Content uploaded by Brian Miller
Author content
All content in this area was uploaded by Brian Miller on Dec 19, 2019
Content may be subject to copyright.
Original Research
Exercise is Medicine®: Knowledge and Awareness among Exercise Science and
Medical School Students
RACHEL N. MEALY*1, LAURA A. RICHARDSON‡1, BRIAN MILLER‡1, MELISSA SMITH‡1
and JUDITH A. JUVANCIC-HELTZEL‡1
1School of Sport Science and Wellness Education, The University of Akron, Akron, OH, USA
*Denotes undergraduate student author, ‡Denotes professional author
ABSTRACT
International Journal of Exercise Science 12(3): 505-514, 2019. The purpose of this exploratory study
was twofold: to determine whether exercise science and medical students are aware of the Exercise is
Medicine® (EIM®) program and to construct a tool that would permit assessment of EIM® variables with students
enrolled in both programs. The study consisted of a quantitative, cross-sectional design, using a self-report
electronic questionnaire. An Exploratory Factor Analysis (EFA) using principal component analysis extraction
method with Varimax factor rotation was employed to validate the survey instrument based on the expected
constructs, which posited five (5) contending factors: Value, Familiarity, Preparedness, Curricular Perceptions, and
Opinions. A pairwise comparison was then performed to compare elements of the EIM® scale identified from the
factor analysis by student type (medical and exercise science student) using multiple independent sample t-tests.
Based on the pairwise comparisons, there were statistically significant differences of all EIM® factors by student
type with the exception of Opinions (p = 0.109). Based on the trends observed in the data, exercise science students
had a more positive report for each EIM® factor compared to medical students. These findings suggest a
discrepancy in the delivery, acceptance, and implementation of the EIM® initiative between exercise professionals
and medical healthcare providers. Future investigation is warranted to validate this experimental instrument and
study the differences in EIM® factors among current medical and exercise professionals.
KEY WORDS: Physical activity, healthcare, non-communicable diseases, American College of
Sports Medicine, American Medical Association, medical students, exercise prescription,
exercise counseling, exercise education
INTRODUCTION
Advances in modern medicine have led to the development of many pharmacological agents
that control and treat non-communicable and lifestyle-induced diseases like hypertension,
dyslipidemia, obesity, and type II diabetes (16, 17, 25). As a result, prescription medications are
often the first line of treatment for these conditions with little consideration given to lifestyle
change as a viable solution. However, amidst what epidemiologists have identified as an
“inactivity epidemic,” the importance of widespread lifestyle alteration is becoming
increasingly clear (1, 2, 5, 6). For example, routine exercise is not only as effective as
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
506
pharmacological treatments in reversing these conditions, but also offers a long-term solution
to the problem and helps to reduce the growing financial burden of healthcare (8, 10, 20, 21).
Therefore, the role that physicians have in prescribing exercise, providing lifestyle counseling,
and connecting patients with qualified exercise professionals has received increased attention
in recent years.
In 2007, the American College of Sports Medicine (ACSM) and the American Medical
Association (AMA) established the Exercise is Medicine® (EIM®) campaign. Created just over a
decade ago, this campaign marks a new direction in healthcare and attempts to bridge the gap
between medicine and fitness through a focus on healthy living as the cornerstone of
preventative and curative treatment for a variety of medical conditions (3, 24). According to the
EIM® website, the goal of the initiative is to “[encourage] primary care physicians and other
healthcare providers to include physical activity when designing treatment plans and to refer
patients to evidence-based exercise programs and qualified exercise professionals, especially
those with the EIM® credential” (15). Since its establishment, little research has been conducted
to determine whether healthcare providers are aware of the program, receptive to its
implementation in the clinical setting, and adequately trained to provide physical activity
instruction to patients (4, 23, 24).
The purpose of this exploratory study was to evaluate awareness of the EIM® solution among
exercise science and medical students in order to determine whether the EIM® campaign is
effectively reaching future healthcare providers. This study also examined educational
differences between exercise science and medical students and whether their academic and
experiential preparation would allow them to prescribe exercise both confidently and effectively
(7, 13, 23). It was hypothesized that the responses to the questionnaire instrument would
indicate a discrepancy in the delivery, adoption, and implementation of the EIM® initiative
between future medical and exercise professionals. An additional objective of the study was to
develop an instrument that could be used as a tool to evaluate the adoption of EIM® among
healthcare providers.
