Introduction
Youth offenders aged 15–24 years account for 40% of criminal justice apprehensions despite representing only 15% of the population, and individuals aged 17–24 years offend more than any other age group (Lambie, 2018). Adolescent development extends until the mid-20 (Lees et al., 2020), meaning their executive functions are still developing. Therefore, individuals in this age group are yet to fully develop impulse control, the ability to have awareness of the consequences of their actions and psychosocial maturity. This is of particular concern in the youth Corrections population because of greater exposure to biological, psychological and developmental factors that impact cognitive development compared with the general population.
Internationally, there is growing evidence of the impact of traumatic brain injury (TBI) on incarceration rates and criminal behaviour, with TBI considered an independent predictor of future criminal behaviour. Kaba et al. (2014) found at least one head injury was reported by 259 (67.4%) of the 384 inmates screened in a youth justice service in the United States. National research from New Zealand shows that a substantial proportion of individuals do not present to hospital following sustaining an acute TBI (Feigin et al., 2013). This is particularly concerning in the youth justice setting, as outcomes are likely to be worsened without appropriate medical attention, including increased risk for future criminal and violent behaviours (Colantonio et al., 2014; Hawley & Maden, 2003; Leon-Carrion & Ramos, 2003).
Even mild TBI can have a significant impact on cognitive abilities, particularly in the developing brain. Hessen and colleagues (2007) found that children that sustained a mild complicated TBI were more vulnerable to development of mild neuropsychological dysfunction than adults that sustained similar head injuries when both were measured around 23 years following the initial head injury. Moreover, Anderson et al. (2012) found that children aged 3–12 years had persisting cognitive difficulties following a TBI, particularly in executive functions, along with behavioural, social and adaptive skills. These problems were more pronounced in those with severe TBI. These findings were replicated by Petranovich et al. (2020), who found that cognitive problems were also evident in mild to moderate TBI. That study noted weaknesses across multiple domains such as visual learning and memory, working memory, visual skills, visual-motor skills and processing speed. Furthermore, executive skills had the most pronounced difficulties, including inhibitory control.
Prenatal exposure to alcohol can lead to physical, mental, behavioural and cognitive abnormalities. Burd et al. (2010) suggested that foetal alcohol spectrum disorder (FASD) is common in adolescents and adults in the Corrections setting. A systematic review identified FASD as the leading cause of preventable birth disorders and developmental disabilities, with those exposed to alcohol in gestation more likely to commit criminal acts compared with others. It has also been suggested that FASD entails higher financial costs compared with the normal population related to support measures to individuals and family members, in addition to the social impact of criminal offences committed (Sessa et al., 2022). Bower et al. (2018) found that 36% of those in an Australian youth detention population met the criteria for FASD. This is concerning given that potential cognitive impairments caused by FASD are apparent across all areas of cognitive functioning, including executive skills, verbal and visual skills, attention and processing speed (Lange et al., 2019).
Trauma is another prevalent factor in prison populations that impacts neuropsychological abilities. Bevan (2017) identified high rates of lifetime exposure to potential traumatising events, with 57% of prisoners of a New Zealand adult corrections population having experienced sexual and/or family violence. In an Australian study, qualitative interviews identified a trajectory to violent offending and incarceration related to childhood/adolescent trauma, lack of support or treatment for trauma experiences and substance abuse to mask pain (Honorato et al., 2016). Such impacts are particularly apparent in neuropsychological outcomes among youth exposed to early life trauma. A meta-analysis demonstrated the impact of trauma on development of executive skills and suggested that trauma exposed youth had low levels of executive skills, including inhibition, working memory and cognitive flexibility (Op den Kelder et al., 2018).
Interestingly, trauma history was found to be more prevalent in those with attention deficit hyperactivity disorder (ADHD) compared with the general population; however, ADHD was a better predictor of poorer psychosocial functioning into adulthood (Rucklidge et al., 2006). A metanalysis by Young et al. (2015) highlighted that the prevalence of an ADHD diagnosis was five-fold higher in youth prison populations (30.1%) and 10-fold higher in adult populations (26.2%) compared with the general population. This is of particular concern, since ADHD is also associated with poor development of executive skills (Molitor et al., 2019) and other areas of intellectual functioning during adolescence (Faedda et al., 2019).
