Archives of Physical Medicine and Rehabilitation
Volume 90, Issue 6 , Pages 939-946, June 2009

Caregivers' Well-Being After Traumatic Brain Injury: A Multicenter Prospective Investigation

  • Jeffrey S. Kreutzer, PhD

      Affiliations

    • Department of Physical Medicine and Rehabilitation, Virginia Commonwealth University, Richmond, VA
    • Corresponding Author InformationReprint requests to Jeffrey S. Kreutzer, PhD, Dept of Physical Medicine and Rehabilitation, Virginia Commonwealth University, Box 980542, Richmond, VA 23298-0542
    • ,
  • Lisa J. Rapport, PhD

      Affiliations

    • Department of Psychology, Wayne State University, Detroit, MI
    • ,
  • Jennifer H. Marwitz, MA

      Affiliations

    • Department of Physical Medicine and Rehabilitation, Virginia Commonwealth University, Richmond, VA
    • ,
  • Cynthia Harrison-Felix, PhD

      Affiliations

    • Craig Hospital, Englewood, CO
    • ,
  • Tessa Hart, PhD

      Affiliations

    • Moss Rehabilitation Research Institute and Department of Rehabilitation Medicine, Jefferson Medical College, Philadelphia, PA
    • ,
  • Mel Glenn, MD

      Affiliations

    • Spaulding Rehabilitation Hospital, Boston, MA
    • ,
  • Flora Hammond, MD

      Affiliations

    • Carolinas Rehabilitation, Charlotte, NC

Article Outline

Abstract 

Kreutzer JS, Rapport LJ, Marwitz JH, Harrison-Felix C, Hart T, Glenn M, Hammond F. Caregivers' well-being after traumatic brain injury: a multicenter prospective investigation.

Objective

To describe frequency and magnitude of caregivers' emotional distress and life satisfaction using standardized assessment procedures; compare distress levels among spouses, parents, and other caregivers; and identify risk factors.

Design

Prospective collaborative cohort study.

Setting

Six Traumatic Brain Injury Model System Centers providing neurotrauma care, rehabilitation, and outpatient follow-up.

Participants

Caregivers (N=273) of patients who were 1, 2, or 5 years postinjury.

Interventions

Acute neurotrauma care, inpatient interdisciplinary brain injury rehabilitation, and postacute services.

Main Outcome Measure

Brief Symptom Inventory-18 Depression, Anxiety, and Somatic dimensions.

Results

Levels of Depression, Anxiety, and Somatic symptoms were equally prevalent, with 1 in 5 caregivers scoring above the cutoff in each area. The proportion of participants with 1, 2, and 3 elevations was 17.9%, 5.5%, and 10.6%, respectively. Conversely, approximately two thirds (65.9%) had no scores exceeding cutoffs. Distress levels among spouses, parents, and other caregivers were comparable. Higher caregiver distress was associated with caring for survivors who had worse functional status, received more supervision, were less satisfied with life, and used alcohol excessively.

Conclusions

Depression, Anxiety, and Somatic symptoms are common among caregivers. Findings substantiate the importance of clinical care systems addressing the needs of caregivers in the long term as well as survivors.

Key Words: Brain injuries, Caregivers, Family, Rehabilitation, Stress, psychological

List of Abbreviations: ANCOVA, analysis of covariance, ANOVA, analysis of variance, BRFSS, Behavioral Risk Factor Surveillance System, BSI-18, Brief Symptom Inventory-18, DRS, Disability Rating Scale, GCS, Glasgow Coma Scale, GOAT, Galveston Orientation and Amnesia Test, GSI, Global Severity Index, PTA, posttraumatic amnesia, SCL-90, Symptom Checklist-90, SRS, Supervision Rating Scale, SWLS, Satisfaction with Life Scale, TBI, traumatic brain injury, TBIMS, Traumatic Brain Injury Model System

 

FAMILY MEMBERS OFTEN assume a primary caregiving role for persons with brain injury after hospital discharge. Concerns have been expressed that rehabilitation professionals focus predominantly on survivors' needs at the expense of family members.1 Increasing concern about family members' welfare has led to expanded clinical research during the last decade. Many studies have focused on functioning of the family system.2, 3, 4, 5, 6, 7, 8 Other studies have focused on caregivers' needs, distress, and burden.9, 10, 11, 12, 13, 14, 15, 16

Recent studies have identified alarmingly high rates of depression ranging from 25% to 61% among persons with TBI.17, 18, 19, 20 Concerns about the emotional well-being of survivors has sparked similar concerns about caregivers. Unfortunately, caregivers' emotional distress has received scant attention. Mauss-Clum and Ryan21 were among the first to investigate family members' emotional distress. Their survey of support group members provided evidence that emotional distress was more common to wives than survivors. Nearly 80% of wives described themselves as depressed, whereas 57% of survivors were described as depressed. Later researchers began studying family members using standardized measures including the SCL-9022 and derivatives. For example, Linn et al23 used the SCL-90 to compare 60 spouses and survivors participating in a weekend retreat. Depression was predominant, with 73% of spouses and 70% of survivors exceeding the scale cutoff. With regard to Anxiety scale scores, 55% of spouses and 50% of survivors exceeded cutoffs. The investigators did not find a relationship between spousal distress and injury severity as measured by PTA and coma duration.

