Event-Level Drinking Intentions, Social Drinking Contexts, and Motives Affect Alcohol Consumption Among Same-Sex Female Couples

Abstract

Objective:

Young sexual minority women (SMW; e.g., lesbian, bisexual women) are at increased risk for heavy drinking, although SMW in relationships may be at less risk than those who are single. The current study examined how SMW partners influence drinking in the moment, if their own intentions predict drinking beyond partner influence and if this association is moderated by partner influence, and if their own drinking motives moderate partner influence.

Method:

Young adult same-sex female couples (N = 162 couples) were recruited to participate in a 14-day daily diary study where they reported each morning their drinking intentions, motives, and behavior as well as whether their partner was present during drinking occasions.

Results:

Findings from multilevel structural equation models revealed that own intentions and select partner influences (i.e., intentions, drinking) were uniquely linked to a greater quantity of drinking at the within-subject level. Significant interactions between own intentions and partner intentions, and between partner intentions and select drinking motives (social and enhancement), on same-day alcohol use were found at the within-subject level.

Conclusions:

Findings suggest that partner influences have a strong effect on drinking at the daily level and can be strengthened or weakened when drinking for specific reasons. These findings expand on prior daily diary research on the effects of drinking motives by examining how these processes unfold in same-sex female couples and their intersection with partner intentions.

Young sexual minority women (SMW; e.g., lesbian, bisexual women) are at increased risk for heavy drinking compared with their heterosexual counterparts (for a review, see Hughes et al., 2020). Among a sample of young adult women, 23% of SMW were classified as heavy drinkers, compared with only 6% of heterosexual women (Gruskin et al., 2001). SMW are also 11 times more likely to meet the criteria for alcohol use disorder than heterosexual women (Drabble & Trocki, 2005) and more likely to experience negative consequences from problematic drinking behavior (Wilsnack et al., 2008). Within this high-risk population, SMW in committed relationships are at a slightly lower risk for heavy drinking and consequences than those who are single, suggesting that being in a committed relationship may be a protective factor for problematic drinking among SMW (Veldhuis et al., 2019). Married SMW also are more likely to drink at similar levels as their partners (i.e., behavioral concordance) than heterosexual men and women in different-sex marriages (Holway et al., 2018).

Although alcohol consumption has been examined across women in same-sex relationships, less is known about partner influences on consumption at the daily level. One study examining daily alcohol consumption among married partners in same-sex and different-sex relationships found that women in same-sex marriages do not change their daily consumption levels based on how much their spouses drank the previous day, whereas women in different-sex marriages decrease their daily drinking if their spouse drank more the previous day (Pollitt et al., 2020), suggesting that women in same-sex relationships may be less affected by partner drinking. Taken together, SMW represent a high-risk population for heavy drinking, and identifying risk factors that could elevate drinking is crucial. Much of what is known about partner influence comes from studies of different-sex couples. Given this nascent literature, it is important to examine SMW in relationships with other women to identify when partner influence is strongest on drinking. As such, the current study seeks to examine associations between partner influences (e.g., partner drinking) on drinking at the daily level among SMW couples, including when this influence may be stronger or weaker.

Drinking intentions are theoretically and empirically linked to alcohol consumption and may be subject to social influence from relationship partners. Classical models of health behavior engagement, such as the theory of planned behavior (Ajzen, 1991) or the theory of reasoned action (Fishbein & Ajzen, 1975), often describe behavioral intentions as a precursor to health behavior engagement. However, much of what we know about drinking intentions and their association with alcohol consumption comes from studies of college students (Rimal & Mollen, 2013; Testa et al., 2009), and these studies typically comprise samples that are majority heterosexual (Freibott et al., 2022) and that typically have not examined the roles of relationship status or partner influence, which may be important. For example, higher group average drinking intentions before entering a nightclub are associated with higher individual blood alcohol concentration when exiting the club, suggesting that others' intentions predict greater personal consumption (Miller et al., 2013). Further, greater close individuals' (e.g., best friends, relationship partners) intentions for participants to get drunk when celebrating their 21st birthday (i.e., prointoxication intentions) were associated with greater alcohol consequences when pro-safety intentions were low (Fillo et al., 2017). Moderation of others' drinking intentions on associations between personal intentions and drinking has not yet been examined, but greater quantities of drinks consumed by others strengthened the association between daily intentions and drinking (Hamilton et al., 2022). These findings suggest that others' drinking intentions (including their relationship partners) are associated with personal consumption, and drinking by others strengthens associations between personal intentions and drinking. These associations have not yet been examined among SMW, nor have any studies explored how partner influences (intentions, presence, and drinking) moderate associations between personal intentions and drinking daily.

