psychiatric disorders following tbi

7/28/2019 Psychiatric Disorders Following TBI

1/9

J Head Trauma Rehabil

Vol. 24, No. 5, pp. 324332Copyright c 2009 Wolters Kluwer Health | Lippincott Williams & Wilkins

Psychiatric Disorders Following

Traumatic Brain Injury: Their Nature

and FrequencyRochelle Whelan-Goodinson, DPsych; Jennie Ponsford, PhD ; Lisa Johnston, PhD;

Fiona Grant, PhD

Objectives: To retrospectively establish the nature and frequency of Axis I psychiatric disorders pre- and post-TBI.Participants: One hundred participants who were 0.5 to 5.5 years post mild to severe TBI and 87 informants,each evaluated at a single time point. Main Measure: The Structured Clinical Interview for DSM-IV Disorders(SCID-I). Results: Preinjury, 52% received a psychiatric diagnosis, most commonly substance use disorder (41%),followed by major depressive disorder (17%) and anxiety (13%). Postinjury, 65% received a diagnosis, of which

major depression became the most common (45%), followed by anxiety (38%) and substance use disorder (21%).Frequency of depression, generalized anxiety disorder, posttraumatic stress disorder, panic disorder, and phobiasrose from preinjury to postinjury. More than two-thirds of postinjury depression and anxiety cases were novel andshowed poor resolution rates. Few novel cases of substance use disorder were noted. Psychotic disorders, somatoformdisorders, and eating disorders occurred at frequencies similar to those in the general population.Conclusions: Ahigh frequency of postinjury psychiatric disorders was evident up to 5.5 years postinjury, with many novel cases ofdepression and anxiety. Individuals with TBI should be screened for psychiatric disorders at various time points post-injury without reliance on history of psychiatric problems to predict who is at risk, so that appropriate interventioncan be offered. Keywords: anxiety, brain injuries, depression, psychotic disorders

TRAUMATIC BRAIN INJURIES (TBI) are amongthe leading causes of death and disability in indi-viduals under the age of 45 years, most commonly youngmales.1 Brain injury is often diffuse and bilateral, com-monly including frontotemporal regions, limbic system,basal ganglia, and hippocampus and causing cognitive,behavioral, and emotional changes. Such changes dis-rupt the lives of these young people, affecting their abil-ity to establish independence, a vocation, and relation-ships, potentially leading to loss of self-esteem.24

A proportion of those with TBI develop psychiatricproblems postinjury. Variable frequencies of psychiatricdisorders have been reported at various time points fol-lowing TBI, those for depression and anxiety ranging

from 14% to 77%,2,513 and for substance use from4.9% to 28%.2,7,10,11,14 High rates of current depres-sion and anxiety have also been reported, from 10% to46%,2,7,10,11 as have high rates of current substance use

Author Affiliations: School of Psychology, Psychiatry and PsychologicalMedicine, Monash University, Melbourne, Australia (Drs Whelan-Goodinson and Ponsford and Ms Grant), Monash-Epworth Rehabilitation

Research Centre, Epworth Hospital, Melbourne, Australia (Drs Ponsfordand Johnston), and National Trauma Research Institute, Melbourne,

Australia (Dr Ponsford).

Corresponding Author: Jennie Ponsford, PhD, Department of Psychology,

Monash University, Clayton, Victoria 3800, Australia ([email protected]).

disorders, from 8% to 18%.11,15 In the Australian generalpopulation, prevalence rates for depression and anxiety

over a 12 month interval are approximately 5.7% and9.7%, respectively, and 2.2% to 3.5% for substance usedisorders.16 Comorbidity between anxiety and depres-sion is high; 1 in 3 people with an anxiety disorder alsohas an affective disorder. Few TBI studies have examineda range of psychiatric disorders, most having focusedon depression.2,7,10,14 Two studies have attempted to as-sess Axis II personality disorders in TBI participants.7,17

However, brain injury is an exclusionary criterion forDiagnostic and Statistical Manual-IV (DSM-IV) diagno-sis of a personality disorder. Furthermore, certain per-sonality changes may be symptoms of frontal lobe

injury.Some studies2,7,10,11,1820 have identified high fre-

quencies of preinjury psychiatric disorders, ranging from18% to 51%; however, others have excluded people witha preinjury psychiatric diagnosis.9,2123 Only 3 knownstudies have used semistructured or structured clini-cal interviews to establish preinjury psychiatric diag-noses, and all found higher rates of such diagnosespostinjury.7,10,11 Participants in these studies were inter-viewed at wide-ranging follow-up intervalson average3, 8, and 48 years postinjury, respectively. Retrospective

recall of preinjury psychiatric symptoms may be unre-liable at such long times after injury, particularly given

324


2/9

Psychiatric Disorders Following Traumatic Brain Injury 325

the likely presence of cognitive impairments. Arguably,significant others should be consulted to improve relia-bility, which these studies appear not to have done.

