Motivations to combine alcohol and nicotine in college students: A validation of the Alcohol and Nicotine Motives Scale.

Abstract

Introduction:

Research on combining alcohol and nicotine (ALCNIC) has shown this risky behavior results in significantly more consequences than using either alcohol or nicotine alone. No measures currently exist to assess ALCNIC motives limiting intervention and prevention efforts. The present study developed a psychometrically sound and multidimensional measure of ALCNIC motives (the ANMS).

Methods:

An initial item pool of ALCNIC items was developed from literature on college student drinking, focus groups, and individual interviews. Study 1 involved students from a northeastern university who completed an online survey on the ALCNIC items (N=55; 57.1% female; Mage=20.3). Analysis focused on reliability (exploratory factor analysis). Study 2 involved a cross-validation national sample of college students (N=336; 49.7% female; Mage=21.2) completing the same survey items. Confirmatory factor analysis, criterion-related validity (ALCNIC/weekend drinking), and discriminant validity (social desirability) were assessed using structural equation modeling.

Results:

Results across two studies revealed three factors to engage in ALCNIC: antagonistic (party longer), synergistic (enhanced effects), and social (peer pressure); and one factor to avoid ALCNIC: negative effects (feeling anxious) (all alphas > 0.7). In study 2, criterion-related validity revealed that synergistic motives were significantly positively associated with ALCNIC use; and negative effects motives were significantly negatively associated with ALCNIC use. Discriminant validity showed ALCNIC subscales were not significantly associated with social desirability (except social).

Conclusions:

The study developed a reliable and valid measure of motives for ALCNIC use. Results were robust to cross-validation across two samples of college students. These measures provide targets for intervention and prevention efforts.

1. Introduction

Universities have focused their efforts on preventing risky drinking, despite evidence that alcohol is often not used in isolation and co-use of alcohol and other substances results in significantly more negative consequences than alcohol use alone (Lee et al., 2020; Mallett, Turrisi, Hultgren, Reavy, & Cleveland, 2017; Mallett, Turrisi, Trager, Sell, & Linden-Carmichael, 2019; Roberts et al., 2018). Nicotine use, either by cigarettes or e-cigarettes (e-cigs), has consistently been associated with risky alcohol use among college students (e.g., Jackson, Colby, & Sher, 2010; Littlefield, Gottlieb, Cohen, & Trotter, 2015; McKee, Hinson, Rounsaville, & Petrelli, 2004; Roys et al., 2020). Research examining cigarette use and alcohol indicates that students often smoke and use alcohol simultaneously within the context of an episode of alcohol consumption (Jackson et al., 2010; McKee et al., 2004). Research examining e-cigs, found that college students and emerging adults who engage in simultaneous use of alcohol and e-cigs are at high risk for engaging in problematic alcohol use and/or experiencing negative alcohol-related outcomes (Hefner, Sollazzo, Mullaney, Coker, & Sofuoglu, 2018; Littlefield et al., 2015; Mallett et al., 2017; Roberts et al., 2018). These studies provide substantive evidence of the high prevalence and problematic effects of simultaneous alcohol and nicotine use (ALCNIC) among college students and support further investigation into ALCNIC use to understand these alarming trends.

1.1. Alcohol and Nicotine Motives

A wealth of research has focused on the relationship between drinking motives and drinking behavior. Cooper’s (1994) confirmatory factor analysis of the Drinking Motives Questionnaire among a large sample of adolescents resulted in a four-factor model of drinking motives (e.g., social, coping, enhancement, conformity). Cross-sectional and longitudinal studies have shown consistent significant associations between these drinking motives and alcohol use among adolescents, young adults, and college students (Cooper, 1994; Cooper, Frone, Russell, & Mudar, 1995; Read, Wood, Kahler, Maddock, & Palfai, 2003).

