J Head Trauma Rehabil Vol. 31, No. 2, pp. E12–E25 c 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright 

The Impact of Traumatic Brain Injury on Self-Identity: A Systematic Review of the Evidence for Self-Concept Changes Elizabeth Jane Beadle, DPsy, BSc(Hons); Tamara Ownsworth, PhD, BA(Hons); Jennifer Fleming, PhD, BOccThy(Hons); David Shum, PhD, BA(Hons) Objectives: This review systematically appraised the evidence for changes to self-identity after traumatic brain injury (TBI) in adults and investigated associations between self-concept changes and neurocognitive and psychosocial functioning. Methods: Systematic searches of 4 databases (PsycINFO, PubMed, CINAHL, and Cochrane Systematic Review Database) were undertaken from January 1983 to July 2014. Empirical studies were included if they used a quantitative measure of pre-/postinjury changes in self-concept after TBI or compared levels of self-concept between TBI and control participants. Results: Fifteen studies met the review criteria and, despite methodological differences, provided mostly evidence of negative changes to self-concept. However, stability in self-concept and positive changes to sense of self were also reported in some studies. Furthermore, levels of self-esteem and personality characteristics did not significantly differ between participants with TBI and orthopedic/trauma controls. Negative self-concept changes were associated with emotional distress in 3 studies. Conclusions: People with TBI most commonly experience negative changes in self-identity; however, such changes are also reported after other traumatic events or injuries. Greater consistency in measurement of self-identity change and use of longitudinal designs is recommended to improve understanding of factors contributing to self-concept changes after TBI and to guide clinical interventions. Key words: self-concept, self-esteem, self-identity change, systematic review, traumatic brain injury

Author Affiliations: School of Applied Psychology and Behavioural Basis of Health Program, Griffith Health Institute, Griffith University, Mt Gravatt, Queensland, Australia (Drs Beadle, Ownsworth, and Shum); School of Health and Rehabilitation Sciences, University of Queensland, St Lucia, Queensland, Australia (Dr Fleming); and Princess Alexandra Hospital, Woolloongabba, Queensland, Australia (Dr Fleming). Dr Beadle was awarded a Centre for Research Excellence in Brain Recovery PhD scholarship. There are no financial relationships that affect this body of work. There is no conflict of interest relating to any of the authors. Clinical trial registration: Nil. Contributors’ statements: Dr Beadle contributed to conceptualization and design of the study, conducted all searches, reviewed all studies as the primary rater for study inclusion and exclusion, appraised methodological quality of included studies, drafted the initial manuscript, and approved the final manuscript as submitted; Associate Professor Ownsworth contributed to conceptualization and design of the study, reviewed all studies as the second rater for study inclusion and exclusion, and appraised methodological quality, as well as reviewing all drafts of the manuscript and approved the manuscript as submitted; Associate Professor Fleming was involved in conceptualization of the study and critically reviewed and approved the final manuscript as submitted; and Professor Shum was involved in conceptualization of the study and critically reviewed and approved the final manuscript as submitted. The authors declare no conflicts of interest.

S

ELF-IDENTITY IS A NEBULOUS term often used interchangeably with “sense of self” and selfconcept. Philosopher William James1 conceptualized the self as a complex personal construction and “the most puzzling puzzle with which psychology has to deal”1(p30) due to the elusive nature of the term. Since James, diverse perspectives on self-identity have evolved within philosophy, sociology, and neuroscience.2–4 In the psychological literature, self-identity is broadly defined as the collective characteristics we perceive as our own, which endure over time and are continuously under construction.4 Self-identity research typically focuses either on individual characteristics (ie, what makes us unique) or on those shared with others on the basis of common social memberships.5 There is a general Corresponding Author: Elizabeth Jane Beadle, DPsy, BSc(Hons), School of Applied Psychology and Behavioural Basis of Health Program, Griffith Health Institute, Griffith University, Mt Gravatt, QLD 4122, Australia ([email protected]). DOI: 10.1097/HTR.0000000000000158

E12 Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

The Impact of TBI on Self-Identity consensus that self-identity develops from and is shaped by an interplay of internal (biological, cognitive) and external (culture, social environment, life events) factors.6–8 Although terminology differs,4,5 efforts have been made to operationalize the construct of self-identity for measurement. In developmental psychology, measures of self-concept and self-esteem are commonly used to infer people’s sense of self or self-identity.4,9,10 Self-concept refers to the overarching thoughts and feelings a person has about him or herself, whereas self-esteem represents an evaluative component or judgments about one’s own worth or value.11 Developmental accounts propose that self-concept develops in a hierarchical manner with lower-level self-perceptions (eg, situation-specific self-appraisals) having a bottomup influence on domain-specific areas (eg, physical, cognitive, social, and behavioral), which are subjectively weighted in importance and contribute to higher-level self-representations or global self-concept.12 Hence, self-concept is viewed by developmental theorists as both unitary and multidimensional with global and domain-specific components.9,12 Although lower-level self-perceptions can be altered by situational influences to produce state-like variations,13 research supports that global self-concept is relatively stable across the life span.14,15 Moreover, the degree of stability is similar to that observed for personality traits.15 Researchers from diverse psychology fields have investigated the biological and social conditions for changes to self-identity and self-concept.6,16 Such changes have been attributed to many life experiences4 , including (1) a neurological event,17,18 for example, the onset of dementia or acquiring a brain injury; (2) major life transitions and stressful events,14 for example, becoming a parent; (3) social evaluation,6,19 for example, external feedback; and (4) psychotherapy and other interventions targeting self-schema.20,21 Of particular relevance to this review, changes in selfconcept and self-esteem have been investigated in the context of illness and trauma for many populations such as chronic pain,22 spinal cord injury,23 stroke,24 and cancer.25 The domain of self-concept investigated in these studies has typically varied according to the functional consequences of the condition (eg, body image and sexual self-concept after spinal cord injury). However, a consistent finding in the chronic illness literature is that global and domain-specific self-concepts can be altered by the experience of serious illness, which, in turn, influences emotional adjustment and quality of life.22 The impact of TBI on self-identity and social identity has received greater attention in recent literature.4,26–29 A meta-synthesis30 of 23 qualitative studies on subjective experiences of recovery after TBI found that changes in

