This cross-sectional, online survey evaluated prescription stimulant NMU among college students in the United States. Pathways of initiation of prescription stimulant NMU were presented, as well as characteristics of non-oral use.

All respondents to this survey engaged in NMU of stimulants, while nearly one in four (23%) reported non-oral use. Non-oral use was consistently associated with increased likelihood of serious, non-alcohol substance use. Although other studies have found ADHD diagnosis, gender, race, depression, anxiety, marijuana and alcohol use to be associated with prescription stimulant NMU (34), in this study age, race, recent alcohol binge drinking, and marijuana use did not differentiate oral and non-oral NMU. Univariate analyses revealed associations of major factors with reports of non-oral ROA, especially being male, lifetime diagnoses of ADHD, alcohol or other substance disorder as well as conduct/oppositional defiant disorder, and history of prescription opioid NMU.

The finding that women students were less likely overall to report non-oral NMU practices, may be an underlying feature common to many of the other factors investigated. For instance, women are less likely to be diagnosed with ADHD and, women diagnosed with ADHD appear to exhibit lower levels of conduct problems than their male counterparts (35), both factors that were associated with non-oral ROA NMU in this survey. Women were significantly less likely to inject, even if they reported snorting prescription stimulants, however, oral-only ROA vs. snorting but not injection were not significantly different by gender. Injecting may be associated with higher risk-taking behavior in men with ADHD (35). On the other hand, other studies of gender differences in adults with ADHD (36) have found women with an ADHD diagnosis to be much more likely than women without one to express suicidal ideation, which may be a very different phenomenon than male risk-taking behavior.

Of those reporting non-oral prescription stimulant NMU, the large majority (88%) snorted the medication, with a quarter reporting smoking and 15% injecting. Our findings are consistent with those from the few studies investigating injection of prescription stimulants, suggesting that injection of these medications is relatively rare (6, 20, 23). Because the small number of college students who inject prescription stimulants in this sample resulted in point estimates that may not be stable, the findings reported for non-oral injection users should be considered preliminary and used to stimulate further research. However, even though injection may be rare, it is associated with much higher risk and understanding what motivates this behavior may inform prevention and treatment strategies.

Compared with oral-only users, respondents reporting any non-oral route were more likely to have more serious substance use histories and drug-use patterns. Non-oral use is consistently associated with serious adverse consequences (10), increased prevalence of overdose (37), addictive potential (38) and significant risk of morbidity and mortality (23). Note that in the current study, the large majority of those who reported NMU of prescription stimulants via non-oral routes snorted the drug. Further research into this type of non-oral prescription stimulant NMU is warranted, as such work may clarify a role that ADF formulations can have in reducing the serious adverse consequences of non-oral prescription stimulant NMU (16, 19, 23, 24).

Consistent with findings of earlier research (7, 10, 39), two-thirds of NMU respondents obtained the prescription stimulant from a family member or friend (either bought, given, or stolen). About 40% obtained the stimulant through a legitimate prescription. Non-oral users, on the other hand, endorsed other sources, especially getting the drug from a “dealer,” and online purchases, while also obtaining prescription stimulants through the usual sources of friends, family, and providers.

The motivation most often reported for NMU by oral routes and snorting was performance enhancement, as others have noted (3, 7, 10, 40). Such oral NMU of prescription stimulants for performance enhancement may signal the lack of adequate resources or supports for students in a highly artificial (e.g., memory-based performance testing), high stress environment, where academic outcomes can have life-altering consequences. NMU for performance enhancement by students with ADHD using oral routes also raises questions regarding the perceptions by these patients as to the adequacy of their treatment (16). It may also be relevant that learning problems often associated with an ADHD diagnosis may be more responsive to treatments other than medication (35). Although not among the most often reported motivations by those who inject, performance enhancement was reported as a motivation for NMU by nearly a quarter of those respondents. Those reporting snorting and injection were significantly more likely than the oral-only group to report motivations to get high or to enhance the effects of other drugs. These motivations and the higher likelihood of polysubstance use in non-oral users are indicative of higher risk behaviors often associated with having or developing significant substance use problems.

More than 90% of prescription stimulant NMU was observed in the context of polysubstance use (not including alcohol); a finding consistent with results of a population-based survey (41). Among polysubstance users, 18.7% reported marijuana as the only other drug used. While 80.6% of polysubstance users reported marijuana use, and 69.6% of those reported marijuana use prior to prescription stimulant NMU, marijuana was considered separately in this study, given its legal status in many states for use by at least some individuals in the examined age range as well as changing views of marijuana use by the public (42, 43). Nevertheless, it is noteworthy that prescription stimulant NMU preceded prescription opioid NMU 45% of the time and preceded illegal drug use 32% of the time. Our pathways analyses are conceptually similar to the so-called “gateway hypothesis” (29), which proposes that a child or adolescent's early experimentation with or exposure to alcohol, tobacco, or cannabis leads to more addictive illicit drugs later in adulthood (44, 45). Empirical examination of the gateway hypothesis has yielded conflicting and ambiguous results (45, 46), suggesting that a simple, causal role of early exposure to abusable or addictive substances is unlikely. Interpretation of early prescription stimulant NMU in the pathway toward use of illegal drugs, NMU of other prescription drugs or use by non-oral routes is complex. There is evidence from this survey and other research (47) that an ADHD diagnosis maybe a risk factor for later serious drug involvement (i.e., polysubstance use and non-oral ROA use). Specifically investigating exposure to ADHD medications in childhood and adolescence, McCabe et al. (48) found that children who initiated stimulant medication for ADHD early (aged 9 or less) and for longer duration (6 or more years) did not differ between population controls (youth without ADHD and unmedicated youth with ADHD).

The pathway findings among the 90% of survey respondents reporting polydrug use in this study support other studies that suggest primary care physicians largely underestimate the potential for NMU and diversion of ADHD medications by their patients (16). Thus, the findings presented here suggest that prescription stimulant NMU coincided with a more general pattern of substance misuse/abuse. Prescribers should at least be cognizant of the potential for current or later escalation of substance use by their patients. This study shares the limitations of other online surveys, including self-selection bias, self-report of drug use and medical diagnoses, and the use of an online panel. Although efforts were made to establish a representative sampling frame from which to draw respondents, the final sample may not be representative of college students who engage in NMU. While many of the findings are consistent with those from other studies, generalization should be approached cautiously. Use of a cross-sectional survey to explore retrospective recall of pathways of drug initiation across time is, by definition, limited, and can only be considered suggestive. In agreement with other authors (9), our use of the Internet panel survey is intended to complement, not replace, national probability studies.