A cross-sectional survey was administered to first-year medical students (M1), second-year medical students (M2), and third-year medical students (M3) at a single California medical school in 2024. An email was sent to all these students, and 121 students among all three classes opted to complete the survey. The inclusion criteria were that the participant was to be 18 years of age or older and to be able to provide consent for participation. All 121 participants met the criteria, and no responses were excluded. The final sample included 54 M1s, 45 M2s, and 22 M3s.

This study received ethical approval from the California University of Science and Medicine Institutional Review Board (approval: HS-2024‐51) on July 11, 2024. Informed consent for primary data collection and secondary analyses of the research data was obtained from all individual participants included in the study. Participants were entered into a raffle where they could win one of twenty US $20 gift cards. Participant privacy was protected by de-identiying all information obtained by preventing the collection of direct identifiers such as email addresses and instead using a study ID for the raffle.

The survey collected demographic characteristics and self-reported use of coffee, tea, energy drinks, sodas, and over-the-counter caffeine preparations (Multimedia Appendix 1). Participants were asked about only standardized serving sizes and not how much caffeine each serving contained.

To explore maladaptive caffeine-related behaviors, we adapted the four-item CAGE questionnaire to caffeine use. This tool has not been validated for caffeine; therefore, it was used only for exploratory screening.

The primary outcomes included (1) the daily caffeine intake (mg/day), (2) the proportion of participants consuming ≥400 mg/day, a widely accepted upper safe limit for adults [4]; and (3) the endorsement of ≥1 adapted CAGE item that was obtained using the descriptive data only.

Descriptive statistics are reported as mean (SD) or median (IQR). Group differences across training years were evaluated using ANOVA or Kruskal-Wallis tests as appropriate. Analyses were descriptive and not adjusted for confounding variables such as sleep, stress, or baseline stimulant use.

Table 1 presents the sample characteristics. The cohort included 61.2% (74/121) female medical students, with a mean age of 24.1 (SD 1.7) years.

The mean daily caffeine intake was 114.3 (SD 625.7) mg/day. Intake generally increased by training year (M1: 111.4 mg/day; M2: 106.1 mg/day; M3: 138.4 mg/day), although the differences were not statistically significant (P=.09). 1/121 of participants consumed ≥400 mg/day of caffeine.

Coffee was the primary source of caffeine (96/121), followed by tea (85/121), sodas (51/121), and energy drinks (33/121). Over-the-counter caffeine products were used by 5/121.

63/121 endorsed ≥1 adapted CAGE item. The most common was “felt the need to cut down” (42/121 or 35%). Because the CAGE-caffeine tool is unvalidated, these findings are interpreted only as exploratory descriptors and not diagnostic indicators.

In this single-institution cross-sectional study, medical students reported moderate caffeine intake, with one student exceeding the daily upper limit of 400 mg recommended for adults. Caffeine intake increased across training years, although the differences did not reach statistical significance. These findings are consistent with prior research showing rising caffeine use with progressing academic demands [3].

A key limitation is the lack of validated screening tools for problematic caffeine use. Our adapted CAGE-caffeine items were included only to explore behavioral tendencies; they cannot indicate caffeine use disorder or dependence. Future work should employ validated caffeine behavior instruments or include psychometric testing. Other limitations include potential recall bias, the absence of important confounders such as sleep and stress, restricted generalizability due to the single-site sample, and potential self-selection bias. The potential for winning gift cards could have potentially attenuated some self-selection bias by providing a generalized incentive for participation. Despite these limitations, the study provides descriptive data on caffeine use patterns in medical training and highlights the need for validated tools to assess maladaptive caffeine behaviors.

Medical students commonly consume caffeine, with a minority exceeding recommended daily limits. Exploratory findings suggest some students may wish to reduce intake. Future studies should incorporate validated instruments and longitudinal designs to better evaluate changes over time.