Traumatic Brain Injury and Its Relationship to Previous Convictions, Aggression, and Psychological Functioning in Dutch Detainees

ABSTRACT

Objective: Higher prevalence of traumatic brain injury (TBI) has been reported for detained individuals. TBI may result in erroneously interpreting situations in everyday life, impair problem-solving abilities through dialogue, or negotiation, which may increase aggressive behavior. Knowledge of TBI and its consequences in Dutch penitentiaries is lacking, as virtually no screening, supervision, or treatment is currently provided.

Method: The current manuscript assesses differences in self-reported aggression, impulsivity, indication for mild intellectual disability, history of childhood trauma, and emotion regulation in detainees (n = 283) and non-incarcerated controls (NIC; n = 51), with and without a history of TBI.

Result: A total of 45.2% of detainees reported a TBI, compared to 29.4% in the NIC group. ANCOVA results revealed no interaction between group and TBI history, indicating that TBI history did not affect behavior in detainees and NIC differently. The main effects revealed higher levels of aggression in detainees compared to NICs, and in participants with TBI compared to those without. Separate ANCOVA’s per group reveal that TBI history in detainees was associated with more aggression – also when correcting for violence as the cause of TBI. A logistic regression revealed that detainees with TBI have a higher chance (odds ratio = 3.3) of having at least one prior conviction.

Conclusions: TBI history was related to aggressive behavior in Dutch inmates, future studies should assess if TBI history is predictive of in-prison infractions and recidivism rates, and if screening and providing neuro-rehabilitative measures or improved supervision may improve TBI related deficits.

KEYWORDS:

• Acquired brain injury
• prisoners
• violence
• conviction

Severe head injuries often cause trauma to the brain, and these traumatic brain injuries are the leading cause of handicap, hospitalization, and death in children and adolescents (Rusnak, 2013). Traumatic brain injuries (TBI) can have prolonged and detrimental effects on behavior and social functioning during adulthood (Durand et al., 2017). TBI is defined as an alteration in brain function, or other evidence of brain pathology, caused by an external cause (Menon et al., 2010). They are often the result of falls, sports-related injuries, violence, and/or traffic accidents. Children below the age of five, male adolescents and male young adults are especially at risk (Yates et al., 2006). TBI prevalence in the general (adult) population is estimated to be around 12% in developed countries (Frost et al., 2013). Most often, TBIs are a “mild” form of brain trauma, where patients typically either do not lose consciousness or lose consciousness for a maximum of 30 minutes. Patients who have experienced a TBI may show deficits in various domains, including emotion processing, self-control, impulsivity, IQ, and aggression (Hartikainen et al., 2010; Levin & Hanten, 2005; Parker & Rosenblum, 1996; Wall et al., 2006). Reduced emotion regulation abilities following TBI have received relatively little attention but seems to be impaired after TBI (Rowlands et al., 2019). A decrease in such functions may result in erroneously interpreting situations in everyday life, and may impair problem-solving abilities through dialogue or negotiation, which in turn may increase the chance of violent and aggressive behavior (Leon-Carrion & Ramos, 2003).

An overrepresentation of people with TBI within the criminal justice system has been reported (Durand et al., 2017; Farrer & Hedges, 2011; Shiroma et al., 2010). A large Swedish cohort study showed that experiencing TBI was associated with violent offending, even when compared to unaffected siblings (Fazel et al., 2011). TBI is related to increased criminal behavior in a longitudinal cohort study and may represent a risk factor for offending (McKinlay et al., 2014). Furthermore, a recent review estimates that between 40% and 60% of inmates have suffered at least one TBI and that having experienced multiple TBIs occurs relatively frequent (Durand et al., 2017). Detainees with a history of TBI have a higher risk of recidivism (Ray & Richardson, 2017), have been convicted more often (Williams et al., 2010), and have increased risk of in-prison infractions including vocal and physical aggression (Shiroma et al., 2010). Moreover, female – but not male – prisoners with TBI more often report having experienced childhood physical and sexual abuse, but not neglect (Colantonio et al., 2014).

