Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 1.
Published in final edited form as: Psychol Addict Behav. 2021 Apr 1;36(1):44–53. doi: 10.1037/adb0000698

Romantic relationship status, stress and maturing out of problematic drinking

Stephen Armeli a, Hannah R Hamilton b, Constance Hammen c, Howard Tennen d,*
PMCID: PMC8484339  NIHMSID: NIHMS1648631  PMID: 33793278

Abstract

Objective:

The current study examined the unique influences of romantic relationship status and episodic and chronic stress associated with relationships in predicting changes in alcohol consumption and drinking motivations from college to post-college life.

Method:

Moderate to heavy college student drinkers reported their drinking level and drinking motives using an Internet-based daily diary for 30 days in college and again five years later. At the post-college wave, participants also completed a semi-structured phone-based interview to assess romantic relationship stress.

Results:

Multiple regression analysis indicated that chronic relationship stress and relationship dissolution stress were uniquely related to mean daily levels of post-college drinking to cope (DTC) motivation, but not to mean daily levels of enhancement motivation. Some evidence was found for the effect of relationship status, but not stress, on mean daily heavy drinking levels. We also found evidence that chronic stress moderated the effect of relationship dissolution stress, with individuals showing higher mean daily chronic stress levels displaying a stronger positive association between relationship dissolution and post-college DTC motivation.

Conclusions:

Results are discussed in term of how negative reinforcement processes might be an important mechanism underlying commonly found associations between romantic relationship status and problematic drinking during young adulthood.

Keywords: Drinking to cope motivation, Romantic Relationships, Stress, Maturing out

Alcohol use disorder (AUD) symptoms and heavy episodic drinking (5 or more drinks on the same occasion) peak between the ages of 18 and 25 and then steadily decrease for most individuals (Jackson & Sartor, 2016). However, some individuals fail to “mature out” of risky drinking and are at greater risk for developing AUDs (Jochman & Fromme, 2010). Multiple mechanisms have been posited to underlie the pace at which individuals mature out of problematic drinking, if at all. First, role incompatibility theory (Yamaguchi, 1990) states that individuals either choose social roles that allow them to maintain heavy drinking patterns, or experience demands and social influences upon entering adult social roles that result in reduced drinking (Jochman & Fromme, 2010). Second, evidence suggests a key role of life stress in that continued problematic drinking throughout young adulthood is more strongly related to drinking for coping reasons than for other reasons (Littlefield et al., 2010). In the present study, we apply these two theoretical frameworks in examining how romantic relationship status and stress are related to changes in drinking motivations and drinking level during the critical transition from college to post college life (Arria et al., 2016; Gotham et al., 1997). We focused on romantic relationships given evidence of their role in maturing out of problematic drinking (Jochman & Fromme, 2010; Staff et al., 2010) and because it is an especially deleterious source of stress for emerging adults (e.g., Vrshek-Schallhorn et al., 2015).

Romantic relationship status and alcohol use

Considerable evidence shows that romantic relationship status is associated with alcohol use and related problems, especially during the critical period of young adulthood. Findings indicate that getting married – especially for individuals in their twenties – is a robust correlate of reduced alcohol use (Bachman et al., 1997; Bennett et al., 1999; Gotham et al., 1997; Jochman & Fromme, 2010). Studies have also shown that decreased drinking might precede marriage, possibly due to what can be termed an “engagement effect” or simply being in a committed relationship (Bachman et al., 1997; Gotham et al., 1997). Consistent with this notion, Fleming et al. (2010) found that after controlling for adolescent drinking levels, young adults who were married, cohabitating, or in non-cohabitating dating relationships showed lower drinking levels compared to individuals not in dating relationships. These protective effects for cohabitation are at odds with previous findings (e.g., Horwitz & White, 1998) and might represent a historical shift in that cohabitation has become more normative and less associated with risky behavior.

Fleming et al. (2018) extended this line of research by repeatedly surveying young adults (ages 18-24) monthly for twelve months to examine how changes in relationship status correlated with alcohol use. They found that individuals who were single but not actively dating and individuals in stable relationships (e.g., married or cohabitating) reported lower drinking levels during this period compared to individuals who were single but dating, or who had ended a relationship, or who ended and then started a new relationship. Examination of monthly changes for individuals who reported relationship dissolution also indicated that drinking was higher post-breakup compared to pre-breakup. The authors posited that relationship stability – whether being single or in a committed relationship – rather than being in a relationship per se, seemed to be an important correlate of drinking level.

Romantic relationship stress and alcohol use

The effects of romantic relationship status instability on drinking level could be due, in part, to negative reinforcement processes – i.e., drinking to cope (DTC) – associated with such situations. Stress emanating from relationship instability might be especially relevant during the important developmental phases of emerging and early adulthood. Theories have long emphasized the importance of developing healthy intimate relationships for emerging adults (Erikson, 1968; Hazan & Zeifman, 1994). Consistent with this notion, evidence indicates that interpersonal forms of stress, especially romantic relationship stress, are stronger predictors of depression among emergent adults compared to non-interpersonal stressors (Vrshek-Schallhorn et al., 2015).

Findings from longitudinal studies also highlight the important role that life stress might play in maturing out of problem drinking among young adults. For example, Littlefield et al.’s (2010) long-term longitudinal study of college student drinkers (ages 18 to 35) showed that changes in DTC motivation, but not enhancement drinking motivation, correlated with changes in drinking-related problems. Other studies have also found that problematic drinking among young adults was more strongly associated with DTC motivation compared to other reasons for drinking (Perkins, 1999; Patrick & Schulenberg, 2011). Although coping-related drinking might arise from a variety of antecedent factors, interpersonal stress is considered to be a key correlate of this reason for drinking, especially among young adults (Lambe et al., 2015; Levitt & Leonard, 2015). Indeed, Carrigan et al. (1998) found that among young adults, drinking associated with interpersonal conflict explained over a third of the variance in overall DTC motivation levels. More specific to romantic relationships, Patrick et al. (2018) found that young adults reported higher levels of DTC motivation in months when they ended romantic relationships.

Stress associated with relationship dissolution is not the only possible source for coping-related drinking related to romantic relationships. Individuals might also drink to cope with dissatisfaction with their current romantic relationship status; for example, being single, but wanting to be in a relationship. This could partly explain higher levels of drinking found among single but actively dating individuals compared to individuals who are single and not dating (e.g., Fleming et al., 2018), i.e., such individuals might be seeking relationships, but are not satisfied with their current prospects. Furthermore, relationships characterized by high levels of conflict could also elicit drinking for coping reasons. Indeed, Collibee and Furman’s (2015) longitudinal study found an interactive effect of age and romantic relationship satisfaction in predicting substance use. Specifically, relationship satisfaction was unrelated to substance use when participants were adolescents (high school years), but it was significantly negatively related to substance use when they were young adults in their early twenties. This finding is also consistent with other research (e.g., Littlefield et al., 2010; Patrick & Schulenberg, 2011) suggesting that stress-related factors become more closely linked to alcohol use as individuals progress through emerging adulthood.

