Journal of Clinical Lipidology (2014) 8, 199–205
Integrating nutrition education into the cardiovascular curriculum changes eating habits of second-year medical students Eric J. Vargas, BS, Robert Zelis, MD* College of Medicine, Penn State University, Hershey, PA, USA (Mr. Vargas); and Lipid Clinic, Heart and Vascular Institute, College of Medicine, Penn State University, 500 University Drive, H047, Hershey, PA 17033–0850, USA (Dr. Zelis) KEYWORDS: Nutrition education; Medical curriculum; Eating habits; Personal cardiovascular risk; Rate Your Plate questionnaire
BACKGROUND: Survey of medical curricula continues to show that nutrition education is not universally adequate. One measure of nutritional educational competence is a positive change in student eating habits. OBJECTIVE: The objective of this study was to evaluate whether integrating nutrition education within the second-year cardiovascular course for medical students, using the ‘‘Rate Your Plate’’ (RYP) questionnaire, coupled with knowledge of student personal 30-year risk of a cardiovascular event was useful in changing students’ eating behaviors. METHODS: Thirty-two students completed an unpublished 24-item questionnaire (modified-RYP) about their eating habits in the spring of their first year. The same students then completed the questionnaire in the spring of their second year. Paired t test was used to analyze the difference in RYP scores. Pearson correlation coefficients were calculated for the Framingham 30-year cardiovascular event risk and change in RYP score to examine whether risk knowledge may have changed eating habits. RESULTS: Mean scores at baseline and 1 year later were 57.19 and 58.97, respectively (paired t test, P , .01). Correlation coefficient between 30-year relative risk, adjusted for family history, and change in RYP score was –0.322. CONCLUSION: Although medical students were eating healthy at baseline, integration of nutrition education within the second-year cardiovascular medical curriculum was associated with improved heart healthy eating habits. Because student attitudes about prevention counseling are influenced by personal eating habits, this suggests that students with a more healthy diet will be more likely to recommend the same for their patients. Ó 2014 National Lipid Association. All rights reserved.
Despite the recommendation of the National Academy of Sciences in 1985, which stated a minimum of 25 hours of nutrition education be provided to medical students over the 4-year curriculum, Adams et al1 found a 12% decline between 2004 and 2009 and noted only 27% of the 105 schools met the minimum.1 In 2001, .14,000 US * Corresponding author. E-mail address: [email protected] Submitted July 2, 2013. Accepted for publication November 15, 2013.
medical students completed the Association of American Medical Colleges graduation questionnaire; 56% felt their nutrition-related experiences were inadequate.2 Furthermore, 66% felt their clinical decision-making and care skills about nutrition were poor. Similar experiences have been noted among Canadian medical students.3 A number of suggestions have been made to combat this trend, such as development of a competency-based certification program modeled after the American Heart Association program in Cardiac Life Support.4 Rather than draft
1933-2874/$ - see front matter Ó 2014 National Lipid Association. All rights reserved. http://dx.doi.org/10.1016/j.jacl.2013.11.006
200 new curricula, wider distribution of already existing curricula has been advocated,4 such as teaching modules developed by the Nutrition in Medicine Project,5 which conform to the consensus guidelines of the American Society of Clinical Nutrition.6 Another curriculum guide was developed by the National Institutes of Health-supported Nutrition Academic Award Program.7,8 It is with this background that our institution appointed a multidisciplinary task force to determine how and when nutrition was taught in our curriculum and to make recommendations to ensure that nutrition education was adequately incorporated. The report, presented to faculty in 2006, was approved by the curriculum committee and distributed to the course committees for implementation. Many of the learning objectives of this task force mirrored those of the Nutrition Academic Award,7 especially objectives related to the cardiovascular system and were incorporated into the nutrition component of the organ system-based integrated second-year cardiovascular block. One nutritional objective not addressed in most guidelines is to promote changes in student eating habits toward more heart healthy behavior. Multiple studies have shown that first-year medical students tend to consume a lower fat diet with more fruits and vegetables and to exercise moderately.9–11 However, those eating habits begin to decline during the clinical years.10 This is important because medical students ‘‘preach what they practice.’’12 Students and physicians who have a healthy lifestyle are more likely to see the importance of diet and exercise in disease prevention and to counsel their patients accordingly.12–17 Unfortunately, the importance of prevention counseling declines over the 4 years of medical school.13,14,16 The purpose of this study was to determine whether integrating nutrition within the second-year cardiovascular course, plus 2 small changes to the curriculum, would result in more heart healthy personal dietary choices by medical students. The first was to introduce the Rate Your Plate (RYP) dietary questionnaire into the curriculum as a means for students to assess personal eating habits. This is one of the counseling tools presented in one of the papers in the Nutrition Academic Award symposium on Innovative Teaching Strategies.18 The second was to provide a free fasting lipid panel for study participants to calculate their personal 30-year risk of developing a cardiovascular event, as a means to stimulate interest to improve personal risk factors. The 2 hypotheses of our study were (1) integrating nutrition education while adding the RYP dietary assessment to the curriculum would improve heart healthy eating habits and (2) students with the higher perceived cardiovascular risk would improve their eating habits more.
Methods Study rationale The presentation of nutrition education within the cardiovascular course used a variety of components. There
Journal of Clinical Lipidology, Vol 8, No 2, April 2014 were 2 hours of lecture on lipoprotein transport and a third hour on practical nutritional advice, and 2 separate ProblemBased Learning (PBL) cases that had dietary elements. During the hypertension PBL case, students discussed the Dietary Approach to Stop Hypertension diet,19 and the coronary artery disease PBL case had a learning objective to use the Therapeutic Lifestyle Change diet recommended by the Third Adult Treatment Panel of the National Cholesterol Education Program.20 There was also a separate nutritional exercise. Students in the PBL groups of 8 students were required to analyze the eating habits of 2 patients, one with heterozygous familial hypercholesterolemia and the other with metabolic syndrome. Half the students were assigned to one or the other problem and were required to take specific dietary components, visit grocery stores and fast food and upscale restaurants, and determine what was ‘‘good’’ and ‘‘poor’’ about their patient’s dietary selections. They were then asked to recommend better choices to achieve lipid goals that were based on individual patient Framingham risk assessment. They returned approximately 2 weeks later for discussion of their findings and recommendations within the full PBL group. The learning objectives for these activities covered most objectives present in the Nutrition Curriculum Guide for Training Physicians,7 under the cardiovascular system (section D) and all those ranked in the top one-third of importance. An additional assignment at the same time was a third ‘‘patient’’, themselves. Most did little on this component for 2 reasons. Ten-year cardiovascular risk evaluation had little meaning for students in their mid-20s. This problem was resolved when in 2009 the Framingham Heart Study published a simple algorithm to quantify 30-year risk.21 The second problem was that students found quantifying their own 24-hour food consumption formally in terms of nutritional composition was too tedious and did not seem practical. After reviewing various dietary self-assessment tools, it seemed that the RYP questionnaire might be particularly suitable for our purposes. It was easy to complete, used language familiar to students, and dealt specifically with dietary components likely to affect cardiovascular health.18,22,23 RYP is a semiquantitative food-frequency questionnaire that provides a single numerical ranking of the quality of food choices and a ranking into 1 of 3 qualitative groups. It also provides positive reinforcement for healthy choices and specific recommendations on how less healthy choices can be improved, thus making it a useful tool for patients. RYP has been directly compared with the Willett food-frequency questionnaire and found favorable.22 The validation was stronger when RYP was administered before the Willett questionnaire, than when given after. The researchers suggested that the much longer 136-question Willett may have led to a more haphazard filling out of the brief RYP. Moreover, as noted above, RYP has been advocated as a strategy to train primary care physicians in nutritional skills.18 However, we were concerned that using identical questionnaires 3 times to train students and to assess student eating habits before and after training might confound the
Vargas and Zelis
Changing medical student eating habits
201
results. The RYP used in the second-year course could have been reviewed by some students before taking the follow-up RYP after training and might have influenced the responses. This problem was avoided by using an unpublished updated RYP for assessment. The unpublished version of the RYP questionnaire was dissimilar in a number of ways, including the reversal of the order of response. This allowed us to use a different instrument for before and after testing while using the earlier published version of RYP for teaching. The study design entailed testing a volunteer group of first-year medical students in the spring before the fall cardiovascular course and retesting the same students the next spring, 6 months after the cardiovascular course ended.
coded identification number and recorded in the database in a manner to preserve participant confidentiality. The following 2 study end points were preselected for statistical analysis: (1) the initial and follow-up RYP scores were compared by 2-sided paired t test and (2) the change in RYP score was compared with the individual relative risk by using the Pearson correlation coefficient after adjusting relative risk for a family history of premature vascular disease. Data are presented as mean 6 SD. P values , .05 were considered significant. The changes in score for each of the 24 RYP items were scanned to determine which items contributed (610%) to a change in the mean score but were not analyzed for significance. The 2005 and 2009 versions of RYP were supplied by the author of the questionnaire22,23 and used with permission of the copyright holder (Institute for Community Health Promotion, Brown University, Providence, RI). The 2005 version was used by the entire class for education and was available in print22 and online at their website at the time of the study. The 2009 version was used as the evaluation instrument and replaced the 2005 version online after the study was completed and now appears with a 2010 copyright.