E13

identity, including loss of self, social disconnection, and reconstruction of self-identity, were dominant themes. Similarly, a scoping review31 of 19 qualitative studies highlighted the common experience of loss of identity and stigmatized identity, and the challenging process of reconstructing one’s sense of self in rehabilitation. Many individuals reported losing something of themselves that they valued, but this was difficult to define.32 More recently, using grounded theory, Levack and colleagues26 conceptualized self-identity after TBI according to 3 components: (1) self-coherence, or having a satisfying and complete sense of self; (2) feeling respected, validated, and accepted by others; and (3) having a valued place in the world. Participants described a sense of loss in each of these aspects of self-identity. These insights into the lived experience of self-identity change highlight the importance of assessing changes to sense of self after TBI. In research and clinical practice, the impact of TBI on self-identity is typically assessed using 2 main quantitative approaches. First, people’s responses on validated measures of current self-concept or self-esteem may be compared with relevant norms or matched controls to indicate people’s level of self-esteem or perceived competency across a range of domains.33,34 Although not referred to as measures of self-concept per se, personality inventories have also been used to examine selfperceived changes in characteristic patterns of emotions, behavior, and thinking.35 However, unless measures are conducted very early in recovery and readministered over time,24 it is not possible to determine changes in self-concept or personality arising from TBI. The second approach involves asking people to retrospectively rate their preinjury self on a broad list of attributes and then rate their current self on the same attributes to indicate changes in self-concept since the injury.36 Various issues potentially influence retrospective ratings (eg, memory impairment and overly positive views of preinjury self4 ); nonetheless, this approach seeks to capture people’s own subjective understanding of changes to their sense of self. Related to this point, the construct of self-identity and its measurement needs to be distinguished from “self-awareness” or accuracy of self-appraisal after brain injury, for which self-ratings of functioning are compared with a presumably more objective standard such as significant others’ ratings.37 Rather, the focus of the second approach is on assessing “self-discrepancy” or changes in self-concept since TBI, which has important implications for psychological adjustment and rehabilitation endeavors. Informed by various social psychology theories (eg, cognitive dissonance theory,38 balance theory39 ), selfdiscrepancy theory19 proposes that emotional distress is experienced when we hold 2 or more conflicting selfbeliefs. This theory argues that discrepancy between www.headtraumarehab.com

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

E14

JOURNAL OF HEAD TRAUMA REHABILITATION/MARCH–APRIL 2016

one’s actual self (ie, who I actually am) and ideal self (ie, who I want to be) underlies depression whereas discrepancy between one’s actual self and ought self (ie, who I should be) elicits anxiety. The concept of selfdiscrepancy has been applied to theories of adjustment in the TBI literature.40,41 In the context of TBI, preinjury self represents a salient standard for comparison with actual or current self (ie, abilities, attributes, and behavior), whereby the perception of negative changes between one’s preinjury and current selves are proposed to contribute to emotional distress.41 Self-discrepancy between one’s preinjury and postinjury selves is also proposed to influence engagement and progress in rehabilitation.41 Negative self-discrepancies may motivate people to set goals that focus on regaining their abilities and returning to their former self.42 Conversely, these may propel people to search for meaning in, and adapt to, changes brought about by the injury.43 Over time, heightened distress associated with persisting negative self-discrepancies may produce a sense of hopelessness and lead to maladaptive coping and disengagement from rehabilitation.4,41 Conversely, lack of self-discrepancy may be related to neurologically based awareness deficits and/or defensive coping (eg, minimization and denial), both of which pose a barrier to active participation in rehabilitation.44 In a recent systematic review of intervention studies Ownsworth and Haslam29 found mixed support for the efficacy of rehabilitation for improving self-concept after TBI. They argued the need for improved assessment and intervention methodology to improve understanding of changes to sense of self in the context of rehabilitation. Ownsworth4 presented a biopsychosocial conceptualization of factors contributing to self-identity change after brain injury. Drawing on the cognitive awareness model,45 which was developed to account for impaired self-awareness in Alzheimer disease, the Ownsworth4 model asserts that self-perceptions of personal abilities and attributes are updated by experiences of relative success and failure in daily living (eg, perceived performance on occupational tasks and during social interaction). Higher-order cognitive processes support people to identify a mismatch between stored self-representations (ie, preinjury self) and current self. This involves recalling past experiences, forming broader generalizations from patterns in one’s experiences, and assimilating these with existing notions of self to produce modified selfschema (ie, the old and new me). This account suggests that despite marked functional impairments, individuals with severe memory and executive dysfunction may perceive fewer changes to their personal abilities and attributes, thus maintaining preinjury notions of self.4 The Ownsworth4 model proposes that psychosocial factors, including personality, coping style, and the social environment (eg, feedback), affect how people

make sense of and respond to their injury-related impairments. Therefore, neurocognitive and psychosocial factors may influence the extent to which individuals experience self-discrepancy after TBI and hence report changes in self-identity.

STUDY RATIONALE AND AIMS Qualitative studies highlight that people with TBI experience complex changes to their sense of self and find the process of reconstructing their self-identity challenging.30,31 In quantitative research, the impact of TBI on self-identity is typically examined using psychometric measures of current self-concept or selfdiscrepancy (ie, comparison of preinjury vs postinjury self-concept). However, it is unclear whether findings differ according to the approach to measurement or aspect of self-concept investigated (ie, global vs domainspecific areas). A review of assessment methods and research investigating self-identity changes after TBI may improve understanding of psychological adjustment and guide clinical interventions. Accordingly, the primary objective of this review was to examine the evidence for self-identity change after TBI by appraising quantitative studies that (a) examined pre-/postinjury selfdiscrepancies, or (b) compared current levels of selfconcept of people with TBI and matched controls (ie, healthy controls and/or orthopedic/psychiatric controls). A secondary objective was to investigate factors related to self-identity change.

METHODS Search strategy The following electronic databases were searched for the period of January 1983 to July 2014: Cochrane Systematic Review Database, CINAHL, PsycINFO, and PubMed. Bibliographies of included studies were manually searched for any further relevant studies. Two sets of key words were entered to search each database: 1. Neurotrauma terms: Brain injur∗ ; Head injur∗ ; Brain damage; Head trauma. 2. Identity change terms: Self discrepancy; Self concept; Self esteem; Identity; Discontinuity; Continuity; Self change∗ ; Possible sel∗ ; Self perception; Personhood; Personality change. The search strategy was tailored for each specific database, according to search limit options. The full search strategy for each database is available from the first author on request. One of the authors (E.B.) screened all abstracts and titles for inclusion, whereas 2 of the authors (E.B. and T.O.) were involved in the review of the full texts for potentially eligible studies.