Psychiatric morbidity is higher among prisoners than in the general population in New Zealand, including conditions such as psychosis, major depression, bipolar disorder, substance misuse and dependence (Indig et al., 2016). These disorders are known to be associated with cognitive impairment, including in executive functions, information processing and new learning (Kim et al., 2018). Interestingly, improvement in psychiatric symptoms was found following cognitive remediation that targeted cognitive impairment (Kim et al., 2018). This suggested there is a reciprocal relationship between psychiatric and cognitive abilities.
Binge drinking, heavy alcohol consumption and polysubstance use, including cannabis and methamphetamines, in adolescence are associated with an increased likelihood of poor cognitive functioning and mental health (Cyrus et al., 2021; Lees et al., 2020). Previous research suggested heavy drinking and drug use were reported by 79% of youth offenders compared with 27% of non-offenders, and may precipitate and maintain offending; 65.5% of offenders aged 17–24 years had used methamphetamine in the past year (Lambie, 2018). Neuroimaging studies focused on brain development have shown that ongoing alcohol and polysubstance use during adolescence accelerated decreases in grey matter volume, attenuated increases in white matter volume and density, and resulted in poorer white matter integrity (Lees et al., 2020). The critical neurodevelopmental period of adolescence is associated with higher order cognitive functioning, meaning that the brain is particularly vulnerable to neurotoxin exposure from alcohol and polysubstance use. This affects healthy brain development in terms of cognitive, emotional and social functioning (Lees et al., 2020).
Adolescence reflects a heightened period of vulnerability for cognitive development, particularly for the youth offender population who are exposed to a greater range of risk factors that impact neurological development compared with non-offending populations. A systematic review demonstrated that comprehensive assessment was superior to brief screening for neurodevelopmental disorders in the youth justice setting (Holland et al., 2021). For example, an understanding of neurodevelopmental disorders can lead to individualised and tailored interventions that consider individuals’ emotional and cognitive needs to improve capacity to engage with employment and education, and better engage with provision of social needs including housing. Therefore, a better understanding of the factors that impact cognitive development and their neuropsychological sequelae is imperative in reducing recidivism rates of youth offenders, which are greater compared with adult prison populations (Lambie, 2018).
The high prevalence of factors that impact cognitive development in the prison population suggests there is need to better understand to cognitive profiles of those in the justice system. Interestingly, Syngelaki et al. (2009) found that in a group of youth offenders aged 12–18 years, IQ scores were lower than the general population, with higher rates of perseveration in responding. There were also specific impairments in areas such as problem solving, working memory and planning. In the New Zealand context, Barnfield and Leatham (1998) found that in a sample of 50 prisoners, all participants performed below population norms on verbal memory and abstract thinking, with differences by severity of TBI or level of substance use. An Australian study found that skills in understanding ambiguity, making inferences and understanding figurative language were associated with risk and protective factors for youth offending (S. A. Anderson et al., 2022).
To date, no New Zealand study has explored the neuropsychological profiles of those in a youth Corrections setting. The aim of the present study was to retrospectively review the neuropsychological screening results of those seen at the Youth Unit at the Christchurch Men’s prison to establish a neuropsychological profile and understanding of the prevalence of risk factors. It was hypothesised that those presenting in youth justice would present with a neuropsychological profile that is significantly different from that of the general population. It was also hypothesised that established risk factors would be highly prevalent in the youth justice population.
Method
Research Setting and Ethical Approval
This study was reviewed and formally approved by research ethics committees from both Ara Poutama Aotearoa (NZ Department of Corrections) and Oranga Tamariki (NZ Ministry for Children). This study was conducted in a youth Corrections setting, as part of the medium-high security section of Christchurch Men’s Prison, located in the South Island of New Zealand.
Procedure
Youth who resided in the facility were provided the opportunity to complete a neuropsychological assessment. Participants were informed of the general purpose of the neuropsychological assessment by the activities officer, case leader or neuropsychologist. An appointment was then scheduled for the participant and neuropsychologist. The neuropsychologist outlined the purpose of the neuropsychological assessment as an approach to better inform clinical and rehabilitation intervention. Each participant was provided with a consent form describing the purposes of the assessment and research. Participants also had the option to consent for their data to be used for research purposes. If a participant provided consent, an assessment was completed in a single session. Some participants completed the assessment over two sessions when factors such as fatigue, stress and attention impacted their ability to complete the assessment in a single session. A standard neuropsychological clinical interview was completed followed by a battery neuropsychological of tests. In some cases, the assessment was shortened because of the individual’s poor attentional capacity, or additional testing measures were added to provide diagnostic clarification (e.g. to identify a specific learning disorder).
