If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Rehabilitation Medicine, University of Washington, Seattle, WADepartment of Neurological Surgery, University of Washington, Seattle, WADepartment of Psychiatry and Behavioral Medicine, University of Washington, Seattle, WA
Department of Rehabilitation Medicine, University of Washington, Seattle, WADepartment of Neurological Surgery, University of Washington, Seattle, WADepartment of Biostatistics, University of Washington, Seattle, WA
Fogelberg DJ, Hoffman JM, Dikmen S, Temkin NR, Bell KR. Association of sleep and co-occurring psychological conditions at 1 year after traumatic brain injury.
To compare individuals' sleep 1 year after traumatic brain injury (TBI) with that of a healthy comparison group, and examine the relationship between sleep, co-occurring conditions, and functional status in those with TBI.
Longitudinal assessment of a prospectively studied sample of individuals with moderate to severe TBI. Assessment of sleep occurred at 1 year after TBI.
Inpatient acute rehabilitation for TBI and community follow-up at 1 year postinjury.
Individuals with TBI (N=174) were recruited from consecutive admissions to an inpatient rehabilitation unit and enrolled into the TBI Model Systems study. Participant mean age was 38, and mean Glasgow Coma Scale score on admission was 9.3. Seventy-eight percent of the sample were men.
Main Outcome Measures
Sleep was assessed with the Pittsburgh Sleep Quality Index (PSQI). Depression, anxiety, and pain were measured with the Patient Health Questionnaire-9, the Generalized Anxiety Disorder-7 Scale, and an analog pain rating scale, respectively.
Participants with TBI reported significantly greater sleep difficulties than the healthy comparison group. Forty-four percent of participants with TBI reported significant sleep problems (PSQI>5). Participants with 1 or more co-occurring conditions (depression, pain, or anxiety) had significantly worse sleep than those without such a condition. The highest level of sleep problems was reported by participants with multiple co-occurring conditions. Sleep problems were also associated with poor functional status.
Sleep difficulties are a frequent problem at 1 year after TBI, and often co-occur with depression, anxiety, and pain. Assessment and treatment of sleep difficulties should be included in clinical practice. Future research on the potential causal relationship among co-occurring conditions may assist in additional intervention planning.
found that individuals had difficulty in sleep initiation, maintenance, quality, adequacy, and somnolence. Sleep disturbances are not confined to the newly injured, but are an enduring problem for a significant number of individuals with TBI. In a survey of individuals who were between 15 and 24 years postinjury, Colantonio et al
found that sleep problems continued to affect 27% of the sample.
Although the body of literature examining sleep disturbance among adults with TBI has grown in recent years, the picture remains unclear and fragmented. The diversity of findings is evident in a survey of studies of disrupted sleep after TBI by Zeitzer et al,
who reported prevalence rates ranging between 3% and 84%. Most studies have included small sample sizes, and those with larger sample sizes contain a wide range of injury severity and enrollment at many different times postinjury.
The picture is further complicated by the various forms of sleep disturbance measured. Additionally, there is little consistency in the methods used to measure sleep disturbance; these have ranged from endorsement of a single item on a questionnaire to full polysomnographic assessment.
There are multiple reasons to suspect that sleep disturbance in its various forms could be associated with poor functional status. Within the general population, the consequences of sleep disturbance can be severe, and even relatively short periods of sleep disturbance can potentially impair cognitive functioning, in particular, executive functions including decision-making and judgment of risk.
Given the direct impact TBI can have on these functions, the potential of sleep disturbance, which can further compromise them, is of concern. Furthermore, disorders of sleep may be associated with various co-occurring conditions seen in individuals with TBI. For example, sleep deprived individuals experience greater levels of fatigue and sleepiness,
have suggested that fatigue may not only result from sleep disturbance, but may actually contribute to long-term insomnia by encouraging irregular sleep habits. Chronic pain, a condition found in up to 76% of individuals up to 1 year after TBI, interferes with sleep.
The relationship between many of these conditions and sleep has not been well studied in individuals with TBI, and most of the studies conducted have had relatively small sample sizes.
The goals of the current study were to examine sleep difficulties in individuals with TBI compared with a healthy comparison group, and to examine the relationship between sleep difficulties and common co-occurring disorders including depression, anxiety, and pain at 1 year postinjury. Our sample included adults with TBI enrolled at the time of acute rehabilitation care and longitudinally followed and tested at 1 year after injury. In addition, we examined the relationship between sleep problems and functional status at 1 year after injury.
