Nonmedical use of prescription stimulants as a “red flag” for other substance use

Abstract

Background.

Nonmedical use of prescription stimulants (NMPS) has increased on college campuses during the past two decades. NMPS is primarily driven by academic enhancement motives, and normative misperceptions exist as well. However, large, nationwide studies have not yet been conducted to generalize findings more broadly and gain a deeper understanding of the relationship between NMPS and other substance use (e.g., alcohol use, marijuana, etc.). The present study was conducted to lay the foundation for prevention efforts related to NMPS by establishing NMPS prevalence, practices surrounding NMPS, and other substance use.

Methods.

N=2,989 students from seven universities around the U.S. completed a web-based survey assessing NMPS practices and related behaviors. Prevalence and factors associated with NMPS were explored.

Results.

Analyses revealed a 17% past-year prevalence of NMPS with associated widespread misperceptions of peer use. NMPS was significantly related to alcohol use, binge drinking, and marijuana use, as well as skipped classes and affiliation with Greek life.

Conclusions.

Although most college students do not report NMPS, those who do also are more likely to report alcohol use, binge drinking, and marijuana use, and NMPS could be a “red flag” for other risk behaviors worth exploring. Implications for prevention and intervention are discussed.

Nonmedical use of prescription stimulants (NMPS) is defined as the use of a medication usually prescribed to treat Attention-Deficit/Hyperactivity Disorder (ADHD) without a prescription or in a manner that was not directed by a physician (Substance Abuse and Mental Health Services Administration, 2017). Prescription stimulants can be beneficial for the treatment of ADHD (Faraone, Biederman, Spencer, & Aleardi, 2006) but have become an increasing public health concern due to their abuse potential. NMPS is most common among adolescents and young adults, particularly college students (Substance Abuse and Mental Health Services Administration, 2017). McCabe and colleagues (2014) found that past-year prevalence of NMPS among college students nearly doubled from 2003 to 2013 (5.4% to 9.3%), and, according to the National Survey on Drug Use and Health, 2.4% of full-time college students engaged in NMPS during the past month (Center for Behavioral Health Statistics and Quality, 2020). Prevalence estimates of lifetime NMPS vary widely, but researchers studying college students have found the range to be between 5% and 35% (Weyandt et al., 2013). There is also a higher prevalence of past-year nonmedical use of Adderall^© among college students when compared with their non-college peers [8.4% vs. 5.8%, respectively; (Schulenberg et al., 2020)].

NMPS among college students is associated with being white, male, and a member of a Greek organization (McCabe, Knight, Teter, & Wechsler, 2005). McCabe and colleagues (2005) also found NMPS to be more common at colleges in the Northeast and at schools with highly competitive admission standards when compared with other universities. In a longitudinal study of college students at one large, mid-Atlantic university, increasing alcohol use, marijuana use, and skipping class, as well as decreased grade point average (GPA), were all predictors of NMPS (Arria et al., 2013). NMPS is also associated with heavy alcohol use and other drug involvement, and researchers suggest NMPS seldom occurs in isolation (Arria et al., 2013; McCabe & Teter, 2007).

Students who engage in NMPS are at increased risk for several negative outcomes compared with those who do not engage in NMPS or those who use prescription stimulants medically (i.e., take the prescription stimulant as prescribed), including being more likely to use other drugs, binge drink, and drive after binge drinking (McCabe et al., 2005). McCabe and colleagues (2017) recently found that adolescents who engage in NMPS have more substance use disorder symptoms during adulthood compared with their non-using peers, suggesting that NMPS is not only associated with coincident substance use-related problems, but with problems later in adulthood as well.

Students who engage in NMPS spend less time studying and more time socializing with peers, which are patterns of behavior that would appear to hinder academic performance (Arria, O’Grady, Caldeira, Vincent, & Wish, 2008). Cross-sectional results show a similar relationship between NMPS and poor academic outcomes (McCabe et al., 2005; McCabe, Teter, & Boyd, 2006). In addition, recent longitudinal results indicated that college students who engage in NMPS show no statistically significant increases in their GPA and gain no detectable academic advantages over their peers (Arria et al., 2017).