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

The Impact of TBI on Self-Identity Selection criteria The review included quantitative studies (ie, those using formal psychometric measures and statistical analyses) that investigated self-identity change or compared levels of current self-esteem, self-concept, or personality characteristics of adults with TBI with matched controls (healthy or orthopedic/trauma samples). For the purposes of this review, self-identity change was defined as subjective appraisal of postinjury changes in emotional, personality, and behavioral attributes as compared with preinjury functioning. Further eligibility criteria included the following: samples with mixed causes of brain injury comprised at least 70% with TBI, and a validated approach to measuring self-esteem, selfconcept, self-discrepancy, or personality was used. The exclusion criteria included the following: more than 30% of participants had a brain injury from other causes (eg, stroke), pediatric sample, only qualitative data, and publication in a language other than English. Data extraction A 5-point scale was developed on the basis of the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines46 to rate the

E15

methodological quality of the articles selected. Articles were scored 0 (does not meet criteria) or 1 (meets criteria) on the following criteria: (1) representative and adequate sampling (eg, consecutive admissions and n ≥30); (2) longitudinal design and/or matched control sample; (3) used a measure of self-concept, self-esteem, or personality that has been validated for TBI and/or is well validated in the general population; (4) reporting of means and standard deviations in text, table, or graph; and (5) adequate sample homogeneity, as indicated by TBI participants with similar chronicity (time since injury) or severity of injury. Higher scores represented stronger methodological quality. Two reviewers (E.B. and T.O.) independently appraised each study according to the 5 criteria. Discrepancies were resolved through discussion to determine a final score. RESULTS As shown in Figure 1, a total of 1278 articles were initially identified from the search process across 4 databases. Of these, 288 duplicate articles were excluded. The title and abstract were reviewed for 990 articles, and a further 920 were removed for reasons including not relevant to the population (eg, not TBI,

Figure 1. Flow diagram of search strategy and outcomes of systematic review.

www.headtraumarehab.com Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

E16

JOURNAL OF HEAD TRAUMA REHABILITATION/MARCH–APRIL 2016

pediatric); not relevant to self-identity; and review article/opinion/editorial/book chapter. A full-text review of the remaining 62 articles was conducted, and 47 were excluded on the basis of lack of measurement of current self-concept or self-identity change (or use of relative’s reports only), only qualitative data, nonvalidated or single-item measures, and case study. There was 100% agreement on the inclusion/ exclusion of each study between the 2 reviewers; however, judgment of eligibility for one study was deemed “uncertain” by both reviewers and therefore the opinion of the third author (J.F.) was sought and supported inclusion. Two articles were identified as using the same data set34,47 and so were considered as a single study.34 A final set of 15 studies was included for review. The characteristics and findings of these studies are summarized in Table 1. Methodological quality There was substantial interrater agreement in ratings of methodology quality (κ = 0.90). As shown in Table 2, one study met 1 of 5 criteria,40 7 studies met 2 of 5 criteria,36,48–52,57 2 studies met 3 of 5 criteria,34,53 3 studies met 4 of 5 criteria,33,35,56 and 2 studies met 5 of 5 criteria.54,55 Five of the studies used a representative sampling approach, such as consecutive hospital admissions or discharges.33,51,54–56 There was considerable heterogeneity in sample characteristics, with only 6 studies recruiting TBI participants who were similar in terms of chronicity or injury severity.33,35,51–54 Chronicity ranged from 2 months to 40 years postinjury, whereas injury severity ranged from mild to very severe. One study55 was noteworthy as the only one to measure self-identity change longitudinally. This study retained 21 of 36 patients (58%) over a 2.5-year follow-up and found that negative self-discrepancies persisted longterm. Another study50 examined stability of ratings of current self-esteem over a 2-week period but not pre- and postinjury self-discrepancy. Seven studies33–35,53,54,56,57 examined the impact of TBI on self-identity by comparing the current self-concept or personality characteristics of people with TBI with healthy controls or orthopedic/trauma controls. Six of these studies used a matched control sample, with matching on at least 2 characteristics such as age, gender, race, education, and chronicity.33–35,53,54,56 Thirteen studies used measures that either were specifically developed or validated for TBI or are well established in the general population.33–36,48–50,52–57 Four studies reported reliability coefficients for the measure of identity change in the study sample.36,48–50 As shown in Table 2, most studies (73%) reported descriptive data (means and standard deviations) on the measures of self-concept or personality. Other noteworthy strengths

related to the analysis included adjusting for multiple comparisons,33,49,56 careful matching of control samples,33–35,53,54,56 and conducting a statistical comparison between participants retained and those lost to follow-up.55 Overall, because of the significant sample heterogeneity and variability in the study design and measurement of self-identity, meta-analysis was not used to synthesize the findings. The findings of each study are instead described qualitatively in Table 1. The following section describes the sample characteristics and assessment approaches used in the studies. Sample characteristics The majority of studies were conducted within the United Kingdom.36,48–50,52 Other countries include Australia,33,34,53,54 the United States,40,56 Hong Kong,57 and Belgium.35 Eleven studies were conducted in a community outpatient setting, whereas the other 4 did not specify the setting.35,36,50,56 Sample sizes varied from 8 to 865 participants, with most studies describing the use of convenience sampling.34–36,40,48–50,52,53,57 Five studies used a more representative sampling approach such as consecutive hospital admissions.33,36,51,54,55 Four studies included a brain injury sample with mixed causes.47,50,51,53 Severity of TBI varied within and between studies; 2 mainly recruited people with mild TBI,55,56 4 predominantly included people with severe TBI,36,48,52,54 and 5 recruited people with mild to severe TBI.33,35,40,49,50,57 There was also considerable heterogeneity in chronicity within and between studies (eg, 2-15 months postinjury36 ; 2-40 years postinjury35 ). Measurement of self-identity change Table 3 summarizes the 10 measures used to investigate the impact of TBI on self-identity, including measures of self-concept change, current self-concept, and personality characteristics. Information concerning reliability and validity is included where available. Measures of self-concept change The most commonly used measure of pre-/postinjury self-discrepancy was the Head Injury Semantic Differential Scale (HISD).36 The original HISD was specifically designed to capture subjective changes to sense of self in individuals with brain injury. Three studies36,52,55 used the original version of the HISD. Two revised versions of the HISD have subsequently been developed, which entailed minor wording changes.4 Andrewes et al48 used the second version (HISD-II), whereas the most recent version (HISD-III) was used by Carroll and Coetzer.49 Cantor et al40 developed a modified version of the Selves Interview58 and Selves Adjective Checklist (SAC59 ). The adjective checklist had a forced-choice

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

8 patients with severe TBI (aged 16-57 y); chronicity = 6 mo; community 74 patients with mild to severe TBI (77% TBI; aged 16-54 y); chronicity = 6-30 mo; community 20 patients with mixed brain injury (85% TBI; mean age = 32.45 y); chronicity: M = 15.09 (SD = 13.66); community 50 “minor” patients with TBI (aged 17-86 y); chronicity = 6 mo at time 1; community

68 patients with moderate to severe TBI (aged 15-65 y); chronicity: 6 mo; setting not specified 88 patients with severe TBI (aged 14-75 y); chronicity = 1-5 y; community 120 patients with mixed severity TBI (aged 19-65 y); chronicity not reported; community

McWilliams (1991)52 /UK

Lannoo et al (1997)35 /Belgium

Man et al (2003)52 /Hong Kong

Curran et al (2000)54 /Australia

Wright and Telford (1996)55 /UK

Crisp (1996)53 /Australia

Malia et al (1995)51 /UK

25 patients with severe TBI (aged 17-34 y); chronicity = 2-15 mo; setting not specified

