Abstract
A study of people with TBI in Japan suggests that interventions for apathy and disinhibition, as well as management of psychiatric conditions, may increase employment status among this population.
Researchers have investigated potential prognostic factors as they pertain to the employment status of people who have sustained a traumatic brain injury (TBI) and found that, in general, employment outcomes are associated with injury severity, preinjury occupational status, functional status, physical condition, cognitive function (intelligence, memory, and executive function), involvement in vocational rehabilitation services, emotional status, and demographic factors (age, education level, gender, and race; Ownsworth & McKenna, 2004; Saltychev et al., 2013; Scaratti et al., 2017).
In these studies, however, some potentially notable prognostic factors were not included, for example, psychiatric conditions, apathy, and disinhibition. The psychiatric condition of people who have sustained a TBI has attracted a great deal of attention (Fujii & Ahmed, 2002, 2014; Guerreiro et al., 2009; Lane-Brown & Tate, 2009; Ownsworth et al., 2011; Takahata et al., 2019 ; Zgaljardic et al., 2015; Zhang & Sachdev, 2003). Statistically speaking, the risk of psychiatric comorbidity among people with TBI is twofold compared with those without TBI (Yeh et al., 2020). Psychiatric diagnoses, however, have been considered in only a few studies of employment status (Catalano et al., 2006; DiSanto et al., 2019 ; Grauwmeijer et al., 2012). In particular, the association between post–TBI psychiatric conditions and employment status was addressed in only one study (Grauwmeijer et al., 2012). Another study reported that preinjury psychiatric disorders, but not postinjury psychiatric disorders, were related to employment outcome on the basis of their association with mood and cognitive and behavioral changes (Schönberger et al., 2011). One factor that may explain the dearth of research on this topic is the long latency period associated with the development of psychiatric conditions (Fujii & Ahmed, 2001 ; Takahata et al., 2019), which sometimes extends to 50 yr after TBI and has a mean latency period of 4.1 yr (SD = 6.6; Fujii & Ahmed, 2002). Although there might be other causes of late-life conditions, a direct association between these conditions and TBI has been investigated; one study found that increased τ deposits in the brain’s white matter is associated with late-onset neuropsychiatric symptoms after TBI (Takahata et al., 2019).
Apathy is a syndrome characterized by a lack of motivation and consequential reduced goal-directed behavior that is not attributable to an alteration of consciousness or a disturbance of intellect or emotion (Le Heron et al., 2019 ; Marin, 1991). Apathy is a common problem after TBI; it can have a major impact on activities of daily living (Arnould et al., 2013; Cattelani et al., 2008; Lane-Brown & Tate, 2009; Worthington & Wood, 2018). Again, few researchers have studied the issue of employment status and apathy in people with TBI.
Disinhibition is a lack of restraint manifested in impulsivity; disregard of others and social norms; aggressive outbursts; misconduct and oppositional behaviors; disinhibited instinctual drives, including risk-taking behaviors; and hypersexuality (Boller & Grafman, 2002). Disinhibition is one of several major problems people with TBI struggle with in the workplace (Wehman et al., 1989), but it has been addressed in only one study concerning employment status after TBI (Simpson & Schmitter-Edgecombe, 2002).
Given the lack of pertinent research, we investigated employment status after TBI by using psychiatric and neuropsychological assessments that might be associated with employment status. We hoped to shed light on the most effective rehabilitation methods. In this article, “psychiatric conditions” are limited to those included in the International Classification of Diseases, 11th Revision (ICD–11; World Health Organization, 2019), Section F06—mental disorders due to brain damage—because this study focused on mental health issues closely associated with TBI. We hypothesized that these psychiatric and neuropsychological data, together with data from the other factors, would better explain employment status than data from the other factors, for example, intelligence, memory function, or executive function.
Method
Participants
Ethical aspects of this study were reviewed and approved by the Human Research Ethics Committee at Ashikaga Red Cross Hospital, Ashikaga, Japan. Informed consent was obtained from all participants. Study participants of working age (18–65 yr) who had TBI were recruited from the outpatient cognitive dysfunction clinics associated with Ashikaga Red Cross Hospital and Edogawa Hospital (Tokyo, Japan) from March 2015 to March 2020. Both facilities specialize in the assessment and rehabilitation of people with cognitive dysfunction (e.g., TBI and cerebrovascular diseases). All study participants were Japanese.