Drinking motives (i.e., reasons for drinking) are robust predictors of drinking (for a meta-analysis, see Bresin & Mekawi, 2021). Significant indirect effects for drinking to cope on associations between negative experiences (e.g., microaggressions) and consumption among SMW have been observed cross-sectionally (Kalb et al., 2018; Kelley et al., 2018). Higher daily social, coping, and enhancement motives are also linked to greater same-day alcohol use among SMW (Dworkin et al., 2018). Further, higher daily enhancement drinking motives strengthen the association between SMW drinking with heterosexual individuals only (compared to drinking with SM individuals only or with both heterosexual and SM individuals) and quantity. This limited evidence suggests that daily drinking motives may differ in terms of their effect on alcohol consumption and may affect how much the people with whom SMW individuals are drinking matter. However, no study to date has investigated the link between drinking motives and drinking within same-sex female relationships at the daily level. Further, whether partner influences strengthen associations between personal drinking motives and drinking among SMW has not been examined.

The current examination was a follow-up analysis of a study that examined how minority stress affected alcohol use in SMW's daily lives (Lewis et al., 2021). It used a daily diary design to assess both members of same-sex female couples, allowing the current examination to assess how partner influences (e.g., drinking intentions, physical presence, drinking behaviors), as well as individual daily drinking motives, affect consumption among same-sex female couples. The current study had several hypotheses. We expected that higher own intentions to drink would be associated with more drinking on a given day over and above partner influence (Hypothesis [H] 1a). In addition, we examined which types of partner influence (intentions, presence, drinking) were uniquely associated with drinking after controlling for the other types of partner influence (H1b). Further, we examined which types of partner influence moderated the associations between own intentions and drinking, controlling for other types of influence (H2). We expected that the associations between participant intentions and own drinking would be strengthened if partner intentions were higher, if the partner was present, and if the partner was drinking. Finally, we expected that participants' own coping, social, and enhancement motives would moderate the associations between partner influences and the number of drinks consumed by the individual (H3), such that partner influences would be stronger when participants were drinking for social or enhancement reasons but weaker when participants were drinking for coping reasons.

Method

Participants

Young adult same-sex female couples were recruited. Inclusion criteria for all individuals included (a) ages 18–35 years, (b) cisgender woman, (c) in a relationship with a woman for at least 3 months, (d) in-person contact with their partner at least once a week, (e) partner is willing to also participate in the study, and (f) willing and able to complete daily surveys between 6 A.M. and 12 noon for 14 days. In addition, at least one member of the couple had to meet the following inclusion criteria: (a) reporting being only or mostly attracted to women; (b) consuming alcohol on at least 3 days during the past 2 weeks; and (c) reporting consuming at least four drinks in one sitting in the past 2 weeks. The final sample consisted of 321 SMW. Participants had an average age of 27.57 (SD = 3.65) and were mostly White (71.5%; see Table 1 for more demographic information). The average survey compliance rate was 89.9% (12/14 surveys). For more detailed information on the study protocol of the 163 same-sex couples who were recruited for participation in Project Relate, see Heron et al. (2019).

Table 1.

Demographic characteristics from Project Relate

Continuous variables	M (SD)
Age	27.57 (3.65)
Relationship duration	40.99 (31.24)

