A Longitudinal Analysis: The Moderating and Mediating Effect of Alcohol Outcome Expectancies on Traditional Masculine Gender Norms and Alcohol Use among Latino Immigrant Men

Abstract

Latino immigrants in the United States are at risk for problematic drinking behaviors, with Latino men engaging in higher rates of drinking than their female counterparts. Guided by Gender Schema Theory and the Alcohol Expectancy Model, the present study examined the longitudinal associations between traditional masculine gender norms (i.e., traditional machismo and caballerismo), alcohol outcome expectancies (AOEs), and alcohol use in a sample of 276 Latino immigrant men. We tested competing linear mixed effects models to examine the hypothesized moderation and mediation pathways. Alcohol use significantly decreased by the third wave compared to baseline. Greater endorsement of traditional machismo was associated with higher positive AOEs, which in turn were associated with greater alcohol use. Positive AOEs moderated the relationship between traditional machismo and alcohol use such that stronger positive AOEs attenuated the negative association between traditional machismo and alcohol use. Negative AOEs also moderated this relationship: higher negative AOEs amplified the inverse association between traditional machismo and alcohol use. Traditional machismo was inversely associated with alcohol use, whereas endorsement of positive AOEs was positively associated with alcohol use. Higher education was associated with lower alcohol use. Implications for effective prevention and intervention strategies are discussed, along with recommendations for future research.

Latinos represent the largest ethnoracial minority group in the United States (U.S.), accounting for 19.5% of the total U.S. population (U.S. Census Bureau, 2024). Among Latino adults, approximately 80% report lifetime alcohol use, 63% report past-year use, and 46% report past-month use (Substance Abuse and Mental Health Services Administration [SAMHSA], 2023a, 2023b, 2023c). Latino men are at increased risk for alcohol use and alcohol-related problems relative to Latinas and to individuals from other ethnoracial groups (Castañeda et al., 2019; Witbrodt et al., 2014). Alcohol use has been linked to alcohol-related problems and other substance use (Azpeitia et al., 2019; Torres et al., 2024). Despite this, limited work has examined how traditional masculine gender norms (TMGNs) relate to alcohol use among Latino immigrant men over time, or how this association may operate through cognitive processes such as alcohol outcome expectancies (AOEs). The present study addresses this gap by examining (a) the longitudinal impact of AOEs on the relationship between TMGNs and alcohol use and (b) the indirect effect of TMGNs on alcohol use via AOEs. Understanding these pathways can inform prevention and intervention efforts for Latino immigrant men.

Gender Schema Theory

Gender schema theory proposes that individuals learn and internalize gendered traits early in life and subsequently use these schemas to guide behavior and self-regulation (Bem, 1981). Informed by social cognitive framework, this theory suggests that individuals develop the ability to identify with “masculine” or “feminine” schemas through social learning. Learned characteristics associated with either masculine or feminine schemas can help individuals categorize information, make decisions, and regulate behavior based on their gender identification (Starr & Zurbriggen, 2017). Recent meta-analytic work suggests that conformity to masculine norms is generally related to alcohol use (Zamboanga et al., 2024b). For Latino men, aspects of TMGNs have been linked to alcohol use, although findings are mixed.

Traditional Masculine Gender Norms and Alcohol Use

In Latino cultural contexts, TMGNs are often described using the construct of machismo, conceptualized as two dimensions (Arciniega et al., 2008). Traditional machismo reflects dominance, hypermasculinity, and risk-taking. Caballerismo reflects chivalry, responsibility, and a strong emphasis on family and emotional connectedness (Arciniega et al., 2008).

Prior studies have generally found that traditional machismo is positively associated with alcohol use and, in some cases, with binge drinking among Latino men (Balagopal et al., 2022; Perrotte & Zamboanga, 2021). However, other studies report null or even inverse associations (Kissinger et al., 2013; Rojas et al., 2023). Findings for caballerismo are also inconsistent. Some work suggests caballerismo may be protective among Latino men including reduced binge drinking (Gonzalez & Barnett, 2023; Ojeda et al., 2014; Perrote et al., 2018), yet other studies have not observed significant associations (Kissinger et al., 2013; Valdez et al., 2019). Taken together, evidence suggests that the influence of TMGNs on alcohol use is complex and may depend on how masculinity is expressed and which alcohol outcomes are assessed.

Mixed findings may be partly due to variability in how alcohol is measured (e.g., frequency/quantity vs. binge drinking vs. alcohol-related problems; Perrote & Zamboanga, 2021; Zamboanga et al., 2024b). Recent meta-analytic work indicates that associations between masculine norms and alcohol use differ depending on the outcome and tend to be stronger for alcohol quantity/frequency than for binge drinking (Zamboanga et al. 2024b). Cognitive processes may also influence these associations. In particular, AOEs, which reflect beliefs about the anticipated effects of drinking, are a plausible mechanism (Leigh & Stacy, 1993).

Alcohol Outcome Expectancies (AOEs) as Moderation and Mediational Pathways

AOEs refer to cognitive processes that involve cognitive representations or anticipated beliefs about the effects of alcohol use (Leigh & Stacy, 1993). According to the Alcohol Expectancy Model, AOEs function as proximal cognitive influences that not only predict alcohol use but also mediate the influence of more distal factors (e.g., individual or environmental learning experiences) on drinking behaviors (Smith & Goldman, 1994). According to Smith and Goldman (1994), distal factors may include pre-drinking experiences, such as observable learning events of familial drinking or other vicarious learning regarding the properties and effects of drinking. Both positive and negative AOEs have been identified in the literature; positive AOEs include favorable consequences (e.g., social enhancement, reduced tension, increased sexual drive), whereas negative AOEs include unfavorable consequences associated with drinking (e.g., physical illness, cognitive/behavioral impairment; Leigh & Stacy, 1993). Empirical findings consistently demonstrate a positive association between endorsement of positive AOEs and alcohol use (Crisafulli et al., 2024; Dunne & Katz, 2015; Pinquart & Borgolte, 2024). However, findings concerning negative AOEs are more variable, with studies reporting mixed associations between these expectancies and alcohol use (Des Rosiers et al., 2012; Pinquart & Borgolte, 2024).