Tyerman and Humphrey (1984)36 /UK

TBI sample characteristics and setting

Between-group: TBI vs 40 gender- and chronicity-matched orthopedic controls Between-group: TBI vs 146 healthy, nonmatched controls

Self-concept (Chinese version of ASSEI)

Self-esteem (RSES)

Personality (NEO-FFI)

Self-concept change (HISD)

Longitudinal: 36 complete HISD at time 1; 21 at 3 y (n = 21) postinjury Between-group: TBI vs 28 age-, gender-, and education-matched trauma controls

Self-esteem (SEI)

Personality (LOT; easygoing disposition)

Semantic Differential Scale

Self-concept (20-item Semantic Differential Scale)

Between-group: TBI vs disability groups matched on age, education, and unemployment

Cross-sectional cohorts at 6, 12, 18, 24, and 30 mo postinjury

Cross-sectional

Cross-sectional

Design

Self-identity, self-concept, or personality measures Key significant findings

No differences between the TBI and orthopedic control groups. Lower self-esteem was associated with higher depression and anxiety. TBI participants were more satisfied than controls in their law-abidingness, spiritual convictions, and sense of culture. They were less satisfied in their physical abilities, grooming, and appearance. (continues)

Present self was rated more negatively than preinjury self. Return to past self was anticipated. Past/present and past/future self-discrepancies were related to psychological distress and persisted at 2.5 y of follow-up. No differences between the TBI and trauma control groups on personality dimensions.

Postinjury personality change was reported on all measures at 6 and 12 mo; on “easygoing disposition” at 18 mo; on all measures except optimism at 24 mo postinjury changes. No personality changes were reported at 30 mo. Individuals with chronic pain and other physical conditions reported higher self-esteem than those with brain injury or a psychiatric condition.

Present self was rated as more negative than past self. Individuals expected to return to preinjury self within 1 y. Present self was viewed less positively than a “typical” person and more positively than a “typical head-injured” person. Current self-concept was viewed more negatively than preinjury self. Return to preinjury self was expected within 1 y.

Participant characteristics, methodology, and key findings of included studies

Authors/country

TABLE 1

The Impact of TBI on Self-Identity E17

www.headtraumarehab.com

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

20 patients with mild TBI (mean age = 37.21 y), 39 moderate/severe (mean age = 38.75 y); chronicity: “a few” months; setting not specified 22 patients with mixed brain injury (73% TBI; aged 18-70 y); chronicity = 16-348 mo; setting not specified 29 patients with mild to severe TBI (aged 22-64 y); chronicity: 2-40 y; community

865 patients with mild to severe TBI (mean age = 34.7 y); chronicity = 1–20 y; community 41 patients with mixed brain injury (83% TBI; aged 18-73 y); chronicity = 1-20 y; community 10 patients with severe TBI (mean age = 38.3 y); chronicity (M = 4.2 y); outpatient

Rush et al (2006)56 /USA (Florida)

Downing et al (2013)33 /Australia

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. Self-concept change (HISD-III)

Cross-sectional

Self-concept (TSCS-2) and self-esteem (RSES) Self-concept change (HISD-II)

Between-group: TBI vs 41 age-, gender-, and education-matched healthy controls Within-group (preintervention data)

Self-esteem (RSES)

Self-esteem (RSES)

Cross-sectional with a repeated-measures component (reassessment after 2 wk)

Between-group: TBI vs 142 age- and gender-matched healthy controls

Personality (NEO-PI-R)

Self-discrepancies (SI and SAC)

Within- and between-group: TBI vs 31 gender- and race-matched orthopedic controls; 1- to 2-y of follow-up

Cross-sectional

Design

Self-identity, self-concept, or personality measures

Individuals with TBI reported poorer self-esteem and global self-concept than matched controls. Lower self-concept and self-esteem were correlated with higher levels of anxiety and depression. No difference between ratings of past, current, and future self.

Current self-esteem was rated lower than preinjury self-esteem. Higher current self-esteem was associated with poorer cognitive and executive functioning and lower self-awareness. Current self-esteem ratings showed temporal stability over 2-wk interval. Current self was more negative than preinjury self. Negative self-discrepancy was associated with depression, grief, and low self-esteem. Higher current self-esteem was related to poorer awareness. Individuals with TBI reported poorer global self-esteem than matched healthy controls.

On average, individuals reported negative changes to self after TBI. Negative associations between depression, anxiety, and self-discrepancies on the SAC but not the SI. No differences between TBI and controls on personality scales. No consistent evidence of change in pre- and postinjury personality ratings for either the TBI or control group. No differences between premorbid personality and current personality.

Key significant findings

Abbreviations: ASSEI, Adult Source of Self-concept Inventory; HISD, Head Injury Semantic Differential Scale; LOT, Life Orientation Test; NEO-FFI, NEO Five-Factor Inventory; NEO-PI-R, NEO Personality Inventory–Revised; RSES, Rosenberg Self-Esteem Scale; TBI, traumatic brain injury; SAC, Selves Adjective Checklist; SEI, Self-esteem Inventory; SI, Selves Interview; TSCS-2, Tennessee Self Concept Scale, Second Edition.

Andrewes et al (2014)48 /UK

Kelly et al (2013)47 / Australia Ponsford et al (2014)34 /Australia

Carroll and Coetzer (2011)49 /UK (Wales)

Cooper-Evans et al (2008)50 /UK

21 adults with mild to severe TBI (aged 22-77 y); chronicity: 2-33 y; community

Cantor et al (2005)40 /USA (New York)

TBI sample characteristics and setting

Participant characteristics, methodology, and key findings of included studies (Continued)

Authors/country

TABLE 1

E18 JOURNAL OF HEAD TRAUMA REHABILITATION/MARCH–APRIL 2016

The Impact of TBI on Self-Identity TABLE 2

E19

Ratings of methodological quality of included studies

Study

Measure Longitudinal validated for Representative and/or TBI and/or and adequate matched well validated sampling control in the general (n ≥30) design population

Andrewes et al48 Cantor et al40 Carrol and Coetzer49 Cooper-Evans et al50 Crisp53 Curran et al54 Downing et al33 Lannoo et al35 Malia et al51 Man et al57 McWilliams52 Ponsford et al34 Rush et al56 Tyerman and Humphrey36 Wright and Telford55

Reporting of mean and SD for self-concept in text, table, or graph

Sample homogeneity: All TBI and similar chronicity or severity

Total

a

0

0

1

1

0

2

0 0

0 0

0 1

1 1

0 0

1 2

0

0

1

1

0

2

0 1 1

1 1 1

1 1 1

1 1 1

0 1 0

3 5 4

0 1 0 0 0

1 1 0 0 1

1 0 1 1 1

1 0 1 0 1

1 0 0 1 0

4 2 2 2 3

1 0

1 0

1 1

0 0

1 1

4 2

1

1

1

1

1

5

Abbreviation: TBI, traumatic brain injury. a Confirmed through correspondence with the authors.