Similarly, numerous studies have examined the relationship between smoking motives and smoking (i.e., nicotine use). For example, non-daily (or intermittent) smokers differ from daily smokers on the Wisconsin Inventory of Smoking Dependence Motives (Piper et al., 2004), with non-daily smokers emphasizing motives associated with acute, situational smoking, and daily smokers emphasizing dependence-related motives (Shiffman, Dunbar, Scholl, & Tindle, 2012). Among adolescents, social and affect regulation motives have been found to be significant predictors of nicotine use with non-daily smokers tending to smoke almost exclusively with friends (Li et al., 2003; Piko, Varga, & Wills, 2015). Among college students, latent class analysis has been used to identify nicotine use groups by motives: 1) addicted smokers, who use due to pleasure and habit/addiction; 2) stress smokers, who use to relax, reduce stress, or regulate mood; and 3) social smokers, who use because of social factors to fit in (Rosa, Aloise-Young, & Henry, 2014).

Despite the high prevalence of ALCNIC use among college students (Harrison et al., 2009; McKee & Weinberger, 2013; Reed et al., 2007), limited literature exists examining motives for ALCNIC. The present study was designed to fill this gap by developing a psychometrically reliable, valid, and multidimensional ALCNIC Motives Scale (the ANMS). The theoretical model guiding the present study is similar to the rationale offered in Cooper (1994) in the development of the Drinking Motives Questionnaire: people are motivated to use substances (e.g., ALCNIC) to attain certain outcomes and meet different needs or functions. Further, themes that came out of the research on combining alcohol and energy drink motives (e.g., energy/endurance, intoxication-reduction, hedonistic; Droste et al., 2014) and marijuana motives (e.g., social, conformity; Patrick, Fairlie, & Lee, 2018) helped guide our hypotheses on the specific types of motives that would emerge. However, one of the unanswered questions was motives students had to avoid combining alcohol and nicotine, which was added and tested in the current study.

We hypothesize there are theoretically four different types of ALCNIC motives that fit the categories of attaining intended outcomes and meeting different needs or functions: 1) antagonistic - individuals seek to offset sedative effects of alcohol only; 2) synergistic - individuals seek to experience agonist effects that result from ingesting ALCNIC together; 3) social - individuals seek to fit in with their peers’ use of ALCNIC; and 4) negative effects – avoid feelings of anxiety from ALCNIC. Our goal is to create a measure that captures broad motives regarding ALCNIC behavior to disseminate in future research focusing on ALCNIC.

1.2. Development of the Alcohol and Nicotine Motives Scale (ANMS)

The research to develop the ANMS was conducted across two studies. Study 1 examined a focus group sample (n=22) to help generate items for the measure and a small pilot sample (n=55) to examine the preliminary psychometric data (reliability and validity) to capture college students’ motives for ALCNIC use.

Principal components factor analysis was used to create the subscales for the measure. We hypothesized specific ALCNIC motives would emerge: antagonistic, synergistic, social, and negative effects. Reliability was assessed by examining the internal consistency of each subscale. Study 2 examined a larger national sample of college students (n=336) to confirm the measure developed in Study 1. A confirmatory factor analysis was conducted using structural equation modeling to assess the measure in Study 2. Reliability and validity (criterion-related, discriminant) were assessed by examining paths in the structural equation models. We hypothesized that the: 1) factor structure observed in Study 1 would be replicated, 2) ALCNIC motives would be directly associated with ALCNIC use and indirectly associated with weekend drinking (criterion-related validity), and 3) ALCNIC motives would not be associated with social desirability (discriminant validity).

2. Materials and Methods

2.1. Study 1

2.1.1. Participants & Procedures

Focus Group Sample.

An initial pool of new items for ALCNIC motives was created based on a focus group and an open-ended online survey with college students recruited from a large northeastern university. Students were recruited via emailed invitations through the university’s registrar list (n=350). To be eligible for participation in the study, students had to meet the following criteria: 1) be between the ages of 18–23, 2) be currently enrolled as a traditional freshman, sophomore, junior, or senior, 3) report consuming alcohol at least once in the past 90 days, and 4) report a lifetime history of combining alcohol and another substance(s) so that the effects overlapped. Eligible participants (77.3% female; 90.9% white; M_age=20.3) were either invited to a focus group in-person (n=7) or an online open-ended survey (n=15). Participants were asked two separate open-ended questions as follows: “Please list all the reasons you can think of why students may [choose to use/avoid using] both alcohol and nicotine.” To complement the items that came out of the focus groups and open-ended surveys, we conducted a review of the literature on college students’ motives to combine alcohol with other commonly used substances (e.g., energy drinks and marijuana). We added items to the initial pool of items that fell under the categories of energy/endurance, intoxication-reduction, hedonistic, social, and conformity as found in Droste et al. (2014) and Patrick et al. (2018).