In general, a robust body of research suggests that AOEs play a critical role in the initiation and maintenance of drinking behaviors over time (Smit et al., 2018). Moreover, AOEs have been identified as key mediators and moderators in the relationship between individual/environmental factors, such as personality and social connectedness, and alcohol use (Crisafulli et al., 2024; Mezquita et al., 2015). Given this framework, we propose that AOEs may also help to not only explain the relationship between TMGNs and alcohol use, but also that the level of endorsement of both positive and negative AOEs may impact the relationship between TMGNs and alcohol use among Latino immigrant men. Learned experiences and expectations regarding alcohol use and masculinity may influence drinking patterns among Latino men (Valdez et al., 2019). In Latino male cultural contexts, gender norms, particularly those associated with traditional machismo and caballerismo, may shape beliefs about the expected positive and negative consequences of alcohol consumption, ultimately influencing drinking patterns (Perrotte & Zamboanga, 2021; Monk & Heim, 2013). From this perspective, AOEs may function as both moderation and mediation pathways by which TMGNs (i.e., traditional machismo and caballerismo) relate to alcohol use among Latino immigrant men.

To illustrate, Iwamoto et al. (2014) conducted a cross-sectional study demonstrating that endorsement of certain masculine gender norms, specifically playboy (e.g., belief that men should be sexually driven) and winning (e.g., belief that men should be dominant and competitive)—factors associated with traditional machismo (Arciniega et al., 2008; Zamboanga et al., 2024a)—were indirectly related to heavier alcohol use via heightened positive AOEs in a sample of ethnoracially diverse college men. Importantly, their findings revealed no measurement invariance by ethnicity, suggesting the model operated similarly across ethnoracial groups. Similarly, another cross-sectional study by Crisafully et al. (2024) found that positive AOEs explained the relationship between need to belong and social connectedness and alcohol use among ethnoracially diverse college students. These psychosocial constructs have been previously linked to adherence to TMGNs among Latino men (Lee et al., 2019; Perrote & Zamboanga, 2021), reinforcing the idea that AOEs may serve as critical cognitive mechanisms within this cultural and gendered framework.

AOEs have been identified as key mediators and moderators in the relationship between individual/environmental factors, such as personality and social connectedness, and alcohol use (Crisafulli et al., 2024; Mezquita et al., 2015). Drawing from this framework, AOEs may help explain (mediation) and shape (moderation) the relationship between TMGNs and alcohol use among Latino immigrant men. Within Latino cultural contexts, masculine norms may inform expectations about alcohol’s perceived benefits and costs, which in turn influence drinking (Valdez et al., 2019). For example, prior research has shown that masculine norms tied to sexual prowess and competitiveness—concepts consistent with traditional machismo—relate indirectly to heavier drinking through stronger positive AOEs (Perrotte & Zamboanga, 2021). These patterns suggest that AOEs may function as sociocognitive mechanisms through which culturally informed masculine norms relate to alcohol involvement.

Among Latino men, both positive and negative AOEs may interact with traditional machismo and lead to distinct associations with alcohol use. Positive AOEs may exacerbate preconceived traditional machismo beliefs regarding alcohol use that lead to increased drinking, whereas negative AOEs may attenuate such beliefs and reduce drinking (Perrote & Zamboanga, 2021). In contrast, although the association between caballerismo and AOEs has not been previously explored, existing theoretical frameworks provide a foundation for informed hypotheses. Given that caballerismo has been described as a cultural protective factor for Latino men that may deter excessive drinking (Perrote et al., 2018), it is plausible that greater endorsement of negative AOEs could amplify the protective influence of caballerismo beliefs against alcohol use (Anthenien et al., 2017). Conversely, increased endorsement of positive AOEs, which are generally associated with increased alcohol use (Pinquart & Borgolte, 2024), may inhibit the protective nature of caballerismo against alcohol use.

Furthermore, Latino men who endorsed increased caballerismo may be more likely to internalize negative AOEs, thereby contributing to decreased alcohol use. This supposition is based on research linking caballerismo to proximal alcohol specific masculine gender norms in drinking behaviors (Perrote et al., 2020), which have been associated with decreased engagement in alcohol use (Zamboanga et al., 2023) and alcohol-related problems (Perrote et al., 2020). Considering the moderate to strong association between caballerismo and controlled drinking norms among Latino men (Perrote et al., 2020), it is reasonable to infer that caballerismo may be associated with greater endorsement of negative AOEs. From this perspective, increased alcohol use may be perceived as a threat to self-regulation and emotional composure, traits central to caballerismo and its emphasis on respect and familial responsibility (Arciniega et al., 2008).

Of note, previous research has indicated that AOEs are prevalent among Latino men prior to immigration, and that both TMGNs and AOEs may shift post-immigration due to changes resulting from adaptation in the U.S. (Lee et al., 2019). These post-immigration changes may help explain the inconsistent findings regarding associations between gender norms and AOEs among Latino men. Such variability underscores the complex and multifaceted nature of cultural influences and the heterogeneity within the Latino population in the U.S. (Arciniega et al., 2008; Westrick et al., 2023). Despite these nuances, cultural factors related to gender norms remain underexplored in the AOEs literature (Monk & Heim, 2013), limiting our understanding of how they intersect with alcohol use and related outcomes. One such cultural factor is caballerismo, which reflects a uniquely Latino conceptualization of masculinity where culture and gender norms intersect (Perrotte & Zamboanga, 2021). However, the relationship between caballerismo and AOEs among Latino men remains unclear and warrants further investigation.

Although there has been an emergence of research examining the association between TMGNs and alcohol use among Latino men, this area remains predominantly cross-sectional in nature (Perrote & Zamboanga, 2021; Westrick et al., 2023). To our knowledge, only one longitudinal study has explored these associations over time to date (Kissinger et al., 2013), highlighting a critical gap in the literature. Research examining more complex associations between TMGNs and alcohol use among Latino men, including studies incorporating moderation and mediational pathways, have remained largely understudied. This is particularly true for culturally specific constructs such as caballerismo, which has not been previously examined in relation to AOEs and alcohol use within an integrated framework. The omission of caballerismo in studies examining AOEs and alcohol-related behaviors is consequential, given its conceptual distinction from traditional machismo and its potential to exert both protective and risk-related influences. The lack of research exploring how TMGNs (i.e., traditional machismo and caballerismo) intersect with AOEs undermines our understanding of the cultural and psychological factors that shape drinking behavior among Latino immigrant men. These gaps in the literature have real-world implications. Without longitudinal and mechanistic insights, our ability to develop effective prevention and intervention strategies aimed at reducing alcohol use and misuse among Latino men remains limited (Castañeda et al., 2019).

The Present Study

Guided by Gender Schema Theory, the Alcohol Expectancy Model, and prior work on Latino men’s drinking, we examined longitudinal associations between TMGNs, AOEs, and alcohol use among Latino immigrant men. Specifically, we assessed (a) interactions between TMGNs (traditional machismo, caballerismo) and positive and negative AOEs on alcohol use and (b) indirect effects of TMGNs on alcohol use through AOEs. To our knowledge, this is the first longitudinal study to compare competing moderation and mediation models linking these constructs in a community sample of Latino immigrant men.