Categorical variables	n (%)
Education level
Some high school	3 (0.9)
High school graduate	20 (6.1)
Some college	48 (14.7)
Associate's degree	16 (4.9)
Bachelor's degree	156 (47.9)
Master's degree	67 (20.6)
Doctoral/professional degree	16 (4.9)
Ethnicity
Not Hispanic, Latina or Spanish origin	288 (88.3)
Hispanic, Latina or Spanish origin	38 (11.7)
Race
African American or Black alone	28 (8.6)
American Indian or Alaskan Native alone	2 (0.6)
Asian, Asian American, Native Hawaiian or Pacific Islander alone	19 (5.8)
European American, Caucasian, or White alone	233 (71.5)
Multiracial	34 (10.4)
Other	10 (3.1)
Sexual identity
Lesbian	274 (84.0)
Bisexual	67 (20.6)
Queer	122 (37.4)
Gay	98 (30.1)
Pansexual	17 (5.2)
Asexual	2 (0.6)
Questioning	1 (0.3)
Other	2 (0.6)
Relationship status
Single, exclusively dating one person	5 (1.6)
Partnered, in a relationship that is not exclusive	17 (5.3)
Partnered, in an exclusive relationship	210 (65.4)
Partnered, married or in a civil union	83 (25.9)
Other	6 (1.9)
Cohabitation status
Yes	246 (75.7)
No	79 (24.3)

Notes: For sexual identity, percentages do not sum to 100% as participants could select more than one response option. Relationship duration is in months.

Recruitment and procedures

The research team (Heron et al., 2019; Lewis et al., 2021) worked with a market research firm that specializes in web-based research with SM individuals. The market research firm recruited individuals and conducted an initial screening for who was likely to qualify using less stringent criteria. The research team conducted a second screening to ensure that exact eligibility criteria were met. After both members of the couple screened eligible and provided informed consent, they were emailed a link to the baseline survey (i.e., demographic questions and self-report measures). After both members of the couple completed the baseline survey, participants began receiving their daily surveys via email for 14 days, including items about their cognitions, beliefs, and behaviors for the previous day, as well as their drinking intentions for the current day.

Measures

Alcohol use. Each individual was asked, (a) “Did you drink alcohol yesterday?” (1 = yes, 0 = no); (b) “Did your partner drink yesterday?” (1 = yes, 0 = no); and (c) “How many standard drinks did you consume yesterday?” (1–20+ standard drinks consumed). Participants were asked about the number of drinks they had consumed only if they endorsed drinking the previous day.

Drinking intentions. All individuals were asked “What is the likelihood that you will drink in the next 24 hours?” Participants selected a number from 0 to 100 on a sliding visual analog scale, with 0 reflecting that the participant definitely does not intend to drink and 100 indicating that they definitely do intend to drink. This question was asked every day, regardless of the previous day's drinking.

Partner drinking context. Multiple items assessed partner influence. On days participants reported that they drank, they were asked, “Who was present when you were drinking?” and if they selected “partner/spouse,” this was coded as 1 = partner present or 0 = partner not present. If they indicated that their partner was present, they were then asked, “Was your partner …” 1 = present and drinking or 0 = present and not drinking.

Daily drinking motives. Modified items from the Drinking Motives Questionnaire (DMQ; Cooper, 1994) assessed individual motives for daily drinking (only when participants endorsed drinking the previous day). The original DMQ consists of 20 items and four subscales: social, enhancement, coping, and conformity. To ease participant burden for daily assessment, text from the original five items for each subscale was combined to create one item each to measure social, enhancement, and conformity motives. Four items were used to assess coping motives. Three of the original five items for the coping subscale were unmodified; text for the remaining two items—“to forget your worries” and “to forget your problems”—were combined into a single item (i.e., “to forget your worries and problems”). The response scale was also modified slightly for daily assessment, with participants rating their agreement for each statement from 1 = not at all to 5 = definitely (rather than the original response scale from 1 = almost never/never to 5 = almost always/always). We recoded all motives into dichotomous versions where 1 = not at all was recoded as 0 = no, and anything higher than that (2–5) was recoded as 1 = yes because of nonnormality.

Demographics. Demographic questions asked about age, education status, ethnicity, racial identity, relationship status, relationship duration (in months), and cohabitation status (Table 1).

Analysis approach

Three-level multilevel modeling was used given the nested structure of the data (i.e., repeated daily diary reports [Level 1], within partners [Level 2], within couples [Level 3]). A latent variable decomposition approach was used (via multilevel structural equation modeling), reflecting that each day's values are an imperfect indicator of typical behavior, particularly among behaviors with low intraclass correlations (i.e., varies substantially from day to day, like alcohol use; see Preacher et al., 2010). All models included a random intercept and random slopes. Level-1 (within-person) predictors reflected daily reports, and Level-2 predictors were the latent person-level aggregate. By including a Level-2 equivalent of the Level-1 predictor, we examined within-person effects and between-person effects separately using latent variable decomposition. Means and variances were estimated at Level 3 to account for couple-level nesting.