The present study

Research indicates that both romantic relationship status and stress are important and inter-related factors that are associated with maturing out of problematic drinking among young adults (Fleming et al., 2018; Patrick et al., 2018). In addition to role selection and socialization processes associated with changes in romantic relationship role status, stress emanating from such changes, and other factors associated with romantic relationships, such as major arguments or dissatisfaction with current status, might contribute to problematic drinking through negative reinforcement processes (i.e., DTC-motivated drinking). Few studies, however, have simultaneously examined these processes in predicting problematic drinking during this crucial transition period. In the current study, we examined romantic relationship status and recent life stress associated with romantic relationships as predictors of drinking levels and DTC motivation among young adults transitioning from college to post-college life.

We furthered this line of research in several ways. Notably, we assessed persisting stress associated with romantic relationships in general (i.e., chronic stress) and the stressfulness of recent discrete events (i.e., episodic stress) in this social role (e.g., relationship dissolution, major argument) using a semi-structured interviewer-based protocol. We modeled our episodic stress assessment after the Brown and Harris (1978) “contextual threat” method, and used a parallel approach to assess ongoing chronic stress. Interviews that include contextual information are considered superior to other approaches for stress assessment, such as event checklists or brief appraisal measures (Dohrenwend, 2006; Monroe, 2008), and allowed us to incorporate into our models non-self-report (i.e., rater-based) assessments of the severity (i.e., stressfulness) of discrete events such as relationship dissolution as well as other stressors. Previous studies examining role changes such as relationship dissolution have generally treated them as homogeneous events that occurred or not, possibly attenuating observed associations with outcomes. For example, Patrick et al. (2018) found that romantic relationship dissolution was related to increased DTC motivation, but they did not evaluate the context – and stressfulness – of relationship status changes. It would be expected that mutually agreed upon and expected dissolutions of a short-term relationships (e.g., low stress scenarios) would have less of an effect on DTC-motivated drinking compared to unwanted and unexpected dissolutions of long-term relationships (i.e., high stress scenarios). More generally, we predicted that romantic relationship-related episodic events (e.g., relationship dissolution or other stressors such as major arguments) rated by the interviewers as more stressful will be associated with higher levels of post-college heavy drinking and DTC motivation controlling for levels in college.

In addition, our assessment of overall romantic relationship chronic stress allowed us to further explore the notion that role instability (e.g., Fleming et al., 2018), regardless of whether one is in a relationship, is also an important determinant of problematic drinking. Specifically, we operationalized chronic stress for individuals not in relationships as the degree to which they were dissatisfied with their current status and were actively seeking partners; for individuals in relationships, chronic stress was indicated by higher levels of conflict and lower levels of support and communication from their partners. We predicted that higher levels of chronic stress would be associated with higher levels of post-college heavy drinking and DTC motivation controlling for levels in college.

Simultaneous assessment of romantic relationship status, episodic stress, and chronic stress also allowed us to evaluate their unique and possibly interactive effects on changes in DTC motivation. One possibility is that the association between the stressfulness of recent episodic events and post-college heavy drinking and DTC motivation will be stronger in the positive direction for individuals with greater levels of ongoing romantic relationship chronic stress.

Finally, to examine if our effects are unique to DTC motivation, we also examined whether the effects of romantic relationship status and stress showed parallel effects on drinking to enhance positive emotions. Patrick et al. (2018) provided some evidence for this possibility in that individuals who more often reported instances of romantic relationship dissolution showed higher overall levels of enhancement motivation.

Methods

Participants and Procedure

Participants were moderate to heavy drinking college students who were originally recruited for a larger study of college student alcohol use, daily experiences, and health-related behavior (see Armeli et al., 2014). Prospective participants for the parent study were recruited across nine semesters through the psychology research pool, as well as campus-wide broadcast emails and flyers at large eastern state university. Only students who reported drinking alcohol at least twice in the past month and had not received treatment for alcohol use (measured during prescreening) were eligible (see Armeli et al., 2018 for additional details about the parent study). For the longitudinal (5-year follow up) component of the study, only participants (N = 1141) who at wave 1 reported at least one heavy drinking day (≥ 4 drinks for women and ≥ 5 drinks for men) in both a 30-day retrospective assessment and a 30-day daily diary reporting phase (each assessment covering two different time points in college) were eligible. Given the focus on changes in drinking from college life to post-college life, participants also needed to have graduated or to no longer be working towards an undergraduate degree (see Armeli et al., 2018 for additional details about wave 2 recruitment). Prospective participants were recruited by phone. The institutional review board approved all procedures.

At both time points, participants first logged on to a secure website where they provided informed consent and completed a baseline survey including questions about romantic relationship status. Approximately two weeks later, participants completed the 30-day daily diary portion of the study in which alcohol use and drinking motives were assessed. In Wave 1, daily surveys were completed each day between 2:30 pm and 7:00 pm; in Wave 2, daily surveys were completed each day between 4:00 pm and 8:30 pm. These time windows were selected to coincide with most participants’ naturally occurring end of school/work day but before typical evening activities begin (including drinking). In addition, at Wave 2 we assessed life stress in the past year for all participants using the UCLA Life Stress Interview (LSI; Hammen, 1991; Hammen et al., 1989), which was conducted by phone; completion times for the interview ranged from 45 to 90 minutes. The LSI was completed once the baseline survey was done, but several weeks before the daily diary phase.

We obtained complete data (i.e., baseline survey, daily diary [meeting minimum daily adherence of 15 days] and UCLA LSI data) from 898 participants (78.7% of the eligible participants who completed wave 1 also completed wave 2). In the final sample, daily diary completion adherence was high in the college (85%) and post-college (93%) assessments. The final sample was 54.5% women, predominantly Caucasian (86.6%) with a mean age of 19.2 years (SD = 1.3) during college and 24.6 years (SD = 1.3) post-college. At follow up, the majority were employed full-time (78.4%), 97.6% had completed their bachelor’s degree, and 42.3% were enrolled in a graduate program; only 1.2% had children. We compared our final sample to the remainder of the initial 1141 participants contacted for Wave 2. Our final sample was not different in terms of age, t(1, 1139) = −1.25, p = .23), but did differ in terms of sex, χ2(1) = 18.50, p < .001, and ethnicity (White vs. others), χ2(1) = 11.42, p = .001. Individuals not included in our analyses were more likely to be men (61% of excluded vs. 46% of completers) and minorities (22% of excluded vs. 13% of completers).

Measures

Romantic relationship chronic and episodic stress.

We used the UCLA Life Stress Interview (UCLA LSI) developed by Hammen and colleagues (Hammen, 1991; Hammen et al., 1989) to measure romantic relationship chronic and episodic stress. The LSI interview obtains information about the details and context of the person’s life, attempting to evaluate circumstances that permit assignment of the objective stressfulness of life events and chronic ongoing conditions in life roles (e.g., Hammen, 2016).

