Abstract
Criminal justice system involvement among US women is associated with increased risk for HIV/STIs, yet effects of different forms of criminal justice involvement on intimate relationships are not well understood. This study examined associations between arrest, probation, and jail incarceration on the number of sexual partners, sexual concurrency, and consistent condom use for drug-using women (n=631) in Oakland, California. We used logistic and negative binomial regression and adjusted for demographics, sex exchange and drug use. Probation was associated with higher rates of sexual partnership and concurrency (IRR 1.87, 95% CI [1.11, 3.15]; OR 3.64, 95% CI [1.08, 12.20]). Incarceration lasting over 12 weeks was associated with higher rates of sexual partnership (IRR 2.23, 95% CI [1.41–3.51]). Women incarcerated once in the past year had higher odds of concurrency (OR 2.15, 95% CI [1.01–4.57]). Our results reinforce the need for risk-reduction interventions and criminal justice diversion for women who use drugs.
Keywords: Incarceration, Sexual health, Sexual behavior, Drug use
Resumen
Involucramiento de mujeres estadounidenses en el sistema de justicia penal está asociado con riesgo aumentado de VIH / ITS, aunque los efectos de las formas diferentes de involucramiento en la justicia penal en las relaciones íntimas no se comprenden bien. Este estudio examinó las asociaciones entre el arresto, la libertad condicional y el encarcelamiento de la cárcel y el número de parejas sexuales, la concurrencia sexual, y el uso consistente de condones para mujeres que consumen drogas (n = 631) en Oakland, California. Utilizamos regresión logística y binomial negativa y se ajustamos para la demografía, el intercambio sexual y el uso de drogas. La libertad condicional se asoció con tasas más altas de parejas sexuales y concurrencia (IRR 1.87, IC del 95% [1.11, 3.15]; OR 3.64, IC del 95% [1.08, 12.20]). El encarcelamiento más de 12 semanas se asoció con tasas más altas de parejas sexuales (IRR 2.23, 95% CI [1.41–3.51]). Las mujeres que habían sido encarceladas una vez en el último año tenían mayores probabilidades de concurrencia (OR 2.15, IC del 95% [1.01–4.57]). Nuestros resultados refuerzan la necesidad de intervenciones enfocadas en la reducción de riesgo y el desvío del sistema de justicia penal para las mujeres que consumen drogas.
Introduction
Mounting evidence demonstrates that involvement in the criminal justice system is either an independent risk factor for sexual risk behaviors and sexually transmitted infections (STIs), or is a marker for other, potentially unmeasured characteristics that shape risk. Women who have been arrested or incarcerated are more likely to report high risk partnerships, multiple partnerships, condomless sex, and transactional sex than women who have never been incarcerated (1–6). Consistent with this, incarcerated women have high rates of chlamydia (722%) and gonorrhea (3–9%), more than double those of the general population (7–13). The risk of HIV and other STIs extends across partnerships, and sexual networks and communities with high rates of criminal justice involvement have elevated HIV/STI risk, even among those who are not themselves involved in the criminal justice system (14–16).
The mechanisms underlying these associations are not well understood, particularly among criminal-justice involved women. Research on incarcerated men’s relationships suggests incarceration may lead to relationship disruption and thus the acquisition of new sexual partners and may also result in the concentration of risk into sexual networks where a large proportion of both men and women have been incarcerated (15, 17, 18). Indeed, approximately half of men’s primary relationships end during a period of incarceration, and recently incarcerated men report that they wish to increase their number of partners once they are released from prison or jail, to account for the time they were incarcerated (17, 19). In addition, incarcerated men report that once they are released from prison they are seen as healthier, more desirable, and with more potential as romantic partners, perhaps in part due to high rates of male incarceration which reduce the overall number and quality of potential partners (15, 18). There has been comparatively little work on the effects of women’s incarceration on their intimate relationships. Substance abuse, emotional needs, perceptions of risk and self-worth, partner pressure and violence, and economic self-sufficiency have all been identified as factors that affect HIV risk both before and after incarceration (20). In addition, the exchange of sex for money or drugs places many formerly incarcerated women at heightened risk for sexual violence and condomless sexual encounters (21).
This study aimed to examine the temporal associations between arrest, probation and jail incarceration on the number of sexual partners, rates of sexual concurrency, and consistent condom use in a sample of women who use drugs in Oakland, California, and to determine whether engagement in sex exchange or use of particular drugs could account for these relationships. Previous work has demonstrated that experiencing an arrest in the past six months is associated with increased sexual risk, but no previous studies have examined whether there is a dose-response relationship (6). Few prior studies have collected sufficiently granular data on incarceration to measure an effect of timing and duration, with the notable exception of Khan and colleagues who reported that incarceration episodes lasting one day to one month may increase risky partnerships (defined as having multiple partners or engaging in sex exchange) more than episodes lasting greater than one month (1, 6).
