In Tijuana, FSWs are required to obtain permits if they wish to work without prosecution in the Zona Roja, but in practice, more than half of FSWs operate without and work within and outside of the Zona Roja. In Cd. Juarez, a permit is not required and two Zonas Rojas exist. In both cities, FSWs operate out of cantinas, bars, hotels, nightclubs and street corners that cater to men of different cultural backgrounds. Carrier identified nine different types of FSWs in Mexico, ranging from “street walkers” and call girls to companions for parties and vacations (Carrier, 1989). Although some FSWs in these cities are relatively high income earners, most are from lower socioeconomic strata. Most enter prostitution out of economic necessity (Bucardo et al., 2004); for example, due to a failed relationship or intolerable working conditions in maquiladoras (Grossman et al., 1985), (i.e., assembly plants that use their low labor costs to attract manufacturing contracts from foreign firms). Factors related to their mobility, such as extreme poverty and low social support, may heighten their vulnerability to HIV/STIs (Fernandez, 1998, Rangel et al., 2006).

Tijuana and Cd. Juarez are central corridors for smuggling illicit drugs into the U.S. After Colombia, Mexico is the second most important source of heroin entering the U.S., accounting for 30% of all heroin sold in the U.S. and 98% of heroin sold west of the Mississippi river (Bucardo et al., 2005). Approximately 90% of all methamphetamine entering the U.S. is produced in Mexico, and 70% of all cocaine entering the U.S. passes through Mexico en route from South America (Brouwer et al., 2006a). Illicit drug use—particularly injection drug use—has increased in Tijuana and Cd. Juarez over the past ten years as local drug consumption markets have emerged along trafficking routes (Magis-Rodríguez et al., 2002).

The highest consumption of illegal drugs in Mexico is in Baja California, whose largest city is Tijuana. Tijuana has one of the fastest growing injection drug user (IDU) populations in Mexico (Magis-Rodríguez et al., 2002, Secretaría de Salud, 2002). The proportion of the general population in Tijuana aged 12–65 years who reported having ever used an illegal drug was 15%, three times the national average. In Mexico overall, men were 13 times more likely than women to have ever used illicit drugs, but in Tijuana the ratio was 6:1, indicating a high proportion of female drug users (Secretaría de Salud, 2002). In 2003, officials estimated there were ~6000 IDUs attending ‘picaderos’ (shooting galleries) in Tijuana (Morales et al., 2004), but the total number is thought to be closer to 10,000. Among Mexican cities, Cd. Juarez is ranked second only to Tijuana in the number of illicit drug users, which is twice the national average (Lloyd, 2003). In 2001, it was estimated that there were ~6,000 ‘heavy heroin users’ in Cd. Juarez (Cravioto, 2003).

Many FSWs in Tijuana and Cd. Juarez turn to stimulants like methamphetamine or cocaine—alone or in combination with heroin—to help them cope or stay awake (Patterson et al., 2006). FSWs who inject drugs (FSW-IDUs) may experience an elevated risk of acquiring both sexually transmitted infections (STIs) and blood borne infections. In cases where drugs are injected, women who share injection equipment are at risk of acquiring HIV or other blood borne infections such as viral hepatitis. Although HIV prevalence is generally low in Mexico, the prevalence of hepatitis C virus infection among IDUs in Tijuana and Cd. Juarez exceeds 90%, whereas prevalence of hepatitis B exceeds 80% (White et al., 2007).

In contrast, STI prevalence among FSWs in Mexico has been high for decades. The prevalence of active syphilis, Chlamydia and gonorrhea among FSWs in Mexico City was 23.7%, 12.8% and 11.6%, respectively (Valdespino-Gómez et al., 1998), and prevalence of HSV-2 and oncogenic HPV strains was 60% and 43%, respectively (Uribe-Salas et al., 1999, Juarez-Figueroa et al., 2001). Among FSWs on the Mexico-Guatemala border, prevalence of syphilis, Chlamydia, gonorrhea, HSV-2 and HIV was 9%, 14%, 12%, 86% and 0.6%, respectively; low HIV prevalence was attributed to the low prevalence of injection drug use, which likely differs from Mexico's northern border with the U.S.