format, which was considered more suitable than openended questions for individuals with cognitive impairment. Measures of current self-concept Current global self-esteem was assessed using the Rosenberg Self Esteem Scale60 (RSES) in 4 studies.33,34,49,54 Cooper-Evans et al50 also asked participants to provide retrospective ratings of self-esteem on the RSES prior to their TBI and calculated the degree of discrepancy between preinjury and current self-esteem ratings. Two studies33,53 used other wellestablished measures to assess global self-esteem (ie, Coopersmith Self-Esteem Inventory61 ) or multiple selfconcept domains (Tennessee Self Concept Scale, Second Edition62 ). Although the psychometric properties of these tools are well established in the general population, reliability and validity for the TBI population have only been investigated for the RSES (see Table 3). Measures of personality Two studies35,56 used different versions of the NEO personality inventory, which is based on the 5-factor

model of personality.63 In contrast, Malia et al51 used multiple measures of psychosocial functioning to assess perceived changes (ie, pre- and postinjury ratings), 2 of which were measures of personality—the Life Orientation Test,64 and the Easygoing dimension of the Health and Daily Living Form.65 These personality measures have not been validated for use in the TBI population, although internal consistency was reported by Malia et al51 for their sample. In summary, the impact of TBI on current selfconcept and changes in self-concept was examined using 10 different self-report measures across the 15 studies, with the most common measure involving the HISD. Evidence of self-identity change after TBI As summarized in Table 1, 11 of the 15 studies provided evidence of changes to self-concept after TBI.33,34,36,40,49–53,55,57 Such studies demonstrated on at least one measure that individual’s current selfconcept was viewed more negatively than their preinjury self-concept36,40,48–52,55 or that self-concept was more negative for individuals with TBI than for matched controls.33,34,53,57 www.headtraumarehab.com

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

HISD-I and HISD-II: 20 items; HISD-III: 18 items

20 items; importance of and satisfaction with various aspects of self-esteem

25 items

ASSEI67

CSEI61

At least 6 descriptors for “current actual self,” “current ideal self,” and “current ought self” 38 items; different Selves Adjective versions for “current Checklist40,59 actual self,” “current ideal self,” and “current ought self” Measures of current self-concept or self-esteem

Selves Interview40,58

HISD (I-III)36

Measures of self-concept change (self-discrepancy)

Items, domains

Dichotomous scale

11-point scale (0-10)

Forced-choice bipolar adjective scale

Open-ended interview

7-point semantic differential scale of bipolar adjective pairs

Scale

Rate the importance they attach to and satisfaction with various aspects of self-concept. The scale covers physical, social, ethical, familial, and intellectual aspects of self-concept. Respond to statements by endorsing “Like me” or “Unlike me.” Scores range from 0 to 100.

Generate at least 6 descriptors for each self and rate the degree of applicability (1 = slightly true of me to 5 = extremely true of me). Select from a list of adjectives and indicate which adjective is most descriptive of each self.

Rate each “self” (past/present/ future) according to bipolar adjective pairs (eg, “bored” to “interested,” “unhappy” to “happy,” and “inactive” to “active”).

Administration (respondent’s task)

Quantitative approaches to measuring the impact of TBI on self-identity

Instrument

TABLE 3

Reliability not examined in TBI. In general population: test-retest reliability = 0.67-0.69; total ASSEI satisfaction scores were related to the RSES (r = 0.37; P < .001).68 Reliability not examined in TBI. In general population: split half reliabilities = 0.71-0.74; test-retest reliability = 0.80-0.82.69 (continues)

Reliability not examined. More negative self-discrepancy was associated with depression and anxiety.42

HISD-I and HISD-II: Internal consistency: α = .88-.93; split half reliabilities = 0.87-0.93; convergent validity with Frankfurt Self-Concept Scale. HISD-III: internal consistency: α = .92-.93; convergent validity with the RSES and the Brain Injury Grief Inventory66 Reliability not examined in TBI. In general population: 3-y test-retest = 0.56; interrater = 0.86

Reliability and validity in TBI samples

E20 JOURNAL OF HEAD TRAUMA REHABILITATION/MARCH–APRIL 2016

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

82 items

TSCS-262

5-point scale

5-point scale

5-point scale

5-point Likert scale

4-point Likert scale

Scale

Rate the extent to which they agree with each item, from 0 (strongly disagree) to 4 (strongly agree). Higher values (0-24) indicate greater optimism. Rate on a 5-point scale (strongly disagree to strongly agree) their agreement with statements regarding the domains of Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness to Experience. Rate how well each adjective (eg, easygoing, calm, happy) describes one’s self on a 5-point scale (“not at all accurately” to “quite accurately”).

Rate how true each statement is about one’s self. The total summary score for self-concept comprises 6 domains: physical, moral, personal, family, social, and academic/work. Items are rated from 1 (always false) to 5 (always true).

Respond to questions regarding feelings about self, from 1 (strongly disagree) to 4 (strongly agree).

Administration (respondent’s task)

TBI: internal consistency = 0.53-0.7851 (authors did not differentiate between participantand relative-rated versions)

Reliability not examined in TBI. In general population: internal consistency of NEO-PI-R63 = 0.86-0.92; NEO-FFI63 = 0.79-0.83.

In general population: internal consistency = 0.7664 ; internal consistency in TBI sample = 0.7070

High internal consistency (α = .89)50 ; convergent validity with HISD49 ; test-retest reliability = 0.86. Scores negatively associated with anxiety and depression.50 Reliability not examined in TBI. In general population: internal consistency = 0.73-0.95; test-retest reliability = 0.47-0.82; convergent validity with the RSES in TBI sample.34

Reliability and validity in TBI samples

Abbreviations: ASSEI, Adult Source of Self-concept Inventory; CSEI, Coopersmith Self-esteem Inventory; HISD, Head Injury Semantic Differential Scale; NEO-FFI, NEO Five-Factor Inventory; NEO-PI-R, NEO Personality Inventory–Revised; RSES, Rosenberg Self-Esteem Scale; TBI, traumatic brain injury; TSCS-2, Tennessee Self Concept Scale, Second Edition.

