A large 2025 survey of over 65,000 U.S. college students found that those who use cannabis are about 27% more likely to also report excessive social media use, with the association growing stronger among higher-risk users. However, the study’s cross-sectional design means we cannot tell whether cannabis drives social media overuse, whether social media overuse drives cannabis use, or whether an unmeasured factor like depression or impulsivity fuels both.

Cannabis use and excessive social media consumption are both rising among U.S. college students, and each behavior is independently linked to mental health deterioration, academic impairment, and sedentary lifestyle risks. Campus health practitioners are increasingly asked to address both, yet virtually no large-scale data have examined whether these behaviors co-occur in the same students or remain separate clinical targets. Understanding their overlap is a prerequisite for designing efficient screening and intervention programs in a population where healthcare resources are often stretched thin.

Cannabis use among U.S. college students has risen steadily over the past decade, with recent national data showing that nearly one in four undergraduates reports use in the past 30 days. Simultaneously, excessive social media consumption has emerged as an independent concern on college campuses, linked to poorer mental health, reduced academic performance, and increased sedentary behavior. Drawing on the theory of compensatory internet use, which posits that individuals may turn to online platforms to cope with negative emotional states, the authors of this 2025 study hypothesized that cannabis users might be more likely to engage in excessive social media use. Their investigation used two waves of the American College Health Association National College Health Assessment (ACHA-NCHA-III), deploying both a binary cannabis use variable and the validated ASSIST-based Substance Specific Involvement Score to stratify cannabis risk.

Among 65,052 undergraduates, 23.2% reported current cannabis use. After adjusting for age, sex, race and ethnicity, parental education, enrollment status, class format, and other college-related factors, current cannabis users had 27% higher adjusted odds of excessive social media use compared to non-users (AOR = 1.27, 95% CI 1.20 to 1.34). In the 25,431-student sub-sample with cannabis risk scores, a gradient emerged: moderate risk was associated with 21% higher odds (AOR = 1.21, 95% CI 1.12 to 1.30) and high risk with 82% higher odds (AOR = 1.82, 95% CI 1.53 to 2.16) of excessive social media use. The authors acknowledge that the cross-sectional design precludes causal inference and that key potential confounders, notably depression, anxiety, and impulsivity, were not included as covariates. They call for longitudinal research to establish temporal sequencing and to inform targeted campus interventions.

Cannabis Use and Excessive Social Media Use Occur Together in College Students, but Which Comes First?

Nearly one in four U.S. college students reports using cannabis in the past month, and a similar proportion spends more than two hours a day on social media. A new study of 65,000 undergraduates finds these two behaviors overlap far more than chance would predict, and the more important question, which one leads to the other, remains unanswered. To the paper’s credit, it goes beyond the usual binary “did you use cannabis?” question and deploys the WHO’s validated ASSIST-based risk score to stratify students into low, moderate, and high cannabis risk categories. That choice is genuinely valuable, because it reveals something a simple yes-or-no measure would have missed: a clear gradient. Students at high cannabis use risk had nearly double the odds of excessive social media use compared to their low-risk peers, an effect size large enough to warrant clinical attention. The study’s sample is enormous, its confidence intervals are narrow, and the direction of findings is consistent across multiple models. These are real strengths. But the paper then overreaches. Its introduction frames cannabis as a potential “precursor” to excessive social media use, citing the theory of compensatory internet use. That theoretical model is plausible, but the study’s cross-sectional design is structurally incapable of testing it. Photographing two people standing next to each other does not tell you which one arrived first, or whether someone else invited them both.

The central methodological concern here is unmeasured confounding, and it is not subtle. Depression, anxiety, loneliness, impulsivity, and boredom proneness are all independently associated with both cannabis use and excessive social media consumption in this population. None of them appeared in the covariate models. Finding that people who carry umbrellas are more likely to get wet does not mean umbrellas cause rain; they both follow from the same upstream event. In clinical terms, it is entirely plausible that a student with untreated anxiety turns to both cannabis and social media scrolling as parallel coping strategies, and that neither behavior causes the other. The study cannot distinguish that scenario from a direct causal pathway in either direction. The paper also uses a 16-hours-per-week threshold for excessive social media use rather than the more standard 14 hours, due to fixed response categories in the survey instrument. That difference may seem small, but it means the study’s definition of “excessive” is stricter than the reference literature, potentially underestimating the prevalence and weakening comparisons across studies.

What would I say to a patient? If you are using cannabis regularly and also spending a lot of time on social media, it is worth exploring whether both habits might be serving the same underlying need, like managing stress or escaping boredom, and whether there are healthier ways to meet that need. To a colleague, I would note that this study adds to a growing pattern of co-occurring risk behaviors in college students, but the missing mental health covariates and cross-sectional design mean we cannot yet tell campus health programs whether addressing one behavior would change the other. To a policymaker, I would urge caution: the co-occurrence is real and worth monitoring, but policies should await longitudinal evidence before targeting either behavior as a lever for changing the other. A large sample makes an association statistically unmistakable but does not make it causally interpretable; the most important scientific question, what comes first and why, remains entirely open after this paper.