Using the Brief Symptom Inventory,24 Kreutzer et al7, 8 surveyed 62 family members of survivors referred for neuropsychologic services. Overall, 23% of family members exceeded the cutoff on the Depression scale, and 47% met the definition for clinically significant levels of emotional distress (caseness). The percentage of participants who exceeded cutoffs for Anxiety, Somatization, and GSI were 32%, 25%, and 33%, respectively. Spouses showed significantly greater levels of distress than parents. Analysis did not substantiate a relationship between injury severity and emotional distress. Gervasio and Kreutzer25 completed a retrospective study using a convenience sample of spouses and parents. Twenty-two percent exceeded the cutoff on the BSI Depression scale. Less than a third of participants exceeded cutoffs for the Anxiety (32%), Somatization (17%), and GSI (23%). Forty-four percent met the criteria for caseness, with spouses showing significantly greater levels of distress than parents. Injury severity was unrelated to distress levels. Using the National Institute of Mental Health Diagnostic Interview Schedule Revised, Gillen et al26 evaluated 39 mothers and 20 spouses of persons averaging 19 months postinjury. Nearly half met the criteria for a major depressive disorder. No differences between spouses and parents were found, and injury severity and time postinjury were unrelated to diagnosis of depression.

Douglas and Spellacy27 used the Self-Rating Depression Scale28 to assess spouses and parents of survivors. Seventy-seven percent of parents were found to be depressed, while only 47% percent of spouses met the criteria. The investigators identified a relationship between depression and functional status as measured by the DRS.29 In another single-center study, no differences were found between spouses and parents with regard to caseness on the BSI.30 Sixty family members were surveyed and 37% met the criteria for caseness. Ergh et al30 did not find a relationship between injury severity (GCS) or patients' alcohol abuse, and family members' emotional distress. Australian researchers6 found no significant differences between spouses and parents with regard to prevalence of depression or anxiety. The authors assessed 143 caregivers as part of their standard of care, finding 22% of family members were depressed and 25% were anxious as indicated by scores on the Leeds Scales of Anxiety and Depression.31 Anxiety and depression levels were unrelated to measures of injury severity (eg, PTA, GCS) or patients' functional status as measured by the Craig Handicap Assessment and Reporting Technique.32

In summary, researchers have found varying rates of clinically significant emotional distress among caregiving family members, with incidence estimates ranging from 22% to 77%. Variability may be attributable to methodologic factors. For example, prior studies used single-center data collection, and most used small convenience samples (eg, n<100). Comparisons between parents and spouses have yielded equivocal findings. The relationships among caregivers' emotional distress, injury severity, patients' functional status, and demographic factors remains uncertain.

The present study was intended to overcome methodologic limitations of prior studies by using a prospective design, collecting data at multiple research centers across the United States, incorporating standardized distress measures, and using a substantially larger and more representative sample of caregivers. The primary objectives of the investigation were (1) to describe the frequency and magnitude of caregivers' emotional distress with regard to anxiety, depression, and somatic complaints using standardized assessment tools; (2) to characterize the prevalence of poor life satisfaction among caregivers; (3) to compare the emotional distress and life satisfaction levels of parents, spouses, and other caregivers; (4) to examine the relationships among injury-related factors (eg, severity, time postinjury, functional status), socioeconomic factors (eg, survivors' employment, income level), and caregivers' well-being; and (5) to identify factors that distinguish between caregivers with and without abnormally high levels of emotional distress.

Back to Article Outline

Methods 

Participants 

Caregivers were identified by patient participants in the National Institute on Disability and Rehabilitation Research–funded TBIMS program, a collaborative multicenter project initiated in 1987.33 Caregivers included 226 (82.8%) women and 47 men, a total of 273 persons ranging in age from 18 to 99 (mean ± SD, 51.6±14.2) years. The sample was composed of spouses (33.7%), parents (46.9%), and nonspouse/nonparent caregivers (19.4%) who included siblings, adult children, other relatives, romantic partners, and friends of the survivors. Self-reported racial/ethnic identity of the sample was 73.3% white, 23.4% black, 1.5% Hispanic, 1.1% Asian/Pacific Islander, and 0.7% other. Average years of education, mean ± SD, was 13.1±2.5, and most participants (75.5%) were living with the survivor. Seventy-seven percent typically spent time with the survivor every day of the week, and 93.4% spent 2 or more days a week with the survivor. On days spent with the survivor, most caregivers (46.5%) typically spent 1 to 4 hours caring for, supervising, or helping the survivor, whereas 11.0% spent 5 to 8 hours, and 13.9% spent more than 9 hours. The remainder of the group (28.6%) made themselves available to survivors and were identified as caregivers but did not routinely provide supervision or physical care.

Persons with TBI were 28% women and 72% men, ranging in age from 16 to 88 years (mean ± SD, 36.6±16.7) years. Time postinjury ranged from 1 to 5 years, with 38.2% providing data at 1 year, 26.5% at 2 years, and 35.3% at 5 years. Education ranged from less than 8 years to doctoral-level degrees; most of the survivors had completed high school or a General Equivalency Diploma (29.3%) or some college (32%), with 34.2% completing 11 or fewer years of schooling.

Measures 

Data collected from caregivers included demographics, employment status, days a week spent with survivor, hours a day spent caring for survivor, and whether the caregiver had received treatment for psychologic problems in the past year. In addition, caregivers completed the BSI-1834 and the SWLS,35 characterizing themselves.