Research protocol The research project was reviewed and approved by the institutional review board (protocol 36289). First-year medical students were recruited from the class of 2014 by an e-mail invitation sent to the entire class of 140 students. Interested students were asked to attend a meeting in which the project was discussed. A complete written description of the protocol was given to each student. Participation was voluntary and self-selected. Students who wished to participate were given a numerical identification code by a student co-investigator to preserve anonymity. This was used to logon to a secure website to answer the RYP questionnaire and demographic questions. The students were told which of the latter specific questions would be asked and were instructed how to secure the answers before logon. When all information had been recorded, students were issued a voucher for a fasting lipid panel, the results of which were made available to them subsequently at the secure site. At a second login, students were presented with the results of the lipid panel and their personal relative and absolute 30-year risk. They were given a 1-page explanation to aid in the interpretation of the lipid results, 30-year risk, and interpretation of their blood pressure, weight, body mass index (BMI; calculated as weight divided by height square; kg/m2), and activity level. It was explained that a family history of a premature vascular event in a first-degree relative doubles their risk; a similar history in a grandparent increases risk by approximately 50%.24 They were also told that a positive family history of hypertension or diabetes doubles their chance of developing that problem.25 We listed the major components of the metabolic system, and its risk for diabetes was noted. Finally, they were given a list of suggested dietary changes to mitigate risk. They were given the simple dietary advice to lower low-density lipoprotein (LDL) cholesterol by eliminating trans fats and reducing saturated fats, while substituting unsaturated fat and using a plant stanol/sterol-ester margarine. Blood samples were drawn by the clinical laboratory of the Hershey Medical Center whereby serum triglyceride, total, and high-density lipoprotein (HDL) cholesterol levels were determined, and LDL and non-HDL cholesterol values were calculated. All reports used the
Results Characteristics of study population Of the 44 first-year medical students who completed the questionnaire, 32 (73%) completed the follow-up study 1 year later, approximately one-quarter of the class. The 32 students in the study group had the following characteristics: equal sex distribution; age, 23.9 6 4.1 years; weight, 153.5 6 32.8 pounds; waist circumference, 31.1 6 3.5 inches; BMI, 23.9 6 4.1; blood pressure (systolic/diastolic), 116.5 6 9.2/73.5 6 6.1 mm Hg. With the use of BMI criteria, 4 men and 1 woman were overweight; and 1 man and 3 women were obese. Five women were using hormones. No student was taking antihypertensive or lipid-modifying medication. None had diabetes; none smoked. Only 2 students had a first-degree relative with a premature coronary event; 10 students had a grandparent with a premature coronary event. Nineteen students had relatives with a history of type 2 diabetes and 22 had relatives with hypertension. The only personal data that changed over the course of the year was a 2.6-pound (1.7%) weight increase to 156.1 6 35.2 pounds, which was close to being significant (P 5 .052). Of the 31 students with complete information about their daily exercise, at baseline 19 exercised $30 minutes, 7 exercised 15 to 30 minutes, and 5 exercised ,15 minutes. The only insignificant changes at the time of the second questionnaire were 2 fewer students in the lower group and 2 more in the middle group. Twenty-eight students had lipid information drawn at the beginning of the study. The mean values were total cholesterol of 165.6 6 20.0 mg/dL, HDL cholesterol of 50.8 6 10.8 mg/dL, triglycerides of 72.1 6 23.8 mg/dL, LDL
202
Journal of Clinical Lipidology, Vol 8, No 2, April 2014
Table 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Responses to the individual items that comprise the RYP questionnaire at 2 time points Individual RYP dietary questions
RYP 1
SD 1
RYP 2
SD 2
Change
Change. %
Meat cuts Chicken and turkey Ground meat and poultry Processed meat and poultry Portion size of meat and poultry Fish, shellfish Cooking method (meat, poultry, fish) Meatless meals Whole eggs Milk (yogurt, cream), % fat Cheese (included in meals), fat type Servings of dairy foods Whole grains Fruits and vegetables Cooking method (vegetables, pasta) Fat type in cooking/baking Salt from processed foods Table spreads Salad dressings, mayonnaise Snack foods Nuts, seeds Frozen desserts Sweets, pastries, candy Eating out/take out
2.63 2.81 2.56 2.53 2.06 1.63 2.72 2.47 2.63 2.69 2.00 2.50 2.28 2.09 2.31 2.91 2.06 2.34 2.41 2.59 2.25 2.25 2.09 2.38
0.61 0.59 0.56 0.51 0.72 0.75 0.52 0.76 0.71 0.64 0.80 0.62 0.68 0.59 0.54 0.39 0.62 0.90 0.76 0.61 0.80 0.72 0.69 0.75
2.78 2.69 2.66 2.53 2.19 1.63 2.81 2.50 2.63 2.69 2.13 2.41 2.34 2.41 2.53 2.97 2.25 2.75 2.31 2.63 2.13 2.53 2.22 2.28
0.42 0.54 0.55 0.51 0.74 0.79 0.40 0.76 0.61 0.69 0.79 0.71 0.70 0.56 0.57 0.18 0.72 0.62 0.78 0.49 0.79 0.62 0.61 0.73
0.16 –0.13 0.09 0.00 0.13 0.00 0.09 0.03 0.00 0.00 0.13 –0.09 0.06 0.31 0.22 0.06 0.19 0.41 –0.09 0.03 –0.13 0.28 0.13 –0.09
5.95 –4.44 3.66 0.00 6.06 0.00 3.45 1.27 0.00 0.00 6.25 –3.75 2.74 14.93 9.46 2.15 9.09 17.33 –3.90 1.20 –5.56 12.50 5.97 –3.95
RYP 1, baseline score before intervention; RYP 2, score 1 year later and 6 months after intervention. A value of 1.0 is most heart unhealthy and 3.0 is most heart healthy. Results are mean and SD.
cholesterol of 100.4 6 19.5 md/dL, and non-HDL cholesterol of 114.8 6 19.8 mg/dL. Three men had HDL , 40 mg/dL, and 5 women has HDL , 50 mg/dL; none had a triglyceride value . 150 mg/dL, one had LDL cholesterol . 130 mg/dL, and one had a non-HDL cholesterol . 160 mg/dL. With the use of the lipid data and demographic information, the calculated Framingham 30-year risk for a cardiovascular event was 3.67% 6 1.33% (range, 1.4%–7.0%). Average risk for comparable age and sex was 3.06% 6 0.80%, and optimal was 2.53% 6 0.67%. Relative risk for the group was 1.22 6 0.44. When family history was included, relative risk was 1.29 6 0.72.