METHODS
Participants
Undergraduate exercise science students and graduate-level medical students were recruited
via email from two different midwestern universities. Individuals were eligible to participate
in the study if they were currently enrolled in an exercise science program or a medical school
program at one of the two universities. The sample included 116* participants (39 exercise
science and 77 medical student respondents). The overall mean age was 23.40 years (SD = 8.75,
range = 17-47 years) and 33.62% of respondents were male while 66.38% were female. The mean
age for exercise science students was 19.74 years (SD = 6.37, range = 17-24 years) and 23.08%
were male and 76.92% female. The average age of the medical students was 25.25 years (SD =
8.97, range = 20-47 years) and 38.96% were male and 61.04% female. This study was reviewed
and approved by the university’s institutional review board (IRB).
*Nine (9) students did not report a type and were excluded from the data analysis.
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
507
Protocol
Participants received an email with an embedded link to the electronic questionnaire (consisting
of 30 items) which was created using the Qualtrics© survey platform (Provo, UT, USA, 2018).
Prior to beginning the questionnaire, it was stated that the students’ decision to continue to the
next page (i.e. the first block of questions) served as the informed consent. Participants were
asked to complete a block of questions related to their familiarity with the EIM® initiative. This
series of questions also included questions related to their perception of physical activity as a
viable therapeutic treatment, as well as their level of formal education and experience with
exercise prescription. These items were scored using a five-point Likert scale. The second
section included questions related to participant demographics, including age, gender identity
and educational level/background. The average time for completion of the questionnaire was
six minutes.
Statistical Analysis
Instrument Creation: An a priori sample size minimum was based on a subject to variable ratio
of 5:1 and an optimal sample size based on a subject to variable ratio of 10:1 (11, 19). The original
instrument had 21 items, thus a minimum sample size n = 105 and optimal sample size was n =
210 was determined to be necessary. Factor Analysis (FA) using principal component analysis
extraction with Varimax factor rotation was employed to parse items within the survey
instrument based on the expected constructs. A range of 4-7 identified factors were expected
based on the calculated range of v/3 to v/5, with v = 21, where v = number of variables (12). Prior
to FA, initial item-reduction was performed based on inter-item correlations r ≥0.90.
Assumptions of the FA procedure that could be statistically tested included sampling adequacy
via Kaiser-Meyer-Olkin (KMO) statistic with values above 0.70 indicative of sampling adequacy
and pattern of correlations yielding reliable factors. Additionally, sphericity was assessed via
the Bartlett’s test of Sphericity with statistical significance indicating that the correlation matrix
was not an identity matrix. Varimax rotation was used to create orthogonal factors. Model fit
criteria was based on the χ² measure of goodness of fit with non-significance indicative of model
fit. Subsequently, factors with Eigen values greater than 1 were allowed and expressed as
variance explained after rotation. The item loadings within each factor were based on the
highest loading absolute value. Loading values <0.50 were excluded from the final instrument.
Finally, parallel analysis confirmed the results of the factor analysis. For each factor, reliability
was assessed using inter-item internal consistency via Cronbach’s α. Subscales were quantified
as [
𝑆𝑢𝑏𝑠𝑐𝑎𝑙𝑒)𝑡𝑜𝑡𝑎𝑙 = )
∑
𝑋
/
0
] where N = number of scale items and X = individual Likert scale
response. EFA and reliability analysis was performed using SPSS version 22 (IBM, Chicago, IL,
2013).
Comparison of Student Type by EIM® Factors: Using multiple independent sample t-tests, a
pairwise comparison evaluated elements of the EIM® scale by student type (medical student and
exercise science student), as identified by the factor analysis. The assumption of homogeneity
of variance was tested using Levene’s Test of Equal Variance, with degrees of freedom (df)
adjusted for significant tests. Finally, the pairwise comparisons of EIM® factors by student type
were reported as mean difference (exercise science – medical student) and 95% confidence
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
508
intervals (CI) with significance set at α ≤ 0.05. Statistical analyses were performed using SPSS
version 22 (IBM, Chicago, IL, 2013).
RESULTS
Factor Analysis
Factor Analysis yielded six contending factors; however, Factors 5 and 6 were combined due to
similarity of content. Therefore, the final analysis considered the following five factors: (1) Value
of EIM® in future career, (2) Familiarity with the EIM® initiative, (3) Preparedness and confidence
in prescribing exercise, (4) Perception of exercise topics in current curricula, and (5) Opinions
about the role of EIM® as a medical tool. Factor loadings are illustrated in the rotated component
matrix in Table 1. The Kaiser-Meyer-Olkin (KMO) statistic indicated that the patterns of
correlations were not problematic (KMO = 0.72); thus, factor analysis was appropriate based on
the number of observations and the a priori sample size target. Furthermore, Bartlett’s test of
Sphericity reached significance [Χ²(210) = 898.39, p < 0.001], which indicated probable
relationships within the correlation matrix; thus, factor analysis was appropriate. The model fit
criteria did not reach statistical significance, χ²(99) = 99.807, p = 0.458, thus the factors in the
model were able to adequately explain covariance. The FA procedure identified six factors that
explained a total of 63.54% of the variance. After rotation, each factor was able to account for
14.67%, 11.61%, 11.29%, 11.02%, 7.70% and 7.25% of the variance for factors 1, 2, 3, 4, 5, and 6,
respectively. A parallel analysis confirmed the presence of the six factors, but as stated above,
Factors 5 and 6 were merged based on the similarity of content. Descriptive statistics for each
scale are illustrated in Table 2 below. Following factor analysis, reliability via internal
consistency was calculated, with each factor exceeding Cronbach’s Alpha threshold of
acceptability (α = 0.60), with the exception of Opinions (α = 0.42). Interestingly, when the items
related to the role of exercise and medical professionals in physical activity prescription were
removed (Q3 and Q5), the internal consistency improved to α = 0.80. Possible explanations for
this occurrence are described in the Discussion section.
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
509
Table 1. Rotated component matrix
FACTOR 5:
OPINIONS
0.70
0.69
0.51
*Item Q14 was reverse-coded to achieve consistent phrasing and scaling with other items
0.74
0.59
FACTOR 4:
PERCEPTIONS
0.85
0.83
FACTOR 3:
PREPAREDNESS
0.86
0.77
0.64
0.54
FACTOR 2:
FAMILIARITY
0.86
0.78
0.76
FACTOR 1:
VALUE
0.79
0.79
0.74
0.67
0.66
ITEM
Q21 As a future health professional, would you be interested in becoming
involved with EIM®?
Q16 Do you believe it adds value to your profession to receive education in
physical activity prescription?
Q19 Would you be interested in learning more about the EIM® initiative?
Q15 How likely are you to enroll in courses related to physical activity and
disease prevention/treatment?
Q23 How strongly do you believe that becoming involved with EIM® would
increase your credibility as a health professional?
Q20 Are you aware that the American College of Sports Medicine offers an
EIM training for both physicians and exercise professionals?
Q18 How familiar are you with the EIM® initiative from the American
College of Sports Medicine?
Q22 Are you aware that the American College of Sports Medicine offers
physicians a referral network between exercise professionals and physicians?
Q11 Based on your education, how prepared do you feel to prescribe physical
activity to patients?
Q12 How confident are you that you can effectively prescribe physical
activity to future patients?
Q10 How much formal training or education have you received in prescribing
physical activity?
Q14* How many college courses related to the role of exercise in disease
prevention have you taken so far?
Q17 What is your perception of the physical activity-related content being
offered in medical schools?
Q13 What is your opinion about the level of physical activity-related
education/training present in your current curriculum?
Q2 Do you believe it is a physician's role to prescribe physical activity to
patients?
Q7 Do you believe physical activity is a valid component of the treatment
plan?
Q6 Do you believe physical activity is a necessary component of preventative
medicine?
Q5 How often do you believe physicians should discuss physical activity with
patients?
Q3 Do you believe it is an exercise professional's role to prescribe physical
activity to patients?
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
510
Table 2. Scale descriptive statistics
Scale
Scoring
Range
Minimum
Maximum
Mean
SD
Items
α*
Value
5-25
5
21
9.87
3.45
5
0.78
Familiarity
3-15
3
9
8.13
1.54
3
0.62
Preparedness
4-20
4
19
12.64
3.65
4
0.74
Perceptions
2-10
2
10
5.71
2.40
2
0.81
Opinions
5-25
5
13
7.46
1.78
5
0.42
*α represents the indicated scale internal consistency via Cronbach's Alpha
Comparison of Student Type by EIM® Factors
Undergraduate exercise science students and graduate-level medical students were recruited
via email from two different midwestern universities. The total sample included N = 116*
participants and consisted of 39 exercise science and 77 medical students. Based on violations of
homogeneity of variance, df were adjusted for the t–tests of Value, Familiarity, Preparedness,
and Curricular Perceptions factors. Descriptive statistics revealed differences between student
type for each EIM® factor (Table 3). Pairwise comparisons by student type revealed statistically
significant differences for all EIM® factors, with the exception of Opinions, p = 0.109 (Table 4).
*Students that did not report a type were excluded, n = 9.
Table 3. Descriptive statistics of EIM® factors by student type
Exercise
Medical
Factor
Mean
±
SD
Mean
±
SD
Value
7.23
±
2.11
10.96
±
3.39
Familiarity
7.03
±
2.02
8.70
±
0.76
Preparedness
11.13
±
4.05
13.17
±
3.25
Perceptions
3.69
±
1.38
6.69
±
2.20
Opinions
7.77
±
1.68
7.22
±
1.75
Exercise Student n= 39, Medical Student n= 77
Table 4. Pairwise testing of EIM® factors by student type
95% CI
Levene's Test
Factor
d
t
df
p-value
Lower
Upper
F
p-value
Value*
-3.73
7.28
109.05
0.000
-4.75
-2.71
4.89
0.029
Familiarity*
-1.68
5.00
43.57
0.000
-2.35
-1.00
84.45
0.000
Preparedness*
-2.04
2.73
63.56
0.008
-3.53
-0.55
3.67
0.058
Perceptions*
-3.00
8.96
108.79
0.000
-3.66
-2.33
19.00
0.000
Opinions
0.55
1.62
114.00
0.109
-0.12
1.22
0.00
0.975
*Indicates that the df were adjusted
Mean differences (d) and Confidence Intervals (CI) are reported as Exercise Science – Medical Student
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
511
DISCUSSION
The purpose of this exploratory study was to evaluate awareness of the EIM® solution among
exercise science and medical students, examine educational differences between the two groups,
and develop a tool for the evaluation of EIM® implementation among healthcare providers.
Therefore, by surveying students in these areas, this study sought to determine whether the
EIM® campaign is effectively reaching future healthcare providers and whether their academic
and experiential preparation would allow them to prescribe exercise both confidently and
effectively. It was hypothesized that the responses to the questionnaire instrument would
indicate a discrepancy in the delivery, adoption, and implementation of the EIM® initiative
between future medical and exercise professionals.
Factor analysis of the questionnaire items revealed five major factors: (1) the value of EIM® in
students’ future careers, (2) their familiarity with the EIM® initiative, (3) their preparedness and
confidence in prescribing exercise, (4) their perception of exercise topics offered in current
curricula, and (5) their opinions about the role of Exercise is Medicine® as a medical tool. The
main finding of this study was the discovery of a significant difference in the perceived value of
the EIM® program between the two groups of participants. Exercise science students reported
more positively towards the Value factor, whereas medical students responded more
negatively. Notable differences were also observed for the Preparedness and Familiarity factors,
with exercise science students reporting a higher level of confidence prescribing exercise than
medical students. Exercise science students were also more aware of and familiar with the EIM®
initiative than their medical student counterparts. Possible explanations for these results include
differences in the curriculum and course requirements for exercise science and medical students.
In addition, the amount, duration, and type of exposure and engagement with exercise-related
topics and exercise prescription likely differs between the two groups.
The results of this study indicate that underlying challenges may exist for the Exercise is
Medicine® initiative in reaching its target audiences and medical doctors in particular. As
previously stated, the overarching goal of the EIM® program is to encourage a collaborative
effort between physicians and exercise professionals towards incorporating physical activity in
the treatment plan (15). However, this cannot be accomplished without widespread adoption
of the EIM® principles among healthcare providers on both sides of the proposed collaboration.
The inclusion of exercise prescription training using the EIM® protocol in medical school
curriculums would increase exposure to and proficiency in these areas and help ensure the
success of the EIM® initiative.
This investigation is not without limitations. First, the amount and type of participants that
completed the questionnaire could be improved. The study included 116 student respondents
from two universities. A larger sample size of participants from more institutions would help
make these findings more generalizable. Also, this exploratory study specifically evaluated
students and therefore does not provide an accurate picture of the situation among current
exercise and medical professionals in their respective fields. Future studies should test this
instrument among different groups of health professionals currently working in the field.
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
512
Second, the cross-sectional design of this study did not capture whether opinions and
perspectives related to EIM® change throughout the course of a person’s academic and
professional career. Therefore, administering this questionnaire at various points during a
person’s professional development would provide insight into changes over time. Another
possible limitation of this study was the self-report method of data collection, which can
introduce an unwanted level of subjectivity in reporting one’s ability to prescribe exercise. The
incorporation of a standardized measure for proficiency in exercise prescription would improve
this limitation. Finally, from a statistical standpoint, this investigation was exploratory in nature
and aimed to generate a preliminary model and metric to evaluate EIM®. Therefore, in order to
make generalizations about this instrument, further investigation is warranted to evaluate its
external validity. Validity could be assessed using confirmatory factor analysis and mixed
methods, which would allow for qualitative explanation of the quantitative responses obtained
from multiple health professional populations.
Following factor analysis, the Opinions factor did not meet Cronbach’s Alpha threshold of
acceptable internal consistency of α = 0.60. However, when items Q3 and Q5, which assessed
the perceived role of exercise and medical professionals in physical activity prescription, were
removed, internal consistency improved to α = 0.80. Further analysis of these items revealed
dissonance between the two groups. Exercise students recognized an equal role for exercise
professionals and physicians in prescribing physical activity (OR 1.03) while medical students
were almost twice as likely to not acknowledge the role of exercise professionals in prescribing
physical activity (OR 1.80). This finding indicates that future investigation on the Opinions
factor is warranted and shows that the questionnaire items included under this factor are
relevant when discussing the effectiveness of Exercise is Medicine®. For example, evaluating
opinions about which types of healthcare providers should have the largest role in prescribing
exercise would provide valuable insights about the obstacles that prevent EIM® adoption and
implementation. Similarly, determining whether healthcare providers consider physical
activity to be a legitimate therapeutic approach would offer additional information about the
likelihood of their utilizing the EIM® strategy. Inclusion of the Opinions factor items would
ultimately serve to strengthen the proposed instrument’s evaluative quality.
Future research should continue to determine the breadth and effect of the Exercise is Medicine®
initiative among students in health professions programs. However, this instrument should
also be adapted and repeated in future studies that evaluate the influence of the EIM® initiative
among other groups. For example, it would be beneficial to test this instrument with current
medical and exercise professionals in order to create a more relevant picture of the Exercise is
Medicine® program. Although this study only evaluated future medical doctors, future studies
should also evaluate whether there are significant differences between medical doctors and
osteopathic doctors and their involvement with EIM®.
REFERENCES
1. American College of Sports Medicine. Exercise is Medicine: Fact sheet. Retrieved from:
http://www.exerciseismedicine.org, 2018.
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
513
2. American College of Sports Medicine. Exercise is Medicine: The miracle drug. Retrieved from:
http://exerciseismedicine.org/assets/page_documents/EIM_HCP_1_Page_Summary.pdf, 2018.
3. American College of Sports Medicine. Exercise is Medicine: Healthcare providers’ action guide. Retrieved
from: http://exerciseismedicine.org/assets/page_documents/HCP_Action_Guide%285%29.pdf, 2018.