Given that reports of drinking (amount, motives, partner presence, etc.) were retrospective for the previous day, but intentions to drink were for the current day, the intentions variables were lagged by one day to match the timeframe for drinking (e.g., intentions reported on study day one relate to drinks consumed on study day one, reported on study day two). Also, whereas partner presence was a perception reported by the participant, drinking and intentions were directly reported by both members of the couple. Partner responses were read into each participant's line of data. For example, if ID1 and ID2 are partners, ID1's intentions and whether they drank were read into ID2's data as partner intentions and partner drinking. ID2's intentions and whether they drank were read into ID1's data as partner intentions and partner drinking.

To examine H1a (own intentions predicting own drinking, controlling for partner influence) and H1b (unique partner influence on own drinking, controlling for other forms of partner influence), all aspects of partner influence were included in one comprehensive model. Similarly, one comprehensive model was conducted with all forms of partner influence present in the same model to examine if partner influence moderated the association between own intentions and own drinking (H2). To examine if own drinking motives moderate the impact of partner influence controlling for own intentions (H3), three models were conducted with own intentions, partner influence (separate models for each type), and their interactions with motives as the predictors, and own drinking as the outcome variable. The outcome for all models was the number of drinks consumed by the participant.

For all significant interactions, simple slopes were calculated to examine the effect of the predictor of interest across each level of the moderator (e.g., influence of own intentions when a partner was present vs. not). Analyses were conducted using Mplus 8.4 (Muthén & Muthén, 1998–2017). Maximum likelihood estimation with robust standard errors and two-tailed tests were used.

Results

Hypothesis 1: Unique effects of own intentions (Hypothesis 1a) and partner influence (Hypothesis 1b) on alcohol consumption

As seen in Table 2, when examining H1a, the participant's own intentions were still a significant predictor of greater same-day drinking (within subject) but not at the person level (between subject), controlling for all forms of partner influence. Examining H1b, greater partner intentions and greater partner drinking both uniquely predicted greater same-day drinking by the participant (within-subject) over and above other forms of partner influence, but not at the person level. Partner presence was not a significant predictor at either level.

Table 2.