RYP questionnaire responses The RYP evaluation questionnaire consisted of 24 items. Three responses were presented for each item; each response was preceded by a verb–adverb combination such as ‘‘usually eat,’’ ‘‘sometimes eat,’’ ‘‘rarely eat,’’ ‘‘usually use,’’ and so forth. The responses went from an eating pattern of most heart unhealthy to most heart healthy and corresponded to values of 1, 2, or 3, with 3 representing the most heart healthy pattern. When the numbers for all 24 responses were added, a single score was derived from 24 to 72. For the evaluation instrument, the 3 qualitative assessment categories were as follows: 58 to 72, ‘‘you are
making many healthy choices’’; 41 to 57, ‘‘there are some ways you can make your eating habits healthier’’; 24 and 40 ‘‘there are many ways you can make your eating habits healthier.’’ The mean baseline RYP score was 57.19 6 5.65. Eighteen students were in the most heart healthy eating category, 14 in the middle, and none in the lowest. Thus, at baseline, these first-year medical students appeared to be eating quite well with respect to heart healthy choices. One year later, 6 months after the cardiovascular course ended, a significant improvement was found in the RYP score to 58.97 6 4.97, a change that was significant by paired t analysis (P 5 .008). Table 1 displays the values for each of the 24 items about which inquiry was made, at baseline and 1 year later. Absolute and percentage of changes are indicated. The lowest possible response is 1.0, the highest 3.0. These data are presented primarily to form an impression what might have contributed to the significant change in mean score. Statistical analyses were not performed because of the severe P value correction penalty for 24 comparisons. For most categories the change was positive, and 3 showed improvements of .10%: fruits and vegetables daily frequency improved 15%; table spreads qualitative type improved 17%; and frozen desserts qualitative type and frequency improved 12.5%.
Vargas and Zelis
Changing medical student eating habits
Figure 1 Change in RYP dietary score related to 30-year Framingham relative risk of a cardiovascular event adding the effect of a positive family of premature vascular event (doubling for parent, 50% increase for grandparent). Premature is ,55 years of age for a man and ,65 years of age for a woman. RYP, Rate Your Plate.
Change in RYP score related to personal cardiovascular risk Because 4 students did not have a lipid panel drawn, data from 28 students were available for analysis. Figure 1 shows individual data points that relate the adjusted relative 30-year risk of a cardiovascular event to change in RYP score. The relative risk was increased by 1.0 for students with a parent and 0.5 for a grandparent with a premature vascular event. The change in RYP score did not correlate with personal risk (Pearson’s correlation coefficient, –0.322). The change in score was also not significantly correlated to unadjusted risk (correlation coefficient, –0.223).
Discussion The main finding of this study was that dietary eating habits of medical students as measured by the RYP dietary questionnaire modestly improved over the 1-year period that spanned their second year. We feel that nutrition education within an integrated cardiovascular curriculum contributed to this change in the group of students who volunteered to be evaluated before and after completion of the cardiovascular course. It is quite possible that other components of the preclinical curriculum contributed to this change; however, dietary fat consumption tends to increase and fruit and vegetable consumption decreases as students progress through the medical curriculum, especially the clinical years.10,13,14 A formal course in nutrition has been shown to be associated with an improvement in eating habits, measured by dietary questionnaire before and right after the course.26 It has been documented in a large culturally diverse group that the addition of the RYP to a feedback tip sheet that documented personal cholesterol level, goals, and recommendations led to an
203 additional 3- to 5-mg/dL reduction in mean cholesterol level at the end of 1 year than the feedback tip sheet alone, with the greater reduction seen in the 60% with elevated cholesterol levels.27 Although it cannot be concluded definitely that the teaching of nutrition within an integrated cardiovascular course that incorporated the RYP questionnaire was the cause of the modest improvement in RYP score, it probably contributed somewhat. A second finding, confirmatory of work by others,9,10,13 was that these medical students followed a healthy lifestyle and with some exceptions, were healthy. Nine students (28%) had a BMI . 25 and 4 were obese.10 As a group, weight increased 2.6 pounds. Eight students (29%) had a low HDL cholesterol for their sex, a finding similar to Clair et al9 who found a 42% prevalence of low HDL in a 6-year cross-sectional study of 662 first-year medical students. Two of the students with low HDL in the present study were obese; one of these students exercised ,15 minutes daily. All others had a normal BMI and exercised .30 minutes daily. None of the eight students had elevated triglycerides, LDL, or non-HDL cholesterol values.
Magnitude of improvement The improvement in the RYP score was small. However, it was highly significant, despite starting from a high baseline for heart healthy eating. Although the change in the mean score was only 1.78, it has to be evaluated in the context of what improvement was possible. Because student eating habits were quite good at baseline, the score could improve no more than 15 points; therefore, the increase of 1.78 represented nearly 12% of the maximum that could be achieved (from 57 to 72 points), a modest increase. It is difficult to know what accounted for the improvement. The Penn State curriculum in cardiovascular nutrition has always been heavily committed to using techniques of active adult learning, 1 of 2 fundamental principles that Krebs and Primak28 have stressed as fundamental to the University of Colorado Medical Nutrition and Education program. The 2 changes that were made to the curriculum for this study were to introduce the published RYP questionnaire for self-analysis of diet and to assess personal 30-year cardiovascular risk. Because knowledge of personal risk did not correlate with change in RYP score, we have to conclude that integrating nutritional education within the context of an organ system medical curriculum had some influence on the significant improvement seen in the eating habits.
Other studies on motivating change in eating habits A number of studies have been performed to assess the effect of personal knowledge of coronary calcium score to improve heart healthy behavior.29–32 Orakzai et al29 reported that a favorable change in diet was directly related to the magnitude of coronary calcium detected on electron beam computed tomography, determined by questionnaire
204 before and after a mean interval of 3 years. Schwartz et al30 found that 33% of subjects with coronary calcium questioned 6 years after scanning changed their diet. However, 21% without coronary calcium did so as well. The only other factor that was significantly correlated with changing diet was the presence of dyslipidemia. Forty-four percent of subjects with a lipid problem changed their diet, whereas 31% did not. The presence or absence of hypertension or a family history of premature heart disease did not influence dietary behavior.30 Wong et al31 found the presence of coronary calcium to correlate with decreasing intake of dietary fat 1 to 2 years later, but a family history of heart disease did not. Those populations were considerably older (mean ages, 60, 64, and 54 years, respectively) than the present study (24 years), and all were referred for coronary calcium screening. In a younger population (mean age, 42 years) of healthy active-duty military personnel undergoing routine health examinations who were offered coronary calcium scans as part of a study, O’Malley et al32 found that the knowledge of scan result was not associated with a change in 10-year Framingham risk score. However, intensive case management did result in an improvement. In a critique of the O’Malley et al32 study, Schwartz et al30 noted that there was a low prevalence of coronary calcification; however, more important for understanding the results of the present study, there was a low prevalence of modifiable risk factors.32 As in our study, the younger age and low absolute risk may be key factors to decrease motivation to change behavior. The population of the present study was quite healthy. Fifty-eight percent were exercising $30 minutes per day compared with 15% of average American adults33; 56% were classified by the RYP questionnaire in the highest quality dietary category: you are making many healthy choices. This makes the significant change in mean RYP score more striking. More importantly, and in agreement with O’Malley et al,32 knowledge of risk has little effect on changing risk if initial risk is low. Perhaps the PBL group discussion served the same purpose as the case management.32
Journal of Clinical Lipidology, Vol 8, No 2, April 2014 healthy. Minor changes were made to conform with additions to nutritional selections. More important changes were more precise definition and sampling of serving sizes. Nonetheless, it is still possible that students could have referred to the teaching version of RYP used in the course 6 months previously before or while filling out the follow-up questionnaire during phase 2 of the study. A separate validation study was not performed, and we have no information on how truthfully or accurately students answered questions. Both questionnaires were administered at the same time of year to minimize seasonal variations in diet. However, the entire second-year curriculum took place between the 2 questionnaires, and other components may have influenced the follow-up responses. Students were self-selected volunteers that may have been more health conscious and motivated to obtain a free fasting lipid panel and to receive personal information on risk; therefore, they may be more motivated to change diet. Except for the lipid panel, all of the information used for the calculation of risk was self-reported; this includes information about family medical history. However, these are medical students and were probably more attuned to family medical history than the average young adult. They self-reported current weight, height, waist circumference, and blood pressure. Nonetheless, they were given detailed instructions on how to measure waist circumference and blood pressure with the use of conventional standards. The last limitation is that the questionnaire was directed at sampling nutritional habits as they affected cardiovascular health and was not designed to sample overall nutritional choices. The questionnaire heavily samples fat and oil consumption; however, these have been stressed in most publications that advocate a heart healthy diet. Only 1 question was devoted to salt and one to sweets, pastries, and candy. There is no sampling of other sugar consumption or consumption of refined carbohydrates nor is there a sampling of alcohol consumption. These items, however, and all other learning objectives indicated as highest priority by the Nutrition Academic Award Program7 were considered in the cardiovascular lectures and PBL discussions.
