Methods. We used clinical and classification data on a census of “special needs” inmates (those with behavioral health disorders) in New Jersey (n=3189) and a census of all special needs inmates released to New Jersey communities over a 12-month period (n=974).

Results. Virtually all adult inmates with special needs had at least 1 Axis I mental disorder, and 68% of these had at least 1 additional Axis I mental disorder, a personality disorder, or addiction problem (67% of all male and 75% of all female special needs inmates). Of those special needs inmates released, 25% returned to the most disadvantaged counties in New Jersey (27% of all male and 18% of all female special needs inmates).

Conclusions. Two types of clustering were found: gender-specific clustering of disorders among inmates and spatial clustering of ex-offenders in impoverished communities. These findings suggest a need for gendered treatment strategies within correctional settings and need for successful reentry strategies.

Patterns of behavioral health disorders differ by gender within and across the general and prison populations. Men and women in the general population, although having roughly equal rates (20%) of active disorder, have dissimilar clusters of behavioral health disorders.6 Men, for example, have higher active rates of antisocial personality disorder and alcohol and drug abuse than women. By contrast, women are more likely to have active disorders of somatization, depression, generalized anxiety, panic, and phobia.6

Although relatively little is known about mental health problems within prison populations, what is known deviates from patterns found between genders in the general population. In general, rates of mental illness are higher among female inmates than among men in state and federal prisons and local jails.2 According to national data collected by the Bureau of Justice Statistics, 20% of women confined in state prisons are diagnosed with a mental disorder, compared with 16% of men.2 Moreover, the prevalence of substance abuse and dependence are higher in the prison population, for both men and women, than in the general population.6^,7

These data, although limited in their details and methodological rigor, have been used to demonstrate the need for behavioral health interventions for inmates during their incarceration and to facilitate their reintegration into the community after release.2^,8 It is indisputable that inmates have substantial behavioral health needs, which if left untreated or only marginally treated will influence their behavior in the institution and later in the community after release. Yet these statistics lack the specificity necessary to guide the development of treatment for people with behavioral health and criminal justice problems in correctional settings. To the extent that inmates have clusters of co-occurring problems that differ by gender, therapeutic interventions will have to be similarly differentiated to be effective.

We explored gender patterns of behavioral health disorders among a cohort of inmates in the New Jersey prison system. Emphasis was on gender-specific patterns of Axis I mental disorders, personality disorder, and addiction disorders, which suggest particular treatment needs when these individuals are in prison and related needs and identifiable risks when they return to the community. Of particular concern was whether inmates, either male or female, with known behavioral health disorders are returning to disadvantaged communities after completing their sentences and whether they are being released without supervision. Socially disorganized communities may be least suited to respond to the complex needs of these individuals, making it even more important for them to have formal assistance. Findings from this study provide some of the details that are necessary for designing and situating effective therapeutic and reentry interventions for men and women during and after their incarceration.

Of the 1267 female inmates in New Jersey, 474 (37%) were classified as having special needs (with 471 having an Axis I mental disorder), compared with 25% nationwide.6 This discrepancy in rates for women may be accounted for by the comprehensive screening and treatment of inmates in New Jersey prisons resulting from a monitored court settlement (from a class action lawsuit) regarding the delivery of mental health services to inmates.10^,11 Typical procedures for screening female inmates for mental disorders are not sufficiently sensitive to the fact that a significant proportion of women in prison have experienced some form of childhood abuse (physical, mental, and sexual)12 and that childhood abuse is often associated with anxiety disorders and mood disorders in adulthood.13 Comprehensive screening includes such childhood abuse–related disorders. In addition, because women tend to avail themselves of mental health treatment when it is available, as it is in New Jersey prisons, this higher-than-average rate of mental disorder within the New Jersey female prison population may provide a more accurate picture of disorder within the female prison population nationwide than do estimates drawn from national data based on less-comprehensive screening of inmates for mental disorders.

The second data set that was compiled by the New Jersey Department of Corrections and used in this study includes demographic and offense data on all released adult special needs inmates from New Jersey prisons over a 12-month period, August 1, 2003, through July 31, 2004 (n = 974).