The variability in reported frequencies of preinjuryand postinjury disorders may be related to variable tim-ing of assessment, often within the same study. Studieshave included participants with injuries sustained from

1 to 227 days,23 1 to 37 years,10 and 27 to 48 years7previously, while other studies have focused only withinthe first year postinjury.2,13,24,25 Jorge12 studied symp-toms of depression between 1 and 12 months postinjuryand found that for 40% of those who were initially de-pressed, depression resolved within the year, while 18%of those not depressed at initial interview had devel-oped depression by 1 year. They suggested that acuteonset depression may be associated with injury-relatedbiological changes in the brain, whereas delayed onsetdepression may be associated with a growing awarenessof injury-related disability. Studies conducted over only1 year after injury arguably do not fully capture long-term emotional issues.

Ashman14 found that the incidence of mood and anx-iety disorders rose from preinjury to the first year postin-jury and was then relatively stable 2 and 3 years postin-jury. Two other studies have shown psychotic disordersto have an average latency to onset of 41/2 years fol-lowing TBI26,27 although the method of diagnosis wasunclear in these studies. Studies of post-TBI substanceuse indicate an initial decline in frequency of use inthe first year postinjury but a return to levels similar to

preinjury in subsequent years.18,20,2830

Hibbards10

andKoponens7 findings of high frequencies of psychiatricdisorders many years postinjury suggest that such disor-ders develop and persist over very long periods of time.Hence it would seem important to sample various timepoints after injury.

The severity of injuries hasalso varied widely from onestudy to another, as have the measures used to establishinjury severity. Studies of depression in groups of per-sons with predominantly mild TBI have generally foundlower frequencies of DSM-diagnosed depression, rang-ing from 12.8% to 16.6%. Sample size has also limited

the generalizability of findings from some studies thathave used samples of fewer than 50 participants.9,11,31,32

Previous studies have been conducted in NorthAmerica, the United Kingdom, and Finland. There maybe cultural differences in coping styles, emotional ex-pression, stigma associated with injury and mental ill-ness, and attitudes toward substance use, as well asdifferences in availability of treatment. This may alsocontribute to variability in findings.

Rating scales document clinical symptoms, someof which may be the direct result of the TBI rather

than of depression or anxiety, including sleep distur-bance, concentration problems, fatigue, or psychomo-

tor retardation,3335 potentially leading to overestima-tion of prevalence of certain disorders. On the otherhand, cognitive deficits leading to lack of self-awarenessor denial may result in underdiagnosis of disorders.36

The clinical interview is important to establish accurateand reliable diagnoses by distinguishing symptoms dueto brain injury from those due to a psychiatric disorder.

DSM-based clinical interviews have been shown to havehigh sensitivity and specificity in identifying depressedTBI participants, particularly in comparison with a de-pression rating scale.34 Such measures have not beenfrequently employed in studies of individuals with TBI,possibly because of the time involved in administrationand scoring.

In summary, the wide range in frequency of psy-chiatric disorders reported across studies may be at-tributable to variability in diagnostic instruments, studydesign, cultural differences, or personal and injury-

related characteristics of the participants, including pres-ence of preinjury psychiatric disorders, injury severity,and time postinjury. The aim of this study, which is thefirst in a series, was to examine the frequency of pre- andpost-TBI Axis I psychiatric disorders, established on thebasis of structured clinical interview based on DSM-IVcriteria and verified by a significant other. Participantshad mild to severe TBI sustained 0.5 to 5.5 yearspreviously.

METHODS

ParticipantsOne hundred participants with mild to severe TBI sus-

tained between 6 months and 5.5 years previously wererecruited from the database of all admissions of patientswith head injury to the referring hospital, which pro-vided rehabilitation under a no-fault accident compen-sation system. The majority had incurred moderate tovery severe TBI, and most were injured in road accidentsor workplace accidents. All had been discharged from in-patient care. Eligibility criteria were: (1) minimum age 17years at time of injury and maximum of 75 years at timeof interview; (2) having sufficient proficiency in Englishto complete the interview; (3) no history of previous TBIor serious neurological disorder such as stroke, epilepsy,brain tumor, or neurodegenerative disease. Patients witha premorbid psychiatric history were not excluded.

Demographics

Demographic information for study participants isgiven in Table 1. The average participant was a 37-year-old male with just less than 12 years of education, whowas now 3 years postinjury. Mean length of hospital stay

was 41.59 days (SD, 27.59; range, 5134). Participantshad a mean lowest Glasgow Coma Scale (GCS) score of

www.headtraumarehab.com


3/9

326 JOURNAL OF HEAD TRAUMA REHABILITATION/SEPTEMBEROCTOBER 2009

TABLE 1 Demographic data by mean years postinjury; mean, SD, and range

Year postinjury

1 2 3 4 5 Totala

Variable (0.51.49) (1.52.49) (2.53.49) (3.54.49) (4.55.50) (0.55.5)

Glasgow Coma ScoreMean 8.53 8.50 8.53 8.40 9.53 9.10SD 4.35 4.29 4.35 4.22 3.85 4.12Range 314 314 314 314 314 314

PTA (days)Mean 23.90 19.40 13.30 23.28 24.20 20.77SD 22.53 17.77 14.09 17.34 15.68 17.85Range 177 162 149 163 256 177