Although incongruous with research regarding actual academic outcomes associated with NMPS, pressure to excel academically during college might contribute to the high prevalence of NMPS among college students (Forlini & Racine, 2009; Partridge, Bell, Lucke, & Hall, 2013). Researchers consistently find that the main motivation college students report for NMPS is to improve academic performance (DeSantis, Webb, & Noar, 2008; Rabiner et al., 2009). Given the high levels of perceived academic benefit of NMPS among college students (Arria et al., 2018), social norms on college campuses might become consistent with this misperception.

Perceptions of peers’ behavior have been shown to have a stronger influence on substance use behavior than family, religious, moral, and cultural factors (Borsari & Carey, 2001). Adolescents and young adults are particularly influenced by their perceptions of peer norms, even if those perceptions are not accurate (Kandel, 1985). There is a well-documented, positive relationship between perception of peer substance use and actual substance use behavior across demographic and geographic groups (McAlaney et al., 2015; Otten, Mun, & Dishion, 2017). Otten and colleagues (2017) recently found that perceptions of friends’ substance use was significantly correlated with tobacco, alcohol, and marijuana use at age 17, and perceptions of friends’ substance use at age 22 was associated with initiation of illicit drug use by age 27. Researchers have begun to document the role of perceived prevalence of peer NMPS among college students (Kilmer, Geisner, Gasser, & Lindgren, 2015; McCabe, 2008). Recently, Helmer and colleagues (2016) studied perceived peer NMPS among university students in seven European countries. Consistent with prior research on alcohol and other drug use, students who believed their peers engaged in NMPS were more likely to engage in NMPS themselves.

Although researchers have documented the prevalence and correlates of NMPS, much of the research has been limited by relatively small sample sizes and single-campus studies. Additionally, researchers have looked at concurrent or polysubstance use among students in various combinations, but without a focus that necessarily examines NMPS as a “red flag” or marker for other risky substance use.

The current study was launched to document trends surrounding NMPS, perceptions of NMPS, and engagement in other substance use to eventually inform norms-based prevention program development. The study included a large sample of students drawn from multiple colleges and universities across the US. Because of the possible overlap of marijuana use and NMPS (detailed above), recruitment for this study intentionally included four campuses in states with legal personal/nonmedical cannabis (compared with three campuses without). Thus, the data afforded the opportunity to look at any differences in the potential for past-year NMPS being a “red flag” for marijuana use as a function of marijuana access. Therefore, the aims of the current study were to: (1) further document the prevalence of NMPS along with perceived prevalence of peer NMPS among a sample of college students; (2) examine factors associated with NMPS (some previously examined in research mentioned above), including Greek life affiliation, skipping class, alcohol use, and marijuana use; and (3) describe the overlap between NMPS, binge drinking, and marijuana use, with a special emphasis on the overlap between NMPS and marijuana use as a function of cannabis legislation status.

Materials and Methods

Data collection

Data were collected during the 2015–2016 academic year as part of a study documenting the nonmedical use and diversion of prescription stimulants as well as alcohol and other drug use across seven US colleges and universities (four from states that legalized marijuana in 2012 for people 21 years old or older). Table 1 summarizes the characteristics, sample sizes, and response rates for each school included in this study. All students at each school received an email from their campus administration introducing a new research project being conducted on their campus entitled Project PHARM (Personalized Health Assessment Related to Medications) and informing them that they might receive an invitation to participate in the coming weeks. This served to affirm the legitimacy of a potential forthcoming invitation from Project PHARM to participate in the study. The cumulative response rate across the seven schools was 20.7%; this is comparable with other web-based national surveys of college students such as the National College Health Assessment (NCHA), which in fall of 2019 had a mean response rate of 14.1% (American College Health Association, 2020).

Table 1.

Sample size and response rate, by campus

School	Description	N	Response rate (%)
A	Large public university, Pacific Northwest	1,590	28.7
B	Small private college, Mountain-West	131	32.8
C	Small private college, Northeast	107	20.8
D	Large public university, Southeast	403	10.1
E	Medium-sized public university, Mountain-West	138	21.4
F	Large public university, Southeast	464	15.6
G	Small private college, Pacific Northwest	156	39.0
OVERALL		2,989	20.7

A random sample was obtained from the Registrar’s Office at each campus, and consisted of either 6% of the enrolled student population who were between the ages of 18 and 25 at each school or n=200 (whichever number was higher). Selected students were emailed initial invitations to complete a 20-minute online, confidential survey, and students were sent up to eight reminder emails to complete the survey during the recruitment period. Participants were compensated with a $10 Amazon.com gift card for their participation. All participants provided informed consent online prior to completing the survey. This study was approved by the lead university’s IRB. Participation agreements from each of the other six schools were obtained. Participants received further protections under a federal Certificate of Confidentiality.

