Prevalence and correlates of disability among a sample of victimized women on probation and parole

Abstract

The purpose of this exploratory study was to establish the prevalence of disability as measured by self-reported Social Security Disability Insurance (SSDI) receipt among a sample of women on probation and parole who have experienced interpersonal victimization in childhood and/or adulthood. Women receiving SSDI in this sample were more likely to be older, White, to live alone, and to score lower on measures of social support compared to women nor receiving SSDI. SSDI-recipients were also more likely to report poorer health, chronic pain, and more frequent healthcare service utilization. High rates of adverse childhood experiences, rape, adult victimization, and an overall greater severity of post-traumatic stress disorder symptomatology were observed for women receiving SSDI. Groups had similar overall mental health profiles and diverged primarily on trauma variables. Findings support the need for trauma-informed care and highlight the possibility that some criminal justice system-involved women likely qualify for SSDI yet are not receiving it.

Introduction

As of 2014, there were 1.2 million women involved in the U.S. criminal justice system (CJS), a 700% increase from 1980 (Carson, 2015). Of these women, approximately 1.06 million were on probation/parole, with the remainder residing in correctional facilities (Carson, 2015). The public health consequences of such substantial increases in incarceration rates for this marginalized population, who is already medically underserved and at elevated risk for poor health outcomes, may not be fully realized for years to come (Dumont, Brockmann, Dickman, Alexander, & Rich, 2012; Nowotny, Belknap, Lynch, & DeHart, 2014; Schnittker & John, 2007). Although prevalence of morbidities among CJS-involved women has been established as being higher than those found among the general population (Bronson, Maruschak, & Berzofsky, 2015), the long-term health profile of CJS-involved women is unclear. Disability, conceptualized as chronic health-related impairments which limit everyday functioning, is an underexplored phenomenon within the literature. Increases in overall life expectancy suggests that disability rates will likely rise for aging CJS-involved women who present with myriad health conditions, but who are living longer than in previous generations (Williams et al. 2006).

Chronic disability can impinge upon the quality of life of the individual (Wittchen, Carter, Pfister, Montgomery, & Kessler, 2000) and is a considerable public health cost (Bloom et al., 2012). CJS-involved women comprise a uniquely vulnerable population (Salisbury & Van Voorhis, 2009) and it is likely that by more accurately establishing disability prevalence among this group additional vulnerabilities will be revealed which have thus far remained unexplored. Given the potential avenues for service delivery and intervention prior to the onset of disability, it is important that this under researched area is advanced. Identifying correlates of and elucidating potential causal pathways to disability among CJS-involved women may help inform interventions that lessen the likelihood that health conditions deteriorate into chronic disability.

Background

The majority of women who enter the CJS have a complex and complicated history, defined in large part by cumulative disadvantage and threats to safety and well-being (Chesney-Lind, 2003; Rosseger et al., 2009; Salisbury & Van Voorhis, 2009; Simpson, Yahner, & Dugan, 2008; Tjaden & Thoennes, 2000). Many CJS-involved women will have experienced a range of stressors and traumatic episodes prior and subsequent to CJS-involvement, including abuse, sexual victimization, intimate partner violence (IPV), homelessness, low socioeconomic status (SES), social exclusion, mental and physical health problems, and substance use disorder (SUD) (Aday, Dye, & Kaiser, 2014; Bennett, Holloway, & Farrington, 2008; James & Glaze, 2006; Kushel, Hahn, Evans, Bangsberg, & Moss, 2005; El-Bassel et al., 2004; Green, Miranda, Daroowalla, & Siddique, 2005; Schnittker & John, 2007; Weiser et al., 2009). Oftentimes, these factors are co-occurring and reinforcing insofar as initial vulnerabilities (e.g. economic disadvantage, SUD) increase the likelihood for further victimization, adversity, and poor health (e.g. IPV, homelessness, SUD relapse) (Koob, 2009; Logan, Walker, Jordan, & Leukefeld, 2006; Gutierres & Van Puymbroeck, 2006; Lynch, DeHart, Belknap, & Green, 2012; McHugo et al., 2005).

Of importance to this discussion is the association between adverse events and victimization in childhood and continued adversity and victimization during adulthood, including poor health outcomes (Edwards, Holden, Felitti, & Anda, 2003; Felitti et al., 1998; Min, Minnes, Kim, & Singer, 2013; Tripodi & Pettus-Davis, 2013). Adverse childhood experiences (ACEs) have been found to be associated with multiple health risks in later life, including SUD, depression, attempted suicide, sexually transmitted diseases, obesity, heart disease, and criminal offending, (Anda et al., 2006, 2007; Chapman et al., 2004; Dube et al., 2003; Felitti et al., 1998; Lynch et al., 2017; Mason et al., 2016) in a manner such that the number of ACEs produces a graded relationship with poor adult outcomes. This strong graded relationship has additionally been found among individuals reporting disabilities (Harkonmäki et al., 2006; Rose, Xie, & Stineman, 2014).

