Regular physical activity (PA) is key for the prevention and management of chronic conditions such as cardiovascular disease, obesity, type 2 diabetes mellitus, and some types of cancer (34). In order to increase the low proportion of adults who comply with current PA recommendations, it is recommended that physicians provide PA counseling (8,18,35). But it is known that physicians counsel about PA at insufficient rates (35) and that this is unlikely to spontaneously change soon, since limited training is provided on this topic, as demonstrated by one recent survey that found a PA curriculum in only 13% of U.S. medical schools (15).
There is a small though growing body of literature that demonstrates a positive relationship between physicians' personal health practices and physicians' patient counseling (5-9). The specific relationship between physicians' personal PA behavior and PA counseling of patients has been established in the U.S. among women physicians (12,13) and practicing (38) and primary care physicians of both genders (24,36). Other health practices or factors associated with patient exercise counseling have included PA level (6-10), clinical specialty (6-9), age (6-9), stress levels (7), ethnicity (6,7), and obesity (7,8).
But the evolution of physicians' personal PA practices and counseling practices, and the evolution of the relationship between the two sets of practices, is unknown. There is one small study assessing the personal health behaviors of a single class of medical students; this study suggests that the personal health practices of physicians, including PA, are brought to medical school rather than established during medical school (20). And while PA levels of U.S. medical students have been studied in a few investigations (N = 105 (20); N = 835 (25)), and we have previously examined PA habits of our cohort at freshman orientation (14), none has examined PA and related counseling practices (or their relationship) in a representative sample of U.S. medical students through medical school. Our "Healthy Doc = Healthy Patient" study measures multiple personal health and patient counseling practices, including personal and clinical PA practices and important correlates in a representative sample of U.S. medical students from 16 medical schools during 4 yr. The purpose of this study is to determine the factors predicting PA levels and PA counseling practices of U.S. medical students and measure whether and how this changes during medical school.
All medical students in the class of 2003 at 16 U.S. schools were eligible to complete three questionnaires administered at different times during their years of medical training (at freshman orientation, orientation to wards, and during their senior year). School participation was encouraged by offering the summary use of school-specific data (in aggregate and without student identifiers).
Our sampling frame of schools was designed to reflect all U.S. medical schools in terms of age (freshman average = 24 yr old vs 24 nationally), school size (average number of students per school = 563 vs 527 nationally), NIH research ranking (school average = 64 vs 62 nationally), private/public school balance (51% private schools vs 41% nationally), underrepresented minorities (13% blacks, Hispanics, and Native Americans, vs 11% nationally), gender (45% women vs 43% nationally), and geographic distribution (2-4,26).
Using an IRB-approved protocol, confidential questionnaires were administered to students outside of formal classroom or training time. Students were instructed that their participation was voluntary.
The overall response rate across all three time periods was 80.3%; nonresponse rates for individual questionnaire items were a median of 3%. Students' responses were linked across time using a unique identifier consisting of mother's initials at her birth and father's first two initials. At freshman orientation, 2080 students were eligible to complete the survey and 1846 responded; 1982 were eligible at entry to wards and 1630 responded; 1901 were eligible at senior year and 1469 responded. Of the 2316 students who provided responses, 71.6% (N = 1658) did so at more than one time point; 971 responded at three time points, 687 at two, and 658 at one. School response rates were 48-98%; including responses from the 17th, protocol-noncompliant school, gives a conservative figure of 80.3% responding overall. Not all students were eligible to respond at all three survey points; for example, because of students leaving or returning from pursuing another degree.
Students' current exercise habit was queried as the frequency and duration of minimal, moderate, and strenuous exercise as defined by Godin et al. (16). Both total minutes of exercise and Godin's metabolic equivalent-based score per week were calculated. As well, a dichotomous measure was created for whether or not the student met the CDC PA recommendation (must either engage in vigorous PA for at least 20 min on 3 d·wk−1, or at least 30 min of moderate PA on 5 d·wk−1) (35). Independent variables, chosen a priori based on past literature concerning PA related counseling behaviors, covered the general areas of demographics, personal health-related habits, physical and mental health status, opinions on prevention, intended specialty, and school environment related to personal health promotion.
