Health Services and Outcomes Research

Global Variation in the Prevalence of Elevated Cholesterol

in Outpatients With Established Vascular Disease or 3
Cardiovascular Risk Factors According to National Indices
of Economic Development and Health System Performance
Lakshmi Venkitachalam, PhD; Kaijun Wang, PhD; Avi Porath, MD; Ramon Corbalan, MD;
Alan T. Hirsch, MD; David J. Cohen, MD, MSc; Sidney C. Smith, Jr, MD; E. Magnus Ohman, MD;
Ph. Gabriel Steg, MD; Deepak L. Bhatt, MD, MPH; Elizabeth A. Magnuson, ScD; on behalf of the REACH
Registry Investigators

Background—Elevated serum cholesterol accounts for a considerable proportion of cardiovascular disease worldwide. An
understanding of the relationship between country-level economic and health system factors and elevated cholesterol
may provide insight for prioritization of cardiovascular prevention programs.
Methods and Results—Using hierarchical models, we examined the relationship between elevated total cholesterol (⬎200
mg/dL) in 53 570 outpatients from 36 countries, and tertiles of several country-level indices: (1) gross national income,
(2) total expenditure on health as percentage of gross domestic product, (3) government expenditure on health as
percentage of total expenditure on health, (4) out-of-pocket expenditures as percentage of private expenditure on health,
and the World Health Organization indices of (5) Health System Achievement and (6) Performance/Efficiency. Overall,
38% of outpatients had total cholesterol ⬎200 mg/dL (⬎5.18 mmol/L), and 9.3% of the total variability in elevated
cholesterol was at the country level; this proportion was higher for patients with (12.1%) versus without (7.4%) history
of hyperlipidemia. Among patients with history of hyperlipidemia, countries in the highest tertile of gross national
income or World Health Organization Health System Achievement had lower odds of elevated cholesterol than lower
tertiles (P⬍0.001, for both). Countries in the highest tertile of out-of-pocket health expenditures had higher odds of
elevated cholesterol than those in the lowest tertile (P⬍0.001). No significant associations were found for patients
Downloaded from http://ahajournals.org by on December 31, 2019

without history of hyperlipidemia.

Conclusions—Global variations in the prevalence of elevated cholesterol among patients with history of hyperlipidemia
are associated with country-level economic development and health system indices. These results support the need for
strengthening efforts toward effective cardiovascular disease prevention and control and may provide insight for health
policy setting at the national level. (Circulation. 2012;125:1858-1869.)
Key Words: cardiovascular disease 䡲 hypercholesterolemia 䡲 global trends 䡲 health system performance
䡲 national health expenditures

T he World Health Organization (WHO) proposed the

2008 to 2013 Action Plan for the Global Strategy for the
Prevention and Control of Noncommunicable Diseases to
establish and strengthen initiatives for the surveillance, pre-
vention, and management of chronic diseases around the
world.1 The core objective of this effort was to “raise the
priority accorded to chronic diseases while encouraging
governments and health agencies to integrate preventive and

Received August 24, 2011; accepted March 1, 2012.

From the University of Missouri-Kansas City School of Medicine, Kansas City, MO (L.V., D.J.C., E.A.M.); Saint Luke’s Mid America Heart Institute,
Kansas City, MO (K.W., D.J.C., E.A.M.); Ben Gurion University of the Negev, Beer Sheva, Israel (A.P.); Cardiovascular Division, Pontificia Universidad
Catolica de Chile, Santiago, Chile (R.C.); Division of Epidemiology and Community Health, University of Minnesota School of Public Health,
Minneapolis, MN (A.T.H.); Center for Cardiovascular Science and Medicine, University of North Carolina, Chapel Hill, NC (S.C.S.); Division of
Cardiovascular Medicine, Duke University Medical Center, Durham, NC (E.M.O.); INSERM U-698, Université Paris-Diderot, Hôpital Bichat, Assistance
Publique-Hôpitaux de Paris, Paris, France (P.G.S.); VA Boston Healthcare System, Brigham and Women’s Hospital, and Harvard Medical School,
Boston, MA (D.L.B.).
A Complete List of the REACH Registry Investigators appears in JAMA. 2006;295:180 –189.
Guest Editor for this article was Wayne D. Rosamond, PhD.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.
111.064378/-/DC1.
Correspondence to Elizabeth A. Magnuson, ScD, Director, Health Economics and Technology Assessment, Saint Luke’s Mid America Heart Institute,
4401 Wornall Rd, Kansas City, MO 64111. E-mail emagnuson@saint-lukes.org
© 2012 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.111.064378

1858
 Venkitachalam et al
treatment efforts into healthcare policies.” According to
global estimates from WHO, nearly 1 of every 3 deaths in
2004 was attributed to cardiovascular disease with nearly
80% of these deaths occurring in low- and middle-income
countries.2 In an effort to guide the National Heart, Lung, and
Blood Institute in setting priorities for investment in global
cardiovascular health, the Institute of Medicine recently
outlined key barriers to and evidence-based solutions for the
control of cardiovascular disease across the world.3 An
improved understanding of the relationship between levels of
national economic development, healthcare investment,
health system characteristics, and key modifiable cardiovas-
cular risk factors may provide insights for the prioritization of
cardiovascular disease prevention programs.
Clinical Perspective on p 1869
Elevated serum cholesterol is a modifiable risk factor that
is associated with an estimated 4.4 million deaths each year
and accounts for a considerable proportion of ischemic
strokes and heart disease worldwide.4 Therapeutic lifestyle
changes (reduced dietary intake of saturated fats and choles-
terol, weight control, and increased physical activity) form
the core of all cholesterol-lowering initiatives. Supplemental
therapy with lipid-lowering medications has been shown to
safely reduce the long-term incidence of major cardiovascular
events in secondary prevention and, more recently, high-risk
primary prevention trials, and is universally recommended in
patients with established or at high predicted risk of cardio-
vascular disease.5,6 Numerous reports have documented the
Downloaded from http://ahajournals.org by on December 31, 2019
underrecognition and undertreatment of elevated cholesterol
levels in developed and developing countries.7–11 Whereas most
studies have examined the patient- and physician-level factors
responsible for these variations,9,12,13,14 associations with
country-level factors have not been systematically explored.
A primary objective of the international Reduction of
Atherothrombosis for Continued Health (REACH) Regis-
try9,15 was to describe the global distribution and management
of risk factors for atherothrombotic events across the spec-
trum of patients with established atherosclerotic vascular
disease. As such, REACH provides a unique source for
gaining insight into country-level factors that might influence
effective cardiovascular risk factor control. We used data
from REACH to examine the relationship between estab-
lished national indices of overall health system expenditure
and performance and country-specific prevalence rates of
elevated total cholesterol at baseline. Our analyses allowed
the effect of country-level factors to differ for patients with
versus without a history of hyperlipidemia, because factors
influencing the prevalence of elevated cholesterol may differ
between these 2 populations.
Methods
Study Population
The rationale, design, and baseline characteristics of the REACH
Registry have been previously published.9,15 In brief, REACH is an
international, prospective registry of 68 236 outpatients, aged 45
years or older, at risk of atherothrombosis due to established
coronary artery disease (CAD), cerebrovascular disease (CVD), or
peripheral arterial disease (PAD), or ⱖ3 cardiovascular risk factors.
Patients from 39 countries were enrolled between December 2003 to
Elevated Cholesterol and National Indices 1859
2004. Presence of established CAD, CVD, and PAD was ascertained
by physicians using well-defined criteria (see online-only Data
Supplement Methods for definitions).9 Patients without established
CAD, CVD, or PAD could be enrolled given the presence of 3 or
more of the following risk factors: treated diabetes mellitus, diabetic
nephropathy, ankle-brachial index ⬍0.9, asymptomatic carotid ste-
nosis ⱖ70%, carotid intima media thickness that exceeds twice the
measure at neighboring sites, systolic blood pressure ⱖ150 mm Hg
despite therapy for at least 3 months, hypercholesterolemia treated
with medication, current smoking of at least 15 cigarettes per day,
men aged 65 years or older, or women aged 70 years or older.
Patients already in a clinical trial, hospitalized patients, or those who
might have difficulty returning for a follow-up visit were excluded
from enrolment in REACH.
With the goal of yielding a representative sample of patients at the
national level, the selection of sites was based on the best available
data regarding the burden of atherothrombosis in at-risk populations
within each country, and healthcare delivery settings most typically
used by those patients. Site characteristics considered in the site
selection process included overall patient profiles (eg, CAD, CVD,
or PAD, or primary prevention), physician profiles (general practi-
tioners, internists, cardiologists, neurologists, endocrinologists, vas-
cular surgeons), healthcare environments (rural, suburban, urban),
and medical practices (office-based, hospital-based). Data for pa-
tients enrolled in the registry were collected centrally via use of a
standardized international case report form, completed at the study
visit. The REACH protocol was submitted to the institutional review
board in each country according to local requirements, and signed
informed consent was obtained for all patients.
The analysis cohort was derived from the patients enrolled in
REACH. Of the 68 236 patients from 39 countries enrolled in
REACH, 1576 patients from 3 countries (Taiwan, Hong Kong, and
Panama), for which national health system and economic indices
were unavailable, were excluded from the current analysis. Of the
remaining 66 660 patients, 13 020 (20%) did not have total choles-
terol values (16% and 32% of patients with and without a history of
hyperlipidemia, respectively), and an additional 70 patients were
missing documentation regarding history of hypercholesterolemia.
The remaining 53 570 patients (80.4%) from 36 countries formed the
analytic cohort for this study.
Elevated Total Blood Cholesterol and History
of Hyperlipidemia
Total blood cholesterol levels available at the time of enrollment
(measurements must have been taken within the preceding 12
months) were recorded. For this analysis, elevated cholesterol was
defined as total cholesterol ⬎200 mg/dL (⬎5.18 mmol/L).5 History
of hyperlipidemia was defined as a documented past diagnosis of
hyperlipidemia or reported use of lipid-lowering therapy at study
enrollment.
Country-Level Indices
Information on national health expenditures for the year 2001 was
obtained from the 2004 World health report16 for the following
indices: total expenditure on health as percentage of gross domestic
product (GDP), general government expenditure on health as per-
centage of total expenditure on health, and out-of pocket expenditure
as percentage of private expenditure on health. For country-level
income, we used the 2003 gross national income (GNI) per capita
(Atlas method) as developed by the World Bank.17 For country-level
data pertaining to health system performance, we used 2 indices
developed by WHO: overall Health System Achievement (HAI), and
Health System Performance/Efficiency (HPEI).18 HAI is a weighted
linear aggregate of measures of health level, health inequality,
responsiveness, responsiveness inequality, and fairness of financial
contribution that ranges from 0 to 100, with 100 being the highest
possible level of achievement. HPEI is a ratio of the observed level
of population health (healthy life expectancy) to the maximum that
could be achieved with the observed resources (function of average
years of schooling and per capita health expenditure) and ranges
 1860 Circulation April 17, 2012