Partner effects on number of drinks consumed on drinking days

Variables	Within-subjects				Between-subjects
	B	SE	p	[95% CI]	B	SE	p	[95% CI]
H1a and H1b: Partner influences
Own intentions	0.023	0.002	<.001	[0.020, 0.027]	0.015	0.013	.246	[-0.010, 0.041]
Partner intentions	0.008	0.002	<.001	[0.005, 0.012]	0.004	0.012	.738	[-0.019, 0.027]
Partner presence	0.167	0.144	.247	[-0.115, 0.449]	-7.807	116.545	.947	[-236.235, 220.620]
Partner drinking	0.364	0.128	.005	[0.112, 0.615]	-2.488	3.005	.408	[-8.378, 3.403]
H2: Own Intentions × Partner Influences
Own intentions	0.256	0.041	<.001	[0.175, 0.337]	0.343	0.286	.230	[-0.217, 0.903]
Partner intentions	0.057	0.017	.001	[0.023, 0.091]	0.058	0.129	.651	[-0.195, 0.312]
Partner presence	0.173	0.117	.140	[-0.057, 0.404]	-3.680	33.053	.911	[-68.463, 61.103]
Partner drinking	0.385	0.133	.004	[0.125, 0.646]	-2.677	1.944	.168	[-6.486, 1.132]
Own × Partner Intentions	0.016	0.006	.010	0.004, 0.029]	0.128	0.393	.745	[-0.643, 0.898]
Own Intentions × Partner Presence	-0.021	0.037	.567	[-0.094, 0.052]	-0.426	0.383	.266	[-1.176, 0.324]
Own Intentions × Partner Drinking	-0.031	0.043	.474	[-0.116, 0.054]	-0.191	0.450	.672	[-0.692, 1.073]
H3: Own Motives × Partner Intentions
Own intentions	0.189	0.018	<.001	[0.154, 0.225]	0.092	0.119	.438	[-0.141, 0.325]
Partner intentions	-0.026	0.036	.476	[-0.096, 0.045]	-0.073	0.194	.708	[-0.453, 0.307]
Own social motives	0.923	0.112	<.001	[0.704, 1.143]	-0.587	0.622	.346	[-1.806, 0.633]
Own enhance motives	0.812	0.113	<.001	[0.589, 1.034]	0.848	0.335	.011	[0.191, 1.506]
Own conform motives	0.252	0.273	.356	[-0.284, 0.788]	-0.078	0.819	.924	[-1.683, 1.527]
Own cope motives	0.222	0.090	.013	[0.047, 0.398]	-0.054	0.344	.876	[-0.728, 0.621]
Social Motives × Partner Intentions	0.075	0.031	.017	[0.014, 0.136]	0.683	0.450	.129	[-0.199, 1.564]
Enhance Motives × Partner Intentions	0.055	0.036	.120	[-0.014, 0.125]	-0.215	0.261	.410	[-0.726, 0.296]
Conform Motives × Partner Intentions	0.051	0.073	.485	[-0.093, 0.195]	0.234	0.240	.330	[-0.237, 0.704]
Cope Motives × Partner Intentions	-0.003	0.029	.910	[-0.059, 0.053]	0.014	0.176	.934	[-0.331, 0.360]
H3: Own Motives × Partner Presence
Own intentions	0.217	0.019	<.001	[0.180, 0.254]	0.068	0.091	.454	[-0.110, 0.246]
Partner presence	0.327	0.176	.063	[-0.018, 0.672]	-1.712	23.820	.943	[-48.400, 44.975]
Own social motives	0.968	0.141	<.001	[0.692, 1.245]	0.819	1.625	.614	[-2.366, 4.004]
Own enhance motives	0.984	0.175	<.001	[0.641, 1.328]	1.916	0.867	.027	[0.216, 3.615]
Own conform motives	0.130	0.330	.693	[-0.516, 0.776]	0.325	2.190	.882	[-3.969, 4.618]
Own cope motives	0.291	0.165	.077	[-0.032, 0.614]	-1.053	0.878	.231	[-2.774, 0.668]
Social Motives × Partner Presence	0.024	0.158	.881	[-0.286, 0.334]	-1.997	2.709	.461	[-7.306, 3.313]
Enhance Motives × Partner Presence	-0.166	0.179	.354	[-0.517, 0.185]	-2.198	1.422	.122	[-4.985, 0.590]
Conform Motives × Partner Presence	0.298	0.389	.443	[-0.463, 1.060]	-0.393	2.957	.894	[-6.189, 5.404]
Cope Motives × Partner Presence	-0.088	0.181	.627	[-0.443, 0.267]	1.461	1.006	.146	[-0.511, 3.432]
H3: Own Motives × Partner Drinking
Own intentions	0.021	0.002	<.001	[0.017, 0.025]	0.006	0.009	.501	[-0.012, 0.024]
Partner drinking	-0.097	0.196	.620	[-0.482, 0.287]	-2.244	1.674	.180	[-5.526, 1.037]
Own social motives	0.777	0.181	<.001	[0.422, 1.131]	-1.504	1.152	.192	[-3.762, 0.755]
Own enhance motives	0.532	0.173	.002	[0.193, 0.871]	0.987	0.728	.175	[-0.439, 2.413]
Own conform motives	0.552	0.402	.169	[-0.235, 1.340]	0.691	1.616	.669	[-2.476, 3.859]
Own cope motives	0.151	0.164	.358	[-0.171, 0.473]	-0.423	0.696	.544	[-1.787, 0.942]
Social Motives × Partner Drink	0.253	0.212	.233	[-0.162, 0.669]	1.810	1.350	.180	[-0.835, 4.456]
Enhance Motives × Partner Drink	0.437	0.213	.040	[0.019, 0.855]	-0.569	1.051	.588	[-2.629, 1.491]
Conform Motives × Partner Drink	-0.295	0.454	.516	[-1.184, 0.595]	-0.490	2.284	.830	[-4.966, 3.987]
Cope Motives × Partner Drink	0.083	0.176	.636	[-0.262, 0.428]	0.596	0.809	.461	[-0.990, 2.182]

Notes: Significant values are bold. SE = standard error from models using robust maximum likelihood estimation; H = hypothesis; Social = social drinking motives; Enhance = enhancement drinking motives; Conform = conformity drinking motives; Cope = coping drinking motives.