Interviewers were doctoral-level clinical psychology graduate students and a masters level research associate with experience in diagnostic interviewing, who were trained to evaluate chronic stress during the past year in a variety of life domains, including romantic relationships, and to query participants about discrete events that occurred in a variety of life domains. For the romantic relationship domain, participants were first asked if they were currently in a romantic relationship (yes = have steady romantic partner; no = not in relationship or in brief non-committed relationship[s]). For participants in relationships, interviewers probed for stability, mutual supportiveness, and conflict resolution. Interviewers then rated chronic stress using the following 5-point scale (using half-point increments): 1 (exceptional relationship: high stability, supportive, mutual, good conflict resolution ability), 2 (stable positive relationship: adequately resolve conflicts, close, confiding, trusting, reciprocal), 3 (relationship has some significant problems, e.g. lacking greatly on 1 or 2 “quality” factors, but basic strong foundation is present), 4 (deteriorating relationship or severe problems, e.g. lacking greatly on 3 or more “quality” factors, more negative than positive, unstable, poor conflict resolution) and 5 (abusive relationship: physically and/or emotionally, negative conditions, lack of communication and/or one-sided relationship).

For individuals not in relationships, interviewers probed for the possibilities to meet people; how lonely or content they were without a partner; whether they were actively seeking a partner; and if dating, how often and how many different people. Ratings were made on the following 5-point scale: 1 ([not dating]: completely satisfied without partner, happy and not lonely and not looking for a partner; has other life plans for now; adequate social life; [dating]: frequent dating, perceive partner[s] to be excellent potential for future relationship, excellent qualities), 2 ([not dating]: happy at present, not lonely, but would like to have a partner in the future, not having a partner does not cause concern; possibilities exist if desire; [dating]: some dating, some good prospects, but not certain), 3 ([not dating]: somewhat happy about not having a partner; looking for someone and spends time thinking about how to find someone, limited possibilities and occasional distress about not having a partner, not preoccupied, has friends who are singles; [dating]: some dating, partner[s] not appropriate or ideal), 4 ([not dating]: unhappy about not having someone, causes distress more often than not, but not constantly concerned about not having a partner, limited prospects, most friends are in relationships; [dating]: potential partners disappointing, or infrequent, concern about limited opportunities), and 5 ([not dating]: extremely unhappy and lonely without a partner, ongoing concern about never having a partner, no attempts or unsuccessful attempts to find other; friends all have relationships or experiences pressure from others; [dating]: extremely adverse experiences, mistreated or highly inappropriate dates). To establish reliability, we selected a random sample of 98 participants (~10% of the sample) and had a second rater read the interviewer’s description of the interview and rate the chronic stress levels; intra-class correlation (ICC) was .81, indicating good inter-rater reliability.

In the process of eliciting information about chronic stress, interviewers asked if there had been specific negative life events in that domain. If an event was determined to have occurred within the last year, interviewers used general probes to elicit relevant information regarding the nature and surrounding context of the event. Interviewers were trained to take brief notes during the interviews and then to complete a more detailed narrative after completing the interview. Interviewers also classified events using the romantic relationship categories included in the Paykel life events checklist (Paykel et al., 1971). Independent rater teams then met biweekly to review and rate each event; the interviewer who conducted the interview provided a description of the event and two separate raters assigned a severity score using a 5-point scale ranging from 1 (nonevent, or no significant threat or negative impact) to 5 (a very severe event, maximal negative impact or threat) in half-point increments, indicating the expected amount of impact for a typical individual given a similar context). To establish reliability, a subset of 447 randomly chosen episodic events (across all life domains) were rated independently by two teams; the intra-class correlation (ICC) was .73, indicating good interrater reliability.

For our sample, 46.5% of the participants reported at least one romantic relationship episodic event and 531 discrete events were reported overall (21% of individuals with an event reported more than one). We had two clinical psychology graduate students read the interviewer descriptions and categorize them into the romantic relationship categories outlined in Paykel et al.’s (1971) life events checklist ; 97% of the events fell into four domains: (1) advancing existing relationships (e.g., began cohabitation, engagement or marriage), (2) major arguments, (3) starting new relationships, and (4) relationship dissolution (break-ups). The remaining events (less than 3%) could not be classified into these categories, thus were not retained for analysis. Raters agreed on 96% of the ratings (kappa = .92), indicating very strong agreement. Disagreements were discussed and a final code was agreed upon. Of the 531 discrete events reported, the most frequently reported were dissolution (35%), followed by advancing relationships (28%), starting new relationships (28%) and major arguments (9%). We created four separate episodic stress variables corresponding to the four main categories. If individuals reported these events, their episodic stress rating was equal to the severity rating of their reported event1; otherwise they received a value of 1.0 (non-event/no stress).

Romantic relationship status.

In the baseline survey at wave 2 (post college), participants were asked if they were currently single or married and whether they were cohabitating with a partner. In addition, participants were asked in the UCLA LSI if they were currently married or in a committed romantic relationship (i.e., not currently dating others). This information was used to create a relationship status variable with the following categories: (1) married; (2) cohabitating with committed partner, but not married; (3) in a committed relationship, but not cohabitating, and (4) not in a relationship.2 A plurality of participants were in a committed relationship, but not cohabitating (N = 338; 37%), followed by not in a relationship (N = 315, 35%), cohabitating with committed parnter/not married (N = 204, 23%), and married (N = 41, 5%). Given the relatively small number of married individuals, we combined them with cohabitating with committed partner for the primary analyses.3

Daily alcohol use and motives.

On each of the 30 daily diary days, participants logged in to a secure Internet-based daily survey to report how many drinks (responses: 0 to >15) they had in social (interacting with others) and non-social (alone; not interacting with others) contexts separately for the previous evening (i.e., after they completed the prior day’s survey) and for today (up to the reporting time). One drink is listed as “one 12-oz. can or bottle of beer, one 4-oz. glass of wine, one 12-oz. wine cooler or 1-oz. of liquor straight or in a mixed drink.” Given our focus on problematic drinking, we summed the total number of drinks consumed and calculated a binary heavy drinking day code (≥ 4 for women and ≥ 5 drinks for men). We then aggregated these values over the number of available reporting days to create a proportion of heavy drinking days.

On days when participants reported alcohol use, they were then queried about their reasons for drinking using a slightly modified version of Cooper’s (1994) Reasons for Drinking (RFD) scale. Participants were asked whether they drank for the following reasons (responding with a 3-point scale [0 = no, 1 = somewhat, 2 = definitely]). DTC motivation was assessed with “to forget my ongoing problems/worries” “to feel less depressed,” “to feel less nervous,” “to avoid dealing with my ongoing problems “to cheer up, “because I was angry, and “to feel more confident/sure of myself” and enhancement motivation was assessed with “because I like the pleasant feeling” and “to have fun.” To balance out the response burden, on days when participants did not drink, they were queried about reasons for not drinking.

Close to real-time assessment of drinking motives minimizes various recall errors and biases (Bolger et al., 2003; Tennen et al., 2006) and previous analysis of these scales using this data (Armeli et al., 2014; Armeli et al., 2018) provide evidence for the day-level validity and reliability of these measures. In the present study we then aggregated reports over all drinking days (i.e., calculated mean daily report). If participants did not have any reports of drinking during a reporting wave, we assigned them a value of zero. Reliability estimates for these person-level values were derived from intercept-only models (calculated using HLM software; Raudenbush et al., 2004) separately for each Wave. Estimates were as follows: college Wave (DTC = .84, enhancement = .77), post-college (DTC =.86, enhancement =.85).

