Adolescent Precursors of Young Adult Drinking Motives

Abstract

Introduction

Numerous cross-sectional and shorter-term longitudinal studies have supported the role of drinking motives as potent proximal predictors of alcohol phenotypes (e.g., alcohol use, heavy episodic drinking). However, missing from this literature is a focus both on the stability of drinking motives across young adulthood and on adolescent precursors of drinking motives.

Methods

We investigated the adequacy of using a latent trait-state model (LTSM) to investigate three-wave data on social, enhancement, and coping motives for drinking with a community sample of young adults (N=1004) at the mean ages of 23.8 years, 28.9 years, and 33.5 years. We further investigated adolescent (M age=16.73 years) predictors of young adult drinking motives using data collected on the sample approximately seven years prior to the first young adult data collection.

Results

Findings indicated that all three drinking motives across young adulthood were modeled adequately via the LTSM, and that drinking motives manifested high stability (i.e., rank order) across individuals. Significant common (e.g., being male, alcohol-using peers, stressful life events, boredom susceptibility) and specific (e.g., depressive symptoms for coping motives; heavy episodic drinking for enhancement motives) adolescent precursors of young adult drinking motives were identified.

Conclusions

Common and unique adolescent factors predicted trait-like drinking motives during young adulthood. These findings suggest the utility of intervening during the teen years to prevent or interrupt the development of cognitive motivations that encourage alcohol use for the purpose of affect regulation.

1. Introduction

The investigation of drinking motives (e.g., coping, enhancement, and social motives) is a vibrant area of both basic research (Cooper, Kuntsche, Levitt, Barber, & Wolf, 2016) and practice (Brown, Anderson, Schulte, Sintov, & Frissell, 2005; Conrod, Castellanos-Ryan, & Mackie, 2011; Watt, Stewart, Birch, & Bernier, 2006). Yet, despite these important research and applied foci, studies have not focused on adolescent predictors of young adult drinking motives, even though their identification would provide information relevant to their etiologic bases and facilitate the identification of targets for interventions. As such, in the current study, we tested ten adolescent predictors of young adult social, coping, and enhancement drinking motives (see a description of predictors in Section 2.2 Measures).

1.1 Selection of Adolescent Predictors and Study Hypotheses

To guide our selection of adolescent predictors, we relied on three literatures. First, there is a substantial literature on adolescent risk factors that are associated with adolescent and young adult alcohol and other substance use outcomes (Biglan, Brennan, Foster, & Holder, 2004; Cleveland, Feinberg, Bontempo, & Greenberg, 2008; Hawkins, Catalano, & Miller, 1992). This literature has suggested that variables across multiple levels (e.g., individual, family, peers, neighborhood) are important developmental precursors of adolescent and young adult alcohol use. Because a corresponding literature on predictors of drinking motives does not exist, we selected putative adolescent risk factors of alcohol use as precursors of young adult drinking motives because drinking motives are among the most robust proximal predictors of alcohol use and disorders (Cox & Klinger, 1988) and are closely linked to actual drinking behaviors (Cooper et al., 2008; Littlefield, Sher, & Wood, 2010; Patrick & Schulenberg, 2011). We therefore reasoned that precursors of drinking motives may overlap with precursors of drinking behaviors.

The second literature included research on personality characteristics and drinking motives (Arbeau, Kuiken, & Wild, 2011; Kuntsche, Knibbe, Gmel, & Engels, 2006). Cox and Klinger’s (1988) cognitive-motivational model of alcohol use hypothesizes that distal (e.g., personality) and proximal (e.g., current affective states) factors influence alcohol-specific reinforcement motivations to use alcohol for purposes of affect regulation (Cooper, Frone, Russell, & Mudar, 1995; Woicik, Stewart, Pihl, & Conrod, 2009). For example, extraversion is positively associated with enhancement drinking motives, with the use of alcohol to enhance positive affective states (Cooper et al., 1995; Kuntsche et al., 2006). Coping motives are positively associated with neuroticism and its constituent components (e.g., depressed affect, anxiety), and there is empirical support for the use of coping motives to mitigate negative affect (Cooper et al., 2016; Kuntsche et al., 2006).