This work was motivated by hypotheses outlined below regarding the effects of women’s criminal justice involvement on sexual risk behavior.
Increasing frequency of arrests and episodes of incarceration for women will be associated with increasing numbers of sexual partners and increasing odds of having concurrent relationships, defined as sexual partnerships that overlap in time, due to temporary relationship disruption.
Increasing duration of probation and incarceration for women will be associated with an increase in the number of sexual partners due to permanent relationship disruption. Previous analyses that used a composite outcome of numbers of partners and engagement in sex exchange might have missed this effect on non-transactional relationships.
Increasing duration of probation and incarceration will be associated with a decrease in the rates of concurrency because the longer time away will be more likely to end a relationship permanently.
Any criminal justice involvement will be associated with less consistent condom use.
Methods
Study Design
The Oakland Women’s Health Study consisted of a cross-sectional survey which aimed to measure unmet healthcare needs and the range and accumulation of criminal justice involvement in a population of women who use illicit drugs. The study is described in detail elsewhere (22). Targeted sampling methods were used to recruit 631 adult women in Alameda County, California, who had used crack/cocaine, methamphetamine, or heroin within the 30 days prior to recruitment. The survey included questions about sexual risk behaviors and a detailed inventory of involvement in the criminal justice system. The interviewers and other study staff were trained to refer respondents to local services and provide resources for support in addition to the study incentive.
Measures
Sexual risk.
Respondents were asked about the total number of male and female sexual partners they had in the past six months. These were added together for the total number of sexual partners in the past six months. Concurrency was assessed by asking “did any of [these sex partners] overlap? In other words you had sex with partner A, then partner B, and then partner A again.” Sexual exchange was measured with two questions by asking respondents whether, in the past six months, any male sex partners had given them money in exchange for sex or had given them drugs in exchange for sex. Respondents were asked whether, in the last six months, they used condoms with male partners. The responses “never” and “sometimes” were coded as inconsistent condom use, and “always” was considered consistent use.
Criminal justice involvement.
Respondents were asked whether they had ever been arrested, how old they were at the time of first arrest, and how many times they had been arrested. They were also asked whether they had been arrested in the past year.
Jail incarceration experiences were measured by asking whether the individual had ever spent time, even one day, in city or county jail. This was followed up asking the number of lifetime jail incarcerations and the total time spent in jail as an adult, and how old they were the first time they were sent to jail as an adult. Respondents were then asked how many times they had been to jail in the past year, and the total amount of time in the past year they had spent in jail. This same series of questions was repeated for prison incarceration.
Respondents were also asked whether they had ever been on probation, and then whether they had been on probation in the past year. This was followed by asking the number of months and weeks spent on probation in the past year. The same questions were repeated for parole. Fewer than ten women reported being in prison or on parole in the past year, so no analyses were performed using these variables.
Partner incarceration was measured by asking respondents whether a steady partner had ever been in jail, and whether they had been in jail in the past year. These same questions were repeated for prison incarceration. A steady partner was defined for respondents as either a husband, wife, girlfriend, boyfriend, fiancé or fiancée. Fewer than ten women reported partners with prison incarcerations in the past year, and so only the jail variables were used in the analyses.
Demographics.
Age was asked directly of the respondents and was confirmed using the respondent’s date of birth. Respondents were asked about whether they had a high school diploma or GED. Respondents were asked for their race (White, Black, Asian, Pacific Islander, or Native American, and could indicate multiple categories) and ethnicity (Latina or Not Latina). Asian/Pacific Islander and Native American respondents made up a very small portion of the sample, and so only dichotomous variables for Black race and White race were used for analysis, and other racial categories were considered non-Black and non-White.
Drug use.
Respondents were asked to indicate whether, in the past six months, they had used marijuana, crack or rock cocaine, powder cocaine by itself, heroin by itself, speedball (heroin plus cocaine or crack), other opiates (Oxycontin, codeine, Percocet, etc) without a prescription, benzodiazepines (Valium, Klonopin, Xanax etc.), methamphetamine (crystal), or methadone without a prescription. Binary variables were created for each drug and were added to the models separately.
Statistical Approach
Descriptive statistics for the study sample were generated to provide a foreground of respondents’ experiences with the criminal justice system and their demographics, sexual behavior, and drug use. Bivariate logistic (concurrency) and negative binomial (number of sexual partners) regression models were created to calculate unadjusted odds ratios with the independent variables of demographics, criminal justice involvement, sex exchange and drug use. Bivariate logistic regression models were also used to identify relationships between criminal justice involvement and both sex exchange and drug use, in order to explore potential confounding relationships. Then, multivariable models were constructed (again using logistic and negative binomial regression) containing those variables which were significant at the 0.05 level in the bivariate analyses.