Among FSWs who injected drugs within the last month (N=114), polydrug use was common. The majority reported injecting heroin (93.9%), cocaine/heroin combinations (i.e., “speedball”; 50.0%), cocaine alone (36.0%) or methamphetamine (21.1%). Not unexpectedly, FSW-IDUs were also more likely to report non-injection use of other drugs, including marijuana/hash, heroin and methamphetamine (Table 1). In the past month, 52.6% reported using a needle someone else had used, 73.7% had passed on their own needle to someone else, and 70.3% had shared injection paraphernalia (i.e., cottons, cookers, rinse water). One fifth reported sharing needles with a sex trade client (18.9%).

Compared with FSWs who did not inject drugs within the past month FSW-IDUs were slightly younger (30 vs. 33 years), more likely to be married/common-law, speak English, mainly work on the street rather than a bar, and have been a sex worker for a longer duration. In contrast, compared to other FSWs, FSW-IDUs were less likely to have had children, live in their own home, and reported living with fewer people (Table 1).

In terms of social influence, FSW-IDUs were more likely to report having a steady partner who was also an IDU and fellow FSWs who injected drugs, but were no more likely to report having a pimp or to have clients who used drugs compared to non-IDUs. Compared to other FSWs, FSW-IDUs had greater numbers of clients, and were paid significantly less for sex with a condom. There was no significant difference between FSWs and FSW-IDUs for average amount earned for sex without a condom, or in the median difference earned for vaginal sex with and without a condom (Table 1).

In terms of risk behaviors, the groups did not differ in terms of the proportions who had vaginal sex without a condom, although FSW-IDUs were marginally less likely to use a condom for anal sex. FSW-IDUs were less likely to use alcohol before or during sex, but were more likely to use illicit drugs in these situations. FSW-IDUs were also more likely than other FSWs to report at least one IDU sex partner in the last month and have male clients that currently use or inject drugs (Table 1).

Compared to other FSWs, FSW-IDUs were more likely to have an STI at baseline including HIV, syphilis, Chlamydia, and gonorrhea. HIV prevalence was 12.3% among FSW-IDUs, relative to 5.1% among other FSWs. Overall, 44.7% of FSW-IDU had at least one of these STIs, compared to 24.4% among other FSW (p<0.01) (Table 1).

We examined correlates associated with being an FSW-IDU in univariate logistic regression models. A generally risky pattern emerged. FSW-IDU were more likely to describe oneself as a street worker (OR=5.05, 95% CI 3.00-8.51 ), be married/common-law (OR=2.35, 95% CI 1.56-3.55), and live in Tijuana as opposed to Ciudad Juarez (OR=1.98, 95% CI 1.30-2.99). The odds of being an FSW-IDU were lower for those who lived in their own home and those who worked mainly in a bar. Sex workers whose steady partner was an IDU (OR=12.19, 95% CI 6.24-23.79) and who reported that most fellow FSWs were IDUs were more likely to be FSW-IDUs (OR=1.80, 95% CI 1.50-2.16). Sex workers who often or always used drugs before sex and those who had more clients with whom they had unprotected vaginal sex acts were much more likely to be FSW-IDUs. Finally, FSW-IDUs were more likely to be HIV positive (OR=2.63, 95% CI 1.38-4.99) and to test positive for at least one STI (Table 2).

In the final multivariate model, a number of factors were independently associated with recent injection drug use within this FSW population. After controlling for site and all other factors, FSW-IDUs were more likely to describe themselves as a street worker (OR=3.85), speak English (OR=2.79), be married/common law (OR=2.01), have worked in the sex trade for more than four years (OR=2.12), to have lived for a longer duration in the study location and to report that some of their fellow sex workers were IDUs (OR=2.42) (Table 3). In contrast to other FSWs, those who injected drugs were younger, less likely to live in their own home, and less likely to have had children. They were also less likely to live and work in the same location, and were more likely to have been paid less than $30 US on average for sex with a condom.

Variables that did not retain significance at the 5% level in the multivariate model included being born in the state where they now live, education level, total number of people they live with, mainly working in a bar, whether their steady partner was an IDU, having at least one IDU sex partner in the last month, having male clients who currently use drugs or have injected drugs, median number of clients in the last six months, number of unprotected vaginal sex acts with clients in the last six months, often or always having protected anal sex, and often or always using alcohol before vaginal sex.