Measures

Sample characteristics

Participants provided information on their age, gender [choices included male, female, transgender, and other], race/ethnicity, and past-year housing status. Participants also indicated whether or not they were involved in Greek life (i.e., a fraternity or sorority).

Class attendance

Participants were asked approximately how many class sessions they routinely skipped per week during their last full quarter or semester, and how many of these class sessions were skipped due to “use of alcohol or other substances (e.g., hungover, getting high, etc.)”.

Nonmedical use of prescription stimulants (NMPS)

NMPS was defined for participants as “use of a prescription stimulant that was not prescribed to you, that you took only for the experience or feeling it caused, or that you overused (e.g., too much or too frequently) if you have a prescription”. This definition was based on the work of McCabe, West, & Cranford (2011) and the subsequent question intended to not only assess frequency of use of prescription stimulants among those without a prescription, but to also encompass use among those with valid prescriptions who used their medications in higher doses or more frequently than prescribed. Participants were then asked, “In the past 12 months, on how many days have you used an ADHD prescription stimulant medication (e.g., Ritalin^©, Dexedrine^©, Adderall^©, Concerta^©, methylphenidate) nonmedically?” Similar questions were also asked for during the past six months and during the past 30 days.

Perceived NMPS by peers

Participants were asked to estimate the percentage of students at their school who used ADHD prescription stimulant medications that were not prescribed to them during the past 12 months (McCabe, 2008). Similar questions were also asked for during the past six months and during the past 30 days. Participants were instructed to provide a response between 0% and 100%.

Motives for NMPS

Participants were asked, “In the past six months, please rate how often you have used ADHD prescription stimulant medications (either without a prescription or in a way not prescribed)” for 17 different reasons in a measure developed by Teter and colleagues (2005), with responses of “Never/Almost Never,” “Rarely/Sometimes,” “Sometimes/Half the time,” “Often/Most of the time,” and “Always/Almost always.”

Alcohol use

The Quantity/Frequency/Peak Index (Dimeff, Baer, Kivlahan, & Marlatt, 1999) was used to measure the highest amount of alcohol consumed on a single occasion (ranging from 0 to 25 or more drinks) during the past month, and is a well-established, valid measure of peak drinking (McKenna, Treanor, O’Reilly, & Donnelly, 2018). Responses were recoded into two variables representing any past-month alcohol use and past-month binge drinking. Although we did not have an explicit item asking about exceeding binge drinking thresholds in a set time period, we recoded the past-month peak drinking occasion to compute a binge drinking variable. Binge drinking was defined as consuming five or more drinks for men or four or more drinks for women on a single occasion.

Marijuana use

Participants provided information on how many days they used marijuana during the past 30 days and during the past 12 months.

Data analytic approach

For the purposes of evaluating aims, analyses were primarily descriptive, though chi-square analyses were conducted to evaluate differences in prevalence across groups and Pearson’s r correlations tested relationships between variables of interest. SPSS version 19 was used for all analyses.

Results

Sample

A total of 2,989 participants completed the survey. Of participants sampled, 60.3% were female and the average age was 20.34 years. Table 2 summarizes the demographic variables (age, gender identity, race/ethnicity, housing status, Greek affiliation, and marijuana legal status in state of attendance) of participants.

Table 2.