Hypothesis 2: Moderating effects of own intentions and partner influence on alcohol consumption

Examining H2, there was a significant interaction between own intentions and partner intentions at the within-subject level only but not the between-subject level, such that partner intentions strengthened the association between higher own intentions and greater same-day drinking by the participant. Simple slopes revealed that when partner intentions were low (B = -0.315, SE = 0.213, p = .140, 95% CI [-0.734, 0.103]), own intentions were not associated with same-day own drinking. When partner intentions were average (B = 0.256, SE = 0.041, p < .001, 95% CI [0.175, 0.337]) or high (B = 0.827, SE = 0.235, p < .001, 95% CI [0.366, 1.288]), own intentions were significantly associated with greater drinking levels by the participant, but this was much stronger when partner intentions were high. There were no significant interactions between participant's own intentions and partner presence or partner drinking on their own drinking at either level (Table 2).

Hypothesis 3: Drinking motives, partner influence, and alcohol consumption

As seen in Table 2, greater own social, enhancement, and coping drinking motives were linked to the participant consuming more drinks the same day when controlling for own intentions and partner influences. There was a significant within-subject interaction between partner intentions and social motives. However, simple slopes revealed that regardless of whether the participant endorsed social motives (B = 0.049, SE = 0.037, p = .186, 95% CI [-0.024, 0.122]) or not (B = -0.026, SE = 0.036, p = .476, 95% CI [-0.096, 0.045]), the association between partner intentions and same-day own drinking was not significant. There were no significant interactions between partner intentions and any other motives at either level. There were no significant interactions between any motives and partner presence at either level. For partner drinking, enhancement motives significantly moderated the association between partner drinking and number of drinks consumed by the participant. Simple slopes revealed that the positive association between partner drinking and own number of drinks consumed was only significant when drinking for enhancement reasons (B = 0.340, SE = 0.173, p = .049, 95% CI [0.002, 0.678]) versus not (B = -0.097, SE = 0.196, p = .620, 95% CI [-0.482, 0.287]).

Discussion

The current study examined (a) whether own drinking intentions were related to daily drinking by the participant over and above partner influences (i.e., presence, drinking [yes/no], intentions; H1a), (b) if partner influences were uniquely related to daily drinking by the participant, controlling for other forms of influence (H1b), (c) if these same partner influences moderated the impact of own drinking intentions (H2), and (d) if own drinking motives moderated the impact of partner influences among SMW couples (H3). We found that own intentions, partner intentions, and partner drinking were associated with greater same-day drinking by the participant, suggesting that both own intentions and partner influences are meaningful. In a comprehensive model that controlled for all types of partner influence (such as if they were present and if they drank) and their interactions with own intentions, partner intentions significantly moderated own intentions such that the effect of own intentions on consuming more drinks that day was stronger if the partner also had greater drinking intentions, indicating a synergistic effect. Enhancement motives significantly moderated the association between partner drinking and drinks consumed the same day such that partner drinking only affected own drinking if the participant was drinking for enhancement reasons. Although the association between partner intentions and own drinking was significantly stronger when drinking for social reasons versus not, the association was not significant on either type of day. Notably, nearly all main effect associations and interactions were only significant at the within-subjects level. Taken together, this suggests that although own intentions are a powerful predictor of drinking, there are select drinking contexts in which SMW are more susceptible to partner influence.

Given that drinking intentions are a robust predictor of drinking (e.g., Rimal & Mollen, 2013; Testa et al., 2009), our findings that own drinking intentions are linked to drinking on a particular day and the number of drinks consumed are not surprising. However, it is a novel finding that same-sex partner's drinking intentions are linked to these drinking outcomes over and above own intentions. SMW in same-sex marriages are more resistant to partner influence on drinking levels than women in different-sex marriages (Pollitt et al., 2020). As intentions are theoretically considered a precursor to behavior (Ajzen, 1991; Fishbein & Ajzen, 1975), it could be that SMW are resistant to influence about how much others are drinking but are susceptible to cues about whether someone is drinking or has plans to drink. However, when controlling for interactions between all sources of partner influence and own intentions, only partner intentions significantly moderated the association between own intentions and same-day drinks consumed. This finding suggests that partner intentions may be a stronger source of influence on own drinking among SMW. Further, because these associations were only significant at the within-subjects level, this suggests that daily fluctuations in drinking intentions among individuals in same-sex relationships have stronger effects on their drinking than average person-level intentions and drinking across couples, underscoring the role of daily context in drinking. Prior research by Pollitt et al. (2020) only examined within-subject associations; therefore, the current study builds on prior research by not only expanding the assessment of types of partner influence and moderation by drinking motives but also by revealing that only daily context and not person-level patterns are meaningful.