Results
The study sample consisted of 631 women in Alameda County, California who had used drugs in the past 30 days. The most commonly reported drugs used in the past six months were crack (81%), marijuana (64%), heroin (40%), cocaine (35%), other opioids (32%), and methamphetamine (26%), as shown in Table I. The average age was 46 years, and respondents ranged from 18–68 years old. The women in our sample predominantly identified as Black (86%), with much smaller proportions identifying as Latina (6%), White (5%), Native American (2%), or Asian/Pacific Islander (1%).
Table I.
Description of the study sample
| Variable | Percent (N) | |
|---|---|---|
| Age | ||
| 18–24 | 3.2 (20) | |
| 25–34 | 13.9 (88) | |
| 35–44 | 21.9 (138) | |
| 45–54 | 37.1 (234) | |
| 55–64 | 21.1 (133) | |
| 65 and over | 2.9 (18) | |
| Race | ||
| Black | 86.1 (543) | |
| White | 5.1 (32) | |
| Native American | 2.4 (15) | |
| Asian | 0.3 (2) | |
| Pacific Islander | 0.3 (2) | |
| Latina Ethnicity | 6.0 (38) | |
| Educational attainment | ||
| Less than high school/GED | 36.8 (232) | |
| High school/GED or more | 63.2 (399) | |
| Drug use in the past 6 months | ||
| Marijuana | 63.9 (402) | |
| Crack | 80.8 (508) | |
| Cocaine | 35.5 (223) | |
| Heroin | 40.2 (253) | |
| Speedball | 13.4 (84) | |
| Other opiates | 32.3 (203) | |
| Benzodiazepines | 21.9 (138) | |
| Methamphetamines | 26.4 (166) | |
| Methadone | 10.4 (65) | |
| Totala | 100 (631) | |
Categories may not sum to the total if there was missing data or participants were allowed to select more than one option (race/ethnicity and drug use)
Approximately half (47%) endorsed having a steady partner (Table II). The mean number of male partners in the past six months was twelve. Although the majority reported having only one partner in the past six months, there were 22 women (2%) who reported more than 100 partners in the past six months (range 0–500). Over half of respondents (55%) reported exchanging sex for money or drugs with men in the past six months. Approximately 21% (n=107) had a positive gonorrhea, chlamydia, or trichomonas test at the time of their last test. The prevalence of HIV in the sample was 2.7% (n=17), compared with an overall prevalence for women in Alameda County of 0.1% (23). Approximately one third of women reported never using condoms and another third reported only sometimes using condoms.
Table II.
Sexual partnership characteristics of study sample
| Variable | Percentage (N) |
|---|---|
| Currently have a steady partner | 47.1 (297) |
| Number of male partners in the past six months | |
| 0 | 19.7 (124) |
| 1 | 33.4 (211) |
| 2–3 | 18.2 (115) |
| 4–10 | 11.4 (72) |
| >10 | 17.3 (109) |
| Number of female partners in the past six months | |
| 0 | 89.4 (564) |
| 1 | 5.7 (36) |
| 2–3 | 3.3 (21) |
| 4–10 | 0.3 (2) |
| >10 | 1.3 (8) |
| Concurrent male sexual partners | 39.8 (250) |
| Sex exchange in the past six months (male partners) | 55.3 (260) |
| Condom use in the past six months (male partners) | |
| Never | 31.4 (148) |
| Sometimes | 34.4 (162) |
| Always | 34.2 (161) |
| Steady partner was in jail in the past year | 29.1 (66) |
| Totala | 100 (631) |
Categories may not sum to the total if there was missing data
The majority of respondents had been arrested (86%) or on probation (71%), and most had also spent some amount of time in jail (83%). Nearly one quarter (22%) had spent time in prison. Details of respondents’ criminal justice involvement are shown in Table III.
Table III.
Frequency of criminal justice involvement as an adult in the study sample, and extent of involvement for women who had experienced each type
| Variable | Percent (N) |
|---|---|
| Ever arrested | 86.4 (539) |
| Ever on probation | 69.9 (436) |
| Ever been in jail | 83.2 (519) |
| Ever been in prison | 21.5 (116) |
| Arrested in the past year | 26.3 (140) |
| Jail in the past year | 23.1 (146) |
| Mean (Range) | |
| Number of arrests – past yeara | 1.5 (1–7) |
| Age at first arresta | 24.5 (18–57) |
| Total time on probation – past year (months)a | 8.3 (0.25–12) |
| Number of jail episodes – past yeara | 1.9 (1–25) |
| Total time spent in jail – past year (weeks)a | 8.5 (1–52) |
| Number of prison episodes – lifetimea | 4.1 (1–40) |
| Total time spent in prison – lifetime (years)a | 4.7 (0.1–28) |
Calculated for only those respondents who had experienced the outcome.