Limitations of the study To assess risk, values for total cholesterol and HDL cholesterol were entered into the program.21 However, if these were not available we provided an alternative, BMI. The latter approach was used in the curriculum for teaching purposes. If the values for risk calculated during the course and risk presented to students with the use of lipid values months before differed, it may have diminished the motivational effect of personal risk knowledge. Another limitation of the study is that the RYP questionnaire was used as a teaching instrument, and its effect on diet was evaluated by a similar questionnaire. However, the 2 versions of the questionnaire were dissimilar in a number of significant ways that may have minimized the confounding influence. In the 2009 evaluation questionnaire, the first choice is most heart healthy; in the published 2005 version that the students used in the course, the first choice is least heart
Conclusions Students in their first year of medical school have heart healthy eating habits as assessed by the RYP questionnaire. When a variation to this questionnaire was added to the nutritional component that was integrated within the cardiovascular curriculum, this was associated with a small but significant improvement in eating habits measured a year later. The improvement in dietary choices was not related to knowledge of personal 30-year cardiovascular risk assessed by the Framingham heart algorithm. Therefore, it is likely that changes were related to the integration of cardiovascular nutrition within the cardiovascular curriculum. Because heart healthy dietary behavior in medical students and physicians is associated with a greater likelihood for using preventive strategies to counsel their patients, this suggests that faculty should not
Vargas and Zelis
Changing medical student eating habits
205
only teach nutrition within the curriculum but also should encourage students to practice healthy nutrition. Because healthy behavior and the importance of prevention decline as students’ progress through the clinical years, a future challenge will be to extend nutrition education longitudinally across the curriculum.34
14. Spencer EH, Frank E, Elon LK, Hertzberg VS, Serdula MK, Galuska DA. Predictors of nutrition counseling behaviors and attitudes in US medical students. Am J Clin Nutr. 2006;84:655–662. 15. Frank E, Carrera JS, Elon L, Hertzberg VS. Predictors of US medical students’ prevention counseling practices. Prev Med. 2007;44:76–81. 16. Frank E, Tong E, Lobelo F, Carrera J, Duperly J. Physical activity levels and counseling practices of U.S. medical students. Med Sci Sports Exerc. 2008;40:413–421. 17. Lobelo F, Duperly J, Frank E. Physical activity habits of doctors and medical students influence their counselling practices. Br J Sports Med. 2009;43:89–92. 18. Eaton CB, McBride PE, Gans KA, Underbakke GL. Teaching nutrition skills to primary care practitioners. J Nutr. 2003;133: 563S–566S. 19. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. N Engl J Med. 2001;344:3–10. 20. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA. 2001;285:2486–2497. 21. Pencina MJ, D’Agostino RB Sr., Larson MG, Massaro JM, Vasan RS. Predicting the 30-year risk of cardiovascular disease: the framingham heart study. Circulation. 2009;119:3078–3084. 22. Gans KM, Sundaram SG, McPhillips JB, Hixson ML, Linnan L, Carleton RA. Rate Your Plate: an eating pattern assessment and educational tool used at cholesterol screening and education programs. J Nutr Educ. 1993;25:29–36. 23. Gans KM, Hixson ML, Eaton CB, Lasater TM. Rate Your Plate: a dietary assessment and educational tool for blood cholesterol control. Nutr Clin Care. 2000;3:163–169. 24. Lloyd-Jones DM, Nam BH, D’Agostino RB Sr., et al. Parental cardiovascular disease as a risk factor for cardiovascular disease in middleaged adults: a prospective study of parents and offspring. JAMA. 2004; 291:2204–2211. 25. Stamler R, Stamler J, Riedlinger WF, Algera G, Roberts RH. Family (parental) history and prevalence of hypertension. Results of a nationwide screening program. JAMA. 1979;241:43–46. 26. Conroy MB, Delichatsios HK, Hafler JP, Rigotti NA. Impact of a preventive medicine and nutrition curriculum for medical students. Am J Prev Med. 2004;27:77–80. 27. Gans KM, Burkholder GJ, Risica PM, Harrow B, Lasater TM. Costeffectiveness of minimal contact education strategies for cholesterol change. Ethn Dis. 2006;16:443–451. 28. Krebs NF, Primak LE. Comprehensive integration of nutrition into medical training. Am J Clin Nutr. 2006;83:945S–950S. 29. Orakzai RH, Nasir K, Orakzai SH, et al. Effect of patient visualization of coronary calcium by electron beam computed tomography on changes in beneficial lifestyle behaviors. Am J Cardiol. 2008;101:999–1002. 30. Schwartz J, Allison M, Wright CM. Health behavior modification after electron beam computed tomography and physician consultation. J Behav Med. 2011;34:148–155. 31. Wong ND, Detrano RC, Diamond G, et al. Does coronary artery screening by electron beam computed tomography motivate potentially beneficial lifestyle behaviors? Am J Cardiol. 1996;78:1220–1223. 32. O’Malley PG, Feuerstein IM, Taylor AJ. Impact of electron beam tomography, with or without case management, on motivation, behavioral change, and cardiovascular risk profile: a randomized controlled trial. JAMA. 2003;289:2215–2223. 33. Klein RJ, Proctor SE, Boudreault MA, Turczyn KM. Healthy People 2010 criteria for data suppression. Healthy People 2010 Stat Notes. 2002;(24):1–12. 34. Tobin B, Welch K, Dent M, Smith C, Hooks B, Hash R. Longitudinal and horizontal integration of nutrition science into medical school curricula. J Nutr. 2003;133:567S–572S.
Acknowledgments We thank Penny Kris-Etherton, PhD, for advice about nutritional assessment and comments on review of the data, Robert Steckbeck for creating and maintaining the secure computerized database necessary to collect and analyze data while preserving participant anonymity, and Nakomis Maher for secretarial assistance. We also thank the Penn State students in the medical class of 2014 who participated in this study.
Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jacl.2013.11.006.
References 1. Adams KM, Kohlmeier M, Zeisel SH. Nutrition education in U.S. medical schools: latest update of a national survey. Acad Med. 2010;85:1537–1542. 2. Kushner RF. Denon Institute Award for Excellence in Medical/Dental Nutrition Education Lecture, 2002. Will there be a tipping point in medical nutrition education? Am J Clin Nutr. 2003;77:288–291. 3. Collier R. Canadian medical students want more nutrition instruction. CMAJ. 2009;181:133–134. 4. Frantz DJ, Munroe C, McClave SA, Martindale R. Current perception of nutrition education in U.S. medical schools. Curr Gastroenterol Rep. 2011;13:376–379. 5. Adams KM, Kohlmeier M, Powell M, Zeisel SH. Nutrition in medicine: nutrition education for medical students and residents. Nutr Clin Pract. 2010;25:471–480. 6. Nutrition in Medicine. 2013. Available at: http://www.nutritioninmedicine. net/. Accessed December 24, 2013. 7. Nutrition Academic Award Program Curriculum Committee. Nutrition Curriculum Guide for Training Physician Practice Behavior Skills and Attitudes Across the Curriculum. 2002. Available at: http://www.nhlbi. nih.gov/funding/training/naa/curr_gde.pdf. Accessed December 24, 2013. 8. Walker WA. Innovative teaching strategies for training physicians in clinical nutrition: an overview. J Nutr. 2003;133:541S–543S. 9. Clair JH, Wilson DB, Clore JN. Assessing the health of future physicians: an opportunity for preventive education. J Contin Educ Health Prof. 2004;24:82–89. 10. Frank E, Carrera JS, Elon L, Hertzberg VS. Basic demographics, health practices, and health status of U.S. medical students. Am J Prev Med. 2006;31:499–505. 11. Heimburger DC, Ullman DO, Ramsey MJ, et al. Dietary habits of firstyear medical students assessed during clinical nutrition course. Nutrition. 1994;10:214–220. 12. Wells KB, Lewis C, Leake B, Ware J Jr. Do physicians preach what they practice? A study of physicians’ health habits and counseling practices. JAMA. 1984;252:2846–2848. 13. Delnevo CD, Abatemarco DJ, Gotsch AR. Health behaviors and health promotion/disease prevention perceptions of medical students. Am J Prev Med. 1996;12:38–43.
