Gender Differences in the Relationship between Methamphetamine Use and High-risk Sexual Behavior among Prisoners: A Nationwide, Cross-sectional Survey in Japan

ABSTRACT

This study aimed to examine the relationship between methamphetamine (MA) use and high-risk sexual behavior among newly incarcerated prisoners in Japan. We conducted a secondary analysis using cross-sectional survey data, which constituted a nationwide prison sample surveyed between July and November 2017. In total, 699 participants (462 male and 237 female prisoners), who completed the self-administered questionnaire during imprisonment were recruited for the study. The prevalence of high-risk sexual behaviors was as follows: inconsistent condom use (78.4% male, 81.7% female), multiple sexual partners (61.3% male, 41.3% female), history of sexually transmitted diseases (14.1% male, 23.6% female), and trading sex for drugs (15.6% male, 17.7% female). A logistic regression analysis showed that MA use during sex increased the likelihood of high-risk sexual behavior in both male (adjusted odds ratio [AOR] = 5.86; 95% confidence interval [CI] = 3.41–10.07) and female prisoners (AOR = 2.58; 95% CI = 1.33–5.00). Additionally, in female prisoners, a history of sharing MA injection (AOR = 1.60; 95% CI = 1.06–2.42) increased high-risk sexual behavior. Our results indicate that harm reduction programs tailored according to the specific necessities of each gender to reduce high-risk sexual behaviors among MA users should be included in Japanese drug policies.

KEYWORDS:

• Gender differences
• Japan
• methamphetamine
• prisoners
• sexual behavior
• sexually transmitted diseases

Introduction

Methamphetamine (MA) is a highly addictive stimulant and widely used as a recreational drug internationally. Japan has traditionally maintained a zero-tolerance policy toward illicit drugs, and the use of drugs such as MA has been strictly prohibited by the Stimulants Control Act. According to a nationwide general population survey on drug use in Japan, MA is the third most prevalent illicit drug after cannabis and inhalants. The survey conducted in 2019 reported that the lifetime prevalence of MA use is 0.4% of the population, i.e., 350,000 people are estimated to have used MA in their lifetimes, and there are more male than female MA users (about 300,000 males and about 50,000 females) (Shimane et al. 2020).

In this study, we focused on prisoners who use MA rather than those who use other illicit drugs. This is because cases of MA use-related disorder are most common in clinical settings. According to a nationwide mental hospital survey, the proportion of cases of MA use disorders has continued to be the highest since 1993 (Matsumoto et al. 2019). However, there are a few cases related to other drugs including cannabis and inhalants. Historically, Japanese drug users have administered MA through injection; in the 1990s, smoking emerged as a new form of drug use. According to a study on MA users in a psychiatric hospital, the routes of administration were as follows: injection only (49.1%); smoking only (36.2%); and initially smoking, later injecting (14.7%) (Matsumoto et al. 2002). Japanese drug policy is strengthening health services for drug users, including MA users. For instance, in 2006, the psychosocial program based on the Matrix model (Obert et al. 2000) was developed and implemented in psychiatric hospitals (Kobayashi et al. 2007).

MA use and high-risk sexual behavior

The risk of contracting infectious diseases such as viral hepatitis and HIV infection is one of the health problems related to MA use. Studies from Myanmar and China have revealed that a large proportion of MA users engaged in high-risk sexual behaviors such as inconsistent condom use (Saw et al. 2018), multiple sexual partners (Liu et al. 2013; Saw et al. 2018), and exchange of sexual partners during MA use (Liu et al. 2013). Regarding gender, a study from China reported that male users were more likely to have had multiple sex partners (male 96.9%, female 77.3%) and exchanged sex partners during MA use (male 72.9%, female 46.4%) than female users (Liu et al. 2013).

The relationship between MA use and high-risk sexual behavior in the Japanese population is not well documented. This may be because there are only a few reports of HIV infection with injection drug use as a transmission route. Sexual contact in men who have sex with men (MSM) is a major route for HIV infection transmission in Japan. According to a sexual behavior survey of the general population conducted in 1999, 2.1% of females had unprotected vaginal sex with casual partners and 8.1% had multiple sex partners in the past year (Ono-Kihara et al. 2010). However, MSM have a higher prevalence of high-risk sexual behavior than the general population (unprotected sex 22.1%, multiple sex partners 58.1%) (Hidaka et al. 2006).