3 items

NEO PI-R: 240 items NEO-FFI: 60 items

NEO (PI-R and -FFI)63

Health and Daily Living Form65 – Easygoing dimension

12 items: 4 filler, 4 optimistic, 4 pessimistic

Life Orientation Test64

Measures of personality

10 items

RSES60

Items, domains

Quantitative approaches to measuring the impact of TBI on self-identity (Continued)

Instrument

TABLE 3

The Impact of TBI on Self-Identity E21

www.headtraumarehab.com

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

E22

JOURNAL OF HEAD TRAUMA REHABILITATION/MARCH–APRIL 2016

In the only study to examine both global and domainspecific self-concept, Ponsford et al34 found that TBI participants rated their global self-concept and social, family, academic, and personal self-concept as significantly lower than healthy controls. There were no significant differences on physical or moral self-concept. However, the findings of 3 studies35,54,56 with strong methodology (4-5/5; see Table 2) indicated that changes in self-concept or personality characteristics were not specific to TBI. Specifically, they reported no significant differences in personality or self-esteem ratings between their TBI and orthopedic/trauma control samples. Furthermore, Rush et al56 found no significant changes in pre- and postinjury personality ratings for either the TBI or orthopedic control sample. A study by Andrewes et al48 also did not find significant differences between ratings of pre- and postinjury self-concept for their TBI sample; however, this may have been due to the small sample size (n = 10). The 7 studies that examined the impact of TBI on different emotional, behavioral, or personality attributes (ie, on separate measures, subscales, or item-by-item analysis) revealed that not all changes were negative and that particular traits were stable.34,36,49,51,52,56,57 For example, Tyerman and Humphrey36 found that individuals perceived themselves as just as friendly as before their injury, and Malia et al51 reported that ratings of optimism were similar to preinjury levels for individuals with TBI at 18, 24, and 30 months postinjury. In addition, positive changes were evident in some studies for particular attributes; for example, people with TBI rated themselves as more mature and appreciative since their injury,52 or were more satisfied than controls in their law-abidingness, spiritual convictions, and sense of culture.57 In summary, despite methodological differences, a consistent finding across 11 studies is that people with TBI mainly reported negative changes in selfconcept. The longitudinal study55 found that negative self-discrepancies persisted at 2.5 years of follow-up. Three studies with strong methodology found no significant differences in personality characteristics or selfesteem between TBI and orthopedic/traumatic control samples.35,54,56 Factors related to self-identity change Three studies investigated associations between self-concept change and measures of emotional adjustment.40,49,55 Two studies found that negative self-discrepancy was related to greater depressive symptoms.40,49 One study reported significant associations among measures of grief, adjustment, self-esteem, and self-concept change, with negative self-discrepancy related to increased sense of loss and poorer adjust-

ment and self-esteem.49 Furthermore, negative selfdiscrepancy was associated with increased anxiety, mental health concerns, and psychological distress in 2 studies.40,55 Most studies did not examine neurocognitive correlates of current self-concept or self-discrepancy. However, 2 studies found that higher current self-esteem was related to lower IQ, poorer executive functioning, and greater awareness deficits.49,50 Carroll and Coetzer49 found that self-concept change (ie, self-discrepancy) was not significantly related to awareness of deficits (although there was a trend in the expected direction), which may have been partly due to the small sample size (n = 29) and associated power issues.

DISCUSSION The aims of this review were to examine the impact of TBI on self-identity and investigate factors related to changes in self-concept. Overall, 11 of the 15 studies indicated that TBI has a predominantly negative impact on self-identity.33,34,36,40,49–53,55,57 Therefore, despite differences in sampling characteristics and approaches to measurement, the review yielded largely consistent findings. However, there was also evidence of stability in particular traits (eg, friendliness and optimism) and positive psychological changes (eg, more mature and appreciative) after TBI.34,36,49,51,52,56,57 Importantly, 3 of the 4 studies that indicated no significant impact of TBI on self-identity were those that compared self-esteem or personality characteristics of participants with TBI with orthopedic/trauma (ie, not neurological) controls.35,54,56 Such findings suggest that changes in self-identity after TBI do not arise from neurological damage per se but are related to the broader psychosocial consequences of a traumatic event leading to hospitalization and adjustment to injury or loss of function. This explanation is further supported by the finding that personality characteristics did not differ between participants with mild and moderate/severe TBI.56 Nonetheless, firm conclusions about neurological influences on self-identity cannot be drawn from this review due to the lack of studies investigating neurocognitive factors related to self-concept changes. Intriguingly, higher levels of current self-esteem were related to lower IQ, poorer executive functioning, and greater awareness deficits.49,50 These findings provide tentative support for the Ownsworth theoretical proposition that people with more severe cognitive deficits and reduced self-awareness are less likely to make global negative selfevaluations. However, because of the methodological issues associated with these 2 studies (ie, cross-sectional design and small and heterogeneous samples), caution is needed in interpreting these findings.

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

The Impact of TBI on Self-Identity Three studies identified that negative self-discrepancy was related to greater emotional distress and mental health problems.40,49,55 These findings support the “Y-shaped” model,27 which proposes that discrepancies between people’s current self and aspired-to self (the top of the “Y”) pose threats to self and result in unhelpful coping reactions and emotional distress. According to this model, rehabilitation supports people to resolve negative self-discrepancies (convergence at the center of the “Y”) by the therapist and the client forming a strong alliance and engaging in experiential learning processes (eg, goal setting, planning, performing activities, and reflecting on performance) that promote more adaptive and realistic self-representations (the vertical trunk of the “Y”). Positive identity transition after TBI is proposed to entail assimilation of both continuous and changed aspects of self (ie, the “old” and “new” me) and consolidating the updated identity through regular participation in meaningful activities.16,30 This framework depicts self-identity change as process that occurs over time and the one that is strongly influenced by a person’s social environment. This review identified that most studies used a crosssectional design with measures of self-concept change based upon retrospective ratings of preinjury self, obtained many years postinjury. Two noteworthy studies reported retrospective ratings of preinjury self within a month postinjury and compared these with ratings of postinjury self at 2 to 3 years postinjury.55,56 One of these studies55 reported negative self-discrepancy on the HISD, whereas the other study58 reported no significant difference in pre- and postinjury personality ratings. In future research, there is a need for prospective longitudinal studies to investigate changes in self-concept during the early phase of recovery (eg, rehabilitation admission and discharge) and between the postacute and long-term adjustment phases. Vickery and colleagues24 found significant improvements in self-esteem during acute stroke rehabilitation, with younger patients and those with higher levels of functional status at admission reporting greater improvement. A recent systematic review29 identified largely mixed findings concerning the impact of rehabilitation on self-concept after TBI. Interestingly, despite minimal overlap with studies included in the present review, the same issue emerged in their review concerning lack of consistency in approaches to measuring self-concept change. To achieve greater consistency in measurement of self-identity after TBI, the use of common terminology would be beneficial. Ponsford et al34 distinguished between global self-esteem and multidimensional aspects of self-concept and emphasized the need for clinicians to consider relevant domains of self-concept to address as a focus of intervention. Ownsworth and Haslam29