Data analysis

We used multiple regression analyses to test the primary hypotheses regarding the unique effects of romantic relationship status and stress in predicting mean daily levels of post-college DTC motivation and enhancement motivation and the proportion of heavy drinking days (each assessed in different models). We created two Helmert codes for relationship status: one comparing individuals in relationships (coded 2) versus individuals not in relationships (coded −1) and another comparing married/cohabitating individuals (coded 1) versus individuals in relationships, but not cohabitating (coded −1); not in relationship was coded 0. All models controlled sex (men = 0, women = 1), age post-college, and the level of the dependent variables during college. We entered the predictors in steps to assess whether (a) the relationship status contrasts and episodic relationship stressors, along with the control variables, showed unique effects (step 1) and (b) if these unique effects remained after controlling for relationship chronic stress (step 2). Finally, we tested the moderating effect of relationship chronic stress on the association between each of the episodic stressor predictors and the dependent variables by creating four product terms (i.e., multiplying chronic stress by each episodic stressor). These four product terms were included into the models in step 3.

Results

Descriptive Statistics and Correlations

Table 1 shows the descriptive statistics and correlations of the study variables. Relationship chronic stress was positively associated with all types of episodic relationship stress except for a negative association with advances in relationships. Chronic stress was also positively associated with college and post-college levels of DTC motivation, but not enhancement motivation. The episodic stress ratings were weakly associated with each other. Finally, proportion of heavy drinking days showed no positive assocations with chronic stress, episodic stress or DTC motivation, however, at both waves it was positively associated with enhancement motivation and was higher among men compared to women.

Table 1.

Descriptive statistics and correlations

M SD 1 2 3 4 5 6 7 8 9 10 11
1. Relationship Chronic stress 2.13 .64
2. Advances in relationships 1.07 .21 −.11*
3. Arguments 1.04 .18 .18* −.06
4. Starting new relationships 1.05 .17 .15* −.07* −.03
5. Relationship dissolution 1.16 .40 .34* −.09* .01 .13*
6. DTC motivation college .21 .29 .19* −.07* .01 .01 .12*
7. Enhancement motivation college 1.24 .50 .06 .00 −.06 .03 .05 .26*
8. DTC motivation post-college .10 .17 .22* −.08* .00 .06 .18* .36* .12*
9. Enhancement motivation post-college .85 .48 .06 −.05 −.04 .01 .03 .04 .39* .24*
10. Heavy drinking college .16 .12 −.02 −.01 .01 .03 −.09* .05 .17* .05 .14*
11. Heavy drinking post-college .14 .13 .03 −.02 .01 .00 −.02 −.05 .13* .12 .29* .35*
12. Sex -- -- .03 .05 −.01 −.03 .02 .10* .05 −.03 .03 −.18* −.13*
Open in a new tab

Note. DTC = drinking to cope, Sex: men = 0, women = 1.

*

p ≤ .05

Table 2 shows the means for the study variables for each relationship status group. We found significant differences for all of the stress variables across status. Specifically, married/cohabitating individuals showed the lowest levels of chronic stress. In addition, married/cohabitating individuals showed the lowest level of relationship dissolution stress, but higher levels of relationship advancement stress – possibly due to the fact that these events were differentially prevalent across these groups.4 We also found post-college differences in drinking motivation, with individuals not in relationships showing higher levels of DTC motivation compared to indivduals in relationships, and married/cohabitating indivdiuals showing lower levels of post-college enhancement motivation compared to individuals not in relationships.

Table 2.

Relationship status and stress, drinking motivation and heavy drinking

Married/
Cohabitating 		Non-cohab.
Relationship 		Not in
Relationship
M SD M SD M SD F p r 2
Relationship Chronic stress 1.87a 0.49 2.18b 0.60 2.30c 0.70 35.22 <.001 .073
Advances in relationships 1.25a 0.32 1.01b 0.09 1.00b 0.00 170.36 <.001 .276
Major arguments 1.05a 0.21 1.06a 0.21 1.01b 0.08 7.48 .001 .016
Starting new relationships 1.01a 0.11 1.11b 0.22 1.02a 0.11 32.33 <.001 .067
Relationship dissolution 1.03a 0.20 1.12b 0.36 1.31c 0.50 40.37 <.001 .083
DTC motivation college 0.18 0.24 0.22 0.29 0.23 0.31 2.06 .128 .005
DTC motivation post-college 0.08a 0.14 0.09a 0.15 0.13b 0.20 7.13 .001 .016
Enhancement motivation college 1.21 0.48 1.24 0.49 1.25 0.51 0.40 .672 .001
Enhancement motivation post-college 0.78a 0.47 0.85a,b 0.48 0.89b 0.48 4.10 .017 .009
Heavy drinking college 0.16 0.12 0.16 0.12 0.17 0.12 0.82 .440 .002
Heavy drinking post-college 0.13 0.14 0.13 0.12 0.15 0.12 2.78 .062 .006
Open in a new tab

Note. F-values df = 2, 895. Means with different superscripts are significantly different at the .05 alpha level based on Tukey-Kramer follow up tests.

Multiple Regresion Results

Results from the multiple regressions are shown in Table 3. For post-college DTC motivation levels, the predictors in step 1 indicated that college levels of DTC motivation was a unique positive predictor. The relationship status contrasts were not uniquely significant and only relationship dissolution accounted for unique variance in post-college levels of DTC motivation. Specifically, more stressful dissolutions were associated with higher DTC motivation post-college. Additional analysis did show that the relationship vs. no relationship contrast was significant (b = .012, beta = . 100, p = .002) without the episodic stressor variables in the model. In step 2, relationship chronic stress was a unique positive predictor of post-college DTC motivation and the effect of dissolution stress remained significant, but was reduced in size.

Table 3.

Multiple regression results predicting post-college DTC motivation, heavy drinking and enhancement motivation

DTC motivation Heavy drinking Enhancement motivation
Predictors B SE p Δr2 B SE p Δr2 B SE p Δr2
Step
 1 DV at Wave 1 (college) .207 .019 <.001 .118 .345 .034 <.001 .103 .369 .030 <.001 .145
Sex −.020 .011 .062 .003 −.016 .008 .053 .004 .025 .030 .395 .001
Age −.003 .004 .414 .001 .001 .003 .740 .000 .005 .012 .675 .000
Not in relationship vs. others .007 .004 .075 .003 .005 .003 .101 .003 .024 .012 .042 .004
Not cohab. vs. married/cohab. −.007 .008 .371 .001 −.001 .006 .853 .000 .034 .022 .115 .002
Advances in relationships −.027 .030 .364 .001 −.003 .023 .888 .000 −.024 .084 .775 .000
Major arguments −.001 .030 .979 .000 .008 .022 .725 .000 −.040 .083 .629 .000
Starting new relationships .056 .033 .090 .003 −.005 .025 .826 .000 .006 .092 .951 .000
Relationship dissolution .045 .014 .001 .010 .000 .011 .973 .000 −.020 .039 .603 .000
 2 Not in relationship vs. others .004 .004 .299 .001 .005 .003 .159 .002 .022 .012 .067 .003
Not cohab. vs. married/cohab. −.012 .008 .129 .002 −.002 .006 .718 .000 .031 .022 .160 .002
Advances in relationships −.034 .030 .249 .001 −.005 .023 .842 .000 −.028 .085 .739 .000
Major arguments −.026 .030 .385 .001 .002 .023 .914 .000 −.057 .086 .505 .000
Starting new relationships .041 .033 .214 .001 −.009 .025 .732 .000 −.004 .093 .968 .000
Relationship dissolution .032 .014 .026 .005 −.003 .011 .760 .000 −.029 .040 .468 .001
Relationship Chronic stress (RCS) .035 .009 <.001 .013 .007 .007 .297 .001 .022 .026 .392 .001
 3 RCS × Advances in relationships −.067 .050 .180 .002 .042 .039 .273 .001 .089 .143 .534 .000
RCS × Major arguments −.068 .042 .108 .002 −.024 .033 .470 .001 −.089 .121 .460 .001
RCS × Starting new relationships −.037 .038 .327 .001 −.035 .029 .232 .001 .102 .107 .340 .001
RCS × Relationship dissolution .069 .018 <.001 .014 .015 .014 .265 .001 .020 .050 .690 .000
Open in a new tab