Preliminary Psychometric Pilot Sample.

Students were recruited via emailed invitations through the university’s registrar list at a large northeastern university (n=1988). Participants were routed directly to a baseline screening survey to assess their eligibility (the same eligibility criteria for the focus group participants). A total of 237 (11.9%) met the eligibility criteria. Of these, 55 individuals who reported a history of ALCNIC use (defined as combining alcohol and a nicotine product on one occasion so their effects overlap) were included to complete the motives measure surveys (see Table 2).

Table 2.

Demographics and Descriptives

	Study 1 (n=55)		Study 2 (n=336)
	Mean (SD) or Percentage	Range	Mean (SD) or Percentage	Range
Demographics
Age	20.3 (1.3)	18–23	21.2 (1.5)	18–23
Sex (% female)	57.1%	-	49.7%	-
Race (% white)	92.9%	-	72.9%	-
Ethnicity (% Hispanic)	1.8%	-	11.9%	-
Social Desirability	2.8 (0.5)	1–5	2.4 (1.0)	0–5
Substance Use
Typical Weekly Drinking	11.5 (8.4)	0–40	13.1 (9.6)	0–52
ALCNIC use	2.0 (0.8)	1–4	6.2 (5.1)	0–15

Note for Study 1: Response options for Social Desirability are as follows: 1=Strongly disagree, 2=Moderately disagree, 3=Neither disagree nor agree, 4=Moderately agree, 5=Strongly agree. ALCNIC use defined as past month frequency of combined alcohol and nicotine use so the effects overlapped: 1=Never, 2=Once or twice, 3=Weekly, 4=Daily or almost daily.

2.1.2. Measures

2.1.2.1. Motives for ALCNIC

Participants received 24 items to assess their motives to combine ALCNIC and were prompted with, “Below is a list of reasons people sometimes give for combining alcohol and nicotine (so that their effects overlap). Thinking of all the times you combined alcohol and nicotine, how often would you say that you combined alcohol and nicotine for the following reasons? (e.g., feel energized)”. Nicotine use was described to participants as cigarette, chewing tobacco, cigars, vaping, etc. In addition, participants received 20 items to assess their motives to avoid combining ALCNIC within the past month or 90 days and were prompted with, “Below is a list of reasons people sometimes give for NOT combining alcohol and nicotine so that their effects overlap. How often would you say that you avoided combining alcohol and nicotine for each of the following reasons? (e.g., makes you feel anxious)”. Response options were measured on a 5-point scale (1=Almost Never/Never, 2=Some of the time, 3=Half of the time, 4=Most of the time, 5=Almost Always/Always).

2.1.3. Analytic Procedures

2.1.3.1. Creation of Measure

Principal components analysis was conducted to identify distinct factors within the ALCNIC motives measure. Factor loadings and inter-item correlations were examined to determine if items should be retained and Cronbach’s alpha was used to test the internal consistency of the subscales. For the creation and development of the subscales, items were removed if the: 1) factor loadings were <0.7, 2) inter-item correlations <0.5, and 3) Cronbach’s alphas would improve to be >0.7.

2.1.4. Results

2.1.4.1. Principal Components Analysis (PCA)

The identification of factors and item reduction was determined through PCAs performed separately for motives to engage in and avoid ALCNIC. Analyses across both samples revealed three factors to engage in ALCNIC: 1) antagonistic (party longer), 2) synergistic (enhance effects), and 3) social (peer pressure). One factor was generated to avoid ALCNIC: 1) negative effects (feel anxious). Items were kept for each factor if they met the retention criteria described in the analytic procedures section (e.g., factor loadings >0.7; correlations >0.5), resulting in 11 of the original 44 item set (24 to engage in and 20 to avoid ALCNIC). See Table 1 for subscales, items, factor loadings, and alphas for the measure.

Table 1.

