Correlates of Current Methamphetamine Use and Opioid Co-Use Among Latina Women in a Low-Income Community

Abstract

Using data from Proyecto SALTO, a 15-year follow-up study of a cohort of Mexican American women in a low-income community in San Antonio, Texas, this study examines emerging patterns of current methamphetamine (MA) use, including opioid co-use, among this understudied population. A bivariate analysis compared individuals with and without current MA use and identified sociodemographic correlates and co-occurring mental health and substance use. A secondary analysis compared those with current MA use, opioid use, and concurrent MA and opioid use. Nineteen percent of the sample had current MA use. MA use was associated with having a lower income (OR = 7.04–1.93, SE = 1.59–5.46), residential instability (OR = 5.19, SE = 1.99), and suicidal ideation (OR = 2.62, SE = 0.93). Participants with MA use had more than four times the odds of using opioids than those without MA use. Women with concurrent MA and opioid use differed in sociodemographics and behavioral risks compared to those with only MA or only opioid use. These findings explore the social, mental health, and structural inequities that exacerbate risks and harms associated with high-risk substance use among marginalized Latino populations. Prevention and intervention strategies should adopt a holistic approach that considers and addresses polysubstance use, mental health, and the sociocultural contexts in which individuals live.

While much attention has focused on the opioid epidemic in the United States, rates of stimulant use, primarily methamphetamine (MA), have drastically increased in recent years, characterizing the fourth wave of the drug overdose crisis (Ciccarone 2021). Rates of MA-related deaths alone increased more than 450% in both men and women between 2011 and 2018 (Han et al. 2021), and MA use is associated with increased drug-related morbidity, especially when injected or when used concurrently with opioids (Glick et al. 2021). Generally, MA use has been linked to a variety of social and structural factors, including other substance use disorders, financial and life stressors, drug stigma, drug markets and community contexts, and barriers to drug treatment (Boden et al. 2023; Cumming et al. 2016; Lopez et al. 2021). Considering the risks associated with MA use and with opioid co-use (Al-Tayyib et al. 2017; Cano and Huang 2021; Glick et al. 2021; Novack et al. 2020), more research is needed to better identify and understand risks factors for MA use among vulnerable populations.

Recent changes in the drug markets leading to an increasingly adulterated drug supply and psychostimulant accessibility is having a disproportionate impact on Black and Hispanic communities (Hoopsick, Homish, and Leonard 2021; Larochelle et al. 2021). While MA-related deaths have increased across racial and ethnic groups, MA-related deaths have more than tripled in Latinos ages 25 to 54 years (Han et al. 2021). In recent studies of MA patterns, Latinos reported the second highest rates of past 12-month MA use (6.7 per 1,000) after Whites (Jones, Compton, and Mustaquim 2020), and the second highest past 12-month MA-related treatment admissions, after Native Americans/Alaskan Natives (Jones, Olsen, et al. 2020). Despite these rates and limited access to drug treatment and harm reduction services among Latinos (Dayton et al. 2020; Friedman and Hansen 2022; Goedel et al. 2020), the use of MA has been largely underexamined in this population (Handley and Sudhinaraset 2017; Rawson et al. 2005). Of the limited research of MA among U.S. Latino populations, the majority of studies have examined Latino men who have sex with men and examine MA use in the context of sexual risk behaviors (Díaz, Heckert, and Sánchez 2005; Loza et al. 2020; Wohl, Frye, and Johnson 2008; Young and Shoptaw 2013). While research indicates that women have lower rates of MA use compared to men overall (Jones, Compton, and Mustaquim 2020), some studies indicate that women may have more-severe patterns of use in different contexts compared with men (Hernández et al. 2009; Simpson et al. 2016).