We hypothesized that H1: Greater endorsement of traditional machismo would be associated with increased alcohol use and that greater endorsement of caballerismo would be associated with decreased alcohol use. H2: We expected that traditional machismo paired with stronger positive AOEs would predict higher alcohol use, whereas traditional machismo paired with stronger negative AOEs would predict lower alcohol use. H3: We expected caballerismo combined with positive AOEs to predict higher alcohol use, and caballerismo combined with negative AOEs to predict lower alcohol use. For mediation, H4: We hypothesized that traditional machismo would predict higher positive AOEs, which in turn would predict higher alcohol use, and that H5: Caballerismo would predict higher negative AOEs, which in turn would predict lower alcohol use.

Methods

Study Design and Population

Data were drawn from an NIH-funded longitudinal study on alcohol use trajectories among Latino immigrants (De La Rosa et al., 2020). The parent study enrolled both men and women. For the present analyses, we restricted the sample to Latino immigrant men because of our focus on masculine gender norms. The sample included 276 Latino immigrant men who had been living in the U.S. for approximately 10 years at the current analytic baseline. Participants were primarily of Cuban and Central American descent, with the remainder of South American descent, and averaged 35 years of age at baseline. Most were married or partnered, had completed at least high school, and reported authorized immigration status. Most (85%) did not report a family history of substance use.

We analyzed data from waves 5–7 of the parent study (April 2017–June 2020). These waves were treated as baseline, wave 2, and wave 3 in the current study because AOEs and TMGNs were first assessed at wave 5. Assessments were conducted annually. Most data were collected prior to COVID-19; thirteen participants completed Wave 3 after March 2020 via phone rather than in-person due to pandemic restrictions.

Respondent driven sampling (RDS) was used for recruitment of participants (Heckathorn, 1997). Specifically, each participant (the seed) was asked to refer to 3 individuals within their social network who met the inclusion criteria of the parent study: (a) being between the ages of 18 and 34, (b) having immigrated to the U.S. from a Latin American country within one year prior to the parent study’s baseline assessment that was collected between 2008 – 2009, (c) self-identifying as Latino, (d) living in Miami-Dade County, Florida, and (e) having the intention to stay in the U.S. The method of recruitment was followed up to three legs per seed to avoid skewing the sample. Detailed information on RDS has been published elsewhere (De La Rosa et al., 2020; Heckathron, 1997). The study was approved by the Institutional Review Board of a large public university in South Florida.

Measures

All assessments were administered in Spanish. Measures that were not already available in Spanish underwent translation and back-translation (Behling & Law, 2000).

Machismo Gender Norms

The Traditional Machismo and Caballerismo Scales (Arciniega et al., 2008) were used to examine machismo gender norms (includes two subscales: traditional machismo and caballerismo). Traditional machismo subscale reflects hypermasculine, aggressive, and power-driven traits, whereas caballerismo is characterized by nurturing, family-centeredness, and chivalry. Items are rated on a 7-point Likert scale from 1 (not at all) to 7 (very much so). Higher scores indicate stronger endorsement of each dimension. Mean subscale scores were computed. In the present study, Cronbach’s alpha for traditional machismo were at .89, .66, .70 at baseline, wave 2, wave 3, respectively; Cronbach’s alpha for caballerismo were at .69, .83, .73 at baseline, wave 2, wave 3, respectively.

Positive and Negative Alcohol Outcome Expectancies

The 34-item Alcohol Outcome Expectancies Scale (Leigh & Stacy, 1993) was used to measure positive and negative AOEs. This scale, which has been previously used with Latino samples (Westrick et al., 2023), asks participants to rate how likely each of the statements happen to them when they drink alcohol on a scale ranging from 1 (no chance) to 6 (certain to happen). Mean scores were calculated for positive and negative AOEs subscales. In the present study, Cronbach’s alpha for positive AOEs were at .97, .95, .97 at baseline, wave 2, wave 3, respectively; Cronbach’s alpha for negative AOEs were at .94, .94, .95 at baseline, wave 2, wave 3, respectively.

Alcohol Use

The Alcohol Use Disorders Identification Test-Consumption Subscale (AUDIT-C; Saunders et al., 1993) is a three-item measure of past-year drinking frequency and quantity. Items are scored 0–4 and summed (range 0–12), with higher scores indicating heavier alcohol use. A clinical cut off score of 4 is used to detect at risk drinking. The AUDIT-C has been validated across diverse populations (Frank et al., 2008). In the present study, Cronbach’s alpha for the AUDIT-C were at .82, .78, .79 at baseline, wave 2, wave 3, respectively.

Covariates

We included baseline age, education, income, marital status, immigration status, country of origin, and family history of alcohol use as covariates based on prior literature linking these factors to alcohol use among Latino men (e.g., Castañeda et al., 2019; Rojas et al., 2023; Westrick et al., 2023).

Analytic Plan

Preliminary analyses consisted of generating descriptive statistics across waves. We performed paired t-test for continuous variables compared to the baseline values, and McNemar's test for marital status and immigration status compared to the baseline frequencies, as well as the Stuart-Maxwell test for immigration status compared to the baseline. Main analyses included the use of linear mixed models (LMMs) to assess the association between TMGNs and alcohol use, in addition to the potential moderation effects of AOEs on the relationship between TMGNs and alcohol use. We first standardized the continuous variables before conducting LMM. We then performed crude LMMs for alcohol use, including each covariate, TMGNs (i.e., traditional machismo and caballerismo), and positive and negative AOEs, respectively. We then performed a multivariable model with TMGNs and AOEs, controlling for covariates. Interaction terms between each TMGN and both positive and negative AOEs were added to the model.

To test for mediation, we used a bootstrap approach within LMM to account for the random effects of repeated measures. Three LMMs were estimated for each bootstrap resample: (1) two separate models for each mediator (positive and negative AOEs) with traditional machismo and caballerismo as the main independent variables, controlling for demographic covariates; (2) a third model for alcohol use with both mediators, main independent variables, and demographic covariates. The average of the bootstrapping effect estimates and percentile confidence intervals were computed. Significance of mediation was assessed using two-tailed empirical p-values, calculated as the proportion of bootstrap estimates falling on the opposite side of zero from the mean indirect effect, multiplied by two.