Age at assessmentMean 38.00 35.00 40.95 35.10 36.85 37.18SD 16.96 11.82 16.61 11.97 13.31 14.19Range 1967 1960 1974 2061 2165 1974

Education, yMean 12.10 11.72 11.58 11.40 11.65 11.70SD 2.71 3.04 2.80 2.09 2.74 2.65Range 717 818 616 816 616 618

GenderMale 75% 70% 75% 70% 65% 71%Female 25% 30% 25% 30% 35% 29%

Depression present (n) 4 10 11 10 11 2 = 7.01, df= 4,P= .14

Anxiety present (n) 4 7 10 9 8 2 = 4.50, df= 4,P= .34

Any disorder present (n) 9 12 14 15 15 2 = 5.71, df= 4,P= .22

Abbreviation: PTA, posttraumatic amnesia.aFor n = 100, average time postinjury was 2.98 years, SD = 1.47, range 0.55.5.

9.10, with 35% scoring 13 to 14, 20% scoring 9 to 12, and45% scoring 3 to 8. Mean duration of posttraumatic am-nesia (PTA) was 20.77 days (SD, 17.85; range, 177). Ninepercent had a PTA duration of less than 24 hours, 20%had a PTA of 1 to 7 days, 42% a PTA of 8 to 27 days, and29% hada PTA greater than 28 days. There were no statis-tically significant differences among each year postinjurygroup (see Table 1), or between the present sample andthe 57 participants who declined, or between the present

sample and the main database group on gender (22 =.025, P = .876), education (t = 1.705, df= 596, P =.089, 2-tailed), PTA (U= 27052.50, N1 = 99, N2 = 570,P= .860, 2-tailed), GCS (U= 28334.00, N1 = 96, N2 =592, P= .963, 2-tailed), or age (U= 29273.50, N1 = 100,N2 = 620, P= .371, 2-tailed), indicating that the currentsample was a representative group of participants basedon demographic- and injury-related variables.

Measures

Demographic- and injury-related information and

psychiatric history were initially obtained via asemistructured interview. With consent, further details

were obtained from all participants medical files, whichmost often included a full neuropsychological assess-ment and psychiatric history. The clinical computerizedversion of the Structured Clinical Interview forDSM-IVdisorders (SCID-I) was used to assess frequency, comor-bidity, and resolution over time of psychiatric disorders.It was administered twicefirst, retrospectively to deter-mine lifetime preinjury psychiatric diagnoses, and sec-ond, to identify postinjury psychiatric diagnoses, both

current and resolved.The SCID-I contains over 37 Axis I diagnoses, cover-

ing mood disturbances, anxiety disorders, schizophre-nia and other psychotic disorders, substance use dis-orders, somatoform conditions, eating disorders, andadjustment disorders.37 Depressive disorders refers todysthymia and major depressive disorders; DSM-basedspecifiers of mild, moderate, and severe were docu-mented. The substance use disorder category refers toalcohol and nonalcohol abuse or dependence disorders.A novel disorder refers to a specific disorder occurringpost-TBI that has never occurred before in that personslifetime.


4/9


Procedures

Ethics approval was obtained from the hospitalthrough which participants were recruited. An indepen-dent researcher identified patients injured between July2000 and July 2005, who were 0.5 to 5.5 years postinjury.Of the 720 participants in the database, 550 met eligibil-

ity criteria. In order to have a sample representative of arange of time points postinjury, eligible participants weredivided into 5 groups who were at different time pointspostinjury (0.51.49, 1.52.49, 2.53.49, 3.54.49, and4.5 to 5.5 years postinjury). Within each year level,individual Statistical Product and Service Solutions(SPSS) codes were entered into a random number gen-erator program from the Web site www.random.org. Par-ticipants were contacted sequentially until there were 5equal groups of consenting participants, with 20 partic-ipants in each group on average 1 to 5 years postinjury(see Table 1). Fifty-seven people refused participation or

did not return messages. The primary researcher thencontacted each consenting person and arranged a meet-ing either at home or at the hospital. Written informedconsent was obtained from all participants. Participantsidentified a significant other (someone who knew thepatient well prior to and postinjury), who was also in-terviewed about the survivors past and current emo-tional state using the SCID-I, either at the same time orby phone. Thirteen people either declined to nominatea significant other or the significant other declined tobe interviewed. In the 87 cases where significant otherswere interviewed, while not every symptom reported wasidentical, there was 100% agreement between diagnosesobtained from interviews with the pairs of participants.

To determine interrater reliability, 12 of the 100 partic-ipants were also assessed in person by a clinical psychol-ogist (L.J.) trained in administering the SCID-I. Bothadministrators had completed psychopathology coursesas part of their doctoral training and both completed a2-day training program in administration of the SCID-Iunder the supervision of an experienced clinical psy-chologist (F.G.). Administration took between 30 and150 minutes, depending upon the complexity of the

interviewees responses. The 12 participants obtaineddiagnoses within the categories of substance-use dis-orders and anxiety disorders, and the two administra-tors obtained perfect agreement within these categories,both for current and for preinjury diagnoses (Cohens = 1.0).