Participants were a subset of individuals with a TBI, who were recruited from consecutive admissions to a single inpatient rehabilitation unit and were enrolled into the Traumatic Brain Injury Model Systems (TBIMS) and followed at 1 year after injury. The TBIMS is a longitudinal database and research program sponsored by the National Institute on Disability and Rehabilitation Research. Inclusion criteria were the same as for the TBIMS: (1) a diagnosis of TBI with at least 1 of the following: Glasgow Coma Scale (GCS) score <13 on emergency department admission not due to intubation, sedation, or intoxication; loss of consciousness >30 minutes; posttraumatic amnesia >24 hours; or intracranial abnormality on neuroimaging; (2) at least 16 years of age; (3) arrival to the system's acute hospital within 72 hours of acute injury; and (4) received both acute medical and rehabilitation care within the same system. Over 90% of eligible subjects with TBI admitted to rehabilitation consented to participate in the TBIMS. Of the 174 individuals who were consecutively enrolled between August 2004 and October 2007, a total of 157 (90%) were followed at 1 year after injury. In addition to assessments specific to the TBIMS, additional questionnaires were given to assess sleep and mood. For this study, we examined the 129 who completed the Pittsburgh Sleep Quality Index (PSQI)
at their scheduled 1-year follow-up. A comparison of the 129 who completed the PSQI with the 45 who were enrolled but did not complete the measure showed no difference on age, sex, cause of injury, or injury severity. The 2 groups did differ significantly on race such that those who were not followed had fewer individuals who endorsed being white (64% vs 84%, P<.001). The primary reason that participants did not complete this measure was insufficient time for administration during telephone follow-up. The TBIMS study and the additional measures were approved by the Human Subjects Division of the University of Washington.
Demographic and injury data
Data were gathered through medical chart review and interviews conducted with the subject or their family and included age, sex, race/ethnicity (white, black, Asian/Pacific Islander, Native American, and Hispanic), cause of injury (vehicular, violence, sports, fall, pedestrian, and other), and GCS score
This is a 19-item questionnaire, which yields scores for 7 components of sleep (subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction), each of which has a range of 0 to 3, with higher scores indicating more problematic sleep. These scores can be summed to produce a global score that has a range of 0 to 21. The PSQI's global score has been shown to differentiate between good sleepers (≤5) and poor sleepers (>5) in the general population,
Scores on this 7-item measure range from 0 to 21, with a score of 10 or higher indicating anxiety. Pain scores were based on participant report of the average intensity of their overall pain during the past week. This average is reported on an analog scale ranging from 0 (no pain) to 10 (pain as bad as can be). The 0 to 10 numerical rating scale has consistently shown its validity as a measure of pain intensity through its strong association with other measures of pain intensity, as well as its sensitivity to detect changes in pain associated with pain treatments.
scores were used to measure functional outcomes and have been collected as part of the TBIMS National Database since 1994. Scores on this measure range from 0 to 29, with higher scores indicating a greater degree of functional impairment. This is not a true interval measure, and its sensitivity decreases at the lower end of the scale.
were also used to indicate subjects' functional outcomes. This measure includes items assessing both physical and cognitive abilities, and is widely used in rehabilitation. Total scores range from 18 to 126, with lower scores indicating a greater level of care needed for the individual. Life satisfaction, a third outcome, was measured using the Satisfaction With Life Scale (SWLS).
Total scores on this 5-item measure range from 5 to 35, with higher scores indicating a greater degree of life satisfaction. This measure has been found to have good reliability and construct validity.
Simple descriptive statistics were used to describe the population. Mann-Whitney U tests and Fisher exact tests were used to compare those with TBI with and without sleep disturbance on demographics, co-occurring disorders, and functional status. Independent sample t tests were used to compare the study population with the results from Buysse et al.
Buysse presented mean and SDs on all subscale and global scores for 52 healthy individuals without identified sleep disorders. The 52 healthy comparison subjects included 40 men and 12 women with an average age of 59.9 years (range, 24–83y). Correlations were calculated among the global sleep score and the measures of co-occurring conditions, in order to further examine the relationship among these variables. Independent sample t tests compared those with and without co-occurring conditions with each other and with the healthy comparison subjects. Given the number of comparisons conducted, P<.001 was considered significant. All other P values between .05 and .001 were considered potential trends.