Table. Patient Characteristics Among Those With (First Entry) Versus Without (Second Entry) History of Hyperlipidemia,* by Region/Country
North America Latin America Eastern Europe

Region
Country Total USA Canada Brazil Chile Mexico Bulgaria Hungary Lithuania Romania Russia Ukraine
History of Hyperlipidemia* n⫽42 955 n⫽19 257 n⫽1449 n⫽263 n⫽156 n⫽478 n⫽555 n⫽679 n⫽70 n⫽1286 n⫽552 n⫽252
No History of Hyperlipidemia* n⫽10 615 n⫽2540 n⫽166 n⫽92 n⫽48 n⫽148 n⫽356 n⫽173 n⫽18 n⫽571 n⫽400 n⫽286
Age, y‡
Mean 68.0 69.4 67.9 65.0 65.8 66.2 63.2 64.5 63.6 62.3 60.6 60.4
69.8 73.0 70.2 66.3 69.7 69.9 64.9 67.9 62.7 65.2 63.2 62.9
SD 10.0 10.1 9.7 9.6 9.1 10.2 8.9 9.6 8.2 9.1 8.3 9.1
10.2 10.5 10.2 9.8 9.6 9.8 8.8 9.5 9.6 9.6 8.9 10.0
Men, %‡ 63.4 58 69.7 60.1 73.7 64.2 62.9 59.6 61.4 64.0 76.8 78.6
64.9 52.9 66.9 59.8 58.3 66.2 66.0 62.4 94.4 64.4 67.0 68.2
College or technical 33.8 35.2 30.2 30 39.1 46.2 69.9 41.8 72.9 55.4 90.8 77.4
education, %‡ 30.5 29.2 27.7 16.3 27.1 36.5 61.5 30.1 72.2 44.5 75.3 67.1
History of hypertension, %‡ 83.8 88.3 77.4 79.8 80.8 72.6 91.4 87.0 92.9 79.5 83.9 80.6
78.6 85.8 75.3 81.5 85.4 74.3 93.3 85.0 72.2 82.8 82.8 78.7
History of diabetes, %‡ 47.1 53.4 46.7 42.6 41.0 48.7 27.2 43.3 17.1 29.2 20.1 24.2
42.2 50.4 47.0 45.7 60.4 45.9 34.8 46.8 16.7 24.5 18.8 18.2
BMI ⱖ30 kg/m2, %‡ 32.8 42.3 33.1 27.4 21.2 27.2 28.3 32.3 38.6 27.4 32.1 38.1
23.0 35.7 34.9 17.4 18.8 25.0 24.4 31.8 22.2 21.9 32.3 35.3
Current smoking, %‡ 14.2 13.8 14.9 5.7 9.0 7.3 27.2 19.6 10.0 17.4 22.8 15.5
17.1 15.2 16.3 13.0 2.1 12.2 23.9 19.7 5.6 17.2 27.3 19.9
Documented vascular 78.9 71.9 81.2 86.3 85.9 84.9 89.4 90.6 98.6 96.1 95.5 97.6
disease at baseline, %‡ 86.0 78.6 80.1 89.1 97.9 87.2 94.7 92.5 100.0 98.1 97.5 97.9
Downloaded from http://ahajournals.org by on December 31, 2019

CAD 63.1 60.2 69.9 76.0 67.3 60.5 79.8 67.6 95.7 72.6 90.9 89.3
46.8 50.0 55.4 43.5 58.3 37.8 58.4 43.9 100.0 49.7 80.0 73.8
CVD 22.1 18.3 17.1 14.8 23.1 27.4 24.7 30 25.7 42.5 13.9 20.6
43.0 34.0 31.3 32.6 33.3 52.7 54.8 53.8 0.0 75.0 30.0 29.0
PAD 10.9 8.8 7.4 6.8 14.7 7.9 8.3 17.7 1.4 12.5 11.8 5.2
13.7 10.3 9.6 22.8 25.0 13.5 15.7 17.9 0.0 11.7 13.5 11.2
Cholesterol-related
characteristics
Total cholesterol at
baseline, mg/dL†‡
Mean 191.7 180.8 174.1 197.1 199.4 197.6 245.1 215.8 227.7 228.1 219.5 227.9
196.6 187.0 191.5 187.7 189.4 185.7 214.8 199.7 202.0 197.2 221.5 214.5
SD 48.0 43.8 39.9 53.5 49.9 49.8 53.4 51.1 44.8 55.1 49.9 51.4
41.1 39.1 39.0 38.2 35.3 39.5 45.8 39.3 43.9 41.8 51.1 45.7
TC ⬎200 mg/dL, %‡ 37.0 26.9 23.7 41.4 43.6 47.1 81.1 63.3 72.9 67.6 62.9 26.9
43.0 41.3 35.5 27.2 29.2 30.4 59.3 46.8 44.4 45.5 65.8 32.6
Among patients with history
of hyperlipidemia*
Past diagnosis of 95 97.3 97.2 89.7 94.2 90.4 96.6 92.8 98.6 83.7 92.4 69.4
hyperlipidemia, %
On lipid-lowering therapy 98.6 98.2 98.4 99.6 100 98.5 96 99.3 98.6 99.6 99.1 100
at baseline, %
Statins 91 90.3 93.6 94.3 93.6 93.3 81.8 92.9 87.1 92.8 95.8 96
Others 16.4 22.6 9.7 8.4 10.9 18.4 15.5 10.9 28.6 11.4 5.4 4.8
UAE indicates United Arab Emirates; UK, United Kingdom; USA, United States of America; CAD, coronary artery disease; CVD, cerebrovascular disease; PAD,
peripheral arterial disease; TC, total cholesterol.
*History of hyperlipidemia was defined as a documented past diagnosis of hyperlipidemia or use of lipid-lowering medications before study entry at baseline.
†To convert mg/dL to mmol/L, please divide by 39.
‡Data provided in the second row relate to patients without history of hyperlipidemia.
 Venkitachalam et al Elevated Cholesterol and National Indices 1861

