model 1 | |||||||
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) |
| Yes | 2.71 (2.37-3.11) | 0.92 (0.47-1.78) | 5.37 (4.26-6.77) | 2.44 (1.24-4.78) | 2.09 (1.17-3.75) | 2.35 (1.87-2.96) | 2.21 (1.33-3.65) |
| Model 2 | |||||||
| No | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) | 1.00 (ref.) |
| Yes | 1.63 (1.42-1.88) | 0.87 (0.45-1.72) | 2.02 (1.58-2.58) | 1.55 (0.77-3.13) | 1.64 (0.90-3.01) | 1.56 (1.23-1.98) | 1.58 (0.93-2.67) |
Hazard ratios and their 95% confidence limits were derived from Cox proportional hazard models. P-values for sex interaction 0.45-0.93.
Model 1 adjusted for age, sex, and socioeconomic status. Model 2 adjusted for age, sex, socioeconomic status, night/shift work, smoking, alcohol intake, body mass index, physical activity, diagnosed somatic disease, use of pain killers, depression, anxiety, and use of anxiolytics.
DISCUSSION
Self-reported sleep disturbances were associated with increased risk of all cause and cause-specific work disability and with delayed return to work. Severe sleep disturbances predicted incident work disability due to mental disorders and diseases of the nervous system, musculoskeletal system, circulatory system, and injuries and poisonings, but not disability due to malignant tumors. Probability of not returning to work after disability due to musculoskeletal disorders was increased in men and women who experienced severe sleep disturbances at baseline. The same was true for men with work disability due to mental disorders. The main results were replicated after excluding those with comorbid conditions at baseline or disability within the first two years of follow-up and by using sleep medication as a proxy indicator for sleep disturbances.
Predicting Cause-Specific Work Disability
Our findings are in agreement with previous studies that reported sleep disturbances to be associated with increased sickness absence and retirement due to declines in health.6–13 They are also consistent with cross-sectional studies linking sleep disturbances to ill-health as measured by the presence of mental disorders,24 neurological diseases,25,26 musculoskeletal problems,27 cardiovascular disease events, and mortality.28,29 Given that sleep disturbances predicted (with the exception of malignancies) disability due to all main categories of disease outcomes evaluated, some underlying processes are likely to be general, and may include factors such as immune dysfunction, impaired host defenses, increased levels of cortisol secretion, dysregulation of the autonomic nervous system,30,31 and alterations in metabolism.32 However, it has been suggested that inflammation33 and impaired energy metabolism34 due to sleep loss contribute, especially to musculoskeletal symptoms or problems in tissue recovery,27 and there is evidence that low-grade inflammation may precede the onset of depressive symptoms35 and coronary events.36 Lack of sleep increases sensitivity to pain37; in diseases that inflict acute or chronic pain, such as many musculoskeletal diseases, pain may disturb sleep and sleep disturbances further augment pain.38
The mechanism underlying the association between sleep disturbances and work disability due to external causes, such as injuries and poisonings, is likely to be different than that for disabling diseases analyzed in this study. Injuries and accidents may be attributable to more direct consequences of sleep disturbances, such as sleepiness and decreased vigilance caused by related sleep loss.39 In this study, women (but not men) with sleep disturbances had an increased risk of work disability due to injuries and poisonings. This finding is in agreement with a recent study reporting difficulties initiating sleep to be associated with occupational injury in women.40 However, some data suggest a stronger association of any sleep disturbances with occupational accidents and injuries among men compared to women.40,41 Those studies focused on occupational injuries and included less severe events with shorter duration of work disability,40 and self-reported minor scratches or cuts, which might not have led to disability.41 In contrast, we focused on long-term work disability due to more severe injuries and poisonings which were not necessarily work-related.
