Mental health and sleep habits during preclinical years of medical school.

2.1. Participants and study approach

An invitation to participate in this non-interventional study was delivered to all enrolled students in the class of 2023 and the class of 2024 in a medical school (in the United Stated, accredited by the HLC and COCA) with two separate campuses, the inclusion criteria was being a medical student currently enrolled at the specific level (first or second year). Follow-up survey invitations were emailed to those students that signed the informed consent to participate in this study and completed the first survey. Exclusion criteria was students that did not complete a consent form or did not meet the requirements for enrollment for the semester. An outline of the study is presented in Fig. 1 .

No compensation was provided for participation, which was voluntary. All participants consented prior to participating. The project was reviewed by the Rocky Vista University Institutional Review Board with the corresponding IRB# 2019-0070. The project was deemed exempt from a full IRB review. No specific funding was received to conduct this study.

2.2. Surveys

Surveys assessed underlying chronic disease, sleep quantity and quality, exercise habits, and depression and anxiety. A chronic disease was specified as one diagnosed by a physician, this was done to reduce the risk of including self-diagnosed conditions. Sleep quantity was assessed as average hours of sleep per night over the last month. Sleep quality was assessed the first year of the study with three questions that incorporated graded responses to quantify sleep quality changes (SQ-3). Although the sleep quality questions were common questions used in other studies to quantify sleep quality, they have never been used in this format. These specific questions are (rated on a likert scale): “On average, how many times throughout the week do you feel the inclination to fall asleep during the day?” “On average, how often do you feel tired or fatigued after you sleep?” “During your waking time, do you feel tired, fatigued, or not up to par?”. The Pittsburgh Sleep Quality Index (PSQI) [27] was added in the survey for the Summer ‘20 to supplement the evaluation of sleep quality. Exercise was assessed as days per week and minutes per session with total weekly minutes being calculated. To assess mental health, validated questionnaires were used to analyze levels of anxiety and depression, Generalized Anxiety Disorder-7 (GAD-7) and Patient Health Questionnaire-9 (PHQ-9), respectively [28,29]. Therefore, all subjective measurements were assessed with previously validated questionnaires. Three additional questions were added to evaluate for self-reported effects of COVID-19 in the summer survey of 2020 (Likert scale (strongly disagree, disagree, neither agree or disagree agree, and strongly agree). These questions were: “COVID-19 has caused me to experience higher levels of depression over the two weeks before orientation.” “COVID-19 has caused me to experience higher levels of anxiety over the two weeks before orientation.” “COVID-19 has caused a decrease in my quality of sleep during the past month.” All data provided by the participants was self-reported. The survey used is available as a supplementary file.

2.3. Statistical analysis

Data was analyzed using Generalized Linear Mixed Models (GLMMs) on the numeric responses of SQ-3, GAD-7, PHQ-9 and PSQI. These models included fixed effects of class year (class of 2023/class of 2024), COVID-19 period (Pre-/ Post-), specific time points (Summer ‘19 through Spring ‘21), chronic disease diagnosis (Yes/No), average number of sleep hours (hours), and total exercise per week (minutes). Repeated measurements by participants were included in the models as a random effect. All models assumed errors to be normally distributed with mean zero and variance of σ². All modeling and descriptive statistical evaluations were performed with SAS v.9.4 (SAS Institute, Cary, NC). Statistical significance was declared at a confidence level of 95%.

The analysis was performed on three separate cohorts defined by the time points and class years. The first modeling analysis was performed on the full dataset which included both classes of 2023 and 2024, this cohort did not analyze COVID-19 specific questions. The second modeling analysis did include the COVID-19 specific questions. The third, and last modeling analysis, was performed only on the class of 2023 cohort, this was done to separate the effect of participants from the class of 2024 that had only been in medical school during the pandemic. This separate cohort approach was devised to reduce any possible bias introduced by environmental confounders derived from the COVID-19 pandemic. This study was ongoing while the pandemic developed and thus decisions were made based on available data.

