There is a clear and inextricable link between various manifestations of mental health and offending which has been examined for many years. Volumes have been written, and careers have been dedicated to the aim of disentangling and clarifying the relationship of these two broad and multifaceted concepts. The seminal contribution entitled ‘Offenders, Deviants or Patients?’ by Herschel Prins, a former colleague who passed away in 2016, provides an excellent summary of the challenge. Herschel was a social worker/probation officer/academic/author/and forensic mental health expert who dedicated his career to understanding the intersection of mental health, crime and importantly, societies’ varied responses to manifestations of these. In some contributions (e.g. Prins, 1994), Prins showed an appreciation of the developmental aspects of both offending and mental health, which is perhaps not surprising considering there is a suggestion that Donald West, the original Director of the Cambridge Study in Delinquent Development, was influential early in his career.

Much previous work on the relationship between mental health and offending has focused on improving our understanding of the potential causal ordering of these outcomes. For example, does offending increase the likelihood of mental health conditions, do mental health conditions increase the likelihood of offending, or do both offending and mental health conditions have similar antecedents? Also, to what extent do some mental health conditions fully overlap with norm-breaking/offending behaviour (e.g. conduct disorder), and therefore can the two be meaningfully separated?

The answers to these questions are not only of considerable academic and theoretical interest, but also of critical practical importance. In the progressively more pressurized environments of criminal justice and mental health provision, it is crucial to understand which risk factors or symptoms should be prioritized to both support the individual, and mitigate their risks to themselves and to those around them. At the same time, answering questions about the causal order and reciprocal dynamics between offending and mental health symptoms remains a daunting task even with the best longitudinal data that we currently have. The reasons are manifold: Current theories rarely allow us to formulate more precise expectations about the time delay between a change in mental health symptoms and a change in offending; the links between mental health and offending may be mediated in complex ways through processes such as labelling and stigma, change in social networks or interventions by formal agencies—all of which can be difficult to measure; the direction and strength of causal associations between mental health and offending may vary depending on the examined manifestation of poor mental health (e.g. internalising versus externalising) as well as the type of offending (e.g. violent versus property crime), and finally, our data are often not precise enough to capture the nuances of within-individual change that are required to address issues of causal effects.

Adopting a developmental approach to the questions about the causal link between offending and mental health further increases the complexity for research and practice. This is because the relationship between mental health and offending might be different depending on important age-graded considerations such as when different life events are experienced or exposure to stressors at critical points in development. These considerations and many others touch on the complexity that we hoped to articulate and address in this special issue.

We are pleased to introduce the contributions to our first Special Issue as co-editors of The Journal of Developmental and Life-Course Criminology. The number of compelling expressions of interest and the subsequent high-quality contributions in this volume are a testament to the growing appreciation of the importance of adopting an age-graded and developmental approach when considering the complex link between mental health and offending.

The paper by Wilkinson and Thompson takes a unique approach to examining the relationship between mental health and self-reported offending by considering the impact of the label of a mental health condition. This label was operationalised as a self-report of whether the participant had ever been told by a health practitioner that the participant had attention problems (e.g. ADD/ADHD), depression or anxiety disorders. The timing of the application of these labels was also considered (childhood, adolescence, adulthood). The results suggested that the label of depression or anxiety was associated with an increase in the likelihood of self-reported offending controlling for mental health symptoms, sociodemographic and so-called stress-process mediators (e.g., mental health treatment, criminal history, alcohol/drug use, social relationships and self-control). Interestingly, for those labelled with ADHD, it appeared that the symptoms, and not the label, were associated with the later observed increase in offending. Moreover, the timing of the application of these labels was found to be important. Labels applied in adolescence had a relatively strong impact on later offending, while the independent effect of labels for depression and anxiety applied in adulthood was still significant but small.

In their paper, Wiesner and colleagues used the Oregon Youth Study to investigate the bi-directional relationship between mental health broadly defined (measured using a 53-item self-report of symptoms ranging from somatization to psychoticism), to self-reported offending measured at six time points (from age 19–20 to 29–30). Important co-variates (such as parent criminality, parent mental health problems, participants’ antisocial propensity and internalising symptoms) were controlled in their random intercept cross-lagged panel model (RI-CLPM). The bi-variate results illustrated the considerable rank-order stability of mental health conditions over time and also that the concurrent correlations between mental health and offending were evident, but not strong. The results of the RI-CLPM illustrated a strong positive relationship for mental health and offending between individuals. However, while cross-lagged effects were evident between mental health and offending within individuals, the findings suggested that offending was inversely related to subsequent mental health problems. While this counter-intuitive result was only evident for two timepoints in the mid-twenties, there was no evidence of cross-lagged effects of mental health problems on offending. The authors highlight the implication of this by suggesting that for some offence subcategories, the link between early adult mental health and offending appears spurious, and therefore, interventions to prevent mental health conditions (i.e. those that target internalising disorders) and conduct problems might be required to reduce the incidence of mental health conditions and offending. The counterintuitive finding may encourage readers to reflect on how well statistical models may capture true causal effects. This is especially the case when there is uncertainty whether the ordering of time lags and reference periods in the data available matches the real-world causal lags in the processes under study (Vaisey & Miles, 2017).