Third, we used the substance use typology literature on internalizing and externalizing pathways to alcohol use (Hussong, Jones, Stein, Baucom, & Boeding, 2011; Windle, 2015; Zucker, 2006, 2008) to propose three developmental pathways by which adolescent precursors predict each of the three drinking motives in young adulthood. Supported by literature on an internalizing pathway (Hussong et al., 2011; Wills & Shiffman, 1985), the first pathway hypothesized was from adolescent depressive symptoms, stressful life events, and lower family support to coping motives. There has been support for this internalizing—coping motives pathway (Cooper et al., 2016; Hussong et al., 2011; Wills & Shiffman, 1985; Windle & Windle, 2015), and a strong and consistent association of lower family emotional support with higher adolescent alcohol use and problems (Barnes, Reifman, Farrell, & Dintcheff, 2000; Windle, 1992).

Likewise, substantial evidence has accumulated for an externalizing pathway to substance use (Windle, 2015; Zucker, 2008), with characteristics of this disinhibited pathway associated with enhancement motives (Kuntsche et al., 2006; Schulenberg, Wadsworth, O’Malley, Bachman, & Johnston, 1996). Based on this literature, it was hypothesized that enhancement motives in young adulthood would be predicted uniquely by adolescent heavy episodic drinking (HED), delinquent activities, marijuana use, and boredom susceptibility. The third pathway was for social motives (a normative social pathway) and there were no unique predictors hypothesized, though there were some common predictors hypothesized to predict all three motives.

Three variables hypothesized as common predictors of the three drinking motives were adolescent alcohol use, density of alcohol-using friends, and male sex. Findings have indicated that alcohol use and alcohol-using friends are significant correlates and predictors of all three drinking motives (Kuntsche, Knibbe, Gmel, & Engels, 2005; Kuntsche & Stewart, 2009). With regard to sex, early adolescent males and females report similarities in drinking motives but during later adolescence and early adulthood sex differences emerge, with males reporting higher levels of social and enhancement motives and weaker sex differences for coping motives, with males slightly higher than females (Cooper, 1994; Kuntsche et al., 2006). Hence, we hypothesized that male sex would be a general predictor of all three motives.

Finally, few studies have investigated the relationship between drinking motives and academic performance. However, of those that have, higher levels of coping and enhancement motives have been predictive of academic problems and lower school performance (Merrill & Read, 2010; Merrill, Wardell, & Read, 2014; Windle & Barnes, 1988). Based on these findings, we hypothesized that lower grades in high school would predict higher levels of coping and enhancement motives in young adulthood.

1.2 Summary of Study Goals

The aims of this study were two-fold. First, utilizing three waves of data on coping, social, and enhancement motives collected approximately five years apart during young adulthood, we evaluated the adequacy of a latent trait-state model (LTSM) to represent the data. The LTSM allows for the estimation of trait and state (or occasion-specific) components of an individual difference variable (Jackson & Sher, 2003; Windle & Dumenci, 1998), thereby making inferences about the stable, trait-like nature of a variable while accounting for possible situational state influences. Second, we specified ten adolescent predictors to account for variation in drinking motives during young adulthood.

2. Method

2.1 Participants and Procedures

The data were collected as part of a multi-wave panel design focused on risk factors and alcohol and other substance use among 1205 teens during the high-school years (with four waves of assessment at six-month intervals occurring from 1988–1990) (for details, see Windle & Wiesner, 2004). Survey data were collected within the adolescents’ high-school settings and the student participation rate was 76%. The sample consisted of high-school sophomores (52%) and juniors (48%) recruited from two homogeneous suburban public high-school districts (a total of three high schools) in Western New York. The average age of the respondents at Wave 1 was 15.54 years (SD=0.66), 98% were White, and 50.8% were females. Sample retention across the first four waves of measurement was uniformly high, in excess of 90%.