Data analysis

Data were analyzed using SPSS 14 for Windows. Fre-quency measures were obtained for preinjury and postin-jury disorders, both current and resolved. Chi-square

analyses were used to compare frequencies of psychiatricdisorders from 1 to 5 years postinjury.

RESULTS

Frequency of preinjury disorders

Table 2 displays the number of participants with prein-jury Axis I diagnoses. Among those with a documentedpre-TBI psychiatric history, all but one participants self-report of preinjury psychiatric history were consistent

with medical file reports. However, some participantswithouta documented history reported premorbid psy-chiatric symptoms, which warranted a retrospective di-agnosis. It seems, therefore, that some individuals withTBI had undiagnosed preinjury psychiatric disorders.

Prior to injury, 52% had a psychiatric disorder; 28%had only 1 diagnosis, whereas 24% had more than 1 di-agnosis. Preinjury alcohol dependence disorder was themost common single diagnosis (29%); however, othersubstance use disorders were also common, with a totalof 41% falling into these categories. Major depressivedisorder was the second most frequent preinjury disor-der (17%). Preinjury anxiety disorders were also common(13%).

Frequency of postinjury disorders

Following TBI, 65% of the current sample met crite-ria for at least 1 diagnosis. Twenty-seven percent receiveda single diagnosis, whereas 38% received multiple diag-noses. Table 2 displays the breakdown of postinjury di-agnoses. Postinjury major depression was the most com-mon diagnosis (45%), whereas there was only one case

of dysthymia.Anxiety (38%) was the second most common diag-

nosis. Frequencies of individual disorders were greaterthan the overall figure of 38%, as some participants hadmore than 1 anxiety disorder. Generalized anxiety dis-order (GAD) was the most commonly diagnosed anx-iety disorder (17%), followed by posttraumatic stressdisorder (PTSD; 14%). Specific phobia (7%), panic dis-order (with or without agoraphobia, 6%), and socialphobia (6%) occurred with similar frequency. Only 1person was diagnosed with obsessive-compulsive dis-order (OCD) and 1 with agoraphobia. Three people

received postinjury diagnoses of a psychotic disorder,1 had an eating disorder, and 1 a somatoform disorder.

Twenty-one percent met criteria for a postinjury sub-stance use disorder. Alcohol dependence was the mostcommon (14%), followed by nonalcohol substance de-pendence (7%), alcohol abuse (3%), and nonalcohol sub-stance abuse (2%). Marijuana was the most commonlyused drug (45%), followed by stimulants (20%), opioids(15%), and cocaine (5%). The remaining 15% of thegroup were polydrug users.

Comparison of the rate of disorders in the first year

after injury with that in the subsequent years indicatedthat the frequency of depressive disorders in the first year



5/9


TABLE 2 Number of people (N= 100) with SCID-diagnosed psychiatric disorders pre-and post-TBI, novel post-TBI disorders, and breakdown of disorders

Psychiatric disorder Pre-TBI Post-TBI Novel disorders

Number of participants with disorderDepression 17 46 33

Any anxiety disorder 13 38 28Any psychotic disorder 1 3 3Substance use disorders 41 21 3Somatoform disorder 0 1 1Eating disorders 2 1 0

Number of disordersMajor depression 17 45 32Dysthymia 0 1 1GAD 5 17 13PTSD 4 14 10Specific phobia 0 7 7Panic disorder 1 6 5Social phobia 2 6 4OCD 1 1 1

Agoraphobia 1 1 0Substance-induced anxiety disorder 1 0 0Substance-induced psychotic disorder 1 0 0Schizoaffective disorder 0 1 1Psychotic disorder NOS 0 2 2Alcohol abuse disorder 7 3 2Alcohol dependence disorder 29 14 1Nonalcohol substance abuse disorder 5 2 0Nonalcohol substance dependence disorder 12 7 3

Abbreviations: GAD, generalized anxiety disorder; NOS, not otherwise specified; OCD, obsessive-compulsive disorder; PTSD, posttrau-

matic stress disorder; SCID, Structured Clinical Interview for DSM-IVdisorders; TBI, traumatic brain injury.

postinjury was significantly lower than in subsequentyears (2 = 6.80, df= 1, P = .012). This result ap-proached significance for anxiety (2 = 3.43, df= 1,P= .075) and for any disorder (2 = 4.40, df= 1, P=.064).

Novel disorders following TBI

Numerous participants experienced depressive disor-ders and anxiety disorders for the first time followinginjury (see novel disorders column, Table 2). Of the

46 people who experienced a depressive disorder postin-jury, 33 had developed depression for the first time. Afurther 7 were depressed at the time of the accident, and6 had a preinjury history of depression, which was inremission at the time of injury.

Of the 38 people who experienced anxiety disorderpostinjury, 28 were new onset. For GAD, 13 were newcases, 3 were ongoing, and 1 was a relapse of a previouslyresolved disorder. Ten people experienced PTSD as anovel disorder; the other 4 cases were current at the timeof injury. All diagnoses of specific phobia and psychotic

disorders were novel, and 5 of 6 cases of panic disorderwere novel, as were 4 of 6 cases of social phobia.