CJS-involved women also experience high rates of pervasive stressors in adulthood, such as economic and occupational uncertainty, barriers to accessing services, homelessness, psychological abuse, social marginalization, and community instability (e.g. neighborhood violence, environmental hazards), all of which can contribute to declines in health (Adler & Newman, 2002; Benach et al., 2014; Berkman & Glass, 2000; DeLongis, Folkman, & Lazarus, 1988; Folkman, 2013; Galster, 2012; Greenberg & Rosenheck, 2008; Hatzenbuehler, Phelan, & Link, 2013; Kawachi, Subramanian, & Kim, 2008; Pico-Alfonso, 2005; McEwen & Morrison, 2013; Van Olphen, Eliason, Freudenberg, & Barnes, 2009). It is unclear, though, what the culmination of a lifetime of stress and adversity produces in terms of CJS-involved women’s long-term health outcomes. For instance, higher rates of some specific types of chronic and infectious disease (e.g., HIV, hepatitis, bipolar disorder) have been observed among CJS-involved women (Binswanger, Krueger, & Steiner, 2009; Dumont et al., 2012; Meyer, Springer, & Altice, 2011), however, insufficient attention has been given to considering women’s health trajectories parallel to any arc of CJS-involvement across the lifespan.

Thus far, national estimates of disability prevalence among CJS populations have been produced by government-led survey initiatives. Such surveys have used different criteria for operationalizing the term disability and may have failed to capture many important respondent characteristics, such as social history, psychological symptom onset, and morbidity type (Bronson et al. 2015; Wilper et al., 2009). In determining disability prevalence among incarcerated individuals, the 2011-12 National Inmate Survey (NIS-3) operationalized ‘disability’ similarly to the U.S. Census Bureau’s American Community Survey, in which a disability is constituted by any impairment falling within the categories of cognitive, vision, hearing, ambulatory, self-care, or independent living. Any difficulties reported in these areas by incarcerated individuals qualified them as ‘disabled’ for NIS-3 purposes. However, the authors noted limitations of such measurement, including the fact that self-report was restricted to individuals who had the capacity to complete the survey (e.g. those with severe impairments were excluded) and that mental health disorders were not included. Moreover, etiology was not determined.

Despite these limitations, the NIS-3 provides some of the most recent national estimates of disability among incarcerated individuals. Overall, the health of incarcerated individuals is poor compared to the general population, with 13-16% of individuals indicating multiple disabilities as (Bronson et al., 2015). Using 2002 Survey of Inmates in Local Jails and 2004 Survey of Inmates in State and Federal Correctional Facilities data, Wilper and colleagues (2009) estimated that between 39% and 43% of incarcerated individuals had a chronic physical or mental health condition and noted that mental health conditions among incarcerated individuals is likely higher than the estimates produced by their analyses given the prevalence of undiagnosed mental health symptoms. Using the same survey data, Binswanger et al. (2009) found that incarcerated individuals were at increased odds for chronic conditions such as arthritis, hypertension, asthma, hepatitis, and cancer compared to the general population.

CJS-involved women have higher rates of disability than their male counterparts, with 40% of women in prison and 49% of women in jail reporting at least one disability in 2012, compared to 31% and 39% of men, respectively (Bronson et al., 2015). Serious mental health conditions (e.g. bipolar disorders, major depressive disorder) are also significantly higher among incarcerated females compared to males (Baillargeon, Hoge, & Penn, 2010; Binswanger et al., 2009; Steadman, Osher, Robbins, Case, & Samuels, 2009). CJS-involved women can be frequent consumers of health services, yet somewhat paradoxically, this group often experiences disruptions in and barriers to accessing healthcare and receiving adequate treatment (Ramaswamy, Upadhyayula, Chan, Rhodes, & Leonardo, 2015; Sered & Norton-Hawk, 2013; Wipler et al., 2009).

Beyond studies examining health and disability among geriatric individuals who are incarcerated and investigations exploring the institutional barriers to quality of life among incarcerated elderly individuals, there are limited data describing the prevalence, characteristics, and correlates of disability among CJS-involved women (Fazel, Hope, O’Donnell, Piper, & Jacoby, 2001; Human Rights Watch, 2012; Loeb & AbuDagga, 2006; Williams et al., 2006, 2009). To date, there are no studies examining disability prevalence among non-elderly women on probation and parole. Given what we do know about CJS-involved women, it is likely that there are CJS-involved women in the community with functional impairments who are in need of targeted services yet whose needs are under-recognized and who are under-served. Many important questions remain regarding the nature of disability generally and for CJS-involved women specifically. This is partially because disability as a construct can be gauged using multiple metrics and yet still not fully capture the essence of what it means to experience disability (Barbotte, Guillemin, & Chau, 2001). Indeed, the proportion of individuals who are functionally disabled in the U.S. is unknown (National Institutes of Medicine, 2007) and there remains no universal definition for what establishes ‘disability’ proper (Brault, 2012). These epidemiological and conceptual issues make researching disability among CJS-involved women challenging, but no less important.