HIV prevalence among FSW-IDUs was more than twice that of FSWs who did not inject drugs (12% vs. 5%). This suggests that the window of opportunity for HIV prevention is closing rapidly in these cities, since IDU-associated HIV epidemics are often explosive and quickly become generalized (Strathdee et al., 1998, Aceijas et al., 2004). The high prevalence of syphilis titers consistent with active infection is worrisome, because untreated syphilis can cause debilitating illness and congenital blindness. Syphilis, gonorrhea and Chlamydia can also lead to infertility and are known cofactors of HIV transmission (Laga et al., 1993). The states of Baja California and Chihuahua have the highest rates of adult and congenital syphilis among Mexico's 32 states (CENSIDA, 2006). Although cases of active syphilis, gonorrhea and Chlamydia are eligible for free treatment under Mexico Ministry of Health guidelines regardless of health insurance status, the tendency to rely on a passive system for treatment, rather than an active system involving contact tracing and partner notification, may contribute to these elevated infection rates.

In terms of sociodemographic characteristics, FSW-IDUs were more likely to live in Tijuana than Cd. Juarez, were younger, more likely to speak English and less likely to have had children or live in their own home. The fact that FSW-IDUs were more likely to speak English may reflect the tendency for Mexicans who are more acculturated to a North American lifestyle, to have higher rates of drug use (Grant et al., 2004). The higher proportion of FSWs who reported injection drug use in Tijuana is consistent with the fact that Baja California has historically had the highest prevalence of substance abuse across Mexico (Magis-Rodríguez et al., 2002, Secretaría de Salud, 2002).

We observed that FSWs who injected drugs tended to have social networks that may perpetuate their drug use. FSW-IDUs were more than twice as likely to report that most of the sex workers they knew injected drugs, which could suggest a group dynamic through which women buy and use drugs together. Although studies on female drug users in Mexico are sparse, a study of young female IDUs in Baltimore suggested that women initiated other women to inject drugs (Doherty et al., 2000). FSW-IDUs were also twice as likely as other FSWs to be married or live in a common-law relationship. A qualitative study among male and female IDUs in Tijuana and Cd. Juarez found that female IDUs often used drugs within the context of a sexual relationship with a drug using male partner (Cruz et al., 2007). International literature suggests that female IDUs tend to have greater overlap in their sexual and drug use networks relative to males (Latkin et al., 2003, Sherman and Latkin, 2001, Cruz et al., 2007), which can contribute to being “second on the needle” (Cruz et al., 2007, Harvey et al., 1998). In a study of Hispana female IDUs on the U.S. side of the U.S.-Mexico border, Andrade and Estrada highlight the importance of sociocultural and gender norms to be taken into account when considering their vulnerabilities (Andrade and Estrada, 2003). The extent to which network and sociocultural influences may interact with needle sharing or sexual behaviors require further study.

In general, we observed a riskier pattern of sexual behaviors among FSW-IDUs compared to other FSWs in these cities. FSW-IDUs were marginally more likely to have unprotected anal sex with clients compared to other FSWs, although this association did not persist after adjusting for other confounders. FSW-IDUs were almost seven times more likely to use illicit drugs before sex than other FSWs. This is troubling because sex work conducted under the influence of drugs has been associated with lower rates of condom use in other settings (Plant et al., 1989, Paone et al., 1999, de Graaf et al., 1995). In a study conducted in the United Kingdom (Gossop et al., 1995), FSWs' drug use was associated with having unprotected sex with clients in exchange for more money, which is not surprising since women suffering from withdrawal symptoms may be more likely to acquiesce to demands for unsafe sex to support their drug habit. On average, we found that FSW-IDUs earned less money for sex with and without condoms than other FSWs, which in turn may lead them to have greater numbers of clients and/or more unprotected sex.

This study was limited because we lacked a number of unmeasured contextual variables (e.g., relating to recruitment venue and study site) that may differ between FSW-IDUs and FSW non-IDUs. We were also lacking information on the context of sexual behaviors with clients, particularly the extent to which they exchanged sex for drugs rather than money, which has been reported elsewhere (Edlin et al., 1994, Sterk, 1998, Inciardi, 1995, McMahon et al., 2006). However, nearly half of all the FSWs in our study reported that their clients used drugs and one fifth reported sharing needles with clients. These circumstances may predispose FSWs to using drugs during sex, and could contribute to their difficulties negotiating safer sex.

Since women were recruited through convenience sampling and higher risk women were selected for the purposes of testing an intervention study, our findings may not reflect the experience of FSWs who consistently use condoms with clients. Future studies using respondent driven sampling would improve upon these estimates (Heckathorn, 2002). Finally, small cell sizes precluded us from comparing ever, never and former injectors using ordinal regression. However, results were similar in a re-analysis comparing FSWs who ever versus never injected.