Sample characteristics (N=2,989; missing data omitted from percentage calculations)

	M	SD
Age	20.34	1.63
	N	%
Gender
Female	1,791	60.3
Male	1,135	38.2
Transgender or other identity not listed	43	1.5
Missing	20	--
Race/ethnicity
White	1,941	65.5
Asian/Asian American	603	20.3
Black or African American	128	4.3
Native Hawaiian or Other Pacific Islander	11	0.4
Alaskan Native or American Indian	10	0.3
Multiracial	186	6.3
Other race or ethnicity	86	2.9
Missing	24	--
Past-year housing status
Off campus	1,250	41.9
On campus or in an off-campus residence hall owned by the college	1,078	36.2
At home with parents	347	11.6
Fraternity/sorority house	278	9.3
Living elsewhere	29	1.0
Missing	7	--
Involved in Greek Life
Yes	643	21.6
No	2335	78.4
Missing	11	--
Marijuana legal status in state of school attendance
Attends school in state with legal nonmedical/personal marijuana use over age 21	2,015	67.4
Attends school in state without legal marijuana	974	32.6

Prevalence of NMPS

Among the 2,735 study participants with valid data for NMPS (254 had missing data), 17.2% (n=471) engaged in NMPS at least once during the past year. Of those engaging in past-year NMPS, use was generally a low frequency behavior, with more than half (55.4%) engaging in NMPS on four or fewer occasions during the past year. As was the case with prior studies, there was a higher prevalence of NMPS among those who identified as male (21.1%) and as White/Caucasian (20.7%).

Perceived NMPS by peers

Although the majority of students had not engaged in NMPS during the past year, students perceived NMPS on their campus to be much higher than it actually was. In comparison with the 17.2% past-year prevalence of NMPS, students estimated that about 30% of students engaged in NMPS during the past year, with estimates ranging from 0% to 98%. One in five students (21%) believed that 50% or more of the students on their campus engaged in NMPS at least once during the past year. NMPS across different time frames was significantly positively correlated with perceived NMPS in those same time frames. Namely, past-year use was significantly correlated with perceived past-year norms (r=.107, p<.001), and past-6-month use was significantly correlated with perceived past-6-month norms (r=.141, p<.001).

Variables associated with NMPS

Motives for NMPS were overwhelmingly academic in nature, with eight of the nine most endorsed motives/reasons related to academic performance. Among those with past six month NMPS, the most frequently endorsed motives (any endorsement of “sometimes/half the time,” “often/most of the time,” or “always/almost always”) were “to concentrate better while studying” (54.0%), “to be able to study longer” (52.8%), “to feel less restless while studying” (35.0%), “because it helps increase my alertness” (28.9%), “to concentrate better in class” (18.7%), “to keep better track of assignments” (13.9%), “to feel less restless in class” (11.2%), “to feel better” (10.7%), and “to prevent others from having an academic edge” (9.4%). Motives related to partying were much less frequently endorsed (e.g., “to get high,” 9.1%, “to prolong the intoxicated effects of alcohol/substances,” (8.6%), “to counteract the effects of other drugs” (4.3%), as were motives related to weight loss (5.9%).

NMPS was significantly associated with Greek life affiliation; while only 12.9% of students not involved in Greek life engaged in past-year NMPS, 32.4% of students involved in Greek life engaged in past-year NMPS (χ²(1)=124.93, p<0.001). As shown in Table 3, past-year NMPS was significantly associated with skipping class, alcohol use, and marijuana use. Skipping class was more common among students who engaged in NMPS (χ²(1)=61.05, p<0.001), such that 54.1% skipped at least one class during their last full quarter/semester, compared with 34.9% of students who did not engage in NMPS. Further considering those with at least one skipped class, skipping for reasons related to the use of alcohol or other substances (e.g., being hungover, getting high, etc.) was more than four times greater for those who engaged in NMPS (39.6%) compared with those who did not engage in NMPS (8.9%; χ²(1)=131.72, p<0.001).

Table 3.

Variables associated with past-year nonmedical use of prescription stimulants (NMPS; n=471)

	No NMPS	NMPS
	(%)	(%)	χ2 value
Skipping Class			61.05*
Yes	34.9	54.1
No	65.1	45.9
Skipping Class Due to Alcohol and/or Drug Use			131.72*
Yes	8.9	39.6
No	91.1	60.4
Past-month Alcohol Use			122.26*
Yes	71.8	96.2
No	28.2	3.8
Past-month Binge Drinking			241.87*
Yes	47.0	87.1
No	53.0	12.9
Past-year Marijuana Use			335.37*
Yes	38.8	86.0
No	61.2	14.0
Past-30-day Marijuana Use			331.74*
Yes	23.0	66.2
No	77.0	33.8

^*Statistically significant at the p<0.001 level.