Purpose

Poor health outcomes have been observed for CJS-involved women, yet it is uncertain how many CJS-involved women are considered disabled outside of institutional settings, such as women on probation/parole. It is unclear what factors differentiate disabled CJS-involved women from non-disabled CJS-involved women, despite potential similarities. The current research may help explicate some of the significant factors associated with disability in a population with high rates of socioeconomic disadvantage, mental and physical illness, SUD, and a history of victimization. Using data from a study of women on probation and parole in Kentucky, the purpose of this exploratory study was to establish the prevalence of disability status using Social Security Disability Insurance (SSDI) receipt and to compare women receiving SSDI to those not receiving SSDI across demographic characteristics, health, psychological functioning, victimization, and service utilization.

Methods

A primary data analysis was conducted using baseline data obtained from a sample of women on probation and/or parole who have experienced victimization (N=406). Between July 2010 and January 2013 adult women (age >18) on probation (N=307) and/or parole (N=92) in the Louisville metro area (Jefferson County) were recruited via in-person contacts, flyers, direct mailing, and electronic/print media. Screening eligibility was conducted via telephone and in person. Study eligibility criteria included: a) current probation/parole status, b) sexual activity with a male partner in the 30 days prior to study involvement, and c) a history of physical and/or sexual abuse. Victimization experiences were not limited to a particular developmental period or type, thus women who had reported experiencing physical and/or sexual abuse from a parent, caretaker, intimate partner, or non-intimate partner as a child or adult (or both) were included in the sample. Of those screened for study inclusion, 81% were eligible to participate. Computer-assisted interviews (ACASI; NOVA Research Company, 2003) were conducted in person by trained female staff. Extensive information was provided to the respondents during the process of informed consent on possible risk and safeguarding of confidentiality. In addition, the women were made aware that the research team had received a certificate of confidentiality from the National Institutes of Health. Participants were compensated $35 for their time and were debriefed upon completing the interviews. The study was approved by the University of Louisville Institutional Review Board.

Measures

Demographic information was captured using age, educational attainment (high school diploma/GED), employment (i.e. has ever been employed, current employment status), living situation (e.g. living alone vs. living with others, homelessness), marital status (e.g. single, married or in a relationship, and divorced, separated, or widowed), and race was self-reported by study participants.

Disability status was determined by self-report of past-30-day receipt of SSDI benefits. The Social Security Administration (SSA) defines disability as: “the inability to engage in any substantial gainful activity by reason of any medically determinable physical or mental impairment(s) which can be expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months” (SSA, 2017a, para 4).

Health status was operationalized using HIV status along with items adapted from the Medical Outcomes Study: 36-Item Short Form Survey Instrument (Ware & Sherbourne, 1992). Items consisted of five categorical variables measured via statements describing perceived general health (e.g. “In general, would you say your health is: excellent, very good, good, fair, poor?”), bodily pain, (e.g. “How much bodily pain have you had during the past 4 weeks?”), and limitations in activity (e.g. difficulty walking, difficulty doing activities such as carrying groceries or moving a table).

Healthcare utilization was measured using past 12-month number of healthcare visits, any lifetime episodes of SUD treatment, and any past-12-month psychiatric service medication utilization adapted from the MOS 36-Item Short-Form Health Survey (SF-36) Selection (Ware & Sherbourne, 1992).

Childhood adversity and victimization was operationalized using questions adapted from the Revised Conflict Tactics Scale (Straus, Hambly, Boney-McCoy, & Sugarman, 1996), Tolman’s Psychological Maltreatment of Women Inventory revised to address the construct of childhood victimization as opposed to adult women, and questions from Mullings et al. (2003) study involving incarcerated women (Tolman 1989, 1999). These adapted questions yielded 11 dichotomous variables pertaining exclusively to psychological abuse (e.g. “Did your parents or caretaker ever insult, shame, or humiliate you in front of others?”), physical abuse, (e.g. “Did your parents or caretaker ever physically hurt you on purpose (including grabbing, slapping, burning, scalding, punching, choking, throwing you around, or harshly spanking you?”), sexual abuse (e.g. “Did your parents or caretaker ever force or threaten you to do sexual things other than intercourse?” and “Was your first sexual intercourse consensual?”), emotional neglect (e.g. “While growing up, did you frequently feel unloved?”), economic insecurity/neglect (e.g. “While growing up, did you often did not have enough food to eat?”), and family instability (e.g. parents marital status and “Did you ever see anyone abuse your mother?”).