Few recent studies have specifically examined MA use among Latina women (Cheney et al. 2018; Hernández et al. 2009). Existing research on MA use among women in the United States has historically focused on child and pregnancy outcomes associated with perinatal MA use (Gorman et al. 2014; Perez et al. 2022; Smid, Metz, and Gordon 2019; Terplan et al. 2009) and associations of MA with sexual risk behaviors (Johnson et al. 2016; Lorvick et al. 2012; Semple, Grant, and Patterson 2005). Further, Latina women are often not represented in MA studies among women (Copes et al. 2016; Johnson et al. 2016). Despite this lack of research, Latina women make up 46% of all MA-related treatment admissions among Latinos, the highest proportion of women among any racial/ethnic group (Guerrero et al. 2023). A qualitative study among Latinas in residential substance use treatment in Los Angeles found that living in low-income and socially marginalized communities heightens women’s vulnerability to physical abuse and emotional distress, which are in turn associated with initiation and continuation of MA use (Cheney et al. 2018). Other low-income and marginalized Mexican American communities, including that of the present study in San Antonio, Texas, have had documented patterns of inter-generational heroin use fueled by readily accessible opioid markets (Bullington 1977; Casavantes 1976; Cepeda, Nowotny, and Valdez 2016; Garcia 2010; Maddux and Desmond 1981; Moore et al. 1978; Valdez and Cepeda 2008; Valdez, Kaplan, and Cepeda 2000; Valdez, Mikow, and Cepeda 2006). Given the historically low rates of MA use among Latinos, research is needed to better understand the emergence of MA-use patterns in communities where opioid use has been a primary drug of choice.

The present study aims to examine profiles of Latina women who currently use MA. Among a cohort of women living in a low-income community in San Antonio, TX, we assess how sociodemographics, co-occurring substance use, mental health, and other behavioral risk factors are associated with MA use. To further understand polysubstance use patterns, we also conducted a secondary analysis to disentangle the profiles of women who concurrently use both MA and opioids. In this community, in which MA has historically not been widespread, understanding risk profiles for MA use can inform interventions and prevention efforts to more effectively address drug-related harms.

Materials and methods

The San Antonio Latina Trajectory Outcome (Proyecto SALTO) study is a 15–20 year follow-up study conducted between 2015 and 2020 examining the long-term health and drug use among a cohort of U.S. born, adult Latina women (ages 32 to 40 years). The sample includes 97 of 150 women who were originally recruited and interviewed as adolescents between 1999 and 2002. Of the original 150 women, 92% were successfully relocated and 70% of those relocated were enrolled in Proyecto SALTO and provided informed consent. During Proyecto SALTO, an additional 128 women were recruited from the original sampling frame for a total sample size of 225. Sampling and recruitment protocols of the original study can be found in previous publications (Petersen and Valdez 2005; Valdez 2007; Yin et al. 2000). Eligibility criteria for Proyecto SALTO and the original study included being a Mexican American female, being between ages 14 and 18 at the time of the original study (1999–2002), and being associated (brother, boyfriend, close friend) with a male member of one of 27 known male street gangs from the study’s catchment area. San Antonio is in the top 10 cities in the United States with the largest number of people living in socioeconomically distressed zip codes (Economic Innovation Group 2016). The catchment area of this study included the San Antonio’s Westside, which includes the most distressed zip codes in the city.

The Institutional Review Board at the University of Southern California approved all study protocols, and all participants provided written informed consent. During Proyecto SALTO, consenting participants completed private interviews with a trained interviewer consisting of standardized questionnaires and life history calendars. Participants were asked to complete drug metabolite urinalysis, of which 202 participants completed. Only one participant did not complete either the drug urinalysis or the self-reported drug-use questionnaire and was thus excluded from the final analytical sample (n = 224).

Measures

Sociodemographics.

Sociodemographic data considered in this analysis included age, sexual orientation (heterosexual, gay/lesbian, or bisexual), educational attainment (eighth grade or below, less than a high school degree or GED, high school degree or GED, some college or more), current full- or part-time employment, income in the last 30 days (less than $1,000, $1,000–$1,999, $2,000–$3,999, $4,000 or more), marital status (single, never married; married or living together; separated/divorced/widowed), number of children, and current health insurance status (insured, uninsured). It should also be noted that the sexual orientation question was originally not included in the survey; it was added later after data monitoring identified the error. Thus, only 143 women completed the sexual orientation question and were included in the analysis of this variable.

Current drug use and concurrent use.