To examine the suppression effect of TMGNs on alcohol use, we performed additional models for alcohol use using gender norms: reduced linear mixed-effects models with and without the AOEs. To further explore whether decreases in the endorsement of TMGNs were associated with reductions in alcohol use, a post-hoc analysis was performed. Specifically, we calculated changes in alcohol use, TMGNs, and AOEs between baseline and wave 2, as well as between wave 2 and wave 3. We then used a linear mixed-effects model to test whether reductions in TMGNs and changes in AOEs predicted decreases in alcohol use, controlling for baseline levels of alcohol use. A two-sided p-value < 0.05 was considered statistically significant. Statistical analyses were performed using the statistical software SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Preliminary Analyses

As shown in Table 1, average alcohol use score in the present sample was 4.4 (SD = 2.3) at baseline and declined over time, reaching 3.9 (SD = 2.1) with significant decreases from baseline to wave 2 (p = 0.003) and wave 3 (p < 0.001). The mean traditional machismo score was 2.6 (SD = 1.4) at baseline and decreased over time, with a mean score of 1.7 during the second and third wave of assessments. Traditional machismo scores decreased significantly from baseline to later waves (ps < 0.001). Caballerismo showed smaller changes but was significantly lower by the third wave 6.0 (SD = 0.8) compared to baseline 6.6 (SD = 0.6) (p < 0.001). The mean positive AOEs was 4.1 (SD = 1.2) at baseline and decreased to 2.4 (SD = 1) by wave 3 (ps < 0.001). In contrast, the mean negative AOEs were 2.8 (SD = 1.1) at baseline, 3 (SD = 1.2) at wave 2, and 2.3 (SD = 0.9) at wave 3, which was significantly lower than that at baseline (p < 0.001).

Table 1.

Descriptive Statistics and Preliminary Analyses

Variable	Baseline			Wave 2			p	Wave 3			p
	N	Mean (SD.)	Min- Max	N	Mean (SD.)	Min- Max		N	Mean (SD.)	Min- Max
Age at baseline	276	35 (4.9)	24-44	-	-	-		-	-	-
Annual income (thousand of US dollars $k)	275	42.3 (57.9)	2-625.1	270	42.1 (15.1)	3.8-128	0.933	266	58.4 (243.1)	0.04-4000	0.283
AUDIT subscale score: Alcohol consumption	275	4.4 (2.3)	0-10	270	4.1 (2.1)	0-10	0.003	266	3.9 (2.1)	0-10	<0.001
Traditional machismo	275	2.6 (1.4)	1-7	270	1.7 (0.7)	1-3.7	<0.001	265	1.7 (0.8)	1-5.3	<0.001
Caballerismo	275	6.6 (0.6)	2.8-7	270	6.7 (0.5)	1-7	0.063	265	6.0 (0.8)	1-7	<0.001
Positive alcohol outcome expectancies	275	4.1 (1.2)	1-6	270	2.9 (0.9)	1-5	<0.001	266	2.4 (1)	1-5.1	<0.001
Negative alcohol outcome expectancies	275	2.8 (1.1)	1-6	270	3 (1.2)	1-5.2	0.071	266	2.3 (0.9)	1-5.3	<0.001
Variable	n	%		n	%		p	n	%		p
Country of origin
Cuban	107	39.1		-	-			-	-
South American	59	21.5		-	-			-	-
Central American	108	39.4		-	-			-	-
Family history of alcohol use
Yes	42	15.2		-	-			-	-
No	234	84.8		-	-			-	-
Education							0.894				0.756
Less than high school	4	1.5		2	0.7			1	0.4
High school diploma	154	56.2		152	56.3			144	54.1
Some training/college after high school	94	34.3		96	35.6			100	37.6
Bachelor's (4-5 year college) degree	19	6.9		16	5.9			18	6.8
Graduate/Professional studies	3	1.1		4	1.5			3	1.1
Marital status							0.011				0.002
Single or separated	91	33.0		74	27.4			72	27.1
Married/In a relationship	185	67.0		196	72.6			194	72.9
Immigration status							0.059				0.008
Unauthorized	16	5.8		11	4.1			9	3.4
Authorized	259	94.2		259	95.9			257	96.6

Note.

SD. = Standard Deviation

Min-Max = Minimum to Maximum

p are p-values from paired-t test for continuous variables compared to the baseline values, and McNemar's test for marital status and immigration status compared to the baseline frequencies, as well as Stuart-Maxwell test for immigration status.

Moderation Analysis

Covariates

Compared to individuals with a high school diploma or less, participants with some post-high school training reported 0.2 standard deviations lower alcohol use (β = −0.22, 95% CI = [−0.39, −0.04], p = 0.016). People with a college degree or higher had 0.4 standard deviations lower alcohol use (β = −0.43, 95% CI = [−0.67, −0.19], p = 0.001). Other covariates were not significantly related to alcohol use (ps > 0.05).

Changes in Alcohol Use Over Time

There was a significant decrease in alcohol use from baseline to the third wave (β = −0.13, 95% CI = [−0.26, −0.01], p = 0.041), while controlling for covariates in the model (Table 2). However, no significant difference in alcohol use was found between baseline and the second wave (p > 0.05).

Table 2.

Estimates from Linear Mixed-effects Model for Alcohol Consumption among Male Participants from Baseline to Wave 3 (N= 807)

Independent Variable	Zero-Ordered Models				Final Model
	Estimate	95% Confidence Interval		p	Estimate	95% Confidence Interval		p
		Lower	Upper			Lower	Upper
Time:
Wave 2 vs. Baseline	−0.13	−0.22	−0.03	0.007	−0.06	−0.18	0.05	0.290
Wave 3 vs. Baseline	−0.23	−0.32	−0.13	<0.001	−0.13	−0.26	−0.01	0.041
Age	−0.03	−0.14	0.07	0.537	−0.04	−0.15	0.07	0.440
Education:
Some training after high school vs. High school or less than high school	−0.25	−0.42	−0.07	0.005	−0.22	−0.39	−0.04	0.016
College or above vs. High school or less than high school	−0.51	−0.74	−0.28	<0.001	−0.43	−0.67	−0.19	0.001
Annual income in thousands of dollars	0.00	−0.04	0.05	0.903	0.01	−0.03	0.06	0.632
Marital status: Married/In a relationship vs. Single or separated	−0.12	−0.29	0.05	0.177	−0.08	−0.26	0.10	0.388
Immigration status: Authorized vs. Unauthorized	−0.33	−0.68	0.02	0.062	−0.20	−0.61	0.20	0.280
Family history of alcohol use: Yes vs. No	−0.26	−0.56	0.04	0.084	−0.20	−0.49	0.09	0.173
Country of Origin:
Central American vs. Cuban	0.06	−0.18	0.30	0.646	0.01	−0.23	0.25	0.926
South American vs. Cuban	−0.19	−0.47	0.10	0.203	−0.11	−0.39	0.17	0.454
Traditional machismo	0.02	−0.03	0.06	0.449	−0.11	−0.17	−0.05	0.001
Caballerismo	0.00	−0.04	0.05	0.845	−0.04	−0.09	0.01	0.144
Positive alcohol outcome expectancies	0.13	0.08	0.17	<0.001	0.14	0.07	0.21	<0.0001
Negative alcohol outcome expectancies	0.05	0.01	0.10	0.023	−0.01	−0.07	0.04	0.621
Traditional machismo*Positive alcohol outcome expectancies	-	-	-	-	0.10	0.04	0.16	0.002
Traditional machismo*Negative alcohol outcome expectancies	-	-	-	-	−0.06	−0.11	−0.01	0.031
Caballerismo*Positive alcohol outcome expectancies	-	-	-	-	0.00	−0.07	0.07	0.900
Caballerismo*Negative alcohol outcome expectancies	-	-	-	-	−0.02	−0.08	0.05	0.617