4. American College of Sports Medicine. Exercise is Medicine: Scorecard. Retrieved from:
http://www.exerciseismedicine.org/assets/page_documents/EIM%20Scorecard%20August%202016.pdf,
2016.
5. Barnes, P. M., & Schoenborn, C. A. Trends in adults receiving a recommendation for exercise or other physical
activity from a physician or other health professional. Natl Health Stat Report, (86): 1–8, 2012.
6. Blair, S. N. Physical Inactivity: The biggest public health problem of the 21st century. Br J Sports Med, 43: 1–3,
2009.
7. Cardinal, B. J., Park, E. A., Kim, M., & Cardinal, M. K. If exercise is medicine, where is exercise in medicine?
Review of U.S. medical education curricula for physical activity-related content. J Phys Act Health, 12(9): 1336–
1343, 2015.
8. Carlson, S. A., Fulton, J. E., Pratt, M., Yang, Z., & Adams, E. K. Inadequate physical activity and healthcare
expenditures in the United States. Prog Cardiovasc Dis, 57(4): 315–323, 2015.
9. Coleman, K. J., Ngor, E., Reynolds, K., Quinn, V. P., Koebnick, C., Young, D. R., … Sallis, R. E. Initial validation
of an exercise “vital sign” in electronic medical records. Med Sci Sports Exerc, 44(11): 2071–2076, 2012.
10. Coombes, J. S., Law, J., Lancashire, B., & Fassett, R. G. “Exercise is medicine”: Curbing the burden of chronic
disease and physical inactivity. Asia Pac J Public Health, 27(2): NP600-5, 2015.
11. Costello, A. B., & Osborne, J. W. Best practices in exploratory factor analysis: Four recommendations for
getting the most from your analysis. Practical assessment, research & evaluation, 10(7), 1-9, 2005.
12. Dimitrov, D. M. Quantitative research in education. New York, NY: Whittier; 2013.
13. Dirks-Naylor, A. J., Griffiths, C. L., Gibson, J. L., & Luu, J. A. (2016). The prevalence of exercise prescription-
related course offerings in United States pharmacy school curricula: Exercise is medicine. Adv Physiol Educ,
40(3): 319–322, 2016.
14. Dusenbury, C., Gaziano, T.A., Koplan, J.P., Nugent, R., Puska, P. & Willett, W.C. Disease control priorities in
developing countries. 2nd ed. New York, New York: Oxford University Press; 2006.
15. Exercise is Medicine. Website: http://www.exerciseismedicine.org/, 2018.
16. HealthyPeople.gov. 2020 Topics & objectives: Physical activity overview, 2014.
17. HealthyPeople.gov. 2020 Topics & objectives: Physical activity objectives, 2014.
18. Heath, G. W., Kolade, V. O., & Haynes, J. W. Exercise is medicine???: A pilot study linking primary care with
community physical activity support. Prev Med Rep, 2: 492–497, 2015.
19. Lomax, R. G., & Schumacker, R. E. A beginner's guide to structural equation modeling. New York, NY:
Routledge Academic; 2012.
Int J Exerc Sci 12(3): 505-514, 2019
International Journal of Exercise Science http://www.intjexersci.com
514
20. Martinson, B. C., Crain, A. L., Pronk, N. P., O’Connor, P. J., & Maciosek, M. V. Changes in physical activity
and short-term changes in healthcare charges: A prospective cohort study of older adults. J Prev Med, 37(4):
319–326, 2003.
21. Naci, H., & Ioannidis, J. P. A. Comparative effectiveness of exercise and drug interventions on mortality
outcomes: Metaepidemiological study. Br J Sports Med (Clinical Research Ed.), 347(4): f5577, 2013.
22. Office of Disease Prevention and Health Promotion (ODPHP), 2008 Physical activity guidelines for Americans
summary, 2008.
23. Sanchez, A., Bully, P., Martinez, C., & Grandes, G. Effectiveness of physical activity promotion interventions
in primary care: A review of reviews. Prev Med, 76(S): S56–S67, 2015.
24. Vuori, I. M. Role of primary healthcare in physical activity promotion. Dtsch Z Sportmed, 64(6): 176–182, 2013.
25. World Health Organization (WHO). Global health risks: Statistical report, 2009.