Integrating nutrition education into the cardiovascular curriculum changes eating habits of second-year medical students Eric J. Vargas, BS, Robert Zelis, MD* College of Medicine, Penn State University, Hershey, PA, USA (Mr. Vargas); and Lipid Clinic, Heart and Vascular Institute, College of Medicine, Penn State University, 500 University Drive, H047, Hershey, PA 17033–0850, USA (Dr. Zelis) KEYWORDS: Nutrition education; Medical curriculum; Eating habits; Personal cardiovascular risk; Rate Your Plate questionnaire
BACKGROUND: Survey of medical curricula continues to show that nutrition education is not universally adequate. One measure of nutritional educational competence is a positive change in student eating habits. OBJECTIVE: The objective of this study was to evaluate whether integrating nutrition education within the second-year cardiovascular course for medical students, using the ‘‘Rate Your Plate’’ (RYP) questionnaire, coupled with knowledge of student personal 30-year risk of a cardiovascular event was useful in changing students’ eating behaviors. METHODS: Thirty-two students completed an unpublished 24-item questionnaire (modified-RYP) about their eating habits in the spring of their first year. The same students then completed the questionnaire in the spring of their second year. Paired t test was used to analyze the difference in RYP scores. Pearson correlation coefficients were calculated for the Framingham 30-year cardiovascular event risk and change in RYP score to examine whether risk knowledge may have changed eating habits. RESULTS: Mean scores at baseline and 1 year later were 57.19 and 58.97, respectively (paired t test, P , .01). Correlation coefficient between 30-year relative risk, adjusted for family history, and change in RYP score was –0.322. CONCLUSION: Although medical students were eating healthy at baseline, integration of nutrition education within the second-year cardiovascular medical curriculum was associated with improved heart healthy eating habits. Because student attitudes about prevention counseling are influenced by personal eating habits, this suggests that students with a more healthy diet will be more likely to recommend the same for their patients. Ó 2014 National Lipid Association. All rights reserved.
Despite the recommendation of the National Academy of Sciences in 1985, which stated a minimum of 25 hours of nutrition education be provided to medical students over the 4-year curriculum, Adams et al1 found a 12% decline between 2004 and 2009 and noted only 27% of the 105 schools met the minimum.1 In 2001, .14,000 US * Corresponding author. E-mail address: [email protected] Submitted July 2, 2013. Accepted for publication November 15, 2013.
medical students completed the Association of American Medical Colleges graduation questionnaire; 56% felt their nutrition-related experiences were inadequate.2 Furthermore, 66% felt their clinical decision-making and care skills about nutrition were poor. Similar experiences have been noted among Canadian medical students.3 A number of suggestions have been made to combat this trend, such as development of a competency-based certification program modeled after the American Heart Association program in Cardiac Life Support.4 Rather than draft
1933-2874/$ - see front matter Ó 2014 National Lipid Association. All rights reserved. http://dx.doi.org/10.1016/j.jacl.2013.11.006
200 new curricula, wider distribution of already existing curricula has been advocated,4 such as teaching modules developed by the Nutrition in Medicine Project,5 which conform to the consensus guidelines of the American Society of Clinical Nutrition.6 Another curriculum guide was developed by the National Institutes of Health-supported Nutrition Academic Award Program.7,8 It is with this background that our institution appointed a multidisciplinary task force to determine how and when nutrition was taught in our curriculum and to make recommendations to ensure that nutrition education was adequately incorporated. The report, presented to faculty in 2006, was approved by the curriculum committee and distributed to the course committees for implementation. Many of the learning objectives of this task force mirrored those of the Nutrition Academic Award,7 especially objectives related to the cardiovascular system and were incorporated into the nutrition component of the organ system-based integrated second-year cardiovascular block. One nutritional objective not addressed in most guidelines is to promote changes in student eating habits toward more heart healthy behavior. Multiple studies have shown that first-year medical students tend to consume a lower fat diet with more fruits and vegetables and to exercise moderately.9–11 However, those eating habits begin to decline during the clinical years.10 This is important because medical students ‘‘preach what they practice.’’12 Students and physicians who have a healthy lifestyle are more likely to see the importance of diet and exercise in disease prevention and to counsel their patients accordingly.12–17 Unfortunately, the importance of prevention counseling declines over the 4 years of medical school.13,14,16 The purpose of this study was to determine whether integrating nutrition within the second-year cardiovascular course, plus 2 small changes to the curriculum, would result in more heart healthy personal dietary choices by medical students. The first was to introduce the Rate Your Plate (RYP) dietary questionnaire into the curriculum as a means for students to assess personal eating habits. This is one of the counseling tools presented in one of the papers in the Nutrition Academic Award symposium on Innovative Teaching Strategies.18 The second was to provide a free fasting lipid panel for study participants to calculate their personal 30-year risk of developing a cardiovascular event, as a means to stimulate interest to improve personal risk factors. The 2 hypotheses of our study were (1) integrating nutrition education while adding the RYP dietary assessment to the curriculum would improve heart healthy eating habits and (2) students with the higher perceived cardiovascular risk would improve their eating habits more.
Methods Study rationale The presentation of nutrition education within the cardiovascular course used a variety of components. There
Journal of Clinical Lipidology, Vol 8, No 2, April 2014 were 2 hours of lecture on lipoprotein transport and a third hour on practical nutritional advice, and 2 separate ProblemBased Learning (PBL) cases that had dietary elements. During the hypertension PBL case, students discussed the Dietary Approach to Stop Hypertension diet,19 and the coronary artery disease PBL case had a learning objective to use the Therapeutic Lifestyle Change diet recommended by the Third Adult Treatment Panel of the National Cholesterol Education Program.20 There was also a separate nutritional exercise. Students in the PBL groups of 8 students were required to analyze the eating habits of 2 patients, one with heterozygous familial hypercholesterolemia and the other with metabolic syndrome. Half the students were assigned to one or the other problem and were required to take specific dietary components, visit grocery stores and fast food and upscale restaurants, and determine what was ‘‘good’’ and ‘‘poor’’ about their patient’s dietary selections. They were then asked to recommend better choices to achieve lipid goals that were based on individual patient Framingham risk assessment. They returned approximately 2 weeks later for discussion of their findings and recommendations within the full PBL group. The learning objectives for these activities covered most objectives present in the Nutrition Curriculum Guide for Training Physicians,7 under the cardiovascular system (section D) and all those ranked in the top one-third of importance. An additional assignment at the same time was a third ‘‘patient’’, themselves. Most did little on this component for 2 reasons. Ten-year cardiovascular risk evaluation had little meaning for students in their mid-20s. This problem was resolved when in 2009 the Framingham Heart Study published a simple algorithm to quantify 30-year risk.21 The second problem was that students found quantifying their own 24-hour food consumption formally in terms of nutritional composition was too tedious and did not seem practical. After reviewing various dietary self-assessment tools, it seemed that the RYP questionnaire might be particularly suitable for our purposes. It was easy to complete, used language familiar to students, and dealt specifically with dietary components likely to affect cardiovascular health.18,22,23 RYP is a semiquantitative food-frequency questionnaire that provides a single numerical ranking of the quality of food choices and a ranking into 1 of 3 qualitative groups. It also provides positive reinforcement for healthy choices and specific recommendations on how less healthy choices can be improved, thus making it a useful tool for patients. RYP has been directly compared with the Willett food-frequency questionnaire and found favorable.22 The validation was stronger when RYP was administered before the Willett questionnaire, than when given after. The researchers suggested that the much longer 136-question Willett may have led to a more haphazard filling out of the brief RYP. Moreover, as noted above, RYP has been advocated as a strategy to train primary care physicians in nutritional skills.18 However, we were concerned that using identical questionnaires 3 times to train students and to assess student eating habits before and after training might confound the
Vargas and Zelis
Changing medical student eating habits
201
results. The RYP used in the second-year course could have been reviewed by some students before taking the follow-up RYP after training and might have influenced the responses. This problem was avoided by using an unpublished updated RYP for assessment. The unpublished version of the RYP questionnaire was dissimilar in a number of ways, including the reversal of the order of response. This allowed us to use a different instrument for before and after testing while using the earlier published version of RYP for teaching. The study design entailed testing a volunteer group of first-year medical students in the spring before the fall cardiovascular course and retesting the same students the next spring, 6 months after the cardiovascular course ended.
coded identification number and recorded in the database in a manner to preserve participant confidentiality. The following 2 study end points were preselected for statistical analysis: (1) the initial and follow-up RYP scores were compared by 2-sided paired t test and (2) the change in RYP score was compared with the individual relative risk by using the Pearson correlation coefficient after adjusting relative risk for a family history of premature vascular disease. Data are presented as mean 6 SD. P values , .05 were considered significant. The changes in score for each of the 24 RYP items were scanned to determine which items contributed (610%) to a change in the mean score but were not analyzed for significance. The 2005 and 2009 versions of RYP were supplied by the author of the questionnaire22,23 and used with permission of the copyright holder (Institute for Community Health Promotion, Brown University, Providence, RI). The 2005 version was used by the entire class for education and was available in print22 and online at their website at the time of the study. The 2009 version was used as the evaluation instrument and replaced the 2005 version online after the study was completed and now appears with a 2010 copyright.