Main Effects

Traditional machismo was associated with decreased alcohol use (β = −0.11, 95% CI = [−0.17, −0.05], p = 0.001). Participants who scored one standard deviation higher in traditional machismo reported a 0.11 standard deviation lower in alcohol use, when both positive and negative AOEs were 0. Positive AOEs were also positively associated with increased alcohol use (β = 0.14, 95% CI = [0.07, 0.21], p < 0.0001). Results indicate that one standard deviation higher in positive AOEs was associated with a 0.14 standard deviation increase in alcohol use, when traditional machismo is 0. Notably, caballerismo and negative AOEs were not significantly associated with alcohol use (p > 0.05).

Moderation Effects

There were significant interactions between traditional machismo and both positive (β = 0.10, 95% CI = [0.04, 0.16], p = 0.002) and negative (β = −0.06, 95% CI = [−0.11, −0.01], p = 0.031) AOEs. When positive AOEs were high, the inverse association between traditional machismo and alcohol use weakened. In contrast, higher negative AOEs strengthened the inverse association between traditional machismo and alcohol use (See Figure 1). The simple slope estimates are presented in Table 3. Overall, these patterns indicate that AOEs shape how traditional machismo relates to alcohol use.

Figure 1. Moderation Effects of Alcohol Outcome Expectancies on Traditional Machismo and Alcohol Consumption.

Table 3.

Simple Slopes of Traditional Machismo by Levels of Positive and Negative Alcohol Outcome Expectancies

Alcohol Outcome Expectancies	Standardized Value	Simple Slope of Traditional Machismo	95% Confidence Interval		p
			Lower	Upper
Positive Alcohol Outcome Expectancies	−1	−0.21	−0.32	−0.10	<0.001
	0	−0.11	−0.17	−0.05	<0.001
	1	−0.01	−0.07	0.05	0.774
Negative Alcohol Outcome Expectancies	−1	−0.05	−0.12	0.02	0.190
	0	−0.11	−0.17	−0.05	0.001
	1	−0.17	−0.26	−0.08	<0.001

Mediation Analysis

Results indicated that traditional machismo had a significant indirect effect on alcohol use through positive AOEs (indirect effect = 0.03, 95% CI = [0.02, 0.05]). Specifically, stronger traditional machismo predicted stronger positive AOEs (β = 0.21, 95% CI = [0.15, 0.27], p < 0.001), which in turn predicted higher alcohol use (β = 0.13, 95% CI = [0.06, 0.19], p < 0.001), while accounting for covariates (Figure 2).

Figure 2. Mediation Effects of Alcohol Outcome Expectancies on Traditional Masculine Gender Norms and Alcohol Consumption.

Note. * p < 0.05, ** p <= 0.01, *** p <= 0.001. The bootstrapped indirect effect of Traditional Machismo on Alcohol Consumption was 0.03 with 95% CI: [0.017, 0.052] through Positive Alcohol Use Outcome Expectancies, and −0.003 with 95% CI: [−0.010, 0.002] through Negative Alcohol Use Outcome Expectancies. The bootstrapped indirect effect of Caballerismo on Alcohol Consumption was -0.002 with 95% CI: [−0.012, 0.009] through Positive Alcohol Use Outcome Expectancies, and −0.003 with 95% CI: [−0.003, 0.011] through Negative Alcohol Use Outcome Expectancies. Linear mixed effects models controlled for time of assessment, age, education, income, marital status, documentation status, family history of alcohol use, and country of origin.

In addition, results showed that the direct effect of traditional machismo on alcohol use was negative (β = −0.06, 95% CI = [−0.11, −0.01], p = 0.020), suggesting that higher traditional machismo is significantly associated with reduced alcohol use. The indirect pathway via negative AOEs was not significant. Caballerismo did not show significant indirect effects through AOEs.

Results from the models for the mediation analysis show that time, education, and traditional machismo was correlated with AOEs (Table 4). There was a significant decrease in positive AOEs over time (wave 2 vs. baseline: β = −0.74, 95% CI = [−0.88, −0.59], p < 0.0001; wave 3 vs. baseline: β = −1.13, 95% CI = [−1.29, −0.98], p < 0.0001) and an increase in negative AOEs (wave 2 vs. baseline: β = 0.26, 95% CI = [0.09, 0.44], p = 0.003; wave 3 vs. baseline: β = −0.41, 95% CI = [−0.59, −0.23], p < 0.0001). Higher education (college or above compared to high school or less than high school) was associated with lower positive (β = −0.37, 95% CI = [−0.59, −0.15], p = 0.001) and negative AOEs (β = −0.44, 95% CI = [−0.71, −0.18], p = 0.001), as well as less alcohol use (β = −0.47, 95% CI = [−0.71, −0.24], p < 0.0001). A family history of alcohol use was associated with lower positive AOEs (β = −0.24, 95% CI = [−0.40, −0.07], p = 0.004).

Table 4.

Estimates from Linear Mixed-effects Models for the Mediation Analysis among Male Participants from Baseline to Wave 3 (N= 801).