The BSI-1834 is an 18-item self-report instrument designed to quantify psychologic distress. The instrument is a psychometrically sound, abbreviated version of the SCL-90, Revised.34 Frequency ratings for items in 3 primary symptom dimensions are added to yield scores for Somatization, Depression, and Anxiety. The GSI reflects the sum of scores for the 3 primary symptom dimensions. T scores for each dimension and the GSI are calculated based on community norms. The developers of the BSI-1834 use the term caseness to identify individuals with clinically significant emotional distress levels. Specifically, caseness refers to having either a GSI T score greater than 62, or having 2 or more symptom dimension T scores greater than 60. For the 3 symptom dimensions and the GSI, internal consistency is in the acceptable range, with alpha coefficients ranging from .74 to .89. Test-retest reliability is .90 for the GSI and ranges from .68 to .84 for the symptom dimensions.34

The SWLS is a global measure of life satisfaction consisting of 5 items rated on a 7-point Likert scale.35 The 5 item scores are summed to yield a total score. Internal consistency for the SWLS is good, with alpha coefficients exceeding .80. Likewise, test-retest reliability has been generally acceptable. The measure has been used by others to examine brain injury samples.36, 37 For the purposes of this study, classification of the percent of participants exceeding a normative cutoff was based on data from Pavot and Diener38 using a score of T less than 40. A higher score indicates higher life satisfaction levels.

Information on persons with TBI (eg, age, ethnicity, level of education, employment status, family income) was derived from the TBIMS National Database.39 Length of PTA, an injury severity indicator, was measured as the number of days from injury to emergence from PTA. Emergence from PTA was determined using the GOAT. The GOAT has demonstrated validity as an instrument for assessing cognition serially during the subacute stage of recovery from brain injury.40 The TBIMS National Database determines PTA emergence when 2 consecutive GOAT scores of 76 or greater occur within a period of 24 to 72 hours.

Functional status of persons with TBI was measured using the FIM instrument41 and the DRS.29 The FIM is made up of 18 items that measure the level of independent functional performance for self-care, transfers, sphincter control, locomotion, communication, and social cognition.41 Each item has a score range from 1 (totally dependent) to 7 (totally independent); thus, a total score ranges from 18 to 126. A higher FIM score reflects greater functional independence. The FIM also includes 2 subscale scores: a cognitive score that includes 5 items, and a motor score that includes 13 items. Numerous studies have shown clinically appropriate validity and reliability for the FIM, with median subscale reliability scores ranging from .78 to .95 and median interrater reliability of .95 for the total FIM.42, 43, 44, 45

The DRS was developed to track functional changes over time (eg, from coma to community).29 Items address impairment, activity limitation, and participation, with a total score ranging from 0 (no disability) to 30 (death), with higher scores signifying greater disability. Interrater reliability of the DRS has been demonstrated with Pearson correlations of .97 to .98.29 Test-retest reliability was established with a Spearman rho correlation of .95.46

The SRS47 was used to measure the level of supervision that caregivers provided to the person with TBI. Ratings were based on the level of supervision received, not on how much supervision a person was judged or predicted to need. Scores range from 1 (independent) to 13 (full-time direct supervision, the person is in restraints). Prior research has demonstrated the interrater reliability of the SRS (intraclass correlation=.86; weighted κ=.64).47

The TBI National Database was the source of information on survivors' SWLS score and alcohol use. Drinking patterns were assessed using the 4 alcohol use questions from the Centers for Disease Control's BRFSS core survey.48 Recent research has shown that the BRFSS provides findings comparable to those obtained using the National Health Interview Survey.49

Procedure 

Six TBIMS centers designed and participated in the present investigation, with local institutional review boards approving each center's operations. Each of the TBIMS centers included emergency medical services, intensive and acute medical care, inpatient rehabilitation, and a spectrum of community rehabilitation services. An individualized, comprehensive program of adult (ie, age >16y) inpatient rehabilitation was provided to each patient. The following services were provided at each center: nursing, occupational therapy, physiatry and related medical services, physical therapy, psychology and neuropsychology, therapeutic recreation, social services, and speech and language therapy. Information about medical aspects of injury was obtained from hospital records. Admission FIM was scored within 72 hours of rehabilitation admission and discharge FIM was scored within 72 hours of discharge by the interdisciplinary team members using standard protocols.50

TBIMS follow-up data collection involved phone interviews with participants at 1, 2, and 5 years postinjury, and follow-up rates were approximately 89%. At the time of routinely scheduled follow-up interviews, the participant was asked to identify the person best described as the caregiver. The identified person was then approached for consent to participate. Informed consent was obtained from both the person with TBI and the identified caregiver. Project staff estimated that fewer than 5% of caregivers declined participation. Data were collected primarily via telephone interviews between June 2004 and December 2005 from a cross-sectional sample of the TBIMS. Patients without identified caregivers were not included in the study.

Data Analysis 

Descriptive statistics including means, SDs, and percentages were computed for all relevant variables. Frequency and magnitude of emotional distress and life satisfaction were evaluated by comparing the sample to appropriate normative data for the BSI-18 and SWLS. Relationships between caregiver well-being and survivor level of functioning were evaluated using ANOVA for categorical variables and simple linear regression for continuous variables. Variables that showed significance at alpha equal to .10 were then included in the final adjusted analyses to examine better the combined effect of predictors of overall emotional distress (BSI-GSI) and satisfaction with life (SWLS). ANOVA, Mann-Whitney, and chi-square tests were used to examine characteristic differences between caregivers experiencing acute emotional distress and those not in distress.

Back to Article Outline

Results 

Frequency/Magnitude of Emotional Distress and Low Satisfaction With Life 

Descriptive and clinical classification statistics for the BSI-18 and the SWLS are presented in table 1. Mean T scores for the Depression, Somatization, and Anxiety symptom dimensions were relatively similar, ranging from 49.1 to 50.4. The proportion of caregivers exceeding BSI-18 clinical cutoffs ranged from 17.6% to 23.8%. Many persons exceeding cutoffs were receiving or had received counseling or pharmacologic treatment for distress within the past year (45.5%–53.2%).