In the initial bivariate analyses of the number of sexual partners, age greater than 45 years was a significant predictor of having fewer sexual partners (incident rate ratio (IRR) = 0.49, 95% CI [0.36, 0.66]). The age variable was initially divided into ten-year categories, however there were not statistically significant differences between the younger age groups, and so this was collapsed into age less than 45 years and age greater than 45 years. Latina women reported fewer sexual partners than non-Latina identified women (IRR = 0.44, 95% CI [0.21, 0.77]), but there was not a significant effect of Black or White racial identity on the number of sexual partners. Women with a high school education, general equivalency diploma (GED), or more education had fewer sexual partners than women who had not completed high school/GED (IRR 0.65, 95% CI [0.48, 0.89]). Engaging in sexual exchange was associated with an extremely high incidence rate ratio for sexual partnership (IRR = 22.2, 95% CI [17.30, 28.47]). Of the drug use variables, crack use, methamphetamine use, and methadone use in the past six months were associated with higher rates of sexual partnership (IRRs 2.49, 3.17, and 2.12, respectively, and 95% CIs [1.72, 3.60], [2.25, 4.46], and [1.24, 3.61]). Benzodiazepine use was associated with having fewer partners (IRR 0.57, 95% CI [0.39, 0.85]). Heroin use was not significantly associated with the number of sexual partners. The unadjusted incidence rate ratios for all of the arrest-, probation-, and jail-related variables were significant, and predicted higher numbers of sexual partners (shown in Table IV, Column 1).
Table IV.
Unadjusted and adjusted incident rate ratios (IRR) for the number of sexual partnersa and odds ratios (OR) for concurrent partnershipb
| Sexual Partners | Sexual Partners | Concurrency | Concurrency | |||||
|---|---|---|---|---|---|---|---|---|
| Unadjusted | Adjusted | Unadjusted | Adjusted | |||||
| Variables | IRR | CI (95%) | IRRα | CI (95%) | OR | CI (95%) | AORβ | CI (95%) |
| Arrested in the past year (Ref=No) | 2.54*** | (1.77, 3.65) | 1.24 | (0.91, 1.68) | 2.47*** | (1.67, 3.67) | 1.73 | (0.90, 3.32) |
| Number of arrests in the past year | 1.36* | (1.01–1.85) | 1.00 | (0.80, 1.26) | 1.21 | (0.91–1.60) | 1.26 | (0.83, 1.93) |
| Age at first arrest | 0.92*** | (0.90, 0.94) | 0.99 | (0.97, 1.01) | 0.95*** | (0.92, 0.97) | 1.02 | (0.98, 1.07) |
| Probation in the past year (Ref=No) | 2.59*** | (1.60, 4.21) | 1.55* | (1.04, 2.31) | 2.18** | (1.30, 3.67) | 2.01 | (0.83, 4.83) |
| Total time on probation in the past year – months (Ref =0) | ||||||||
| 1–11 | 1.83 | (0.94, 3.58) | 1.24 | (0.74, 2.09) | 1.92 | (0.94, 3.97) | 1.48 | (0.45, 4.92) |
| 12 | 3.38*** | (1.77, 3.48) | 1.87* | (1.11, 3.15) | 2.84** | (1.38, 5.87) | 3.64* | (1.08, 12.20) |
| Jail in the past year (Ref = No) | 2.48*** | (1.75, 3.52) | 1.20 | (0.89, 1.61) | 2.18*** | (1.50, 3.19) | 1.60 | (0.86, 3.00) |
| Total time in jail in the past year – weeks (Ref=0–1) | ||||||||
| 2–3 | 1.75 | (0.44, 6.90) | 1.06 | (0.39, 2.92) | 2.22 | (0.49, 10.05) | 1.17 | (0.11, 12.39) |
| 4–12 | 1.20 | (0.63, 2.28) | 0.79 | (0.46, 1.33) | 1.58 | (0.80, 3.13) | 1.46 | (0.45, 4.73) |
| Over 12 | 4.53*** | (2.41, 8.56) | 2.23*** | (1.41, 3.51) | 2.95** | (1.46, 5.96) | 1.05 | (0.36, 3.03) |
| Number of times admitted to jail in the past year (Ref=None) | ||||||||
| Once | 2.64*** | (1.73, 4.03) | 1.16 | (0.82, 1.64) | 2.44*** | (1.54, 3.87) | 2.15* | (1.01, 4.57) |
| More than once | 2.22** | (1.32, 3.75) | 1.27 | (0.84, 1.92) | 1.82* | (1.03, 3.20) | 0.97 | (0.39, 2.41) |
| Steady partner in jail in the past year (Ref=No) | 3.72*** | (2.25, 6.16) | 1.48 | (0.97, 2.25) | 2.53** | (1.41, 4.54) | 2.07 | (0.75, 5.72) |
Boldface indicates significance
p<0.05;
p<0.01;
p<0.001.