MA use among prisoners in Japan

According to the Ministry of Justice, a total of 4,378 new inmates (3,811 male and 567 female) related to the use of MA were imprisoned in Japan, in 2018. These inmates were 95% native Japanese, 2% Korean, and 3% nationals of other countries, and a majority of them were Japanese male. Prisoners who used MA had a high frequency of recidivism, and the recidivism rate within 5 years was 48.5% (Research Department, Research and Training Institute, Ministry of Justice, Japan 2018). In prison, all substance users including MA users are required to participate in intervention programs to recover from drug addiction based on the Japanese law; Act on Penal Detention Facilities and the Treatment of Inmates and Detainees.

A study on criminal justice clients with substance abuse problems from Sweden revealed that having a partner with substance abuse issues was one predictor of criminal recidivism among female clients (Mannerfelt and Håkansson 2018); a lifetime history of substance use was associated with criminal recidivism among male clients. Hence, we hypothesized that there are gender differences in relationships between MA use and high-risk sexual behaviors among prisoners. Therefore, this study evaluated the relationship between MA use and high-risk sexual behavior among prisoners throughout Japan.

Methods

Sample

The present study was conducted as a secondary analysis of the original project, titled “Health survey for drug offenders 2017,” collected by the Research and Training Institute, Ministry of Justice, Japan. The primary project aimed to collect basic statistics about the health of MA offenders. Data were collected between July and November 2017, at all 78 prisons throughout Japan. Inclusion criteria were: (1) newly incarcerated prisoners who had violated the Stimulants Control Act; (2) prisoners who have used MA at least once in their lifetime; (3) prisoners who were incarcerated due to MA use; (4) prisoners who were able to read and write Japanese; (5) prisoners who provided informed consent. Inmates placed in medical prisons were excluded from the study due to the severity of their condition. Newly incarcerated prisoners were defined as individuals who were admitted to prison within 20 days of data collection.

We approached 806 prisoners (542 males and 264 females). Of them, 107 refused to participate in the survey. In total, 699 prisoners (462 males and 237 females) participated in the self-administered survey, giving a response rate of 86.7%. Previous studies have confirmed the representativeness of our samples (Shimane et al. 2019). If the entire MA user inmate population was compared with the participants of our study, the ratio of age groups and the number of recidivism due to MA use were quite similar.

In the original project, informed consent was obtained from each participant. The National Center of Neurology and Psychiatry signed a Material Transfer Agreement with the Research and Training Institute, Ministry of Justice, and took over an anonymized dataset for a secondary analysis. The study protocol was reviewed and approved by the Ethics Committee of the National Center of Neurology and Psychiatry (approval number: A2017–107).

Measures

The questionnaire, which was designed by the Ministry of Justice, and National Center of Neurology and Psychiatry, included scales whose reliability and validity have been confirmed, and consisted of 15 and 16 sections for male and female participants, respectively.

Dependent variables

We defined sexual activity as follows: “Sexual activity refers to not only vaginal intercourse but also anal or oral intercourse.” We evaluated the high-risk sexual behaviors of participants using the following four indicators. All indicators were self-reported evaluations.

Inconsistent condom use (ever) was assessed with the question “Have you ever had sex without a condom, due to the influence of the drug you or your sexual partner used?” Response options were “none,” “sometimes,” and “mostly.” This variable was dichotomized into “No: none” and “Yes: sometimes, mostly, and unknown.”

Multiple sexual partners (the preceding year) were assessed with the question, “During the year before you entered prison, with how many different people have you had sex?” Response options were “none,” “only one person,” “2‒4 people,” “5‒9 people,” and “more than 10 people.” Responses were dichotomized into No: “none,” or “only one person” and Yes: “having multiple sexual partners” if the participant had had more than two sexual partners in the preceding year.

Sexually transmitted disease (STD) history (ever)Was assessed by asking the participants about histories of hepatitis A, hepatitis B, chlamydia, syphilis, HIV, and gonorrhea, which had been diagnosed by a medical professional. Responses were categorized into having (Yes) or not having a history of STD (No).

Trading sex for drugs (the preceding year) was assessed with the question “During the year before you entered prison, did you trade sex for drugs?” Response options were “Yes,” “No,” and “No sexual activity in the preceding year.” Responses were dichotomized as Yes: having traded sex for drugs in the preceding year and No: no sexual activity or not having traded sex for drugs in the preceding year.