Table. Continued
Western Europe

Region
Country Austria Belgium Denmark Finland France Germany Greece Netherlands Portugal Spain Switzerland UK
History of Hyperlipidemia* n⫽993 n⫽211 n⫽250 n⫽170 n⫽2842 n⫽3454 n⫽485 n⫽185 n⫽169 n⫽1484 n⫽464 n⫽364
No History of Hyperlipidemia* n⫽272 n⫽94 n⫽79 n⫽37 n⫽522 n⫽890 n⫽105 n⫽36 n⫽27 n⫽498 n⫽103 n⫽46
Age, y‡
Mean 67.7 69.8 65.1 65.5 68.7 67.4 67.7 64.2 67.6 66.6 68.1 68.6
70.7 72.9 68.8 69.0 71.3 70.5 70.8 68.3 67.6 70.9 73.3 70.3
SD 9.5 9.3 9.4 9.0 9.7 8.9 9.1 8.9 9.0 9.8 9.5 8.3
10.0 9.2 11.5 9.4 9.7 9.2 8.3 11.0 8.9 9.7 9.0 8.5
Men, %‡ 64.5 71.6 71.2 71.8 73.0 69.1 68.2 74.1 68.6 72.8 72.8 62.9
60.3 63.8 73.4 45.9 73.0 61.6 70.5 61.1 77.8 68.5 69.9 67.4
College/technical 26.5 52.1 28.8 37.6 25.0 18.9 27.2 22.2 26.0 20.1 33.0 30.8
education, %‡ 21.0 48.9 12.7 43.2 17.4 14.4 21.9 19.4 11.1 12.4 36.9 41.3
History of hypertension, %‡ 86.0 80.6 52.0 61.8 77.0 90.9 87.2 60.3 80.5 71.8 84.5 69.2
84.2 76.6 40.5 59.5 77.4 85.4 83.8 55.6 81.5 67.9 78.6 58.7
History of diabetes, %‡ 39.5 33.2 16.0 28.2 40.2 45.8 52.4 23.8 42.0 43.2 32.8 23.4
45.6 30.9 10.1 18.9 48.7 52.0 39.0 11.1 40.7 45.8 43.7 23.9
BMIⱖ30 kg/m2, %‡ 30.6 27.5 22.4 24.1 27.3 29.4 30.1 23.2 26.6 33.2 29.7 31.6
29.8 28.7 19.0 18.9 28.7 27.6 20.0 11.1 22.2 27.9 28.2 19.6
Current smoking, %‡ 15.7 17.1 23.6 12.4 14.1 14.8 27.6 26.5 8.9 12.3 16.6 14.6
16.5 17.0 26.6 13.5 18.8 19.3 24.8 30.6 22.2 13.7 22.3 10.9
Documented vascular 83.1 83.9 99.2 84.7 73.6 89.83 69.7 94.6 68.6 85.6 84.9 92.0
disease at baseline, %‡ 74.6 93.6 100.0 86.5 71.8 82.5 85.7 94.4 74.1 87.8 81.6 95.7
Downloaded from http://ahajournals.org by on December 31, 2019

CAD 65.5 64.5 62.4 70.63 56.0 70.2 53.0 63.2 50.9 64.8 69.2 79.7
36.8 61.7 19.0 43.2 35.1 44.9 44.8 36.1 25.9 34.9 46.6 63.0
CVD 24.8 27.0 45.6 18.8 15.2 26.6 22.7 16.2 18.9 23.7 17.2 20.3
35.3 33.0 74.7 45.9 26.1 35.5 41.9 36.1 44.4 52.8 25.2 28.3
PAD 14.2 15.2 10.4 4.1 17.7 22.8 8.9 40.0 7.7 11.3 16.6 7.7
19.5 21.3 13.9 2.7 24.3 26.3 18.1 58.3 11.1 12.4 25.2 8.7
Cholesterol-related
characteristics
Total cholesterol at
baseline, mg/dL†‡
Mean 206.8 203.1 195.8 173.9 196.9 204.2 226.5 195.8 204.9 198 193.1 182.8
204.2 209.5 195 187.8 200.6 212.5 211.6 208.1 190.9 191.8 202.6 192.8
SD 46.9 48.1 46.6 37.9 42.7 52.5 49.3 55.9 43.3 45.7 42.7 44.6
36.7 21.1 45.6 29.6 36.1 48.4 28.5 28.6 27.6 39 38.4 48.7
TC ⬎200 mg/dL, %‡ 52.0 50.7 42.4 22.9 41.2 48.2 67.6 35.7 50.9 43.2 40.5 30.0
46.7 60.6 45.6 29.7 46.9 61.6 65.7 55.6 40.7 40.0 43.4 10.5
Among patients with
history of hyperlipidemia*
Past diagnosis of 97.0 93.4 94.0 97.6 94.5 95.8 97.1 91.9 94.1 90.2 95.7 94.2
hyperlipidemia, %
On lipid-lowering 100 96.7 98.8 100.0 98.8 97.4 100.0 98.4 97.0 97.3 99.4 99.2
therapy at baseline, %
Statins 93.9 86.3 98.8 98.8 84.9 92.2 93.6 95.7 91.7 94.6 97.2 95.6
Others 9.2 11.8 1.6 3.5 15.6 11.0 8.2 6.6 8.3 5.9 6.5 6.6
 1862 Circulation April 17, 2012