Demographics and Injury Data
Demographic and injury data collected during inpatient rehabilitation are included in table 1.
Table 1Demographic and Injury Data Comparison for Those With and Without Disordered Sleep
Differences Between Individuals With and Without Sleep Disturbance
Table 1 also contains comparisons between participants with and without sleep disturbance. At 1 year after TBI, 44% of the participants had global PSQI scores greater than 5, indicating the presence of a clinically significant sleep disturbance.
Those participants who were classified as poor sleepers were more likely to be nonwhite and have significant levels of depression, anxiety, or pain. In addition, they also had significantly worse scores on all 3 functional status measures (FIM, DRS, and SWLS).
Sleep Disturbance at 1 Year After TBI
Mean scores for each of the sleep components and the global score of the PSQI are shown in table 2. The mean global score was 5.54, which indicates a significant sleep problem. Significant differences between the current sample and the healthy comparison group were found in 4 of the 7 sleep components measured by the PSQI, as well as for the global scores. Specifically, participants with TBI took significantly longer to fall asleep (sleep latency), used more sleep medications (medication use), and reported worse effects on daytime functioning (daytime dysfunction) compared with the healthy comparison group. The participants rated the overall quality of their sleep (subjective sleep quality) lower than the comparison group in the Buysse et al study.
There was a trend toward those with TBI reporting sleep more disturbed by awakening during the night, respiratory problems, pain, or being too hot or too cold than the healthy comparison group. No significant differences were found in number of hours slept (sleep duration) and the ratio of hours slept to hours spent in bed (habitual sleep efficiency).
Table 2Comparison of PSQI for TBI and Healthy Group
Association of Sleep Disturbance With Co-Occurring Conditions
In this sample, global sleep scores, depression, anxiety, and pain were all highly correlated (r=.72, .39, and .56 between sleep and depression, anxiety, and pain, respectively; P≤.001 for each). In order to better understand the relationship between sleep and these co-occurring conditions, the sample was divided into groups based on co-occurring conditions: no co-occurring conditions (n=79), a single co-occurring condition (anxiety n=1, depression n=6, pain n=13), and 2 or more co-occurring conditions (depression and pain, n=6; n of other 2 pairings was 0, all 3 n=16). Nine participants with missing data on 1 or more variables were not included in this analysis. Total PSQI scores for each of the groups represented within the sample are depicted in figure 1. Those participants with 2 or 3 co-occurring conditions reported the worst sleep.
A series of independent sample t tests were conducted comparing subjects with 1 or more of the co-occurring disorders with subjects without any of these disorders (table 3). These 2 groups differed significantly on 5 of the dimensions (latency, disturbance, daytime dysfunction, quality, and medication use), as well as the global score, with a trend toward difference in duration.
The differences between subjects with 1 or more co-occurring conditions and the healthy comparison group were significant (P<.001) on 6 of the dimensions. Those subjects with TBI and no co-occurring disorders had trends toward differences from the healthy comparison group on 3 dimensions (latency, habitual efficiency, and medication use), as well as the global score.
The results of this study indicate that individuals with TBI have significantly more problems with sleep than previously published data on healthy individuals, a finding that is consistent with existing literature.
Of the 7 components of sleep measured by the PSQI, those with TBI indicated worse functioning on 4 (sleep quality, latency, sleep medication use, and daytime dysfunction), as well as on the global score. Sleep problems were most pronounced among those subjects who reported significant levels of pain, depression, or anxiety, with the highest scores on the PSQI reported by those individuals who reported a combination of 2 or more of these conditions. There were no significant differences found between those subjects with TBI who had no co-occurring conditions and the healthy comparison group, but trends were observed for sleep latency, sleep efficiency, medication use, and the global score. Additionally, the functional status of individuals with and without sleep disturbance were significantly different, with those reporting sleep problems describing more functional limitations, higher reliance on others, and diminished satisfaction with their quality of life. Given that the healthy sample was older, which may increase the likelihood of sleep problems, the level of sleep disturbance in those with TBI may be underestimated. While we were unable to determine whether there is any causal link between these outcomes and sleep, the outcomes do suggest the need to address sleep difficulties, which may impact function.