Current drug use was measured by both self-report and drug urinalysis. Participants self-reported past-30-day nonmedical use of MA, opioids (prescription, heroin, synthetic), cannabis, cocaine, and sedatives (barbiturates, benzodiazepines). Drug metabolites were assessed using the iCup AD 8-panel urine test from Alere Toxicology (Richmond, VA) and detected cocaine, marijuana/THC, opiates, amphetamines, MA, PCP, benzodiazepines, and barbiturates. Variables for MA, opioids, cannabis, cocaine, and sedatives were coded dichotomously, with current use identified with a positive urinalysis result or self-reported past-30-day use. Both Medication for Opioid Use Disorder (MOUD; such as methadone or buprenorphine) and prescription drugs used as prescribed for medical purposes were reported separately from illicit drug use or prescription misuse. Neither self-report MOUD or appropriate medical prescription drug use were included in the final drug variables. No participants self-reported medical use of prescription opioids and tested positive for opioid metabolites in the drug urinalysis. Lifetime drug dependence for MA was measured using the five-item Severity of Dependence Scale (SDS) (Gossop et al. 1995). The scale was summed and a cutoff score of four or more was used to indicate MA dependence (Topp and Mattick 1997).

We also assessed individual and concurrent MA and opioid use. Concurrent use was indicated if a participant self-reported past-30-day use of or tested positive in urinalysis for both MA and opioids. Concurrent use indicates use of both substances in the same time period (e.g., past 30 days) but does not capture simultaneous use (use of both substances together—e.g., goofball) or unintentional exposure (e.g., fentanyl contamination of MA). Two other dichotomous variables were created indicating whether individuals had current MA use without opioid use (“MA only”) and current opioid use without MA use (“opioid only”). A binary variable of history of an opioid overdose indicated whether a participant had ever experienced an overdose while using opioids in their lifetime. This variable was specific to an opioid overdose and was thus only included in the analysis examining concurrent use.

Mental health.

Mental health indicators included depression, psychological distress, posttraumatic stress disorder (PTSD), and suicidal ideation. Depression symptoms were measured using the eight-item Center for Epidemiologic Studies–Depression Scale (CES-D) (Melchior et al. 1993). A binary variable for depression was created using the recommended cutoff of 16 or higher (Radloff 1977). Psychological distress and suicidal ideation were measured from the 28-item General Health Questionnaire (GHQ-28) (Goldberg and Hillier 1979). Psychological distress was indicated by a cutoff of 24 or higher. Suicidal ideation was indicated by a positive response to any of following four items from the GHQ-28: “found that the idea of taking your own life kept coming into your mind;” “found yourself wishing you were dead and away from it all;” “thought of the possibility that you might make away with yourself;” and “felt that life isn’t worth living.” Lastly, PTSD symptoms were measured using the PTSD Checklist-Civilian Version (PCL-C), which assesses symptoms using the Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) criteria (Conybeare et al. 2012). The PCL-C is scored using a continuous, summed severity total, with higher scores indicating greater severity of PTSD symptoms.

Other risks.

Participants reported the number of places they lived in the last year, and a binary variable of residential instability was created to indicate living in three or more places. Participants reported the number of incarceration episodes lasting 30 days or longer over the lifetime and whether they had at least one felony conviction record. Lastly, participants reported the number of sexual partners (male or female) they had in the last year.

Analysis

Bivariate tests of associations were used to examine study variables among those with and without current MA use. Independent t-tests or one-way ANOVAs for continuous variables and chi square tests or Fisher’s exact tests for categorical variables compared study variables by current MA status. Fisher’s exact tests were used for categorical variables when greater than 20% of cells had a frequency of less than five (Kim 2017). Binary logistic regression was used to calculate odds ratios (OR). These bivariate tests of association and pairwise comparisons were then used in a secondary analysis to assess study variables among those with only MA use, only opioid use, and concurrent MA and opioid use.

Results

The mean age of the sample was 33.3 (SD = 2.3) with the majority of participants identifying as heterosexual (81.1%) (Table 1). Appoximately 60% reported having attained at least a high school diploma or GED and nearly 60% reported being employed either full- or part-time at the time of the interview. Over 30% reported an income of less than $1,000 in the past 30 days, and almost 20% reported having experienced residential instability in the past year.

Table 1.