The sensitivity analyses with the healthy subcohort and those free of work disability within the first two years of follow-up suggest that our findings are not attributable to reverse confounding bias. Indeed, the association between sleep disturbances and diseases of the circulatory system was slightly strengthened after excluding subjects with baseline indications of depression, chronic somatic condition, or sleep apnea. These findings fail to support the suggestion that the association between disturbed sleep and cardiovascular events would be confounded by poor health in general and depression in particular.42 Instead, the association between disturbed sleep and cardiovascular events could be conveyed through other mechanisms, such as dysfunction of the autonomous nervous system or inflammation.36,42
Outcome of Work Disability
Investigation of sleep disturbances as a predictor of the outcome of work disability is a unique feature of the present study. Baseline sleep disturbances prior to any disability predicted delayed return to work after the onset of work disability caused by musculoskeletal diseases. Our findings are consistent with clinical studies that have shown that experience of restorative sleep and less trouble falling asleep at baseline predict faster resolution of chronic widespread pain and improved musculoskeletal health.16 While those findings support the value of sleep disturbances as a prognostic marker in patients with a disease, our results suggest that sleep disturbances may provide prognostic information prior to the onset of disabling disease.
We found an association between baseline sleep disturbances and increased probability of not returning to work from mental health related disability among men; no such relationship was observed in women. While sleep problems have been associated with poorer prognosis of anxiety and depression43 and good sleep with faster return to work after clinical burn-out,15 previous studies do not provide strong support for sex differences in the association between sleep disturbances and morbidity in non-work context. Female employees tend to take sick leave at earlier phases of a disease than males44; thus, baseline sleep disturbances in working men may indicate a more severe mental disorder with a poorer prognosis. Although sleep disturbances and mental disorders are more strongly related to alcohol abuse in men than women,45,46 our findings suggest that baseline health and health behaviors may explain only part of the observed sex differences, as the association remained after statistically adjusting for baseline health and excessive alcohol use. Further research is needed to understand the observed sex differences in the ability of sleep disturbances to predict the outcome of disabling mental disorders.
Surprisingly, participants with severe baseline sleep disturbances were more likely to return to work after disabling nervous system disease than those free of sleep disturbances. This apparently paradoxical finding might relate to a higher prevalence of mononeuropathies of upper limb (e.g., carpal tunnel syndrome) or other treatable diseases among those with severe sleep disturbance. In those who returned to work during follow-up, the prevalence of work disability due to mononeuropathies was 50% in the group with severe sleep disturbances and 35% in those with no sleep disturbances. Mononeuropathies disturb sleep because the symptoms are typically amplified in rest, but they can be effectively treated with surgery that quickly eliminates the pain and thus may lead to fast return to work.
Limitations
This study benefitted from a large sample size, prospective study design, and reasonably high response rate. To minimize subjectivity bias, we retrieved objective data on work disability from independent national registers, which reliably cover all long-term sickness absences and disability pensions with diagnoses. However, the study population was comprised exclusively of Finnish public sector employees, who were predominantly female (80%) and racially homogeneous (white). Further research covering various ethnic groups and employees in private sector is needed to assess the generalizability of our findings, especially in the light of marked racial and ethnic group differences in sleep.47
Data from one measurement of sleep disturbances may have underestimated the effect of long-term sleep disturbances on work disability as such measurement does not enable us to distinguish between chronic and temporary sleep disturbances. Further research with repeated measurements could inform whether the associations with work disability are specific to acute or chronic sleep disturbances and their relationship to primary sleep disorders not measured in the present study.
With observational data, we are unable to exclude several alternative explanations for our findings. Instead of serving as a causal mechanism for developing a disease, disturbed sleep may be an early symptom of an undiagnosed or subclinical disease, or it may be a factor affecting only the course of an existing illness. A further alternative is that employees with co-occurring sleep disturbances and illness may more likely take sick leave from work than those with an equally severe illness but no sleep disturbance. Intervention studies are needed to further test causality between sleep disturbances and work disability. Such causality will be supported if successful treatment of sleep difficulties reduces work disability and precipitates the return to work.
CONCLUSION
The findings from this large contemporary cohort of Finnish employees support the hypothesis that sleep disturbances may contribute to the development of disabling diseases and adversely affect the outcome of work disability due to various disorders.
DISCLOSURE STATEMENT
This was not an industry supported study. Dr. Salo has financial interests in a private sleep clinic in Turku, Finland. The sleep clinic had no involvement in this study. The other authors have indicated no financial conflicts of interest.
ACKNOWLEDGMENTS
This study was supported by the Academy of Finland (projects 124271, 124322, 129262, and 132944), the Social Insurance Institution of Finland, and the participating organizations. Mika Kivimäki is supported by the NIH/National Institute on Aging (R01AG034454) and the National Heart, Lung, and Blood Institute (R01HL036310), USA, and the BUPA Foundation, UK. The sponsors had no role in designing the study, analyzing or interpreting the data, or preparing the manuscript.The work was performed at the Finnish Institute of Occupational Health.