Alcohol and Nicotine Motives Scale (Principal Components & Confirmatory Factor Analysis)

Subscale	Item	Study 1 (N=55)		Study 2 (N=336)
		Factor Loadings	Alpha	Factor Loadings	Alpha
1. Antagonistic	Prevent feeling tired/falling asleep	0.89	0.92	0.79	0.86
	To feel energized	0.88		0.84
	To stay awake longer	0.81		0.85
2. Synergistic	Nicotine gives a slight head-high	0.86	0.85	0.72	0.83
	To enjoy the buzz	0.84		0.92
	To feel good	0.76		0.74
3. Social	Peer pressure	−0.88	0.83	0.83	0.75
	To be cool	−0.92		0.72
4. Negative Effects	Makes you feel shaky	0.91	0.92	0.79	0.81
	Makes you feel anxious	0.95		0.79
	Because you won’t sleep well	0.91		0.70

2.2. Study 2

2.2.1. Participants & Procedures

The second sample was recruited through Amazon Mechanical Turk using the same eligibility criteria as Study 1. A national sample was obtained by recruiting participants in five waves by U.S. region: 1) northeast, 2) southeast, 3) southwest, 4) northwest, and 5) central. A total of 5,899 people completed screening with the following demographic breakdown: 53.9% female, 69.6% white, 12.1% Hispanic, 42.7% current students, 73.0% drank in past 90 days, 39.6% endorsed lifetime ALCNIC use, and M_age=30.6. Of those, 511 (8.7%) met eligibility criteria and consented to participate in the main survey. Check questions were included in the survey to detect participants who answered inconsistently (e.g., check ‘True’ for the “The sky is blue”). To ensure that the respondents were indeed U.S. college students, participants were required to provide the full name of the U.S. college/university, school mascot, and city/state in both the screening and main survey. Participants were removed from the final analytic sample if the information provided was incorrect or any discrepancies were detected between answers provided in screening versus the main survey. Only participants who answered the check questions and U.S. college/university questions correctly and consistently, and reported a history of ALCNIC use were included in the final analytic sample (n=336; see Table 2).

2.2.2. Measures

2.2.2.1. Motives for ALCNIC

The 11 items generated from the principal components analysis in Study 1 were presented to participants in Study 2. See Table 1 for items.

2.2.2.2. Alcohol Use

Typical weekend drinking was assessed with the Daily Drinking Questionnaire (DDQ; Collins, Parks, & Marlatt, 1985), which asks respondents to estimate how many drinks they consume on each day of a typical week. To assist with drink estimates, a standard drink was defined as 12 oz. of beer, 8–9 oz. of malt liquor, 5 oz. of table wine, or a 1.5 oz. shot of distilled spirits (NIAAA, 2016). Item responses for Friday and Saturday were summed to obtain an estimate of typical weekend drinking. Participants were asked to report typical weekend drinking in the past 90 days (M_drinks=8.3; SD_drinks=5.4; see Table 2 for typical weekly drinking).

2.2.2.3. ALCNIC Use

To capture ALCNIC use, participants were asked, “During the past 90 days, when you were drinking alcohol, how often did you use alcohol in combination with nicotine (that is, so that their effects overlap)?” Response options ranged from (0) 0 times to (15) 15+ times (see Table 2).

2.2.2.4. Social Desirability

Social desirability was assessed using five items adapted from the Balanced Inventory of Desirable Responding (BIDR; Paulhus, 1991). Typical alphas for this scale range from 0.67 to 0.85. Example items include “I never swear” and “I take out my bad mood on others now and then.” Response options were (1) True and (0) False. Items were summed to compute a composite social desirability score (see Table 2).

2.2.3. Analytic Procedures

2.2.3.1. Model Fit

A variety of fit indices were evaluated. The root mean square error of approximation (RMSEA) was tested to determine if RMSEA<0.05 and the p-value for the test of close fit was non-significant. The comparative fit index (CFI) was examined to assess variance explained in the outcome. The root mean square residual and the largest absolute discrepancy between a predicted and an observed correlation were also observed. Inspection of the residuals and modification indices were assessed to determine any ill-fit in the model.

2.2.3.2. Confirmatory Factor Analysis

The measurement model from Study 1 was examined in a confirmatory factor analysis to assess associations between the latent constructs and the observed measures of each construct. Path coefficients and residuals were assessed to determine whether Study 2 data provide evidence to support the factor structure in Study 1.