Participants
In total, 50 participants completed the assessment, with those assessed ranging in age from 18 to 21 years. Participants’ ethnicity was NZ European (45%), Māori (40%), Pasifika (5%), Māori/Pasifika (5%) and other (3%).
All participants who completed testing gave consent for their data to be used for research purposes. Data collected through neuropsychological testing were deemed to be a valid reflection of the young person’s abilities as all measures of effort were passed and there were no significant secondary gains evident through the assessment. Participants who were assessed presented as motivated to do well. Although education is an important factor associated with cognitive development, educational data were not included in this study because of variability of accurate reflection of educational engagement (e.g. high levels of school truancy and different types of schooling such as alternative schooling in the Corrections system).
Measures
A standard neuropsychological battery was administered, with some variability in testing provided related to the clinical focus of the assessment. The battery was administered to participants, with alterations made depending on the presenting history and engagement with client. The measures used were: the Weschler Adult Intelligence Scale Fourth Edition subtests (Block Design, Similarities, Vocabulary, Digit Span, Matrix Reasoning, Visual Puzzles, Coding, Symbol Search, Comprehension, Arithmetic); Weschler Memory Scale (WMS) Fourth Edition (Logical Memory I & II, and Visual Reproduction I & II); Trails A and B; Depression, Anxiety and Stress Scale – 21 items (DASS-21); Rey Complex Figure Test (RCFT); Stroop Colour and Word Test (STROOP), Controlled Oral Word Association Test (COWAT); Rey Auditory Verbal Learning Test; and Test of Pre-morbid Functioning. Embedded and stand-alone validity measures included reliable digits, WMS-4 Logical Memory Recognition, and the Dot Counting Test.
Prevalence data were collated from reports that were based on self-report from interviews, medical records and collateral informants.
Results
The results of neuropsychological measures used to assess the youth unit population were compared with population norms. Some measures were excluded from analysis because the test was used qualitatively as opposed to quantitatively, or too much data were missing for that measure (e.g. RCFT, STROOP, DASS-21). Assumptions for normality and distribution were not met in the data sample, so non-parametric analysis using the Wilcoxon signed rank test with continuity correction was employed. Table 1 presents the means and standard deviations of neuropsychological testing measures transformed into z-score based on population age-based norms (Weschler, 2008; Pearson Assessment, 2009; Rickert & Senior, 1998; Tombaugh, 2004; Wechsler, 2009).
The results in Table 2 are based on information collected from reports, which comprised that verified with medical information, formal diagnosis, corroborative history and self-reported information.
Post-hoc Analyses
Post-hoc analyses were conducted using the two variables that were considered the most reliably reported. These two variables were ethnicity and incidence of head knock. The incidence of ADHD, mental health, trauma and exposure to alcohol in gestation were less reliably reported and thought to be potentially underestimated given the nature of the clinical assessment and lack of reliable collateral information. Therefore, ethnicity and incidence of head knock were selected for the post-hoc analyses.
Ethnicity was split into two categories: NZ European and those who identified as Māori or Māori/Pasifika (NZ European n=19, Māori n=24). The non-parametric Wilcoxon rank sum test was used for the analysis. The results showed there were no statistically significant differences in scores between the two ethnicity categories across all neuropsychological measures. Incidence of head knock was split into two categories: those who reported no history of head knock, and those who reported one or more head knock. The non-parametric Wilcoxon rank sum test was used. The results showed there were no statistically significant differences between those with and without a history of head knock across all neuropsychological measures (Table 3).
Discussion
This study aimed to explore the cognitive profile and prevalence of risk factors for poor cognitive development for youth offenders. The performance of the whole group was compared with age-matched norms for measures of general intellectual abilities, visual and verbal skills, executive functions, attention, processing speed, new learning and memory. The hypothesis concerning cognitive abilities of those in a youth sample in corrections was supported as study participants performed significantly below population norms. The only subtest on which the two populations did not significantly differ was Trails A, which was not unexpected. This is a simple measure of processing speed, and has the lowest cognitive demands from the test battery of neuropsychological measures used. As such, it would be the most likely test for those with lower cognitive abilities to perform within expected levels.