In previous studies, inconsistent condom use and multiple sexual partners were used as direct measures, and STD history was used as an indirect measure (Browne et al. 2009; Saw et al. 2018). We added “trading sex for drugs” to these three indicators because previous studies had shown that MA users, particularly female MA users, traded sex for drugs or money (Lion et al. 2017; Molitor et al. 1998; Semple et al. 2011).

Independent variables

Demographic variables of the prisoners were collected from the prisons in which they were incarcerated, and included gender (in the family register), age, marital status, employment status, and criminal record. The variable age was referenced to the data at the time of the survey. The variables marital status and employment status were referenced to the data at the time before prison incarceration. Marital status was categorized into “unmarried” and “other,” based on five categories: unmarried, married, divorced, bereaved, and unknown. Employment status was categorized as “employed” or “unemployed.” Recidivism includes incarceration for all crimes, not just those related to drug use. Recidivism was categorized as “first time” or “second or more times” based on the number of times in prison.

MA use-related variables were selected based on a previous study, which included age at MA initiation (Saw et al. 2017), length of MA use (Liu et al. 2013), reasons for MA use (Cheng et al. 2009), frequency of MA use (Saw et al. 2018), MA use during sex (Saw et al. 2018), sharing MA injection (Lorvick et al. 2006), and having partners who used MA (Liu et al. 2013), DAST-20 and AUDIT (Trenz et al. 2016).

Participants indicated the age at which MA was used for the first time, and the age at which MA was used for the last time. The difference between last use and initiation was calculated and defined as “length of MA use.” Participants selected the reasons for using MA within a year preceding incarceration. There were 17 items and the participant could select all that applied. Frequency of MA use was defined as the average number of days MA was used in a month within a year preceding incarceration. The participant reported the average number of days that MA was used on the questionnaire.

To assess MA use during sex, we included the question “Have you ever used MA during sex or 2 hours before starting sex?” Responses were dichotomized into having (yes) or not having (no) a history of MA use during sex. To assess the history of sharing MA injections, we included the question “Have you ever used MA with a shared syringe?” Response options were “never,” “sometimes,” and “frequently.” Responses were dichotomized into having (yes) or not having (no) a history of sharing MA injections. To assess having partners that used MA, we included the question “Currently, do you have partners using MA?” We have defined a partner as a “spouse or live-in partner.” Response options were “yes,” “no,” “unknown,” and “I currently have no partners.” Responses were dichotomized into having (yes) or not having (no) a partner using MA.

To measure the severity of drug use-related problems, we used the Drug Abuse Screening Test (DAST-20, score range: 0‒20) (Skinner 1982). The severity of alcohol use-related problems was measured using the Alcohol Use Disorders Identification Test (AUDIT, score range: 0‒40) (Bohn, Babor, and Kranzler 1995). Both screening tests evaluate the severity within a year preceding incarceration. We used the Japanese test versions that had their reliability and validity previously confirmed (Hiro and Shima 1996; Shimane et al. 2015).

In this study, we evaluated the prisoners’ past history of MA use and high-risk sexual behaviors, not current behaviors. Japanese prisoners are in a highly controlled environment, which make it very difficult for them to engage in drug use and sexual activities. Therefore, we evaluated drug use and sexual behaviors when participants were out of prison.

Statistical analysis

First, we examined gender differences in demographic parameters by bivariate analysis. Next, we calculated the prevalence of variables related to high-risk sexual behavior and MA use and conducted bivariate analysis to examine gender differences. In these statistical analyses, significant differences were confirmed using t-test for continuous variables and Pearson’s chi-square test for categorical variables. The threshold for statistical significance was set at P < .05 (two-tailed).

Finally, we performed logistic regression analysis using a generalized estimating equation (GEE) to identify covariates associated with high-risk sexual behavior. Independent variables for the final model were selected based on previous studies (Mannerfelt and Håkansson 2018; Saw et al. 2018). The GEE model is robust with regard to inflated alpha. To address this problem, we adopted the most conservative unstructured correlation design. This approach assumed that the four indicators were correlated and equivalent to each other, not independent. This is the same approach that was applied in previous studies (Browne et al. 2009; Saw et al. 2018). Covariates including “age” and “age at MA initiation” were checked for multicollinearity before performing the final models. Data were analyzed using the Statistical Package for the Social Sciences (SPSS) version 26 (SPSS Inc., Chicago, IL, USA).