TABLE 1— Characteristics of Special Needs Inmates in New Jersey Prisons (2002) and Special Needs Inmates Released From New Jersey Prisons (2003–2004), by Gender

In comparison with male inmates, a greater percentage of female inmates were incarcerated for drug-related offenses, but a lower percentage were incarcerated for violent types of offenses (i.e., those involving the use of force or the threat of using force). In all, about half of male special needs inmates had a violent crime as their most serious type of conviction.

The composition of the special needs inmate population was quite similar to the special needs released inmate population in terms of race/ethnicity and age. One noteworthy difference exists: women with special needs were overrepresented among the released inmate population (21%) compared with their representation within the universe of special needs inmates (15%). This difference can be explained by the fact that special-needs women are less likely to be incarcerated for violent crimes (which carry longer sentences) (Table 1 ).

The matrix grouping was undertaken in 2 steps. The first step classified the Axis I mental disorders into 5 types: psychotic disorders or dementia (type 1); bipolar (type 2); all other mood disorders (type 3); posttraumatic stress disorder (type 4); and anxiety disorders (type 5). A sixth classification was made for those with no Axis I mental disorder. Unique cases with and without multiple diagnoses were assigned to the disorder type with the lowest number (ranging from 1 to 5). For example, cases with schizophrenia and depression were assigned to type 1. Cases with a nonbipolar mood disorder and an anxiety disorder were assigned to type 3. All cases in the data set were exclusively assigned to 1 of the 5 Axis I types (or the no Axis I group), which are represented by rows in the matrix (Tables 2 and 3 ).

TABLE 2— Behavioral Health Problems Among Male Special Needs Inmates Within the New Jersey Prison Population, 2002 (n = 2715)

TABLE 3— Behavioral Health Problems Among Female Special Needs Inmates Within the New Jersey Prison Population: 2002 (n = 474)

The second step of constructing the matrix involved the calculation of the percentages of special needs inmates in each of the exclusive Axis I–type categories (1 through 5) who had (1) an additional type of Axis I mental disorder, (2) a co-occurring personality disorder (Axis II), and (3) a co-occurring addiction disorder. These percentages are calculated out of the total number of cases in each row in Tables 2 and 3 and arrayed across each of the relevant columns in these tables.

The second objective of this study was to estimate the percentage of male and female special needs ex-offenders who entered disadvantaged communities on release. For the purposes of conducting this analysis, 2 of the 21 counties in New Jersey, Camden and Essex, were designated as the most disadvantaged communities.

According to a recent Urban Institute Justice Policy Research Report,14 31% of adult inmates released from New Jersey prisons in 2002 returned to Essex and Camden Counties (16% to Essex and 15% to Camden). These individuals were found to concentrate in a small number of communities within these counties, with 13% of all released inmates returning to Newark (Essex County) and an additional 10% returning to the city of Camden (Camden County). Essex and Camden counties, and especially the cities of Newark and Camden, are significantly more disadvantaged than other parts of the state. The 2003–2004 release data (n = 974) were used to examine the extent to which male and female special needs ex-offenders differentially relocated to either of these counties.

Within the special needs inmate population (n = 3189), men were more likely to have a diagnosis of psychotic disorders (28% for men, compared with 12% for women; χ²=47.002; P < .001), whereas women were more likely to have depressive disorders (58% for women, compared with 45% for men; χ² = 24.615; P < .001). The results for men and women appear in Tables 2 and 3 , respectively. The percentages reported in both tables were calculated from the total number of cases reported for each row (i.e., from the total number of special needs inmates with a particular Axis I diagnosis).

The findings shown in Table 2 demonstrate that approximately 30% of special needs male inmates had a diagnosis of schizophrenia, psychotic disorder, delusional disorder, or dementia, which is labeled type 1. The most common comorbid Axis I disorder for this special needs type was some form of mood disorder, type 3 (20.3%), but not bipolar (type 2). Roughly one quarter of type 1 male inmates had some form of personality disorder (Axis II), most commonly cluster B (i.e., antisocial personality disorder, borderline personality disorder, or narcissistic personality disorder). An addiction disorder was diagnosed in approximately 40% of male inmates with type 1 disorder.