Our primary outcomes of interest were two variables concerning medical students' patient counseling on PA: 1) perceived relevance of this counseling in the student's intended practice, and 2) self-reported (though validated (11)) frequency of counseling a "typical general medicine patient." For the validation, as part of a lengthy questionnaire, 88 senior medical students answered these five separated questions: "With a typical general medicine patient, how often do you actually talk to patients about 1) nutrition; 2) exercise/physical activity; 3) weight; 4) smoking cessation (among smokers); and 5) alcohol?" And as part of their internal medicine subinternship final exam, students clinically assessed four SP cases with predetermined risk factors (poor diet, exercise, alcohol, and/or cigarette smoking habits). The possible questionnaire responses for relevance were "not at all/somewhat/highly" and for frequency were "never-rarely/sometimes/usually-always." Counseling relevance was queried at all time points, while frequency was measured only on the senior year survey. To maximize sample size, outcome variables were collapsed to "highly" versus "less than highly" for relevance and "usually/always" versus "less than usually/always" for frequency.
Cross-tabulations with the independent covariates were generated for both percent compliance with the CDC PA recommendation and median Godin exercise score. To assess significant relationships in the cross-tabulations, bivariate logistic regressions were performed for CDC PA recommendation adherence, and bivariate linear regressions were performed with exercise scores. Finally, both series were tested for time-point interactions.
Multivariate logistic regression procedures were used to model the outcome variables. Analyses for the relevance models were restricted to entry to wards and senior year because potentially significant environmental measures, including the school health promotion score, were not measured at the freshman time point. Continuous/ordinal variables were checked for linearity. Stepwise model selection procedures were used in selecting the final models. SUDAAN (32) was used to account for the correlated structure of the data within school and subject over time. Variables were dropped out of the final model individually until all variables remaining in the model were significant at a P < 0.10 level. A priori determined interaction terms were tested in the models until all highest-order terms were significant at a P < 0.10 level. Gender was brought back in as a control variable. Models were checked for multicollinearity. Residual analyses and influence diagnostics were performed to check for model lack of fit. Variables that were present in either of the final models were cross-tabulated with perceived relevance and reported frequency of exercise counseling. Chi-square and t-tests are reported.
Table 1 shows the distributions of important variables related to the students' PA habits. Maintaining relatively stable levels over time, an average of 61% of students were adherent with the CDC PA recommendations. Black/African American students consistently reported the lowest rate, and Hispanic students reported the highest rate of compliance with PA recommendations. Students intending to specialize in primary care had the lowest, while those intending to subspecialize had the highest rates of adherence with PA recommendations. Also, students who reported less stress and fewer days of "bad mental health" were more likely to meet PA recommendations and reported higher PA levels (P < 0.01).
Students who felt more positive about their schools' (P = 0.02) and their classmates' attitudes (P = 0.007) towards exercise promotion were more likely to be adherent with the PA recommendations. Exercising with classmates on a regular basis was also associated with increased adherence to PA guidelines. Strongly agreeing with the statements, "In order to effectively encourage a patient, a physician must also adhere to a healthy lifestyle," "I will be able to provide more credible and effective counseling if I exercise and stay fit," and "Medical school faculty members should set a good example by practicing a healthy lifestyle," were associated with an increased likelihood that the student would be adherent with the PA recommendations. Finally, a self-assessment of adequate personal PA levels showed a strong positive linear association with PA recommendations adherence (P < 0.0001).
More students perceived exercise counseling to be highly relevant to their intended practices at entry to wards than at senior year (69 vs 53%, P < 0.01, from Table 2). A perception that PA counseling would be highly relevant to their intended practice was more likely among women, nonwhites, students intending to specialize in primary care, students who complied with the PA recommendations, and those who reported stress in the last 12 months. PA counseling was also reported as more relevant by students who strongly agreed with the statements, "My classmates encourage each other to exercise," "I will be able to provide more credible and effective counseling if I exercise and stay fit," and "Medical school faculty members should set a good example by practicing a healthy lifestyle," and by those who strongly disagreed with the statement, "Prevention is less interesting to me than treatment." Students with similar characteristics were also more likely to report currently offering PA counseling: those who intended to specialize in primary care, complied with PA recommendations, strongly agreed with the statement, "I will be able to provide more credible and effective counseling if I exercise and stay fit," or that "Medical school faculty members should set a good example by practicing a healthy lifestyle," or they disagreed with the statement, "Prevention is less interesting to me than treatment." The mean school health promotion and exercise scores for students who reported usually or always offering PA counseling were higher than for those who never or sometimes offered PA counseling.