There was a seven-year gap between the Wave 4 assessment in adolescence and the Wave 5 data collection that occurred when the average age of the young adults was 23.8 years, and then five-year gaps between Wave 6 (age=28.9 years) and Wave 7 (age=33.5 years). Participants were paid $40 to complete a computer-assisted personal interview that lasted approximately 2 hours. At all waves of the study, informed consent was used and confidentiality was assured with a U.S. Department of Health and Human Services Certificate of Confidentiality. This study was reviewed and approved by the Institutional Review Board of the University at Buffalo.

In the current study we used data from 1004 participants, and 53% were females. These subjects participated at least once during adolescence (i.e., Waves 3 and/or 4) and at least once during young adulthood (i.e., Waves 5–7; 55.8% participated at all three waves, 24.6% at two-waves, and 19.6% at one wave). In addition, to be included in the current study, young adults had to have reported drinking alcohol during at least one wave of data collection during Waves 5–7. Missing values were estimated using full information maximum likelihood estimation for about 20% of the individual responses.

Attrition analyses were conducted for those participants in the current study (n=1004) and those who did not meet inclusion criteria (n=201). A chi-square test indicated that significantly more females (86.4%) than males (78.8%) were included in the study (χ² with 1 df=12.30, p <.001). Because males and females differ on a number of variables used in this study, attrition analyses were conducted using adolescent data separately for males and females. One-way ANOVAS were conducted for each sex group across14 variables: three sociodemographic variables (family income, number of children in the family, and primary caregiver’s highest educational attainment), family cohesion, percentage of alcohol and drug using friends, respectively, past month use of tobacco, alcohol, marijuana, and other illicit drug use, frequency of binge drinking, delinquency, depressive symptoms, and number of stressful life events in the last six-months. Of these 28 comparisons (14 for each of the sex groups), only two were statistically significant. For males, primary caregiver’s educational level was lower for the non-included group (ES=.22); for females, delinquency was higher for the non-included group (ES=.42). The other 26 comparisons indicated no significant group differences. Based on these findings, we concluded that attrition bias was minimal for the included and non-included groups.

2.2 Measures

Many of the adolescent measures used in this study consisted of the average of two scores that were assessed six-months apart at Waves 3 and 4. This procedure was adopted to increase the reliability and validity of assessment across a one-year period. Given fluctuations in alcohol use and other problem behaviors across adolescence, we sought to capture variation among adolescents who engaged in the targeted behaviors during the last year of the adolescent portion of the study. Alcohol and substance use variables were non-normally distributed and log transformed scores were used in the analyses.

Quantity-Frequency Index (QFI)

QFI of alcohol use was derived from questions related to the quantity and frequency with which participants consumed various types of alcohol (beer, wine, hard liquor) in the past six months. After applying standard conversion formulas (see Armor and Polich, 1982) for the average amount of ethanol in each of the various beverage types, we obtained a measure of the average ounces of absolute alcohol consumed per day over the past six months.

Heavy Episodic Drinking (HED)

The frequency of HED was assessed by items asking for the number of days during the past six months that the person reported consuming six or more drinks during a drinking episode. (Note that the use of this threshold for HED was used prior to more recent recommended cut-points of 5 drinks for males and 4 drinks for females.)

Density of Alcohol-Using Friends

Adolescents were requested to indicate the number of persons they considered friends and were then requested to indicate how many of these friends drank alcohol regularly. Percentage scores were calculated by dividing the number of regularly-drinking friends by the total number of friends and multiplying by 100, with a possible range of 0% to 100%.

Delinquent Activities

Sixteen items were used to assess adolescents’ frequency of involvement in various delinquent activities over the past 6 months (Elliott, Huizinga, & Menard, 1989). A composite score was calculated by summing the 16 items; alphas ranged from .78–.82.