The findings in Table 2 also show that of those male inmates diagnosed with bipolar disorder (type 2), slightly more than one third were also diagnosed with a personality disorder, and nearly 56% had an addiction disorder. For those individuals diagnosed with major depression or a mood disorder (type 3), about one quarter were also diagnosed with an anxiety or adjustment disorder (type 5). About 3 percent of special needs male inmates were diagnosed with posttraumatic stress disorder (type 4) as their most serious diagnosis.

Table 3 shows that approximately 67% of special needs female inmates had a diagnosis of bipolar disorder or depression (n = 234). Moreover, roughly 30% of women with type 1 (psychotic) or type 2 (bipolar) disorders also had a depression diagnosis (type 3). The most common comorbid Axis I disorder for female inmates with depression (type 3) were panic and adjustment disorders (type 5). Fewer than 1 in 7 female inmates had a diagnosis of personality disorder. However, an active addiction diagnosis was present in half to three quarters of women in the various Axis I types.

Table 4 compares inmates entering these 2 counties with those entering all other counties in New Jersey on a number of selected characteristics by gender. A greater percentage of men (27%) than women (18%) were released to these counties (χ² = 6.8; P < .05), as was the percentage of inmates aged 30 years and younger (30%) compared with older inmates (23%; χ² = 5, P < .05).

TABLE 4— Characteristics of Inmates Entering Essex and Camden Counties Compared With Inmates Entering Other Counties in New Jersey Prisons, by Gender: 2003–2004

In comparison with Whites and Hispanics, a greater percentage of released special needs Black inmates entered the 2 disadvantaged counties (17%, 23%, and 32%, respectively); this difference was statistically significant (χ² = 30; P < .001). However, there were no statistically significant differences between inmates entering Camden and Essex counties and those entering other counties on types of offense (violent vs non-violent, drug-related vs non–drug-related). Gender differences of ex-offenders released to Essex and Camden counties were not statistically significant, and this may be partially because of the relatively small sample of women released there.

No information is available about the communities in which these inmates resided before imprisonment. Research shows, however, that 95% of people admitted to prison return to their communities of origin on discharge.14 Roughly 60% of men and 50% of women released to Camden and Essex counties had completed their sentences and were released there without parole supervision.

Our findings suggest that virtually all adult inmates with special needs in New Jersey prisons (99%) had at least 1 Axis I mental disorder, and 68% of these had at least 1 additional Axis I mental disorder, a personality disorder, or addiction problem (67% of all male and 75% of all female special needs inmates).

There is, however, a gender-specific clustering of disorders among inmates: schizophrenia or some form of psychotic disorder and personality disorder were more likely within the male population, whereas depressive disorders and addiction problems were more likely in the female population. In addition, there is some evidence of spatial clustering. Specifically, of those special needs inmates released over a 12-month period, 25% (27% of all male and 18% of all female special needs inmates) returned to the most disadvantaged counties in New Jersey.

Before discussing the implications of this study, we note several limitations. The data examined here, and therefore, any conclusions drawn from the findings, apply to the adult inmate and ex-offender population in New Jersey. However, New Jersey’s incarceration and reentry trends are similar to those observed at the national level,14 as are their rates of mental illness in the male prison population.2 The higher rate of mental disorder among female inmates in New Jersey suggests the possibility of greater variation in disorders among the female inmates, which itself may be sensitive to the screening for and delivery of mental health services within correctional settings.

Reentry communities are represented by counties, which vary spatially in their economic and crime conditions. Although it is possible for ex-offenders released from New Jersey prisons to locate to the relatively well-off areas of generally impoverished counties (Essex and Camden), such a pattern is not consistent with extant evidence.14

National data suggests that in general, ex-offenders are more likely to return to impoverished areas with high levels of social disorganization.15^–19 Specifically, Lynch and Sabol20 found in 2001 that 66% of state prisoners released in 1996 went into a core county, defined by a central city of a metropolitan area, and these individuals further concentrated in relatively few neighborhoods within the central cities of the core counties.8 In 1998, Rose and Clear21 found similar patterns for offenders released from Florida prisons, as did Travis et al.22 in 2001 for the counties of Essex and Camden.

Our findings reveal only part of the story but are consistent with the broader literature on the spatial concentration of offenders in socially disadvantaged communities.