Adjusted for age, race and ethnicity, educational attainment, sex exchange for money or drugs, crack use in the past 6 months, benzodiazepine use in the past 6 months, meth use in the past 6 months, and methadone use (not from a program) in the past 6 months
Adjusted for age, educational attainment, sex exchange for money or drugs, coke use in the past 6 months, and meth use in the past 6 months
In the bivariate analyses of the odds of concurrent partnerships, odds were decreased for women over the age of 45 (OR 0.40, 95% CI [0.29, 0.56]), and for women who had at least a high school education or GED (OR 0.57, 95% CI [0.41, 0.79]. There was no significant effect of the racial or ethnic identity variables on the odds of concurrent partnerships. As with the number of sexual partners, engaging in sex exchange was a significant predictor of having concurrent partnerships (OR 26.33, 95% CI [16.08, 43.12]). Cocaine use and methamphetamine use in the past six months were also associated with higher odds of concurrent partnerships (ORs 2.80 and 2.91, respectively, with 95% CIs [1.38, 5.70] and [1.98, 4.26]). All of the arrest-, probation-, and jail-related variables were significantly associated with increased rates of concurrent partnerships in the unadjusted analyses (shown in Table IV, Column 3).
Engaging in sex exchange was associated with increased odds of consistent condom use (OR 2.55 95% CI [1.70, 3.83]). No other variables, and in particular none of the criminal justice involvement variables, were associated with condom use, and so a multivariable model was not constructed.
Bivariate analyses of the relationships between criminal justice involvement and specific types of drug use demonstrated that women who used methamphetamines had greater odds of arrest in the past year (OR 3.28, 95% CI [2.18, 4.94]), of more frequent arrest (OR 1.75, 95% CI [1.24, 2.47]), of being on probation in the past year (OR 2.17 95% CI [1.29, 3.68]), and of having spent the entire past year on probation (OR 2.97 95% CI [1.48, 5.95]). Methamphetamine use was also associated with increased odds of jail incarceration in the past year (OR 2.88 95% CI [1.94, 4.28]), having one jail incarceration (OR 3.17 95% CI [1.97, 5.07], reference = none), having two or more jail incarcerations (OR 2.47, 95% CI [1.38, 4.42], reference = none), spending 4–12 weeks in jail (OR 5.52 95% CI [2.71, 11.2], reference = 0–1 weeks), and having a partner in jail in the past year (OR 3.07 95% CI [1.67, 5.66]). Crack use was associated with decreased odds of having a partner in jail in the past year (OR 0.50, 95% CI [0.27, 0.94]). Other specific substances (benzodiazepines, methadone, and other forms of cocaine) were not significantly associated with criminal justice involvement in this sample.
The bivariate relationships between sex exchange and criminal justice involvement were statistically significant for arrest in the past year (OR 1.86, 95% CI [1.18, 2.94]), jail incarceration in the past year (OR 1.87, 95% CI [1.21, 2.91]), having one jail incarceration (OR 1.69, 95% CI [1.00, 2.85], reference = none), having two or more jail incarcerations (OR 2.26, 95% CI [1.12, 4.57], reference = none), and spending over 12 weeks in jail in the past year (OR 3.03, 95% CI [1.27, 7.19], reference = 0–1 weeks). The relationships between sex exchange and frequency of arrest in the past year, probation in the past year, amount of time on probation in the past year, and having a partner in jail in the past year were not statistically significant.
Once age, race and ethnicity, educational attainment, sex exchange for money or drugs, crack, benzodiazepine, methamphetamine, and methadone use in the past six months were included as covariates in the models of rates of sexual partnership over the past six months, many of the associations with the criminal justice involvement variables were diminished and were no longer statistically significant (shown in Table IV, Column 2). The variables that remained statistically significantly associated with an increased number of sexual partners were experiencing probation in the past year, spending the entire past year on probation, and spending more than 12 weeks in jail in the past year (Table IV, Column 2).