2.2.3.3. Criterion-Related Validity (ALCNIC and Weekend Drinking)

Direct effects were observed for the following associations: 1) each ALCNIC motive and ALCNIC use, and 2) ALCNIC use and weekend drinking. Indirect effects of each ALCNIC motive on weekend drinking (through ALCNIC use) were also observed. Statistically significant associations between ALCNIC motives and ALCNIC use (and weekend drinking) provide support for criterion-related validity of those latent variables. The model controlled for age, birth sex, race, ethnicity, and social desirability.

2.2.3.4. Discriminant Validity (Social Desirability)

Direct effects of each ALCNIC motive on social desirability were observed. Non-significant associations between ALCNIC motives and social desirability provide support for discriminant validity of those latent variables. The model controlled for age, birth sex, race, and ethnicity.

3. Results

Using a structural equation-modeling framework, each ALCNIC motive was conceptualized as a latent variable with multiple indicators (i.e., items from the subscale). The structural and measurement models implied by this approach are illustrated in Figure 1. The model is statistically over identified and a variety of fit indices were evaluated. The RMSEA was 0.05, with 90% confidence intervals of 0.03 and 0.06. The p-value for the test of close fit (RMSEA<0.05) was 0.78. The comparative fit index (CFI) was 0.95. The root mean square residual was 0.06. The indices point toward good model fit.

Figure 1.

Structural and Measurement Model (Study 2; N=336)

3.1. Confirmatory Factor Analysis

Examination of the measurement model revealed strong associations between the latent constructs and the observed measures of each construct (see Table 1 for estimates of paths; all path coefficients >0.7). These data provide evidence supporting the factor structures observed in Study 1.

3.2. Criterion-Related Validity

Together the four latent factors accounted for 17% of the variance in ALCNIC use and 14% of the variance in weekend drinking. Examination of the direct effects of ALCNIC motives reflective of criterion-related validity revealed a statistically significant positive association between synergistic motives and ALCNIC use. A significant negative association was also observed between negative effect motives and ALCNIC use. Direct effects of ALCNIC use on weekend drinking revealed a statistically significant positive association. Further, indirect effects (through ALCNIC use) showed the same significant positive association between synergistic motives and weekend drinking, and a negative association between negative effect motives and weekend drinking. No other statistically significant effects were observed. See Figure 1 for the betas associated with each path.

3.3. Discriminant-Validity

The analysis examined the relationship between social desirability and the four different ALCNIC motives, using a similar structural equation-modeling framework described above. The RMSEA was 0.05, with 90% confidence intervals of 0.03 and 0.06. The p-value for the test of close fit (RMSEA<0.05) was 0.72. The CFI was 0.96 and the root mean square residual was 0.05. The indices point toward good model fit. Examination of the unique influences of the ALCNIC motives reflective of discriminant-related validity only revealed a statistically significant positive association between social motives and ALCNIC use (b=−0.24; 95% CI [−0.49, −0.04]). No other statistically significant effects were observed. Correlations between factors and social desirability were all small ranging from −0.19 to 0.04.

4. Discussion

The present research was designed to fill a gap in the literature by developing a reliable and valid ALCNIC Motives Scale (the ANMS). Study 1 hypothesized that we would observe four theoretically different types of ALCNIC motives: 1) antagonistic (to stay awake or party longer); 2) synergistic (feel good in a way that would be different from alcohol-only or nicotine-only use); 3) social (peer pressure), and 4) negative effects (feel anxious). We observed empirical support for four internally consistent subscales consistent with our hypotheses (factor loadings were >0.7; all Cronbach’s alphas >0.7). Study 2 hypothesized that the: 1) factor structure would be replicated, 2) ALCNIC motives would be directly associated with ALCNIC use and indirectly associated with weekend drinking (criterion-related validity), and 3) ALCNIC motives would not be associated with social desirability (discriminant validity). The confirmatory factor analysis provided additional empirical support by replicating the high internal consistency of the factors.