Although TBI appears to be relevant for the etiology of – violent – criminal behavior, both in society and within the prison, several issues limit strong conclusions regarding this link (Durand et al., 2017). Most studies assessing TBI in a criminal justice setting lack an appropriate control group (preferably containing subjects who suffered TBI, but with no prison history). Such comparison is further hindered by several confounding factors – including socioeconomic status, substance (ab)use, and frequent risk-taking behavior – which are related to the risk of experiencing a TBI, and are also more prevalent in the prison versus general populations (Durand et al., 2017). Furthermore, preexisting aggression and offending behavior are strong predictors of post-TBI aggression and offending, which prevents causal conclusions regarding this relationship. Nevertheless, TBI was still predictive of offending behavior while correcting for pre-TBI offending (Elbogen et al., 2015). Despite the problems related to drawing causal conclusions, and comparisons between detained and non-incarcerated control (NIC) groups, the association between TBI and violent behavior justifies addressing this issue upon (re)entering the criminal justice system. Screening for TBI history and providing care or adjusted supervision might be a target for reducing aggressive behavior in prison and after release (Durand et al., 2017; Leon-Carrion & Ramos, 2003).

Unfortunately, this is currently not the case in many countries, including the Netherlands, and this manuscript, therefore, aims to answer the following issues: (1) describe the prevalence and causes of TBI within Dutch penitentiaries, (2) evaluate if TBI is more prevalent in detainees (DN) compared to non-incarcerated controls (NIC), (3) assess if DN with TBI have higher rates of aggressive behavior, more previous convictions, longer or more sentences, deficits in emotion regulation, impulsivity, or lower scores on a screener for intellectual disability, compared to those without TBI, (4) assess whether the number of experienced TBIs is related to any of the above.

Methods

This article is based on data that were collected for a larger research project, investigating the neurobiological and psychosocial characteristics in Dutch detainees (Den Bak et al., 2018).

Participants

A total of 281 male detainees were recruited from six different geographically dispersed penitentiaries in the Netherlands. Only detainees from the general prison population and houses of custody were eligible for participation, and therefore prisoners from maximum security or psychiatric wards were excluded. Subjects were required to have been housed in prison for at least 3 weeks because psychophysiological (e.g. heart rate and skin conductance) data were collected and initial stress of entering the prison system would otherwise have confounded this data (see (Den Bak et al., 2018)). Detainees were recruited through poster advertisements, internal newsletters, and oral personal communication.

Additionally, 51 male non-incarcerated controls were recruited through online advertisement, social media, and personal communication. An effort was made to focus on people with lower than average education levels to increase comparability with the detained participants. None of the participants reported a history of incarceration or any previous convictions. The study was approved by the Ethical Commission of the psychology department of the University of Leiden and participants signed an informed consent form after a thorough explanation of the research procedure – consistent with the declaration of Helsinki – before participating in the study. All participants were remunerated for their participation; for detainees, a total of €10 was added to their in-prison bank-account.

Questionnaires

Demographic information and TBI history

A demographic questionnaire was administered, which assessed current age, education level (low, medium, high) and nationality (Dutch, Turkish-Dutch, Moroccan-Dutch, Antillean-Dutch, Surinam-Dutch, Surinam, Other, Other-Dutch). In this questionnaire, all participants were asked whether they have had ever experienced “severe trauma to the head.” After receiving an affirmative response, participants could indicate how often this had occurred and what the cause of the injuries was.

Aggression and sentencing history

The Dutch version of the Aggression Questionnaire (AQ) was administered, consisting of 12 items, each of which was answered on a 5-point Likert scale (Buss & Perry, 1992; Hornsveld et al., 2009). The questionnaire provides information on the general level of aggression (total score), and four subscales on physical and verbal aggression, as well as anger and hostility. Additionally, detainees self-reported current sentence length (in months) and if – and how often – they had received prior convictions, and/or incarcerations during the past 5 years.

