Neurobiological Factors as Predictors of Prisoners’ Response to a Cognitive Skills Training
Introduction
Throughout the world, more than ten million people are confined in penal institutions (Walmsley, 2013). Incarcerating people with criminal behavior is the most widely used strategy to protect society against crime, but the recidivism rate after confinement is high. For this reason, several rehabilitation models have been introduced to develop effective interventions aimed to reduce antisocial behavior and, eventually, to reduce recidivism rate. Of these models, the Risk-Need-Responsivity (RNR) model is currently most prominent for treating offenders (e.g. Andrews et al., 2011, Ward et al., 2007).
The RNR model was developed in the 1980s and is primarily based on personality and socio-psychological perspectives on human behavior (Andrews & Bonta, 2010). According to this model, the assessment and treatment of offenders should be based on three principles. The risk principle proposes that the level of treatment intensity should correspond to the offender’s risk level; the need principle determines which specific criminogenic needs should be targeted in treatment; and the responsivity principle suggests that cognitive/behavioral interventions work best for offenders and prescribes that the intervention should be tailored to the offender’s learning style, motivation, abilities and strengths.
There is strong meta-analytic evidence suggesting that current behavioral, cognitive-behavioral and multimodal intervention strategies are successful in influencing factors that are known to predict recidivism (e.g. Andrews and Bonta, 2010, Genoves et al., 2006, Lipsey and Cullen, 2007, Pearson et al., 2002). For instance, cognitive-behavioral therapy (CBT) aims to ameliorate dysfunctional (i.e. antisocial) thinking processes by improving specific cognitive skills such as empathy, moral reasoning, planning and problem solving (McDougall et al., 2009, Sadlier, 2010, Vaske et al., 2011). Examples of well-known CBT programs are Reasoning and Rehabilitation (Ross & Hilborn, 2008), Aggression Replacement Training (Goldstein, Glick, & Gibb, 1998) and Enhanced Thinking Skills therapy (Clark, 2000).
Nevertheless, response rates of these intervention programs vary widely between different effect studies. For example, the effectiveness of CBT varies between less than 10% up to almost a 50% reduction of criminal recidivism (Lipsey and Cullen, 2007, Lipsey et al., 2007, McDougall et al., 2009). Additionally, the rates of treatment non-completion range from 20% to 40% (Hollin et al., 2008, Olver et al., 2011, Polaschek, 2010). These high percentages are concerning, especially since ‘non-completers’ are six to eight times more likely to reoffend compared to treatment ‘completers’ (e.g. Dowden and Serin, 2001, Hollin et al., 2008, Seager et al., 2004). This implies that non-completers may represent the harder-to-treat cases that are especially in need of treatment (Wormith & Olver, 2002).
According to the RNR model, several factors are assumed to affect treatment outcome: gender, ethnicity, age, clinical status, verbal intelligence, motivation and personality (Andrews & Dowden, 2007). In addition, factors such as treatment integrity, program setting, and different offender’s characteristics, such as a prior offense history and drug abuse, have been suggested as explanations for the wide variability in treatment outcome (Lipsey et al., 2007, Serin and Kennedy, 1997, Sterling-Turner et al., 2002). Nevertheless, it remains unclear which mechanisms exactly underlie a wide treatment response variety and which factors can ‘predict’ whether the offender is likely to adhere to and complete therapy. According to Lipsey and Cullen (2007), “(…) there are many questions about the sources of variability in the effects of rehabilitation treatments that have not been adequately addressed by the research available to date” (p. 313). This indicates the need to better understand why some individuals respond well to correctional treatment and others do not, for both the eventual improvement of treatment selection and success, and the reduction of recidivism rates.
In recent years, more attention has been paid to a neurobiological view on antisocial behavior, which has become a valuable additional perspective for its understanding (Glenn & Raine, 2014). The increasing neurocriminological knowledge has led to the suggestion that specific impairments in neurobiological systems, such as poor frontal brain functioning, may disrupt the types of cognitive or emotional processing that usually play a prominent role in therapeutic interventions (Fishbein et al., 2006, Van Goozen and Fairchild, 2008). In addition, Vaske et al. (2011) argue that CBT is effective in reducing antisocial behavior because it targets specific cognitive deficits and corresponding brain areas associated with these cognitive deficits. Therefore, information about underlying neurobiological mechanisms related to effective CBT is what eventually may improve our understanding of why some offenders benefit from CBT while others do not.
To illustrate, cognitive and emotional empathy are central concepts to CBT and to criminology in general (Jolliffe and Farrington, 2004, Van Langen et al., 2014). In addition, neuropsychological studies have shown that both types of empathy are associated with activation in specific brain regions, such as the medial prefrontal cortex, temporo-parietal junction and cingulate cortex1 (Vaske et al., 2011). It is likely that effective CBT does not only change behavioral aspects of empathy, but also changes frontal brain functioning associated with cognitive and emotional empathy. In addition, not only might CBT change brain functioning, but it is also very likely that a reciprocal relationship exists between the outcome of CBT on behavior and brain functioning (CBT ← → brain functioning) (Vaske et al., 2011). In other words, individual differences in brain functioning may moderate the effectiveness of CBT. This raises the question whether brain functioning, and perhaps other neurobiological factors, may present a responsivity concern to correctional therapy.
