The relationship between academic performance and recreation use among first-year medical students

Figures & data

Fig. 1.  Distribution of number of CR visits 21 days before an exam. The distribution was derived from approximately 20 exam periods each for 408 first-year medical students across four cohorts from 2006 to 2010.

Table 1 Characteristics of first-year medical students (N=408)

	N (%) or Mean (Std. Dev)
Changes in exam scores and CR visits between exam dates
ΔExam score (percentage points)	0.16 (0.12)
ΔCR visits 21 days before exam	−0.08 (0.02)
ΔCR visits 21 days after exam	0.01 (0.02)
Demographic and baseline academic characteristics
Asian	122 (29.9%)
African American	24 (5.9%)
Hispanic	42 (10.3%)
Caucasian	200 (49%)
Missing race or ethnicity	20 (4.9%)
Male	231 (56.6%)
In-state resident	289 (70.8%)
Age at matriculation	23.95 (2.62)
Undergraduate GPA	3.56 (0.35)
MCAT Score–Biological Sciences	10.18 (1.62)
MCAT Score–Physical Sciences	9.95 (1.86)
MCAT Score–Verbal Reasoning	9.34 (1.34)

ΔExam score refers to changes in aggregate exam score from one exam date to the next, which is usually in 2–3 weeks (though exam dates can be closer together). Changes in exam scores are interpreted as percentage point differences between the scores on one exam date and the next. ΔCR visits refers to the changes in CR visits 21 days before or after an exam between one exam date and the next.

Fig. 2.  Average exam scores and CR visits by month of the academic year. Each point corresponds to the mean percentage exam score (in blue), and number of campus recreation (CR) visits (in red) for 408 first-year medical students for exams given in each month of the academic year. Number of CR visits was calculated as the number of times a student visited a CR facility in the 21-day period prior to an exam date.

Fig. 3.  Conditional means of CR visits and exam performance. An exam score of 60% represents the passing level for most exams. The data are based on approximately 20 exam sessions each for 408 first-year medical over 4 years. In (a), the number of CR visits is calculated as the number of times a student visited a CR facility in the 21-day period after an exam date. In (b), the number of CR visits is calculated as the number of times a student visited a CR facility in the 21-day period prior to an exam date.

Table. 2. Results of Ordinary Least Squares (OLS) regression models estimating the reciprocal relationship between changes in academic performance and changes in CR use

	(1)		(2)
	ΔCR visits 21 days after exam		ΔExam score
	Beta	Std. error	Beta	Std. error
ΔExam score	0.007***	0.002
ΔCR visits 21 days before exam			0.180**	0.079
Asian	0.025	0.023	−0.052	0.073
African American	0.088**	0.037	0.220	0.153
Hispanic	0.059**	0.029	−0.053	0.103
Missing race or ethnicity	−0.017	0.061	0.037	0.098
Male	0.000	0.017	0.031	0.061
In-state resident	−0.001	0.024	0.032	0.067
Age at matriculation	0.003	0.002	0.022**	0.011
Undergraduate GPA	0.015	0.027	−0.016	0.097
MCAT Score–Biological Sciences	0.005	0.008	−0.021	0.023
MCAT Score–Physical Sciences	−0.007	0.008	−0.022	0.021
MCAT Score–Verbal Reasoning	0.003	0.005	0.001	0.016

Caucasian is the reference category for race and ethnicity. Year indicator variables are included in both models.

CR: campus recreation. ***p<0.01, *p<0.05.

Fig. 4.  Predicted changes in CR use and academic performance from regression models. Graph (a) plots the average predicted values of ΔCR visits from the regression shown in Column 1 of Table 2. Graph (b) plots the average predicted values of ΔExam Score from the regression shown in Column 2 of Table 2.