The ORs for concurrent partnerships were adjusted for age, educational attainment, sex exchange for money or drugs, and cocaine and methamphetamine use in the past six months. With this adjustment, many of the odds ratios were reduced and no longer reached statistical significance. The associations with increased odds of concurrent partnerships that remained statistically significant were those with spending the entire past year on probation and experiencing one jail incarceration in the past year (Table IV, Column 4).
Discussion
The relationship between high risk sexual behavior and involvement in the criminal justice system is a complex one. It is clear that many factors which lead to arrest, probation, and jail incarceration, such as using drugs and exchanging sex for money or drugs, also lead to high risk sexual behaviors. In these analyses, we were able to look at the effect of criminal justice involvement in the past year controlling for these other risk behaviors. We found that even after accounting for sex exchange and drug use, women who had been on probation, and in particular those who had been on probation for the full year, reported a higher number of sexual partners compared with women who were not on probation. In addition, women incarcerated for more than 12 weeks in the past year, but not those who had been incarcerated for shorter durations, reported a higher number of sexual partners in the past six months compared with women who were not incarcerated. This supports our hypothesis that a longer cumulative duration of probation and incarceration would lead to an increase in the number of sexual partnerships, perhaps because of relationship disruption.
We also found that women who had been on probation during the entire year, and those who had been incarcerated once in the past year, but not more than once, had higher rates of concurrent partnerships even after controlling for sex exchange and drug use; this may be suggestive of temporary, rather than permanent disruptions in relationships in the setting of a single, short episode of incarceration or supervision in the community. We did not find evidence to support our hypothesis that longer duration of probation or incarceration would decrease rates of concurrency.
We also found that in our sample, methamphetamine use in particular is strongly associated both with two measures of sexual risk (the number of partners and rates of concurrency) and with criminal justice involvement, making it an important confounder in our analyses. Indeed, the inclusion of the drug use variables in our multivariable models reduced the magnitude of our effect sizes and in many cases reduced statistical significance as well. This is consistent with prior work that demonstrates that women who use methamphetamine have high rates of sexual risk behavior (24). It is notable that even with the inclusion of drug use variables, there was a significant effect of both probation and jail time. This suggests that even once the effects of drug use are accounted for, the temporary and permanent disruptions of criminal justice surveillance and supervision may result in additional increases in the number of partners and rates of concurrent partnerships.
With regard to sex exchange, our results suggest that it may act as a partial mediator between criminal justice involvement and our measures of sexual risk (25). Sex exchange meets the criteria set forth by Baron and Kenny, in that it is significantly associated with both criminal justice involvement and our measures of sexual risk, and when it is included in the multivariable models, the effect sizes and statistical significance of the relationships between criminal justice involvement, number of sexual partners, and rates of concurrency are reduced. It may be that women increase their engagement in sex work during probation or after incarceration to account for economic hardships that come with criminal justice involvement. There are, however, multiple potential confounding relationships between sex exchange and criminal justice involvement and, without longitudinal data, our ability to draw firm conclusions about mediation relationships is limited (26).
Our findings diverge somewhat from previous work on the effects of criminal justice involvement on an individual’s number of sexual partners. It has previously been reported that incarceration episodes lasting from one day to one month, more so than episodes lasting longer than one month, increased the odds of reporting multiple partners or engagement in sex exchange among African American women in North Carolina with a low prevalence of substance use (1). By contrast, we found that longer duration of incarceration was associated with a higher number of sexual partners. There are several potential explanations for this; first, it may be that because our study separated out jail incarcerations, which are on average quite short, while previous work combined both jail and prison incarcerations, that we are seeing the same phenomenon but with different categorical cut points. Second, it is important to note that this earlier study used sex exchange as an outcome, rather than as a covariate, and so the underlying conceptual model is slightly different. Thirdly, these samples represent quite disparate groups of women. Taken together, the two studies suggest that individuals who are incarcerated for very brief periods are more likely to be engaged in sex work, and as a result have more sexual partners, but that separate from sex exchange, episodes of incarceration that last more than twelve weeks are most likely to disrupt other relationships and result in the acquisition of new sexual partners.
Arrest, as measured in our study, may simply be serving as a marker for other risk behaviors. The associations between arrest and high risk sexual behavior were almost entirely accounted for by engagement in sex exchange and drug use. This is consistent with prior work, which has shown that arrest in the past six months is associated with high risk sexual behavior, when sex exchange was included in that measure of high risk behavior (6). For women who exchange sex or use drugs or both, the presence of police and the possibility of arrest may be daily realities, and the frequency with which an individual is arrested likely reflects her exposure to the criminal justice system due to these behaviors.