Table. Continued
Middle East Australasia

Region Saudi South

Country Israel Lebanon Arabia UAE Australia China Indonesia Japan Malaysia Philippines Singapore Korea Thailand
History of Hyperlipidemia* n⫽304 n⫽86 n⫽167 n⫽100 n⫽1326 n⫽328 n⫽302 n⫽2205 n⫽361 n⫽520 n⫽568 n⫽235 n⫽385
No History of Hyperlipidemia* n⫽41 n⫽19 n⫽18 n⫽19 n⫽150 n⫽236 n⫽124 n⫽1947 n⫽43 n⫽148 n⫽70 n⫽213 n⫽80
Age, y‡
Mean 69.7 63.5 60.5 62.5 71.8 65.5 62.5 69.6 61.7 64.3 63.3 64.0 66.9
73.2 70.1 66.9 68.0 73.4 65.7 62.1 71.0 64.8 66.8 65.4 64.6 69.4
SD 9.4 10.0 9.5 9.8 8.6 8.9 9.8 8.6 9.2 9.8 9.5 8.5 9.9
8.8 8.6 10.5 11.9 8.8 9.3 10.0 8.8 10.4 11.3 10.7 9.0 8.8
Men, %‡ 71.4 68.6 74.7 68.0 66.4 73.8 63.2 61.9 77.3 48.8 64.6 65.1 63.1
73.2 78.9 83.3 84.2 62.7 66.9 80.6 76.7 86.0 50.0 64.3 69.0 60.0
College/technical 35.5 37.2 29.3 10.0 34.3 36.6 43.4 25.0 27.1 63.7 18.7 35.7 28.3
education, %‡ 31.7 26.3 22.2 15.8 32.0 31.4 48.4 23.4 20.9 50.7 11.4 27.7 23.8
History of 83.5 77.9 80.2 76.0 78.7 79.9 72.5 72.9 83.1 87.7 82.0 78.7 81.6
hypertension, %‡ 92.7 78.9 66.7 63.2 67.7 70.8 71.0 69.3 74.4 83.1 70.0 68.1 75.0
History of 42.4 57.0 67.7 61.0 36.5 28.0 50.3 49.3 59.3 51.0 60.6 60.0 54.5
diabetes, %‡ 51.2 63.2 72.2 26.3 31.3 19.9 46.0 41.3 53.5 43.9 57.1 46.5 56.3
BMIⱖ30 kg/m2, %‡ 30.9 19.8 35.9 31.0 29.6 4.0 9.6 4.9 16.6 9.2 12.1 3.8 4.9
34.1 26.3 27.8 15.8 28.0 4.7 4.8 3.0 9.3 7.4 10.0 5.6 5.0
Current smoking, %‡ 13.2 26.7 18.6 4.0 6.6 12.5 4.6. 13.7 10.2 6.2 15.7 20.9 5.7
17.1 15.8 5.6 5.3 8.0 14.8 11.3 18.1 11.6 8.8 12.9 14.6 6.3
Documented vascular 79.9 72.1 93.4 92.0 86.9 97.0 84.4 79.2 86.1 83.5 75.2 80.0 85.2
disease at baseline, %‡ 75.6 68.4 100.0 94.7 92.0 98.7 89.5 89.0 88.4 89.9 80.0 85.4 92.5
Downloaded from http://ahajournals.org by on December 31, 2019