Table 1. Well-Being of Caregivers of Persons With TBI
Well-Being IndexT Score (Mean ± SD)RangeExceeding CutoffExceeding Cutoff Who Received Treatment
BSI-18
BSI Depression⁎49.1±9.440–8119.0(53)50.9(27/53)
BSI Somatization⁎50.4±10.541–7923.8(66)45.5(30/66)
BSI Anxiety⁎49.3±10.138–8117.9(50)46.0(23/50)
BSI-GSI⁎49.5±10.833–8117.6(47)53.2(25/47)
BSI caseness† 17.6(47)53.2(25/47)
SWLS‡49.9±11.121–6722.1(61)54.1(33/61)

NOTE. Values are % (n) or as otherwise indicated.

⁎BSI-18 Depression, Somatization, Anxiety, and GSI subscales greater than 60 T.

†BSI-18 caseness is equal to 2 subscales greater than 60 T or BSI-GSI greater than 62 T.

‡Classification based on normative data from Pavot and Diener.38 Cutoff equals T less than 40.

To gauge the breadth of caregivers' emotional distress, the number of symptom dimensions on which BSI-18 cutoffs were exceeded was calculated for each participant. The proportion of participants with 1, 2, and 3 elevations was 17.9%, 5.5%, and 10.6%, respectively. Conversely, approximately two thirds (65.9%) had no scores exceeding cutoffs.

With regard to the SWLS, the sample mean (mean ± SD, 23.6±7.1) is very similar to normative data for nonclinical samples (mean ± SD, 23.7±6.4).38 The mean SWLS for the sample corresponds to an average item endorsement indicating “neither agree nor disagree” to “slightly agree” (mean ± SD, 4.7±1.4) regarding the positive statements about life satisfaction. Additionally, an independent t test indicated that caregivers endorsed higher levels of life satisfaction than did the survivors for whom they care (mean ± SD, 21.9±8.1; t473=2.19; P=.015).

Comparisons were made among spouses, parents, and other caregivers. ANOVAs indicated that spouses, parents, and nonfamily caregivers did not differ on any of the BSI-18 symptom dimensions or the SWLS (P>.30 for all comparisons). Furthermore, the percent of caregivers exceeding the cutoff for clinically significant depressive symptoms (BSI-Depression) did not differ significantly for spouses (21.7%), parents (20.3%), and other caregivers (11.3%; χ21,N=273=2.62; P=.270). Similar results were observed with regard to Anxiety, Somatization, and BSI caseness (P>.44 for all comparisons).

BSI-18 symptom dimension scores obtained at 1, 2, and 5 years postinjury were compared and were found to be very similar. For example, mean Depression T scores ranged from 48.3 to 49.7, Anxiety scores ranged from 48.2 to 49.9, and Somatic scores ranged from 49.8 to 50.9. As expected, ANOVAs did not indicate time-related differences (P>.43 for all comparisons). Scores on the SWLS obtained at different times postinjury were similar as well (P=.138), ranging from 22.1 to 24.2.

Relationships Between Caregiver Well-Being and Survivor Level of Functioning 

Simple linear regressions were completed to identify survivor characteristics predictive of caregiver well-being (BSI-GSI, SWLS). The number of hours caregivers spent caring for/supervising the survivor, as well as survivors' DRS, FIM total, SRS, and SWLS, were analyzed. As shown in table 2, all variables examined showed a significant effect on caregiver SWLS, while all but FIM total showed a significant effect on BSI-GSI.

Table 2. Effects of Predictors on Caregiver SWLS and BSI-GSI
Survivor Characteristics at Follow-UpCaregiver SWLSCaregiver BSI-GSI
Unadjusted EffectsAdjusted EffectsUnadjusted EffectsAdjusted Effects
Continuous PredictorsSlopeSEPSlopeSEPSlopeSEPSlopeSEP
DRS−0.6840.149<0.001−1.7280.486<0.0010.6630.2440.0070.2620.7840.739
FIM total0.0750.0270.006−0.1100.0740.1410.0740.0440.0970.1330.1210.273
SRS−0.4600.1920.0170.5490.3350.1030.8060.2950.0070.7130.5480.195
SWLS0.3240.053<0.0010.2450.060<0.001−0.4420.084<0.001−0.3750.097<0.001
Hours spent care/supervision−1.4240.289<0.0010.0580.3870.8811.8470.466<0.0010.5910.6320.351
Categorical PredictorsMean ± SEPMean ± SEPMean ± SEPMean ± SEP
Employment status <0.001 0.198 0.008 0.661
Competitively employed25.607*±0.751 23.745±0.802 47.512*±1.166 48.768±1.321
Unpaid vocational role25.355*±1.236 24.132±1.193 46.903*†±1.920 49.214±1.922
Retired23.212*±0.692 24.601±0.817 50.181†±1.103† 49.227±1.344
Unemployed20.078†±0.964 21.856±1.037 53.460†±1.512† 51.392±1.677

NOTE. Means with the same superscript (* or †) do not differ at P<.05.

Vocational status of survivors was significantly associated with caregiver well-being. ANOVA tests examined group differences on SWLS and BSI-GSI for caregivers of survivors who were competitively employed (n=84), retired (n=99), or unemployed (n=51), and a group of survivors who identified specific but unpaid vocational roles (n=31; eg, identified themselves as students, primary house/family care, with sheltered workshop or volunteer work). As shown in the unadjusted effects columns of table 2, caregivers of survivors who were employed had significantly better BSI-GSI scores than retired and unemployed survivors, and caregivers of survivors in unpaid vocational roles had better BSI-GSI scores than unemployed survivors.

Prediction of Caregiver Satisfaction With Life and Overall Emotional Distress 

To examine caregiver SWLS better, an ANCOVA model was fit including effects for variables that were significant in univariate analyses (number of hours caregivers spent caring for the survivor, DRS, FIM total, SRS, survivors' SWLS, survivors' employment status). This model accounted for 24% of the variations in caregiver SWLS (F8,214=8.47; P<.001). DRS (P=.001) and survivor SWLS (P<.001) were significant predictors of caregiver SWLS, while number of hours caring, FIM total, SRS, and survivor employment status were not (see adjusted column in table 2). A 1-unit increase in DRS was associated with a 1.7-unit (SE=.06) decrease in caregiver SWLS. A 1-unit increase in survivor SWLS was associated with a 0.25-unit (SE=.06) increase in caregiver SWLS.

A second ANCOVA model was fit to examine overall caregiver distress (BSI-GSI). This model accounted for 16% of the variations in caregiver BSI-GSI (F8,211=4.85; P<.001). Survivor SWLS (P=.001) was a significant predictor of BSI-GSI, while number of hours caring, DRS, FIM total, SRS, and survivor employment status were not (see adjusted column in table 2). A 1-unit increase in survivor SWLS was associated with a 0.38-unit (SE=.10) decrease in caregiver BSI-GSI scores. Residual diagnostics indicated that model assumptions, including normality of the residuals, were adequate.

Characteristics of Caregivers Who Endorsed Significant BSI-18 Symptoms 

ANOVA, Mann-Whitney, and chi-square tests examined differences between caregivers experiencing acute emotional distress (meeting the definition for caseness on the BSI-18, n=46) and those not in distress. As can be seen in table 3, caregivers experiencing significant levels of emotional distress were caring for survivors with lower levels of functioning (DRS, FIM), who received more supervision (SRS, hours spent caregiving), who reported lower satisfaction with life (SWLS), and more frequently used alcohol excessively. With regard to acute emotional distress, no difference was found between caregivers who were living with and not living with survivors.

Table 3. Caregiver Emotional Distress Group and Characteristics of Survivors
CharacteristicsNo BSI CasenessBSI CasenessTotaldfFP
Length of PTA (d)⁎30.3±35.833.4±26.430.8±34.511920.61.436
DRS (discharge)⁎6.3±3.67.0±3.56.5±3.612701.29.257
DRS (follow-up)⁎2.1±2.73.1±3.12.3±2.812676.44.012
FIM (follow-up)⁎115.2±15.3111.0±18.0114.5±15.912675.38.021
SRS (follow-up)⁎2.6±2.63.7±2.82.8±2.612675.73.017
SWLS (follow-up)22.7±8.017.4±7.121.9±8.1122813.52<.001
h/d caregiving (follow-up)†‡2.2±1.32.9±1.82.3±1.4 −2.95.003
Living with survivor (follow-up)§ (%)77.065.9 2.31.138
Any use of alcohol (follow-up)§ (%)36.945.7 1.22.269
Excessive use of alcohol (follow-up)§ (%)12.327.3 6.15.011

NOTE. Values are mean ± SD or as otherwise indicated.

⁎Square root or log-transformation used in parametric analyses.

†Categorical index: 1=0h/d; 2=1–4h/d, 3=5–8h/d; 4=9–12h/d.

‡Mann-Whitney U test, 2-tailed.

§Chi-square test, df=1.

Differences between depressed and nondepressed caregivers (T>60) were examined. Results indicate that caregivers reporting clinically significant levels of depression (n=52) were caring for survivors with lower levels of functioning (DRS, F1,267=7.52, P=.007; FIM, F1,269=8.62, P=.004), who received more supervision (SRS, F1,267=7.52, P=.007; hours spent caregiving, Mann-Whitney U test, 2-tailed=–3.09, P=.002), and who reported lower satisfaction with life (F1,228=17.23; P<.001). Additionally, caregivers reporting clinically significant levels of depression were caring for survivors who were more likely to use alcohol excessively (27.5%) than were survivors of caregivers who did not report significant levels of depression (11.8%; χ2=7.98; df=1; 2-tailed P=.005).

Similar comparisons were made for caregivers with clinically significant anxiety (n=50) and those without on the BSI-18. Compared with caregivers who did not endorse clinically significant levels of anxiety, anxious caregivers were caring for survivors with lower levels of functioning (DRS, F1,267=5.10, P=.025; FIM, F1,267=4.00, P=.047) who received more supervision (SRS, F1,267=5.71; P=.018) and more caregiving (Mann-Whitney U test, 2-tailed=–3.81; P<.001). Anxious caregivers provided care for survivors who more often used alcohol excessively (χ2=5.19; df=1; 2-tailed P=.026) and reported lower satisfaction with life (F1,228=12.32; P=.001).

Finally, comparisons were made between caregivers scoring high on the Somatization scale and those scoring within normal limits. Caregivers reporting clinically significant somatic symptoms (n=63) cared for survivors with lower levels of functioning on the DRS (F1,267=7.06; P=.008), who received more supervision (SRS, F1,267=5.84; P=.021), and who reported lower satisfaction with life (F1,228=8.40; P=.004).

Back to Article Outline

Discussion 

Multicenter, prospective data derived from the present investigation substantiate concerns about caregivers' emotional well-being.7, 8, 23, 26 Nearly a quarter of caregivers reported problematic levels of somatic symptoms related to emotional stress. A similar number were characterized by high anxiety levels. About 20% exceeded cutoff scores for depressive symptomatology. The prevalence of depression was modestly higher than expected in the general population.51 Still, depression prevalence rates were lower in our study than in several others. For example, Mintz et al52 and Gillen et al26 reported a prevalence rate of nearly 50%. The most obvious reason for differential findings is that our sample was selected from a prospectively followed cohort, rather than recruited from clinical case lists, which are more likely to draw distressed participants. Previous studies based on prospective samples have found emotional distress rates more comparable to our sample (eg, 37% in Ergh et al30).

Our findings suggest that emotional distress is 1-dimensional for some caregivers (17.9%) and multidimensional for others. In fact, one third of the caregivers exceeded clinical cutoffs in at least 1 symptom dimension, and 1 in 10 exceeded cutoffs in all 3 dimensions. As a benchmark, note that the prevalences of mood and anxiety disorders in the general population are 9.5% and 18.1%, respectively.51 On the other hand, our findings also suggest that a substantial number of caregivers cope adequately with their situation. Nearly two thirds did not exceed cutoffs in any symptom dimension. Further research is needed to improve our appreciation of adaptive coping strategies.

Relationship of the caregiver to the survivor has been a focus of interest in prior investigations and an area of inquiry in the present investigation. We sought to clarify the conflicting findings in previous literature stating variously that spouses were more distressed than parents,7, 8, 25 parents were more distressed than spouses,27 or kinship made no difference.6, 26, 30 Our findings are consistent with the last group of studies showing that the caregivers' relationship to the survivor was not a significant predictor of emotional distress. These data suggest that classifying relationships based on simple categories is insufficient to characterize the nature of interactions between individuals.

Some researchers have found increases in caregiver distress over time,53, 54, 55, 56 especially in cases in which social support is inadequate.30 Other researchers have found little evidence of change over time.25, 26 For the present investigation, scores on emotional distress and life satisfaction measures were remarkably similar at the 3 follow-up intervals. The findings provide support for contentions that distress levels remain high for years postinjury, making a case for the availability of long-term supports. The present investigation used a cross-sectional design, limiting the extent to which change could be fully appreciated. Future studies should incorporate longitudinal methods to ascertain better the course of change. For some family members, distress levels may increase, and for others, distress levels may decrease. Identifying caregiver and survivor characteristics that mediate distress levels can facilitate the development of effective interventions.

What did predict emotional dysfunction to a highly significant degree was the functional status of the survivor at the time of caregiver evaluation. Previous studies have been equivocal: some have found a relationship with survivors' level of disability27 and others have not.6 More consistent evidence has identified the importance of survivors' cognitive and behavioral functioning rather than their overall functional limitation.6, 30 In the current study, caregivers' emotional function and distress were inversely related to survivors' functional status on multiple indices of activity, participation, and employment. For example, the caregivers of survivors engaged in competitive or volunteer work were less likely to report emotional distress. Factors leading to diminished productivity may be important contributors to caregivers' distress. With regard to psychosocial factors, increased caregiver distress was associated with excessive alcohol use by survivors.

Study Limitations 

Limitations of the present investigation include the select sample, reliance on self-report telephonic data collection methods, and the potential for response bias. A select group was examined: caregivers of persons with moderate and severe TBI who received inpatient rehabilitation. Findings of emotional well-being may be different among persons with mild TBI, or those who are admitted to acute care but do not receive inpatient rehabilitation. Representativeness of the sample might be further limited by the fact that only 6 of the 16 TBIMS centers participated in the project. However, data were collected from centers in the western, southern, eastern, midwestern, and northeastern United States, and the sample included a substantial proportion of nonwhite caregivers. Emotional distress was measured using self-report questionnaires, and there is a possibility of response bias. Caregivers may have chosen to downplay emotional distress, desiring to appear well for the survivor's benefit. Underreporting distress would lead to underestimates of distress. Future studies should incorporate a combination of standardized measures, in-person interviews, and observational methods, allowing for integration of different perspectives. Additionally, future studies may help us better identify factors mediating distress by including measures of caregiver health, perceived burden, and ways of coping.

The present findings suggest that clinical care systems should address the emotional needs of caregivers. Clinicians have recently been encouraged to develop emotional distress screening and treatment systems for survivors.18 Caregivers are likely to benefit from the development of similar systems to meet their individual needs. Most clinical programs have limited resources and must wisely allocate care. Findings of the present investigation suggest that caregivers of persons with the most severe disabilities and persons who use alcohol excessively are at greatest risk for emotional distress. Spouses, parents, and other caregivers seem equally at risk. Unfortunately, caregivers may be reluctant to discuss their apprehensions in the presence of survivors. Telephone contacts may allow them opportunities to disclose their concerns more comfortably. Distressed caregivers should be offered educational materials and referrals including counseling, support groups, medications, and respite.

Rehabilitation programs that address survivors' level of function, alcohol problems, and productive activity in the long term have the potential to benefit their caregivers as well. Employment, school, or volunteer activity all may provide not only satisfaction for the participant but also reduced stress for the caregiver. Help from a personal care assistant may reduce caregiver distress. Day recreational programs with peers of similar age can provide satisfaction for the person with TBI while greatly diminishing caregiver supervision time.

Back to Article Outline

Conclusions 

Although most caregivers appear to cope adequately, the present findings suggest that one third are at risk for depression, anxiety, or other forms of psychologic distress. Those who care for survivors who are more disabled, who are unemployed or otherwise uninvolved in productive activity, or who have problems with alcohol abuse are at greatest risk. Greater hours spent in caregiving are another risk factor. Clinicians are encouraged to seek out caregivers to assess their level of distress and mitigate their symptoms. Those who are suffering should be offered counseling and support groups, or be encouraged to seek out community supports such as religious institutions and friends. A better appreciation of prevalence and risk factors will emerge from research relying on multimethod assessment approaches.