Appendix 1. Disease categories of the work disability cases among the Finnish public sector employees.
| ICD-10 Code | ICD-10 Category label | Work disability cases N |
|---|---|---|
| C00-C97 | Malignant neoplasms | 398 |
| C00-C75 | Malignant neoplasms, stated or presumed to be primary, of specified sites, except of lymphoid, haematopoietic and related tissue | 346 |
| C76-C80 | Malignant neoplasms of ill-defined, secondary and unspecified sites | 11 |
| C81-C96 | Malignant neoplasms, stated or presumed to be primary, of lymphoid, haematopoietic and related tissue | 41 |
| C97 | Malignant neoplasms of independent (primary) multiple sites | 0 |
| F00-F99 | Mental and behavioural disorders | 823 |
| F00-F09 | Organic, including symptomatic, mental disorders | 4 |
| F10-F19 | Mental and behavioural disorders due to psychoactive substance use | 3 |
| F20-F29 | Schizophrenia, schizotypal and delusional disorders | 20 |
| F30-F39 | Mood [affective] disorders | 626 |
| F40-F48 | Neurotic, stress-related and somatoform disorders | 155 |
| F50-F59 | Behavioural syndromes associated with physiological disturbances and physical factors | 9 |
| F60-F69 | Disorders of adult personality and behaviour | 3 |
| F70-F79 | Mental retardation | 1 |
| F80-F89 | Disorders of psychological development | 2 |
| F90-F98 | Behavioural and emotional disorders with onset usually occurring in childhood and adolescence | 0 |
| F99 | Unspecified mental disorder | 0 |
| X60-X84 | Intentional self-harm (Suicide) | 18 |
| G00-G99 | Diseases of the nervous system | 174 |
| G00-G09 | Inflammatory diseases of the central nervous system | 1 |
| G10-G13 | Systemic atrophies primarily affecting the central nervous system | 5 |
| G20-G26 | Extrapyramidal and movement disorders | 22 |
| G30-G32 | Other degenerative diseases of the nervous system | 12 |
| G35-G37 | Demyelinating diseases of the central nervous system | 22 |
| G40-G47 | Episodic and paroxysmal disorders | 31 |
| G50-G59 | Nerve, nerve root and plexus disorders | 59 |
| G60-G64 | Polyneuropathies and other disorders of the peripheral nervous system | 6 |
| G70-G73 | Diseases of myoneural junction and muscle | 5 |
| G80-G83 | Cerebral palsy and other paralytic syndromes | 3 |
| G90-G99 | Other disorders of the nervous system | 8 |
| I00-I99 | Diseases of the circulatory system | 252 |
| I00-I02 | Acute rheumatic fever | 0 |
| I05-I09 | Chronic rheumatic heart diseases | 0 |
| I10-I15 | Hypertensive diseases | 14 |
| I20-I25 | Ischaemic heart diseases | 97 |
| I26-I28 | Pulmonary heart disease and diseases of pulmonary circulation | 2 |
| I30-I52 | Other forms of heart disease | 42 |
| I60-I69 | Cerebrovascular diseases | 77 |
| I70-I79 | Diseases of arteries, arterioles and capillaries | 7 |
| I80-I89 | Diseases of veins, lymphatic vessels and lymph nodes, not elsewhere classified | 12 |
| I95-I99 | Other and unspecified disorders of the circulatory system | 1 |
| M00-M99 | Diseases of the musculoskeletal system | 1522 |
| M00-M25 | Arthropathies | 530 |
| M30-M36 | Systemic connective tissue disorders | 19 |
| M40-M54 | Dorsopathies | 629 |
| M60-M79 | Soft tissue disorders | 322 |
| M80-M94 | Osteopathies and chondropathies | 15 |
| M95-M99 | Other disorders of the musculoskeletal system and connective tissue | 7 |
| S00-T98 | Injuries and poisonings | 363 |
REFERENCES
- 1.Ohayon MM. Epidemiology of insomnia: What we know and what we still need to learn. Sleep Med Rev. 2002;6:97–111. doi: 10.1053/smrv.2002.0186. [DOI] [PubMed] [Google Scholar]