Psychological functioning was operationalized using current psychological distress, locus of control (LOC), and self-esteem. Current psychological distress was determined using the 18-item Brief Symptom Inventory (BSI 18) (Derogatis, 1993) which measured the degree to which nine symptom dimensions (i.e. somatization; obsessive compulsive; interpersonal sensitivity; depression, anxiety; hostility; phobic anxiety; paranoid ideation; psychoticism) were experienced over the past seven-day period (0= not at all; 4=extremely). A global severity index (GSI) was also produced as an indicator of the overall recent psychological distress experienced. Alpha reliabilities for the BSI subscales range from .80 to .91. LOC was measured using the 40-item Adult Nowicki-Strickland Internal-External Control Scale (ANSIE) (Nowicki & Duke, 1974) comprised of questions to which participants can endorse a ‘yes’ or ‘no’ response, including 15 reversed coded questions. Greater number of ‘yes’ responses (and ‘no’ responses for reverse coded questions) yield higher scores indicating a greater external LOC. Alpha reliability for the ANSIE ranges between .56 and .83 (Nowicki & Duke, 1974; Roueche & Mink, 1976). Self-esteem was measured using the Rosenberg Self-esteem Scale (RSE) (Rosenberg, 1965) which ranges from 0-30, with lower scores reflecting lower self-esteem. The mean reliability estimate for the RSE (derived from Heise 1969 single item scale equation and comparable to Cronbach’s alpha) is .75 (Robins, Hendin, & Trzesniewski, 2001).

Trauma was operationalized using diagnostic criteria for PTSD and PTSD severity which were measured using the 49-item Posttraumatic Stress Diagnostic Scale (PDS; Foa 1995; Foa, Cashman, Jaycox, & Perry, 1997). The PDS has been found to be reliable and valid in its diagnostic capabilities using DSM-IV criteria (Foa, 1995; Griffin, Uhlmansiek, Resick, & Mechanic, 2004; Powers, Gilliham, Rosenfield, Jerud, & Foa, 2012).

Finally, social support was measured using a 4-item version of the Medical Outcomes Study-Social Support Survey (MOS-SSS) which assesses practical and emotional supports as well as affection (e.g. “Someone to help with daily chores if you were sick”, “Someone to love and make you feel wanted) and which has been determined as a reliable and valid measure of social support among victimized CJS-involved women (Higgins, Marcum, Golder, Hall, & Logan, 2015).

Data Analysis

Data were analyzed using SPSS-24. Descriptive statistics were examined for all relevant variables in order to establish prevalence of SSDI among the sample. Chi-square and t-tests were utilized to determine statistically significant differences between those receiving SSDI and those who were not.

Results

In the sample of 406 women, 83 (20.4%) were receiving SSDI. Thus, one-fifth of the sample was conceptualized as disabled. Demographic and health variable differences between women receiving SSDI and those who were not found are reported in Table 1. Women receiving SSDI were older ($\overline{x}$ = 41.6 vs. 36.1), more likely to have higher past-month income (p<.001), to live alone (p<.001), and to be White (p<.05). No significant differences were found among comparisons of women receiving SSDI and non-SSDI recipients on variables of education, lifetime employment, homelessness, or marital status.

Table 1.

Means and proportions of demographic and health variables

	All	SSDI recipients	Non-SSDI recipients	p value
N	406	83 (20.4%)	323 (79.6%)
Age***	37.2	41.6	36.1	.000
High School/GED	73	70.0	74.0	.432
Has ever been employed in their lifetime	97.6	92.0	99.5	.001
Past-month Income***
<$500	52.5	15.7	63.5	.001
$500-$999	31.5	68.7	22.3
$1,000-$1,999	12.3	15.7	11.7
>$2,000	2.0	0.0	2.5
Living Alone***	16.5	31.3	12.7	.001
Homeless	34.1	26.5	36.0	.103
Marital Status
Single	44.5	43.4	44.8	.170
Married, in relationship	16.8	10.8	18.3
Divorced, separated, widowed	38.8	45.8	36.9
Race*
White	50.6	59.8	46.7	.035
Non-White
General Health*
Excellent	6.6	7.2	6.5	.022
Very Good	18.7	10.8	20.7
Good	40.9	33.7	43.7
Fair	26.8	37.3	24.1
Poor	6.9	10.8	5.9
Bodily Pain***
None/Very Mild	34.7	25.3	37.1	.001
Mild/Moderate	40.3	31.3	42.7
Severe/Very Severe	25.0	43.4	20.2
Walking
No Limitations in Activity	51.1	43.2	53.1	.273
Limitations in Activity <3 months	22.0	24.7	21.3
Limitations in Activity >3 months	26.9	32.1	25.6
Moderate Activity**
No Limitations in Activity	52.6	37.0	56.6	.004
Limitations in Activity <3 months	20.6	23.5	19.8
Limitations in Activity >3 months	26.8	39.5	23.6
Vigorous Activity
No Limitations in Activity	47.0	39.0	49.1	.061
Limitations in Activity <3 months	18.9	15.9	19.7
Limitations in Activity >3 months	34.1	45.1	31.3
HIV positive***	2.7	8.4	1.2	.001
Number of healthcare visits past 12 –months***	6.2	9.6	7.2	.001
Past 12-month Psychiatric Service Utilization***	39.4	59.0	34.4	.000
Past 12-month Psychiatric Medication***	37.4	61.4	31.5	.000
SUD treatment ever	66.6	69.9	65.7	.475