Independent Variable	Model for Positive Alcohol Use Outcomes Expectancies				Model for Negative Alcohol Use Outcomes Expectancies				Model for Alcohol Use
	Estimate	95% Confidence Interval		p	Estimate	95% Confidence Interval		p	Estimate	95% Confidence Interval		p
		Lower	Upper			Lower	Upper			Lower	Upper
Time:
Wave 2 vs. Baseline	−0.74	−0.88	−0.59	<0.0001	0.26	0.09	0.44	0.003	−0.04	−0.16	0.07	0.454
Wave 3 vs. Baseline	−1.13	−1.29	−0.98	<0.0001	−0.41	−0.59	−0.23	<0.0001	−0.14	−0.26	−0.01	0.028
Age	0.07	0.00	0.13	0.047	0.06	−0.02	0.14	0.119	−0.04	−0.15	0.07	0.469
Education:
Some training after high school vs. High school or less than high school	0.06	−0.06	0.19	0.325	0.003	−0.15	0.15	0.972	−0.219	−0.39	−0.05	0.013
College or above vs. High school or less than high school	−0.37	−0.59	−0.15	0.001	−0.44	−0.71	−0.18	0.001	−0.47	−0.71	−0.24	<0.0001
Annual income in thousands of dollars	−0.04	−0.09	0.02	0.208	−0.05	−0.12	0.02	0.152	0.01	−0.03	0.05	0.657
Marital status: Married/In a relationship vs. Single or separated	−0.09	−0.22	0.05	0.222	−0.17	−0.34	−0.01	0.040	−0.08	−0.26	0.09	0.348
Immigration status: Authorized vs. Unauthorized	0.29	0.00	0.57	0.046	0.20	−0.14	0.53	0.251	−0.22	−0.57	0.12	0.207
Family history of alcohol use: Yes vs. No	−0.24	−0.40	−0.07	0.004	−0.15	−0.34	0.05	0.137	−0.19	−0.48	0.11	0.213
Country of Origin:
Central American vs. Cuban	0.02	−0.11	0.15	0.739	−0.03	−0.18	0.13	0.725	0.03	−0.21	0.27	0.801
South American vs. Cuban	0.05	−0.11	0.20	0.567	0.10	−0.08	0.29	0.276	−0.10	−0.38	0.19	0.502
Traditional machismo	0.21	0.15	0.28	<0.0001	0.10	0.03	0.18	0.008	−0.06	−0.11	−0.01	0.020
Caballerismo	−0.01	−0.07	0.05	0.751	−0.12	−0.19	−0.04	0.002	−0.04	−0.09	0.01	0.108
Positive alcohol outcome expectancies	-	-	-	-	-	-	-	-	0.13	0.06	0.19	0.0001
Negative alcohol outcome expectancies	-	-	-	-	-	-	-	-	−0.03	−0.08	0.03	0.334

Note.

Estimates are adjusted fixed effects from the three models for mediation analysis.

Follow-up Analyses

Analyses examining potential suppression effect of TMGNs indicated direct effect of traditional machismo on alcohol use was non-significant when the mediator was not included (β = 0.02, 95% CI [−0.03, 0.06], p = 0.46). However, when positive AOEs were added to the linear mixed-effects model, traditional machismo became significantly negatively correlated with alcohol use (β = −0.05, 95% CI = [−0.10, −0.00], p = 0.043). When we controlled for all remaining covariates in the model, results were consistent that traditional machismo continued to be negatively correlated with alcohol use (β = −0.06, 95% CI = [−0.11, −0.01], p = 0.02). This change of significance in traditional machismo is consistent with suppression, such that the inclusion of a mediator reveals a significant association that is masked in the reduced model.

Post-hoc analysis revealed that greater declines in traditional machismo over time were associated with greater reductions in alcohol use, (β = −0.07, 95% CI: [−0.11, −0.02], p = 0.004), even after controlling for AOEs and baseline alcohol use in the model.

Discussion

This study tested longitudinal moderation and mediation models linking traditional machismo, caballerismo, AOEs, and alcohol use in Latino immigrant men. Several key findings emerged.

First, higher endorsement of traditional machismo was associated with less alcohol use—contrary to our initial hypothesis and most research suggesting a positive relationship (Perrotte & Zamboanga, 2021; Valdez et al., 2019). Concurrently, we found no significant association between caballerismo and alcohol use, suggesting that only certain aspects of TMGNs such as traditional machismo may influence alcohol use. One possible explanation is that the meaning and expression of these norms may shift with time in the U.S. Men may begin to endorse a version of masculinity that blends elements of traditional machismo (e.g., authority, responsibility) and caballerismo (e.g., family obligation), emphasizing self-control and caretaking rather than heavy alcohol use. In the present study, endorsement of both traditional machismo and caballerismo declined overtime. A post-hoc analysis showed that decreases in traditional machismo were associated with reductions in alcohol use. This suggests evolving gender norms may coincide with reduced drinking, perhaps due to the acculturative process.

Notably, these findings underscore the importance of reconsidering traditional machismo as a protective or universal risk factor for alcohol use and indicate the need to explore individual and contextual influences. Factors such as alcohol use environments, familial and peer influences, other social pressures, and cultural values (e.g., respect) may mediate or influence this relationship (Arciniega et al., 2008; Valdez et al., 2019). For example, the contexts and social networks in which Latino men function may influence AOEs related to Latino men gender norms. Latino men’s endorsement of AOEs and drinking behaviors may be different when they socially gather with other Latino men (e.g., playing dominoes, socializing after a soccer game) compared to when they attend family gatherings where they may be viewed as the respectful head figure who everyone looks up to for guidance.

Second, our findings indicated that positive AOEs were related to increased alcohol use, whereas negative AOEs were not directly associated with alcohol use. Prior work has shown that positive AOEs are reliable predictors of drinking (Crisafulli et al., 2024; Pinquart & Borgolte, 2024), while associations involving negative AOEs are less consistent (Pinquart & Borgolte, 2024). One potential explanation for the nonsignificant direct association between negative AOEs and alcohol use is the possible influence or mediating role of extraneous variables not considered in the present study, such as drinking motives—consistent with the tenets of Cox and Klinger’s Motivational Model of Alcohol Use (1988). For example, Bacio (2021) found that negative AOEs were indirectly related to problematic drinking via drinking to cope with negative mood among Latino college students, suggesting that the relationship between AOEs and alcohol use may also be more complex.

Furthermore, we found that endorsement of both positive and negative AOEs significantly decreased from baseline to third wave assessment, with positive AOEs showing a steeper decline compared to negative AOEs. These findings suggest the need to further examine trajectories of AOEs by age and acculturation among Latino men, as well as how these factors may moderate the relationship between TMGNs, AOEs, and alcohol use over time. As discussed by Monk and Heim (2013), the influence of positive and negative AOEs on alcohol use may change over time according to age. As individuals age, drinking motives also tend to shift from social or enhancement motives to coping motives (Kuntsche et al., 2005). Consequently, negative AOEs may become less influential, as drinking is driven more by internal regulation than anticipated outcomes. Moreover, a recent meta-analysis by Pinquart and Borgolte (2024) also found that the association between positive AOEs and alcohol use declined from adolescence to young adulthood, whereas the association between negative AOEs and alcohol use became less negative over time, indicating that negative AOEs exert a weaker deterrent effect on alcohol use in young adulthood compared to adolescence. According to Pinquart and Borgolte (2024), younger individuals may be more susceptible to positive AOEs, while older drinkers may have greater lived experience with alcohol-related consequences and become less influenced by these expectancies. Similarly, acculturation levels may shape these relationships, as prior research suggests that time spent living in the U.S. and broader acculturation processes are associated with shifts in alcohol use patterns among Latino immigrants (Alvarez et al., 2017; Lorenzo-Blanco et al., 2016). Hence, future research should consider the acculturation processes and developmental aspects associated with AOEs and alcohol use across the lifespan (Alvarez et al., 2017; Kuntsche et al., 2005; Monk & Heim, 2013).