Research protocol The research project was reviewed and approved by the institutional review board (protocol 36289). First-year medical students were recruited from the class of 2014 by an e-mail invitation sent to the entire class of 140 students. Interested students were asked to attend a meeting in which the project was discussed. A complete written description of the protocol was given to each student. Participation was voluntary and self-selected. Students who wished to participate were given a numerical identification code by a student co-investigator to preserve anonymity. This was used to logon to a secure website to answer the RYP questionnaire and demographic questions. The students were told which of the latter specific questions would be asked and were instructed how to secure the answers before logon. When all information had been recorded, students were issued a voucher for a fasting lipid panel, the results of which were made available to them subsequently at the secure site. At a second login, students were presented with the results of the lipid panel and their personal relative and absolute 30-year risk. They were given a 1-page explanation to aid in the interpretation of the lipid results, 30-year risk, and interpretation of their blood pressure, weight, body mass index (BMI; calculated as weight divided by height square; kg/m2), and activity level. It was explained that a family history of a premature vascular event in a first-degree relative doubles their risk; a similar history in a grandparent increases risk by approximately 50%.24 They were also told that a positive family history of hypertension or diabetes doubles their chance of developing that problem.25 We listed the major components of the metabolic system, and its risk for diabetes was noted. Finally, they were given a list of suggested dietary changes to mitigate risk. They were given the simple dietary advice to lower low-density lipoprotein (LDL) cholesterol by eliminating trans fats and reducing saturated fats, while substituting unsaturated fat and using a plant stanol/sterol-ester margarine. Blood samples were drawn by the clinical laboratory of the Hershey Medical Center whereby serum triglyceride, total, and high-density lipoprotein (HDL) cholesterol levels were determined, and LDL and non-HDL cholesterol values were calculated. All reports used the
Results Characteristics of study population Of the 44 first-year medical students who completed the questionnaire, 32 (73%) completed the follow-up study 1 year later, approximately one-quarter of the class. The 32 students in the study group had the following characteristics: equal sex distribution; age, 23.9 6 4.1 years; weight, 153.5 6 32.8 pounds; waist circumference, 31.1 6 3.5 inches; BMI, 23.9 6 4.1; blood pressure (systolic/diastolic), 116.5 6 9.2/73.5 6 6.1 mm Hg. With the use of BMI criteria, 4 men and 1 woman were overweight; and 1 man and 3 women were obese. Five women were using hormones. No student was taking antihypertensive or lipid-modifying medication. None had diabetes; none smoked. Only 2 students had a first-degree relative with a premature coronary event; 10 students had a grandparent with a premature coronary event. Nineteen students had relatives with a history of type 2 diabetes and 22 had relatives with hypertension. The only personal data that changed over the course of the year was a 2.6-pound (1.7%) weight increase to 156.1 6 35.2 pounds, which was close to being significant (P 5 .052). Of the 31 students with complete information about their daily exercise, at baseline 19 exercised $30 minutes, 7 exercised 15 to 30 minutes, and 5 exercised ,15 minutes. The only insignificant changes at the time of the second questionnaire were 2 fewer students in the lower group and 2 more in the middle group. Twenty-eight students had lipid information drawn at the beginning of the study. The mean values were total cholesterol of 165.6 6 20.0 mg/dL, HDL cholesterol of 50.8 6 10.8 mg/dL, triglycerides of 72.1 6 23.8 mg/dL, LDL
202
Journal of Clinical Lipidology, Vol 8, No 2, April 2014
Table 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Responses to the individual items that comprise the RYP questionnaire at 2 time points Individual RYP dietary questions
RYP 1
SD 1
RYP 2
SD 2
Change
Change. %
Meat cuts Chicken and turkey Ground meat and poultry Processed meat and poultry Portion size of meat and poultry Fish, shellfish Cooking method (meat, poultry, fish) Meatless meals Whole eggs Milk (yogurt, cream), % fat Cheese (included in meals), fat type Servings of dairy foods Whole grains Fruits and vegetables Cooking method (vegetables, pasta) Fat type in cooking/baking Salt from processed foods Table spreads Salad dressings, mayonnaise Snack foods Nuts, seeds Frozen desserts Sweets, pastries, candy Eating out/take out
2.63 2.81 2.56 2.53 2.06 1.63 2.72 2.47 2.63 2.69 2.00 2.50 2.28 2.09 2.31 2.91 2.06 2.34 2.41 2.59 2.25 2.25 2.09 2.38
0.61 0.59 0.56 0.51 0.72 0.75 0.52 0.76 0.71 0.64 0.80 0.62 0.68 0.59 0.54 0.39 0.62 0.90 0.76 0.61 0.80 0.72 0.69 0.75
2.78 2.69 2.66 2.53 2.19 1.63 2.81 2.50 2.63 2.69 2.13 2.41 2.34 2.41 2.53 2.97 2.25 2.75 2.31 2.63 2.13 2.53 2.22 2.28
0.42 0.54 0.55 0.51 0.74 0.79 0.40 0.76 0.61 0.69 0.79 0.71 0.70 0.56 0.57 0.18 0.72 0.62 0.78 0.49 0.79 0.62 0.61 0.73
0.16 –0.13 0.09 0.00 0.13 0.00 0.09 0.03 0.00 0.00 0.13 –0.09 0.06 0.31 0.22 0.06 0.19 0.41 –0.09 0.03 –0.13 0.28 0.13 –0.09
5.95 –4.44 3.66 0.00 6.06 0.00 3.45 1.27 0.00 0.00 6.25 –3.75 2.74 14.93 9.46 2.15 9.09 17.33 –3.90 1.20 –5.56 12.50 5.97 –3.95
RYP 1, baseline score before intervention; RYP 2, score 1 year later and 6 months after intervention. A value of 1.0 is most heart unhealthy and 3.0 is most heart healthy. Results are mean and SD.
cholesterol of 100.4 6 19.5 md/dL, and non-HDL cholesterol of 114.8 6 19.8 mg/dL. Three men had HDL , 40 mg/dL, and 5 women has HDL , 50 mg/dL; none had a triglyceride value . 150 mg/dL, one had LDL cholesterol . 130 mg/dL, and one had a non-HDL cholesterol . 160 mg/dL. With the use of the lipid data and demographic information, the calculated Framingham 30-year risk for a cardiovascular event was 3.67% 6 1.33% (range, 1.4%–7.0%). Average risk for comparable age and sex was 3.06% 6 0.80%, and optimal was 2.53% 6 0.67%. Relative risk for the group was 1.22 6 0.44. When family history was included, relative risk was 1.29 6 0.72.