Significant User-Non-User differences

^*p<.05

^**p<.01

^***p<.001

As expected women receiving SSDI reported poorer overall health (p<.05) and more severe bodily pain (p<.001). Women receiving SSDI were more limited than non-recipients with regard to moderate activity, but there were no differences between groups in limitations of walking or vigorous activity (e.g., lifting heavy object, sports). Women receiving SSDI were more likely to be HIV positive (p<.001) and to have consumed more health services, as indicated by statistically significant differences in the mean number of past-12-month healthcare visits (p<.001), past-12-month psychiatric service utilization, (p<.001), and past-12-month psychiatric medication utilization (p<.001). However, there were no differences between groups in regard to lifetime SUD treatment.

Table 2 reports population differences of mental health, trauma, and social support variables. Several of the childhood victimization questions did not differ significantly between the groups, however questions regarding psychological abuse (Q1,2), physical abuse (Q3,4) and sexual abuse (Q5,6) were significantly different as women who received SSDI were more likely to report psychological, physical and/or sexual childhood abuse and to report that their first sexual experience was not consensual (p<.01). Additionally, women receiving SSDI were more likely to report childhood food insecurity (Q8; p<.05).

Table 2.

Means and proportions of mental health, trauma, and social support variables

	All	SSDI Recipients	Non-SSDI Recipients	p value
N	406	83 (20.4%)	323 (79.6%)
Childhood victimization
1. Did your parents or caretaker ever treat you like you were stupid or inferior and/or call you names in private?*	54.7	66.3	51.7	.018
2. Did your parents or caretaker ever insult, shame, or humiliate you in front of others?**	61.6	73.5	58.6	.013
3. Did your parents or caretaker ever physically hurt you on purpose (including grabbing, slapping, burning, scalding, punching, chocking, throwing you around, or harshly spanking you)?*	62.5	72.3	60.1	.040
4. Did your parents or caretaker ever beat you up?*	36.9	47.0	34.2	.031
5. Did your parents or caretaker ever force or threaten you to do sexual things other than intercourse (for example, forced petting or forced oral sex)?***	33.5	49.4	29.6	.001
6. Did your parents or caretaker ever force or threaten you to have sexual intercourse and it actually happened?**	22.5	33.7	19.7	.007
7. Did you ever see someone physically abuse your mother?	58.9	62.2	58.1	.504
8. While growing up, did you frequently not have enough food to eat?*	12.6	21.7	10.3	.021
9. While growing up, did you frequently feel unloved?	42.4	44.6	41.8	.371
10. During childhood, did you ever live with both your father and mother in the same home at the same time?	68.3	72.5	67.2	.362
11. Parents were never married or are divorced, separated, or widowed.	69.9	69.7	70.1	.947
Age of sexual intercourse	13.6	13.2	13.7	.293
First sexual intercourse was not consensual**	20.3	31.3	17.5	.006
Mental Health
Locus of Control***	17.0	19.1	16.5	.001
Self-esteem	12.2	12.1	12.2	.694
BSI Global Severity Index	1.18	1.31	1.14	.101
Anxiety	1.2	1.3	1.2	.247
Depression	1.3	1.4	1.2	.217
Somatization	1.0	1.2	1.0	.092
Obsessive-Compulsive	1.4	1.6	1.4	.135
Interpersonal Sensitivity	1.3	1.5	1.3	.164
Hostility	1.0	1.0	0.9	.369
Phobic Anxiety	0.9	1.0	0.9	.203
Paranoid Ideation	1.3	1.5	1.3	.062
Psychoticism	1.2	1.3	1.1	.190
PTSD
Meets diagnostic criteria for PTSD**	48.6	61.4	45.3	.009
Severity Total**	18.1	21.9	17.2	.006
Social Support*	10.5	9.2	10.5	.018

Significant User-Non-User differences

^*p<.05

^**p<.01

^***p<.001

Examination of relevant mental health variables revealed that women receiving SSDI had higher mean LOC scores, indicating a higher external locus of control whereby they were more likely to perceive themselves as less capable of impacting their lives and more likely to attribute success or failure to external forces No differences were found among self-esteem or any of BSI psychopathology scale scores. However, women receiving SSDI were more likely to meet the criteria for PTSD (p<.01) and report more severe PTSD symptoms. Finally, women were receiving SSDI scored lower for measures of social support than women not receiving SSDI (p<.05).