Note. DV at Wave 1 (college) = value of the dependent variable in Wave 1 (college). Sex: men = 0, women = 1; Not in relationship vs. others: Not in relationships = 2 vs. others =−1; Not cohab. vs. married/cohab.: in relationship/not cohabitating = 1, married or cohabitating = −1, not in relationship =0.

The results for post-college enhancement motivation and heavy drinking days are also shown in Table 3. In step one, only college levels of the dependent variables predicted heavy drinking. For enhancement motivation, college levels were predictive as was relationship status, with individuals not in relationships showing higher levels compared to individuals not in relationships. None of the episodic stress variables were predictive in either model. In step 2, relationship chronic stress was not significant in either model.

Step 3 in Table 3 shows the results for the product terms testing the relationship chronic stress × episodic stress interactions. For post-college DTC motivation, we found a significant interaction between relationship chronic stress and dissolution stress. To clarify the effect, we used the Johnson-Neyman approach described in Hayes (2018) to probe for regions of significance. Individuals in the top 33% of relationship chronic stress (values approximately ≥ 2.5 on our rating scale) showed a significant positive association between dissolution stress and post-college DTC motivation; individuals in the bottom 5% on relationship chronic stress (values approximately <1.4 on our rating scale) showed a negative association between dissolution stress and post-college DTC motivation. No interactions were found for heavy drinking and enhancement motivation.

Supplemental analyses

We conducted several supplemental analyses. First, we examined the possibility that relationship status and episodic stressors had indirect effects on post-college DTC motivation through their association with relationship chronic stress. We focused only on post-college DTC motivation because this was the only outcome related to relationship chronic stress. We first estimated a regression predicting chronic stress from the predictors described in step 1 of the previously described regressions.5

Results are shown in Table 4. Individuals not in relationships showed higher levels of chronic stress compared to individuals in relationships. In addition, married/cohabitating individuals showed lower levels of chronic stress compared to individuals in relationships, but not cohabitating. All of the episodic stress variables, except for relationship advancement stress, were positively related to chronic stress. Next, we estimated 95% percentile bootstrap confidence intervals (CI) (specifying 5000 bootstrap estimates) for the indirect effects of relationship status and episodic stress on post-college DTC motivation through relationship chronic stress using procedures outlined by Hayes (2018). Results indicated significant indirect effects (i.e., the CI did not contain zero) for not in relationship vs. in relationship (CI: .001 to .006: beta = .018), non-cohabitating relationships vs. married/cohabitating (CI: .002 to .009: beta = .029), dissolution (CI: .006 to .022: beta = .03); major arguments (CI: .010 to .043: beta = .03), and starting relationship (CI: .003 to .031: beta = .015).

Table 4.

Multiple regression results predicting post-college relationship chronic stress

Predictors B SE p Δr2
DTC motivation (college) .301 .067 <.001 .018
Sex .056 .038 .142 .002
Age .023 .015 .120 .002
Not in relationship vs. others .088 .015 <.001 .030
Not cohab. vs. married/cohab. .143 .028 <.001 .023
Advances in relationships .207 .108 .055 .003
Major arguments .737 .107 <.001 .042
Starting new relationships .430 .118 <.001 .011
Relationship dissolution .385 .050 <.001 .052
Open in a new tab

Note. Sex: men = 0, women = 1; Not in relationship vs. others: Not in relationships = 2 vs. others =−1; Not cohab. vs. married/cohab.: in relationship/not cohabitating = 1, married or cohabitating = −1, not in relationship =0.

Next, we tested for sex differences in (a) the unique (direct) effects of episodic and chronic stress and relationship status in predicting post-college DTC motivation and (b) the indirect effects of episodic stressors and relationship status on post-college DTC through relationship chronic stress. We followed procedures outlined by Hayes’ (2018) for testing the moderating effect of sex on the direct and indirect effects of each predictor of interest separately using same covariates that were included in our previous models. Results for these analyses are shown in the online supplement (see online supplemental tables). Sex only moderated the direct effect of dissolution stress (b = .08, SE = .03 p = .003, Δr2 = .008), with women (b = 16, p <.001, 95% CI: .09 to .25) but not men (b = −.03, p = .52, 95% CI −.13 to .07) showing a significant effect. None of the other direct effects interacted with sex and tests of moderated mediation all indicated no sex differences.

Finally, we examined differences in proportion of drinking days post-college across relationship status groups and found that married/cohabitating (M = .42, SD = .24) showed the higher levels compared to individuals in relationships, but not cohabitating (M = .35, SD = .21) and individuals not in relationships (M = .37, SD = .21). We then re-estimated the model predicting heavy drinking controlling for post-college proportion of drinking days in step 1 and found that the two relationship status contrasts were significant in the predicted direction, with individuals not in relationships compared to others showing higher levels of heavy drinking (b = .008, beta = .086, p = .002) and individuals in relationships, but not cohabitating showing higher levels of heavy drinking compared to married/cohabiting individuals (b = .011, beta = .069, p = .015). Inclusion of the stress predictors in the model did not affect the significance of these status effects.

Discussion

Our study extended previous research on romantic relationship status and stress in several ways. Most notably, we showed that relationship status is correlated with relationship stress, thus highlighting the importance of assessing both factors when trying to understand the role of romantic relationships in the development of problematic alcohol use during young adulthood. Although we did find that relationship status was associated with changes in DTC motivation from college to post-college years – with married or cohabitating individuals showing the lowest levels and individuals not in relationships showing the highest levels – status differences were no longer significant after controlling for relationship stress. This suggests an important role of negative reinforcement processes, in addition to role socialization or selection processes, in understanding the association between relationship status and problematic drinking during this developmental period.