Results

Bivariate analysis

Table 1 shows the demographic variables. Compared with female prisoners, male prisoners were more likely to be older, unmarried, employed, and recidivistic (p < .001 for all).

Table 1. Socio-demographic characteristics^a of methamphetamine-using prisoners by gender (N = 699)

Characteristics	Total (n = 699)		Male (n = 462)		Female (n = 237)
	n	%	n	%	n	%	p-value
Age, mean (SD) (years)	43.5 (10.3)		44.5 (10.7)		41.7 (9.1)		<0.001
Age, range (years)	22–78		22–78		24–70
Marital status							<0.001
Other	511	73.4	311	67.5	200	85.1
Unmarried	185	26.6	150	32.5	35	14.9
Employment status							<0.001
Employed	294	42.7	239	52.4	55	23.6
Unemployed	395	57.3	217	47.6	178	76.4
Recidivism							<0.001
First time	181	25.9	99	21.4	82	34.6
Second or more times	518	74.1	363	78.6	155	65.4

^aSocio-demographic characteristics included age at the time of survey, marital status before incarceration, employment status before incarceration, and recidivism.

Table 2 shows the results related to high-risk sexual behaviors. The high-risk sexual behaviors were “inconsistent condom use” (male 78.4%, female 81.7%), “multiple sexual partners” (male 61.3%, female 41.3%), “STD history” (male 14.1%, female 23.6%), and “trading sex for drugs” (male 15.6%, female 17.7%). Male prisoners were more likely to have had multiple sexual partners than females (p < .001). Female prisoners were more likely to have an STD history than males (p = .002). No gender differences were found for “inconsistent condom use” (p = .310) and “trading sex for drugs” (p = .489).

Table 2. Variables related to high-risk sexual behavior^a by gender (N = 699)

	Total (n = 699)		Male (n = 462)		Female (n = 237)
	n	%	n	%	n	%	p-value
Inconsistent condom use							0.310
No	137	20.4	95	21.6	42	18.3
Yes	533	79.6	345	78.4	188	81.7
Multiple sexual partners							<0.001
No	304	45.6	169	38.7	135	58.7
Yes	363	54.4	268	61.3	95	41.3
STD history							0.002
No	578	82.7	397	85.9	181	76.4
Yes	121	17.3	65	14.1	56	23.6
Trading sex for drugs							0.489
No	559	83.7	368	84.4	191	82.3
Yes	109	16.3	68	15.6	41	17.7

^aHigh-risk sexual behavior was measured using four indicators: 1) inconsistent condom use (ever), 2) multiple sexual partners (the preceding year), 3) STD history (ever), and 4) trading sex for drugs (the preceding year).

Table 3 shows the results related to MA use. Mean age at MA initiation was in the early twenties in both males (22.9 years) and females (22.4 years). Mean length of MA use was 20.6 years for males and 18.9 years for females. Most participants engaged in MA use during sex (male 84.4%, female 88.6%) and shared MA injection (male 66.4%, female 75.7%). Female participants were more likely to frequently use MA than male participants (male 8.0/30 days, female 9.7/30 days). Female participants were more likely to have partners who used MA than male participants (male 29.9%, female 60.3%). Regarding “reasons for MA use,” male participants were more likely to answer “to get sexual pleasure and excitement” (male 60.8%, female 32.9%) than females. Alternatively, females were more likely to answer “to escape reality” (male 37.0%, female 46.0%) and “to lose weight” (male 7.6%, female 43.9%) than males. The mean DAST-20 score was significantly higher in females than in males (p = .003), but the mean AUDIT score was similar in males and females.

Table 3. Variables related to methamphetamine use^a by gender (N = 699)