- 2.Kronholm E, Partonen T, Laatikainen T, et al. Trends in self-reported sleep duration and insomnia-related symptoms in Finland from 1972 to 2005: A comparative review and re-analysis of Finnish population samples. J Sleep Res. 2008;17:54–62. doi: 10.1111/j.1365-2869.2008.00627.x. [DOI] [PubMed] [Google Scholar]
- 3.Sivertsen B, Krokstad S, Mykletun A, et al. Insomnia symptoms and use of health care services and medications: the HUNT-2 study. Behav Sleep Med. 2009;7:210–22. doi: 10.1080/15402000903190199. [DOI] [PubMed] [Google Scholar]
- 4.Daley M, Morin CM, Leblanc M, Gregoire JP, Savard J, Baillargeon L. Insomnia and its relationship to health-care utilization, work absenteeism, productivity and accidents. Sleep Med. 2009;10:427–38. doi: 10.1016/j.sleep.2008.04.005. [DOI] [PubMed] [Google Scholar]
- 5.Linton SJ, Bryngelsson I. Insomnia and its relationship to work and health in a working-age population. J Occup Rehabil. 2000;10:169–83. [Google Scholar]
- 6.øverland S, Glozier N, Sivertsen B, et al. A comparison of insomnia and depression as predictors of disability pension: The HUNT study. Sleep. 2008;31:875–80. doi: 10.1093/sleep/31.6.875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sivertsen B, øverland S, Neckelmann D, et al. The long-term effect of insomnia on work disability: The HUNT-2 historical cohort study. Am J Epidemiol. 2006;163:1018–24. doi: 10.1093/aje/kwj145. [DOI] [PubMed] [Google Scholar]
- 8.Eriksen W, Natvig B, Bruusgaard D. Sleep problems: A predictor of long-term work disability? A four-year prospective study. Scand J Public Health. 2001;29:23–31. [PubMed] [Google Scholar]
- 9.Åkerstedt T, Kecklund G, Alfredsson L, Selen J. Predicting long-term sickness absence from sleep and fatigue. J Sleep Res. 2007;16:341–5. doi: 10.1111/j.1365-2869.2007.00609.x. [DOI] [PubMed] [Google Scholar]
- 10.Westerlund H, Alexanderson K, Åkerstedt T, Magnusson Hanson L, Theorell T, Kivimäki M. Work-related sleep disturbances and sickness absence in the Swedish working population, 1993-1999. Sleep. 2008;31:1169–77. [PMC free article] [PubMed] [Google Scholar]
- 11.Ihlebaek C, Brage S, Eriksen HR. Health complaints and sickness absence in Norway, 1996-2003. Occup Med (Lond) 2007;57:43–9. doi: 10.1093/occmed/kql107. [DOI] [PubMed] [Google Scholar]
- 12.Sivertsen B, øverland S, Bjorvatn B, Maeland JG, Mykletun A. Does insomnia predict sick leave? The Hordaland health study. J Psychosom Res. 2009;66:67–74. doi: 10.1016/j.jpsychores.2008.06.011. [DOI] [PubMed] [Google Scholar]
- 13.Sivertsen B, øverland S, Pallesen S, et al. Insomnia and long sleep duration are risk factors for later work disability. The Hordaland health study. J Sleep Res. 2009;18:122–8. doi: 10.1111/j.1365-2869.2008.00697.x. [DOI] [PubMed] [Google Scholar]
- 14.Vahtera J, Pentti J, Helenius H, Kivimäki M. Sleep disturbances as a predictor of long-term increase in sickness absence among employees after family death or illness. Sleep. 2006;29:673–82. doi: 10.1093/sleep/29.5.673. [DOI] [PubMed] [Google Scholar]
- 15.Sonnenschein M, Sorbi MJ, Verbraak MJ, Schaufeli WB, Maas CJ, van Doornen LJ. Influence of sleep on symptom improvement and return to work in clinical burnout. Scand J Work Environ Health. 2008;34:23–32. doi: 10.5271/sjweh.1195. [DOI] [PubMed] [Google Scholar]