Characteristics of Individuals with Tested or Self-reported Past 30-day MA Use Compared to Those Without Current Use (n=224)^a

	Total Sample (n=224)		Current MA Use (n=42, 18.8%)		No Current MA Use (n=182, 81.3%)
	n	%/ Mean (SD)	n	%/ Mean (SD)	n	%/ Mean (SD)	t/x2 p	OR (SE)	p
Demographics
Age	224	33.3 (2.3)	42	33.2 (2.6)	182	33.3 (2.3)	0.4	0.97 (0.08)
Sexual Orientation (n=143)							7.3**
Heterosexual/Straight	116	81.1	14	60.9	102	85.0		1.0 (REF)
Lesbian/Gay or Bisexual	27	18.8	9	39.1	18	15.0		3.60 (1.81)	**
Educationb							7.6*
8th Grade or Less	19	8.5	4	9.5	15	8.2		4.36 (3.56)	+
Less than High School Degree	68	30.4	16	38.1	52	28.6		5.03 (3.32)	*
High School Degree or GED	85	38.0	19	45.2	66	36.3		4.70 (3.05)	*
Some College or More	52	23.2	3	7.1	49	26.9		1.0 (REF)
Current Employment (n=221)	132	59.7	15	35.7	117	65.4	12.4***	0.29 (0.11)	**
Income in Last 30 Days (n=223)							16.5**
Less than $1,000	70	31.4	24	57.1	46	25.4		7.04 (5.46)	*
$1,000 - $1,999	60	26.9	8	19.1	52	28.7		2.08 (1.71)
$2,000 - $3,999	64	28.7	8	19.1	56	30.9		1.93 (1.59)
$4,000 or more	29	13.0	2	4.8	27	14.9		1.0 (REF)
Marital Status (n=222)							3.5
Single, never married	82	36.9	20	47.6	62	34.4		1.0 (REF)
Married or Living Together	102	46.0	14	33.3	88	48.9		0.49 (0.19)	+
Separated, Divorced, or Widowed	38	17.1	8	19.1	30	16.7		0.83 (0.39)
Number of Kids	224	3.3 (1.8)	42	3.3 (1.9)	182	3.3 (1.7)	−0.1	1.01 (0.10)
Current Health Insurance (n=219)	118	53.9	11	27.5	107	59.8	13.7***	0.26 (0.10)	***
Other Drugs (Tested or self-reported)
Opioids	57	25.5	22	52.4	35	19.2	19.8***	4.62 (1.67)	***
Marijuana	67	29.9	16	38.1	51	28.0	1.7	1.58 (0.57)
Cocaine or Crack	41	18.3	10	23.8	31	17.0	1.0	1.52 (0.63)
Sedatives	39	17.4	10	23.8	29	15.9	1.5	1.65 (0.68)
Injection Drug Use (Last Year) (n=219)	28	12.8	13	33.3	15	8.3	18.0***	5.50 (2.39)	***
Mental Health
Depression (n=215)	45	20.9	11	28.2	34	19.3	1.5	1.64 (0.66)
Psychological Distress (n=209)	131	62.7	24	66.7	107	61.9	0.3	1.23 (0.48)
PTSD Severity Score (n=205)	205	25.0 (19.0)	38	30.2 (21.2)	167	23.8 (18.3)	−1.9+	1.02 (0.01)	+
Suicidal Ideation (n=219)	79	36.1	22	55.0	57	31.8	7.6**	2.62 (0.93)	**
Other Risks
Residential Instability (Last Year)	41	18.3	18	42.9	23	12.6	20.8***	5.18 (1.99)	***
# of Incarcerations >30 days (n=222)	222	1.2 (2.2)	40	1.9 (2.8)	182	1.0 (2.0)	−2.5*	1.17 (0.08)	*
Any Felony Conviction (n=222)	75	33.8	20	50.0	55	30.2	5.7*	2.31 (0.82)	*
# of Sexual Partners (Last Year) (n=218)	218	1.7 (3.9)	40	2.6 (6.3)	178	1.5 (3.0)	−1.7+	1.06 (0.04)

⁺p<.10;

^*p<.05;

^**p<.01;

^***p<.001;

^aAs indicated by the variable name, some analyses included a smaller sample size due to missing data;

^bDue to >20% of cells having frequencies <5, Fisher’s exact test was used and the corresponding p-value is reported.

Current MA use

In total, 32% reported lifetime MA use and 11% met criteria for lifetime MA dependence. Forty-two women (18.8%) were considered to have current MA use. Thirteen discrepancies were apparent, in which participants did not self-report MA use in the past 30 days but the presence of MA was detected in urinalysis. Of these 13 MA discrepancies, six had opioid co-use.