Table 3 shows the modeled estimated odds ratios for the analysis of predictors of perceived relevance of PA counseling. Students who perceived that PA counseling would be relevant to their future practices were more likely to be nonwhite, have experienced stress in the last 12 months, agree with the statements, "My classmates encourage each other to exercise," "Medical school faculty members should set a good example by practicing a healthy lifestyle," and (marginally) that "I will be able to provide more credible and effective counseling if I exercise and stay fit"; disagree with the statement, "Prevention is less interesting to me than treatment"; and adhere with CDC PA recommendations. Specializing in ob/gyn, specializing in nonprimary care or being undecided, and being in the senior year of study decreased the odds that students would perceive exercise counseling as relevant to their intended practice.
Table 4 shows the modeled estimated odds ratios for predictors of reported frequency of PA counseling among seniors. Associated with increased odds of offering PA counseling to patients were having a more health-promoting school environment, agreeing with the statement, "I will be able to provide more credible and effective counseling if I exercise and stay fit," and disagreeing with the statement, "Prevention is less interesting to me than treatment," and reporting more PA. Specializing in ob/gyn (OR = 0.91), nonprimary care (OR = 0.51), or being undecided about their future specialties (OR = 0.60) decreased the odds that students reported offering PA counseling to their patients.
This is the first investigation in which the exercise habits of a representative sample of U.S. medical students has been studied, and it shows a positive relationship between personal PA habits and counseling patients about PA. Previous studies of women physicians (13), primary care physicians (1), and internists (21,37) have also shown that personal PA levels are a significant predictor of patient counseling. We know that if physicians convey their personal positive PA habits, patients find their PA advice more believable and motivating (13), and CDC suggests that health professionals should exercise, not only to benefit their own health but also to make their endorsement of an active lifestyle more credible (29). Our results suggest that strategies to maintain and promote PA among medical students may enhance their frequency and perceived relevance of PA counseling to their patients.
U.S. medical students are more active than their age-matched counterparts in the general population; 61% met the CDC PA recommendation (with a median age of 26) (35). National estimates indicate that 57% of 18- to 24-yr-olds and 50% of 25- to 34-yr-olds meet this recommendation, with somewhat higher rates among those more educated (54% of college-educated 25- to 34-yr-olds) (7,22). Previous investigations carried out in selected medical schools have also shown that medical students are more physically active than their age-matched counterparts in the general population (19,20,28). In our sample, we found a temporary and nonsignificant drop in the percentage of students who met the PA recommendations, from a freshman year rate of 64%, down at to entry to wards (56%), and back up by senior year (62%; P = 0.4). Konen et al. (20) found that between years 2 and 4 of one medical school class (N = 105), the percentage of students regularly exercising decreased, although not significantly (65 to 54%, P = 0.13). A few other earlier small studies of single class of medical students also reported a decline in exercise habits 1 or 2 yr after the initiation of medical school (N = 52 (9), N = 104 (40)). Importantly, though, we found that at all three time points, U.S. medical students are more active than adults of the same age in the general population, and this first large cohort study suggests that any decline at the time of entry to wards is very small, and temporary.
With the exception of Hispanic students, PA levels of U.S. medical students mirror ethnicity patterns seen in the general population, with higher exercise levels among Anglo Americans, and lower levels among African Americans and Asian Americans (6). Previous research has shown that educational attainment and socioeconomic status predict PA behaviors (10,17).
We found that U.S. medical students with higher rates of adherence with CDC PA recommendations reported less short-term (previous 2 wk) or long-term (previous 12 months) stress than their counterparts who did not meet PA recommendations. This inverse relationship between mental health and PA level was evident at all three assessment points and is consistent with prior research. Evidence extracted from four population surveys in North America found that the positive relationship between PA and mental health was independent of current health or socioeconomic status (33). Further, findings from the 2001 Behavioral Risk Factor Surveillance System show that the proportion of adults reporting greater than 14 d of bad mental health during the last month was lower among those adhering to PA recommendations in all age, racial, and sex groups (5). While our study is only an observational design and cannot assign causation, our findings help support the implementation of programs within medical schools to increase and maintain participation in regular PA to help medical students deal with stress and improve their mental well-being.
We found a positive correlation between students' PA levels and their compliance with CDC exercise recommendations and their reports of school and classmate exercise attitudes, and of exercising with classmates. Peer social support promotes physical activity (34), and Helping Relationships is one of the behavioral processes of change in the Transtheoretical Model of Behavior Change to increase PA levels, a theoretical foundation of this study (23,31,34). Again, while we are not able to ascribe causation, our findings suggest that encouragement from the medical school environment promotes higher PA levels among U.S. medical students.