Our findings also highlight the complexity of these posited negative reinforcement processes in that relationship status is associated with qualitatively different types of stressors. One example of this is that relationship dissolution was more prevalent among individuals not in relationships, whereas major arguments were more prevalent among individuals currently in relationships. However, regardless of the types of episodic stressors experienced across the relationship status groups, the association between these distinct stressors and post-college levels of DTC motivation was generally related to their association with relationship stability and satisfaction with current status (i.e., chronic stress). Stated in other words, although relationship status differences in DTC motivation might arise, to some degree, from the experience of distinct types of episodic relationship stressors across these groups, the effects of these stressors on DTC motivation, for the most part, occur through a common pathway: higher levels of chronic relationship stress. We should note that our data are not ideal for testing indirect effects given that all of the variables involved were measured at the same time (Maxwell et al., 2011). However, we take Hayes’s (2018) perspective that our tests of indirect effects are presented merely to enhance our theoretical interpretation of the pattern of findings we present.

Beyond possible indirect effects that might exist, our results indicated that the effect of dissolution stress on post-college DTC motivation remained significant (though substantially reduced in size) after controlling for chronic stress. One possibility is that relationship dissolution has multiple pathways through which it affects DTC motivation. Dissolution might cause individuals to drink due to its effect on persisting (chronic) evaluations of relationship stability and satisfaction. In addition, dissolution might be associated with other stress-related appraisals, such as perceived loss of one’s former partner; this might also explain why dissolution stressors were assessed as more severe compared to other stressors. Relationship dissolution also might be related to role selection and socialization processes, such as the loss of associated peer networks (i.e., loss of common friends after a breakup) which might further exacerbate the stressfulness of such scenarios. Our supplemental analysis indicated that the unique effect of dissolution stress was significant only for women. Although this interaction was not predicted and should be replicated, this could indicate that women might be more reactive to the unique processes associated with dissolution stress in terms of turning to alcohol as a viable coping strategy.

We also found evidence that chronic relationship stress moderated the association between relationship dissolution stress and post-college DTC motivation. Rather than framing this in terms of the vulnerability conferred by high levels of chronic stress, it might be more useful to focus on the protective nature of low chronic stress. For individuals in relationships, low chronic stress in our study was conceptualized as the presence of supportive partners, good conflict resolution ability, and communication. Considerable literature demonstrates the overall benefits of healthy romantic relationships (Loving & Slatcher, 2013) and its role in buffering the effects of life stress (e.g., Busuito et al., 2014; Florian et al., 2002). For individuals not in relationships, low chronic stress was conceptualized as being satisfied and not lonely without a partner, and adequate fulfillment from social life and life plans. Past research indicates that satisfaction with relationship status and general social support serve protective roles in general (e.g., Lehmann et al., 2015) and during times of stress (Cohen & Wills, 1985). Although our findings were consistent with these notions, additional research is needed to better understand the differential protective components at play for individuals in and not in current relationships.

We should note that although there is controversy over whether a variable can act as mediator and as a moderator (e.g., Kraemer et al., 2008), we take the approach outlined by Karazsia and Berlin (2018) that variables can play a role in both processes. For example, discrete episodes of stress (e.g., major arguments, dissolution) might result in lower levels of relationship stability and satisfaction (i.e., increased chronic stress). High levels of chronic stress then might serve as a vulnerability when future episodic stressors occur (e.g., dissolution), resulting in more deleterious outcomes from such events. It should be noted that episodic stressors only accounted for a small to moderate amount of variability in chronic stress, suggesting that a variety of other factors (e.g., finding a new partner, satisfaction with current prospects and/or current single status) probably play a larger role in determining overall chronic stress levels. Our mediation and moderation effects might be capturing processes associated with these different risk and protective factors underlying reports of chronic stress. Nevertheless, consistent with guidelines suggested by Karazsia and Berlin (2018), future studies that repeatedly assess chronic stress prior to and after the occurrence of stressors such as dissolution, along with the sequelae of such events, are needed to better evaluate both mediation and moderation processes.

Unlike previous studies, we did not find clear evidence for relationship status differences in heavy drinking (e.g., Fleming et al., 2018; Patrick et al., 2018). This might be due to sample differences; our sample was comprised of generally high functioning college graduates who were moderate to heavy drinkers in college, whereas previous studies (Fleming et al., 2018; Patrick et al., 2018) used broader community samples. Our findings showing that married/cohabitating individuals drank with greater frequency, but with a lower proportions of heavy drinking days, might be distinct to highly educated, financially secure individuals in committed relationships. Such individuals might be more likely to engage in low-risk drinking in convivial social settings (e.g., drinking wine with dinner, casual get-togethers with friends) where drinking is driven by processes other than negative reinforcement compared to more high-risk type scenarios (e.g., bar-club settings). Another complicating factor was that we were unable to separate out single, not actively dating individuals – a group that showed low drinking levels in Fleming et al. (2018) – from other individuals not in relationships. Inclusion of such individuals in this group might have attenuated our comparisons. Future studies using more diverse samples should investigate the nature of the drinking contexts and more precisely assess current relationship status to examine these possibilities.

Finally, several key limitations of our study merit mentioning. First, although we used a gold-standard approach in assessing recent romantic life stress, we only assessed it for the past year. It is possible that relationship stress prior to this period (e.g., a highly stressful break up) would help to explain variation in recent alcohol use and drinking motives. Future studies should use this life stress assessment protocol repeatedly (e.g., yearly) across emerging adulthood to better understand the unfolding of changes in relationship status and related episodic and chronic stressors during this critical developmental period. Second, it is likely that much of the coping-motivated drinking associated with relationship stressors occurred proximal to such events (e.g., Derrick & Testa, 2017). Our daily reporting was not contingent on the occurrence of these events, thus in many cases assessment of drinking and motives might have occurred months later, thus attenuating our effects of interest. Future studies are needed that coordinate daily reporting with the occurrence of such events. Third, our sample reported generally low levels of recent relationship stress; less than half of our sample reported an episodic stressor and most of these stressors were rated relatively low in severity. This could be due to the fact that this was a high functioning population (highly educated, financially stable) and relatively young (most still in their early twenties). Many of our participants might not have experienced more severe types of romantic relationship stress such as divorce or dissolution of long-term relationships. Fourth, although minority students were actively recruited for the current study, the sample was primarily White, thus limiting the generalizability of our findings. Fifth, chronic stress was operationalized differently for individuals in relationships compared to individuals not in relationships. Thus, differences in how raters used the assessment scale might account for relationship status differences observed in chronic stress. However, as predicted, we did find that married/cohabitating individuals showed lower levels of relationship chronic stress compared to individuals in relationships but not cohabitating – two groups evaluated using common stress ratings anchors. Finally, future research should also assess more nuanced drinking motives related to romantic relationships. For example, among individuals not in relationships, drinking to cope with loneliness and drinking to reduce anxiety when encountering potential dating partners might be commonly reported reasons for drinking that have differential effects on drinking-related problems.

These limitations notwithstanding, our results provide new information about the role of romantic relationships in the development and maintenance of drinking for problematic reasons (i.e., drinking to cope) among young adults. Understanding which factors associated with being in romantic relationships, or the desire to be in such relationships, might serve in a risk (e.g., relationship dissolution) or protective (e.g., supportive partners; satisfaction with being single) capacity with respect to coping-motivated drinking could help to further enhance intervention efforts focusing on the role of interpersonal conflict as a precursor to alcohol-related problems (Rodriguez et al., 2019; Litt et al., 2003).