Discussion

This exploratory study established the prevalence of disability among a sample of women on probation/parole who have experienced victimization through examination of SSDI receipt. Among healthcare and social service providers, SSDI receipt is often used as a proxy for disability status. One advantage in using SSDI as a proxy for disability is that the SSA defines disorder-specific disability in stringent terms, meaning that if one is receiving SSDI, it is for significant functional impairment. A limitation of this method for measuring disability is that it excludes those who have not applied for SSDI, those whose applications have been denied, and possibly those with intermittent impairments and/or conditions with variable symptomatology severity, which may be more readily denied by the SSA. Given that SSDI can serve to corroborate self-reported disability status, it is a useful measure by which to determine disability rates among CJS-involved women, even if it is likely to produce some underestimate in the overall prevalence.

Disability prevalence was expected to be higher among women in this sample (Bronson et al., 2015; SSA, 2016; Steadman et al., 2009). When utilizing SSDI receipt as a proxy for disability, it was determined that one-fifth of the women were disabled in some capacity. This rate, while approximately 11% higher than the prevalence of disability among the general population in Kentucky, is nearly four times higher than the general population nationally (4.7% versus 20.6%) (SSA, 2011; 2016). However, it is still lower than rates of self-reported disability found in the nationally representative sample inmates in the NIS-3 (Bronson et al., 2015) However, utilization of SSDI receipt as a marker of disability more accurately measures the prevalence of severe and chronic disabilities. When excluding vision and hearing disabilities, the prevalence of 20% found in the current study is higher than the rate of ambulatory, self-care, and independent living disabilities reported in the NIS-3 among jail inmates and prison inmates. These three categories are more likely to reflect chronic disabilities that would be most likely to warrant SSDI receipt, since these disabilities include impairments that limit mobility and successful independent living tasks such as shopping (Bronson et al., 2015). The most prominent reason for SSDI receipt among adults in the U.S. is due to physical disorders (SSA, 2016), however mental health disorders comprise the second largest stand-alone disability category (26.5) next to musculo-skeletal system and connective tissue conditions (31.7%). Since the current data does capture the reason for SSDI receipt, it is unknown to the extent which specific disabilities or morbidities are represented in the current data.

Women in both the SSDI group and non-SSDI group were similar in many important areas, such as educational attainment, substance use history, and mental health symptomatology. However significant differences arose when comparing age, income, solitary living, and race. Given general profiles of those reporting disability, it was expected that women receiving SSDI would be older. However, the mean age for women receiving SSDI was 42.6. Nationally, women ages 40-44 comprise only 6.6% of all female SSDI-recipients (SSA, 2016). Women in this sample receiving SSDI were also more likely to have incomes in the $500-$999 range, which is the likely income category for women of their average age according to Social Security Administration data (SSA, 2016). Though these benefit amounts are low, they may serve as a protective factor in the lives of SSDI-recipients, not just economically, but also facilitating opportunities for these women to assume a positive role as provider (Angell, 2011; Hansen, Bourgois, & Drucker, 2014). Women reporting disability were also more likely to be White which is consistent with findings from national CJS-disability survey data (Bronson et al., 2015) but not general population trends (Centers for Disease Control and Prevention, 2015). It is unclear if this is due to qualitatively greater severity in impairments in functioning for White versus minority groups in this sample; greater access to screening and greater likelihood of SSDI benefit approval of for White compared to minority groups; or race-based discrepancies in self-report of SSDI status. Given the limited research in this area, further studies are needed to determine if this constitutes a trend of disability or a trend of bureaucracies.

Women receiving SSDI scored higher for external LOC than those not receiving SSDI. CJS-involved women (Asberg & Renk, 2014; Gussak, 2009), as well as individuals with a history of childhood victimization (Bolger & Patterson 2001), have been found to demonstrate higher external LOC. Higher external LOC has been associated with depressive symptoms and with less adaptive coping strategies against stress (Harrow, Hansford, & Astrachan-Fletcher, 2009). Women receiving SSDI were also more likely to live alone and to have weaker social support. When examining stress, LOC, and social support as predictors of psychological functioning among a sample of CJS-involved women, Asberg & Renk (2014) found that perceived social support accounted for half of the variance in depression and anxiety. Insufficient social support and social integration are associated with poor health outcomes among multiple populations (Berkman, Kawachi, & Glymour, 2014).