E23

similarly recommended that clinicians and researchers be guided by developmental models of self-concept to select assessment and intervention approaches. In particular, they presented a hierarchical framework in which dynamic relationships were proposed between (a) lower level self-evaluations (ie, self-perceptions of competency in specific situations), (b) domain-specific selfconcepts (eg, physical attributes), and (c) higher-level self-representations (ie, global self-concept). Given that it is not practical to administer several tools to assess similar constructs, there is a need to consider both the level and domain(s) of self-concept most relevant to assess in clinical practice or to address particular research questions. There are a number of conceptual and methodological issues related to this review that may limit the conclusions drawn. In particular, self-identity was broadly defined in terms of self-perceived personal characteristics. Self-reported personality attributes were considered to fall within this general definition, as both personality and self-identity refer to enduring characteristics within the individual. However, there is a distinction between these constructs: others can observe and describe someone else’s personality but not his or her self-identity, because this can only be subjectively construed.4 Related to this definitional issue, studies investigating the impact of TBI on social identity were not included. Levack and colleagues26 emphasized the need to assess social and cultural aspects of identity. Development of validated measures of social identity after TBI represents a priority for future research. Overall, the rigorous selection criteria identified a relatively small number of studies that were too heterogeneous to use meta-analytic techniques. Despite this, the key findings largely concur with themes derived from qualitative research.30,31 The present review contributes to the literature by improving conceptual clarity regarding terminology and appraising approaches to measuring the impact of TBI on self-identity. Prospective longitudinal studies are recommended to investigate the process through which self-concept changes over time after TBI and in response to rehabilitation. CONCLUSIONS Overall, this review identified that people with TBI typically experience negative self-discrepancy and poorer self-concept than healthy controls. Levels of self-esteem and personality characteristics were similar between participants with TBI and orthopedic/trauma controls, thus suggesting that the neurological effects of TBI may not be as influential as the psychosocial consequences of a traumatic injury. Greater consistency in self-identity measurement and prospective longitudinal research would advance the field. www.headtraumarehab.com

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

E24

JOURNAL OF HEAD TRAUMA REHABILITATION/MARCH–APRIL 2016

REFERENCES 1. James W. Principles of Psychology. New York, NY: Henry Holt; 1890. 2. Brinthaupt T, Lipka R. The Self: Definitional and Methodological Issues. Albany, NY: State University of New York; 1992. 3. Mathews DJH, Bok H, Rabins PV. Personal Identity and Fractured Selves: Perspectives From Philosophy, Ethics, and Neuroscience. Personal Identity and Fractured Selves: Perspectives from Philosophy, Ethics, and Neuroscience. Baltimore, MD: Johns Hopkins University Press; 2009. 4. Ownsworth T. Self-Identity After Brain Injury. New York, NY: Psychology Press; 2014. 5. Leary M, Tangney J. Handbook of Self and Identity. Vol 2. New York, NY: The Guilford Press; 2012. 6. Jetten J, Haslam C, Haslam SA. The Social Cure: Identity, Health and Well-being. New York, NY: Psychology Press; 2012. 7. Legrand D, Ruby P. What is self-specific? Theoretical investigation and critical review of neuroimaging results. Psychol Rev. 2009;116(1):252–282. 8. Hogg M, Terry D. Social identity and self-categorization processes in organizational contexts. Acad Manage Rev. 2000;25(1):121–140. 9. Harter S. The Construction of the Self. 2nd ed. New York, NY: Guilford Press; 2012. 10. Marsh H. Age and sex effects in multiple dimensions of selfconcept: preadolescence to early adulthood. J Educ Psychol. 1989;81:417–430. 11. Schweitzer RD, Seth-Smith M, Callan V. The relationship between self-esteem and psychological adjustment in young adolescents. J Adolesc. 1992;15:83–97. 12. Marsh H, Byrne B, Shavelson R. A multidimensional, hierarchical self-concept. In: Brinthaupt T, Lipka R, eds. The Self: Definitional and Methodological Issues. Albany, NY: State University of New York; 1992. 13. Markus H, Kunda Z. Stability and malleability of the self-concept. J Pers Soc Psychol. 1986;51(4):858–866. 14. Orth U, Trzesniewski K, Robins R. Self-esteem development from young adulthood to old age: a cohort-sequential longitudinal study. J Pers Soc Psychol. 2010;98:645–658. 15. Trzesniewski K, Donnellan M, Robins R. Stability of self-esteem across the life span. J Pers Soc Psychol. 2003;84:205–220. 16. Rathbone CJ, Moulin CJ, Conway MA. Autobiographical memory and amnesia: using conceptual knowledge to ground the self. Neurocase. 2009;15(5):405–418. 17. Feinberg TE. Neuropathologies of the self: clinical and anatomical features. Conscious Cogn. 2011;20:75–81. 18. Caddell L, Clare L. The impact of dementia on self and identity: a systematic review. Clin Psychol Rev. 2010;30(1):113–126. 19. Higgins ET, Bond R, Klein R, Strauman T. Self-discrepancies and emotional vulnerability: how magnitude, accessibility, and type of discrepancy influence affect. J Pers Soc Psychol. 1986;51(1):5–15. 20. Ylvisaker M, McPherson K, Kayes N, Pellett E. Metaphoric identity mapping: facilitating goal setting and engagement in rehabilitation after traumatic brain injury. Neuropsychol Rehabil. 2008;18(5/6):713–741. 21. Beck A. The current state of cognitive therapy: a 40-year retrospective. Arch Gen Psychiatry. 2005;62:953–959. 22. Harris S, Morley S, Barton S. Role loss and emotional adjustment in chronic pain. Pain. 2003;105:363–370. 23. Fitting M, Salisbury S, Davies N, Mayclin D. Self-concept and sexuality of spinal cord injured women. Arch Sex Behav. 1978;7(2):143–156. 24. Vickery C, Evans C, Lee J, Sepehri A, Jabeen L. Multilevel modeling of self-esteem change during acute inpatient stroke rehabilitation. Rehabil Psychol. 2009;54:372–380.