RYP questionnaire responses The RYP evaluation questionnaire consisted of 24 items. Three responses were presented for each item; each response was preceded by a verb–adverb combination such as ‘‘usually eat,’’ ‘‘sometimes eat,’’ ‘‘rarely eat,’’ ‘‘usually use,’’ and so forth. The responses went from an eating pattern of most heart unhealthy to most heart healthy and corresponded to values of 1, 2, or 3, with 3 representing the most heart healthy pattern. When the numbers for all 24 responses were added, a single score was derived from 24 to 72. For the evaluation instrument, the 3 qualitative assessment categories were as follows: 58 to 72, ‘‘you are
making many healthy choices’’; 41 to 57, ‘‘there are some ways you can make your eating habits healthier’’; 24 and 40 ‘‘there are many ways you can make your eating habits healthier.’’ The mean baseline RYP score was 57.19 6 5.65. Eighteen students were in the most heart healthy eating category, 14 in the middle, and none in the lowest. Thus, at baseline, these first-year medical students appeared to be eating quite well with respect to heart healthy choices. One year later, 6 months after the cardiovascular course ended, a significant improvement was found in the RYP score to 58.97 6 4.97, a change that was significant by paired t analysis (P 5 .008). Table 1 displays the values for each of the 24 items about which inquiry was made, at baseline and 1 year later. Absolute and percentage of changes are indicated. The lowest possible response is 1.0, the highest 3.0. These data are presented primarily to form an impression what might have contributed to the significant change in mean score. Statistical analyses were not performed because of the severe P value correction penalty for 24 comparisons. For most categories the change was positive, and 3 showed improvements of .10%: fruits and vegetables daily frequency improved 15%; table spreads qualitative type improved 17%; and frozen desserts qualitative type and frequency improved 12.5%.
Vargas and Zelis
Changing medical student eating habits
Figure 1 Change in RYP dietary score related to 30-year Framingham relative risk of a cardiovascular event adding the effect of a positive family of premature vascular event (doubling for parent, 50% increase for grandparent). Premature is ,55 years of age for a man and ,65 years of age for a woman. RYP, Rate Your Plate.
Change in RYP score related to personal cardiovascular risk Because 4 students did not have a lipid panel drawn, data from 28 students were available for analysis. Figure 1 shows individual data points that relate the adjusted relative 30-year risk of a cardiovascular event to change in RYP score. The relative risk was increased by 1.0 for students with a parent and 0.5 for a grandparent with a premature vascular event. The change in RYP score did not correlate with personal risk (Pearson’s correlation coefficient, –0.322). The change in score was also not significantly correlated to unadjusted risk (correlation coefficient, –0.223).
Discussion The main finding of this study was that dietary eating habits of medical students as measured by the RYP dietary questionnaire modestly improved over the 1-year period that spanned their second year. We feel that nutrition education within an integrated cardiovascular curriculum contributed to this change in the group of students who volunteered to be evaluated before and after completion of the cardiovascular course. It is quite possible that other components of the preclinical curriculum contributed to this change; however, dietary fat consumption tends to increase and fruit and vegetable consumption decreases as students progress through the medical curriculum, especially the clinical years.10,13,14 A formal course in nutrition has been shown to be associated with an improvement in eating habits, measured by dietary questionnaire before and right after the course.26 It has been documented in a large culturally diverse group that the addition of the RYP to a feedback tip sheet that documented personal cholesterol level, goals, and recommendations led to an
203 additional 3- to 5-mg/dL reduction in mean cholesterol level at the end of 1 year than the feedback tip sheet alone, with the greater reduction seen in the 60% with elevated cholesterol levels.27 Although it cannot be concluded definitely that the teaching of nutrition within an integrated cardiovascular course that incorporated the RYP questionnaire was the cause of the modest improvement in RYP score, it probably contributed somewhat. A second finding, confirmatory of work by others,9,10,13 was that these medical students followed a healthy lifestyle and with some exceptions, were healthy. Nine students (28%) had a BMI . 25 and 4 were obese.10 As a group, weight increased 2.6 pounds. Eight students (29%) had a low HDL cholesterol for their sex, a finding similar to Clair et al9 who found a 42% prevalence of low HDL in a 6-year cross-sectional study of 662 first-year medical students. Two of the students with low HDL in the present study were obese; one of these students exercised ,15 minutes daily. All others had a normal BMI and exercised .30 minutes daily. None of the eight students had elevated triglycerides, LDL, or non-HDL cholesterol values.
Magnitude of improvement The improvement in the RYP score was small. However, it was highly significant, despite starting from a high baseline for heart healthy eating. Although the change in the mean score was only 1.78, it has to be evaluated in the context of what improvement was possible. Because student eating habits were quite good at baseline, the score could improve no more than 15 points; therefore, the increase of 1.78 represented nearly 12% of the maximum that could be achieved (from 57 to 72 points), a modest increase. It is difficult to know what accounted for the improvement. The Penn State curriculum in cardiovascular nutrition has always been heavily committed to using techniques of active adult learning, 1 of 2 fundamental principles that Krebs and Primak28 have stressed as fundamental to the University of Colorado Medical Nutrition and Education program. The 2 changes that were made to the curriculum for this study were to introduce the published RYP questionnaire for self-analysis of diet and to assess personal 30-year cardiovascular risk. Because knowledge of personal risk did not correlate with change in RYP score, we have to conclude that integrating nutritional education within the context of an organ system medical curriculum had some influence on the significant improvement seen in the eating habits.
Other studies on motivating change in eating habits A number of studies have been performed to assess the effect of personal knowledge of coronary calcium score to improve heart healthy behavior.29–32 Orakzai et al29 reported that a favorable change in diet was directly related to the magnitude of coronary calcium detected on electron beam computed tomography, determined by questionnaire
204 before and after a mean interval of 3 years. Schwartz et al30 found that 33% of subjects with coronary calcium questioned 6 years after scanning changed their diet. However, 21% without coronary calcium did so as well. The only other factor that was significantly correlated with changing diet was the presence of dyslipidemia. Forty-four percent of subjects with a lipid problem changed their diet, whereas 31% did not. The presence or absence of hypertension or a family history of premature heart disease did not influence dietary behavior.30 Wong et al31 found the presence of coronary calcium to correlate with decreasing intake of dietary fat 1 to 2 years later, but a family history of heart disease did not. Those populations were considerably older (mean ages, 60, 64, and 54 years, respectively) than the present study (24 years), and all were referred for coronary calcium screening. In a younger population (mean age, 42 years) of healthy active-duty military personnel undergoing routine health examinations who were offered coronary calcium scans as part of a study, O’Malley et al32 found that the knowledge of scan result was not associated with a change in 10-year Framingham risk score. However, intensive case management did result in an improvement. In a critique of the O’Malley et al32 study, Schwartz et al30 noted that there was a low prevalence of coronary calcification; however, more important for understanding the results of the present study, there was a low prevalence of modifiable risk factors.32 As in our study, the younger age and low absolute risk may be key factors to decrease motivation to change behavior. The population of the present study was quite healthy. Fifty-eight percent were exercising $30 minutes per day compared with 15% of average American adults33; 56% were classified by the RYP questionnaire in the highest quality dietary category: you are making many healthy choices. This makes the significant change in mean RYP score more striking. More importantly, and in agreement with O’Malley et al,32 knowledge of risk has little effect on changing risk if initial risk is low. Perhaps the PBL group discussion served the same purpose as the case management.32
Journal of Clinical Lipidology, Vol 8, No 2, April 2014 healthy. Minor changes were made to conform with additions to nutritional selections. More important changes were more precise definition and sampling of serving sizes. Nonetheless, it is still possible that students could have referred to the teaching version of RYP used in the course 6 months previously before or while filling out the follow-up questionnaire during phase 2 of the study. A separate validation study was not performed, and we have no information on how truthfully or accurately students answered questions. Both questionnaires were administered at the same time of year to minimize seasonal variations in diet. However, the entire second-year curriculum took place between the 2 questionnaires, and other components may have influenced the follow-up responses. Students were self-selected volunteers that may have been more health conscious and motivated to obtain a free fasting lipid panel and to receive personal information on risk; therefore, they may be more motivated to change diet. Except for the lipid panel, all of the information used for the calculation of risk was self-reported; this includes information about family medical history. However, these are medical students and were probably more attuned to family medical history than the average young adult. They self-reported current weight, height, waist circumference, and blood pressure. Nonetheless, they were given detailed instructions on how to measure waist circumference and blood pressure with the use of conventional standards. The last limitation is that the questionnaire was directed at sampling nutritional habits as they affected cardiovascular health and was not designed to sample overall nutritional choices. The questionnaire heavily samples fat and oil consumption; however, these have been stressed in most publications that advocate a heart healthy diet. Only 1 question was devoted to salt and one to sweets, pastries, and candy. There is no sampling of other sugar consumption or consumption of refined carbohydrates nor is there a sampling of alcohol consumption. These items, however, and all other learning objectives indicated as highest priority by the Nutrition Academic Award Program7 were considered in the cardiovascular lectures and PBL discussions.