Among the overall sample, 76.0% of students used alcohol during the past month, and alcohol use was significantly associated with past-year NMPS (χ²=122.26, p<0.001). Among students who engaged in past-year NMPS, nearly all (96.2%) used alcohol during the past month, while among students who did not engage in past-year NMPS, 71.8% used alcohol during the past month. Results were similarly discrepant for past-month binge drinking, with past-year NMPS significantly associated with higher prevalence of binge drinking among the overall sample (χ²=241.87, p<0.001), and for both men (χ²=99.65, p<0.001) and women (χ²=141.98, p<0.001) separately. Among students engaging in past-year NMPS, 85.6% of men and 88.4% of women engaged in binge drinking during the past month. Among students not engaging in past-year NMPS, 47.0% of men and 47.1% of women engaged in binge drinking during the past month.

NMPS was also significantly associated with both past-year (χ²=335.37, p<0.001) and past-30-day marijuana use (χ²=331.74, p<0.001). Among students who did not engage in past-year NMPS, 38.8% used marijuana during the past year and 23.0% used marijuana during the past 30 days. However, among students who engaged in past-year NMPS, 86.0% used marijuana during the past year and 66.2% used marijuana during the past 30 days.

The relationship between NMPS and past-year and past-month marijuana use was even more pronounced within campuses in states with legal personal/nonmedical marijuana sales. Among those with any past-year NMPS who lived in a state with legal marijuana, 90.4% reported past-year marijuana use compared with 78.7% of those who lived in a non-legal state. This was not just a function of access based on legal age; in fact, among 18- to 20-year-old participants with past-year NMPS, 92.9% of those in legal states reported any past-year marijuana use compared with 75.5% of those in non-legal states (χ²=184.49, p<0.001), and among 21- to 25-year-old participants with past-year NMPS, 88.6% of those in legal states reported past-year marijuana use compared with 83.1% of those in non-legal states (χ²=151.22, p<0.001).

Overlap between NMPS, binge drinking, and marijuana use

Among those with valid data on all substance use variables (n=2,680, 89.7% of the sample; see Figure 1), 16.8% engaged in past-year NMPS, 46.7% used marijuana during the past year, and 53.8% engaged in past-month binge drinking. There is considerable overlap of students who engage in polysubstance use. Specifically, 17 students (0.6%) engaged in NMPS only, 228 students (8.5%) used marijuana only, and 413 students (15.4%) engaged in binge drinking only. A quarter of students (23.8%) engaged in binge drinking and marijuana use but not NMPS, 1.5% of students engaged in marijuana use and NMPS but not binge drinking, and 1.7% of students engaged in binge drinking and NMPS but not marijuana use. More than one in eight students (12.9%) engaged in all three substance use behaviors during the past year. Strikingly, among the 449 students who engaged in past-year NMPS, 96.2% also engaged in past-month binge drinking and/or past-year marijuana use.

Figure 1.

Overlap between past-month binge drinking, past-year marijuana use, and past-year nonmedical use of prescription stimulants (NMPS) among 2,680 undergraduate students.

Note. This figure is a proportional Venn diagram. Percent of all students with any binge drinking, marijuana use, or NMPS (n=1,727) shown in italics. Percent of all students (n=2,680) with no missing data and valid data for each behavior shown in bolded italics.