Third, AOEs moderated the association between traditional machismo and alcohol use. When men strongly endorsed positive AOEs, the protective (inverse) association between traditional machismo and alcohol use weakened. When men endorsed stronger negative AOEs, the inverse association between traditional machismo and alcohol use strengthened. These findings highlight AOEs as key cognitive influences that can either buffer or intensify the link between TMGNs and alcohol use. Conversely, caballerismo did not show parallel moderation effects, suggesting that its relevance may emerge more clearly for other alcohol-related outcomes (e.g., binge drinking, alcohol-related problems) rather than for general alcohol consumption (Perrote et al., 2018).

Fourth, mediation analyses indicated an indirect effect of traditional machismo on increased alcohol use via positive AOEs. In other words, men higher in traditional machismo tended to endorse stronger positive AOEs, which predicted higher alcohol use. This finding is consistent with a prior cross-sectional study examining the mediating role of positive AOEs in the relationship between masculine gender norms consistent with hypermasculinity (e.g., playboy and winning), such as traditional machismo, and increased drinking (Iwamoto et al., 2014). Notably, the indirect effect of traditional machismo on alcohol use through positive AOEs revealed a suppression effect. Specifically, there is a direct and negative effect of traditional machismo that becomes positive when positive AOEs are added to the model. This finding underscores the importance of examining underlying mechanisms by which TMGNs may relate to alcohol use and cautions against interpreting the influence of TMGNs on alcohol use in isolation. Our research further extends the literature to examine this association over time among Latino immigrant men specifically using Latino culture specific gender norms.

Contrary to expectations, there was no significant indirect effect of caballerismo on alcohol use via negative AOEs. As previously mentioned, an explanation for nonsignificant findings regarding caballerismo may be that endorsement of traditional machismo is more directly relevant to alcohol use than caballerismo. As Perrotte and Zamboanga (2021) noted, the conceptual understanding regarding how caballerismo relates to alcohol use remains unclear. While evidence suggests that caballerismo may act as a protective factor among Latino men health (Galvan et al., 2014; Ojeda & Piña-Watson, 2014; Rivera et al., 2020), caballerismo has received less attention in alcohol use research compared to traditional machismo (Perrotte & Zamboanga, 2021). However, one hypothesis is that caballerismo is often associated with family orientation and interpersonal responsibility (Arciniega et al., 2008), and alcohol use among Latino men who endorse caballerismo may be shaped more by cultural customs and norms in familial and social gatherings, which can vary by country of origin (Chartier et al., 2015; Correa-Fernández et al., 2022). Furthermore, existing measures of TMGNs (i.e., traditional machismo, caballerismo) may not adequately capture aspects of masculinity emergent from the acculturation process and how they relate to alcohol use (Perrotte et al., 2020). Therefore, further exploration of the theoretical framework and measurement of caballerismo as well as its associations with drinking behaviors among Latino immigrant men is warranted.

Additionally, while several covariates were included in the models, education emerged as the only significant predictor of alcohol use. Specifically, Latino men with higher levels of education reported lower alcohol use compared to their peers, suggesting that increased education may be a protective factor against alcohol use and at-risk drinking. This finding is consistent with a recent study on socioeconomic and cultural correlates of alcohol use frequency and quantity among Latino adult men also residing in Miami-Dade County, Florida (Rojas et al., 2023). It also lends support to the notion that limited knowledge about the risks associated with increased alcohol use may contribute to higher rates of alcohol misuse among Latino men (Valdez et al., 2019). Nonetheless, further research is needed to more comprehensively examine the role of socioeconomic factors in shaping alcohol use patterns among Latino immigrants.

Another consideration is that participants in the present study had lived in the U.S. for approximately 10 years at baseline. While the overall effect of time on alcohol use was nonsignificant across all assessment waves, our findings indicated a significant reduction in alcohol use among Latino men between the first and third waves of assessment. This decline may be attributed to participants’ increased time in the U.S.—by the third wave, participants had spent over a decade in the U.S., likely becoming more acclimated to their host country compared to their earlier years. Greater adjustment could reduce their susceptibility to immigration-related stressors which tend to be more pronounced during the initial years in the U.S. (Cabrera Tineo et al., 2020; Romano & Sánchez, 2023). To substantiate this interpretation, earlier analyses from the parent longitudinal study (including women and earlier waves not included in the present study) showed overall declines in alcohol use quantity and frequency from pre-immigration to post-immigration, as captured at baseline (De La Rosa et al., 2020).

Implications

A major strength of the current study lies in its use of competing longitudinal models to examine the associations between TMGNs, AOEs, and alcohol use among Latino immigrant men—variables that have been largely understudied and limited to cross-sectional designs. Our study demonstrates that AOEs may serve as both moderators and mediators in the relationship between TMGNs and alcohol use over time, underscoring the need for further longitudinal analyses examining competing models. Findings from such studies can more adequately inform preventative efforts and treatment interventions targeting alcohol use among Latino men. Although the present findings demonstrated there was a significant decline in alcohol use from baseline assessment to the third wave assessment, over half of the study sample reported engaging in at-risk drinking behaviors, using the threshold score of 4 on the AUDIT-C, with 68%, 63%, and 57% of participants falling into this category at baseline, wave two, and wave three, respectively. These percentages reinforce the public health significance of addressing alcohol use among Latino men, and underscore the need for earlier preventative efforts and treatment interventions following immigration to the U.S. Furthermore, these findings suggest the importance of examining different alcohol outcomes to better understand the associations between TMGNs, AOEs, and alcohol use among these Latino immigrant men.