Psychological variables

Several questionnaires were administered to assess the possible presence of a mild intellectual disability (used as a proxy measurement for intelligence) (SCIL 18+ (Kaal et al., 2016)), impulsivity (BIS-11 (Patton et al., 1995)), emotion regulation (ERQ (Gross & John, 2003)), and history of childhood trauma (CTQ (Spinhoven et al., 2014)). The SCIL 18+ is a Dutch screening instrument for the assessment of a mild intellectual disability. The instrument contains 14 questions and assignments, including education level, language comprehension, and some calculations. The maximum score on this instrument is 28, whereas a score of 19 or lower indicates a possible mild intellectual disability. The Dutch version of the Behavioral Inhibition Scale (BIS-11) assesses impulsivity through a 30-item questionnaire (Patton et al., 1995), and provides a measurement of general impulsivity (total score). The Dutch version of the Emotion Regulation Questionnaire [ERQ (Gross & John, 2003)] was administered to assess emotion regulation ability. The ERQ is a 10 item questionnaire with statements on emotion regulation abilities/problems. Participants score the extent to which they agree with each statement on a 7 point Likert scale. Finally, participants completed the Dutch version of the Childhood trauma questionnaire, containing 25-items on a history of child abuse and neglect which are answered on a 5 point Likert scale ranging from never true to very often true (Spinhoven et al., 2014).

Analyses

Sample size

A sample size calculation was performed, based on a Swedish population-wide study by Fazel et al. (2011). This study revealed that people with a TBI history were 3.3 times more likely to commit a violent crime (Cohen f = 0.33), compared to those without. This effect was attenuated to an odds ratio of 2.0 (Cohen f = 0.19) when correcting for unaffected siblings. Two sample size calculations for an ANCOVA were conducted using G-power (v3.1.9.4) (Faul et al., 2007), with power (1-β) set at 80%, α = 0.05, two groups, numerator df = 1, and 3 covariates. Results revealed that for effect sizes of f =.19 and f = .33, a total sample size of 220 and 75 participants would suffice.

Data analyses

For information on missing data and data preparation, please see Supplementary information 1. First, chi-square tests were performed to assess whether TBI prevalence and education level differed between detainees and NICs. Additionally, several independent t-tests were performed to assess whether groups (DN/NIC) differed in a number of experienced TBIs, age, emotion regulation, experienced childhood trauma, aggression, impulsivity, and intelligence.

To assess whether there is a significant interaction effect between group and TBI history on aggression, impulsivity, emotion regulation, childhood trauma, and intelligence, five separate two-way ANCOVAs were performed in R using the miceadds package, including total score for aggression, impulsivity, emotion regulation, childhood trauma, and intelligence as the dependent variables, group (NIC/DN) and TBI history (yes/no) as fixed between-subject factors, and age, education level (low, medium, high), and nationality as covariates.

As shown in supplementary Table S1, no interaction effect between group (NIC/DN) and TBI history (yes/no) was present for any of the analyses, although significant main effects for the group (NIC/DN) were present for aggression, impulsivity, and intelligence. Additionally, there was a significant main effect of TBI history on aggression. Due to the non-significant interaction between TBI (yes/no) and group (NIC/DN), as well as several differences in sample characteristics between groups (see below), it was decided to conduct the analyses per group. This manuscript will, therefore, focus on differences in sentencing history (current sentence length, previous convictions, previous prison sentences), aggression, impulsivity, emotion regulation, childhood trauma, and intelligence in detainees with and without a history of TBI, whereas the results from the NIC group are presented in supplementary Table S2.

To assess whether TBI history is related to aggression, current sentence length, number of previous convictions, number of previous prison sentences in the past 5 years, impulsivity, emotion regulation, childhood trauma, and intelligence in detainees, eight separate one-way ANCOVAs were performed in R using the miceadds package, including total score for aggression, current sentence length, previous convictions, previous prison sentences, impulsivity, emotion regulation, childhood trauma, and intelligence as the dependent variables, and TBI history (yes/no) as fixed between-subject factor, and age, education level (low, medium, high), and nationality as covariates.

To assess whether previous conviction and previous prison sentence are related to TBI history, two logistic regression analyses were performed with previous conviction (yes/no), and previous prison sentence (yes/no) as the dependent variables, and TBI history as the independent variable, while including age, education level (low, medium, high), and nationality as covariates.