Discussion

Consistent with prior research (Garnier-Dykstra, Caldeira, Vincent, O’Grady, & Arria, 2012), the present study results showed that 17.2% of students sampled from seven US colleges and universities (for the purpose of documenting trends and attitudes surrounding NMPS) engaged in past-year NMPS. Similar to research on other substances (Kilmer et al., 2006), the current study results also indicated widespread normative misperceptions regarding NMPS on campus. Specifically, students estimated on average that 30% of typical students on their campus engaged in past-year NMPS; and 21% of students believed that 50% or more of the students on their campus engaged in past-year NMPS. Those who engaged in past-year NMPS were more likely to be involved in Greek life, skip classes for any reason, skip classes due to alcohol and other drug use, and engage in alcohol and marijuana use. In the current study, the overlap of various substance use behaviors among groups of students provides a unique contribution to the literature on this topic, particularly given how few students engage in NMPS only (0.6%). Among students who engaged in past-year NMPS, almost all (96.2%) also engaged in past-month binge drinking and/or past-year marijuana use. These results indicate that students engaged in NMPS represent a high-risk group on campus. Among students engaging in NMPS, 87.1% also engaged in binge drinking as compared with 47.1% of those who did not engage in NMPS. For marijuana use, 86.2% of students who engaged in past-year NMPS also used marijuana during the past year compared with only 38.8% of students who did not engage in NMPS. Past-30-day use of marijuana was even more pronounced with 66.2% of students who engaged in past-year NMPS also using marijuana during the past 30 days compared with only 23.0% of students who did not engage in NMPS. This study also highlighted that in states with legal marijuana for nonmedical/personal reasons, a greater percentage of students who engaged in NMPS also reported past-year marijuana use. Interestingly, the association between NMPS and marijuana use was more pronounced among 18 to 20 year olds compared with 21 to 25 year olds (i.e., those with legal access).

Arria et al. (2013) presented a model demonstrating how NMPS might be related to marijuana use and/or skipping class. On its own, marijuana use among college students is associated with skipping classes, lower GPA, delayed graduation, and school dropout (Arria et al., 2013; Suerken et al., 2016). Results of the current study provide further support for this model, as more students who engage in past-year NMPS skipped class than those who did not engage in past-year NMPS, and four times more of these students also had substance use-related reasons for skipping. NMPS might be initiated and maintained by these students in an unsuccessful attempt to improve academic performance following the detrimental effects of substance-related skipped classes; thus, addressing the relationship between marijuana use and reducing missed classes might reduce motivation for NMPS among this population.

The findings from this study have a number of implications that can inform prevention, education, and intervention efforts on college campuses. First, with the well-established link between overestimating descriptive norms and people’s own use of alcohol and marijuana, correcting students’ normative misperceptions associated with NMPS as well as misperceptions about the relationship between NMPS and improved academic outcomes could be important steps in reducing the frequency of and consequences associated with NMPS. Additionally, Event Specific Prevention approaches, targeting events such as 21^st birthday drinking, spring break, and other “time sensitive” holidays/activities, have shown potential in reducing high-risk behavior and minimizing related consequences for other substance use behavior (Lee et al., 2014; Neighbors et al., 2012). Adapting such interventions to address high-risk periods for NMPS misuse (e.g., midterms, finals, or when end of the quarter/semester deadlines are approaching) could also be beneficial (though these would need to be evaluated). Additionally, with Screening, Brief Intervention, and Referral to Treatment (SBIRT) efforts having been deemed effective by US Preventive Services, integrating NMPS screening into assessments could serve to detect and address risky or problematic use early (Babor et al., 2007). Further, given the very high co-occurrence of substance use behaviors among those who engage in past-year NMPS (e.g., 87.1% with past-month binge drinking and 86.2% with past-year marijuana use), and the fact that almost all students who engaged in NMPS (96.2%) also engaged in binge drinking, marijuana use, or both; a student’s disclosure of NMPS could serve as a “red flag” for other substance use that warrants subsequent screening and intervention related to alcohol and/or marijuana use. Almost all (92.9%) of 18- to 20-year-old students in states with legal marijuana who reported past-year NMPS also reported past-year marijuana use. Of course, NMPS on its own can be associated with a range of consequences, but the co-occurrence of NMPS with other risky behaviors highlights the importance of exploring these issues further. Incorporating a single item into screening in health centers and counseling centers, (e.g., “How many times during the past year have you used a prescription stimulant that was not prescribed to you, that you took only for the experience or feeling it caused, or that you overused (e.g., too much or too frequently) if you have a prescription?”) could serve to elicit additional conversations about substance use. A student who endorses this might not necessarily have a “stimulant problem,” but instead could be struggling with academics, skipped classes, heavy episodic drinking, or marijuana use.

The alcohol use literature has identified a number of groups at elevated risk for alcohol-related problems, including student athletes and members of fraternities and sororities (Turrisi, Mallett, Mastroleo, & Larimer, 2006). Clinically, given our finding that members of Greek life were more likely to engage in NMPS, focused and more intensive prevention and intervention efforts addressing NMPS within these groups would be a prudent strategy. Although the exact mechanisms surrounding higher engagement in NMPS by fraternity and sorority members is not known, access to stimulants could be higher in an intact social group in a way that would be different from those living in residence halls or independently. Thus, both targeted individual prevention efforts as well as efforts to limit diversion and sharing of stimulant medications within these organizations is warranted.