Similarly, our findings that social, enhancement, and coping drinking motives are linked to greater same-day drinking over and above own intentions and partner influences are in line with prior research in other populations (e.g., college students, Bresin & Mekawi, 2021) as well as among SMW (Dworkin et al., 2018; Kelley et al., 2018). Our findings that higher daily enhancement motives strengthen the association between partner drinking and own drinking are in line with prior research suggesting it may impact the effects of social influence on drinking among SMW (Dworkin et al., 2018), which also found only within-subject associations. Finding only significant within-subject associations suggests that in-the-moment context of drinking among SMW in relationships matters more than typical motives and drinking across couples. Dworkin et al. (2018) found that drinking motives interact with who one drinks with based on sexual identity to affect the amount consumed. We extend these findings to reveal that own motives also have an impact on the influence of whether their partner is also drinking on a drinking day (or in other words, partner influence may only be relevant when SMW are drinking for enhancement reasons).

The current study had several limitations and strengths. The daily surveys were retrospective (except for intentions), introducing the possibility of recall bias. However, because they were sent in the early morning, this risk was minimized. Drinking intentions asked about the next 24 hours, which could have, in theory, included next-day drinking rather than drinking later that day. There were also recruitment challenges for recruiting a diverse sample (for a detailed discussion, see Lewis et al., 2021), which affected the generalizability of our findings to all SMW in relationships. Researchers may wish to oversample SMW who identify with racial and ethnic identities other than White to better understand how partner influences among SMW with intersecting identities affect drinking at the daily level. The current study assessed both members of same-sex female couples, allowing for self-reporting of partner influences (e.g., drinking intentions, physical presence, drinking behaviors). One person reporting on both themself and their partner is common (e.g., Lewis et al., 2018) but could incorporate misperceptions of partner cognitions and behavior. For example, members of the couple were possibly aware of each other's drinking intentions from verbal statements (e.g., “I want to get drunk tonight”), or possibly not if no comments were made. However, the current approach allowed both members of the couple to report their own cognitions and behaviors, which means partner effects in the current study are based on the partners' thoughts and actions, not the other partner's knowledge of them.

Given the current findings that partner qualities (intentions, presence, drinking) affect own drinking among SMW, a couple-focused intervention could have potentially powerful effects on drinking. Strong evidence suggests that couple-focused interventions are effective at reducing alcohol use and encouraging abstinence among different-sex couples (see Powers et al., 2008, for a meta-analysis). This approach may be effective at curbing drinking among SMW, a high-risk group for heavy drinking. However, such an intervention should be tailored to the unique stressors and challenges SMW face (e.g., Braitman et al., 2021) and the unique dynamics of SMW in relationships when it comes to influencing health behavior engagement (including alcohol use; Holway et al., 2018; Reczek & Umberson, 2012). This intervention may wish to highlight how partners may be emotionally close even if not drinking together. It may also be beneficial to discuss the relevance of drinking motives and how individuals may be particularly susceptible to partner influence under select conditions. An appropriately tailored couple-focused alcohol intervention for SMW could be beneficial in shifting alcohol-related cognitions (e.g., intentions, motives) to reduce problematic drinking for both partners.

Conclusions

The present study found that intentions, drinking motives, and several partner influences (i.e., intentions, drinking) are linked to more drinks consumed at the daily level among SMW in relationships. We also found several interactions between own intentions and partner influences, which are linked to increased drinking. Namely, the impact of own intentions to drink on the number of drinks consumed was strengthened when partner intentions were also high (i.e., synergistic effects). Further, enhancement motives strengthen the association between partner drinking and drinks consumed at the daily level, suggesting that partner influence may have a stronger impact when SMW are drinking for specific reasons. Despite prior findings that drinking levels of SMW may be more resistant to partner influence (via how much the partner consumes), the current findings suggest that SMW are not immune to partner influence in the form of partner intentions, and if the partner is present and drinking, especially if they are drinking for social or enhancement reasons.