	Total (n = 699)		Male (n = 462)		Female (n = 237)
	n	%	n	%	n	%	p-value
Age at MA initiation, years
mean (SD)	22.8 (7.3)		22.9 (6.9)		22.4 (7.9)		0.414
Length of MA use, years
mean (SD)	20.0 (10.7)		20.6 (11.0)		18.9 (9.9)		0.053
Reasons for MA use
To get sexual pleasure and excitement	359	51.4	281	60.8	78	32.9	<0.001
To improve concentration	316	45.2	224	48.5	92	38.8	0.015
To deal with a depressed mood or anxiety	311	44.5	201	43.5	110	46.4	0.464
To recover from fatigue	309	44.2	196	42.4	113	47.7	0.185
To escape reality	280	40.1	171	37.0	109	46.0	0.022
To relieve drowsiness	269	38.5	184	39.8	85	35.9	0.308
To get high and feel happy	256	36.6	171	37.0	85	35.9	0.766
To kill time	219	31.3	149	32.3	70	29.5	0.464
To relieve pain and physical symptoms	213	30.5	129	27.9	84	35.4	0.041
To calm down	162	23.2	101	21.9	61	25.7	0.250
To work actively	148	21.2	96	20.8	52	21.9	0.722
To lose weight	139	19.9	35	7.6	104	43.9	<0.001
To sharpen sight and hearing	120	17.2	83	18.0	37	15.6	0.435
To get excited with friends	113	16.2	71	15.4	42	17.7	0.424
To have confidence in yourself	89	12.7	48	10.4	41	17.3	0.009
To talk to people without being shy	44	6.3	23	5.0	21	8.9	0.045
To avoid exclusion by peers	18	2.6	11	2.4	7	3.0	0.651
Frequency of MA use
mean/30 days (SD)	8.6 (9.5)		8.0 (9.0)		9.7 (10.2)		0.040
MA use during sex							0.132
No	99	14.2	72	15.6	27	11.4
Yes	600	85.8	390	84.4	210	88.6
Sharing MA injection							0.013
No	206	30.5	150	33.6	56	24.3
Yes	470	69.5	296	66.4	174	75.7
Having partners using MA							<0.001
No	418	59.8	324	70.1	94	39.7
Yes	281	40.2	138	29.9	143	60.3
DAST-20
mean (SD)	9.6(4.1)		9.3(3.9)		10.3(4.4)		0.003
AUDIT
mean (SD)	6.7(6.9)		6.8(6.4)		6.6(7.7)		0.672

DAST: Drug Abuse Screening Test, AUDIT: Alcohol Use Disorders Identification Test.

^aVariables related to MA use included age at MA initiation, length of MA use, reasons for MA use (the preceding year), frequency of MA use (the preceding year), MA use during sex (ever), sharing MA injection (ever), having partners who use MA (at the time before incarceration), DAST-20 score (the preceding year), and AUDIT score (the preceding year).

Multivariate analysis

After adjusting for the effects of covariates (Table 4), MA use during sex increased the likelihood of high-risk sexual behavior in both male (adjusted odds ratio [AOR] = 5.86; 95% CI = 3.41‒10.07) and female prisoners (AOR = 2.58; 95% CI = 1.33‒5.00). Additionally, in female prisoners, the history of sharing MA injections (AOR = 1.60; 95% CI = 1.06‒2.42) and score of DAST-20 (AOR = 1.07; 95% CI = 1.03‒1.12) increased the likelihood of high-risk sexual behavior. Alternatively, age (AOR = 0.97; 95% CI = 0.95‒0.99) decreased the likelihood of high-risk sexual behavior.

Table 4. Generalized estimating equation for the high-risk sexual behavior^a of methamphetamine-using prisoners by gender

	Male (n = 462)				Female (n = 237)
	Crude OR	95% CI	AOR^b	95% CI	Crude OR	95% CI	AOR^b	95% CI
Marital status
Other	1.00		1.00		1.00		1.00
Unmarried	0.94	0.77–1.14	1.06	0.86–1.31	1.30	0.86–1.96	1.23	0.78–1.94
Employment status
Employed	1.00		1.00		1.00		1.00
Unemployed	1.03	0.86–1.23	0.98	0.80–1.21	1.49	1.06–2.09	1.42	0.99–2.01
Recidivism
First time	1.00		1.00		1.00		1.00
Second or more times	0.94	0.75–1.20	0.83	0.63–1.09	0.99	0.74–1.34	0.96	0.69–1.33
MA use during sex
No	1.00		1.00		1.00		1.00
Yes	5.46	3.48–8.57	5.86	3.41–10.07	4.46	2.46–8.09	2.58	1.33–5.00
Sharing MA injection
No	1.00		1.00		1.00		1.00
Yes	1.45	1.18–1.79	1.18	0.92–1.50	2.01	1.41–2.88	1.60	1.06–2.42
Having partners using MA
No	1.00		1.00		1.00		1.00
Yes	1.58	1.32–1.90	1.19	0.97–1.45	1.13	0.85–1.50	1.04	0.74–1.45
Age	1.00	0.99–1.01	1.00	0.99–1.01	0.97	0.96–0.99	0.97	0.95–0.99
Age at MA initiation	0.98	0.97–0.99	0.99	0.98–1.02	0.98	0.96–0.99	1.01	0.99–1.03
Frequency of MA use	1.02	1.01–1.03	1.01	1.00–1.02	1.03	1.01–1.04	1.01	0.99–1.02
DAST-20	1.07	1.04–1.10	1.03	0.99–1.06	1.10	1.07–1.14	1.07	1.03–1.12
AUDIT	1.00	0.99–1.02	1.00	0.99–1.02	1.02	0.99–1.04	1.01	0.99–1.03