The unique contribution by Tcherni-Buzzeo begins by making a strong case for the use of panel data for developmental criminological research. The author then utilizes 14 years of state-level panel data to examine the potential link between juvenile offending and two different approaches that have been increasingly adopted to address the mental health needs of young people, namely prescribing psychotropic medication and school-based services for children with learning disabilities. The author suggests that, theoretically at least, both methods of intervention address common risk factors for offending such as low self-control and school success. The results suggested that the prescribing of psychotropic drugs to young people was not consistently associated with measures of violence regardless of severity. However, increases in school-based services to children were associated with reductions in forms of minor and moderate violence such as fighting and weapon carrying, while child poverty was strongly related to concurrent and later juvenile violence.

An additional contribution examined the complex inter-play between offending and mental health amongst those formerly criminal justice involved as they transition to adulthood. Using a stratified random sample of 1172 males and 657 females (aged 10 to 18) detained between 1995 and 1998 in Chicago, Leverso and colleagues assessed clinical diagnoses for mood disorders (e.g. major depression, mania, hypomania or dysthymic disorder) and self-reported offending at four time points. The results suggested that mood disorders were not related to between-individual differences in self-reported offending, but that decreases in mood disorders were associated with a greater number of offences within individuals. When mood disorders were examined as the dependent variable, self-reported offending was associated with an increase in mood disorders within individuals. Interestingly, there were important within-individual changes noted when examining offending (e.g. employment, marriage and cohabitation associated with reductions in offending) that were less evident when examining mood disorders. However, functional impairment, or limitations that detrimentally influence a person’s ability to carry out certain activities in their lives, was a strong predictor of mood disorders and offending.

Using the data of 602 13-year-olds (46% female) from the Netherlands, Defoe and colleagues aimed to investigate the evidence for three competing explanations for the possible relationship between internalising symptoms and offending. These were the acting out model, the failure model and strain theory. The acting out model proposes that depression proceeds delinquency as youth commit offences as a result of their inability to manage their depression. The failure model proposes that delinquency, and specifically the societal and social response to delinquency, results in depression. Alternatively, the strain theory proposes a negative relationship between delinquency and depression as delinquency may be used as a way of coping with the emotional strains of depression. While there were important variations by gender, ethnicity and phase of adolescence, the overall results suggested that higher levels of self-reported delinquency predicted lower levels of depressive symptoms, in support of strain theory.

Ward and colleagues adopted an age-graded health-based desistence approach to consider how mental and physical health may influence substance use and offending directly or indirectly through psychosocial maturity using three waves (and earlier baseline data) of the Pathways to Desistence Study. Psychosocial maturity or developmental was conceptualised as including identity, self-reliance and work orientation. The results highlight the negative relation between greater physical health and psychosocial maturity and decreased depressive symptoms. Overall, the authors conclude that poor health states such as depressive symptoms and poor physical health delayed desistence through their deleterious impact on psychosocial maturity.

The contribution of Bergstrom and Farrington focused on a somewhat contentious mental health condition, that of psychopathy. Psychopathy could be considered a mental health condition because it is a defined and measurable constellation of psychological and behavioural traits, and as such is similar to the personality disorders in diagnostic manuals. However, there is no psychiatric diagnosis of psychopathy, and from an academic and practical perspective, the burden of those with psychopathic symptoms generally fails within with the criminal justice system as opposed to mental health. However, considering the social cost of psychopathy, there is no question that more research on the development of this condition is needed. Bergstrom and Farrington used the male and female children of the prospective longitudinal Cambridge Study in Delinquent Development to examine individual, parental, family and socioeconomic risk factors for psychopathy. Importantly, the results suggested that some of the risk factors were similar for the development of psychopathy in both males and females (e.g. poor parental supervision, risk taking), while others were unique to males (e.g. early school leaving) and females (e.g. parental conflict, parental separation).

The collection of manuscripts in this special issue illustrates the range of research being conducted in which the relationship between mental health and offending is being considered through a developmental lens. It is a challenge to integrate the results of these works and draw firm conclusions, but it is clear that mental health labels are important to consider, that both mental health and offending may have similar risk factors, that well-established risk factors may have primacy in explaining offending (i.e. poverty), that different life events may influence offending and mental health differently, that the relationship between mental health and offending might be different at different ages, that psychosocial maturity might be an important concept connecting mental health and offending and that there are intergenerational impacts on the development of psychopathy and these may differ between males and females.

We would encourage scholars to extend on these important results. For example, there is still a considerable amount of work that is needed to move beyond the identification of risk factors towards understanding the causal mechanisms that may link mental health and offending. Also, it is noteworthy that in some of the studies (where data was available), ethnicity was considered and examined, but this was not the focus of any studies. This is an important gap that needs addressing particularly as there is a well-known ethnicity and mental health paradox. That is, individuals of Black ethnicity have higher exposure to risk factors (e.g. stress, poverty), but lower rates of mental health diagnoses (Ruiz et al., 2019). This may reflect differences in access to mental health support, poor mental health literacy (i.e. recognizing mental health symptoms) or negative help-seeking behaviour, all of which could reflect systemic racism and discrimination and the lack of culturally appropriate and responsive services (Lateef, 2021).