To further elucidate the relationship between TBI history and aggression in detainees, four posthoc ANCOVA were performed with anger, hostility, physical, and verbal aggression as dependent variables, TBI history (yes/no) as fixed between-subject factor, and age, education level (low, medium, high), and nationality as covariates. Finally, spearman rho correlation analyses were performed between the number of traumatic brain injuries (which were non-normally distributed), and aggression, current sentence length, previous convictions, previous prison sentences, impulsivity, emotion regulation, childhood trauma, and intelligence.

Results

Characteristics

Sample characteristics for both the detainee and non-incarcerated control groups are presented in Table 1. Detainees were significantly older, more aggressive, more impulsive, have a lower education, and have lower scores on a screener for mild intellectual disability. Furthermore, the average sentence length was 38.5 months, 75% of detainees had prior convictions, and 48% had been sentenced to jail in the past 5 years.

Table 1. Sample Characteristics

	Detainees		Non-incarcerated controls		Significance
TBI history (yes/no)	127/154		15/36		X2(1) = 3.77 p = .05
Education (low/middle/high)	101/147/31		2/14/35		X2(2) = 91.34 p < .001
	Mean	Sd	Mean	Sd
Number of TBIs	1.33	4.02	0.51	0.92	t(229.94) = 1.43 p = .15
Age	38.29	12.39	42.94	17.43	t(229.41) = − 2.15 p = .03
ERQ	35.60	9.51	37.84	7.21	t(226.90) = −1.55 p = .12
CTQ	41.86	16.79	38.91	11.23	t(224.74) = 1.17 p = .24
AVL-AV	27.44	10.67	22.29	8.04	t(229.87) = 3.19 p = .002
BIS-11	64.18	11.96	58.07	9.25	t(93.28) = 2.60 p = .01
SCIL 18+	19.51	5.05	24.65	3.04	t(230) = −6.91 p < .001
Current sentence length	38.52	20.48
Previous convictions	10.13	6.28
Previous prison sentences	3.19	1.86

This table denotes the prevalence of experienced TBI, as well as the means and standard deviations for the number of experienced TBIs, age, emotion regulation (ERQ), childhood trauma (CTQ), aggression (AVL-AV), impulsivity (BIS-11), and scores on a screener for mild intellectual disability (SCIL18+). Please note that TBI history, and education were computed for n = 281 detainees, whereas other variables are computed for n = 183.

Prevalence of TBI

A total of 45.2% of detainees reported a history of severe head trauma, compared to 29.4% in non-incarcerated controls, but a chi-Square test did not reveal a significant difference (X²(1) = 3.77 p = .05). The number of experienced TBIs also did not differ significantly between detainee (m = 1.32 sd = 4.01) and, NIC (m = 0.51 sd = 0.92) groups (t(229) = 1.43 p = .15). When assessing the self-reported causes of TBI (see Figure 1), the injuries in NICs were mostly related to falling or (traffic) accidents, whereas in the DN group, the injuries were also often caused by violence (and less so by falling). This was not statistically tested due to the small number of NIC participants with a TBI history.

Figure 1. Causes of TBI. This figure depicts the self-reported causes of the TBI. Please note that one participant may report multiple causes of TBI

Differences related to TBI history in detainees

Results of the eight separate one-way ANCOVA’s investing detainees with and without a TBI history differed in aggression, current sentence length, previous convictions, previous prison sentences, impulsivity, emotion regulation, childhood trauma, and intelligence only reveal significant differences in aggression levels (see Table 2). Detainees with a history of TBI report significantly (f(1,766858.6) = 30.90 p < .001) more aggression (m = 31.75 sd = 10.52), when compared to detainees without a history of TBI (m = 23.86 sd = 9.45). This effect remained significant when correcting for violence as the cause of the TBI, although this covariate was also significantly related to aggression (f(1,3065189.4) = 8.89 p = .003).

