Energy Drinks, Tobacco and Junk Food: What Really Drinks Drive Dubai Students

Energy drinks have become a fixture of student life across the Gulf, but a new study from Dubai suggests that the factors determining whether a young adult picks up a can at all are not the same as those determining how often they drink one. The research, published in BMC Public Health by a team at Canadian University Dubai, surveyed 599 undergraduate students and found that nearly 40 percent reported using energy drinks. More strikingly, when the researchers separated the question of use from the question of frequency, two distinct pictures emerged, one dominated by diet and tobacco, the other sharpened by sex and heavier tobacco involvement.

The survey was anonymous and administered online, recruiting participants through university and student channels. After a post-hoc integrity audit, five probable repeat submissions were excluded, leaving a primary analytic sample of 599 complete records. Of these, 238 participants, or 39.7 percent, reported energy-drink use, with a 95 percent confidence interval of 35.9 to 43.7 percent. Notably, the use item did not specify a recall period, meaning respondents were reporting on their general pattern rather than a defined window such as the past month.

To understand what distinguished users from non-users, the researchers employed Pearson chi-square tests followed by multivariable binary logistic regression. Among those who did use energy drinks, proportional-odds ordinal regression examined reported consumption frequency, with threshold-specific binary models used to check robustness. This two-tiered analytical approach allowed the team to ask not just who uses energy drinks, but who among users drinks them most often, a distinction rarely made in the literature.

The results for use status pointed clearly to diet and tobacco. A fast-food or junk-food dietary pattern was associated with more than double the odds of energy-drink use, with an adjusted odds ratio of 2.39. Occasional tobacco-product use carried an adjusted odds ratio of 2.12, while regular tobacco-product use nearly tripled the odds at 3.19. The adjusted estimate for male sex was less precise, at 1.35 with a confidence interval spanning 0.90 to 2.04, meaning the data could not rule out no difference between sexes for use status.

Among the 238 users, however, the picture shifted. Male sex emerged as a strong correlate of higher consumption frequency, with an adjusted cumulative odds ratio of 2.67. Regular tobacco-product use also predicted heavier consumption, at an adjusted cumulative odds ratio of 2.34. Both findings held up in threshold-specific sensitivity models, suggesting they were not artefacts of how the frequency categories were cut.

The most consistent factor across both outcomes was tobacco-product use, appearing as a significant correlate whether the question was use or frequency. This overlap is biologically plausible. Nicotine and caffeine share stimulant properties, and students who use tobacco may be more broadly oriented toward stimulant-based coping or social behaviours that include energy drinks. The authors suggest that university health promotion should address energy-drink use alongside tobacco-product use, dietary behaviours and study-related consumption contexts rather than treating these as isolated habits.

The finding that sex mattered more for frequency than for status is particularly interesting from a methodological standpoint. Had the study relied solely on a binary use measure, the sex difference in heavier consumption among users would have been invisible. This supports the authors’ broader argument that assessing frequency among users can identify patterns not captured by a simple yes-or-no measure, and that the correlate profiles for use and for heavy use are overlapping but non-identical.

The study is cross-sectional and uses a non-probability sample, so causal direction cannot be established and generalisation beyond the sampled population is limited. It remains unclear whether energy-drink use drives tobacco use, whether tobacco use prompts energy-drink consumption, or whether both reflect a broader lifestyle cluster. The authors also note that the use item lacked a defined recall period, which limits precision in interpreting the prevalence estimate.

Despite these constraints, the study offers a clear and actionable insight for universities in the region and beyond. Health-promotion programmes that lump all stimulant and dietary behaviours together may miss the students at highest risk of heavy consumption, who are more likely to be male and regular tobacco users. Separating use from frequency, the authors argue, is not a statistical nicety but a practical necessity for identifying the students whose energy-drink habits have moved beyond occasional novelty into a pattern worth addressing.

The methodological distinction at the heart of the study deserves closer attention, because it reflects a broader tension in survey-based nutrition research. Binary indicators of substance or product use are easy to administer and easy to report, but they compress a wide range of behaviour into a single yes-or-no response. A student who drinks one energy drink a year and a student who drinks one before every lecture both count as users, yet the health implications and the underlying drivers of those two patterns may differ substantially. By modelling consumption frequency as an ordered outcome among users, the Dubai team effectively treated intensity of use as a phenomenon in its own right, an approach more common in alcohol and tobacco research than in studies of caffeinated beverages.