Few new onset substance use disorders were evident;only 3 of 21 people developed a postinjury substance usedisorder with no prior history. Two cases of novel alco-hol abuse disorder were found; 1 of these people had apreinjury nonalcohol substance disorder. Of those withpostinjury alcohol dependence, 1 was a novel disorder,10 had alcohol dependence at the time of injury, and 3had a history of alcohol dependence prior to injury butwere not dependent at the time of injury. There wereno new cases of nonalcohol substance abuse; 1 case was

ongoing and 1 had relapsed. Of the 7 people with non-alcohol substance dependence, 3 were novel disordersand 4 were present prior to injury.

Current and comorbid disorders

Table 3 shows Australian prevalence rates for psychi-atric disorders where available.16 Given that the averageparticipant was male, aged 37, prevalence rates for malesof this age group are also provided. Of the 46 partici-pants found to have depression at any time postinjury,34 (74%) were depressed at the time of assessment (see

Table 3); therefore, 12 cases of depression had been re-solved by the time of interview. Of these participants


6/9


TABLE 3 Number of participants (N= 100) with current and resolved psychiatric disor-ders post-TBI and 12-month Australian prevalence rates (where available)

Australian ratesPostinjury disorders

General MalesDisorder Current %a Resolved %b population %c Males %d 3544 %

Number of participants with disorder (N)Depressive 34 12 5.1 3.4 6.0Anxiety 36 2 9.7 7.1 8.3Substance use 17 4 7.7 11.1 12Psychotic 2 1Somatoform 1 0Eating 0 1

Number of disorders (N)GAD 14 3 3.1 2.4PTSD 11 3 3.3 2.3Specific phobia 7 0Panic disorder 6 0 1.3 0.6Social phobia 6 0 2.7 2.4

OCD 1 0 0.4 0.3Agoraphobia 1 0 1.1 0.7Schizoaffective 1 0Psychotic NOS 1 1Alcohol abuse 2 1 3 4.3Alcohol dependence 10 4 3.5 5.1Nonalcohol substance abuse 2 0 3.5Nonalcohol substance dependence 5 2 2.2

Abbreviations: GAD, generalized anxiety disorder; NOS, not otherwise specified; OCD, obsessive-compulsive disorder; PTSD,

posttraumatic stress disorder; TBI, traumatic brain injury.aCurrent % refers to number of participants with current diagnoses divided by total number of participants (N= 100) multiplied by

100.b

Resolved % refers to number of resolved diagnoses divided by total number of participants ( N= 100) multiplied by 100.cRates are for disorders that occurred in the last 12 months, males and females, 18 years and above. 16

dPrevalence rates are for disorders that occurred in the previous 12 months for males 18 years and above. 16

with current depression, 5 were in partial remission,4 had mild symptoms, 12 had moderate symptoms,and 13 had severe symptoms at the time of assessment.Eight people with current depression had a comorbidsubstance use disorder (23.5%), and 25 had a comorbidanxiety disorder (73.5%). Of the 45 participants withcurrent depression and/or anxiety, 23 (51.1%) were re-ceiving medication and/or counseling.

Of the 38 people with postinjury anxiety disorders,only 2 cases had resolved by the time of interview (seeTable 3). All cases of specific phobia, panic disorder,social phobia, OCD, agoraphobia, and somatoform dis-order were current. High current frequencies were alsofound for the remaining anxiety disorders, as 14 of 17cases (82%) of post-TBI GAD were current, and 11 of14 cases (79%) of the PTSD cases were current. Six peo-ple with anxiety (16.7%) had a comorbid substance usedisorder.

Of the 3 diagnosed postinjury psychotic disorders, 2were current. The 1 eating disorder was in remission.Although the overall frequency of substance use disor-

ders fell preinjury to postinjury, postinjury substance usedisorders tended to be current (see Table 3).

Treatment

Twenty-three of the 45 participants with current de-pression and/or anxiety (51.1%) were being treated withmedication and/or psychological therapy, as compared

with 31.3% (n = 5) of those for whom depression oranxiety had resolved at time of assessment.

DISCUSSION

This study aimed to examine the frequency of psychi-atric disorders in an Australian sample of 100 individualswith mild to severe TBI 0.5 to 5.5 years postinjury, usinga DSM-based structured clinical interview to establishdiagnoses. A significant other was also interviewed in87% of cases. Reliability of participants self-report wasdemonstrated.

The current study found a high proportion of prein-jury and postinjury psychiatric disorders (52% and 65%,



7/9


respectively). Preinjury, depression and substance usedisorders were most common (17% and 41%, respec-tively). These findings are consistent with most previousTBI studies.10,11,18 The frequency of preinjury anxietydisorders in this study (13%) was the same as that foundby Hibbard.10 Frequencies of most preinjury psychiatricdisorders in the current study were much higher than

those in the Finnish Koponen7 study, in which therewere no preinjury depressive disorders, GAD, or nonal-cohol substance disorders. However, given that partici-pants in that study were at least 27 years postinjury, itis possible that their retrospective account of premor-bid diagnoses was unreliable. Furthermore, all TBIs oc-curred between 1950 and 1971, at a time when mentalhealth disorders were relatively less well recognized andacknowledged. There may also be cultural differencesinfluencing expression of emotion and/or substance usepatterns.