Table 2. Pooled ANCOVA Results

	Model 1						Model 2
	Variable	SSQ (type2)	Df	F-value	p-value	Eta2	Variable	SSQ (type2)	Df	F-value	p-value	Eta2
Aggression	TBI	2823.40	1.00	29.57	P < .001	0.14	TBI	2823.40	1.00	30.90	P < .001	0.14
(r2 =.22/r2 =.26)	Age	668.01	1.00	6.96	P < .01	0.03	Age	668.01	1.00	7.28	P < .01	0.03
	Education	265.75	2.00	1.38	0.25	0.01	Education	265.75	2.00	1.44	0.24	0.01
	Nationality	772.90	7.00	1.15	0.33	0.04	Nationality	772.90	7.00	1.21	0.30	0.04
	Residual	16198.60					TBI caused by violence	808.11	1.00	8.89	P < .001	0.04
							Residual	15390.49
Impulsivity	TBI	500.43	1.00	2.75	0.10	0.02
(r2 =.11)	Age	472.76	1.00	2.95	0.09	0.02
	Education	770.19	2.00	2.27	0.10	0.03
	Nationality	995.36	7.00	0.74	0.63	0.04
	Residual	23324.36
Emotion regulation	TBI	21.08	1.00	0.19	0.67	0.00
(r2 =.04)	Age	4.13	1.00	0.02	0.87	0.00
	Education	355.33	2.00	1.78	0.17	0.02
	Nationality	341.99	7.00	0.41	0.89	0.02
	Residual	15730.94
Childhood trauma	TBI	1104.54	1.00	3.89	0.05	0.02
(r2 =.09)	Age	7.17	1.00	0.01	0.91	0.00
	Education	119.74	2.00	0.15	0.86	0.00
	Nationality	3347.09	7.00	1.60	0.13	0.07
	Residual	46755.88
Intelligence	TBI	15.64	1.00	0.77	0.38	0.00
(r2 =.25)	Age	29.33	1.00	1.41	0.23	0.01
	Education	789.41	2.00	19.31	0.00	0.17
	Nationality	317.66	7.00	2.20	0.03	0.07
	Residual	3489.67
Number of convictions	TBI	118.97	1.00	1.04	0.31	0.02
(r2 =.09)	Age	138.42	1.00	1.83	0.18	0.02
	Education	90.20	2.00	0.47	0.63	0.01
	Nationality	298.74	7.00	0.33	0.94	0.04
	Residual	6534.04
Sentence length	TBI	116.79	1.00	0.15	0.70	0.00
(r2 =.06)	Age	1162.42	1.00	2.17	0.14	0.02
	Education	599.74	2.00	0.36	0.70	0.01
	Nationality	2949.53	7.00	0.32	0.94	0.04
	Residual	71575.22
Number of incarcerations (past 5 years)	TBI	5.81	1.00	0.79	0.37	0.01
(r2 =.17)	Age	27.98	1.00	3.56	0.06	0.04
	Education	32.33	2.00	1.69	0.19	0.05
	Nationality	41.33	7.00	0.62	0.74	0.07
	Residual	525.93

This table shows the pooled univariate ANCOVA results for the difference between detainees with and without a history of TBI, on aggression, impulsivity, emotion regulation, childhood trauma, and intelligence. In model two, a covariate is added to the model assessing differences between detainees with and without a history of TBI on aggression, which reflects whether TBI was ever the result of violence. Results indicate that detainees with a history of TBI have significantly different levels of aggression, but not in levels of impulsivity, emotion regulation, childhood trauma, and intelligence. Differences in aggression remain significant when adding “TBI caused by violence” as a covariate. TBI = traumatic brain injury, SSQ = sum of squares, significant results are presented in bold letters.

Results of the four posthoc ANCOVAs revealed that detainees with a history of TBI revealed higher levels of anger (f(1,721882000) = 28.15 p < .001), physical (f(1,21902) = 27.34 p < .001) and verbal aggression (f(1,172688823005) = 18.22 p < .001), but not hostility (f(1, 11402516.5) = 5.67 p = .02).

The logistic regression analysis assessing if TBI history was significantly associated with having at least one previous conviction (see Table 3), while correcting for confounding variables, was a significantly better fit compared to an empty model (McFadden pseudo R2 = .15 p < .01). A value of .15 can be considered a medium to good fit (McFadden, 1977). Within the model, TBI history was a significant (T(166.55) = 2.74 p < .01) predictor of having a prior conviction, with an odds ratio of 3.33, indicating that detainees with a TBI history have a 3.33 times higher chance of having a prior conviction compared to detainees without a TBI history.