Significant differences in sociodemographics were noted among the women who had and did not have current MA use (Table 1). Current MA use was positively associated with identifying as lesbian/gay or bisexual (OR = 3.6); having a high school degree/GED versus some college or more (OR = 4.70); and making less than $1,000 in the last 30 days compared to $4,000 or more (OR = 7.04). Current MA use was negatively associated with being currently employed (OR = 0.29); currently having health insurance (OR = 0.26); and being married or living with a partner versus single (OR = 0.49).

Other current drug use did not differ among those who had and did not have current MA use with the notable exception of opioids. Those who had current MA use had 4.6 times the odds of current opioid use compared to those without current MA use (52.4% vs. 19.2%). Similarly, those with MA use were 5.5 times as likely to report injection drug use in the last year (33.3% vs. 8.3%). Although few mental health symptoms differed, those with current MA use were 2.6 times as likely to report experiencing suicidal ideation (55.0% vs. 31.8%) compared with participants with no current MA use. Moreover, residential instability was reported by 42.9% of women who had current MA use compared with 12.6% who had no current MA use (OR = 5.18). Those with current MA use had an increased number of lifetime incarceration episodes longer than 30 days (mean 1.9 vs. 1.0, OR = 1.17) and a prevalence of ever having had a felony conviction (50.0% vs. 30.2%, OR = 2.31). Lastly, the average number of sexual partners in the last year marginally differed by current MA status (mean 2.6 vs. 1.5).

MA use, opioid use, and concurrent use

Among those with any current MA use, 48% had only MA use, while 52% (22 women) had concurrent MA and opioid use. An additional 16% of the sample had only current opioid use. Those with concurrent MA and opioid use had significantly lower rates of current employment compared with those with only MA (18.2% vs. 55%) or only opioid (18.2% vs. 48.6%) use (Table 2). Relatedly, individuals with only opioid use had a higher prevalence of current health insurance compared with those with only MA (60.0% vs. 31.6%) and those with concurrent use (60.0% vs. 23.8%). No other sociodemographics significantly differed between groups.

Table 2.

Characteristics of Individuals with Current MA Use, Opioid Use, and Concurrent MA and Opioid Use (n=224)^a

	No MA or Opioid Use (n=147, 65.6%)		Only MA Use (n=20, 8.9%)		Only Opioid Use (n=35, 15.6%)		Concurrent Use (n=22, 9.8%)			Only MA v. only opioid	Only MA v. concurrent	Only opioids v. concurrent
	n	%/ Mean (SD)	n	%/ Mean (SD)	n	%/ Mean (SD)	n	%/ Mean (SD)	F/x2 p	p	p	p
Demographics
Age	147	33.3 (2.3)	20	33.4 (3.0)	35	33.5 (2.1)	22	33.0 (2.3)	0.2
Sexual Orientation (n=143)b									10.5*
Heterosexual/Straight	85	86.7	3	42.9	17	77.3	11	68.8
Lesbian/Gay or Bisexual	13	13.3	4	57.1	5	23.7	5	31.3
Educationb									13.7+
8th Grade or Less	12	8.2	1	5.0	3	8.8	3	13.6
Less than High School Degree	37	25.2	9	45.0	15	42.9	7	31.8
High School Degree or GED	57	38.8	9	45.0	9	25.7	10	45.5
Some College or More	41	27.9	1	5.0	8	22.9	2	9.1
Current Employment (n=221)b	100	69.4	11	55.0	17	48.6	4	18.2	23.4***		*	*
Income in Last 30 Days (n=223)b									28.6**
Less than $1,000	30	20.6	10	50.0	16	45.7	14	63.6
$1,000 - $1,999	43	29.5	3	15.0	9	25.7	5	22.7
$2,000 - $3,999	48	32.9	6	30.0	8	22.9	2	9.1
$4,000 or more	25	17.1	1	5.0	2	5.7	1	4.6
Marital Status (n=222)									8.8
Single, never married	46	31.3	9	45.0	16	48.5	11	50.0
Married or Living Together	77	52.4	6	30.0	11	33.3	8	36.4
Separated, Divorced, or Widowed	24	16.3	5	25.0	6	18.2	3	13.6
Number of Kids	147	3.3 (1.7)	20	3.3 (1.9)	35	3.2 (2.0)	22	3.3 (1.9)	0.1
Current Health Insurance (n=219)	86	59.7	6	31.6	21	60.0	5	23.8	13.9**	*		**
Other Drugs (tested or self-reported)
Marijuana	33	22.5	8	40.0	18	51.4	8	36.4	13.0**
Cocaine or Crack	20	13.6	5	25.0	11	31.4	5	22.7	7.1+
Sedativeb	18	12.2	3	15.0	11	31.4	7	31.8	10.8*
Injection Drug Use (Last Year) (n=219)b	3	2.1	2	11.1	12	34.3	11	52.4	59.0***	+	**
History of Opioid Overdose (Ever)b	14	9.7	1	5.3	8	22.9	8	38.1	15.8**	+	*
Mental Health
Depression (n=215)b	28	19.9	3	16.7	6	17.1	8	38.1	4.3			+
Psychological Distress (n=209)	80	57.6	10	58.8	27	79.4	14	73.7	6.7+
PTSD Severity Score (n=205)	135	23.1 (18.9)	18	27.4 (22.0)	32	26.7 (15.8)	20	32.8 (20.6)	1.4
Suicidal Ideation (n=219)	41	28.5	8	42.1	16	45.7	14	66.7	13.8**
Other Risks
Residential Instability (Last Year)	18	12.2	10	50.0	5	14.3	8	36.4	22.2***	**		+
# of Incarcerations >30 days (n=222)	147	0.6 (1.5)	19	0.6 (1.1)	35	2.5 (2.9)	21	3.1 (3.4)	13.7***	*	**
Any Felony Conviction (n=222)	37	25.2	5	26.3	18	51.4	15	71.4	23.5***	+	**
# of Sexual Partners (Last Year) (n=218)	143	1.3 (0.8)	19	1.6 (0.9)	35	2.3 (6.6)	21	3.5 (8.7)	2.3+