These findings suggested there is a need to adjust expectations on prisoners to engage in rehabilitation programmes, interventions, employment and education compared with the general population. For example, the utility of group-based interventions should be tailored for youth prisoners. High-level attentional processes are required to take in and process information, which this study showed to be well below expected levels of the general population on Trails B (a measure of multi-task processing). Similarly, lower scores on measures of reasoning (e.g. Similarities and Matrix Reasoning) implied difficulty with abstract thinking, forming connections between abstract concepts and solving patterns (Wechsler, 2008). This may have implications for functioning across several different areas, such as social skills, capacity for self-awareness, understanding of material and intervention and day-to-day functioning in the community (Barnfield & Leatham, 1998).
Across all measures, standard deviations (a measure of the variability of the data) were greater than or equal to 1. The significant variability across the data set demonstrated a high level of individual difference across participants, despite significantly lower abilities overall across almost all subtests compared with the general population. Similarly, standard deviations may suggest variability in the neuropsychological profile of each individual. There are significant differences in how individuals pay attention, learn, understand and reason with information. This suggested it is necessary to understand individual strengths and weaknesses to inform modifications to rehabilitation and interventions for those in prison.
A ‘one size fits all’ approach with respect to expectations on prisoners to engage in programmes is unlikely to be feasible (Popova et al., 2018). An understanding of cognitive strengths and weaknesses allows for weaknesses to be addressed in rehabilitation while reinforcing and using areas of cognitive strength (Bennett, 2001). Similarly, addressing areas of cognitive difficulties decreases the likelihood of an individual returning to previous maladaptive coping strategies. This may lead to increased confidence through reducing the instances of hopelessness after failing to engage with programmes that were not modified to accommodate areas of cognitive difference.
This study also sought to determine the prevalence of established risk factors for cognitive impairment in the youth Corrections population. We found that 27% of youth were exposed to alcohol in gestation, 20% had a history of likely hypoxic brain injury (e.g. from cardiac arrest, cord around neck at birth), 73% had at least one head knock, 48% had a previous diagnosis of ADHD, 20% had an IQ less than 70 or a diagnosis of intellectual disability, 75% had mental health difficulties and 62% had a history of trauma. These findings were largely consistent with research among youth populations in Australia (Bower et al., 2018) and the UK (Hughes et al., 2012). They were also significantly higher than expected compared with the general population, which supported our second hypothesis.
Interestingly, post-hoc analyses revealed no statistically significant differences in any cognitive measure when considering ethnicity or history of head knock. In terms of ethnicity, this was unexpected, as research has demonstrated differing performance on tests depending on individual ethnicity. For example, Ogden et al. (2003) compared 20 Māori and 20 New Zealand Europeans aged 16–30 years. They found that Māori performed worse on tasks that relied on formal western education and concepts, but scored similarly or better on tasks that aligned with concepts valued by Māori, such as visuospatial skills.
There are multiple factors that may account for non-significant findings in terms of head knock. The present study defined head knock as a knock to the head, whereas other research analysed data based on severity of TBI (mild, moderate and severe) or the number of TBI sustained. Previous research demonstrated the impact of both frequency and severity of head injury on long term outcomes, including neuropsychological sequalae (Forslund et al., 2019). Reporting of head knock in the youth Corrections population may not be accurate given the high incidence of trauma and polysubstance use; those that reported no head injuries may have sustained one or more head injuries. Moreover, functional outcomes in mild to severe TBI remain poorly understood in terms of individual differences (Polinder et al., 2015). This may account for the lack of significant findings in the present study. Another factor maybe insufficient power given the relative low cell numbers in the comparison.
Consistent with previous research (Moffitt et al., 2002), the overall findings of this study demonstrate that neurodevelopmental abnormalities, trauma, mental health issues, low intellectual abilities, reading difficulties and poor performance on neuropsychological testing are more prevalent in offending youth populations compared with non-offending youth populations. The present findings highlight the need for comprehensive neuropsychological assessment to reliably identify neurodevelopmental disorders and other cognitive difficulties. This may increase the chance for success of rehabilitation as interventions can be tailored to prioritise needs, focus on strengths, develop compensatory strategies around areas of difficulty, reduce environmental risk factors and promote protective factors to reduce recidivism as well as support provided during the pre-imprisonment justice process and following release.