Supplementary Material

Tables

Public Health Significance Statements: This study indicates that both romantic relationship status and the stressfulness of this social role might help explain using alcohol for coping reasons among young adults.

Acknowledgments

This research was supported by grants P60-AA003510 and 5T32-AA007290-37 from NIAAA. Findings from this study have not been previously reported elsewhere.

Footnotes

1.

If an individual reported more than one of the same types of episodic stressors, we used the highest stress rating as the final value. This was infrequent, occurring only among 7.7% of the stress values across all categories.

2.

All participants who reported being single in the baseline survey also reported not being in a committed relationship in the UCLA interview.

3.

We did assess marital status at wave 1 (in college); only one participant was married. This person was no longer married at wave 2 and was retained for analysis.

4.

Relationships status differences in the episodic stress variables showed evidence of heteroskedasticity, mainly due to the differential prevalence of events occurring in these groups, e.g., individuals not in relationships showed greater frequency of dissolution stress and no variability in advancement stress. Thus, we also examined these differences using Welch’s ANOVA and Games-Howell follow up tests; conclusions were identical. In addition, we converted relationship advancement stress into a binary variable (no stress vs. any stress >1) and used a chi-square test to examine its association with relationship status; results for group differences were substantively identical to those shown in Table 2.

5.

We included DTC motivation at wave 1 (college) because it was found to be related to chronic relationship stress at the bivariate level (see Table 1).