Limitations of the study To assess risk, values for total cholesterol and HDL cholesterol were entered into the program.21 However, if these were not available we provided an alternative, BMI. The latter approach was used in the curriculum for teaching purposes. If the values for risk calculated during the course and risk presented to students with the use of lipid values months before differed, it may have diminished the motivational effect of personal risk knowledge. Another limitation of the study is that the RYP questionnaire was used as a teaching instrument, and its effect on diet was evaluated by a similar questionnaire. However, the 2 versions of the questionnaire were dissimilar in a number of significant ways that may have minimized the confounding influence. In the 2009 evaluation questionnaire, the first choice is most heart healthy; in the published 2005 version that the students used in the course, the first choice is least heart
Conclusions Students in their first year of medical school have heart healthy eating habits as assessed by the RYP questionnaire. When a variation to this questionnaire was added to the nutritional component that was integrated within the cardiovascular curriculum, this was associated with a small but significant improvement in eating habits measured a year later. The improvement in dietary choices was not related to knowledge of personal 30-year cardiovascular risk assessed by the Framingham heart algorithm. Therefore, it is likely that changes were related to the integration of cardiovascular nutrition within the cardiovascular curriculum. Because heart healthy dietary behavior in medical students and physicians is associated with a greater likelihood for using preventive strategies to counsel their patients, this suggests that faculty should not
Vargas and Zelis
Changing medical student eating habits
205
only teach nutrition within the curriculum but also should encourage students to practice healthy nutrition. Because healthy behavior and the importance of prevention decline as students’ progress through the clinical years, a future challenge will be to extend nutrition education longitudinally across the curriculum.34
14. Spencer EH, Frank E, Elon LK, Hertzberg VS, Serdula MK, Galuska DA. Predictors of nutrition counseling behaviors and attitudes in US medical students. Am J Clin Nutr. 2006;84:655–662. 15. Frank E, Carrera JS, Elon L, Hertzberg VS. Predictors of US medical students’ prevention counseling practices. Prev Med. 2007;44:76–81. 16. Frank E, Tong E, Lobelo F, Carrera J, Duperly J. Physical activity levels and counseling practices of U.S. medical students. Med Sci Sports Exerc. 2008;40:413–421. 17. Lobelo F, Duperly J, Frank E. Physical activity habits of doctors and medical students influence their counselling practices. Br J Sports Med. 2009;43:89–92. 18. Eaton CB, McBride PE, Gans KA, Underbakke GL. Teaching nutrition skills to primary care practitioners. J Nutr. 2003;133: 563S–566S. 19. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. N Engl J Med. 2001;344:3–10. 20. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA. 2001;285:2486–2497. 21. Pencina MJ, D’Agostino RB Sr., Larson MG, Massaro JM, Vasan RS. Predicting the 30-year risk of cardiovascular disease: the framingham heart study. Circulation. 2009;119:3078–3084. 22. Gans KM, Sundaram SG, McPhillips JB, Hixson ML, Linnan L, Carleton RA. Rate Your Plate: an eating pattern assessment and educational tool used at cholesterol screening and education programs. J Nutr Educ. 1993;25:29–36. 23. Gans KM, Hixson ML, Eaton CB, Lasater TM. Rate Your Plate: a dietary assessment and educational tool for blood cholesterol control. Nutr Clin Care. 2000;3:163–169. 24. Lloyd-Jones DM, Nam BH, D’Agostino RB Sr., et al. Parental cardiovascular disease as a risk factor for cardiovascular disease in middleaged adults: a prospective study of parents and offspring. JAMA. 2004; 291:2204–2211. 25. Stamler R, Stamler J, Riedlinger WF, Algera G, Roberts RH. Family (parental) history and prevalence of hypertension. Results of a nationwide screening program. JAMA. 1979;241:43–46. 26. Conroy MB, Delichatsios HK, Hafler JP, Rigotti NA. Impact of a preventive medicine and nutrition curriculum for medical students. Am J Prev Med. 2004;27:77–80. 27. Gans KM, Burkholder GJ, Risica PM, Harrow B, Lasater TM. Costeffectiveness of minimal contact education strategies for cholesterol change. Ethn Dis. 2006;16:443–451. 28. Krebs NF, Primak LE. Comprehensive integration of nutrition into medical training. Am J Clin Nutr. 2006;83:945S–950S. 29. Orakzai RH, Nasir K, Orakzai SH, et al. Effect of patient visualization of coronary calcium by electron beam computed tomography on changes in beneficial lifestyle behaviors. Am J Cardiol. 2008;101:999–1002. 30. Schwartz J, Allison M, Wright CM. Health behavior modification after electron beam computed tomography and physician consultation. J Behav Med. 2011;34:148–155. 31. Wong ND, Detrano RC, Diamond G, et al. Does coronary artery screening by electron beam computed tomography motivate potentially beneficial lifestyle behaviors? Am J Cardiol. 1996;78:1220–1223. 32. O’Malley PG, Feuerstein IM, Taylor AJ. Impact of electron beam tomography, with or without case management, on motivation, behavioral change, and cardiovascular risk profile: a randomized controlled trial. JAMA. 2003;289:2215–2223. 33. Klein RJ, Proctor SE, Boudreault MA, Turczyn KM. Healthy People 2010 criteria for data suppression. Healthy People 2010 Stat Notes. 2002;(24):1–12. 34. Tobin B, Welch K, Dent M, Smith C, Hooks B, Hash R. Longitudinal and horizontal integration of nutrition science into medical school curricula. J Nutr. 2003;133:567S–572S.
Acknowledgments We thank Penny Kris-Etherton, PhD, for advice about nutritional assessment and comments on review of the data, Robert Steckbeck for creating and maintaining the secure computerized database necessary to collect and analyze data while preserving participant anonymity, and Nakomis Maher for secretarial assistance. We also thank the Penn State students in the medical class of 2014 who participated in this study.
Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jacl.2013.11.006.
References 1. Adams KM, Kohlmeier M, Zeisel SH. Nutrition education in U.S. medical schools: latest update of a national survey. Acad Med. 2010;85:1537–1542. 2. Kushner RF. Denon Institute Award for Excellence in Medical/Dental Nutrition Education Lecture, 2002. Will there be a tipping point in medical nutrition education? Am J Clin Nutr. 2003;77:288–291. 3. Collier R. Canadian medical students want more nutrition instruction. CMAJ. 2009;181:133–134. 4. Frantz DJ, Munroe C, McClave SA, Martindale R. Current perception of nutrition education in U.S. medical schools. Curr Gastroenterol Rep. 2011;13:376–379. 5. Adams KM, Kohlmeier M, Powell M, Zeisel SH. Nutrition in medicine: nutrition education for medical students and residents. Nutr Clin Pract. 2010;25:471–480. 6. Nutrition in Medicine. 2013. Available at: http://www.nutritioninmedicine. net/. Accessed December 24, 2013. 7. Nutrition Academic Award Program Curriculum Committee. Nutrition Curriculum Guide for Training Physician Practice Behavior Skills and Attitudes Across the Curriculum. 2002. Available at: http://www.nhlbi. nih.gov/funding/training/naa/curr_gde.pdf. Accessed December 24, 2013. 8. Walker WA. Innovative teaching strategies for training physicians in clinical nutrition: an overview. J Nutr. 2003;133:541S–543S. 9. Clair JH, Wilson DB, Clore JN. Assessing the health of future physicians: an opportunity for preventive education. J Contin Educ Health Prof. 2004;24:82–89. 10. Frank E, Carrera JS, Elon L, Hertzberg VS. Basic demographics, health practices, and health status of U.S. medical students. Am J Prev Med. 2006;31:499–505. 11. Heimburger DC, Ullman DO, Ramsey MJ, et al. Dietary habits of firstyear medical students assessed during clinical nutrition course. Nutrition. 1994;10:214–220. 12. Wells KB, Lewis C, Leake B, Ware J Jr. Do physicians preach what they practice? A study of physicians’ health habits and counseling practices. JAMA. 1984;252:2846–2848. 13. Delnevo CD, Abatemarco DJ, Gotsch AR. Health behaviors and health promotion/disease prevention perceptions of medical students. Am J Prev Med. 1996;12:38–43.