Individuals receiving SSDI often report poor general health and high rates of bodily pain, however, both groups had elevated rates of impaired functioning. For instance, approximately 20% of non-recipients reported severe/very severe bodily pain. Additionally, one quarter of this group reported limitations in walking or moderate activity which lasted more than three months. This finding indicates that some women may be functionally impaired, but not receiving SSDI benefits. Many individuals with chronic health conditions will at some point apply for SSDI, but a large portion will be denied (Fremstad & Vallas, 2013). There are multiple qualifying thresholds for SSDI eligibility, the first of which includes having a condition that meets or exceeds items deemed by the SSA to be a medical impairment (e.g. endocrine, neurological, immunological disorders) and which presents with significant severity so as to interfere with “basic work-related activities” (SSA, 2015). Part of this qualification includes that the SSDI applicant’s condition prohibits her from engaging in former work-related activities in addition to other potential forms of gainful employment (SSA, 2015). In other words, SSDI applicants must demonstrate that their condition hinders them from functioning in any occupational capacity. Beyond these criteria, there is significant variability between what medical conditions are deemed eligible for SSDI receipt (SSA, 2015). Between 2006-2008 only 40% of those who applied for SSDI were approved (SSA, 2013). It is possible that despite high rates of disability among CJS-involved women, some fail to meet the qualifying threshold for SSDI, thus placing them in highly vulnerable circumstances.

Similar to other investigations, the women in this exploratory study reported multiple forms of childhood adversity and victimization (Aday et al., 2014; McDaniels-Wilson & Belknap, 2008). Rape as a unique trauma has been associated with mental health disorders and poor health outcomes (Chen et al., 2010; Smith & Breiding, 2011). Approximately 19% of women in the U.S. have reported lifetime instances of rape (Black et al., 2011), which is consistent with the sample as a whole. However, 31.3% of women receiving SSDI in this study reported that their first sexual intercourse was non-consensual, compared to 17.5% of non-recipients. More broadly, actual and/or threatened forced sexual contact was reported by approximately one-third of the sample, and a majority of women reported childhood physical and psychological abuse. Adversity and victimization across the lifecycle are positively correlated with increased risk for acquiring adult morbidities (Danese et al., 2009), including complex presentations of PTSD (Cloitre et al., 2009; Green et al., 2016; Twaite & Rodriguez-Srednicki, 2004), further indicating that women who have experienced victimization are in need gender-specific, trauma-informed care as a possible means for improving chronic health outcomes (Covington, 2008; Messina, Calhoun, & Braithwaite, 2014).

Higher rates for past-12-month health care service utilization, including past-12-month psychiatric medication and psychiatric service utilization by women receiving SSDI suggests that, despite similarities for BSI scores, women with disabilities may be disproportionately impacted by trauma-related mental health disorders. This is supported by the finding that more SSDI-recipients met diagnostic criteria for PTSD and presented with greater PTSD symptom severity. Childhood adversity and PTSD are both independently and conjointly associated with mental and physical health conditions, including chronic pain (Coppens et al., 2017; Dube et al., 2009; Huffhines, Noser, & Patton, 2016; Outcalt et al., 2015; Scioli-Salter et al., 2015) and there is support for the idea that PTSD symptom severity mediates the relationship between childhood adversity and other mental health disorders in CJS-involved adults (Greene, Ford, Wakefield, & Barry, 2014). Harner and colleagues (2013) found that women with the most severe PTSD symptoms were likely to be linked to mental health services during incarceration, whereas only half of women with moderate-severe PTSD and comorbid issues were provided with care. Since prison has become a primary but imperfect provider of mental health services for women, those whose symptoms are not identified or treated while incarcerated may be less likely to access services upon release and may instead utilize higher-cost, emergency care which insufficiently meets their long-term needs (Leukefeld et al., 2006; Staton, Leukefeld, & Logan, 2001). Conversely, it could be that both groups in this sample are similar in their mental health profile, but that women receiving SSDI are somehow more inclined or able to regularly access healthcare (Ramaswamy, et al., 2015; Upadhyayula, Chan, Rhodes, & Leonardo, 2015). The interceding role of availability and continuity of care during and subsequent to incarceration in shaping disability outcomes for CJS-involved women requires further exploration. Future investigations into the determinants of perceived disability and SSDI rates for CJS-involved women should also closely examine the role that childhood trauma plays in adult risk for morbidities and disability.