Marijuana Use

The frequency of using marijuana/hashish during the past 6 months was assessed with a 7-point Likert scale item that ranged from 1=never used to 7=used every day.

Boredom Susceptibility

Adolescents’ propensity to boredom susceptibility was measured with the 10-item Boredom Susceptibility Subscale (ZBS) from Zuckerman’s Sensation Seeking Scale, Form V (SSS-V; Zuckerman, Eysenck, & Eysenck, 1978). Reliability estimates across 21 studies was .61 (Deditius-Island and Caruso, 2002); in this study it was .58.

Depressive Symptoms

Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977). The CES-D consists of 20 self-report items and provides a unitary measure of current depressive symptomatology; alphas exceeded .90.

Stressful Life Events (SLEs)

The list of life events was adopted from the Adolescent Life Change Scales (ALCES; Yeaworth, York, Hussey, Ingle, & Goodwin, 1980). Adolescents evaluated a total of 29 events for their occurrence-nonoccurrence within the past 6 months (scores could range from 0–29).

Family Emotional Support

The 20-item Perceived Social Support from Family measure (PSS-Fa) was used to assess the amount of perceived emotional support provided by the family (Procidano & Heller, 1983); alphas ranged from .95–.96.

Grades

Adolescents were asked “What grades do you usually get in school?”. This item has a 7-point response format from “mostly As” to “mostly Ds and Fs”. The Pearson correlation between adolescents’ self-reports of grades and official high school records (that used a somewhat different measurement scale) was .78.

Young Adult Drinking Motives

Young adults completed a 15-item self-report measure (Cooper, Russell, Skinner, & Windle, 1992) that assessed coping, social, and enhancement motives for drinking. In this study, average alphas across Waves 5–7 were .86 for coping motives, .83 for social motives, and .90 for enhancement motives.

3. Results

Table 1 provides correlations, means, and standard deviations for the adolescent predictors and Waves 5–7 drinking motives. While the inter-correlations among the drinking motives are substantial, prior research has indicated distinct antecedents and alcohol-related outcomes for each motive (Cooper, 1994; Knutsche et al., 2005).

Table 1.

Correlation matrix for latent trait-state model variables

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20
1. Sex	1.0
2. A-QFI	−.10	1.0
3. A-HDE	−.19	.79	1.0
4. A-Pot Use	−.04	.54	.52	1.0
5. A-Density Alcohol-Using Friends	.07	.37	.27	.18	1.0
6. A-Delinquency	−.14	.52	.51	.50	.25	1.0
7. A-Boredom Susceptibility	−.20	.21	.20	.18	.22	.23	1.0
8. A-Depressive Symptoms	.18	.13	.10	.07	.08	.19	.01	1.0
9. A-SLEs	.12	.33	.29	.23	.24	.39	.10	.44	1.0
10. A-Family Support	.04	−.14	−.13	−.11	−.06	−.30	−.23	−.35	−.27	1.0
11. A-Grades	.17	−.29	−.30	−.23	−.16	−.36	−.10	−.23	−.29	.24	1.0
12. Y5-Coping	−.08	.19	.20	.16	.14	.20	.12	.22	.23	−.17	−.12	1.0
13. Y5-Social	−.15	.19	.20	.13	.19	.18	.18	.09	.20	−.08	−.10	.56	1.0
14. Y5-Enhance	−.16	.25	.26	.16	.21	.23	.20	.10	.18	−.07	−.13	.70	.70	1.0
15. Y6-Coping	−.04	.27	.24	.15	.20	.15	.15	.18	.23	−.14	−.14	.63	.41	.49	1.0
16. Y6-Social	−.12	.21	.20	.11	.22	.09	.22	−.00	.17	−.11	−.05	.36	.62	.45	.52	1.0
17. Y6-Enhance	−.12	.31	.30	.19	.24	.17	.23	.06	.22	−.11	−.08	.51	.51	.69	.70	.66	1.0
18. Y7-Coping	−.04	.28	.23	.16	.14	.16	.14	.19	.21	−.18	−.05	.64	.37	.51	.70	.38	.56	1.0
19. Y7-Social	−.08	.24	.20	.13	.20	.17	.18	.11	.19	−.14	−.07	.41	.51	.48	.49	.62	.58	.61	1.0
20. Y7-Enhance	−.08	.32	.28	.23	.21	.23	.19	.11	.21	−.15	−.10	.51	.44	.63	.55	.50	.72	.76	.75	1.0
M	1.53	10.91	1.11	0.98	70.67	7.90	12.96	14.86	5.90	57.25	5.02	8.37	12.84	10.31	8.47	13.23	10.38	9.10	11.85	10.46
SD	0.50	17.10	2.91	0.99	32.21	11.94	1.86	8.67	2.83	12.77	1.25	2.81	3.02	3.52	2.75	2.94	3.42	2.93	2.98	3.03