AOR: Adjusted odds ratio, CI: Confidence interval, DAST: Drug Abuse Screening Test, AUDIT: Alcohol Use Disorders Identification Test.

^aHigh-risk sexual behavior was measured using four indicators: 1) inconsistent condom use (ever), 2) multiple sexual partners (the preceding year), 3) STD history (ever), and 4) trading sex for drugs (the preceding year).

^bAdjusted for age (at the time of survey), marital status (before incarceration), employment status (before incarceration), recidivism, MA use during sex (ever), sharing MA injection (ever), having partners that used MA (before incarceration), age at MA initiation, frequency of MA use (the preceding year), DAST-20 score (the preceding year), and AUDIT score (the preceding year).

Discussion

The present study evaluated the relationship between MA use and high-risk sexual behaviors among Japanese prisoners. Our study provided three major findings. First, we observed gender differences with regards to reasons for MA use. Male participants were more likely to answer “to get sexual pleasure and excitement” than females, whereas female participants were more likely to answer “to escape reality” and “to lose weight” than males. These findings suggest that male MA users mainly use the drug as “sex drugs,” whereas females use MA for coping with negative emotions and for weight control. Thus, a treatment program tailored to deal with specific mental struggles such as stress and body image-related anxiety could be effective for female MA users. Moreover, a treatment program to understand the association between MA use and high-risk sexual behavior could be effective among male users. Our results also showed gender differences in length and frequency of MA use. Female MA users had a shorter length of MA use than males, but had a higher frequency of MA use. This suggests that female MA users were using MA more intensively in a shorter period of time. In addition, female MA users had a significantly higher DAST-20 score than males. These results indicate that female MA users consume MA more intensively, and have more severe MA use-related problems and complications than male MA users.

The second major finding was that a large proportion of MA user prison inmates engaged in high-risk sexual behaviors. Inconsistent condom use was frequently observed in both genders, which implies they are at a high risk of contracting infectious diseases such as HIV infections. Regarding gender differences, male MA users more likely had multiple sexual partners than females. This result can be explained by the fact that male inmates mainly use MA to get sexual pleasure. However, female MA users more likely had a history of a STD than males. Prenatal screening for female users might increase opportunity to diagnose sexually transmitted diseases such as chlamydia. A previous study, conducted in China, reported that more male MA users have had multiple sex partners than female MA users (Liu et al. 2013). However, a study in California reported that female MA users were more likely to have been diagnosed with an STD in the last 2 months (Cheng et al. 2009). The gender differences in our current study were consistent with those observed in these previous studies in other countries.

Lastly, we found gender similarities and differences associated with high-risk sexual behaviors and MA use. The GEE logistic regression analysis revealed that MA use during sex increased the likelihood of high-risk sexual behavior in both male and female prisoners. The association between MA use during sex and high-risk sexual behavior may be influenced by the increase in brain dopamine levels due to MA use. Animal studies have shown that MA increases brain dopamine levels and that increased brain dopamine levels may increase sexual desire or libido (Rudzinskas et al. 2019; Volkow et al. 2007). In human studies, MA use has been reported to increase physical sexual desire and psychological pleasure. Increased sexual desire is associated with high-risk sexual behaviors such as unprotected sex (Liu and Chai 2020). Therefore, the increase in sexual desire caused by MA use may be one of the mechanisms by which MA use promotes high-risk sexual behavior.