Back to Article Outline

References 

  1. Boschen K, Gargaro J, Gan C, Gerber G, Brandys C. Family interventions after acquired brain injury and other chronic conditions: a critical appraisal of the quality of the evidence. NeuroRehabil. 2007;22:19–41
  2. Gan C, Campbell KA, Gemeinhardt M, McFadden GT. Predictors of family system functioning after brain injury. Brain Inj. 2006;20:587–600
  3. Sander AM, Caroselli JS, High WM, Becker C, Neese L, Scheibel R. Relationship of family functioning to progress in a post-acute rehabilitation programme following traumatic brain injury. Brain Inj. 2002;16:649–657
  4. Groom KN, Shaw TG, O'Connor ME, Howard NI, Pickens A. Neurobehavioral symptoms and family functioning in traumatically brain-injured adults. Arch Clin Neuropsychol. 1998;13:695–711
  5. Douglas JM, Spellacy FJ. Indicators of long-term family functioning following severe traumatic brain injury in adults. Brain Inj. 1996;10:819–839
  6. Ponsford J, Olver J, Ponsford M, Nelms R. Long-term adjustment of families following traumatic brain injury where comprehensive rehabilitation has been provided. Brain Inj. 2003;17:453–568
  7. Kreutzer JS, Gervasio AH, Camplair PS. Patient correlates of caregivers' distress and family functioning after traumatic brain injury. Brain Inj. 1994;8:211–230
  8. Kreutzer JS, Gervasio AH, Camplair PS. Primary caregivers' psychological status and family functioning after traumatic brain injury. Brain Inj. 1994;8:197–210
  9. Wade SL, Stancin T, Taylor HG, Drotar D, Yeates KO, Minich NM. Interpersonal stressors and resources as predictors of parental adaptation following pediatric traumatic injury. J Consult Clin Psychol. 2004;72:776–784
  10. Wade SL, Taylor HG, Yeates KO, et al. Long-term parental and family adaptation following pediatric brain injury. J Pediatr Psychol. 2006;31:1072–1083
  11. Campbell CH. Needs of relatives and helpfulness of support groups in severe head injury. Rehabil Nurs. 1988;13:320–325
  12. Engli M, Kirsivali-Farmer K. Needs of family members of critically ill patients with and without acute brain injury. J Neurosci Nurs. 1993;25:78–85
  13. Kreutzer JS, Devany C, Keck S. Family needs following brain injury: a quantitative analysis. J Head Trauma Rehabil. 1995;9:104–115
  14. Moules S, Chandler BJ. A study of the health and social needs of carers of traumatically brain injured individuals served by one community rehabilitation team. Brain Inj. 1999;13:983–993
  15. Marks JP, Daggett LM. A critical pathway for meeting the needs of families of patients with severe traumatic brain injury. J Neurosci Nurs. 2006;38:84–89
  16. Armstrong K, Kerns KA. The assessment of parent needs following paediatric traumatic brain injury. Pediatr Rehabil. 2002;5:149–160
  17. Hibbard MR, Uysal S, Kepler K, Bogdany J, Silver J. Axis I psychopathology in individuals with traumatic brain injury. J Head Trauma Rehabil. 1998;13:24–39
  18. Fann JR, Burington B, Leonetti A, Jaffe K, Katon WJ, Thompson RS. Psychiatric illness following traumatic brain injury in an adult health maintenance organization population. Arch Gen Psychiatry. 2004;61:53–61
  19. Kreutzer JS, Seel RT, Gourley E. The prevalence and symptom rates of depression after traumatic brain injury: a comprehensive examination. Brain Inj. 2001;15:563–576
  20. Seel RT, Kreutzer JS, Rosenthal M, Hammond FM, Corrigan JD, Black K. Depression after traumatic brain injury: a National Institute on Disability and Rehabilitation Research Model Systems multicenter investigation. Arch Phys Med Rehabil. 2003;84:177–184
  21. Mauss-Clum N, Ryan M. Brain injury and the family. J Neurosurg Nurs. 1981;13:165–169
  22. Derogatis LR. SCL-90 administration, scoring and procedures manual - II for the revised version and other instruments of the psychopathology rating scale series. Towson: Clinical Psychometric Research; 1983;
  23. Linn RT, Allen K, Willer BS. Affective symptoms in the chronic stage of traumatic brain injury: a study of married couples. Brain Inj. 1994;8:135–147
  24. Derogatis LR, Melisaratos N. The Brief Symptom Inventory: an introductory report. Psychol Med. 1983;13:595–605
  25. Gervasio AH, Kreutzer JS. Kinship and family members' psychological distress after traumatic brain injury: a large sample study. J Head Trauma Rehabil. 1997;12:14–26
  26. Gillen R, Tennen H, Affleck G, Steinpreis R. Distress, depressive symptoms, and depressive disorder among caregivers of patients with brain injury. J Head Trauma Rehabil. 1998;13:31–43
  27. Douglas JM, Spellacy FJ. Correlates of depression in adults with severe traumatic brain injury and their carers. Brain Inj. 2000;14:71–88
  28. Zung WW. A Self-Rating Depression Scale. Arch Gen Psychiatry. 1965;12:63–70
  29. Rappaport M, Hall KM, Hopkins K, Belleza T, Cope DN. Disability rating scale for severe head trauma: coma to community. Arch Phys Med Rehabil. 1982;63:118–123