Notes: Sex: 1=Male, 2=Female; A=Adolescence; Y5=Young Adult, Wave 5; Y6=Young Adult, Wave 6; Y7=Young Adult, Wave 7; QFI=Quantity-Frequency Index; HDE=Heavy Drinking Episodes; SLEs=Stressful Life Events. In the analyses, log transformed scores were used for QFI, HDE, and Marijuana (Pot) use.

N=1,004

3.1 Latent trait state models (LTSMs)

The LTSM was used to obtain a general latent trait score for each of the three drinking motives across three occasions of measurement. This was accomplished by decomposing repeated measures data into components representing both stable (trait) and fluctuating (state) features of drinking motives while removing random error. As depicted in Figure 1, three LTSMs were specified (one for each motive) so that three-wave manifest indicators of each motive “loaded” on a trait factor, and a parameter corresponding to a state factor was estimated for each occasion of measurement. An autoregressive structure was also estimated for the state factors in the event that there was reliable cross-occasion state covariance that was not captured by the common trait factor. The factor loadings between the manifest indicators and trait factors presented in Table 2 were quite high, and the autoregressive effects for the state factors were, in some instances, statistically significant but in all instances indicated low effect sizes. As such, the decomposition of effects for the three drinking motives indicated a high percentage of trait variance.

Figure 1.

Conceptual Schematic of a Latent Trait-State Model of Young Adult Drinking Motives and Adolescent Predictors

Table 2.

Factor Loadings for Manifest Indicators and Trait Factors, and Autoregressive Effects for State Factors for Coping, Social, and Enhancement Drinking Motives¹

Drinking Motives	Standardized Factor Loadings	Autoregressive Effects
Coping Motives	W5: .77 W6: .79 W7: .74	W5–W6: .02 W6–W7: .10*
Social Motives	W5: .68 W6: .70 W7: .69	W5–W6: .14* W6–W7: .06
Enhancement Motives	W5: .75 W6: .78 W7: .87	W5–W6: .11* W6–W7: −.03

¹Models were constrained to equivalence

N=1004

^*p<.001

3.2 Adolescent Predictors of Young Adult Drinking Motives

The ten exogenous predictors were specified for each of the three LTSMs to evaluate the prediction of the trait factor. Statistically significant adolescent predictors for each of the three drinking motives are provided in Table 3, along with the model estimated R² values. The model fit statistics using MLR estimation for coping motives were χ² with 22 df =74.94, RMSEA=.049, CFI=.965, SRMR=.012. The model fit statistics for social motives using MLR estimation were χ² with 22 df =77.94, RMSEA=.050, CFI=.954, SRMR=.015. The model fit statistics for enhancement motives using MLR estimation were χ² with 22 df =83.20, RMSEA=.053, CFI=.964, SRMR=.013.

Table 3.