There has previously been little description of how probation, a degree of involvement in the criminal justice system between the most superficial (policing and arrest) and the most intense (incarceration in jail or prison), affects sexual risk behavior. In our sample, probation had more effects on the number of partners and concurrency than arrest; any probation was associated with an increase in the number of partners, and a longer duration of probation was associated both with an increased number of partners and increased rates of concurrency. This suggests that this intermediate level of criminal justice supervision and surveillance may also result in relationship disruptions, similar to the effects of incarceration, but because the individual remains in the community, fewer of these relationship disruptions are permanent and both the number of partners and rates of concurrency are increased. This is particularly noteworthy because probation supervision is by far the most common form of criminal justice involvement among women.
The study is not without limitations. The exclusive recruitment of women who use drugs, although resulting in a data set that represents a group that is often excluded from research, means that generalizations of these results to all women involved in the criminal justice system may need to be tempered. The high rates of substance use among criminal justice-involved women, however, mean that analyses of these data represent the majority of women arrested, on probation, and in jail, and certainly many of those at highest risk for HIV and other STIs. Our study sample that skews toward middle age is likely to miss some of the higher risk behaviors and patterns of high risk partnerships that are more common in younger women. The cross-sectional nature of the study also significantly limits the ability to make causal statements about the relationships we observed and to draw conclusions about mediation and other confounding relationships. We did limit our exposures to the last year, and outcomes of sexual partnership to the last six months in order to try to capture a temporal effect, but it remains a possibility that the sexual risk behaviors pre-dated the criminal justice episodes. Finally, because women were asked the cumulative amount of time they were incarcerated or on probation, rather than the duration of their longest episode of incarceration or probation, the distinction between frequency and duration in our data set is somewhat blurred; for example, among the women who had been on incarcerated for 12 or more weeks in the past year, some may have experienced on three month incarceration while others were incarcerated for two weeks on six different occasions.
It is notable that in this sample of women who use drugs in Alameda County, California, involvement in the criminal justice system was a normative experience. The vast majority had been arrested and had spent time in jail, which is consistent with previous work with women who use drugs (27, 28). Indeed, a large number of women in this study experienced an intersection of circumstances that place them at exceptionally high risk for HIV and other STIs, including sex exchange, drug use, and criminal justice involvement. From these results suggesting that there is an independent association between probation, jail incarceration and high risk sexual behavior, we conclude that in the interest of harm reduction we must work toward not only implementing risk-reducing interventions (such as medication-assisted treatment and pharmacologic pre-exposure prophylaxis for HIV) for women involved in the criminal justice system, but also toward reducing the burden of criminal justice involvement experienced by women who use drugs.
Acknowledgements
The work of Drs. Lambdin, Comfort, Kral, and Lorvick on this study was supported by a grant from NIMHD (RO1MD007679). Dr. Knittel received support from the University of North Carolina at Chapel Hill Center for AIDS Research (CFAR), an NIH funded program (P30 AI50410). The authors would like to thank Ms. Reana Thomas for her assistance with manuscript preparation.
References
- 1.Khan MR, Miller WC, Schoenbach VJ, et al. Timing and duration of incarceration and high-risk sexual partnerships among African Americans in North Carolina. Ann Epidemiol 2008;18:403–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Rogers SM, Khan MR, Tan S, et al. Incarceration, high-risk sexual partnerships and sexually transmitted infections in an urban population. Sex Transm Infect 2012;88(1):63–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Widman L, Noar SM, Golin CE, Willoughby JF, Crosby R. Incarceration and unstable housing interact to predict sexual risk behaviours among African American STD clinic patients. Int J STD AIDS 2014;25(5):348–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wise A, Finlayson T, Nerlander L, et al. Incarceration, sexual risk-related behaviors, and HIV infection among women at increased risk of HIV infection, 20 United States cities. J Acquir Immune Defic Syndr 2017;75(S3):S261–S7. [DOI] [PubMed] [Google Scholar]