⁺p<.10;

^*p<.05;

^**p<.01;

^***p<.001;

^aAs indicated by the variable name, some analyses included a smaller sample size due to missing data;

^bDue to >20% of cells having frequencies <5, Fisher’s exact test was used and the corresponding p-value is reported.

While current marijuana, cocaine, and sedatives use did not differ in pairwise comparisons among these three groups, injection drug use in the last year was significantly higher among participants with only opioid use (34.3% vs. 11.1%) and concurrent use (52.4% vs. 11.1%) compared with those with only MA use. As expected, lifetime history of opioid overdose was higher among those with only opioid (22.9% vs. 5.3%) and with concurrent use (38.1% vs. 22.9%) compared with those with only MA use, but did not significantly differ between those with only opioid and concurrent use. While most of the mental health measures fell within a similar range across the sample, all three groups had higher suicidal ideation than those who used neither drug (28.5% vs. 42.1%–66.7%). Those with concurrent use had the highest prevalence of suicidal ideation (66.7%). Residential instability in the last year was also significantly more prevalent among those with only MA (50.0%) and with concurrent use (36.4%) than among those with only opioid use (14.3%). Lastly, compared with those with only MA use, both those with only opioid and those with concurrent use had a significantly greater average number of lifetime incarceration episodes lasting at least 30 days (mean 0.6 vs. 2.5 vs. 3.1) and a greater likelihood of at least one felony conviction (26.3% vs. 51.4% vs. 71.4%). Those with concurrent use also had a greater number of incarceration episodes on average and a higher prevalence of a felony conviction than those with only opioid use, although neither number of incarceration episodes nor prevalence of felony conviction were statistically significant.

Discussion

In the Latino community in which our study sample was recruited, intergenerational heroin use has been widely documented (Cepeda, Nowotny, and Valdez 2016; Valdez 2007; Valdez and Cepeda 2008; Valdez, Kaplan, and Cepeda 2000; Valdez, Mikow, and Cepeda 2006) and emerging patterns of MA use require greater attention. This study contributes to this initial understanding, finding that both MA and concurrent MA and opioid use are prevalent in our sample of Latina women. Similar to other research with nationally representative samples (Jones, Olsen, et al. 2020; Shearer et al. 2020), women in our sample reporting current MA use had generally lower socioeconomic statuses (e.g., income, education, employment, residential stability) than those without MA use and those with only opioid use. This was often consistent among those with concurrent MA and opioid use; for example, concurrent MA and opioid use, but not only MA or only opioid use, was associated with lower employment rates among women. These findings highlight distinct social and economic inequities that may contribute to differing MA and opioid use patterns and behaviors and should be considered in designing prevention interventions.