4.1. Mental health and sleep habits

Medical school has a significant effect on sleep quality, depression, and anxiety [[9], [10], [11], [12], [13], [14],19,21,[30], [31], [32], [33], [34], [35]]. The majority of analyses are cross sectional studies that are an assessment of one time point, relying on other studies to provide context for their findings [6,12,30 ,31,36,35,5]. However, the longitudinal analysis of students in this study uniquely allowed the quantification of the change from pre-medical school and summer break to each semester in medical school throughout years one and two, allowing students to act as their own controls and reducing confounding factors. The results across time points displayed a cyclic change in mental health and sleep quality of medical students that was associated directly with the academic/seasonal calendars.

There is great variability in the literature in sleep quality, depression, and/or anxiety of medical students. The reported percentages of students that suffer from poor sleep quality range from 55.8-76%, anxiety from 41-73%, and depression from 10.4-65% [[11], [12], [13],19,21,[30], [31], [32], [33],36]. This variation is due in part to the lack of standardized assessment tools to assess anxiety, depression, and sleep quality in medical students, leading to a variety of assessments being used. Additionally, it is difficult to account for the different structures and expectations within medical schools around the world. As such, this study does not compare the percentages of total anxiety, depression, and sleep quality to other analyses, but assesses the longitudinal change.

The medical school assessed in this study is structured similarly to the general model of medical schools throughout the United States: 4-year curriculum divided into two years of basic science coursework followed by two years of clinical clerkships [37]. This study evaluated students in their first two years of basic science coursework noting no significant difference in anxiety, depression, and/or sleep quality between year one and year two. However, differing levels of anxiety, depression, and sleep quality have been observed in other studies when all years of training are considered [32,38].

The analysis revealed that class year (class of 2023 versus class of 2024) and exercise habits had no effect on mental health or sleep quality. Although exercise has shown to cause reduced incidence of anxiety and depression in the general population [16], this study suggests that there is no significant effect in medical students.

The presence of chronic disease, diagnosed by a physician (self-reported), corresponded with higher anxiety and depression levels with decreased sleep quality. A systematic review of 20 studies outlined that chronic disease may be interrelated to anxiety/depression, one causing the other or both potentially exacerbating each other [17]. A potential explanation for this finding could be that people have a limited ability to compensate for stress. Thus, when students are living with a chronic illness, their level of stress his already higher than average; when the pressures of medical school are added, they are limited in their reserve to compensate healthily and are more susceptible to developing mental illness.

Fewer hours of sleep were shown to have a negative effect on sleep quality and mental health. Number of hours of sleep was shown to affect depression levels more than anxiety levels. Evidence suggests that sleeping less than six hours leads to the development of chronic disease [39], a risk factor for increased anxiety, depression, and poor sleep quality. One study found that medical students often sacrifice sleep to meet the demands of school [13]. Helping students prioritize sufficient hours of sleep each night could have long term benefits.

This analysis provides insight into areas that medical schools may focus, including sleep patterns and mental health. “Good” sleep improves academic performance [40] and medical students who are sleep deprived have significantly lower GPAs [41]. Anxiety, depression, and stress lead to poor academic performance in medical students [42,43]. Students often recognize these problems and perceive their effect on their academic performance [22], but do not always know how to receive help. At a United States medical school where confidential mental health services were available, only 26.5% of the identified depressed students reported treatment. Notably, 20.4% of the depressed students reported experiencing suicidal ideation [33]. Medical student mental health is a serious and prevalent issue that requires more attention.

Medical schools should focus on student health promotion emphasizing the importance of sleep and mental health [21]. One suggestion to help prevent depression is to educate students on appropriate sleep hygiene and encourage students to pursue professional guidance to treat sleep disturbances [22]. Another suggestion is for medical schools to institute stress coping mechanisms and improvement of interpersonal relationships [11]. A well-being curriculum [44] or targeted interventions that have shown some promising results at the college level [45,46] could also be considered. Curriculum additions and interventions can be challenging to implement because of the rigid and saturated curricula that is often followed for medical education in the United States. However, this study offers understanding of the patterns at which advertisement and outreach of current services can be more appreciated.