Adolescent Predictors of Young Adult Coping, Social, and Enhancement Motives for Latent Trait-State Models

Adolescent Predictors	Coping Motives Trait	Social Motives Trait	Enhancement Motives Trait
Sex1	−.09*2	−.14***	−.12**
Quantity-Frequency Index	.15**	.10	.13*
Heavy Episodic Drinking	.07	.08	.11*
Density of Alcohol-Using Friends	.08*	.17***	.14***
Delinquent Activities	−.05	−.07	−.03
Marijuana Use	.04	.01	.05
Boredom Susceptibility	.07*	.14***	.13***
Depressive Symptoms	.16***	−.01	.03
Stressful Life Events	.13**	.18***	.14***
Family Emotional Support	−.09*	−.07	−.04
Grades	.06	.07	.07*
R2	0.182***	0.190***	0.214***

¹Males=0; Females=1;

²Standardized Coefficients are provided;

N=1004

^*p<.05;

^**p<.01;

^***p<.001

There were several adolescent predictors that were common across the drinking motives, and three of these general predictors supported hypothesized relationships and were consistent with the literature, including male sex, alcohol use (coping and enhancement motives but not social motives), and density of alcohol-using friends. Stressful events were hypothesized to uniquely predict coping motives, but findings indicated that they were universal in predicting all three motives.

In addition to these general predictors, mixed results were evident for the specific hypothesized relationships for coping motives (internalizing pathway) and enhancement motives (externalizing pathway). As hypothesized, coping motives were predicted uniquely by adolescent depressive symptoms and lower family support. The only unique significant hypothesized predictor of enhancement motives was HED. Contrary to hypotheses, grades had a unique positive relationship with enhancement motives.

4. Discussion

With regard to the first study aim, the LTSM provided an adequate statistical model representation of each of three drinking motives across young adulthood, with these three motives manifesting high stability across individuals with regard to rank-ordering. For example, those individuals who were in the upper portion of the distribution for coping motives at age 23.8 years, were also in the upper portion at age 33.5 years. Hence, young adulthood may be a period in the lifespan when motives for drinking are fairly trait-like.

The second study aim was to include ten exogenous variables in the LTSM analyses and test their specific and common prospective relationships to the three drinking motives. For coping motives, as hypothesized, adolescent higher depressive symptoms and lower family emotional support uniquely predicted these motives. Stressful life events also significantly predicted coping drinking motives but, counter to our hypothesis, they also predicted enhancement and social motives. The findings for coping motives are consistent with both theoretical work and empirical evidence that has supported an internalizing alcohol motives pathway to alcohol use (Cooper et al., 2016; Hussong et al., 2011; Windle & Windle, 2015).

With regard to specificity hypotheses related to enhancement drinking motives, the adolescent factors of higher HED and higher school grades uniquely predicted these motives. The HED finding is consistent with our hypotheses and prior research (Cooper et al., 1995; Kuntsche et al., 2006) which suggests that engagement in heavy drinking episodes is perceived to enhance positive emotions and desirable outcomes. The positive association between grades and enhancement motives was contrary to our hypotheses. A possible explanation for this unexpected finding is that a third variable, such as personality (e.g., extraversion, intellect/imagination; Theakston et al., 2004), may reflect a common association between higher grades and higher enhancement motives.

In contrast to our hypothesis that boredom susceptibility would be uniquely associated with enhancement motives, it was a significant general predictor of all three motives. Boredom susceptibility as defined by the ZBS is a dispositional style characterized by “the tendency to be underaroused because of an impoverished or understimulating environment” (Mercer-Lynn, Flora, Fahlman, & Eastwood, 2011, p. 586), and has been associated with higher levels of motor impulsivity and sensitivity to reward, alcohol and drug problems, and gambling (Mercer-Lynn et al., 2011; Vodanovich & Watt, 2016). It may be that feelings of boredom motivate the use of alcohol for emotional regulatory functions to reduce the discomfort of boredom and thus predicts all three drinking motives.