Five other studies have examined sleep disturbances in TBI
which are likely related to the samples studied. Our sample was limited to individuals who were admitted for rehabilitation services. However, by prospectively examining a group enrolled shortly after injury, we avoided the bias created by enrolling individuals presenting to outpatient clinics,
Our study reports on the nature of the sleep difficulties (eg, sleep quality, latency, duration), which prior studies using the PSQI did not. Given the highly variable nature of sleep disturbances, this information provides clues for tailoring interventions to address the individual's specific sleep problem. The results of the current study also extend the work of Rao et al,
who found early difficulties in individuals with TBI in sleep initiation, maintenance, and quality, all of which we found at 1 year after injury. These findings suggest that the sleep difficulties are not temporary problems from which the person with TBI recovers. The ability to identify specific types of sleep difficulties may assist with more specific and effective treatment.
Our results highlight the importance of evaluating the sleep quality of individuals with TBI. Similar to prior research, which has shown associations between chronic pain,
we also found that participants with 1 or more of these conditions had more difficulties with sleep. The strong relationship that was found between sleep disturbances and commonly co-occurring psychological disorders is interesting and raises the question of causal direction. Does the sleep disorder cause or contribute to mental health disorders? Do problems with mental health disrupt sleep? Or is there a reciprocal effect of each? These results indicate the need for not only comprehensive early evaluation to detect such conditions, but additional research to better understand the interactions of these conditions with each other and with sleep, which are likely to have implications with respect to the development of interventions and the ultimate success of those interventions.
We did not measure sleep characteristics or parameters during inpatient rehabilitation early after injury. While no studies specifically address the association of sleep and functional outcomes during hospitalization after TBI, it is the observation of most clinicians that sleep disturbances are extremely common and associated with early behavioral disturbances and poorer cognitive functioning. It is interesting that poor sleepers have significantly worse function even at 1 year after injury. While we were able to consecutively enroll patients from inpatient rehabilitation, our results are limited, because we did not include those with TBI who were not admitted for rehabilitation. Future studies should examine sleep disturbances in more representative cases, as well as examine the predictive value of early sleep disturbances on later functioning and the development of mental health disorders.
Although a significant difference was found between racial/ethnic groups in this study, this finding should be interpreted with caution, because the nonwhite groups were very small. It is also difficult to know the precise purpose for which medications were given to this group. Antidepressants and even neuroleptics are often used to treat sleep complaints as well as mental health disorders in this population.
The data gathered through the PSQI are not detailed enough to allow differential diagnosis of sleep disorders. Also, the PSQI relies on self-report of sleep, and, like all self-report measures, is susceptible to inaccuracies because of poor recall or incorrect attribution. Polysomnographic studies have had conflicting results to date on whether self-reports actually correlate with sleep architecture disturbances.
None of these studies to date have examined the presence of co-occurring disorders in the context of TBI. Ideally, improved means of documenting physiologic sleep parameters in a naturalistic setting will be available in the future to augment self-report, allowing for better understanding of the perception of sleep and the interaction with other disorders.
The evaluation of sleep in a clinical setting should not only involve the use of validated sleep measures, which will improve the identification of sleep disorder types (eg, differentiating sleep latency disorders from circadian rhythm disturbances), but should also routinely include evaluations for commonly co-occurring mental health disorders and painful conditions. This might allow for more precise diagnosis and a more parsimonious use of medications, which may have potential detrimental effects on daytime cognitive functioning. A more detailed knowledge of the sleep disorder type would also enable the selection of appropriate nonpharmacologic treatments. For example, cognitive behavioral therapy has been shown to be effective in the treatment of insomnia
In addition, given the relationship between sleep difficulties and functional outcomes, it will be important for clinicians to consider sleep as a possible factor in any decline of function over time.
This study has demonstrated the common occurrence of sleep complaints in those with TBI, and the association of these complaints with other disorders such as depression, anxiety, and pain syndromes in people with TBI. Future studies are necessary to determine the causal relationships among these co-occurring disorders, their differential relationship with functional status, and response to treatment.
Epidemiology of alcohol and medication as aids to sleep in early adulthood.
Supported by the National Institute on Disability and Rehabilitation Research Traumatic Brain Injury Model Systems (grant no. H133A070032 ).
No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.