4.2. Assessment of the impact of COVID-19

The COVID-19 pandemic arose about halfway through the data collection period of this study causing the incorporation of a virtual learning environment to encourage social distancing [47,48]. The pandemic has had a negative impact on depression, anxiety, and sleep disorders [25,26]. However, other studies report increases in sleep among college students [49].

In this study, the effect of COVID-19 was evaluated in two ways: first, by evaluating directly its effect in the model (pre- and post-COVID-19 period variable) and second, by comparing the first three time points (Summer ‘19 - Spring ‘20), to the corresponding seasons of the following year: time points 4–6 (Summer ‘20 - Spring ‘21). On the full dataset, there were no differences across time points for the same season of the year regardless of pre- or post-COVID-19 period variable, suggesting that COVID-19 did not significantly contribute to anxiety, depression, or sleep quality. To further account for confounding effects of putting together two classes (2023 and 2024), the analysis was also run with only the class of 2023. Findings revealed that the COVID-19 period had no significant effect on any scores. Similarly, a recent report showed the effect of the COVID-19 pandemic on stress and sleep was not necessarily negative and suggested that college students are largely resilient [50] . After the COVID-19 pandemic started, 3 questions were added to the surveys to measure the self-reported effect of COVID-19 on sleep quality, anxiety, and depression. The results indicated worse outcomes in all three categories. However, change in anxiety, depression, and sleep quality as assessed by GAD-7, PHQ-9, and SQ-3, respectively, were not significant. These findings suggest that students are unable to recognize the cause of their decline in mental health but are more likely to blame the pandemic than medical school.

4.3. Limitations and future directions

This study comes with important caveats. First, it is important to note that the cyclical fluctuation that we attributed to the academic year has some confounding with the seasonal calendar. In the northern hemisphere, it is traditional to start the academic year during the fall and end in the summer. It would be interesting to have the opportunity to evaluate these effects in alternate locations where the seasonal calendar has different characteristics. This would allow for a more precise estimation of these effects. Second, the patterns we present are representative of a very specific subpopulation of students. Medical school in the United States comes after the completion of a bachelor's degree. These students are unlikely to be comparable to college students as they are older, more mature and go through a heavy pre-selection process for medical school admission. Third, the strategy and the setup of the COVID-19 assessment was opportunistic. The study was ongoing on the first class when the pandemic occurred. Strategies to accommodate for the changing conditions were introduced with the second class. Unfortunately, under these changing conditions, a true randomized experimental approach was never feasible. Last, a limitation of this study is its representativeness across institutions since data was collected from a single medical school consisting of two campuses. Also, the sleep quality analysis added a standardized assessment, PSQI, that was not utilized in the first year. Future research should expand the population surveyed across multiple universities and to include comparisons to other professions that require a terminal degree (graduate school, dentistry school, law school, among others) to better generalize results. This would expand the inference of our findings to students in similar maturity and age range. Furthermore, studies should incorporate third- and fourth-year medical students to provide data that spans the entire medical school training.

4.4. Conclusion

Based on our sample and considering the inherent limitations of our non-interventional study, we conclude that the first two years of medical school display changes in the levels of depression, anxiety, and poor sleep quality; we attribute this effect mostly to the academic and seasonal cycle. Self-reported chronic medical conditions and decreased total number of sleep hours were identified as risk factors for poor mental health and poor sleep quality among medical students. The COVID-19 pandemic did not have an effect we could isolate on medical student mental health or sleep patterns. Medical schools and students themselves should take an active role in addressing sleep patterns and mental health to achieve better well-being, improved academic performance, and ultimately become better physicians.