The statistical machinery behind the frequency analysis also illustrates why the findings carry weight. Proportional-odds ordinal regression assumes that the relationship between a predictor and the outcome holds consistently across every cut point of the ordered categories, an assumption that can be violated in real data. The authors therefore ran threshold-specific binary models as a sensitivity check, effectively re-estimating the associations at each boundary between frequency levels. The fact that male sex and regular tobacco use remained significant across these models suggests the associations are not driven by how the frequency scale was divided, which strengthens confidence in the core result despite the modest sample of users.

The recruitment strategy also shapes how the findings should be read. Because participants were drawn through university and student channels rather than a population registry, the sample reflects students willing to respond to an anonymous online survey distributed through those networks. The reported prevalence of 39.7 percent should therefore be understood as a description of this particular sample rather than a precise estimate for all undergraduates in Dubai. The wide confidence interval, spanning roughly 36 to 44 percent, is a reminder that even within the sample the true proportion could fall across a meaningful range. Cross-sectional surveys of this kind are valuable for generating hypotheses about behavioural clustering, but they cannot disentangle whether energy drinks attract students already inclined toward stimulant use or whether habitual consumption fosters other stimulant habits over time.

The clustering of energy-drink use with fast-food and junk-food dietary patterns fits a wider body of literature on convenience-oriented consumption among young adults. Energy drinks are typically purchased at the same retail points as snack foods, marketed with similar appeals to speed, performance and sociability, and consumed in comparable contexts such as late-night study sessions, gaming and social outings. A shared convenience orientation may underlie both behaviours, meaning dietary pattern may function less as an independent cause of energy-drink use and more as a marker of a broader consumption style. For university health programmes, this suggests that interventions targeting one convenience behaviour may have spillover effects on others, or conversely that single-behaviour campaigns may fail if they ignore the cluster in which the behaviour is embedded.

The tobacco findings raise a parallel set of questions. The graded pattern in the data, with occasional tobacco use associated with roughly doubled odds of energy-drink use and regular use with nearly tripled odds, hints at a dose-response relationship that would be difficult to explain by chance alone. Among users, regular tobacco use again predicted more frequent consumption. One plausible mechanism is shared social settings, since both products are often consumed in group contexts and marketed to similar demographics. Another is a general stimulant-seeking disposition, in which nicotine and caffeine serve complementary roles in managing alertness and study fatigue. The authors’ recommendation that health promotion address energy drinks alongside tobacco is consistent with either mechanism, and the co-occurrence documented here provides a concrete empirical basis for designing combined interventions rather than parallel, disconnected ones.

The exploratory risk-awareness measures included in the survey add a further dimension worth noting. The study assessed participants’ knowledge of energy-drink contents and potential harms, and the supplementary materials report awareness diagnostics alongside the behavioural findings. Understanding whether users differ from non-users in what they know about these products could inform whether educational messaging alone is likely to shift behaviour, or whether consumption is driven more by habit, social context and product availability than by information gaps. The inclusion of self-reported symptoms in the supplementary tables similarly acknowledges that consumption frequency may carry health consequences that a use-versus-non-use dichotomy would obscure.

Finally, the study’s timing and setting give it regional significance. The Gulf states have among the highest per-capita soft-drink and energy-drink consumption levels globally, and university populations in rapidly urbanising cities like Dubai face particular pressures from demanding academic schedules, part-time work and a 24-hour service economy. Documenting how energy-drink behaviour clusters with tobacco use and dietary patterns in this setting provides local evidence for institutions designing wellness programmes, and the methodological lesson, that frequency among users reveals correlates invisible to binary measures, is one that researchers studying other consumable behaviours in the region would do well to adopt.

Subject of Research: Beyond prevalence: distinct correlates of energy-drink use and consumption frequency in a cross-sectional study of undergraduate students in Dubai

Article Title: Beyond prevalence: distinct correlates of energy-drink use and consumption frequency in a cross-sectional study of undergraduate students in Dubai

Article References: Takshe, A. A., Ahmad, H., Akbar, M., Desmal, A. B., Trivedi, H., & Sagintayeva, M. (2026). Beyond prevalence: distinct correlates of energy-drink use and consumption frequency in a cross-sectional study of undergraduate students in Dubai. BMC Public Health. https://doi.org/10.1186/s12889-026-29978-3

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29978-3

Keywords: Beyond, prevalence, distinct, correlates, energy-drink, consumption, frequency, cross-sectional, undergraduate, students, Dubai, scientific research