Considering the long latency period for the development of psychiatric conditions and the nature of a cross-sectional study, we did not set a limit for the time since the onset of TBI among these participants. Patients who had developmental, psychiatric, cerebrovascular, or degenerative diseases before their TBI were excluded, as were those who had an acquired brain injury (e.g., cerebrovascular disease or another TBI) after the primary TBI. Of the 134 outpatients with TBI who met the above criteria, 6 who were medically unstable or had sustained a TBI in the preceding 6 mo were excluded. Also excluded were those who declined the assessment (n = 8) and those who had a missing medical chart during the acute stage after a head injury (n = 10). Thus, 110 people (103 at Ashikaga Red Cross Hospital and 7 at Edogawa Hospital) participated in this study; 79 (71.8%) were male, and 31 (28.2%) were female.
Assessment Methods
We included as many potential factors as possible that had previously been found to be associated with employment status after TBI (Table 1). The demographic factors we investigated were age, education level, years postinjury, gender, and preinjury regular employment status (employed/student vs. unemployed). Years postinjury were considered because the period from the head injury to the first visit at our clinics varied among participants in this cross-sectional study, although all participants had to be medically stable and ≥6 mo postinjury at the time of detailed neuropsychological assessment. Also considered were the severity of injury, which was determined using the duration of posttraumatic amnesia (Ponsford et al., 2016); physical condition, based on a FIM® Motor subscale related to walking or using a wheelchair (Grauwmeijer et al., 2017 ; Ketchum et al., 2012 ; Uniform Data System for Medical Rehabilitation, 1997 [FIM® is a trademark of the Uniform Data System for Medical Rehabilitation, a division of UB Foundation Activities, Inc.]); incidence of epilepsy (Vilkki et al., 1988); and incidence of psychiatric conditions after TBI.
Potential Prognostic Variables for Employment Status
The Walk/Wheelchair section of the FIM Motor subscale was administered and scored by an occupational therapist. Incidence of epilepsy was included because the incidence of epilepsy after TBI can be as high as 10% (Piccenna et al., 2017), and people with epilepsy have difficulty finding and maintaining employment (Smeets et al., 2007). The incidence of psychiatric conditions was also assessed for participants on their first visit to the clinics. Psychiatric conditions were determined according to the ICD–11, Section F06, “other mental disorders due to known physiological condition” (Table 2), by two neuropsychiatrists, each of whom had >15 yr of experience. Mild cognitive disorder due to brain damage (F06.7) was excluded because this condition involves the cognitive dysfunction resulting from TBI itself. These physical and psychiatric conditions were assessed at each participant’s first visit to the clinics.
Diagnostic Codes for Psychiatric Disorders Due to Brain Damage
World Health Organization (2019).
During the 1-mo period after their first visit to our clinics, all participants completed a battery of standardized neuropsychological tests that covered the following neuropsychological functions: verbal and performance intelligence (Machamer et al., 2005); episodic memory function (Anderson & Schmitter-Edgecombe, 2009); frontal lobe function (Simpson & Schmitter-Edgecombe, 2002), including executive function (Sorg et al., 2014); behavior in regard to disinhibition or impulsiveness; and apathy. The instruments are listed in Table 1. Verbal and performance intelligence was assessed by two speech therapists, and the other instruments were administered by two psychiatrists. All of the instruments were validated for Japanese speakers and have been proven valid and reliable for Japanese culture (Fujii, 2006; Iwata, 2006; Kashima, 2003 ; Kashima & Kato, 1995 ; Watanuki et al., 2002).
Disinhibition, or impulsiveness, was assessed using 5 disinhibition-related questions from the 20-item Behavioural Assessment of the Dysexecutive Syndrome (BADS) questionnaire (Kashima, 2003 ; Stuss & Benson, 1984, 1986; Wilson, 1996). Each item was scored on a 5-point scale, ranging from 0 (never) to 4 (very often), with a maximum total score of 20. Although the full BADS questionnaire has been validated and standardized in Japan, the 5 items we used in this study have not been standardized for use on their own. However, previous research has shown that the score derived from these 5 items correlated with the results of a morality judgment task (Funayama et al., 2019), reflecting that a higher score indicates a greater number of inappropriate behaviors. For neuropsychological tests that have age-corrected norms (verbal IQ, performance IQ, Rivermead Behavioural Memory Test, and Wisconsin Card Sorting Test), age-adjusted quotients or Z scores (the number of standard deviations from the control) were used for statistical analysis.
Employment Status Evaluation
Employment status during the period of the neuropsychological assessment was rated using a 3-point scale on which 3 = regular employment (working full time or working part time), 2 = welfare employment (employed as a person with disabilities or undergoing vocational training in the Japanese welfare employment system, for which a worker is paid under either system), and 1 = unemployment. This scale is based on the one used by the Public Employment Security Office in Japan and is similar to Palm et al.’s (2017) scale.
Data Analysis
No single variable had a correlation >.76 with any other variable, indicating that all variables were relatively independent; more specifically, the correlation of verbal IQ with performance IQ was .76 and that of motivation with disinhibition was .61. The correlations between all other variables were <.60.
We first used analysis of variance to determine whether there were any statistically significant differences in variables among the three levels of employment. Post hoc pairwise comparisons were then made using the Tukey–Kramer test (Tukey, 1949). We used Fisher’s exact test for categorical variables. We then investigated the combined influence of the variables for employment status after TBI. Because employment status was an ordinal variable, a linear discriminant regression analysis was used to determine factors that contribute to categorization of the three employment status levels. Variables significant at p < .05 in the analysis of variance or Fisher’s exact test were included in the linear discriminant regression analysis.
Because the five disinhibition questions had not been standardized in Japan, we repeated these analyses without this assessment. Likewise, because there was a long interval between injury and our assessment in some of the cases, we repeated these analyses using data only from participants <20 yr postinjury (n = 91) and <10 yr postinjury (n = 75). All analyses were performed using Excel 2010 with Mulcel add-in software (OMS Ltd., Tokyo, Japan). The significance level for these analyses was set at p < .05.
Results
Analysis of Variance
Table 3 presents the demographic factors, injury severity and health conditions, and neuropsychological assessments by participant employment status. Only 33 of the 110 participants (30.0%) were employed as a regular worker, which is slightly lower than the postinjury employment prevalence of 42.2% in a meta-analysis (Gormley et al., 2019). Seventeen participants (15.5%) had a psychiatric condition (8 had organic delusional disorder, 2 had organic hallucinosis, 2 had organic mood disorder, 2 had organic emotionally labile disorder, and 3 had organic anxiety disorder). The incidence of epilepsy was excluded from the further analysis because it did not differ among the three groups (Fisher’s exact test, p = .47). The other variables differed among the three groups (p < .05).
Analysis of Variance for Demographic Factors, Injury Severity and Health Conditions, Neuropsychological Assessments, and Employment Status
Note. N = 110. Dash indicates not applicable.
Linear Discriminant Regression Analysis
The linear discriminant regression analysis indicated that the incidence of psychiatric conditions after TBI most significantly contributed to the classification of participants into the three employment levels, followed by level of apathy, level of disinhibition, age, and years postinjury (Table 4). The results of the linear discriminant regression analysis produced ps of <.01 (i.e., p < .001, Wilks’s λ = .34) and correctly classified 72.0% of participants in the unemployment group, 66.7% of those in the welfare employment group, and 90.9% of those in the regular employment group. When conducted without the disinhibition level data, the multiple linear discriminant regression model showed similar results, in which the incidence of psychiatric conditions after TBI, level of apathy, age, and preinjury employment status contributed significantly to the classification in this order (p < .001, Wilks’s λ = .37). When repeated using data only from participants who were <20 yr postinjury (n = 91), it again showed similar results, in which the incidence of psychiatric diagnoses after TBI, level of apathy, and preinjury employment status significantly contributed to the classification (p < .001, Wilks’s λ = .39). In contrast, only the level of apathy (p = .064) had a tendency to contribute to the classification in the analysis when data from only participants <10 yr postinjury (n = 75) were included (p < .001, Wilks’s λ = .45).
Linear Discriminant Regression Analysis for Employment Status Levels
In summary, the five variables, namely, psychiatric diagnoses after TBI, level of apathy, disinhibition, age, and years postinjury, were most closely associated with the participants’ employment status.
Discussion
Our results suggest that the employment status of people with TBI is associated with five attributes, namely, apathy level, incidence of psychiatric conditions after TBI, disinhibition, age, and years postinjury. This is the first report to include both psychiatric and apathy assessments to evaluate post–TBI employment status. Although level of apathy has not been taken into account in previous studies of post–TBI employment status, apathy is a common problem after TBI and can have a major impact on cognitive function, psychosocial outcome, and engagement in rehabilitation (Worthington & Wood, 2018). Similarly, poststroke apathy is considered to negatively affect outcomes of acute rehabilitation (Harris et al., 2014). Given these reports and our current findings, careful assessment of the level of apathy after TBI is crucial when assessing patient outcomes, and treatment or interventions for apathy should be considered (Arnould et al., 2013; Cattelani et al., 2008; Lane-Brown & Tate, 2009). Particular jobs may enhance motivation for a person with TBI, although the right type of job most likely differs from person to person. Disinhibition also influenced employment status, as previous researchers have suggested (Simpson & Schmitter-Edgecombe, 2002 ; Wehman et al., 1989). Although it is difficult to manage disinhibition, a psychotherapeutic approach (e.g., cognitive–behavioral therapy) may be effective in some cases (Demily & Franck, 2016).
Psychiatric conditions after TBI are considered severe sequelae for these patients (Fujii & Ahmed, 2002, 2014; Guerreiro et al., 2009; Lane-Brown & Tate, 2009; Ownsworth et al., 2011; Takahata et al., 2019 ; Zgaljardic et al., 2015; Zhang & Sachdev, 2003). In the current study, only 1 of the 17 patients with a psychiatric condition (5.9%) was regularly employed, and only 3 (17.6%) were employed in a welfare employment system. Having psychiatric disorders alone, without traumatic brain damage, hinders employment prospects (Hakulinen et al., 2020 ; Mattila-Holappa et al., 2016). These results indicate that management of psychiatric conditions is needed if employment status is to be enhanced (Shields et al., 2016).
Although preinjury regular employment status and injury severity based on duration of posttraumatic amnesia were not associated with employment status in the current study, this variable might reach statistical significance in a large-scale study. Physical condition and incidence of epilepsy were not correlated with employment status, possibly because the physical condition of these study participants was generally good and seizures were well controlled. Intelligence, memory, and executive function also were not correlated with employment status. These results indicate that the level of apathy, disinhibition, and the incidence of psychiatric conditions may be more strongly related to employment status than individual cognitive functions.
Study Limitations
Our study has several limitations that should be considered. First, the time to evaluation after injury for our participants covered a wide range, with an average of 9.8 yr (SD = 11.0) postinjury. However, the replicate analyses using data only from participants with a shorter period from the injury to the assessment showed similar results. Second, the economic situation for each participant might have differed, although the employment situation in Japan during the past 40 yr has been relatively stable compared with that of other developed countries, with an unemployment rate of <5% throughout the past 50 yr, the lowest rate among G7 countries (https://www.stat.go.jp). Third, this study is based on a cross-sectional approach, and prospective studies are needed to confirm these results. Fourth, generalizability is limited because participants were recruited from only two facilities. In addition, the number of participants from the second facility was only 7, further limiting generalizability. Fifth, the number of participants in the current study was not large. Sixth, more detailed psychometric assessments, such as depression and anxiety scales, could have been administered. Finally, the assessment we used for disinhibition has not been standardized for Japanese culture. We chose the BADS questionnaire because it has been used worldwide and because there is no standardized disinhibition assessment method for acquired brain injury among Japanese people.
Implications for Occupational Therapy Practice
The results of this study have the following implications for occupational therapy practice: A comprehensive approach to addressing apathy, disinhibition, and psychiatric conditions might facilitate employment of people with TBI. Vocational rehabilitation services designed to address these issues are worth exploring. More comprehensive assessment, including use of brain imaging and vocational rehabilitation services to predict employment status of people with TBI, is advisable.
Conclusion
Our study suggests that the employment status of people with TBI is mainly associated with apathy level, disinhibition, and incidence of post–TBI psychiatric conditions.
Footnotes
Acknowledgments
Ethical aspects of this study were reviewed and approved by Ashikaga Red Cross Hospital Human Research Ethics Committee. The authors report no conflicts of interest and no sources of funding. Michitaka Funayama designed the study, acquired case data, and drafted the manuscript; Yoshitaka Nakagawa, Asuka Nakajima, Hiroaki Kawashima, and Isamu Matsukawa acquired case data; Taketo Takata and Shin Kurose supervised the study. All authors read and approved the final manuscript. The data sets generated and analyzed during this study are available from the corresponding author on request.