We observed partial support for the criterion-related validity hypothesis. The synergistic and negative effects factors were significantly associated with ALCNIC use and weekend drinking, but the antagonistic and social factors were unrelated to ALCNIC use. These findings suggest that synergistic and negative effects may be the key motives to focus on in future studies. Students seem to be most motivated to engage in ALCNIC use based on the synergistic effects such as enjoying the buzz and feeling good. Alternatively, they also tend to avoid ALCNIC use because they don’t want to feel anxious, shaky, or have difficulty sleeping. While these specific motives seem to be on opposite ends of a continuum, the factors were not highly correlated and each had unique influences on ALCNIC. These findings provide empirical support that they are operating independently.

The utility of these two motives of ALCNIC can be explored within the context of targeting theoretical constructs proximal to behavioral outcomes. The associations between motivational constructs and substance misuse (e.g., ALCNIC), and the theoretical mediating effects of decision-making constructs has been the topic of numerous papers and theories (e.g., Gibbons, Gerrard, & Lane, 2003; Gibbons & Gerard, 1997). For example, according to Gibbons and Gerrard (2003) behavioral willingness theory, students’ willingness to engage in ALCNIC use will be theoretically influenced by each of these motives, which will in turn, influence ALCNIC behavior. If students are strongly motivated to engage in ALCNIC use for synergistic reasons, and do not experience the negative effects (shakiness, anxiety, sleep problems), they are theoretically likely to be more willing to engage in ALCNIC. In contrast, if students are strongly motivated to avoid engaging in ALCNIC because of the negative effects (shakiness, anxiety, sleep problems), they are theoretically less likely to be willing to engage in ALCNIC. Understanding these motives in a theoretical context as they relate to students’ decision-making can be important for intervention efforts facilitating reductions in ALCNIC use (e.g., personalized feedback interventions that explore students’ behavioral willingness and ALCNIC motives).

Study 2 demonstrated partial support for discriminant validity by the non-significant associations between the subscales and social desirability. The lone exception was that the social subscale was observed to have a small, but significant association with social desirability. However, this subscale did not significantly predict ALCNIC use. It is not surprising this subscale was related to social desirability considering it focuses on motives to engage in ALCNIC use “to be cool” and as a result of feeling peer pressure.

4.1. Limitations and Future Directions

The current study examined ALCNIC behaviors as the number of times used in the past 90 days. While this is a common approach, it might be beneficial to utilize more nuanced measures of ALCNIC use that consider context. For example, individuals may engage in ALCNIC use when they are with a certain group of friends, if they are in certain locations, or on specific days or times. It is also plausible that the associations between the motives and ALCNIC use might differ between combustible cigarette vs. e-cig users. However, it has been shown that college students use a variety of nicotine products (Enofe, Berg, & Nehl, 2014) so those effects may not vary greatly.

The study used an observational method to generate the plausible motives for ALCNIC. This method only provides readily available reasons and may miss more implicit reasons for engaging in ALCNIC (e.g., personality, social options and motivations). Although the majority of young adults are not regular users of nicotine (Harrison & McKee, 2011; Hefner et al., 2019; Piko et al., 2015; Rosa et al., 2014), it is possible that regular nicotine users are more likely to engage in ALCNIC because of the underlying motivation to maintain regular use.

Lastly, motivations related to the social environment (e.g., interpersonal relationships, friend group dynamics) were not measured in the present study. The non-significant association between social motives and ALCNIC could be due to the way social motives were defined (‘peer pressure’ and ‘to be cool’). Future studies may consider including motives that capture the social environment to examine whether these factors are predictive of ALCNIC use.

4.2. Conclusions

In sum, the current study extends the literature by developing a reliable and valid measure to identify motives to combine alcohol and nicotine. With the recent rise in nicotine use among college student populations, the prevalence of combining alcohol and nicotine is also increasing. While measures exist that assess college students’ motives to consume alcohol-only and nicotine-only use, findings from the current study found that the ANMS is both a reliable and valid measure of motives to engage in and avoid ALCNIC use specifically.

Highlights.

• Reliable measure of motivations to combine alcohol and nicotine

• Cross-validation across two samples of college students

• Scale revealed four factors: Antagonistic, Synergistic, Social, Negative Effects

• Synergistic and Negative Effects associated with combined alcohol and nicotine use