- 5.Khan MR, Epperson MW, Mateu-Gelabert P, et al. Incarceration, sex with an STI- or HIV-infected partner, and infection with an STI or HIV in Bushwick, Brooklyn, NY: A social network perspective. Am J Public Health 2011;101(6):1110–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Epperson MW, Khan MR, Miller DP, et al. Assessing criminal justice involvement as an indicator of human immunodeficiency virus risk among women in methadone treatment. J Subst Abuse Treat 2010;38(4):375–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fox KK, Whittington WL, Levine WC, et al. Gonorrhea in the United States, 1981–1996: Demographic and geographic trends. Sex Transm Dis 1996;25(7):386–93. [DOI] [PubMed] [Google Scholar]
- 8.Joesoef MR, Weinstock HS, Kent CK, et al. Sex and age correlates of Chlamydia prevalence in adolescents and adults entering correctional facilities, 2005: implications for screening policy. Sex Transm Dis 2009;36(2 Suppl):S67–71. [DOI] [PubMed] [Google Scholar]
- 9.Hardick J, Hsieh Y-H, Tulloch S, et al. Surveillance of Chlamydia trachomatis and Neisseria gonorrhoeae infections in women in detention in Baltimore, Maryland. Sex Transm Dis 2003;30(1):64–70. [DOI] [PubMed] [Google Scholar]
- 10.Mertz KJ, Schwebke JR, Gaydos CA, et al. Screening women in jails for chlamydial and gonococcal infection using urine tests: Feasibility, acceptability, prevalence, and treatment rates. Sex Transm Dis 2002;29(5):271–6. [DOI] [PubMed] [Google Scholar]
- 11.Datta SD, Torrone E, Kruszon-Moran D, et al. Chlamydia trachomatis trends in the United States among persons 14 to 39 years of age, 1999–2008. Sex Transm Dis 2012;39(2):92–6. [DOI] [PubMed] [Google Scholar]
- 12.Spaulding AC, Seals RM, Page MJ, et al. HIV/AIDS among Inmates of and Releasees from US Correctional Facilitities, 2006: Declining Share of Epidemic but Persistent Public Health Opportunity. PLoS One 2009;4(11):e7558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kouyoumdjian FG, Leto D, John S, Henein H, Bondy S. A systematic review and metaanalysis of the prevalence of chlamydia, gonorrhoea and syphilis in incarcerated persons. Int J STD AIDS 2012;23(4):248–54. [DOI] [PubMed] [Google Scholar]
- 14.Knittel AK, Snow RC, Riolo RL, Griffith DM, Morenoff J. Modeling the communitylevel effects of male incarceration on the sexual partnerships of men and women. Soc Sci Med 2015;147:270–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Green TC, Pouget ER, Harrington M, et al. Limiting options: sex ratios, incarceration rates, and sexual risk behavior among people on probation and parole. Sex Transm Dis 2012;39(6):424–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Dauria EF, Elifson K, Arriola KJ, Wingood G, Cooper HLF. Male incarceration rates and rates of sexually transmitted infections. Sex Transm Dis 2015;42(6):324–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Khan MR, Behrend L, Adimora AA, et al. Dissolution of primary intimate relationships during incarceration and implications for post-release HIV transmission. J Urban Health 2011;88(2):365–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Dauria EF, Oakley L, Arriola KJ, et al. Collateral consequences: implications of male incarceration rates, imbalanced sex ratios and partner availability for heterosexual Black women. Cult Health Sex 2015;17(10):1190–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Thomas JC, Levandowski BA, Isler MR, Torrone E, Wilson G. Incarceration and sexually transmitted infections: A neighborhood perspective. J Urban Health 2007;85(1):90–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Fogel CI, Gelaude DJ, Carry M, et al. Context of risk for HIV and sexually transmitted infections among incarcerated women in the south: individual, interpersonal, and societal factors. Women & Health 2014;54(8):694–711. [DOI] [PubMed] [Google Scholar]
- 21.Fogel CI, Belyea M. The lives of incarcerated women: Violence, substance abuse, and at risk for HIV. J Assoc Nurses AIDS Care 1999;10(6):66–74. [DOI] [PubMed] [Google Scholar]
- 22.Lorvick J, Comfort M, Kral AH, Lambdin BH. Exploring Lifetime Accumulation of Criminal Justice Involvement and Associated Health and Social Outcomes in a CommunityBased Sample of Women who Use Drugs. Journal of urban health : bulletin of the New York Academy of Medicine 2017. [DOI] [PMC free article] [PubMed]
- 23.HIV Epidemiology & Surveillance Unit. HIV in Alameda County, 2014–2016 Oakland, CA; 2018. [Google Scholar]
- 24.Semple SJ, Grant I, Patterson TL. Female Methamphetamine Users: Social Characteristics and Sexual Risk Behavior. Women & Health 2005;40(3):35–50. [DOI] [PubMed] [Google Scholar]
- 25.Baron RM, Kenny DA. The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations. Journal of Personality and Social Psychology 1986;51(6):1173–82. [DOI] [PubMed] [Google Scholar]
- 26.Maxwell SE, Cole DA. Bias in cross-sectional analyses of longitudinal mediation. Psychol Methods 2007;12(1):23–44. [DOI] [PubMed] [Google Scholar]
- 27.Rich JD, Cortina SC, Uvin ZX, Dumont DM. Women, incarceration, and health. Womens Health Issues 2013;23(6):e333–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Smith VC. Substance-abusing female offenders as victims: Chronological sequencing of pathways into the criminal justice system. Vict Offender 2017;12(1):113–37. [DOI] [PMC free article] [PubMed] [Google Scholar]