25. Munstedt K, Manthey N, Sachsee S, Vahrson H. Changes in self-concept and body image during alopecia induced cancer chemotherapy. Support Care Cancer. 1997;5:139–143. 26. Levack WMM, Boland B, Taylor W, et al. Establishing a person-centred framework of self-identity after TBI: a grounded theory study to inform measure development. BMJ Open. 2014;4:e004630. 27. Gracey F, Ownsworth T. Editorial. Neuropsychol Rehabil. 2008;18(5/6):522–526. 28. Wilson B. ‘Cutting Edge’ Developments in neuropsychological rehabilitation and possible future directions. Brain Impairment. 2011;12(1):33–42. 29. Ownsworth T, Haslam C. Impact of rehabilitation on self-concept following traumatic brain injury: an exploratory systematic review of intervention methodology and efficacy [published online ahead of print November 10, 2014]. Neuropsychol Rehabil. doi:10.1080/09602011.2014.977924. 30. Levack WM, Kayes NM, Fadyl JK. Experience of recovery and outcome following traumatic brain injury: a metasynthesis of qualitative research. Disabil Rehabil. 2010;32(12):986–999. 31. Bryson-Campbell M, Shaw L, O’Brien J, Holmes J, Magalhaes L. A scoping review on occupational and self-identity after a brain injury. Work. 2013;44(1):57–67. 32. Jumisko E, Lexell J, Soderberg S. The meaning of living with traumatic brain injury in people with moderate or severe traumatic brain injury. J Neurosci Nurs. 2005;37(1):42–50. 33. Downing MG, Stolwyk R, Ponsford JL. Sexual changes in individuals with traumatic brain injury: a control comparison. J Head Trauma Rehabil. 2013;28(3):171–178. 34. Ponsford J, Kelly A, Couchman G. Self-concept and self-esteem after acquired brain injury: a control group comparison. Brain Inj. 2014;28(2):146–154. 35. Lannoo E, de Deyne C, Colardyn F, de Soete G, Jannes C. Personality change following head injury: assessment with the NEO Five-Factor Inventory. J Psychosom Res. 1997;43(5):505– 511. 36. Tyerman A, Humphrey M. Changes in self-concept following severe head injury. Int J Rehabil Res. 1984;7(1):11–23. 37. Sherer M, Bergloff P, Boake C, High W Jr, Levin E. The Awareness Questionnaire: factor structure and internal consistency. Brain Inj. 1998;12(1):63–68. 38. Festinger L. A Theory of Cognitive Dissonance. Stanford, CA: Stanford University Press; 1962. 39. Heider F. The Psychology of Interpersonal Relations. New York, NY: Wiley; 1958. 40. Cantor J, Ashman T, Schwartz M, et al. The role of selfdiscrepancy theory in understanding posttraumatic brain injury affective disorders: a pilot study. J Head Trauma Rehabil. 2005;20(6): 527–543. 41. Gracey F, Evans JJ, Malley D. Capturing process and outcome in complex rehabilitation interventions: a “Y-shaped” model. Neuropsychol Rehabil. 2009;19(6):867–890. 42. Turner B, Fleming J, Ownsworth T, Cornwell PL. Perceptions of recovery during the early transition phase form hospital to home following acquired brain injury: a journey of discovery. Neuropsychol Rehabil. 2011;21(1):64–91. 43. Silva J, Ownsworth T, Shields C, Fleming J. Enhanced appreciation of life following acquired brain injury: posttraumatic growth at 6 months postdischarge. Brain Impairment. 2011;12(2):93–104. 44. Ownsworth T, Clare L. The association between awareness deficits and rehabilitation outcome following acquired brain injury. Clin Psychol Rev. 2006;26(6):783–795.

Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

The Impact of TBI on Self-Identity 45. Agnew S, Morris R. The heterogeneity of anosognosia for memory impairment in Alzheimer’s disease: a review of the literature and a proposed model. Aging Ment Health. 1998;2(1):7–19. 46. von Elm E, Altman DG, Eggers M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The strengthening of reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370:1453– 1457. 47. Kelly A, Ponsford J, Couchman G. Impact of a family-focused intervention on self-concept after acquired brain injury. Neuropsychol Rehabil. 2013;23(4):563–579. 48. Andrewes HE, Walker V, O’Neill B. Exploring the use of positive psychology interventions in brain injury survivors with challenging behaviour. Brain Inj. 2014;28(7):965–971. 49. Carroll E, Coetzer R. Identity, grief and self-awareness after traumatic brain injury. Neuropsychol Rehabil. 2011;21(3):289– 305. 50. Cooper-Evans S, Alderman N, Knight C, Oddy M. Self-esteem as a predictor of psychological distress after severe acquired brain injury: an exploratory study. Neuropsychol Rehabil. 2008;18(5/6): 607–626. 51. Malia K, Powell G, Torode S. Personality and psychosocial function after brain injury. Brain Inj. 1995;9(7):697–712. 52. McWilliams S. Affective changes following severe head injury as perceived by patients and relatives. Br J Occup Ther. 1991;54(7):246–248. 53. Crisp R. Community integration, self-esteem, and vocational identity among persons with disabilities. Aust Psychol. 1996;31(2):133– 137. 54. Curran C, Ponsford J, Crowe S. Coping strategies and emotional outcome following TBI: a comparison with Orthopedic patients. J Head Trauma Rehabil. 2000;15(6):1256–1274. 55. Wright JC, Telford R. Psychological problems following minor head injury: a prospective study. Br J Clin Psychol. 1996;35(3): 399–412. 56. Rush BK, Malec JF, Brown AW, Moessne AM. Personality and functional outcome following traumatic brain injury. Rehabil Psychol. 2006;51(3):257–264.

E25

57. Man DW, Tam AS, Li EP. Exploring self-concepts of persons with brain injury. Brain Inj. 2003;17(9):775–788. 58. Strauman TJ. Self-guides and emotionally significant childhood memories: a study of retrieval efficiency and incidental negative emotional content. J Pers Soc Psychol. 1990;59:869–880. 59. Tangney J, Niendenthal P, Covert M, Barlow D. Are shame and guilt related to distinct self-discrepancies? A test of Higgins’ (1987) hypotheses. J Pers Soc Psychol. 1998;75:256–268. 60. Rosenberg M. Society and the Adolescent Self Image. Princeton, NJ: Princeton University Press; 1965. 61. Coopersmith S. Self-esteem Inventories. Palo Alto, CA: Consulting Psychologists Press; 1989. 62. Fitts WH, Warren WL. Tennessee Self-concept Scale, TSCS:2 (2nd Edition) [Manual]. Los Angeles, CA: Western Psychological Services; 1996. 63. Costa P, McCrae R. Revised NEO Personality Inventory (NEO-PIR) and NEO Five-Factor Inventory (NEO-FFI) Manual. Odessa, FL: Psychological Assessment Resources; 1992. 64. Scheier M, Carver C. Optimism, coping, and health: assessment and implications of outcome expectancies. Health Psychol. 1985;4:219–247. 65. Holahan C, Moos R. Life stress and health. Personality coping and family support. J Pers Soc Psychol. 1985;49:739–747. 66. Coetzer R, Vaughn F, Ruddle J. The Brain Injury Grief Inventory [unpublished manuscript]. Bangor, England: North Wales Brain Injury Service, Conwy & Denbighshire NHS Trust; 2003. 67. Elovson A, Fleming J. Rationale for Multidimensional Self-esteem Scale Scoring and Weighting [unpublished manuscript]. Long Beach, CA: California State University; 1989. 68. Davis-Zinner N. Source Differences in Male and Female Self-esteem. Los Angeles, CA: California School of Professional Psychology; 1990. 69. Bedian A, Teague R, Zmud R. Test-restest reliability and internal consistency of a short-form of Coopersmith’s Self Esteem Inventory. Psychol Rep. 1977;41:1041–1042. 70. Peleg G, Barak O, Harel Y, Rochberg J, Hoofien D. Hope, dispositional optimism and severity of depression following traumatic brain injury. Brain Inj. 2009;23(10):800–808.

www.headtraumarehab.com Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.