CAD 72.7 50.0 76.0 73.0 75.4 73.2 57.0 51.3 70.4 45.8 51.2 50.2 61.0
46.3 52.6 61.1 68.4 62.0 33.9 43.5 40.8 46.5 30.4 34.3 37.1 41.3
CVD 17.8 27.9 25.7 26.0 18.6 40.9 40.1 30.7 21.1 43.8 31.7 38.3 27.8
31.7 21.1 61.1 42.1 37.3 80.1 56.5 47.8 32.6 69.6 42.9 54.0 45.0
PAD 5.6 9.3 3.0 5.0 8.0 2.1 8.3 9.7 6.4 4.2 5.3 3.4 6.0
7.3 21.1 0.0 10.5 7.3 3.4 5.6 11.9 27.9 3.4 5.7 2.3 17.5
Cholesterol-related
characteristics
Total cholesterol at
baseline, mg/dL†‡
Mean 183.4 214.7 199.2 182.2 175.6 196.7 221.0 195.9 196.2 215.5 192.5 190.2 191.4
183.2 188.9 190.0 171.5 201.8 191.4 189.1 193.9 180.9 188.5 191.1 184.7 183.9
SD 37.2 47.3 58.8 49.1 36.3 54.1 56.2 35.8 46.1 57.0 44.3 52.2 48.5
29.8 27.2 48.6 34.8 40.1 39.7 40.2 33.4 31.0 44.1 33.9 35.2 30.2
TC⬎200 mg/dL, %‡ 26.0 55.8 38.9 70.6 21.9 44.2 64.2 40.9 43.8 56.0 38.2 32.3 37.9
31.7 36.8 33.3 61.5 46.0 37.7 28.2 42.3 27.9 33.1 32.9 27.7 28.8
Among patients with
history of hyperlipidemia*
Past diagnosis of 98 90.7 96.4 92 96.8 73.8 89.4 91.9 95 93.3 96.7 80 96.1
hyperlipidemia, %
On lipid-lowering 98 100 98.2 100 99.8 100 100 100 100 99.6 99.6 99.1 99.5
therapy at baseline, %
Statins 96.7 84.9 96.4 99 98 85.7 92.4 87.2 94.7 93.3 91.4 84.3 91.4
Others 8.9 20.9 6.6 4 3.5 19.5 16.2 19 9.1 8.8 17.6 18.7 13.8
 Venkitachalam et al
from 0 to 1 (fully efficient). Estimates of average country-level total
fat consumption (grams per person per day) were obtained from the
Food and Agricultural Organization for the year 2003 and were
available for all countries with the exception of Singapore.19
Statistical Analysis
Baseline characteristics of participants in REACH according to
history of hyperlipidemia, both overall and by country, are summa-
rized as means and standard deviations for continuous variables, and
percentages for categorical variables. To estimate the proportion of
total variability in elevated cholesterol (⬎200 mg/dL) that is at the
country level, we obtained the intraclass correlation coefficient from
a hierarchical logistic regression model in which the individual
patient is considered level 1 and country is considered level 2.20 This
analysis was performed for the overall study cohort, and separately
for patients with and without history of hyperlipidemia, as well.
Cubic smoothing spline models were used to estimate smooth
curves describing the associations between country-level health
system and economic indices (continuous variables) and country-
specific prevalence of elevated cholesterol. Hierarchical generalized
mixed-effects (logistic) regression models, with individual patient at
the first level and country (treated as a random effect) at the second
level, were used to estimate the association between the country-
level economic/health system indices and odds of elevated choles-
terol. We explored the use of site as a third possible level, but the
large number of sites (n⫽5267) and relatively low number of
patients per site (mean, 10.2, range,1–20) prohibited model conver-
gence. For these analyses, the national indices of health-related
expenditures and performance were categorized into low, medium,
and high levels, approximately according to tertiles (see online-only
Data Supplement Figure IA through IF). For GNI per capita,
previously established cut points reflecting variation in GNI across
countries were used.21 These analyses were performed on the overall
analytic population, with the inclusion of interaction terms between
Downloaded from http://ahajournals.org by on December 31, 2019
the country-level indices and the indicator for history of hyperlipid-
emia. Two models were fit for each of the country-level indices, with
elevated total cholesterol (⬎200 mg/dL) as the outcome measure.
The first model adjusted for patient-level factors, assessed at the time
of enrollment, including age, sex, educational level, smoking status,
obesity (BMI ⱖ30 kg/m2), established CAD, CVD, or PAD, history
of congestive heart failure, carotid surgery, atrial fibrillation/flutter,
aortic valve stenosis, diabetes mellitus, and treated hypertension, and
current use of antiplatelet and antihypertensive medications; covari-
ates significant at P⬍0.20 were retained in the model. In the second
model, we further adjusted for average country-level fat consump-
tion (as continuous variable)19 to account for the potential influence
of regional variation in dietary patterns on cholesterol levels. An
interaction between fat consumption and history of hyperlipidemia
was also included, and eliminated if nonsignificant (P⬎0.05). From
each of the final models, the relative impact of different levels of
each health system/economic index on elevated cholesterol was
estimated separately for history versus no history of hyperlipidemia
subgroups, and associated odds ratios and corresponding 95%
confidence limits were generated. All analyses were performed by
use of SAS version 9.2 statistical software (SAS Institute Inc, Cary,
NC). P values of ⱕ0.05were considered statistically significant.
Results
In the overall analytic cohort, 20 469 (38%) patients had
baseline total cholesterol ⬎200 mg/dL. The prevalence of
elevated cholesterol varied widely across countries and
ranged from 73% in Bulgaria to 24% in Finland. Among
42 955 patients with a history of hyperlipidemia, 37% had
elevated cholesterol reported at baseline; among the remain-
ing 10 615 patients without history of hyperlipidemia, 43%
had elevated cholesterol (Table). Overall, 9.3% of the total
variability in elevated cholesterol was at the country level.
Elevated Cholesterol and National Indices 1863
This proportion was higher for patients with (12.1%) versus
without (7.4%) history of hyperlipidemia.
Figure 1A through 1F show scatter plots of the country-level
economic/health system indices versus percentage of patients
with elevated cholesterol, stratified by history of hyperlipidemia,
respectively. For patients with history of hyperlipidemia,
country-specific elevated cholesterol percentages tended to in-
crease with decreasing total expenditure on health as percentage
of GDP spent on health, GNI, and WHO HAI and PEI indices,
and increasing levels of out-of-pocket expenditures. For patients
without history of hyperlipidemia, these graphs suggest some-
what different associations, with trends toward high prevalence
of elevated cholesterol with both increasing percentage of GDP
spent on health and increasing percentage of governmental
expenditure as a function of total expenditure on health.
A high prevalence of elevated cholesterol was observed in
patients from eastern European countries (Bulgaria. Lithua-
nia, Romania, Ukraine, Hungary, and Russia); these countries
also ranked relatively low on health system/economic indices.
A relatively low prevalence of elevated cholesterol, similar to
that for the United States, was seen for countries (eg, Finland,
United Kingdom, Israel, Australia, and Canada) with consid-
erably lower total expenditure on health as percentage of
GDP and comparable or slightly more favorable WHO health
system indices.
Figure 2A through 2F presents odds ratios and associated
95% confidence intervals for the relative effect of different
levels of the country-level indices on the risk of elevated
cholesterol, for patients with (top panel) and without (bottom
panel) history of hyperlipidemia at baseline. Within each
panel, results from 2 models (adjusted for patient-level
factors only [blue] and adjusted for both patient-level factors
and country-level fat consumption [red]) are presented. Fur-
ther details relating to the model results, including probability
values for the estimated effects of each of the country-level
indices are presented in online-only Data Supplement Table I.
For the history of hyperlipidemia subgroup, countries in the
highest tertile of GNI had significantly lower odds of elevated
cholesterol than countries in the lowest tertile (P⬍0.001).
Similar results were seen for the WHO HAI. Countries in
the lowest tertile of WHO HPEI had significantly higher
odds of elevated cholesterol than those in the intermediate
or highest tertile (P⬍0.001 for both). Countries in the
highest tertile of out-of-pocket health expenditures had
significantly higher odds of elevated cholesterol than
countries in the lowest tertile (P⬍0.001). No statistically
significant associations between country-level factors and
odds of elevated cholesterol were found for patients
without history of hyperlipidemia.
Higher fat consumption at the country level (average grams
per person per day) was associated with a significant increase
in the odds of elevated cholesterol in the model examining the
association with GNI. For none of the other models was the
estimated effect of fat consumption statistically significant,
and no significant interactions between fat consumption and
history of hyperlipidemia was found in any of the models. As
shown in Figure 2A, adjustment for country-level fat con-
sumption in addition to patient-level factors increased the
magnitude of association between GNI and the odds of
 1864 Circulation April 17, 2012
B vated cholesterol (total cholesterol ⬎200
C WHO, World Health Organization.
Downloaded from http://ahajournals.org by on December 31, 2019
elevated cholesterol. Estimated coefficients for all covariates
for both GNI models (with and without fat consumption) are
included in the online-only Data Supplement Table II. Esti-
mated coefficients describing the association between
patient-level covariates and the odds of elevated cholesterol
varied little across models for different country-level indices.
Discussion
The exponential rise in noncommunicable chronic diseases
over the past decade has placed a tremendous burden on the
health and economic development of countries worldwide,
with unprecedented demands for an effective response from
governments and other stakeholders in global health. An
understanding of how country-level economic and healthcare
indices are associated with disease-specific, modifiable risk
factors may be of potential value for strategic planning and
evaluation of related public health policies and programs. In
this study, we examined data from a large, multinational
registry of individuals with or at high risk of atherothrom-
botic events22,23 to provide insight into the relationship
Figure 1. Scatter plots of the country-level
indices of health system and economic
development versus percentage of ele-
mg/dL or 5.18 mmol/L), in patients with
(left side) and without (right side) history of
hyperlipidemia. *History of hyperlipidemia
was defined as a documented past diag-
nosis of hyperlipidemia or use of lipid-
lowering medications before study entry at
baseline. †Information on gross national
income per capita was not available for
United Arab Emirates for the year 2003.
Smooth lines describe the association
between the 2 variables derived from the
flexible cubic spline models. GDP indi-
cates gross domestic product; UAE,
United Arab Emirates; UK, United King-
dom; USA, United States of America;
between indices of national health systems and economic
development, and elevated total cholesterol, a key risk factor
for cardiovascular disease. The significant associations ob-
served at the country level between these national indices and
prevalence of elevated cholesterol underscore the importance
for countries to maintain, improve, or establish effective
surveillance of chronic disease risk factors such as cholesterol
levels, while also prioritizing population-based efforts aimed
at the prevention and management of chronic diseases.
Comparison With Previous Studies
Numerous reports have consistently demonstrated a substantial
gap between evidence-based guidelines and actual clinical prac-
tice across geographic regions, among different physician spe-
cialties, and independent of the arterial bed affected.7,9 –11,13,24
Data from the multinational WHO MONICA Project demon-
strated wide variations in the prevalence and treatment of
hypercholesterolemia between 1989 and 1997 across 32 popu-
lations from 19 countries.7 Subsequently, using data from
European national surveys conducted during 1995 and 2007,
 Venkitachalam et al
F
Downloaded from http://ahajournals.org by on December 31, 2019
Kotseva et al10 documented dramatic improvements in choles-
terol control over time; however, less than half of the patients on
lipid-lowering therapy reached the recommended target level. In
a recent study of ⬇80 000 adults from 8 high- and middle-
income countries, the proportion of “undiagnosed” individuals
with elevated cholesterol (ⱖ6.2 mmol/L or 240 mg/dL) varied
5-fold (16% in the United States to 78% in Thailand), whereas
the rates of effective cholesterol control among patients on
lipid-lowering medication ranged from 4% in Germany to 58%
in Mexico.11 In our study, which includes patients from primary
care and specialty practices, ⬇4 in 10 individuals had elevated
cholesterol at baseline; this proportion differed significantly
across countries, with country-level factors explaining an appre-
ciable proportion of variability in elevated cholesterol in the
overall cohort.
Recruitment of patients to the REACH Registry was
designed to yield a representative sample of patients, based
on the burden of atherothrombosis in at-risk populations
within each country and healthcare delivery settings most
typically used by those patients. A comparison of the preva-
Elevated Cholesterol and National Indices 1865
Figure 1 (Continued).
lence of elevated total cholesterol observed in 9 countries
included in both the REACH Registry and the EUROASPIRE
II survey25 revealed a high correlation between the 2 studies
(r⫽0.72; see online-only Data Supplement Figure II for
details).26 To the extent that the patterns of risk factors
observed for patients enrolled in REACH are representative
of national patterns for these high-risk populations, findings
from our study reveal the degree to which the prevalence of
elevated cholesterol at the country level may be related to
macroeconomic and healthcare system factors. Moreover, the
association between elevated cholesterol and GNI for patients
with a history of hypercholesterolemia is consistent with
patterns revealed in a recent analysis of data from the
multinational Prospective Urban and Rural Epidemiological
(PURE) study,26 in which a 20-fold difference in the use of
statins for secondary prevention between low- and high-
income countries was found.
The significant associations found between prevalence
rates of elevated cholesterol and total and out-of-pocket
health expenditures, and WHO indices, as well, for patients
 1866 Circulation April 17, 2012
A B
C D
E F
Downloaded from http://ahajournals.org by on December 31, 2019
with, but not without, a history of hyperlipidemia suggest
differential influence of these country-level factors in these
patient subsets. Although the exact reasons for these differ-
ences cannot be determined in this study, the sample of
patients without history of hyperlipidemia in REACH is small
(20% of the overall cohort) and likely comprises patients who
were either never screened or who had been screened but
were found not to be hyperlipidemic before enrollment in the
registry. The potential inclusion of the latter category of
patient may contribute to the lack of association between the
prevalence of elevated cholesterol and country-level indices
in this seemingly underdiagnosed subgroup. In contrast,
elevated cholesterol in patients with history of hyperlipidemia
may be due to prescription of suboptimal therapy (undertreat-
ment) or medication nonadherence related to availability or
affordability issues.
Global Challenges for Cholesterol Management
Optimal management of cardiovascular disease is complex
and country-level variations in risk factor control may arise
from variability in guidelines, and whether, and the extent to
Figure 2. A, Gross national income per
capita (high ⬎$9000; medium $2900 –
$9000; low ⬍$2900). B, Total expendi-
ture on health as percentage of GDP
(high ⬎8; medium 6 – 8; low ⬍6). C, Gov-
ernment expenditures as percentage of
total expenditure on health (high ⬎72;
medium 50 –72; low ⱕ50). D, Out-of-
pocket expenditure on health (high ⱖ90;
medium 70 –90; low ⬍70), E, WHO
health system achievement index (high
ⱖ90; medium 80 –90; low ⬍80). F, WHO
health system performance/efficiency
index (high ⬎0.85; medium 0.75– 0.85;
low ⱕ0.75). Odds ratios and associated
95% confidence intervals for the relative
effect of different levels of the country-
level indices on the risk of elevated cho-
lesterol, for patients with (top panel) and
without (bottom panel) history of hyper-
lipidemia. Within each panel, results from
2 models (adjusted for patient-level fac-
tors only [blue] and, adjusted for both
patient-level factors and country-level fat
consumption [red]) are presented. (Fat
consumption data are not available for
Singapore). *History of hyperlipidemia:
documented past diagnosis of hyperlip-
idemia or use of lipid-lowering medica-
