Yoga as a Complementary Activity for Medical Students Learning Anatomy | BMC Medical Education



In the 2018-2020 academic years at the Stritch School of Medicine, all first-year medical students were required to take anatomy courses in the first semester of medical school concurrently with two longitudinal courses. The anatomy class per week was as follows: an average of 8 hours of in-person lessons with two or three 4-hour in-person labs where students rotated and explored an already dissected body. This averages out to about 16-20 hours of scheduled class time. About every three weeks there was an anatomy review regarding material from the last three weeks. There were a total of four exams for the course.

The study included first-year medical students at the Stritch School of Medicine (SSOM) who were enrolled in the fall semester anatomy course. The study excluded students with a medical condition or physical limitation that would prevent them from participating in yoga. All students enrolled in the anatomy course (335 over the 2 years) were notified of the study by email approximately one month before the start of the anatomy course. A week before the start of anatomy class, interested students were invited to a question and answer session regarding the study. The research study was approved by the Loyola University Chicago Health Sciences Campus (LUCHSC) Institutional Review Board (IRB), and the research protocol was performed in accordance with relevant guidelines and regulations. of the IRB of the LUCHSC. All participation was voluntary and all participants provided informed consent before participating in any study activity.

study design

All participants completed an intake survey to capture basic characteristics, including their age, race, learning style, previous anatomy experience, and previous yoga experience. For this study, traditional students were those who immediately enrolled in medical school, while non-traditional students were those who had a gap of at least one year between graduation and enrolling in medical school. Medical School. Students’ primary learning style was also categorized as visual, auditory, reading/writing, or kinesthetic, as described by Fleming [16] and Leite et al. [17].

Prior to each anatomy exam, all participants also completed the Perceived Stress Survey 4 (PSS-4) which measures the frequency of stress on a scale of 0 to 16 points (where higher scores indicate more frequent stress) [18, 19]. Students assigned to the yoga intervention cohort completed eight yoga sessions. Before and after each session, these students self-reported their confidence in the anatomical content covered in that session ranging from 0 (not confident) to 4 (very confident), stress on an 11-point score scale ranging from 0 ( no stress) to 10 (excessive stress) and completed the stress subscale of the Depression, Anxiety, and Stress Scale Assessment (DASS-21); this subscale ranges from 0 to 42 points (where higher scores indicate more distress) [20].

As for the yoga sessions, a 200hr certified yoga teacher trained in alignment yoga (a form of Hatha yoga) taught each yoga session. The teacher created a script which was used to maintain consistency across grades. These sessions included both beginner and continuous yoga poses. The yoga flow followed the general sequence taught in alignment yoga: pre-yoga, standing poses, inversions, back bends, forward bends, twists, pranayama, and relaxation. Each yoga session included the anatomy class topics for that week. For example, if students were taught upper limbs, yoga poses were described using the anatomical locations, innervations, actions, and functions of upper limb muscles.

Calculation of power and sample size

A a priori The calculation of power and sample size was estimated to test the null hypothesis that the average performance in anatomy and physiology exams was the same between students taking the yoga class (intervention) and those who wait (control). Group sample sizes of 27 people randomly assigned to the yoga condition and 27 people randomly assigned to the control condition achieved 81.2% power to detect a difference of 7 review points ( out of 100) in a design with four examinations (or repeated observations). This calculation assumed an attrition rate of 20% in each group, a compound symmetry covariance structure for correlated student exam responses, and a standard deviation of 8.65 points. These assumptions were based on historical price data between 2016 and 2018. This calculation also assumed that the correlation between observations on the same subject was 0.80 and the alpha level was 0.05 [21,22,23].

statistical analyzes

First-year medical students were randomized to the yoga or waiting cohort using a 1:1 random allocation scheme. For these participants, demographic data are provided as valid counts and percentages for all nominal characteristics and as mean with standard deviation for age. For the main objective, a linear mixed-effects model was used to estimate the mean difference in exam performance between the experimental (Yoga) and control (Wait) groups. An interaction term was used to estimate the average difference between these two cohorts for each exam, and a Sidak correction was used to control for the type 1 error rate. Since students could contribute up to four exam results to the analysis, random interceptions were allowed for each student to account for their correlated exam performance. The degrees of freedom for these comparisons were estimated using the method of Kenward and Roger [24]. The same approach was used to compare group responses to the PSS-4 assessment.

Among students assigned to the yoga cohort, linear mixed-effects models were also used to estimate the mean change in self-reported stress response and DASS stress response from pre- to post-intervention. As before, a Sidak correction was used to control the Type 1 error rate for these comparisons, and since students could contribute up to two responses for each yoga session, random interceptions were allowed for each student to to take their correlated responses into account. The degrees of freedom for these comparisons were also estimated using the method of Kenward and Roger [24].

Finally, we used generalized estimating equations (GEE models) to estimate the odds of reporting greater confidence in each anatomical construct post-session compared to pre-session for students assigned to the intervention cohort of yoga. These models specified a multinomial distribution with a cumulative logit link for the ordinal confidence response and used robust standard errors to account for paired (correlated) student survey responses. As before, a Sidak correction was used to control the type 1 error rate for these comparisons. All analyzes were performed using SAS version 9.4 (Cary, NC).


About Author

Comments are closed.