U.S. medical students who meet CDC PA recommendations are more likely to recognize that physician PA levels influence patient PA levels. They are more likely to agree with the statements, "in order to effectively encourage a patient, a physician must also adhere to a healthy lifestyle"; "I will be able to provide more credible and effective counseling if I exercise and stay fit"; and "medical school faculty members should set a good example by practicing a healthy lifestyle." The relationship between physician PA habits and counseling patients to be more physically active is found among samples of practicing physicians (39), U.S. women physicians (12,13), and primary care physicians (36). Our results indicate that U.S. medical students who are physically active recognize this relationship, and they maintain this understanding throughout medical school training.
More students perceived PA counseling to be highly relevant to their intended practices at entry to wards than at senior year. The latter years of medical school are spent working with clinicians, many of whom do not feel competent to provide patient counseling, including providing counseling about exercise (24,27,39). A survey of 251 internal medicine residents at six training programs found that while only 28% felt confident in their skills to prescribe exercise for patients, 91% felt that training in exercise counseling would be worthwhile (30). Similarly, a survey of competence in and relevance of patient counseling among 58 internal medicine residents found that exercise counseling relevance scores were significantly higher than competence scores; consistent with our data about students planning to work in primary care, to counsel confidently, and to be trained in counseling, and that those having a high priority to exercise more were also more likely to counsel on exercise (13). Increasing the medical school counseling curriculum should improve physicians' counseling competence, confidence, and perceptions of importance and relevance.
Frequency of PA counseling among senior U.S. medical students is predicted by personal PA level, having a more health-promoting school environment, and by agreeing with the statement, "I will be able to provide more credible and effective counseling if I exercise and stay fit." Previous data have also shown that U.S. women physicians (N = 4501) complying with CDC's personal exercise recommendations were more likely to counsel patients on exercise. Likewise, another smaller study has shown that physicians (N = 175) who are more fit (lower resting heart rate and higher self-reported PA level) were more likely to counsel their patients to be more physically active (36). Similarly, in a study of 91 county agents responsible for health promotion in Kansas, those who met CDC exercise recommendations were more likely (P < 0.05) to adopt a PA program for their county (69% adopting) than were agents who were less active (41%) or inactive (11%).
Strengths of this investigation include the large sample size, selection of a representative group of U.S. medical students, prospective design, high response rate, and validation of self-reported exercise counseling frequency with standardized patients (11). Limitations of this study include the potential social-desirability and information biases inherent in self-reported data.
In summary, we found that 61% of U.S. medical students meet current PA recommendations throughout medical school. Higher personal adherence with PA recommendations is related to better mental health, intention to be a subspecialist, greater perceived relevance of PA patient counseling, and higher frequency of PA patient counseling. Having a more health-promoting school environment and coexercising are also correlated with positive PA counseling attitudes and practices. The relationship between personal PA habits and PA counseling practices is already present at the beginning of medical training, and this relationship is also influenced by medical school experiences and environment. These findings support the implementation of programs aimed at increasing the proportion of medical students adopting and maintaining regular PA habits as a strategy to increase exercise and disease prevention in the general population.
The results of the present study do not constitute endorsement by ACSM. We would like to acknowledge the support of the American Cancer Society for this work.
1. Abramson S, Stein J, Schaufele M, Frates E, Rogan S. Personal exercise
habits and counseling practices of primary care physicians
: a national survey. Clin J Sport Med
2. American Association of Medical Colleges. AAMC data warehouse: student records system as of December 12, 2000, reflecting those actively enrolled on October 31, 2000; [cited 2006 April 18]. Available from: http://www.aamc.org/data/facts/archive/famg82001.htm
5. Brown DW, Balluz LS, Heath GW, et al. Associations between recommended levels of physical activity and health-related quality of life. Findings from the 2001 Behavioral Risk Factor Surveillance System (BRFSS) survey. Prev Med
8. Chakravarthy MV, Joyner MJ, Booth FW. An obligation for primary care physicians
to prescribe physical activity to sedentary patients to reduce the risk of chronic health conditions. Teach Learn Med
9. Crapse FJ, Hudgins PM, Baker HH. Lifestyle changes associated with osteopathic medical education
. J Am Osteopath Assoc
10. Crespo CJ, Smit E, Andersen RE, Carter-Pokras OA, Ainsworth BE. Race/ethnicity, social class and their relation to physical inactivity during leisure time: results from the Third National Health and Nutrition Examination Survey, 1988-1994. Am J Prev Med
11. Frank E, McLendon L, Elon LK, Denniston M, Fitzmaurice D, Hertzberg V. Medical students' self-reported typical counseling practices are similar to those assessed using standardized patients. Medscape Gen Med.