tions before study entry; P value for
interaction term between global indices
and hyperlipidemia group indicator was
⬍0.001 for all indices, with the exception
of Government Expenditure as % of
Total Expenditure on Health (P value:
nonsignificant). GDP indicates gross
domestic product; WHO, World Health
Organization.
which, specific initiatives are effectively implemented, as
well. Ferket et al27 reviewed 27 professional guidelines for
cardiovascular risk assessment from Western countries and
documented marked differences in the definition of target
populations, specific screening tests used, and thresholds for
initiating pharmacological treatment. The specific cut point
for defining “desirable” cholesterol level varies as do recom-
mended strategies for subsequent risk assessment and man-
agement.5,6,28 In our study, based on the National Cholesterol
Education Program Adult Treatment Panel III guidelines for
risk assessment,5 we considered total cholesterol ⱖ200
mg/dL to be elevated. This cut point is, for instance, higher
than the thresholds used in guidelines from the European
Society of Cardiology6 and WHO.28 Such variability in the
threshold used to identify elevated cholesterol may explain
some of the variation in the prevalence rates of elevated
cholesterol across countries and, to the extent that it is
correlated with the country-level factors used in this analysis,
may underlie some of the associations found.
Wide variations in procurement prices across countries and
public versus private health sectors have also been demon-
 Venkitachalam et al
strated to impact use of recommended medications across
regions.29 –31 The association between the prevalence of
elevated cholesterol and out-of-pocket expenditure seen for
patients with history of hyperlipidemia in our study is
noteworthy, because it may reflect an inability or unwilling-
ness on the part of some patients in countries with higher
out-of-pocket healthcare expenses to be consistently compli-
ant with prescribed chronic lipid-lowering therapy. The
recent availability of generic statin therapy should make
out-of-pocket expense less of an impediment to use. Recent
results from the United States-based Post-Myocardial Infarc-
tion Free Rx and Economic Evaluation (MI-FREEE) trial
demonstrated that the elimination of copayments for medica-
tions prescribed after myocardial infarction significantly im-
proved adherence rates and rates of major cardiovascular
events, without increasing overall health costs, although, even
with full coverage, adherence rates remained far from opti-
mal, at 55% overall.32 Barriers to achieving adequate choles-
terol control at the population level may also be linked to
agricultural and economic policies that increase access to
low-cost, unhealthy diets.33 In our study, adjustment for
differences in country-level dietary fat consumption in-
creased the estimated magnitude of association between
elevated cholesterol and some country-level indices, suggest-
ing a complex interplay between access to diets rich in highly
saturated fats, national level of economic development, and
cholesterol levels.
Limitations
Downloaded from http://ahajournals.org by on December 31, 2019
Because this analysis was based on data from an observa-
tional registry, causal mechanisms underlying the associa-
tions described in this report cannot be determined. For
example, although the prevalence of elevated cholesterol
tended to increase with country-level out-of-of pocket health-
care expenditures, explicit recommendations for national
healthcare coverage should not be based on these data alone.
Nonetheless, a contemporary study of US outpatients with
established CAD found that patients’ lack of medical insur-
ance is significantly associated with statin underuse.14
The summary WHO indices of health system achievement
and performance examined in this analysis have drawn
criticism relating to the data on and methods by which they
were derived.34,35 In addition, these indices were derived at an
earlier time period than REACH Registry enrolment and may
have limited relevance in more contemporary settings. Nev-
ertheless, the observed relationship of these indices to objec-
tive disease-specific risk factors in high-risk patients suggests
their potential validity for assessing health system perfor-
mance with respect to effective cardiovascular disease pre-
vention and management.
By design, the REACH Registry recruited participants with
access to health care; as a result, the prevalence of elevated
cholesterol in this study may underestimate the overall
prevalence in the countries included in this study, and the
associations found may not be generalizable to individuals
without access to medical care. However, these patterns of
variation in statin use according to country-level income are
similar to those found in the PURE study,26 which used a
community-based sampling scheme.
Elevated Cholesterol and National Indices 1867
Total cholesterol measurements were unavailable for pro-
portionately more patients without a history of hyperlipid-
emia than patients with such history; this was the case for
almost all participating countries; and no evidence of system-
atic differences across countries with respect to other baseline
variables was apparent. The extent to which the lower sample
size for the no history of hyperlipidemia subgroup, and any
systematic bias associated with the availability of total
cholesterol measurements, accounts for the lack of significant
associations found for that group, cannot be determined. Our
models evaluating the association between country-level
indices and elevated total cholesterol, however, adjusted for
all patient-level risk factors. Other limitations include that
fact that information related to lipid-lowering therapy was
abstracted from medical records and may not reflect true
patient compliance. In addition, details related to dosing
strategies, contraindications to the use of, or adverse effects
from lipid-lowering medications were not collected in
REACH, nor were side effects that may have precluded
specific medication use. Finally, whereas other components
of the lipoprotein profile might be considered more optimal
measures of cardiovascular risk, only total cholesterol mea-
surements were recorded in the REACH Registry.
Implications
The above limitations notwithstanding, these results could
serve as impetus for continued national and international
dialogue and targeting of efforts toward improving cardio-
vascular risk factor control. In the past decade, several
countries have made progress in their efforts to strengthen
national capacity for prevention and control of noncommu-
nicable diseases; yet, much of these advancements are seen in
high-income countries.36 WHO has recently proposed a
framework of several cost-effective or low-cost population-
based strategies and individual-level interventions aimed at
reducing behavioral and physiological risk factors for chronic
diseases that may help reduce such disparities between
low/middle- and high-income countries.1 In recognition of the
extent to which patients, physicians, and healthcare systems
all influence adherence to preventive therapies, a framework
for the integration of healthcare system-based approaches
with community-based health promotion efforts within the
United States has also been proposed.37 In addition, as part of
the Million Hearts initiative, the US Centers for Medicare and
Medicaid Services has expanded coverage for CVD preven-
tion.38 Efforts such as these, in combination with ensuring
adequate infrastructure for surveillance of risk factors and
health outcomes, are crucial for controlling the rising, yet
preventable, burden of cardiovascular diseases and may serve
as a template for other countries to build upon.
Conclusions
Global variations in the prevalence of elevated cholesterol
among patients with history of hyperlipidemia are associated
with country-level economic and health expenditure indices,
and WHO indices of healthcare system achievement and
performance/efficiency, as well. These results support the
need for maintaining and strengthening healthcare system
efforts toward effective cardiovascular disease prevention
 1868 Circulation April 17, 2012
and control and may provide potential insights for strategic
public health policy settings at the national level.
Sources of Funding
The REACH Registry is sponsored by Sanofi-Aventis, Bristol-Myers
Squibb, and the Waksman Foundation (Tokyo, Japan). The REACH
Registry is endorsed by the World Heart Federation. Dr Venkitacha-
lam was supported by the American Heart Association Pharmaceu-
tical Roundtable–David and Stevie Spina Outcomes Research Post-
doctoral Fellowship at the Saint-Luke’s Mid America Heart Institute.
Disclosures
Dr Cohen received research grant support (significant) from Boston
Scientific, Abbott Vascular, Edwards Lifesciences, Merck/Schering
Plough, Daiichi-Sankyo, St. Jude medical, and Astra Zeneca. He
received consulting income (modest)from Medtronic and speaking
honoraria (modest) from Eli Lilly/Daiichi-Sankyo. Dr Hirsch re-
ceived consulting fees (modest) from Bristol-Myers Squibb/Sanofi
Aventis partnership, and honorarium (modest) from Bristol-Myers
Squibb/Sanofi Aventis partnership. Dr Ohman received research
grants (significant) from Bristol-Myers Squibb, CV Therapeutics,
Daiichi Sankyo, Datascope, Eli Lilly, Marquet, Sanofi-Aventis,
Schering-Plough, and The Medicines Company; and received fees
for consulting or other services (modest) for Abiomed, AstraZeneca,
CV Therapeutics, Datascope, Gilead Sciences, Liposcience, Mar-
quet, Northpoint Domain, Pozen, Response Biomedical, Sanofi-
Aventis, The Medicines Company, and WebMD (theheart.org). Dr
Steg received research grants (significant to his institution) from
Servier; he was a member of consultancy or advisory boards
(modest) for Astellas, AstraZeneca, Bayer, Boehringer-Ingelheim,
BMS, Daiichi-Sankyo-Lilly, GlaxoSmithKline, Medtronic, Otsuka,
Pfizer, Roche, Sanofi-Aventis, Servier, and The Medicines Com-
pany; and he is a stockholder in Aterovax. Dr Bhatt received research
grants (significant) from Amarin, AstraZeneca, Bristol-Myers
Downloaded from http://ahajournals.org by on December 31, 2019
Squibb, Eisai, Ethicon, Medtronic , Sanofi Aventis, and The Medi-
cines Company. Dr Magnuson received research grants (significant)
and honoraria (modest) from Sanofi-Aventis/Bristol-Myers Squibb
partnership, and research grants (significant) from Eli Lilly, Astra
Zeneca, and Daiichi-Sankyo. Drs Venkitachalam, Wang, Porath,
Corbalan, and Smith had no disclosures.
References
1. World Health Organization (WHO). 2008–2013 Action Plan for the Global
Strategy for the Prevention and Control of Noncommunicable Diseases:
Prevent and Control Cardiovascular Diseases, Cancers, Chronic Respi-
ratory Diseases and Diabetes. Geneva, Switzerland: WHO; 2008. http://
whqlibdoc.who.int/publications/2009/9789241597418_eng.pdf. Accessed
April 13, 2011.
2. World Health Organization (WHO). Cardiovascular Diseases (CVDs).
Fact Sheet No. 317. Geneva, Switzerland: WHO; 2011. http://www.who.
int/mediacentre/factsheets/fs317/en/print.html. Accessed April 13, 2011.
3. Fuster V, Kelly BB, Vedanthan R. Promoting global cardiovascular
health: moving forward. Circulation. 2011;123:1671–1678.
4. World Health Organization (WHO). The World Health Report 2002:
Reducing Risks, Promoting Healthy Life. Geneva, Switzerland: WHO;
2002. http://www.who.int/whr/2002/en/whr02_en.pdf. Accessed April
12, 2011.
5. National Cholesterol Education Program (NCEP) Expert Panel on
Detection Evaluation and Treatment of High Blood Cholesterol in Adults
(Adult Treatment Panel III). 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) final report. Circulation. 2002;106:3143–3421.
6. De Backer G, Ambrosioni E, Borch-Johnsen K, Brotons C, Cifkova R,
Dallongeville J, Ebrahim S, Faergeman O, Graham I, Mancia G, Cats VM,
Orth-Gomer K, Perk J, Pyorala K, Rodicio J, Sans S, Sansoy V, Sechtem U,
Silber S, Thomsen T, Wood D, Albus C, Bages N, Burell G, Conroy R,
Christian Deter H, Hermann-Lingen C, Humphries S, Fitzgerald A, Old-
enburg B, Schneiderman N, Uutela A, Williams R, Yarnell J, Priori SG,
Angeles Alonso Garcia M, Blanc J-J, Budaj A, Cowie M, Dean V, Deckers
J, Fernández Burgos E, Lekakis J, Lindahl B, Mazzotta G, McGregor K,
Morais Jo, Oto A, Smiseth O, Trappe H-J, Budaj A, Agardh C-D, Bassand
J-P, Deckers J, Godycki-Cwirko M, Heagerty A, Heine R, Home P, Priori S,
Puska P, Rayner M, Rosengren A, Sammut M, Shepherd J, Siegrist J,
Simoons M, Tendera M, Zanchetti A. European guidelines on cardiovascular
disease prevention in clinical practice. Eur Heart J. 2003;24:1601–1610.
7. Tolonen H, Keil U, Ferrario M, Evans A; WHO MONICA Project.
Prevalence, awareness and treatment of hypercholesterolaemia in 32
populations: results from the WHO MONICA Project. Int J Epidemiol.
2005;34:181–192.
8. Wood DA. Clinical reality of coronary prevention guidelines: a com-
parison of EUROASPIRE I and II in nine countries. Lancet. 2001;357:
995–1001.
9. Bhatt DL, Steg PG, Ohman EM, Hirsch AT, Ikeda Y, Mas JL, Goto S,
Liau CS, Richard AJ, Rother J, Wilson PW. International prevalence,
recognition, and treatment of cardiovascular risk factors in outpatients
with atherothrombosis. JAMA. 2006;295:180 –189.
10. Kotseva K, Wood D, De Backer G, De Bacquer D, Pyörälä K, Keil U.
Cardiovascular prevention guidelines in daily practice: a comparison of
EUROASPIRE I, II, and III surveys in eight European countries. Lancet.
2009;373:929 –940.
11. Roth GA, Fihn SD, Mokdad AH, Aekplakorn W, Hasegawa T, Lim SS.
High total serum cholesterol, medication coverage and therapeutic con-
trol: an analysis of national health examination survey data from eight
countries. Bull World Health Organ. 2011;89:92–101.
12. Cacoub PP, Abola MTB, Baumgartner I, Bhatt DL, Creager MA, Liau
C-S, Goto S, Rãther J, Steg PG, Hirsch AT. Cardiovascular risk factor
control and outcomes in peripheral artery disease patients in the
Reduction of Atherothrombosis for Continued Health (REACH) Registry.
Atherosclerosis. 2009;204:e86 – e92.
13. Kumar A, Fonarow GC, Eagle KA, Hirsch AT, Califf RM, Alberts MJ,
Boden WE, Steg PG, Shao M, Bhatt DL, Cannon CP, on behalf of the
REACH Investigators. Regional and practice variation in adherence to
guideline recommendations for secondary and primary prevention among
outpatients with atherothrombosis or risk factors in the United States: a
report from the REACH Registry. Crit Pathw Cardiol. 2009;8:104 –111.
14. Arnold SV, Spertus JA, Tang F, Krumholz HM, Borden WB, Farmer SA,
Ting HH, Chan PS. Statin use in outpatients with obstructive coronary
artery disease. Circulation. 2011;124:2405–2410.
15. Ohman EM, Bhatt DL, Steg PG, Goto S, Hirsch AT, Liau C-S, Mas J-L,
Richard A-J, Röther J, Wilson PWF. The REduction of Atherothrombosis
for Continued Health (REACH) Registry: an international, prospective,
observational investigation in subjects at risk for atherothrombotic
events-study design. Am Heart J. 2006;151:786.e1– e10.
16. World Health Organization (WHO). The World Health Report 2004:
Changing History. Geneva, Switzerland: WHO; 2004. Accessed April 12,
2011.
17. World Bank. September 2004. World Development Indicators. http://data.
worldbank.org/indicator. Accessed April 21, 2011.
18. World Health Organization (WHO). The World Health Report 2000:
Health Systems: Improving Performance. Geneva, Switzerland: WHO;
2000. http://www.who.int/whr/2000/en/whr00_en.pdf. Accessed May 26,
2011.
19. Food and Agricultural Organization of the United Nations. http://faostat.
fao.org/site/345/default.aspx. Accessed April 21, 2009.
20. Hedekar D. Multilevel models for ordinal and nominal variables. In: de
Leeuw J, Meijer E, eds. Handbook of Multilevel Analysis. New York, NY:
Springer; 2008:239 –275.
21. Orlandini A, Diaz R, Wojdyla D, Pieper K, Van de Werf F, Granger CB,
Harrington RA, Boersma E, Califf RM, Armstrong P, White H, Simes J,
Paolasso E. Outcomes of patients in clinical trials with ST-segment
elevation myocardial infarction among countries with different gross
national incomes. Eur Heart J. 2006;27:527–533.
22. Steg PG, Bhatt DL, Wilson PWF, D’Agostino R Sr, Ohman EM, Rother
J, Liau C-S, Hirsch AT, Mas J-L, Ikeda Y, Pencina MJ, Goto S, for the
REACH Registry Investigators. One-year cardiovascular event rates in
outpatients with atherothrombosis. JAMA. 2007;297:1197–1206.
23. Alberts MJ, Bhatt DL, Mas JL, Ohman EM, Hirsch AT, Rother J, Salette
G, Goto S, Smith SC Jr, Liau CS, Wilson PW, Steg PG. Three-year
follow-up and event rates in the international REduction of Athero-
thrombosis for Continued Health Registry. Eur Heart J. 2009;30:
2318 –2326.
24. Cacoub PP, Zeymer U, Limbourg T, Baumgartner I, Poldermans D,
Rother J, Bhatt DL, Steg PG. Effects of adherence to guidelines for the
control of major cardiovascular risk factors on outcomes in the REduction
 Venkitachalam et al Elevated Cholesterol and National Indices 1869