References

  1. Armeli S, Dranoff E, Tennen H, Austad CS, Fallahi CR, Raskin S, Wood R, & Pearlson G (2014). A longitudinal study of the effects of coping motives, negative affect and drinking level on drinking problems among college students. Anxiety, Stress, & Coping, 27(5), 527–541. 10.1080/10615806.2014.895821 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Armeli S, Covault J, & Tennen H (2018). Long-term changes in the effects of episode-specific drinking to cope motivation on daily well-being. Psychology of Addictive Behaviors, 32, 715–726. 10.1037/adb0000409 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arria AM, Caldeira KM, Allen HK, Vincent KB, Bugbee BA, & O’Grady KE (2016). Drinking like an adult? Trajectories of alcohol use patterns before and after college graduation. Alcoholism: Clinical and Experimental Research, 40, 583–590. https://doi-org.libaccess.fdu.edu/10.1111/acer.12973 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bachman JG, Wadsworth KN, O’Malley PM, & Schulenberg J (1997). Marriage, divorce, and parenthood during the transition to young adulthood: Impacts on drug use and abuse. In Schulenberg J, Maggs JL, & Hurrelmann K (Eds.), Health risks and developmental transitions during adolescence (pp. 246–279). New York, NY: Cambridge University Press. [Google Scholar]
  5. Bennett ME, McCrady BS, Johnson V, & Pandina RJ (1999). Problem drinking from young adulthood to adulthood: Patterns, predictors and outcomes. Journal of Studies on Alcohol, 60, 605–614. 10.15288/jsa.1999.60.605 [DOI] [PubMed] [Google Scholar]
  6. Bolger N, Davis A, & Rafaeli E (2003). Diary methods: Capturing life as it is lived. Annual Review of Psychology, 54, 579–616. 10.1146/annurev.psych.54.101601.145030 [DOI] [PubMed] [Google Scholar]
  7. Brown GW & Harris T (1978). Social Origins of Depression. A Study of Psychiatric Disorder in Women. Tavistock. London. [Google Scholar]
  8. Busuito A, Huth-Bocks A, & Puro E (2014). Romantic attachment as a moderator of the association between childhood abuse and posttraumatic stress disorder symptoms. Journal of Family Violence, 29, 567–577. https://doi-org.libaccess.fdu.edu/10.1007/s10896-014-9611-8 [Google Scholar]
  9. Cohen S, & Wills TA (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. https://doi-org.libaccess.fdu.edu/10.1037/0033-2909.98.2.310 [PubMed] [Google Scholar]
  10. Collibee C, & Furman W (2015). Quality counts: Developmental shifts in associations between romantic relationship qualities and psychosocial adjustment. Child Development, 86, 1639–1652. 10.1111/cdev.12403 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cooper ML (1994). Motivations for alcohol use among adolescents: Development and validation of a four-factor model. Psychological Assessment, 6, 117–128. doi: 10.1037/1040-3590.6.2.117 [DOI] [Google Scholar]
  12. Derrick JL, & Testa M (2017). Temporal effects of perpetrating or receiving intimate partner aggression on alcohol consumption: A daily diary study of community couples. Journal of Studies on Alcohol and Drugs, 78, 213–221. doi: 10.15288/jsad.2017.78.213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dohrenwend BP (2006). Inventorying stressful life events as risk factors for psychopathology: Toward resolution of the problem of intracategory variability. Psychological Bulletin, 132, 477–495. https://psycnet.apa.org/doi/10.1037/0033-2909.132.3.477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Erikson EH (1968). Identity: youth and crisis. Norton & Co.; Oxford, England. [Google Scholar]
  15. Fleming CB, Lee CM, Rhew IC, Ramirez JJ, Abdallah DA, & Fairlie AM (2018). Descriptive and prospective analysis of young adult alcohol use and romantic relationships: Disentangling between and within-person associations using monthly assessments. Substance Use & Misuse, 53, 2240–2249. 10.1080/10826084.2018.1467455 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fleming CB, White HR, & Catalano RF (2010). Romantic relationships and substance use in early adulthood: An examination of the influences of relationship type, partner substance use, and relationship quality. Journal of Health and Social Behavior, 51, 153–157. https://dx.doi.org/10.1177%2F0022146510368930 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Florian V, Mikulincer M, & Hirschberger G (2002). The anxiety-buffering function of close relationships: Evidence that relationship commitment acts as a terror management mechanism. Journal of Personality and Social Psychology, 82, 527–542. https://doi-org.libaccess.fdu.edu/10.1037/0022-3514.82.4.527 [PubMed] [Google Scholar]
  18. Gotham HJ, Sher KJ, & Wood PK (1997). Predicting stability and change in frequency of intoxication from the college years to beyond: Individual-difference and role transition variables. Journal of Abnormal Psychology, 106, 619–629. https://doi-org.libaccess.fdu.edu/10.1037/0021-843X.106.4.619 [DOI] [PubMed] [Google Scholar]
  19. Hammen C (2016): Depression and stressful environments: Identifying gaps in conceptualization and measurement. Anxiety, Stress, & Coping, 29, 335–351. [DOI] [PubMed] [Google Scholar]
  20. Hammen C (1991). The generation of stress in the course of unipolar depression. Journal of Abnormal Psychology, 100, 555–561. [DOI] [PubMed] [Google Scholar]
  21. Hammen C, Ellicott A, Gitlin M, & Jamison K (1989). Sociotropy/Autonomy and vulnerability to specific life events in patients with unipolar depression and bipolar disorders. Journal of Abnormal Psychology, 98, 154–160. [DOI] [PubMed] [Google Scholar]
  22. Hayes AF (2018). Introduction to mediation, moderation, and conditional process analysis (2nd Ed.). New York: The Guilford Press. [Google Scholar]
  23. Hazan C, & Zeifman D (1994). Sex and the psychological tether. In Bartholomew K & Perlman D (Eds.), Attachment processes in adulthood: Advances in personal relationships (pp. 151–178). London, UK: Kingsley. [Google Scholar]
  24. Horwitz AV, & White HR (1998). The relationship of cohabitation and mental health: A study of a young adult cohort. Journal of Marriage and the Family, 60, 505–514. https://psycnet.apa.org/doi/10.2307/353865 [Google Scholar]
  25. Jochman KA, & Fromme K (2010). Maturing out of substance use: The other side of etiology. In Scheier L (Ed.), Handbook of drug use etiology: Theory, methods, and empirical findings (pp. 565–578). Washington, DC: American Psychological Association [Google Scholar]
  26. Jackson KM, & Sartor CE (2016). The natural course of substance use and dependence. In Sher KJ (Ed.), The Oxford handbook of substance use and substance use disorders, Vol. 1 (pp. 67–131). New York, NY: Oxford University Press. [Google Scholar]
  27. Karazsia BT, & Berlin KS (2018). Can a mediator moderate? Considering the role of time and change in the mediator-moderator distinction. Behavior Therapy, 49, 12–20. 10.1016/j.beth.2017.10.001 [DOI] [PubMed] [Google Scholar]
  28. Kraemer HJC, Kiernan M, Essex M, & Kupfer D (2008). How and why criteria defining moderators and mediators differ between the Baron & Kenny and MacArthur approaches. Health Psychology, 27, S101–S108. 10.1037/0278-6133.27.2(Suppl.).S101 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lambe L, Mackinnon SP, & Stewart SH (2015). Dyadic conflict, drinking to cope, and alcohol-related problems: A psychometric study and longitudinal actor-partner interdependence model. Journal of Family Psychology, 29, 697–707. https://doi-org.libaccess.fdu.edu/10.1037/fam0000098.supp (Supplemental) [DOI] [PubMed] [Google Scholar]
  30. Lehmann V, Tuinman MA, Braeken J, Vingerhoets AJJM, Sanderman R, & Hagedoorn M (2015). Satisfaction with relationship status: Development of a new scale and the role in predicting well-being. Journal of Happiness Studies: An Interdisciplinary Forum on Subjective Well-Being, 16, 169–184. https://doi-org.libaccess.fdu.edu/10.1007/s10902-014-9503-x [Google Scholar]
  31. Levitt A, & Leonard KE (2015). Insecure attachment styles, relationship-drinking contexts, and marital alcohol problems: Testing the mediating role of relationship-specific drinking-to-cope motives. Psychology of Addictive Behaviors, 29, 696–705. https://doi-org.libaccess.fdu.edu/10.1037/adb0000064 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Litt MD, Kadden RM, Cooney NL, & Kabela E (2003). Coping skills and treatment outcomes in cognitive-behavioral and interactional group therapy for alcoholism. Journal of Consulting and Clinical Psychology, 71, 118–128. https://doi-org.libaccess.fdu.edu/10.1037/0022-006X.71.1.118 [DOI] [PubMed] [Google Scholar]
  33. Littlefield AK, Sher KJ, & Wood PK (2010). A personality-based description of maturing out of alcohol problems: Extension with a five-factor model and robustness to modeling challenges. Addictive Behaviors, 35, 948–954. 10.1016/j.addbeh.2010.06.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Loving TJ, & Slatcher RB (2013). Romantic relationships and health. In Simpson JA & Campbell L (Eds.), The Oxford handbook of close relationships. (pp. 617–637). Oxford University Press. [Google Scholar]
  35. Maxwell SE, Cole DA, & Mitchell MA (2011). Bias in cross-sectional analyses of longitudinal mediation: Partial and complete mediation under an autoregressive model. Multivariate Behavioral Research, 46, 816–841. https://doi-org.libaccess.fdu.edu/10.1080/00273171.2011.606716 [DOI] [PubMed] [Google Scholar]
  36. Monroe SM (2008). Modern approaches to conceptualizing and measuring human life stress. Annual Review of Clinical Psychology, 4, 33–52. 10.1146/annurev.clinpsy.4.022007.141207 [DOI] [PubMed] [Google Scholar]
  37. Paykel ES, Prusoff BA, & Uhlenhuth EH (1971). Scaling of life events. Archives of General Psychiatry, 25, 340–347. doi: 10.1001/archpsyc.1971.01750160052010 [DOI] [PubMed] [Google Scholar]
  38. Patrick ME, & Schulenberg JE (2011). How trajectories of reasons for alcohol use relate to trajectories of binge drinking: National panel data spanning late adolescence to early adulthood. Developmental Psychology, 47, 311–317. https://psycnet.apa.org/doi/10.1037/a0021939 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Patrick ME, Rhew IC, Lewis MA, Abdallah DA, Larimer ME, Schulenberg JE, & Lee CM (2018). Alcohol motivations and behaviors during months young adults experience social role transitions: Microtransitions in early adulthood. Psychology of Addictive Behaviors, 32, 895–903. https://psycnet.apa.org/doi/10.1037/adb0000411 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Perkins HW (1999). Stress-motivated drinking in collegiate and postcollegiate young adulthood: life course and gender patterns. Journal of Studies on Alcohol, 60(2), 219–227. 10.15288/jsa.1999.60.219 [DOI] [PubMed] [Google Scholar]
  41. Raudenbush SW, Bryk AS, & Congdon R (2004). HLM 6 for Windows [Computer software]. Lincolnwood, IL: Scientific Software International, Inc. [Google Scholar]
  42. Rodriguez LM, Dell JB, Lee KDM, & Onufrak J (2019). Effects of a brief cognitive reappraisal intervention on reductions in alcohol consumption and related problems. Psychology of Addictive Behaviors, 33, 637–643. https://doi-org.libaccess.fdu.edu/10.1037/adb0000509 [DOI] [PubMed] [Google Scholar]
  43. Staff J, Schulenberg JE, Maslowsky J, Bachman JG, O'Malley PM, Maggs JL, & Johnston LD (2010). Substance use changes and social role transitions: Proximal developmental effects on ongoing trajectories from late adolescence through early adulthood. Development and Psychopathology, 22, 917–932. 10.1017/S0954579410000544 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Tennen H, Affleck G, Coyne JC, Larsen RJ, & DeLongis A (2006). Paper and plastic in daily diary research: A reply to Green et al. Psychological Methods 11, 112–118. [DOI] [PubMed] [Google Scholar]
  45. Vrshek-Schallhorn S, Stroud CB, Mineka S, Hammen C, Zinbarg RE, Wolitzky-Taylor K, & Craske MG (2015). Chronic and episodic interpersonal stress as statistically unique predictors of depression in two samples of emerging adults. Journal of Abnormal Psychology, 124(4), 918–932. https://doi-org.libaccess.fdu.edu/10.1037/abn0000088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Yamaguchi K (1990). Drug use and its social covariates from the period of adolescence to young adulthood. Some implications from longitudinal studies. In Galanter M (Ed.), Recent developments in alcoholism. Volume 8: Combined alcohol and other drug dependence (pp. 125–144). New York: Plenum Press. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Tables
NIHMS1648631-supplement-Tables.docx (36KB, docx)

RESOURCES