In a recent literature review, we have studied what is known about the association between neurobiological factors and different types of behavioral treatment for individuals with antisocial behavior (Cornet, De Kogel, Nijman, Raine & Van der Laan, 2014). Although only ten relevant studies were found, it appears that specific neurobiological factors actually can predict treatment outcome. Especially low levels of physiological arousal, such as a low resting heart rate and low cortisol levels, were predictive of poor treatment outcome. None of the included studies provided a full explanation for this relationship. Yet, one possible reason is that individuals with antisocial behavior and low arousal levels are often characterized by callous, unemotional or psychopathic traits (Cima, Smeets, & Jelicic, 2008). It is known that individuals with high levels of psychopathic traits display several impaired learning processes, such as social learning and error learning, which probably impairs their ability to benefit from behavioral treatment (Blair et al., 2005, Von Borries et al., 2010).
Results from this literature review show that a neurobiological perspective on the treatment outcome of individuals with antisocial behavior may provide additional exploratory value to the current psychological and sociological perspectives central to the RNR model. However, several limitations exist with regard to the studies included in the review. For example, the majority of the studies included a sample of children, while the included studies also differed substantially with regard to antisocial behavior problems, the content of the treatment programs, and treatment outcome measures. Given the newness of this line of research and the limited number of studies, more research is needed.
Therefore, the aim of the present study is to further explore the predictive value of specific neurobiological factors in relation to a cognitive skills training in a sample of convicted adult offenders. Based on the literature review, it was hypothesized that: 1) low levels of heart rate activity are associated with poor treatment outcome and 2) weak neurocognitive functioning, as measured with a variety of neuropsychological tasks, is associated with more benefit from treatment, since there is greater potential for improvement.
Section snippets
Participants
The current sample consisted of 121 male detainees with a mean age of 28.79 (SD = 8.57), who had been selected by the Probation Service to take part in a cognitive skills training aimed at reducing cognitive deficits (see the Cognitive Skills Training Section). Participants were recruited in several prisons in the Netherlands between 2011 and 2013. The only reason for exclusion from participation in the study was an unstable psychological or physical condition at the time of measurement. The
Measure of general intellectual ability
The Dutch version (NLV) of the National Adult Reading Test (Nelson, 1982, Nelson and O'Connell, 1978, Schmand et al., 1992).3 The total NLV score appears to correlate highly with the Wechsler Adult Intelligence Scale total IQ score (.74) and the total verbal IQ score (.85) (Schmand et al., 1992). The NLV score is not valid for subjects who have not grown up with the Dutch language; for
Missing values and imputation strategy
Missing values were detected on predictor variables, with a mean of 7% missing values per variable (range: 0-19%). Missing data on predictor variables was mainly due to incomplete questionnaires or non-response by mail. In addition, technical difficulties (e.g. the heart rate measure was not working), participant’s fatigue or a misunderstanding concerning appointments were other reasons for missing data. Furthermore, an average of 34% (range: 30-37%) missing values per variable was detected on
Results
Descriptive statistics of participants’ results in terms of pre-assessment performance are presented in Table 1. Furthermore, results on treatment outcome measures are also displayed.
Trainers reported a significant positive change in detainees’ behavior following treatment. For those who completed treatment, the mean scores on the ‘behavior during treatment’ questionnaire rose significantly, from 17.56 (SE = 0.39) to 18.64 (SE = 0.34) (t = 3.21, p = .001). However, this positive change in detainees’
Discussion
The present study addressed the predictive value of neurobiological factors in relation to detainees’ treatment outcome, in order to better understand why some individuals respond well to treatment while others do not. In general, various individual characteristics (e.g. background information, behavioral measures and neurobiological factors) were included in the current study, but only a small proportion was associated with treatment outcome. Nevertheless, the present study yielded three
Conclusion
The results of the current study suggest the following: detainees who are perhaps most in need of treatment (e.g. individuals with increased concentration problems, psychopathic traits and early-onset antisocial behavior) may be less likely to benefit from treatment. The incorporation of neuropsychological measures, such as the D2 task, might better detect detainees who are less likely to benefit from treatment. Future research is warranted to gain more insight into the relationship between
Acknowledgement
This work was supported by the Netherlands Society of Scientific Research (NWO: Brain and Cognition: Societal Innovation, Grant: 056-21-011), the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), and the Research and Documentation Centre (WODC), Ministry of Security and Justice. In the process of data collection, the cooperation of the detainees who took part in the study, the prison-staff, the Probation Service, and the Custodial Institutions Agency has been of
References (89)
- et al.
Self-reported trauma, cortisol levels, and aggression in psychopathic and non-psychopathic prison inmates
Biological Psychology
(2008) - et al.
Ambulatory Measurement of Respiratory Sinus Arrhythmia and Respiration Rate
Biological Psychology
(1995) - et al.
Early starters: Which type of criminal onset matters most for delinquent careers?
Journal of Criminal Justice
(2013) - et al.
Neurocognitive skills moderate urban male adolescents' responses to preventive intervention materials
Drug and Alcohol Dependence
(2006) - et al.
Empathy and offending: A systematic review and meta-analysis
Aggression and Violent Behavior
(2004) - et al.
Assessing Aggressive Behavior in Forensic Psychiatric Patients: Validity and Clinical Utility of Combining Two Instruments
Archives of Psychiatric Nursing
(2012) - et al.
A meta-analytic review of the relation between antisocial behavior and neuropsychological measures of executive function
Clinical Psychology Review
(2000) - et al.
Dementia: Estimation of Premorbid Intelligence Levels Using the New Adult Reading Test
Cortex
(1978) - et al.
Being hot-tempered: Autonomic, emotional, and behavioral distinctions between childhood reactive and proactive aggression
Biological Psychology
(2010) - et al.
A comparative study of violence risk assessment tools: A systematic review and metaregression analysis of 68 studies involving 25,980 participants
Clinical Psychology Review
(2011)