This study sits early in the research arc for understanding the intersection of cannabis use and excessive social media consumption. While prior work by Neilson and Lin (2019) linked cannabis frequency to sedentary screen time and Nagata et al. (2025) suggested that heavy internet use is associated with cannabis experimentation, no previous study has examined the association at this scale using a validated cannabis risk instrument. The gradient finding, with high-risk users showing nearly double the odds of excessive social media use, is novel and provides a foundation for more targeted investigation.

From a pharmacological standpoint, cannabis can acutely impair executive function, reduce motivation for physical activity, and alter reward sensitivity, all of which could plausibly increase passive screen-based behavior. However, these mechanisms remain speculative in the context of this study. There are no drug interaction considerations specific to social media use, but clinicians should be aware that students presenting with either high cannabis use risk or excessive social media consumption may be experiencing overlapping mental health challenges. One concrete recommendation: campus health providers conducting ASSIST-based cannabis screening should consider adding a brief social media use assessment for students who score in the moderate-to-high risk range, as these individuals may warrant broader behavioral health evaluation.

This is a cross-sectional secondary data analysis of a large, nationally distributed college health survey. In the evidence hierarchy, cross-sectional studies rank below cohort studies, case-control studies, and randomized trials for establishing causation because exposure and outcome are measured at a single point in time. The single most important constraint on inference is that no temporal sequence can be established: we cannot determine whether cannabis use preceded, followed, or merely co-occurred with excessive social media consumption.

This study extends earlier, smaller investigations that observed correlations between substance use and screen-based behavior in young adults. Neilson and Lin (2019) found cannabis use frequency associated with sedentary screen time, and Nagata et al. (2025) reported that heavy internet and social media use predicted cannabis experimentation, suggesting the relationship may be bidirectional. Notably, Gutierrez and Cooper (2016) found no significant association between time on social media and cannabis use in an earlier college sample, creating a discrepancy this study cannot fully resolve given its own cross-sectional design. The present paper’s gradient by cannabis risk severity is a genuinely new contribution, and it aligns with broader evidence that higher-risk substance users tend to cluster with multiple behavioral risk factors.

The most consequential analytic choice was the set of covariates included in the adjusted models. The authors controlled for demographics and college-related factors but omitted depression, anxiety, impulsivity, and loneliness, which are among the strongest known predictors of both cannabis use and excessive social media consumption. Had these variables been included, the adjusted odds ratios could have attenuated substantially, potentially reducing the AOR for current cannabis use from 1.27 toward the null. Additionally, using the 14-hour-per-week threshold (the standard in prior literature) rather than the 16-hour threshold forced by survey response categories might have reclassified a meaningful number of students, altering both the prevalence of the outcome and the magnitude of the association.

The most likely overinterpretation is reading these results as evidence that cannabis use causes excessive social media use. The study’s introduction frames cannabis as a potential “precursor” to excessive screen time, drawing on the theory of compensatory internet use, but the cross-sectional design is entirely agnostic to which behavior came first. The reverse pathway (social media overuse increasing cannabis exposure through online normalization) and the shared-cause pathway (underlying depression or impulsivity driving both) are equally consistent with the data. An adjusted odds ratio of 1.27 may also be misread as a large effect; in absolute terms, this is a modest increase in odds that may not translate to meaningful individual-level risk differences, particularly when key confounders remain unadjusted.

This study establishes that cannabis use and excessive social media use co-occur more frequently than expected among U.S. college students, with a meaningful gradient by cannabis risk severity. It does not establish that one behavior causes the other, and unmeasured mental health confounders may account for some or all of the observed association. For clinical practice today, these findings are best used as a prompt for broader behavioral screening rather than as a basis for targeted intervention.

Does cannabis use cause excessive social media use?

This study cannot answer that question. It measured both behaviors at the same time, so there is no way to determine which came first. It is equally possible that social media overuse leads to more cannabis exposure, or that a shared factor like anxiety or boredom drives both behaviors simultaneously.

How big is the increased risk?

Current cannabis users had about 27% higher odds of excessive social media use compared to non-users. That is a statistically significant but modest increase. For students at high risk of cannabis use disorder, the odds were roughly 82% higher, which is a more notable finding but still an association, not proof of causation.

Should I be worried if I use cannabis and also spend a lot of time on social media?

Neither behavior in isolation is necessarily harmful in every context, but the combination may signal that both are serving a coping function for stress, loneliness, or other emotional needs. If you find that either habit is interfering with your academic, social, or physical well-being, it is worth speaking with a healthcare provider about what might be driving both patterns.

Does this study apply to all cannabis users?

The study sample was limited to U.S. undergraduate students aged 18 to 24 at four-year institutions. The findings may not generalize to older adults, medical cannabis users, or individuals outside of college settings. The study also did not distinguish between medical and non-medical cannabis use.

Have thoughts on this? Share it: