Predictors of nutrition counseling behaviors and attitudes in US ...

Predictors of nutrition counseling behaviors and attitudes in US
medical students1–4
Elsa H Spencer, Erica Frank, Lisa K Elon, Vicki S Hertzberg, Mary K Serdula, and Deborah A Galuska
ABSTRACT
care physicians include nutrition or dietary counseling in their
Background: Nutrition counseling by physicians can improve pa-
patient visits (6 –11).
tients’ dietary behaviors and is affected by physicians’ nutrition
Although many factors affect counseling rates, one of the least
practices and attitudes, such as the perceived relevance of nutrition
explored factors is the observation that physicians’ healthy per-
counseling.
sonal practices are positively associated with their clinical
Objective: The objective was to provide data on medical students’
prevention-related practices (9, 12–15). Specifically, physi-
perceived relevance of nutrition counseling, reported frequency of
cians’ healthy dietary practices positively correlate with their
nutrition counseling, and frequency of fruit and vegetable intakes.
clinical nutrition counseling attitudes (16) and practices (5, 10,
Downloaded from
Design: Students (n
2316) at 16 US medical schools were sur-
14, 17). Some small studies have shown that training interven-
veyed and tracked at freshmen orientation, at the time of orientation
tions may improve both medical students’ personal dietary be-
to wards, and in their senior year.
haviors (18) and their prevention counseling attitudes (18 –20).
Results: Freshmen students were more likely (72%) to find nutrition
Building on this preliminary association, we implemented the
counseling highly relevant than were students at the time of ward
“Healthy Doc-Healthy Patient” (HD-HP) study to describe not
orientation (61%) or during their senior year (46%; P for trend
only medical students’ attitudes and behaviors regarding per-
www.ajcn.org
0.0003). Those intending to subspecialize had lower and declining
sonal and clinical prevention but also the relation between their
perceptions of counseling relevance (P for trend
0.0009), whereas
the perceived relevance of counseling by primary care specialists
personal and clinical practices (21).
remained high (P for trend
0.5). Students were significantly more
Our primary objective was to describe the characteristics of
likely to find nutrition counseling highly relevant if they were fe-
medical students associated with more frequent nutrition coun-
by on October 17, 2010
male, consumed more fruit and vegetables, believed in primary pre-
seling and a higher perceived relevance of such counseling. The
vention, had personal physicians who encouraged disease preven-
secondary objective was to describe temporal trends in medical
tion, or intended to specialize in primary care. Only 19% of students
students’ perceived relevance of, reported confidence in, and
believed that they had been extensively trained in nutrition counsel-
reported training in nutrition counseling. To achieve our primary
ing, and 17% of seniors reported that they frequently counseled their
objective, we assessed the relation between counseling fre-
patients about nutrition. Students who consumed more fruit and
quency and its predictors in the students’ senior year and exam-
vegetables, believed that they would be more credible if they ate a
ined how the relations between the perceived relevance of nutri-
healthy diet, were not Asian or white, or intended to specialize in
tion counseling and its predictors change during the students’
primary care counseled patients about nutrition more frequently.
training. Novel predictors of interest include whether the health-
Medical students consumed an average of 3.0 fruit and vegetable
promotion efforts (health-promotion score) of medical schools
servings/d, which declined over time.
and the fruit and vegetable intakes of medical students are asso-
Conclusions: The perceived relevance of nutrition counseling by
ciated with the nutrition counseling attitudes and behaviors of the
US medical students declined throughout medical school, and stu-
students. Fruit and vegetable servings per day are the principal
dents infrequently counseled their patients about nutrition. Interven-
dietary component of interest because of their importance in
tions may be warranted to improve the professional nutritional prac-
public health recommendations (22, 23).
tices of medical students.
Am J Clin Nutr 2006;84:655– 62.
1 From the School of Medicine (EHS and EF) and the School of Public
KEY WORDS
Medical students, diet, fruit and vegetable in-
Health, Biostatistics (LKE and VSH), Emory University, Atlanta, GA, and
takes, nutrition counseling, counseling correlates
the Centers for Disease Control and Prevention, Division of Nutrition and
Physical Activity, Atlanta, GA (DAG and MKS).
2 The findings and conclusions in this report are those of the authors and do
INTRODUCTION
not necessarily represent the views of the CDC.
3
Although Americans suffer from increased rates of obesity,
Supported by the American Cancer Society.
4 Address reprint requests to E Frank, Department of Health and Epide-
hyperlipidemia, and diabetes (1–3), some studies have shown
miology, University of British Columbia, 5804 Fairview Avenue, Vancou-
that if physicians advise their patients about nutrition, the inci-
ver, BC, Canada V6T 1Z4. E-mail: efrank@emory.edu.
dence of these diseases will decline (4, 5). Despite the potential
Received October 29, 2005.
for counseling to improve dietary practices, 50% of primary
Accepted for publication May 2, 2006.
Am J Clin Nutr 2006;84:655– 62. Printed in USA. © 2006 American Society for Nutrition
655

---

656
SPENCER ET AL
METHODS
prevention is the best way to eradicate premature cardiovascular
disease (CVD),” “Physicians have a responsibility to promote
Study design
prevention with their patients,” “Patients will adopt a healthier
All medical students in the class of 2003 at 17 participating
lifestyle if counseled to do so,” “I will be able to provide more
schools were eligible for participation at each of 3 HD-HP ques-
credible and effective counseling if I eat a healthy diet,” and
tionnaire sessions: at freshman orientation, at the time of orien-
“How much emphasis has your personal physician placed on
tation to wards, and in the senior year. Our sample was designed
preventing disease?”. The counseling relevance model addition-
to be similar to all US medical schools in terms of age, region,
ally contained a time variable (freshman orientation, ward ori-
school size, National Institutes of Health research ranking, pri-
entation, or senior year).
vate–public school balance, underrepresented minorities, and
Queries on predictors included 8 fruit and vegetable items
sex (24 –27). Our study received approval from the Institutional
(French fries, other potatoes, fruit juice, fruit, vegetable juice,
Review Board at Emory University.
green salad, and vegetable soup, and other vegetables) from a
The HD-HP questionnaires were usually administered after
43-item food-frequency questionnaire within the HD-HP ques-
semimandatory activities to encourage participation, but they
tionnaire (28). Agreement with attitudes was measured on a
were anonymous and participation was voluntary. One school
Likert scale of responses: “strongly agree,” “agree,” “neither
was excluded because of protocol nonadherence and a response
agree nor disagree,” “disagree,” or “strongly disagree.” To eval-
rate of 55% on the first 2 questionnaires. The final sample
uate a student’s exposure to health promotion and prevention at
comprised 16 schools (Appendix A). Our total response rate,
his or her school, we asked 16 questions about the school’s
including all respondents from any of the 3 survey administra-
encouragement of a minimization of stress and about both the
tions and any of the 16 schools, was 80.3%.
school’s and classmates’ encouragement of healthy eating, reg-
Students were tracked across 3 time points throughout 4 y of
ular exercise, responsible alcohol use, and discouragement of
medical school by using an anonymous unique identifier that
smoking. We weighted each topic equally when summing the 16
Downloaded from
used a student’s mother’s initials at her birth and father’s first 2
responses to create an individual school health-promotion or
initials. Of the 2316 students who provided responses at some
“preventive-dose” score, as assessed by one student; this in-
time point, 72% (n
1658) did so at more than one time point,
dividual student’s “preventive dose” assessment of a partic-
and 42% (n
970) were tracked across all 3 survey time points.
ular school was the variable of interest. Higher values for this
variable indicate a stronger perception of a more preventive,
Variables
www.ajcn.org
health-promoting school environment (29). At each time
Within the HD-HP questionnaire, queries on outcomes of self-
point, intended specialties were collapsed into “primary care”
reported relevance and frequency of talking to patients about
(family medicine, general internal medicine, obstetrics/gyne-
nutrition were asked along with 11 other counseling and 9 screen-
cology, pediatrics, and preventive medicine/public health),
ing practices. The questionnaire administered while the students
“subspecialists”
(anesthesia/pathology/radiology,
emer-
by on October 17, 2010
were freshmen contained only one nutrition counseling question:
gency medicine, medical subspecialist, surgery, psychiatry,
“How relevant do you think talking to patients about nutrition
and urology), and “undecided.”
will be in your intended practice?” The response options were
“not at all,” “somewhat,” and “highly.” All subsequent ques-
Statistics
tionnaires also contained questions on confidence and training
in nutrition counseling. The questionnaire administered to
All analyses were conducted with the use of SUDAAN
seniors also included the following question: “With a typical
(30)—a program that accounts for nonindependent observations
general medicine patient, how often do you actually talk to
arising from the clustering of students into schools and the cor-
your patients about nutrition?” The response options were
related responses from each student over time. The counseling
“never/rarely,” “sometimes,” or “usually/always.” Because
outcomes across time are reported in Table 1, only for the subset
of the relatively small cell sizes for subjects reporting “not at
of students responding at all 3 time points. The cross-tabulated
all” relevant and “never/rarely” concerning counseling
associations of relevance and frequency with categorical predic-
(17% for all categories), the less than “highly” relevant and
tors were determined (Table 2) by using the chi-square test; we
less than “usually/always” counseling responses were col-
used a P value
0.01 to test for statistical significance because of
lapsed into one category.
the large number of bivariate associations being tested. To help
Independent predictors were a priori choices based on past
preserve statistical power, we included all observations on the
literature concerning medical student or physician nutrition
predictors and outcome at all time points.
counseling behaviors. Both perceived relevance and reported
Starting with multivariate logistic models that included all the
frequency of nutrition counseling were cross-tabulated with the
potential predictors listed above, final models were selected (Ta-
following medical student variables: dietary practices (servings
bles 3 and 4) via backward elimination and stepwise regression
per day of fruit and vegetables, change in fruit and vegetable
methods, leaving only covariates with a significance of P  0.05.
servings over time, and vegetarianism), demographic character-
Because clustering limited our models to 14 maximal df, multiple
istics (sex and race-ethnicity), physical health (body mass index,
response levels of some variables were collapsed. Meaningful
attempted weight loss, and perceived general health), and clinical
interaction terms were also evaluated (time
intended specialty,
characteristics (intended specialty and a variable measuring a
fruit and vegetable consumption
time, and fruit and vegetable
student’s assessment of his or her school’s health-promotion
consumption
preventive dose). At P  0.01, only time
score). Counseling outcomes were also cross-tabulated with the
intended specialty was bivariately significant and hence offered
level of agreement with several attitudinal questions plus one
into the multivariate relevance model. Models were examined to
mentoring question. These questions were as follows: “Primary
confirm modeling assumptions and to assess fit.

---

MEDICAL STUDENTS’ NUTRITION COUNSELING PREDICTORS
657
TABLE 1
US medical students’ self-reported nutrition counseling behaviors and attitudes throughout medical school: perceived relevance to intended practice,
training, confidence, and frequency1
Time point
Freshman
Orientation to
orientation
wards
Senior year
P
(n
970)
(n
970)
(n
970)
(chi-square)
%
Relevance of nutrition counseling to intended practice
All students
72
2.2
61
2.3
46
2.1
0.0003
Primary care specialty
0.5
Not at all
0
0.2 (425)
0
0.4 (263)
0
0.0 (290)
Somewhat
19
2.7 (425)
21
2.1 (263)
23
2.7 (290)
Highly
81
2.7 (425)
78
2.1 (263)
77
2.7 (290)
Subspecialist specialty
0.0009
Not at all
1
0.4 (349)
6
0.8 (465)
21
1.9 (589)
Somewhat
37
3.3 (349)
44
3.2 (465)
50
2.5 (589)
Highly
62
3.4 (349)
51
3.1 (465)
30
2.2 (589)
Training in nutrition counseling
0.02
None
—2
8
1.5
4
0.9
Some
—2
77
1.8
74
1.8
Extensive
—2
15
2.3
22
2.2
Downloaded from
Confidence in performing nutrition counseling
0.8
Not at all
—2
3
0.6
3
0.5
Somewhat
—2
52
3.6
53
2.2
Highly
—2
45
3.4
44
2.1
Frequency of nutrition counseling with typical general
medicine patient
www.ajcn.org
Never/rarely
—2
—2
15
1.4
Sometimes
—2
—2
68 (1.6)
Usually/always
—2
—2
17 (1.6)
1 Includes predictor and outcome data only for individuals who responded at all 3 time points. n in parentheses.
by on October 17, 2010
2 Not queried about at this time point.
Multiple imputation was used to overcome the large quantity
fruit and vegetable intakes than did the men (P
0.03) (data
of information lost as a result of incomplete data on at least one
not shown).
variable (31). Although the median item nonresponse rate was
Overall, 60% of all 970 students responding to all 3 surveys
3% (range: 0 –15%), the fruit and vegetable variable was missing
perceived nutrition counseling to be highly relevant in their in-
for up to 14% of participants, because this summary variable was
tended practices (data not shown). Freshman-year students were
computed as missing if any of its 8 constituent components were
more likely (72%) to find nutrition counseling highly relevant
missing. The selected multivariate models were then analyzed
than were students at the time of ward orientation (61%) or in
with 5 imputed datasets. Variables with missing rates over 4%
their senior year (46; overall P for trend
0.0003) (Table 1).
were imputed by using a nonnormal Bayesian imputation pro-
Over time, all students (P
0.0003) as well as those intending
cedure (32) or SAS’ PROC MI (33); variance estimates were
to subspecialize (P
0.0009) had declining perceptions of coun-
adjusted for imputation use by the MIANALYZE procedure of
seling relevance, whereas intended primary care specialists’ per-
SAS. Imputed results were found to be consistent with those from
ceived relevance remained high and did not decline (P
0.5).
the unimputed models. The results of the imputation procedures
Only 22% of the students believed that they had been extensively
were selected to represent the final results of our multivariate
trained in nutrition counseling in their senior year. Although the
analyses.
percentage of students who felt extensively trained increased
over time (P
0.02), the percentage who felt highly confident
did not change over time (P
0.8). A similar percentage of
RESULTS
seniors reported “usually/always” (17%) and “never/rarely”
Our response rates were 87%, 78%, and 75% on the 3
(16%) providing nutrition counseling to typical general medicine
questionnaire administrations. The median age at freshman
patients.
orientation was 23 y (range: 17– 45 y); age was not associated
With the exception of intended specialty, predictors of
with any of our outcomes (data not shown). Most of the stu-
counseling relevance were consistent across time and, hence,
dents reported fruit and vegetable intakes of
3 servings/d
the bivariate relations are reported overall (Table 2). There
(median:
2.7 servings per day), and these intakes declined
was a dose-response for higher perceived nutrition counseling
over time (P
0.008). The women’s consumption decreased
relevance by increasing quintile of fruit and vegetable con-
from 2.8 to 2.4 servings/d, and the men’s consumption de-
sumption (P for trend  0.0001). Students intending to spe-
creased from 2.6 to 2.2 servings/d; the women reported higher
cialize in primary care were much more likely to find nutrition

---

658
SPENCER ET AL
TABLE 2
Characteristics of US medical students associated with their perceived relevance and self-reported frequency of clinical nutrition counseling: bivariate
associations
Relevance across all time points1
Frequency at senior year2
Response of
Response of
n
“highly”
P (chi-square)3
n
“usually/always”
P (chi-square)3
%
%
Total sample
4651
60
—
17
—
Demographic characteristics
Sex
—
—
 0.0001
—
—
0.01
Female
2153
69
—
652
19
—
Male
2494
53
—
740
15
—
Ethnicity
—
—
NS (0.04)
—
—
NS (0.02)
White
2962
57
—
887
14
—
Asian
896
60
—
258
18
—
Black, Hispanic, or other
782
67
—
245
26
—
Health behaviors and characteristics
Quintile of fruit and vegetable
—
—
 0.0001
—
—
 0.0001
intakes (servings/d)
Highest
858
69
—
212
23
—
Second highest
880
65
—
230
19
—
Middle
843
62
—
227
22
—
Downloaded from
Second lowest
898
57
—
277
13
—
Lowest
811
49
—
293
12
—
BMI (kg/m2)
—
—
0.01
—
—
NS (0.05)
25, Normal-weight or underweight
3462
62
—
979
18
—
25 to 30, Overweight
923
54
—
313
14
—
30, Obese
165
59
—
56
11
—
www.ajcn.org
Self-defined vegetarian
—
—
NS (0.05)
—
—
NS (0.2)
Vegetarian
430
67
—
112
14
—
Omnivore
4210
60
—
1278
17
—
Self-rated general health
—
—
NS (0.2)
—
—
NS (0.3)
Excellent
1596
62
—
534
19
—
by on October 17, 2010
Good or very good
2928
60
—
799
16
—
Fair or poor
110
54
—
49
14
—
Preventive and school-based attitudes
Primary prevention is effective against
—
—
0.01
—
—
0.01
premature cardiovascular disease
Agree
2225
70
—
946
18
—
Disagree or neither agree nor disagree
1100
59
—
353
12
—
School health-promotion score
—
—
NS (0.1)
—
—
0.01
Highest tertile
981
56
—
533
19
—
Medium tertile
907
50
—
448
16
—
Lowest tertile
865
54
—
356
14
—
Student is more credible if he or she eats
—
—
 0.0001
—
—
 0.001
a healthy diet
Agree
3988
64
—
1135
18
—
Disagree or neither agree nor disagree
588
37
—
223
10
—
Physicians have a responsibility to
—
—
 0.0001
—
—
NS (0.06)
promote prevention with their
patients
Agree
4206
63
—
1211
18
—
Disagree or neither agree nor disagree
356
36
—
141
8
—
Preventive emphasis by student’s
—
—
 0.001
—
—
NS (0.2)
personal physician
Some or a lot
2892
63
—
873
18
—
 Some
1698
56
—
494
15
—
Intended specialty
—
—
 0.0001
—
—
0.01
Primary care specialties
1636
79
—
421
25
—
Subspecialties
2310
45
—
935
13
—
Undecided
664
65
—
14
29
—
1 Includes predictor and outcome data from all 3 time points, with most persons responding at Œ1 time point.
2 Includes predictor and outcome data from senior year only.
3 Nonsignificant P values include 0.01  P  0.05 because of adjustment for testing multiple variables.

---

MEDICAL STUDENTS’ NUTRITION COUNSELING PREDICTORS
659
TABLE 3
TABLE 4
Multivariate predictors for US medical students who perceived the
Multivariate predictors of senior US medical students who self-reported
relevance of nutrition counseling to be “high” throughout medical school
frequent nutrition counseling of typical patients
Variable
Odds ratio1 (95% CI)
Variable
Odds ratio1 (95% CI)
Sex
Ethnicity
Female
1.442 (1.14,1.81)
Asian
1.40 (0.96,2.05)
vs Male (referent)
1.00
Black, Hispanic, or other
2.122 (1.43,3.15)
Food and vegetable intakes (continuous)
1.103 (1.04,1.16)
vs White (referent)
1.00
Primary prevention is effective against
Food and vegetable intakes (servings/d)
1.113 (1.04,1.18)
premature cardiovascular disease
Student will be more credible if he or
Strongly agree
3.633 (2.11,6.27)
she eats a healthy diet
Agree
1.782 (1.12,2.82)
Strongly agree
2.383 (1.46,3.89)
vs Strongly disagree, disagree, or
1.00
Agree
1.44 (0.97,2.14)
neither agree nor
vs Strongly disagree, disagree, or
1.00
disagree (referent)
neither agree nor
Preventive emphasis by student’s
disagree (referent)
physician
Intended specialty
Some or a lot
1.292 (1.06,1.57)
Primary care
2.093 (1.47,2.98)
vs Not much or none (referent)
1.00
vs Subspecialist (referent)
1.00
Time point and intended specialty
1 The odds of answering “usually/always” vs “sometimes” or “never/
Ward orientation vs freshmen
rarely” in the medical students’ senior year in response to a question about the
orientation (referent)
frequency in which they engaged in nutrition counseling.
Among primary care specialists
0.90 (0.64,1.26)
Downloaded from
2 P
0.05.
Among subspecialists
0.672 (0.50,0.89)
3 P
0.001.
Senior year vs freshmen
orientation (referent)
Among primary care specialists
0.84 (0.63,1.13)
DISCUSSION
Among subspecialists
0.264 (0.21,0.32)
By senior year, 46% of the students (compared with 72% of
1
www.ajcn.org
The odds of perceiving nutrition counseling to be highly relevant vs
freshmen) perceived nutrition counseling to be highly rele-
somewhat or not at all relevant across all time points surveyed.
vant in their intended practices. The substantial decline in
2 P
0.05.
perceived nutrition counseling relevance over time was fueled
3 P
0.001.
only by those intending to go into a subspecialty. Moreover,
4 P  0.0001.
an intended primary care specialty consistently predicted both
by on October 17, 2010
higher perceived nutrition counseling relevance and more
frequent counseling. These findings are similar to the findings
of 2 studies in which primary care practitioners, relative to
counseling highly relevant (79%) than were those intending to
subspecialists, counseled their patients more frequently about
subspecialize (45%; P  0.0001) and were more likely to
nutrition (13, 15). However, another physician study did not
counsel (P
0.006). A school’s preventive dose was a sig-
find this specialty-based difference (34). In our study, medical
nificant predictor (P
0.005) of nutrition counseling fre-
school training did not maintain or increase the relatively
quency only. Both perceived nutrition counseling relevance
lower percentage of subspecialists that perceived nutrition
and frequency were predicted by sex, ethnicity, fruit and veg-
counseling to be highly relevant. We found that a school’s
etable intake, BMI, belief in the efficacy of CVD prevention,
preventive, health promotive emphasis was a weak predictor
belief in increased credibility if a healthy diet was consumed,
of nutrition counseling frequency in the bivariate analysis, but
and intended specialty.
not in the multivariate analysis.
Results from the multivariate model showed that students
In their senior year, 25% of the students believed that they
were more likely to find nutrition counseling highly relevant if
had been extensively trained in nutrition, 50% were highly
they were female (P
0.005), consumed more fruit and vege-
confident about their nutrition counseling, and 20% usually
tables (P
0.002), believed that primary prevention was effec-
or always counseled their typical general medicine patients.
tive against premature CVD (P
0.02), or had personal physi-
Regarding training, a 2003 Association of American Medical
cians who had encouraged disease prevention (P
0.02) (Table
Colleges survey of students from all US medical schools
3). The effect of intended specialty was modified by time. For
(n
13 764) similarly reported that 46% believed that “ap-
intended subspecialists, perceived counseling relevance was
propriate time” had been “devoted to nutrition instruction”;
lower than that for intended primary care physicians (regardless
3% believed that the time devoted was “excessive” (35). It is
of when queried about) and decreased only for subspecialists
an interesting conundrum that, despite the improvement in the
(odds ratio: 0.26) over time.
medical students’ perception of feeling adequately trained to
Students were more likely to report frequent nutrition coun-
provide nutrition counseling (P
0.02), the students’ confi-
seling if they were black, Hispanic, or other (compared with
dence in counseling patients did not improve (P
0.8).
Asian or white) (P
0.001), consumed more fruit and vegetables
A broad review of the literature suggests that the nutrition
(P
0.004), believed strongly that they would be more credible
training of many medical students is inadequate; our data suggest
if they ate a healthy diet themselves (P
0.003), or intended to
that the training the medical students in our study received did not
specialize in primary care specialties (P
0.0007) (Table 4).
adequately address the requisite skills for real patient encounters.

---

660
SPENCER ET AL
This finding was also suggested in 2 previous surveys. Similar to
One strength of our study was that we collected data on both
our findings, a 2002 survey (36) of 290 first-, second-, and third-
personal and professional nutrition behaviors from a sample of
year medical students and a 1986 survey (37) of 139 third-year
16 medical schools across the United States. Previous studies of
medical students reported that some students were lacking in
medical students’ dietary intakes have been limited by sample
knowledge about dietary recommendations, healthy BMI,
size (only 1 of 49 was a survey of Œ300 students) and location
CVD risk factors (36), and the ability to deliver nutrition
(only 1 of 50 was at more than one school). Regarding pro-
counseling and education (37). This lack of confidence may
fessional nutrition practices, previous reports have not
have important consequences; according to a popular behav-
assessed how the emphasis of various medical schools on
ioral theory (38), medical students’ feelings of self-efficacy
preventive medicine affects the counseling behaviors of med-
would be important in achieving higher rates of nutrition
ical students differently; we are aware of only one study that
counseling (39). Indeed, the literature suggests that self-
evaluated the counseling and preventive-nutrition attitudes of
efficacy is associated with medical students’ personal health
medical students at many schools (49). Our study was unique
successes (40) and with the amount of effort physicians spend
in that it examined the effect of the general promotion of
on health-promotion strategies (41).
preventive nutrition by numerous medical schools on the
In models, higher fruit and vegetable intakes, at least one
association between students’ personal and clinical nutri-
positive opinion on the importance of disease prevention, sex,
tional attitudes and behaviors.
and ethnicity were consistent predictors of both frequent coun-
The principal strength of our study was that it was the first, to
seling and a high perceived relevance of such counseling. These
our knowledge, to provide a natural history of the entire medical
findings are supported by the literature on physician nutrition
school experience. We examined temporal trends in 3 nutrition
counseling. Several studies have reported associations between
counseling variables and many potential correlates, including
physicians’ healthier diets and increased nutrition counseling of
how changes over time in key correlates affected counseling
Downloaded from
patients (5, 10, 14, 17). Our study is only the third study (10, 14)
frequency and a change in perceived relevance of counseling.
Other than for subspecialist-focused students, there were no dif-
to report this association between healthier dietary intakes
ferent effects by time for predicting perceived nutrition counsel-
that correspond to national dietary guidelines and the first
ing relevance or its observed decrease over time. One conclusion
study to report this association in medical students. As in our
based on this relative absence of a time effect was that most
study, increased clinical nutrition counseling was previously
students’ nutrition counseling attitudes correlated more strongly
shown to be associated with physicians having more positive
with their endogenous attitudes than with the experience of med-
www.ajcn.org
prevention attitudes (42, 43), being female (42, 44), or being
ical school.
black, Hispanic, or other (10). In our study, black physicians
Interventions to improve the professional nutrition practices
counseled their patients about nutrition more frequently than
of students can be built on this study’s foundation. During med-
did other ethnic groups, regardless of specialty, and a higher
ical education training, students were progressively less likely to
by on October 17, 2010
percentage of blacks and Hispanics than of other ethnic groups
find nutrition counseling highly relevant in their intended prac-
intended to specialize in primary care. These 2 phenomena
tices. Training interventions in nutrition counseling are war-
were responsible for the increased rates of counseling by
ranted. These interventions could be targeted at students inter-
blacks and Hispanics.
ested in subspecialties to enhance their perceptions of the
Although national guidelines at the time of this survey
relevance of nutrition counseling in their practices. Because stu-
recommended consuming
5 fruit and vegetable servings/d
dents interested in primary care also expressed modest enthusi-
(23), only 11.4% of medical students reported consuming this
asm for clinical nutrition counseling, it may be beneficial to
amount. Furthermore, fruit and vegetable consumption de-
rethink the paradigm of nutrition education, making nutrition
creased during medical school, perhaps because of normal
more relevant to all disciplines. All students can benefit from
dietary changes during those years (45) or because of a de-
more practical experiences with standardized patients and with
clining interest in or prioritization of personal prevention as
efforts to bolster their confidence in talking to their patients about
training progressed.
nutrition. Future researchers may attempt to quantify the nutri-
A limitation of this study was that our sample of schools was
tion counseling behaviors of medical students in an effort to
not randomly selected. This may have caused our conclusions to
improve their nutrition training and thereby the quality of their
be less generalizable than those derived from a random sample.
clinical counseling practices.
Although our self-reported data could have introduced some bias
toward overreporting, we validated the frequency of nutrition
EHS conducted the analyses, interpreted the data, and wrote and edited the
manuscript. EF designed the study, helped interpret the data, and edited the
counseling in our study population via extensive standardized
manuscript. LKE and VSH helped with the statistical analyses. DAG and
patient (46, 47) testing and found a strong relation (odds ratio:
MKS helped interpret the data and edited the manuscript extensively. None
1.93) between self-report and objective measures (48). Frank et
of the authors had any potential conflicts of interest.
al (29) also found strong correlations between deans’ and stu-
dents’ perceptions of their schools’ health-promotion environ-
ments. Loss to follow-up is a limitation common to longitudinal
REFERENCES
studies of this type. Another limitation was that we were confined
1. US Department of Health and Human Services, Centers for Disease
to qualitative descriptions of “talking to patients about nutrition.”
Control and Prevention, National Center for Health Statistics. 2000.
Therefore, we were unable to describe the quantity, quality, or
Healthy People 2010 Database. Internet: http://wonder.cdc.gov/
data2010/focus.htm (accessed 8 November 2003).
content of the counseling of the students and could not account
2. Mokdad AH, Bowman BA, Ford ES, Vinicor F, Marks JS, Koplan JP.
for the counseling of patients not perceived as “typical general
The continuing epidemics of obesity and diabetes in the United States.
medicine” patients.
J Am Med Assoc 2001;286:1195–200.

---

MEDICAL STUDENTS’ NUTRITION COUNSELING PREDICTORS
661
3. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and
2000. Internet: http://silk.nih.gov/public/cbz2zoz.@www.med.total.
trends in obesity among US adults, 1999 –2000. JAMA 2002;288:
fy2000.dsncc (accessed 13 July 2004).
1723–7.
26. Stoll M. The best medical schools. 2000 ed. New York, NY: Princeton
4. Horrocks P, Blackmore R, Wright AD. A long-term follow-up of dietary
Review Publishing, 1999.
advice in maturity-onset diabetes: the experience of one centre in the
27. American Association of Medical Colleges. Matriculant age at antici-
U.K. Prospective Study. Diabet Med 1987;4:241– 4.
pated matriculation, 1992-2001. Internet: http://www.aamc.org/data/
5. Ammerman AS, DeVellis RF, Carey TS, et al. Physician-based diet
facts/archive/famg112001a.htm (accessed 10 December 2004).
counseling for cholesterol reduction: current practices, determinants,
28. Spencer EH, Elon LK, Hertzberg VS, Stein AD, Frank E. Validation of
and strategies for improvement. Prev Med 1993;22:96 –109.
a brief diet survey instrument among medical students. J Am Diet Assoc
6. Russell NK, Roter DL. Health promotion counseling of chronic-disease
2005;105:802– 6.
patients during primary care visits. Am J Public Health 1993;83:979 –
29. Frank E, Hedgecock J, Elon LK. Personal health promotion at U.S.
82.
medical schools: a quantitative study and qualitative description of
7. Kushner RF. Barriers to providing nutrition counseling by physi-
deans’ and students’ perceptions. BioMed Central Medical Education.
cians: a survey of primary care practitioners. Prev Med 1995;24:546 –
Internet: http://www.biomedcentral.com/1472-6920/4/29 (accessed 29
52.
January 2005).
8. Wechsler H, Levine S, Idelson RK, Schor EL, Coakley E. The physi-
30. SUDAAN. Version 8.0.2. Research Triangle Park, NC: Research Tri-
cian’s role in health promotion revisited—a survey of primary care
angle Institute, 2003.
practitioners. N Engl J Med 1996;334:996 – 8.
31. Allison PD. Missing data. London, UK: Sage Publications, Inc, 2002.
9. Sciamanna CN, DePue JD, Goldstein MG, et al. Nutrition counseling in
32. Rubin DB. Multiple imputation for nonresponse in surveys. New York,
the promoting cancer prevention in primary care study. Prev Med 2002;
NY: J Wiley & Sons, 1987.
35:437– 46.
33. SAS Institute Inc. Statistical analysis software for Windows. Version
10. Frank E, Wright EH, Serdula MK, Elon LK, Baldwin G. US women
8.2. Cary, NC: SAS Institute Inc, 2000.
physicians’ personal and clinical nutrition-related practices. Am J Clin
34. Radecki SE. Patient counseling by specialists in internal medi-
Nutr 2002;75:326 –32.
cine: results of a nationwide survey. Patient Educ Counsel 1986;8:
11. Anis NA, Lee RE, Ellerbeck EF, Nazir N, Greiner A, Ahluwalia JS.
179 – 88.
Direct observation of physician counseling on dietary habits and exer-
35. American Association of Medical Colleges. Medical school graduation
Downloaded from
cise: patient, physician, and office correlates. Prev Med 2004;38:198 –
questionnaire, all schools report. Internet: http://www.aamc.org/data/
202.
gq/allschoolsreports/start.htm (accessed 7 February 2006).
36. Foster KY, Diehl NS, Shaw D, et al. Medical students’ readiness to
12. Lewis CE, Clancy C, Leake B, Schwartz JS. The counseling practices of
provide lifestyle counseling for overweight patients. Eating Behav 2002;
internists. Ann Intern Med 1991;114:54 – 8.
3:1–13.
13. Schwartz JS, Lewis CE, Clancy C, Kinosian MS, Radany MH, Koplan
37. Scott C, Greig LM, Neighbor WE. Curricular influences on preventive-
JP. Internists’ practices in health promotion and disease prevention. A
care attitudes. Prev Med 1986;15:422–31.
survey. Ann Intern Med 1991;114:46 –53.
www.ajcn.org
38. Bandura A. Social learning through imitation. In: Jones MR, ed. Ne-
14. Hyman D, Maibach EW, Flora JA, Fortmann SP. Cholesterol treatment
braska symposium on motivation. Lincoln, NE: University of Nebraska
practices
of
primary
care
physicians.
Public
Health
Rep
Press, 1962.
1992;107:441– 8.
39. Bandura A. Self-efficacy in changing societies. New York, NY: Cam-
15. Frank E, Rothenberg R, Lewis C, Belodoff BF. Correlates of physicians’
bridge University Press, 1995.
prevention-related practices. Findings from the Women Physicians’
40. Zimmerman BJ. Enhancing student academic and health functioning: a
by on October 17, 2010
Health Study. Arch Fam Med 2000;9:359 – 67.
self-regulatory perspective. School Psychol Q 1996;11:47– 66.
16. Shea S, Basch CE, Zybert P. Correlates of internists’ practices in caring
41. Yeager KK, Donehoo RS, Macera CA, Croft JB, Heath GW, Lane
for patients with elevated serum cholesterol. Am J Health Promot 1990;
MJ. Health promotion practices among physicians. Am J Prev Med
4:421– 8.
1996;12:238 – 41.
17. Levine BS, Wigren MM, Chapman DS, Kerner JF, Bergman RL,
42. Price J, Desmond SM, Krol RA, Snyder FF, O’Connell JK. Family
Rivlin RS. A national survey of attitudes and practices of primary-
practice physicians’ beliefs, attitudes, and practices regarding obesity.
care physicians relating to nutrition: strategies for enhancing the use
Am J Prev Med 1987;3:694 –704.
of clinical nutrition in medical practice. Am J Clin Nutr 1993;57:
43. Schucker B, Wittes JT, Cutler JA, et al. Change in physician perspective
115–9.
on cholesterol and heart disease. Results from two national surveys.
18. Heimburger D, Ullmann D, Ramsey MJ, et al. Dietary habits of first-year
JAMA 1987;258):3521– 6.
medical students assessed during clinical nutrition course. Nutrition
44. Frank E, Harvey LK. Prevention advice rates of women and men phy-
1994;10:214 –20.
sicians. Arch Fam Med 1996;5:215–2.
19. Kolasa K, Jobe AC, Miller MG, Clay MC. Teaching medical students
45. Crapse F, Hudgins PM, Baker HH. Lifestyle changes associated with
cancer risk reduction nutrition counseling using a multimedia program.
osteopathic
medical
education.
J
Am
Osteopath
Assoc
Fam Med 1999;31:200 – 4.
1993;93:1051– 4.
20. Kann P, Lane DS. Breast cancer screening knowledge and skills of
46. Dresselhaus T, Peabody J, Lee M, Wang M, Luck J. Measuring
students upon entering and exiting a medical school. Acad Med 1998;
compliance with preventive care guidelines: standardized patients,
73:904 – 6.
clinical vignettes, and the medical record. J Gen Intern Med 2000;
21. Frank E. Physician health and patient care. JAMA 2004;291:637.
15:782– 8.
22. US Department of Health and Human Services. Healthy people 2010.
47. Peabody J, Luck J, Glassman P, Dresselhaus T, Lee M. Comparison of
With understanding and improving health and objectives for improving
vignettes, standardized patients, and chart abstraction: a prospective
health. 2000. Internet: http://www.healthypeople.gov/document/tab-
validation study of 3 methods for measuring quality. JAMA 2000;283:
leofcontents.htm (accessed 16 December 2003).
1715–22.
23. National Institutes of Health, National Cancer Institute, Produce for
48. Frank E, McLendon L, Elon LK, Denniston M, Fitzmaurice D,
Better Health Foundation. 2006. Eat 5 to 9 a day for better health.
Hertzberg V. Medical students’ self-reported typical counseling prac-
Internet: http://www.5aday.gov. (accessed 14 February 2004).
tices are similar to those assessed using standardized patients. Med-
24. American Association of Medical Colleges. AAMC data warehouse:
GenMed 2005;7:2.
student records system as of December 12, 2000, reflecting those ac-
49. Babbott D, Baldwin DC, Killian CD, Weaver SO. Trends in evolution of
tively enrolled on October 31, 2000. Internet: http://www.aamc.org/
specialty choice. J Am Med Assoc 1989;261:2367–73.
stuapps/facts/start.htm (accessed 13 July 2004).
50. Konen JC, Fromm BS. Changes in personal health behaviors of medical
25. National Institutes of Health. Support to US medical schools, fiscal year
students. Med Teach 1992;14:321–5.

---

662
SPENCER ET AL
Appendix A
Characteristics of participating US medical schools
NIH
Location
Public or
research
No. of
No. of
No. of
School
(state)
private
rank1
students1
freshmen1
respondents
Creighton University School of Medicine
NE
Private
108
450
118
105
Duke University School of Medicine
NC
Private
11
379
100
96
Georgetown University School of Medicine
DC
Private
45
678
168
125
Loma Linda University School of Medicine
CA
Private
106
670
174
163
Medical College of Georgia School of Medicine
GA
Public
85
712
182
174
Mercer University School of Medicine
GA
Private
119
208
56
53
Morehouse School of Medicine
GA
Private
80
148
40
37
Texas Tech University Health Sciences Center School of
TX
Public
104
488
125
118
Medicine
Tulane University School of Medicine
LA
Private
79
611
153
91
UCLA School of Medicine/Charles R Drew University of
CA
Public
13
668
170
162
Medicine and Science
University of Colorado School of Medicine
CO
Public
20
524
133
128
UMDNJ2—Robert Wood Johnson Medical School
NJ
Public
62
632
150
132
University of Pennsylvania School of Medicine
PA
Private
2
594
150
142
University of Rochester School of Medicine and Dentistry
NY
Private
31
397
100
99
University of Washington School of Medicine
WA
Public
6
771
101
60
Wayne State University School of Medicine
MI
Public
54
1044
256
221
Downloaded from
1 For the academic year 1999 –2000.
2 University of Medicine and Dentistry of New Jersey.
www.ajcn.org
by on October 17, 2010

---