Indeed, the relationship between trauma and poor health requires that policy makers and healthcare providers reexamine the efficacy of one-size-fits-all assessment and treatment practices for this population. This includes serious consideration of the idea of promoting gender-responsive, trauma-informed care as a standard intervention modality. Targeting PTSD symptomatology and addressing it concurrently with other conditions has been demonstrated as improving outcomes (Hien et al., 2009; Flanagan, Korte, Killeen, & Back, 2016). Shifting healthcare to a trauma-informed service model could improve the lives of CJS-involved women by preventing some health problems from developing, but also by addressing issues that are known to exacerbate serious health conditions that have the potential to deteriorate into chronic disability if overlooked or insufficiently treated (Harris & Fallot, 2001; Mason et al., 2016; Murphy et al., 2016).

Encouragingly, women who experience multiple psychosocial challenges are increasingly seeking treatment (Grella, 2012) and trauma-informed care is advancing as a “best practice” (Machtinger et al., 2015). However, substantial policy changes are being debated that would restrict the accessibility of medical insurance, including Medicaid (Eltorai & Eltorai, 2017; Kim & Richardson, 2012), along with revisions to SSDI aimed at steepening eligibility criteria while reducing benefits (Sperling, 2017). Many social welfare programs which CJS-involved women are likely to utilize have been systematically diminished over the past four decades and insufficient continuity of care between correctional institutions and community-based treatment providers persists (Hansen et al., 2014; Shafer, Prendergast, Melnick, Stein, & Welsh, 2014; Vail, Niyogi, Henderson, & Wennerstrom, 2017). For vulnerable populations with multiple health conditions, SSDI may be one of the few in-roads to relative stabilization. However, SSDI status may be less secure for CJS-involved individuals, as benefits are generally suspended after 30 consecutive days of incarceration and recipients are not automatically eligible upon release (SSA, 2017b). Rather, benefit reinstatement varies among persons and circumstances. Understanding how disruptions in benefit patterns impact the lives of CJS-involved women is another area that warrants exploration.

Limitations

The current study used SSDI status as a proxy to explore disability among CJS-involved women. As a result, disability rates are likely underestimated as only those with the most severe conditions and/or those who had already applied for and begun receiving SSDI benefits at time of study participation were able to be counted as disabled. We were unable to determine women’s functioning prior to entry into the CJS, the age of disability onset, and the circumstances under which impairments in functioning first materialized. Such information will be important in establishing a broader causal narrative regarding disability among CJS-involved women. Additionally, this study was exploratory and did not have information on specific disability type. Future research would benefit by exploring this subject in consideration of childhood adversities and victimization. Finally, findings should be considered in in the context of the state where the study was conducted. Kentucky ranks third in the nation for female incarceration rates, with 108 per 100,000 incarcerated as of 2015 and remains among the top five states for highest SSDI per capita, with approximately 7.3% of residents receiving SSDI (SSA, 2016; The Sentencing Project, 2015). Of those who have applied for SSDI in Kentucky, roughly 70% of initial applications are denied (Disability Benefits Center, 2017). States with differing characteristics may find different prevalence and traits among CJS-involved women receiving SSDI, though instances of adversity and prior victimization would likely continue to be a commonly observed feature across regions. Lastly, it is important to recognize that adversity and victimization may have been experienced by a portion of the women in this sample during childhood as well as adulthood, given that study inclusion criteria could be satisfied with either childhood or adult victimization experiences. Findings should be interpreted with the understanding that even among women receiving SSDI reporting no childhood adversity or victimization, these women may have experienced victimization in adulthood. That only variables related to childhood adversity and victimization were examined without equal consideration for such experiences in adulthood leaves open a gap for future research to fill. Specifically, it will be important to determine what differences exist between groups who have experienced childhood adversity and victimization exclusively verses groups who have experienced both, and to determine what factors mediate the relationship between types of victimization and disability.

Conclusion

This exploratory study described disability prevalence among women on probation and parole as a means of laying a foundation for future work. Consistent with previous investigations of CJS-involved women (Bronson et al., 2015; Wilper et al., 2009), we found high rates of disability. While some physical and mental health limitations linked to structural disadvantage may be more difficult to correct, the high rates of childhood victimization and PTSD among disabled women in this study highlights the importance of intervention for young, at-risk females, as well as the need for policy reforms which could promote accessibility of evidenced-based, gender-responsive, and trauma-informed care within the criminal justice system (Bowen & Murshid, 2016; Dobrow, Goel, & Upshur, 2004; Harner, Budescu, Gillihan, Riley, & Foa, 2015). In addition to establishing more accurate rates of disability prevalence for non-elderly women involved in the CJS, more research is needed in order to establish a better understanding of the possible relationships which exist between early childhood adversity and victimization, criminal offending, and disability among women.