Table 3. Logistic Regression Table for Prior Convictions

Predictor		Odds Ratio	SE	T	df	P	CI: 2.5%	CI: 97.5%
(Intercept)		23.59	0.84	3.77	166.00	<0.001	4.50	123.65
TBI		3.33	0.44	2.74	166.55	<0.01	1.40	7.90
Age		0.95	0.02	−3.00	166.91	<0.01	0.92	0.98
Education
	Medium (vs Low)	0.97	0.46	−0.06	167.29	0.95	0.39	2.39
	High (vs Low)	0.28	0.60	−2.11	162.09	0.04	0.09	0.92
Nationality
	Turkish-Dutch (vs. Dutch)	1.25	1.12	0.20	168.90	0.84	0.14	11.44
	Marrocan-Dutch (vs. Dutch)	0.55	0.78	−0.76	167.58	0.45	0.12	2.58
	Antillian-Dutch (vs. Dutch)	0.47	0.69	−1.11	168.90	0.27	0.12	1.82
	Surinam-Dutch (vs. Dutch)	1.59	1.16	0.40	168.98	0.69	0.16	15.78
	Surinam (vs. Dutch)	1.16	0.89	0.16	169.02	0.87	0.20	6.73
	Other (vs. Dutch)	0.43	0.73	−1.15	162.58	0.25	0.10	1.82
	Other-Dutch (vs. Dutch)	0.31	0.91	−1.29	168.98	0.20	0.05	1.86
Pseudo R^2 =.15	(p < .01)

Summary of the logistic regression analysis, which assessed if TBI history predicts whether or not detainees had any prior convictions, controlling for background variables. Results indicate that detainees with a TBI history were 3.3 times more likely to have prior convictions. SE = standard error, df = degrees of freedom, p = p-value, CI = 95% confidence interval.

The logistic regression analysis assessing if TBI history was significantly associated with having received at least one prior prison sentence (see Table 4) while correcting for confounding variables, was not a significantly better fit compared to an empty model (McFadden pseudo R2 = .06 p = .24). Within the model, TBI history was not associated with having a prior prison sentence (T(166.55) = 1.73 p = .09).

Table 4. Logistic Regression Table for Prior Jail Sentence

Predictor		Odds Ratio	SE	T	df	P	CI: 2.5%	CI: 97.5%
(Intercept)		6.06	0.74	2.45	161.21	0.02	1.42	25.85
TBI		1.90	0.37	1.73	165.55	0.09	0.91	3.95
Age		0.98	0.02	−1.33	165.31	0.18	0.95	1.01
Education
	Medium (vs Low)	0.72	0.41	−0.80	164.17	0.43	0.32	1.63
	High (vs Low)	0.33	0.57	−1.94	162.45	0.05	0.11	1.02
Nationality
	Turkish-Dutch (vs. Dutch)	1.09	0.86	0.10	168.51	0.92	0.20	5.90
	Marrocan-Dutch (vs. Dutch)	0.37	0.66	−1.53	167.74	0.13	0.10	1.34
	Antillian-Dutch (vs. Dutch)	1.22	0.71	0.28	168.50	0.78	0.30	4.93
	Surinam-Dutch (vs. Dutch)	0.99	0.89	−0.01	168.98	0.99	0.17	5.67
	Surinam (vs. Dutch)	1.54	0.86	0.50	168.70	0.62	0.28	8.39
	Other (vs. Dutch)	0.80	0.72	−0.32	153.16	0.75	0.19	3.33
	Other-Dutch (vs. Dutch)	0.95	0.91	−0.06	168.93	0.95	0.16	5.74
Pseudo R^2 =.06	(p = .24)

Summary of the logistic regression analysis, which assessed if TBI history predicts whether or not detainees had any jail sentences, controlling for background variables. Results show no significant effect on TBI history or any of the confounding variables. SE = standard error, df = degrees of freedom, p = p-value, CI = 95% confidence interval.