Two adolescent precursors hypothesized to be uniquely predictive of enhancement motives were delinquency and marijuana use; however, contrary to our expectations, these variables were not significantly associated with enhancement motives. A possible explanation for these findings is that this relatively highly-educated, middle-class sample perpetrated phasic, age normative delinquency (i.e., adolescence-limited; Moffitt, 1993) that did not “spill over” and impact enhancement motives in young adulthood. It is further possible that marijuana use is uniquely related to motives for using marijuana and less so to motives for using alcohol (Cooper et al., 2016).

The findings supported hypothesized relationships with general predictors across the three drinking motives. Consistent with the literature (Cooper, 1994; Kuntsche et al., 2006), being male predicted greater variation in drinking motives, though explanations for male and female divergence in drinking motives during this period of the life-span are unknown. To the extent that drinking motives and drinking behaviors coincide (Cooper et al., 2008; Littlefield et al., 2010; Patrick & Schulenberg, 2011), a range of biological (e.g., hormonal effects) and psychosocial (socialization factors) explanations have been posited for gender differences in alcohol use (Nolen-Hoeksema, 2004; Witt, 2007), and perhaps some of these generalize to motives for drinking. The importance of a higher density of alcohol-using friends during adolescence also predicted higher levels of all three drinking motives in young adulthood and mirrors findings of the integral role that peers play in adolescents’ and young adults’ alcohol consumption (Andrews, Tildesley, Hops, & Li, 2002; Van Ryzin & Dishion, 2014).

There are several intervention implications based on findings from the current study. First, the identification of adolescent precursors of stable drinking motives across young adulthood suggests that interventions during the teenage years aimed at preventing or interrupting the development of cognitive motivations that encourage alcohol use for the purpose of affect regulation may mitigate the development and crystallization of drinking motives in young adulthood. Second, while the LTSMs indicated that drinking motives were best characterized as more trait-like than state-like, it is nevertheless the case that immediate contextual circumstances influence drinking motives and behaviors during specific drinking episodes (Cooper et al., 2016). Research utilizing daily process methodologies is addressing this issue by focusing on both state and trait motivations for drinking during episode-specific situations (Arbeau et al., 2011; O’Hara, Armeli, & Tennen, 2015). This research is affirming the complex nature of the drinking motives—alcohol use relationship, especially to the extent that within-context factors (e.g., current mood and stressors, drinking alone vs. with friends) may moderate this relationship. Therefore, cognitive and behavioral strategies aimed at harm reduction within the context of the drinking situation (e.g., protective strategies such as alternating alcoholic and non-alcoholic drinking, sipping rather than gulping or chugging drinks) may be effective at decreasing alcohol use and, by extension, adverse health consequences associated with heavy use (Martens et al., 2005).

4.1 Study Limitations

A limitation of this study was that the sample was primarily Caucasian and middle-class; findings may not generalize to other racial/ethnic groups or across a broader SES spectrum. Another limitation was that the measures relied upon self-report by the participants and mono-method bias may have influenced the findings. Our longitudinal design with 5-year intervals between assessments may have masked more rapid changes; research utilizing daily data collection methods may be more sensitive to alcohol use and motives during episode-specific drinking situations (Arbeau et al., 2011; O’Hara et al., 2015). Finally, recent literature has suggested the potential utility of decomposing a general coping motives construct into specific constructs of coping motives-depression and coping motives-anxiety (Bravo & Pearson, 2017; Grant, Stewart, & Mohr, 2009; Grant, Stewart, O’Connor, Blackwell, & Conrod, 2007). We did not assess these affect-specific aspects of coping motives and therefore were unable to address this issue in the current study. Future research is needed to address the long-term stability of these separate constructs and their general and specific precursors.

Highlights.

• Drinking motives were highly stable across a ten-year period in young adulthood.

• Adolescent factors were significant predictors of young adult drinking-motives.

• Common predictors included alcohol-using peers, life stressors, and boredom.

• Depression was a unique predictor of coping motives.

• Binge drinking was a unique predictor of enhancement motives.