Following TBI, 65% of participants received at least 1psychiatric diagnosis. Again, depression was the mostcommon diagnosis in 46% in the first 5 years af-ter injury, a frequency consistent with some previousstudies.10,11,14 High current rates of depression werefound (34%), which were substantially higher than theAustralian comparison rates.16 Seventy-two percent ofdepressive disorders were novel disorders, suggestingthat the presence of preinjury psychiatric disorders isnot the only influential factor. As with the general pop-ulation, there was also a high frequency of comorbid-ity of current depressive and anxiety disorders (73.5%).

Only half with current depression and/or anxiety werereceiving treatment. Other studies have reported lowerfrequencies of depression and anxiety,2,18,25 but all in-volved follow-up in the first year after injury. In thecurrent study, there was a trend for the frequency ofpsychiatric disorders to rise between 1 and 4 yearspostinjury. This increase may be associated with factorssuch as improved insight into the effects of the injuryover time, growing despondence at the lack of phys-ical/emotional/vocational progress, financial hardship,or decrease in professional support over time.

The number of participants with anxiety disorders rose

from 13% to 38% postinjury, with a total of 52 diag-noses. Of those with an anxiety disorder post-TBI, 74%had developed novel anxiety disorders, most commonlyGAD and PTSD, followed by specific and social pho-bias and panic disorder. Anxiety disorders showed poorresolution, with 95% of cases current at the time of as-sessment. Frequencies of most current anxiety disorderswere all much higher than in the general population16

but were generally within the ranges reported in previousstudies of persons with TBI.10,11,14 However, the cur-rent reported frequency of OCD (1%) was much lower

than a previous study that reported an incidence of 15%;it was suggested that checking-rechecking behavior may

be a compensatory strategy for poor memory followingTBI.10 However, in order to meet DSM-IV criteria fordiagnosis of OCD, this behavior must be attributed toadherence to a rigid set of rules in order to avoid negativeconsequences (anxiety), rather than memory problemscausing repeated checks to ensure safety.

No other known studies of individuals with TBI have

screened for eating disorders or somatoform disorders.Agoraphobia, eating disorders, somatoform disorders,and psychotic disorders all occurred at frequencies sim-ilar to those in the general population, suggesting thatthese are not common consequences of TBI. However,it is also possible that the current study lacked the powerto detect any change in frequency of such disorders. Inaddition, there were 3 new cases of psychotic disorders,and given previous findings of onset of psychosis morethan 4 years postinjury,26,27 one cannot rule out the pos-sibility of psychotic disorders developing over a longertime frame.

Frequencies of substance use disorders decreasedfrom 41% preinjury to 21% postinjury, with few novelcases. This trend has previously been noted in theliterature.10,20 Frequencies of comorbidity were highof those with current depression and anxiety disorders,23.5% and 16.7%, respectively, also had a substance usedisorder. Current substance use disorders occurred in17% of participants, which is much higher than theiroccurrence in the general population.16 High rates ofalcohol consumption are not unique to the TBI popu-lation and are indeed typical of the young male demo-

graphic population from which they are drawn.20

How-ever, given the severity of brain injury in this group, theamount of alcohol being consumed is potentially muchmore harmful and therefore of concern. It is importantto institute measures to discourage those who engagedin heavy substance use preinjury from returning to this,following TBI.

The findings of this study must be interpreted withinthe context of certain limitations, most notably its retro-spective, cross-sectional design, a limitation also presentin most previous studies. In the present investigation,stratified random sampling was used to ensure inclusion

of equal andrepresentative groups on average 1 to 5 yearsafter injury. However, this design does not allow for pre-cise examination of the timing of onset of disorders.As previously discussed, the referring hospital treatedpatients referred for rehabilitation under a no-fault ac-cident compensation system, so the sample compriseda high proportion of individuals with moderate to verysevere TBI, the majority of whom incurred injuries inmotor vehicle or work-related accidents. Therefore, thefindings of the current study may not be generalizableacross the entire spectrum of severity of TBI, particu-

larly mild TBI, or those not referred for rehabilitation.Despite the use of a stratified random sampling method,


8/9


the possibility of some selection bias cannot be ruledout. It may be that certain people were more or lessmotivated to participate, depending on their emotionalstate. It would also be of interest in future studies witha stratified time postinjury design to consider whetherpatients with a shorter or longer time postinjury weremore or less inclined to participate. This information

was not available for the current study. One could alsoquestion the reliability of retrospective reports of prein-jury symptoms experienced up to 51/2 years previously.The involvement of a significant other in verifying re-ported symptoms mitigated against this. Clearly, it willbe important to follow up these findings with a prospec-tive study, conducted over a longer time frame than thatused in the previous 1-year outcome studies to date.