Another limitation concerns the determination of special needs. The data are limited in terms of measuring behavioral health disorder among inmates. In particular, special needs status may not accurately measure the types and levels of behavioral health disorders within the prison population in New Jersey.

A disorder, as measured by special needs status, requires that inmates report problems and that correctional health care staff diagnose a disorder. Because reporting and diagnosing are a function of the social and fiscal realities within prison, mental illness and addiction disorders are likely to depend on the special needs criteria and the availability of treatment. Our findings, therefore, only characterize those who have been positively diagnosed by correctional health care staff and who are actively in treatment for these problems.

Our findings suggest 2 types of clustering; gender-specific clustering of disorders among inmates and spatial clustering of ex-offenders in impoverished communities. Evidence is strongest for the first type of clustering and suggestive of the latter. Both have implications for correctional care and community reentry planning.

People incarcerated in prison have a constitutional right to treatment.10^,11 At the very minimum, behavioral health treatment in correctional facilities must be responsive to the unique presentation of disorders among male and female inmates. Approximately 95% of state correctional facilities report providing some form of mental health treatment to prisoners.8 Substance abuse treatment, however, has not received equal attention. Roughly 1 in 4 state prisoners received any treatment for substance abuse, with a higher percentage (40%) receiving treatment if they reported drug use at the time of their offense.23

The most common treatment received was self-help group/peer counseling. The style of treatment inside prisons, with separate and underdeveloped substance abuse treatment, is not consistent with the clustering of disorders within the male and female inmate population.

Within a correctional environment, delivering evidence-based, integrated treatment that is gendered is challenging for several reasons. First, evidence-based behavioral health treatment strategies have been proved effective in general populations. These strategies may not achieve equal effectiveness for a correctional population in part because the setting is different (less focused on empowerment) and in part because of the presence of personality disorders. For this reason, more research is needed on treatment strategies for male and female populations with psychiatric disorder, antisocial tendencies, and addiction disorders who are confined in authoritarian environments.

Second, before disorder clusters can be concurrently treated, they must be identified. This requires systematic and comprehensive assessment. Few correctional environments have the fiscal capacity to treat the levels of psychiatric disorder likely to be found within the correctional population.23 Given the limited budget for mental health treatment, correctional staff has little incentive to identify and treat psychiatric disorders among inmates unless inmates create institutional problems (i.e., custody or security issues) or there is evidence of “deliberate indifference”24^,25 (i.e., when the inmate can prove that prison authorities deliberately disregarded his or her need for treatment).

Problems that might underpin depression or addiction behavior, such as past physical or sexual trauma, will go undiagnosed and untreated, even though they confound treatment for other diagnosed problems. This is also true for interactions between addiction and serious mental illness, such as schizophrenia, bipolar disorder, and major depression, which have unique but interdependent recovery trajectories.

Investments in mental health during the incarceration experience are likely to be lost if not continued in the community after individuals are released. Proactive reentry planning, analogous to hospital discharge planning, is needed for individuals with behavioral health problems leaving prison. For reentry planning to be effective, it must recognize the effect of community and the interactions between individual and community resources on therapeutic recovery and prosocial community integration.

Most individuals are released from prison with minimal reentry planning.26 Courts, as a result of class action suits or court settlements, are increasingly ordering correctional authorities to provide reentry planning for inmates with special needs.27^,28 Because treatment for mental illness and substance abuse conditions is critical for both successful reintegration and as a guard against recidivism,29^–35 placing ex-offenders with behavioral health needs in communities without adequate services or in ways that disrupt the process of treatment places the individual and the community at risk.

Services must therefore follow the ex-offender into the community. One way to ensure this is to provide inmates with the ability to buy services. If ex-offenders are returned to the community with “buying power” to meet their therapeutic, housing, and rehabilitation needs, they will not only stimulate the local economy of their host communities but also increase their personal chances of successfully reentering the community, not prison.

Support for this project was provided by the Center for Mental Health Services & Criminal Justice Research through a center grant from the National Institute of Mental Health (grant P20 MH66170).

We thank Christopher Dittus for his research and editorial assistance.

Human Participant Protection
All data were provided deidentified, and their use was approved by the institutional review board at Rutgers University.