- 16.Davies KA, Macfarlane GJ, Nicholl BI, et al. Restorative sleep predicts the resolution of chronic widespread pain: Results from the EPIFUND study. Rheumatology (Oxford) 2008;47:1809–13. doi: 10.1093/rheumatology/ken389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jenkins CD, Stanton BA, Niemcryk SJ, Rose RM. A scale for the estimation of sleep problems in clinical research. J Clin Epidemiol. 1988;41:313–21. doi: 10.1016/0895-4356(88)90138-2. [DOI] [PubMed] [Google Scholar]
- 18.Statistics Finland. Classification of Occupations. Helsinki: Statistics Finland; 2001. [Google Scholar]
- 19.Kujala UM, Kaprio J, Sarna S, Koskenvuo M. Relationship of leisure-time physical activity and mortality: The Finnish twin cohort. JAMA. 1998;279:440–4. doi: 10.1001/jama.279.6.440. [DOI] [PubMed] [Google Scholar]
- 20.WHO Collaborating Centre for Drug Statistics Methodology. Guidelines for ATC Classification and DDD Assignment. Oslo: World Health Organization Collaborating Centre for Drug Statistics; 2004. [Google Scholar]
- 21.Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State-Trait Anxiety Inventory (Form Y) Palo Alto: Consulting Psychologists Press; 1983. [Google Scholar]
- 22.Vahtera J, Poikolainen K, Kivimäki M, Ala-Mursula L, Pentti J. Alcohol intake and sickness absence: a curvilinear relation. Am J Epidemiol. 2002;156:969–76. doi: 10.1093/aje/kwf138. [DOI] [PubMed] [Google Scholar]
- 23.Neovius K, Johansson K, Kark M, Neovius M. Obesity status and sick leave: a systematic review. Obes Rev. 2009;10:17–27. doi: 10.1111/j.1467-789X.2008.00521.x. [DOI] [PubMed] [Google Scholar]
- 24.Jansson-Fräjmark M, Lindblom K. A bidirectional relationship between anxiety and depression, and insomnia? A prospective study in the general population. J Psychosom Res. 2008;64:443–9. doi: 10.1016/j.jpsychores.2007.10.016. [DOI] [PubMed] [Google Scholar]
- 25.Verbaan D, van Rooden SM, Visser M, Marinus J, van Hilten JJ. Nighttime sleep problems and daytime sleepiness in parkinson's disease. Mov Disord. 2008;23:35–41. doi: 10.1002/mds.21727. [DOI] [PubMed] [Google Scholar]
- 26.Elwood P, Hack M, Pickering J, Hughes J, Gallacher J. Sleep disturbance, stroke, and heart disease events: Evidence from the Caerphilly cohort. J Epidemiol Community Health. 2006;60:69–73. doi: 10.1136/jech.2005.039057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Miranda H, Viikari-Juntura E, Punnett L, Riihimäki H. Occupational loading, health behavior and sleep disturbance as predictors of low-back pain. Scand J Work Environ Health. 2008;34:411–9. doi: 10.5271/sjweh.1290. [DOI] [PubMed] [Google Scholar]
- 28.Mallon L, Broman JE, Hetta J. Sleep complaints predict coronary artery disease mortality in males: A 12-year follow-up study of a middle-aged Swedish population. J Intern Med. 2002;251:207–16. doi: 10.1046/j.1365-2796.2002.00941.x. [DOI] [PubMed] [Google Scholar]
- 29.Phillips B, Mannino DM. Do insomnia complaints cause hypertension or cardiovascular disease? J Clin Sleep Med. 2007;3:489–94. [PMC free article] [PubMed] [Google Scholar]
- 30.Hall M, Thayer JF, Germain A, et al. Psychological stress is associated with heightened physiological arousal during NREM sleep in primary insomnia. Behav Sleep Med. 2007;5:178–93. doi: 10.1080/15402000701263221. [DOI] [PubMed] [Google Scholar]
- 31.Hall M, Vasko R, Buysse D, et al. Acute stress affects heart rate variability during sleep. Psychosom Med. 2004;66:56–62. doi: 10.1097/01.psy.0000106884.58744.09. [DOI] [PubMed] [Google Scholar]
- 32.Friese RS. Sleep and recovery from critical illness and injury: A review of theory, current practice, and future directions. Crit Care Med. 2008;36:697–705. doi: 10.1097/CCM.0B013E3181643F29. [DOI] [PubMed] [Google Scholar]