The sample studied was predominantly male (71%).Although this gender imbalance is typical of the TBIpopulation, it may have influenced the frequencyof observed disorders. The majority of studies havefound no gender differences in frequencies of postin-jury depression.2,7,10,13,18,39,40 However, females maybe more likely to be diagnosed with an anxietydisorder.10,13,14 Males have been more often diagnosedwith a substance use disorder in both the TBI and thegeneral population.14,20

Preexisting emotional and substance abuse problemshave been associated with a greater likelihood of TBI.3

Numerous interdependent factors contribute to the like-lihood of having a TBI and to the development ofa men-tal health disorder; it is possible that the same groups

are at risk for both conditions. Young men account for alarge proportion of the population with brain injury, andarguably the frequencies of depression and substanceuse may be higher in this subgroup than the populationnorms against which they have been compared. A demo-graphically matched control group should be consideredin future studies.

Few studies have used measures such as the SCID-Ito establish preinjury and postinjury diagnoses. On thebasis of the high degree of concordance between inter-view results and medical records, the current study hasdemonstrated the appropriateness of thismeasure for usein the TBI population. Given the range of disorders thatoccur in this group, it is recommended that psychiatric

screening cover a broad range of diagnoses.This study has highlighted that TBI creates a risk

for development of several psychiatric disorders, par-ticularly depression and anxiety, in a significant pro-portion of those who had no previous psychiatric his-tory. That this finding was obtained in a group thathad access to comprehensive rehabilitation supports theneed for a greater focus on prevention and treatmentof these problems both within and outside rehabilita-tion programs. Clearly, practitioners cannot rely solelyon preinjury history of psychiatric problems to predictpostinjury problems. Moreover, many who experienceanxiety and depression may not recognize or report it.It is therefore vital that community-based health pro-fessionals are trained to recognize the symptoms andsigns of these conditions in an individual with TBI andare provided with strategies and resources with whichto address them, including the availability of skilledpsychological or psychiatric intervention. Patients andfamilies should be informed regarding symptoms ofdepression and anxiety. TBI follow-up clinics shouldconduct routine long-term screening for such disor-ders. Education about the implications of substance use

following brain injury is also of vital importance. In-vestigation of the factors associated with the develop-ment of these disorders would assist in identifying thosemost at risk, so that they may be targeted for assess-ment and intervention. This represents the focus of thesubsequent study in this series by Whelan-Goodinsonet al.41

REFERENCES

1. Jennett B. Epidemiology of head injury. J Neurol Neurosurg Psychi-atry. 1996;60:362369.

2. Deb S, Lyons I, Koutzoukis C, Ali I, McCarthy G. Rate of psy-chiatric illness 1 year after traumatic brain injury. Am J Psychiatry.1999;156:374378.

3. Ponsford J, Sloan S, Snow P. Traumatic Brain Injury: Rehabilitationfor Everyday Adaptive Living. Hove, Sussex, UK: Psychology Press,

Ltd; 1995.4. MacMillan PJ, Hart RP, Martelli MF, Zasler ND. Pre-injury sta-

tus and adaptation following traumatic brain injury. Brain Inj.2002;16:4149.

5. Varney NR, Martzke JS, Roberts RJ. Major depression in patientswith closed head injury. Neuropsychology. 1987;1:79.

6. Rosenthal M, Christensen BK, RossTP. Depression following trau-

matic brain injury. Arch Phys Med Rehabil. 1998;79:90103.

7. Koponen S, Taiminen T, Portin R, et al. Axis I and II psychiatricdisorders after traumatic brain injury: a 30-year follow-up study.

Am J Psychiatry. 2002;159:13151321.8. Kreutzer JS, Seel RT, Gourley E. The prevalence and symptom

rates of depression after traumatic brain injury: a comprehensiveexamination. Brain Inj. 2001;15:563576.

9. van Reekum R, Bolago I, Finlayson MAJ, Garner S, LinksPS. Psychiatric disorders after traumatic brain injury. Brain Inj.1996;10:319327.

10. Hibbard MR, Uysal S, Kelpler K, Bogdany J, Silver J. Axis I psy-chopathology in individuals with traumatic brain injury. J HeadTrauma Rehabil. 1998;13:2439.

11. Fann JR, Katon WJ, Uomoto JM, Esselman PC. Psychiatric disor-ders and functional disability in outpatients with traumatic brain

injuries. Am J Psychiatry. 1995;152:14931499.



9/9


12. Jorge R, Robinson RG, Starkstein SE, Arndt SV. Depression andanxiety following traumatic brain injury. J Neuropsychiatry ClinNeurosci. 1993;5:369374.

13. LevinHS, Brown SA,Song JX, et al.Depression and posttraumaticstress disorder at three months after mild to moderate traumaticbrain injury. J Clin Exp Neuropsychol. 2001;23:754769.

14. Ashman TA, Spielman LA, Hibbard MR, Silver JM, Chandna T,Gordon WA. Psychiatric challenges in the first 6 years after trau-

matic brain injury: cross-sequential analyses of Axis I disorders.Arch Phys Med Rehabil. 2004;85(4)(suppl 2):S36S42.

15. Jorge R, Starkstein SE, Arndt S, Moser D, Crespo-Facorro B,Robinson RG. Alcohol misuse and mood disorders following trau-matic brain injury. Arch Gen Psychiatry. 2005;62:742749.