In female prisoners, the history of sharing MA injection also increased the likelihood of high-risk sexual behavior. This supports our hypothesis of gender differences in relationships between MA use and high-risk sexual behavior among prisoners. There are several possible explanations for this finding. One explanation is that female MA users share injections because of difficulties to access a clean needle and syringe. In Japan, syringes are regulated as controlled medical devices (class II) by the Medical Devices Act. Thus, they cannot be purchased by individuals and are sold only to medical institutions, veterinary hospitals, and research institutions. Moreover, harm reduction programs such as syringe exchange programs are not provided in Japan. A second explanation is the possibility of female users to share MA injections with a male partner having high-risk sexual behaviors. According to our results, female MA users are more likely to have partners who use MA than male users. At the same time, female users had lower rates of having multiple sexual partners than male users. Although our study did not identify who MA injections are being shared with, these results suggest that female MA users may share syringes with male partners who have high-risk sexual behaviors. Previous studies have reported that female MA users were more likely to be introduced to MA use by a spouse (Brecht et al. 2004) or sex partners (Cheng et al. 2009) than male MA users. Interestingly, in a Chinese study, all female MA users answered that they were “always using MA with heterosexual male partners” (Liu et al. 2013). These findings suggest that male partners who use MA strongly influence female users. However, in our study, “having partners using MA” did not increase high-risk sexual behavior. Hence, “having partners using MA” by itself may not predict high-risk sexual behaviors, but using MA with that partner and sharing injections may increase high-risk sexual behaviors.

Limitations and strengths

Study limitations include the different periods defined for variables of high-risk sexual behaviors. For example, inconsistent condom use and STD history referred to lifetime experience, whereas other indicators referred to the past year. This is because our study was a secondary analysis. A potential problem related to the lifetime experience measurement is the overestimation of high-risk sexual behaviors. However, even in previous studies that evaluated high-risk sexual behaviors using multiple indicators, the period of measurements was not always the same (Browne et al. 2009; Saw et al. 2018).

Second, the cross-sectional study design does not permit causal inference. Also related to the use of cross-sectional data is the presence of confounding factors. In this study, we observed a significant relationship between MA use during sexual intercourse and high-risk sexual behaviors in both genders. However, there may be potential confounding factors, such as the participant’s personality, between the two variables. Previous studies have indicated that personality dispositions, particularly “sensation seeking,” are potentially important confounding factors influencing alcohol-induced sexual behavior associations (Kalichman and Cain 2004; Kalichman et al. 2003). Additionally, it is necessary to mention the endogeneity issue in the analytical model. In the analysis of the cross-sectional data using the GEE model, the four outcome indicators were treated as if they were measured repeatedly. For this reason, it is possible that the dependent variable adversely affects the independent variable, causing the overestimation of causal relationships.

The third limitation is the reliance on self-report, as self-reporting in a prison environment may result in reporting bias. However, previous research on social desirability reports that use of self-report among offenders could predict recidivism after being released (Mills, Loza, and Kroner 2003). Therefore, self-reports may have some predictive validity. Last, participant gender was assessed by the gender registered in the family register. Information on sexual orientation and gender identity was not collected, and sex-related risk behavior may differ between men who have sex with men (MSM) and men who have sex with women.

At the same time, our study was based on nationwide data, and participants were recruited from all prisons throughout Japan. Previous studies on sex-related risk behavior in MA-using offenders had targeted only some states in the USA (Cartier, Greenwell, and Prendergast 2008; Farabee, Prendergast, and Cartier 2002). We also used four indicators of sex-related risk behavior, including “trading sex for drugs.” Last, we identified subgroups of MA-using prisoners who were at higher risk, such as individuals who use MA during sex. Previous studies (Cartier, Greenwell, and Prendergast 2008; Farabee, Prendergast, and Cartier 2002) compared MA users with non-MA users, but it was unclear which subgroup of MA users may be at higher risk.

In conclusion, we believe that our study results can be used for drug policy formulation and program development in Japan. Our results revealed that MA use during sex increased high-risk sexual behavior in both genders. Thus far, harm reduction aimed at reducing secondary harm – including high-risk sexual behaviors associated with drug use – has not been incorporated into drug policies in Japan. Our results imply that treating MA users in prison and clinical settings require careful assessment of not only MA use but also high-risk sexual behavior. Moreover, harm reduction programs tailored to different gender aspects targeted toward reducing MA use-related harms, including high-risk sexual behaviors, need to be included in Japanese drug policies.

Acknowledgments

This work was supported by the Ministry of Health, Labor, and Welfare of Japan: Research grant for the Division of Research, National Center for Addiction Services Administration (2019–08).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

This study is a secondary analysis of the original project of the Ministry of Justice, Japan. The data set is not publicly available due to the contract with the Ministry of Justice, Japan.