A comprehensive prevention and intervention strategy to address NMPS among college students should also involve training of academic advisors, athletic coaches, and fraternity and sorority advisors regarding high-risk periods for NMPS as well as potential warning signs of use. Finally, because our findings indicate norm misperceptions are strongly associated with NMPS, universal prevention strategies that include social norms messages focused on actual prevalence of use across the broader student population might be effective in reducing use (Perkins & Craig, 2002).

Strengths and limitations

This study included a sample comprised of multiple college campuses uniquely and intentionally selected to include schools in states with and without legal marijuana (important, particularly given hypotheses related to marijuana use). Further, the survey included a more thorough measurement of NMPS than is typically utilized, going beyond frequency of use to also assess perceived norms, motives, other substance use, and other trends related to NMPS. Yet, although the current study has several strengths, the study findings need to be viewed in light of limitations. First, there were mixed and low response rates for the current study (ranging from 10% to 39% and an overall response rate of 20.7%), which can reduce generalizability of results. Researchers indicate response rates for web-based surveys are low compared with other assessment methods (Fan & Yan, 2010), and the most frequently endorsed reason for college student survey nonresponse is being too busy to participate (Cranford et al., 2008). Another factor influencing response rates is survey length (Galesic & Bosnjak, 2009), especially in light of the minimal compensation ($10) offered in this survey. Concerns regarding response rates are lessened somewhat by the comparability of the obtained prevalence estimate of NMPS relative to other recent research (Garnier-Dykstra et al., 2012). However, recent study results have shown that increased contact attempts for survey participation are associated with heavy drinking, and late survey respondents are more likely to be binge drinkers than early/immediate respondents (Boniface, Scholes, Shelton, & Connor, 2017; Meiklejohn, Connor, & Kypri, 2012). Based on this work, future research should explore whether or not these findings extend to other substance use such as NMPS. Although the samples were representative of the schools that participated, the aggregate sample cannot be considered representative of all students attending college. Additionally, results are based on cross-sectional data, and thus we cannot establish causal relations between NMPS and other variables. Researchers utilizing longitudinal methods will need to establish temporal stability and causality of NMPS as it relates to academics, normative perceptions, alcohol use, and other substance use.

We did not collect data on simultaneous use of NMPS and other substances, and future studies could explore the degree to which this high-risk form of substance use takes place. A complete assessment of mental health diagnoses and life stressors was outside the scope of this study, and future research could examine the relationship between NMPS and mental health issues. We did not have event-specific motives, nor items associated with the context of use (e.g., with whom they used, when they used, why they used at any given moment), and future diary-based or qualitative studies could provide these factors associated with context of use. Additionally, we could not look at overlapping/co-occurring motives, and future research could explore in greater detail overlap between motives and whether some students have only academic enhancement motives or only party motives.

Finally, data for this study were collected via self-report, which could potentially be inaccurate due to intentional or unintentional influences. However, using objective measures, allowing participants to use their own computers, and use of online safeguards in place to assure confidentiality and minimize risks of data corruption all reduce concerns regarding self-report bias. Under these circumstances, self-report data on substance use has been shown to be both reliable and valid among college students and young adults (Kypri et al., 2016; Simons, Wills, Emery, & Marks, 2015). While the majority of measures used were from published works, further research will need to explore psychometric properties of these surveys.

Conclusions

Despite limitations, the current study results demonstrate relationships among NMPS, perceived norms, skipping class, and other substance use that have useful clinical implications and should lead to future research in this area. Future research is needed to further examine the nature and pattern of relationships among NMPS, skipping classes, and alcohol and other drug use. Particularly in states in which the legal climate surrounding marijuana is changing, monitoring the extent to which any changes in NMPS are associated with increased legal access to marijuana is recommended. Prevention and intervention materials need to be developed and tested to help reduce NMPS and related consequences, including development of screening and intervention materials addressing polysubstance use in the context of NMPS.