[Internet]; [cited 2006 June 13];7(1). Available from: http://www.medscape.com/viewarticle/497041
12. Frank E, Rothenberg R, Lewis C, Fielding B. Correlates of physicians
' prevention-related practices: findings from the Women Physicians
' Health Study. Arch Fam Med
13. Frank E, Schelbert KB, Elon LK. Exercise
counseling and personal exercise
habits of U.S. women physicians
. J Womens Health
14. Frank E, Wright E, Serdula M, Elon L, Baldwin G. Personal and professional nutrition-related practices of US female physicians
. Am J Clin Nutr
15. Garry JP, Diamond JJ, Whitley TW. Physical activity curricula in medical school. Acad Med
16. Godin G, Shephard RJ. A simple method to assess exercise
behavior in the community. Can J Appl Sport Sci
17. He XZ, Baker DW. Differences in leisure-time, household, and work-related physical activity by race, ethnicity, and education. J Gen Intern Med
18. Jacobson DM, Strohecker L, Compton MT, Katz DL. Physical activity counseling in the adult primary care setting: position statement of the American College of Preventive Medicine. Am J Prev Med
19. Kamien M, Power R. Lifestyle and health habits of fourth year medical students at the University of Western Australia. Aust Fam Phys
. 1996;(Suppl. 1):S26-S29.
20. Konen JC, Fromm BS. Changes in personal health behaviors of medical students. Med Teach
21. Lewis CE, Clancy C, Leake B, Schwartz JS. The counseling practices of internists. Ann Intern Med
22. Macera CA, Ham SA, Yore MM, et al. Prevalence of physical activity in the United States: Behavioral Risk Factor Surveillance System, 2001. Prev Chronic Dis
23. Marcus BRJ, Selby V, Niaura R, Abrams D. The stages and processes of exercise
adoption and maintenance in a worksite sample. Health Psychol
24. McKenna J, Naylor PJ, McDowell N. Barriers to physical activity promotion
by general practitioners and practice nurses. Br J Sports Med
25. Najem GR, Passannante MRC, Foster JD. Health risk factors and health promoting behavior of medical dental and nursing students. J Clin Epidemiol
27. Orleans CT, George LK, Houpt JL, Brodie KH. Health promotion
in primary care: a survey of U.S. family practitioners. Prev Med
28. Peterson DF, Degenhardt BF, Smith CM. Correlation between prior exercise
and present health and fitness status of entering medical students. J Am Osteopath Assoc
29. Piazza J, Conrad K, Wilbur J. Exercise
behavior among female occupational health nurses. Influence of self efficacy, perceived health control, and age. AAOHN J
30. Rogers LQ, Bailey JE, Gutin B, et al. Teaching resident physicians
to provide exercise
counseling: a needs assessment. Acad Med
31. Sallis JF, Owen N. Physical Activity & Behavioral Medicine
. Thousand Oaks (CA): Sage Publications, Inc.; 1999.
32. Shah BV, Barnewell BG, Bieler GS. SUDAAN User's Manual
. Research Triangle Park (NC): Research Triangle Institute; 1995.
33. Stephens T. Physical activity and mental health in the United States and Canada: evidence from four population surveys. Prev Med
34. U.S. Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General
. Atlanta (GA): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion
36. Walsh J, Swangard D, Davis T, McPhee S. Exercise
counseling by primary care physicians
in the era of managed care. Am J Prev Med
37. Wells KB, Lewis CE, Leake B, Schleiter MK, Brook RH. The practices of general and subspecialty internists in counseling about smoking and exercise
. Am J Public Health
38. Wells KB, Lewis CE, Leake B, Ware JE. Do physicians
preach what they practice? JAMA
39. Williford HN, Barfield BR, Lazenby RB, Olson MS. A survey of physicians
' attitudes and practices related to exercise promotion
. Prev Med
40. Wolf TM, Kissling GE. Changes in life-style characteristics, health, and mood of freshman medical students. J Med Educ