16. Australian Institute of Health and Welfare. NationalHealthPriorityAreas Report: Mental Health 1998. Canberra, Australian CapitalTerritory, Australia: Australian Institute of Health and Welfare,Cat No PHE 13 Heath and AIHW; 1999.

17. Hibbard MR, Bogdany J, Uysal S, et al. Axis II psychopathologyin individuals with traumatic brain injury. Brain Inj. 2000;14:4561.

18. Fedoroff J P, Starkstein SE, Forrester AW, et al. Depression inpatients with acute traumatic brain injury. Am J Psychiatry.

1992;149:918923.19. Drubach DA, Kelly MP, Winslow MM, Flynn JP. Substance abuse

as a factorin the causality, severity andrecurrence rate of traumaticbrain injury. Md Med J. 1993;42:989993.

20. Ponsford J, Whelan R, Bahar-Fuchs A. Alcohol and drug usefollowing traumatic brain injury: a prospective study. Brain Inj.2007;21:13851392.

21. Dawson DR, Levine B, Schwartz M, et al. Quality of life fol-lowing traumatic brain injury: a prospective study. Brain Cogn.2000;44:3539.

22. Dicker BG. Pre-injury behavior and recovery after a minor headinjury: a review of the literature.J HeadTraumaRehabil. 1989;4:7381.

23. Rapoport MJ, McCullagh S, Streiner D, Feinstein A. The clinical

significance of major depression following mild traumatic braininjury. Psychosomatics. 2003;44:3137.

24. Deb S, Lyons I, Koutzoukis C. Neuropsychiatric sequelae oneyear after a minor head injury. J Neurol Neurosurg Psychiatry.1998;65:899902.

25. Jorge R, Robinson RG, Starkstein SE. Influence of major depres-sion on 1 year outcome in patients with traumatic brain injury. JNeurosurg. 2004;8:726733.

26. Fujii DE, Ahmed I. Risk factors in psychosis secondary to trau-matic brain injury. J Neuropsychiatry Clin Neurosci. 2001;13:6169.

27. Sachdev P, Smith JS, Cathcart S. Schizophrenia-like psychosis fol-

lowing traumatic brain injury: a chart-based descriptive and case-control study. Psychol Med. 2001;31:231239.

28. Sparadeo FR, Gill D. Effects of prior alcohol use on head injuryrecovery. J Head Trauma Rehabil. 1989;4:7582.

29. Corrigan JD, Rust E, Lamb-Hart GL. The nature and extent ofsubstance abuse problems in persons with traumatic brain injury.

J Head Trauma Rehabil. 1995;10:2946.30. Kreutzer JS, Witol AD, Sander AM, et al. A prospective longitu-

dinal multicenter analysis of alcohol use pattern among personswith traumatic brain injury.J HeadTraumaRehabil. 1996;11:5869.

31. Glenn MB, ONeil-Pirozzi T, Goldstein R, Burke D, Jacob L. De-pression amongst outpatients with traumatic brain injury. BrainInj. 2001;15:811818.

32. Robinson RG, Boston JD, Starkstein SE, Price TR. Comparisonof mania and depression after brain injury: causal factors. Am JPsychiatry. 1988;145:172178.

33. Warden D, Labbate L. Posttraumatic stress disorder and other anx-iety disorders. In: Silver J, McAllister TW, Yudofsky SC, eds. Text-book of Traumatic Brain Injury. Washington, DC; London: Ameri-can Psychiatric Publishing; 2005:231243.

34. Jorge R, Robinson RG, Arndt S. Are there symptoms that are spe-cific for depressed mood in patients with traumatic brain injury.

J Nerv Ment Dis. 1993;181:9199.35. Green A, Felmingham K, Baguley I, Slewa-Younan S, Simpson S.

The clinical utility of the Beck Depression Inventory after trau-matic brain injury. Brain Inj. 2001;15:10211028.

36. Robinson RG, Jorge R. Mood disorders. In: Silver J, McAllisterTW, Yudofsky SC, eds. Textbook of Traumatic BrainInjury. Washing-ton, DC; London: American Psychiatric Publishing; 2005:201212.

37. First M B, Spitzer RL, Gibbon M, et al. Computer-Assisted SCIDClinician Version (CAS-CV): Software Manual. Washington, DC:American Psychiatric Publishing; 2004.

38. American Psychiatric Association. Diagnosticand Statistical Manualof Mental Disorders. 4th ed, text rev. Washington, DC: AmericanPsychiatric Association; 2000.

39. Jorge R, Robinson RG, Arndt SV, Starkstein SEForresterAWGeisler F. Depression following traumatic brain injury: a 1year longitudinal study. J Affect Disord. 1993;27:233243.

40. Seel RT, Kreutzer JS, Rosenthal M, Hammond FMCorrigan JD-Black K. Depression after traumatic brain injury: a National In-stitute on Disability and Rehabilitation Research Models Systemsmulticenter investigation. Arch Phys Med Rehabil. 2003;84:177184.

41. Whelan-Goodinson R, Ponsford J, Schonberger M. Predictorsof psychiatric disorders following traumatic brain injury. J HeadTrauma Rehabil. In press.

psychiatric disorders following tbi

Documents