- 33.Irwin MR, Wang M, Campomayor CO, Collado-Hidalgo A, Cole S. Sleep deprivation and activation of morning levels of cellular and genomic markers of inflammation. Arch Intern Med. 2006;166:1756–62. doi: 10.1001/archinte.166.16.1756. [DOI] [PubMed] [Google Scholar]
- 34.Schmid SM, Hallschmid M, Jauch-Chara K, Bandorf N, Born J, Schultes B. Sleep loss alters basal metabolic hormone secretion and modulates the dynamic counterregulatory response to hypoglycemia. J Clin Endocrinol Metab. 2007;92:3044–51. doi: 10.1210/jc.2006-2788. [DOI] [PubMed] [Google Scholar]
- 35.Gimeno D, Kivimäki M, Brunner EJ, et al. Associations of C-reactive protein and interleukin-6 with cognitive symptoms of depression: 12-year follow-up of the Whitehall II study. Psychol Med. 2009;39:413–23. doi: 10.1017/S0033291708003723. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002;347:1557–65. doi: 10.1056/NEJMoa021993. [DOI] [PubMed] [Google Scholar]
- 37.Roehrs T, Hyde M, Blaisdell B, Greenwald M, Roth T. Sleep loss and REM sleep loss are hyperalgesic. Sleep. 2006;29:145–51. doi: 10.1093/sleep/29.2.145. [DOI] [PubMed] [Google Scholar]
- 38.Marty M, Rozenberg S, Duplan B, et al. Quality of sleep in patients with chronic low back pain: A case-control study. Eur Spine J. 2008;17:839–44. doi: 10.1007/s00586-008-0660-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Connor J, Norton R, Ameratunga S, et al. Driver sleepiness and risk of serious injury to car occupants: population based case control study. [Accessed 22 February 2010];BMJ. 2002 324:1125. doi: 10.1136/bmj.324.7346.1125. doi: 10.1136/bmj.324.7346.1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Salminen S, Oksanen T, Vahtera J, et al. Sleep disturbances as a predictor of occupational injuries among public sector workers. J Sleep Res. 2010;19:207–13. doi: 10.1111/j.1365-2869.2009.00780.x. [DOI] [PubMed] [Google Scholar]
- 41.Nakata A, Ikeda T, Takahashi M, et al. Sleep-related risk of occupational injuries in Japanese small and medium-scale enterprises. Ind Health. 2005;43:89–97. doi: 10.2486/indhealth.43.89. [DOI] [PubMed] [Google Scholar]
- 42.Schwartz S, McDowell Anderson W, et al. Insomnia and heart disease: a review of epidemiologic studies. J Psychosom Res. 1999;47:313–33. doi: 10.1016/s0022-3999(99)00029-x. [DOI] [PubMed] [Google Scholar]
- 43.Zammit GK. The prevalence, morbidities, and treatments of insomnia. CNS Neurol Disord Drug Targets. 2007;6:3–16. doi: 10.2174/187152707779940754. [DOI] [PubMed] [Google Scholar]
- 44.Ahola K, Kivimäki M, Honkonen T, et al. Occupational burnout and medically certified sickness absence: A population-based study of Finnish employees. J Psychosom Res. 2008;64:185–93. doi: 10.1016/j.jpsychores.2007.06.022. [DOI] [PubMed] [Google Scholar]
- 45.Marcus SM, Kerber KB, Rush AJ, et al. Sex differences in depression symptoms in treatment-seeking adults: Confirmatory analyses from the sequenced treatment alternatives to relieve depression study. Compr Psychiatry. 2008;49:238–46. doi: 10.1016/j.comppsych.2007.06.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Janson C, Lindberg E, Gislason T, Elmasry A, Boman G. Insomnia in men-a 10-year prospective population based study. Sleep. 2001;24:425–30. doi: 10.1093/sleep/24.4.425. [DOI] [PubMed] [Google Scholar]
- 47.Hall MH, Matthews KA, Kravitz HM, et al. Race and financial strain are independent correlates of sleep in midlife women: The SWAN sleep study. Sleep. 2009;32:73–82. [PMC free article] [PubMed] [Google Scholar]