Finally, a number of TBI prisoners have experienced significantly correlate with aggression (rho = .35 p = .001), anger (rho = .31 p = .005), physical aggression (rho = .28 p = .01), verbal aggression (rho = .32 p = .005), but not with childhood trauma (rho = .21 p = .07), current sentence length (rho = −.02 p = .88), number of previous convictions (rho = −.18 p = .14), number of previous prison sentences (rho = .18 p = .17), impulsivity (rho = .19 p = .13), emotion regulation (rho = .07 p = .57), hostility (rho = .23 p = .05), or intelligence (rho = −.14 p = .23).

Discussion

This study investigated whether the prevalence and causes of a history of TBI differed between detainees and a sample of non-incarcerated controls, as well as if detainees with a TBI history were more aggressive, had more previous convictions, had longer or more previous prison sentences, showed impaired emotion regulation abilities, higher impulsive behavior, and lower scores on a screener for intellectual disability. The results show that 45.2% of the detainees have a history of TBI and that this is related to higher levels of aggression, especially more anger, verbal, and physical aggression. Having experienced more TBIs relates to higher levels of anger, verbal, and physical aggression.

These results were in line with previous studies that report that 40% to 60% of inmates have suffered at least one TBI and that having experienced multiple TBIs occurs relatively frequent (Durand et al., 2017). Although we did not find a significant difference in the prevalence of TBI or the amount of TBIs suffered, between NIC and detainees, the difference in prevalence did border on significance. Additionally, the detainees report less TBIs which were the result of falling, but more TBIs resulting from (traffic) accidents and violence compared to NICs, which corresponds to a study by Perkes et al. (2011). The trauma underlying the reported TBIs might be more severe for TBI resulting from violence or (traffic) accidents compared to falling. Therefore, TBIs reported by the detainees are possibly more severe, compared to the NIC group. Additionally, reporting more benign injuries in the NIC group may inflate prevalence compared to the detainees. Unfortunately, we did not have the possibility to check verified medical records to substantiate this possibility.

Detainees with a TBI history report higher levels of aggression, specifically higher levels of anger, verbal, and physical aggression, and this corresponds to higher aggression levels that have been reported in TBI populations, including Vietnam war veterans (Grafman et al., 1996), young adults at risk for school dropout (Stoddard & Zimmerman, 2011), a Swedish birth cohort study (Fazel et al., 2011), and detainees in England (Pitman et al., 2015). It may also be related to increased risk of in-prison infractions including vocal and physical aggression, as shown by Shiroma et al. (2010). Previous studies have shown (preexisting) aggressive behavior is an important risk factor for TBI occurrence, and this may explain the current results. However, we did show that when correcting for violence as the cause of TBI, detainees with a TBI history were still more aggressive. Furthermore, a previous study shows that acquiring (moderate to severe) TBI predicted subsequent offending even when correcting for pre-TBI convictions (Elbogen et al., 2015). This suggests that acquiring TBI might be a risk factor for aggressive behavior, also in Dutch detainees.

Additionally, the results indicate that detainees with a TBI history are more likely to have a prior conviction compared to detainees without TBI history (odds ratio = 3.3), but are not more likely to have a prior prison sentence. Additionally, having experienced multiple TBIs did not correlate to current sentence length or the number of previous prison sentences/convictions. These results are partly in line with previous studies, which report that TBI history increases the chance of reoffending within 12 months (Ray & Richardson, 2017), and adolescent detainees with TBI have more prior convictions (Williams et al., 2010). Williams et al. (2010) also showed that the higher frequency of TBI related to more prior convictions, and more violent crime.

Although previous studies have shown that TBI might result in higher levels of impulsivity (Votruba et al., 2008), lower intelligence (Raymont et al., 2011), and impaired emotional processing or control (Rowlands et al., 2019), no differences were found in the current study. The detainees showed higher levels of impulsivity and lower intelligence when compared to non-incarcerated controls, but these differences are likely not the result of experienced TBI. Other (childhood) risk factors, including neglect during childhood, paternal mental health issues like substance abuse or antisocial personality disorders, or growing up with a disadvantaged socioeconomic background may explain these differences between the detainee and NIC groups (Lara et al., 2009).