Based on the present findings, we recommend that mental health and alcohol use interventions consider the role of TMGNs and AOEs on drinking behaviors among Latino men. Providers should remain mindful of their own biases related to TMGNs and adopt a strength-based, culturally responsive approach that recognizes the multidimensionality of manhood and gender norms within this population (Lauricella et al., 2021). Engaging Latino men in reflective discussions about their gender identity and belief, such as exploring what “being a man” means to them, can help providers better understand how adherence to specific masculine norms may shape motivations for alcohol use and related behaviors. Similarly, providers should inquire about the cognitive processes, such as AOEs, associated with Latino men's engagement in alcohol use. Exploring motivations for drinking and AOEs will provide critical insights into how Latino men understand and relate to alcohol use, ultimately helping identify and reframe maladaptive beliefs that contribute to problematic drinking patterns. Existing research also shows that masculine norms, AOEs and cognitive factors are closely connected. Traditional masculine beliefs and gender role conflict have been linked to higher alcohol use through coping and enhancement motives (Uy et al., 2014). Conforming to norms such as competitiveness or sexual prowess has also been tied to heavier drinking, especially among men with higher impulsivity (de Andrade et al., 2023). In clinical settings, providers might consider whether Latino men experience gender role conflict, such as difficulty showing vulnerability or pressure to meet masculine expectations, which may lead to drinking as a way to express masculinity and hesitancy in engaging in help seeking behaviors (Groeschel et al., 2010).

We also recommend the implementation of psychoeducational programs that address drinking-related cognitions. Such programs can help Latino men understand how their beliefs and expectations influence their drinking habits. Given that some Latino men may not recognize the health risks of at-risk drinking (Valdez et al., 2019), psychoeducation may be particularly effective, especially in light of our finding that higher educational attainment was associated with lower levels of alcohol use. Another consideration are expectancy challenge interventions (ECIs), which aim to modify maladaptive beliefs about alcohol’s effects by highlighting the discrepancy between expected and actual outcomes. ECIs have been shown to reduce positive AOEs and modestly decrease alcohol use, particularly those related to social, sexual, tension, and arousal (Scott-Sheldon et al., 2012; Gesualdo & Pinquart, 2021). However, ECI’s have been implemented primarily with high school and college populations. Culturally tailoring these approaches may help Latino men critically evaluate and reframe beliefs linking TMGNs, AOEs, and drinking behaviors.

Limitations and Constraints on Generality

The present findings need to be considered within the context of several imitations. The present sample consisted of Latino immigrant men residing in Miami-Dade County, Florida, which may not fully represent the demographics of the overall Latino population in the U.S. and limits the generalizability of the present findings. Although we accounted for country of origin, we were only able to examine Cuban and other regions of the country of origin versus specific Latino nationalities aside from Cuban that may provide a more nuanced understanding of alcohol use among this heterogeneous group. To that effect, another limitation of the present study were the internal consistency estimates for traditional machismo and caballerismo across certain waves (α = .66 and α = .69, respectively), which may have attenuated the observed associations. This could have potentially underestimated the true effects of these gender norm dimensions. While the Traditional Machismo & Caballerismo Scale (Arciniega et al., 2008) has demonstrated strong reliability and validity with Latino populations (primarily Mexican American men), it has not been comprehensively validated in Latinos of diverse national origins (e.g., Cuban, South American). Thus, measurement invariance and construct equivalence among Latino men diverse origins cannot be assumed, and our findings should be interpreted with this context in mind.

Importantly, the present study only examined alcohol use, which does not consider the associations between TMGNs, AOEs, and other types of alcohol outcomes, such as binge drinking and alcohol-related problems. As mentioned previously, it is possible TMGNs may have different influences over specific alcohol outcomes (e.g., general alcohol use vs. binge drinking vs. alcohol-related problems; Zamboanga et al., 2024b). Future studies should explore how TMGNs relate to alcohol expectancy valuations (Monk & Heim, 2013) and the use of alcohol protective behavioral strategies (Cox et al., 2024; Pearson, 2013; Zamboanga et al., 2025) to better understand both risk and protective factors in alcohol use among Latino men. In addition, future research on how masculine gender norms are directly tied to drinking behaviors, such as norms of excess and controlled drinking, is warranted. Recent work using the Masculine Drinking Norms Measure (MDNM; Zamboanga et al., 2024a) has provided valuable insights into how distinct dimensions of masculine drinking norms relate to alcohol consumption patterns among college men. Extending this line of inquiry to Latino populations could enhance understanding of the cultural and gendered factors that shape drinking behaviors in this population.

As noted above, Latinos are a heterogeneous group and may have varying levels of endorsement of TMGNs and alcohol use, depending on several factors not accounted for in the present study (e.g., ethnic identity, acculturation, acculturative stress, pre- and post-immigration trauma; Castañeda et al., 2019; Ruiz et al., 2024; Sanchez et al., 2017). In particular, it would be helpful to understand how acculturation processes may influence the relationships between TMGNs, AOEs, and alcohol use over time among this population. It is possible that acculturation may exert different associations among these variables over time as it is a dynamic process constantly influencing Latino immigrants' experiences in the U.S., and consequently their mental wellbeing and engagement in adaptive and maladaptive behaviors (Alvarez et al., 2017; Des Rosiers et al., 2012; Lorenzo-Blanco et al., 2016; Schwartz et al., 2015). Therefore, we recommend that future research explore more complex structural relationships that account for these evolving influences over time, while accounting for other extraneous factors not included in this study (e.g., acculturative stress, perceived discrimination, and parent–child acculturation gaps; Ruiz et al., 2024)

Conclusion

The present study contributes to the literature by examining the longitudinal associations between TMGNs, AOEs, and alcohol use among Latino immigrant men. To our understanding, the current study is the first longitudinal study examining competing moderation and mediation models of the relationships between TMGNs, AOEs, and alcohol use among Latino immigrant men. The present findings provide critical insights on the importance of AOEs as both moderation and mediation pathways of the link between TMGNs and alcohol use. Results from this study raise important questions about current conceptualizations of machismo gender norms (traditional machismo and caballerismo). Specifically, they suggest the need to expand existing measures of TMGNs to move beyond the negative portrayals of traditional machismo and consider the development of emergent masculine gender norms shaped by acculturation. Additional, future research should explore gender norms more directly tied to drinking behavior, such as masculine gender norms of excess and controlled drinking (Perrote et al., 2020; Zamboanga et al., 2023). Elevated rates of at-risk drinking among Latino men underscore alcohol use as a major public health concern and emphasize the importance of implementing timely, effective prevention and intervention strategies after immigration to the U.S.

Public Significance.

To our knowledge, the present study is the first to examine competing longitudinal models of moderation and mediation pathways of alcohol outcome expectancies (AOEs) on traditional masculine gender norms and alcohol use among Latino immigrant men. AOEs were found to influence and explain the relationship between traditional machismo and alcohol use, underscoring the key role of drinking cognitions on specific gender norms in relation to alcohol use over time in this understudied population.