  30. Ergh TC, Rapport LJ, Coleman RD, Hanks RA. Predictors of caregiver and family functioning following traumatic brain injury: social support moderates caregiver distress. J Head Trauma Rehabil. 2002;17:155–174
  31. Snaith RP, Bridge GW, Hamilton M. The Leeds scales for the self assessment of anxiety and depression. London: Psychological Test Publications; 1976;
  32. Whiteneck GG, Charlifue SW, Gerhart KA, Overholser JD, Richardson GN. Quantifying handicap: a new measure of long-term rehabilitation outcomes. Arch Phys Med Rehabil. 1992;73:519–526
  33. Harrison-Felix C, Newton CN, Hall KM, Kreutzer JS. Descriptive findings from the Traumatic Brain Injury Model Systems National Database. J Head Trauma Rehabil. 1996;11:1–14
  34. Derogatis LR. BSI 18: Brief Symptom Inventory 18. Minneapolis: NCS Pearson; 2000;
  35. Diener E, Emmons RA, Larsen RJ, Griffin S. The Satisfaction With Life Scale. J Pers Assess. 1985;49:71–75
  36. Hart T, O'Neil-Pirozzi TM, Williams KD, Rapport LJ, Hammond F, Kreutzer J. Racial differences in caregiving patterns, caregiver emotional function, and sources of emotional support following traumatic brain injury. J Head Trauma Rehabil. 2007;22:122–131
  37. Corrigan JD, Bogner JA, Mysiw WJ, Clinchot D, Fugate L. Life satisfaction after traumatic brain injury. J Head Trauma Rehabil. 2001;16:543–555
  38. Pavot L, Diener E. Review of the Satisfaction With Life Scale. Psychol Assess. 1993;5:164–172
  39. Traumatic Brain Injury Model Systems National Database Syllabus. Englewood: Craig Hospital, National Database and Statistical Center; 2006;
  40. Levin HS, O'Donnell VM, Grossman RG. The Galveston Orientation and Amnesia Test: a practical scale to assess cognition after head injury. J Nerv Ment Dis. 1979;167:675–684
  41. Guide for the Uniform Data Set for Medical Rehabilitation (including the FIM instrument), Version 5.1. Buffalo: State University of New York at Buffalo; 1997;
  42. Hamilton BB, Laughlin JA, Fiedler RC, Granger CV. Interrater reliability of the 7-level functional independence measure (FIM). Scand J Rehabil Med. 1994;26:115–119
  43. Granger CV. The emerging science of functional assessment: our tool for outcomes analysis. Arch Phys Med Rehabil. 1998;79:235–240
  44. Heinemann AW, Linacre JM, Wright BD, Hamilton BB, Granger C. Relationships between impairment and physical disability as measured by the functional independence measure. Arch Phys Med Rehabil. 1993;74:566–573
  45. Ottenbacher KJ, Hsu Y, Granger CV, Fiedler RC. The reliability of the functional independence measure: a quantitative review. Arch Phys Med Rehabil. 1996;77:1226–1232
  46. Gouvier WD, Blanton PD, LaPorte KK, Nepomuceno C. Reliability and validity of the Disability Rating Scale and the Levels of Cognitive Functioning Scale in monitoring recovery from severe head injury. Arch Phys Med Rehabil. 1987;68:94–97
  47. Boake C. Supervision rating scale: a measure of functional outcome from brain injury. Arch Phys Med Rehabil. 1996;77:765–772
  48. Centers for Disease Control and Prevention: Behavioral Risk Factor Surveillance System user's guide. Atlanta: Centers for Disease Control and Prevention; 1998;
  49. Nelson DE, Powell-Griner E, Town M, Kovar MG. A comparison of national estimates from the National Health Interview Survey and the Behavioral Risk Factor Surveillance System. Am J Public Health. 2003;93:1335–1341
  50. Dahmer ER, Shilling MA, Hamilton BB, et al. A model systems database for traumatic brain injury. J Head Trauma Rehabil. 1993;8:12–25
  51. Kessler RC, Chiu WT, Demler O, Merikangas KR, Walters EE. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:617–627
  52. Mintz MC, van Horn KR, Levine MJ. Developmental models of social cognition in assessing the role of family stress in relatives' predictions following traumatic brain injury. Brain Inj. 1995;9:173–186
  53. Livingston MG, Brooks DN, Bond MR. Patient outcome in the year following severe head injury and relatives' psychiatric and social functioning. J Neurol Neurosurg Psychiatry. 1985;48:876–881
  54. Livingston MG, Brooks DN, Bond MR. Three months after severe head injury: psychiatric and social impact on relatives. J Neurol Neurosurg Psychiatry. 1985;48:870–875
  55. Winstanley J, Simpson G, Tate R, Myles B. Early indicators and contributors to psychological distress in relatives during rehabilitation following severe traumatic brain injury: findings from the Brain Injury Outcomes Study. J Head Trauma Rehabil. 2006;21:453–466
  56. Novack T, Bergquist TF, Bennett G, Gouvier WD. Primary caregiver distress following severe head injury. J Head Trauma Rehabil. 1992;6:69–77

 Supported by the National Institute on Disability and Rehabilitation Research, US Department of Education (grant nos. H133A020516, H133A020509, H133A020510, H133A020513).

 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.

PII: S0003-9993(09)00202-0

doi:10.1016/j.apmr.2009.01.010

Archives of Physical Medicine and Rehabilitation
Volume 90, Issue 6 , Pages 939-946, June 2009

Article Tools