Although childhood neglect and abuse have been related to the occurrence of TBI (Bennett et al., 2011), we did not find that (male) detainees with TBI history reported more childhood trauma. This is in line with a study that showed that female – but not male – prisoners with TBI more often report having experienced childhood physical and sexual abuse, but not neglect (Colantonio et al., 2014). Nevertheless, assessing childhood maltreatment retrospectively through a questionnaire while in prison might result in participants either failing to remember occurrences of childhood trauma or maybe they are simply not willing to disclose such information. This is likely to have affected our results.

Clinical implications

Currently, little attention is directed toward TBI history in Dutch penitentiaries, whereas the current and previous studies show that detainees with a TBI history were more aggressive, were more likely to have a prior conviction, had more in-prison infractions, improved less during treatment, and had a higher reconviction rate (Williams et al., 2018). Since this behavior has a neurological origin, neuro-rehabilitation measures may improve various underlying executive functions and thereby decrease aggressive and offending behavior (Oberholzer & Müri, 2019). Treatment of TBI is often focused on healing the wound, and less so on rehabilitation and supervising the adjustment to the (possibly) acquired impairments. This can lead individuals to erroneously interpret situations in everyday life, and their cognitive deficits may impair normal problem-solving abilities through dialogue or negotiation, which in turn may increase the chance of violent and aggressive behavior (Leon-Carrion & Ramos, 2003; McDonald et al., 2010). Neuro-rehabilitation measures may vary from experience-dependent learning, neurophysiologic stimulation, or a combination of these concepts (Barrett, Oh-Park, Chen & Ifejika, 2013), and should be adapted to the specific needs and abilities of the detainee. Although detainees may benefit from such rehabilitative measures, the prison setting brings some unique challenges, such as restrictions in autonomy, confidentiality, validity of (neuro)psychological testing and possible stigmatization by staff and other detainees (Vanderhoff et al., 2011), which should be addressed when such measures were to be developed or studied.

Limitations and research implications

The current results should be interpreted with some limitations in mind. First, TBI history was assessed through self-report and (length of) loss of consciousness as a result of TBI was not assessed. Due to these limitations, TBI severity could not be factored in, whereas this may be relevant for (aggressive) behavior (Williams et al., 2010). Secondly, we had no access to data on pre-TBI aggression levels, which is known to be predictive of post-TBI aggression. Although an effort was made to correct for pre-TBI aggression by using violence as the cause of TBI as a covariate, this is likely a crude measurement for pre-TBI aggression. Thirdly, all detained participants were required to have been in prison for at least 3 weeks, which automatically results in a sample of detainees with prison sentences of more than 3 weeks, who are likely to have committed more serious (e.g., more aggressive) offenses compared to those with shorter sentences. Fourthly, we did not include detainees from maximum security or psychiatric wards, whereas these detainees may be more aggressive and it is possible that these detainees may have experienced TBI in a similar of possibly even greater extent than other detainees. Future studies should assess TBI in detainees residing in psychiatric and/or maximum security wards. Fifth, we used a screener for mild intellectual disability as a proxy measurement for intelligence, and future research should consider using a better measurement for IQ. Sixth, TBI history was assessed through self-report which is often thought to be problematic especially in forensic populations; however, previous studies show that TBI self-report in detainees is fairly reliable (Schofield et al., 2011). Finally, although we did include a sample of non-incarcerated controls, the sample size was relatively small which might have resulted in a loss-of-power for detecting small effect sizes. Additionally, to assess whether aggression levels are related to TBI status differently in detainees and non-incarcerated controls, future studies should increase sample size and take care in assessing the abovementioned limitations.

Conclusion

In conclusion, our results indicate that TBI is highly prevalent in Dutch detainees and that those with a history of TBI are more likely to report a prior conviction, have higher levels of anger, verbal, and physical aggression. Correcting for violence as the cause of TBI did not seem to affect these results, but no data was present on pre-TBI aggression levels. Future studies should assess if TBI history is predictive of in-prison infractions and recidivism rates in Dutch detainees, and if neuro-rehabilitative measures or improved supportive supervision may improve TBI related deficits.

Disclosure statement

No potential conflict of interest was reported by the author.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This research was funded by the Research and Documentation Centre (grant numbers 2708 and 2914).