of Atherothrombosis for Continued Health (REACH) Registry Europe.
Heart. 2011;97:660 – 667.
25. Kotseva K, Stagmo M, De Bacquer D, De Backer G, Wood D. Treatment
potential for cholesterol management in patients with coronary heart
disease in 15 European countries: findings from the EUROASPIRE II
survey. Atherosclerosis. 2008;197:710 –717.
26. Yusuf S, Islam S, Chow CK, Rangarajan S, Dagenais G, Diaz R, Gupta
R, Kelishadi R, Iqbal R, Avezum A, Kruger A, Kutty R, Lanas F, Lisheng
L, Wei L, Lopez-Jaramillo P, Oguz A, Rahman O, Swidan H, Yusoff K,
Zatonski W, Rosengren A, Teo KK. Use of secondary prevention drugs
for cardiovascular disease in the community in high-income, middle-
income, and low-income countries (the PURE Study): a prospective
epidemiological survey. Lancet. 2011;378:1231–1243.
27. Ferket BS, Colkesen EB, Visser JJ, Spronk S, Kraaijenhagen RA, Steyerberg
EW, Hunink MGM. Systematic review of guidelines on cardiovascular risk
assessment: which recommendations should clinicians follow for a cardio-
vascular health check? Arch Intern Med. 2011;170:27–40.
28. World Health Organization (WHO). Prevention of Cardiovascular Disease:
Guidelines for Assessment and Management of Total Cardiovascular Risk.
Geneva, Switzerland: WHO; 2007. http://www.who.int/cardiovascular_
diseases/guidelines/Full%20text.pdf. Accessed November 7, 2011.
29. Mendis S, Fukino K, Cameron A, Laing R, Filipe A, Khatib O, Leowski
J, Ewen M. The availability and affordability of selected essential med-
icines for chronic diseases in six low- and middle-income countries. Bull
World Health Organ. 2007;85:279 –288.
30. Cameron A, Ewen M, Ross-Degnan D, Ball D, Laing R. Medicine prices,
availability, and affordability in 36 developing and middle-income coun-
tries: a secondary analysis. Lancet. 2009;373:240 –249.
31. van Mourik M, Cameron A, Ewen M, Laing R. Availability, price and
affordability of cardiovascular medicines: a comparison across 36
countries using WHO/HAI data. BMC Cardiovasc Disord. 2010;10:25.
32. Choudhry NK, Avorn J, Glynn RJ, Antman EM, Schneeweiss S, Toscano
M, Reisman L, Fernandes J, Spettell C, Lee JL, Levin R, Brennan T,
Shrank WH. Full coverage for preventive medications after myocardial
infarction. N Engl J Med. 2011;365:2088 –2097.
33. Elinder LSf. Obesity, hunger, and agriculture: the damaging role of
subsidies. BMJ. 2005;331:1333–1336.
34. Almeida C, Braveman P, Gold MR, Szwarcwald CL, Ribeiro JM,
Miglionico A, Millar JS, Porto S, Costa NdR, Rubio VO, Segall M,
Starfield B, Travassos C, Uga A, Valente J, Viacava F. Methodological
concerns and recommendations on policy consequences of the World
Health Report 2000. Lancet. 2001;357:1692–1697.
35. Richardson J, Wildman J, Robertson IK. A critique of the World Health
Organisation’s evaluation of health system performance. Health Econ.
2003;12:355–366.
36. World Health Organization (WHO). Global Status Report on Noncom-
municable Diseases 2010. Geneva, Switzerland: WHO; 2011. http://
whqlibdoc.who.int/publications/2011/9789240686458_eng.pdf.
Accessed July 5, 2011.
37. Ockene JK, Schneider KL, Lemon SC, Ockene IS. Can we improve
adherence to preventive therapies for cardiovascular health? Circulation.
2011;124:1276 –1282.
38. Frieden TR, Berwick DM. The “Million Hearts” initiative—preventing
heart attacks and strokes. N Engl J Med. 2011;365:e27.

CLINICAL PERSPECTIVE
The exponential rise in cardiovascular disease over the past decade has placed a tremendous burden on the health and
economic development of countries worldwide, with unprecedented demands for an effective response from governments
and other stakeholders in global health. From a large, multinational registry of outpatients with established cardiovascular
disease or ⱖ3 risk factors, we used data from 53 570 individuals from 36 countries to examine the relationship between
Downloaded from http://ahajournals.org by on December 31, 2019

country-level economic and health system factors and the risk of elevated cholesterol (total cholesterol levels ⬎200
mg/dL). The analysis was performed separately for patients with versus without previous history of hyperlipidemia; a
higher proportion of the total variability in elevated cholesterol was at the country level for patients with (12.1%) versus
without (7.4%) history of hyperlipidemia. Among patients with history of hyperlipidemia, after adjusting for patient-level
demographic and clinical characteristics and average fat consumption at the country level, countries in the highest tertile
of gross national income or World Health Organization index of health system achievement were found to have
significantly lower odds of elevated cholesterol than those in each of the lower 2 tertiles, and the odds of elevated
cholesterol was higher for countries in higher versus lower tertile of out-of-pocket health expenditures. No significant
associations between country-level factors and elevated cholesterol were found for patients without history of
hyperlipidemia. These results support the need for strengthening efforts toward effective cardiovascular disease prevention
and control and may provide insight for health policy setting at the national level.