Differences among those with MA and opioid use may speak to the differing degrees of stigma or pervasiveness of each drug in the community. For instance, opioid and heroin use has been present in long-established intergenerational social and family networks within this community (Cepeda, Nowotny, and Valdez 2016; Valdez, Kaplan, and Cepeda 2000; Valdez, Mikow, and Cepeda 2006). Thus, it is possible that those with opioid use may have had longer periods of use, more long-term use among peers and family members, and more self-regulated and harm reduction behaviors (Cepeda, Nowotny, and Valdez 2016). As our findings regarding employment, residential stability, and income show, women with opioid use in this community may also be more likely to maintain financial independence (Valdez, Kaplan, and Cepeda 2000) and receive social, familial, and economic support despite their drug use and increased incarceration histories.

Comparatively, the emerging use of MA may be more stigmatized and used in differing social contexts compared with opioids (Baker et al. 2021; Cumming et al. 2016; Lopez et al. 2021). Some research has found that MA is used to maintain alertness or to stay awake in work or unsafe environments, such as among homeless populations or individuals who are involved in sex work (Compton et al. 2018; Schmidt et al. 2019; Von Mayrhauser, Brecht, and Anglin 2001). Other qualitative research has found that people who use opioids sometimes engage in MA use to manage opioid-related risks (Baker et al. 2021; Lopez et al. 2021; Silverstein et al. 2021). Likely compounded for minoritized groups (Friedman and Hansen 2022), people who use MA may face additional treatment barriers due to stigma from providers and social networks related to perceived challenges in addressing the behavioral and mental health problems associated with MA (Dunn et al. 2023). Disparities in treatment access and approaches for MA and opioid use also have particular consequences for the women in our study who have historically had limited access to any culturally responsive, effective drug and mental health treatment (Guerrero et al. 2013; Stahler, Mennis, and Baron 2021). Additional research is needed to fully understand the social contexts in which MA is used individually and concurrently with opioids among Latina women and other diverse populations.

Lastly, our study found that women who use MA were more than 2.5 times as likely to report suicidal ideation as women without MA use. While other mental health indicators did not substantially differ, suicidal ideation is especially concerning given this population’s limited access to health insurance (only 54% of our sample was insured) and mental health services (Buchmueller and Levy 2020; Cepeda et al. 2018; Green et al. 2020; Mongelli, Georgakopoulos, and Pato 2020). While much focus is on drug use patterns and overdose risk, urgent attention is needed regarding the co-occurring mental health conditions of individuals who use drugs. Suicidal ideation screenings and prevention efforts could be beneficial in any system in which these individuals have frequent or regular contact, including housing programs, primary health care and drug treatment, and criminal justice systems.

Limitations

Findings must be considered within the context of study limitations. First, our sample size limited our analysis to bivariate and pairwise comparisons and, due to the cross-sectional design, temporal trends in MA initiation and polysubstance could not be examined. Second, we measured the current and concurrent use of MA and opioids, but we did not specifically measure simultaneous co-use of drugs (e.g., goofball). Differences in the risks and consequences of concurrent versus simultaneous use require further investigation. Moreover, this measurement may also miss unintentional exposures, such as fentanyl contamination, to MA or other drugs. Third, survey data were self-reported and subject to recall bias and social desirability bias. As there were discrepancies between the self-reported use of drugs and urinalysis, we used both measures to help minimize this bias in all drug outcome variables. Finally, the first few years of this study predated the widespread adulteration of fentanyl in other substances in this community and, as such, the instrument did not include questions related to fentanyl use or fentanyl contamination of the drug supply specifically. Despite these limitations, our study is strengthened by the relocation and recruitment of a sample of particularly vulnerable, low-income Latina women who are not often represented in substance use research.

Conclusions

With an evolving overdose crisis that is now disproportionately impacting racial and ethnic minoritized people who use drugs, our findings highlight the patterns and social contexts of MA use and concurrent MA and opioid use among low-income Latina women. Successful prevention and intervention strategies should adopt a holistic approach that addresses not only individual MA use outcomes but also the socioeconomic and mental health contexts these individuals experience. Further considerations are needed of the social, structural, and mental health inequities that exacerbate the risks and harms associated with these drug patterns.

Data availability statement

Data used in this manuscript are available from the senior author upon request.