Professional Documents
Culture Documents
Nej Mo A 2206916
Nej Mo A 2206916
Original Article
A BS T R AC T
BACKGROUND
Five modifiable risk factors are associated with cardiovascular disease and death The authors’ full names, academic de-
from any cause. Studies using individual-level data to evaluate the regional and sex- grees, and affiliations are listed in the Ap-
pendix. Dr. Blankenberg can be contacted
specific prevalence of the risk factors and their effect on these outcomes are lacking. at s.blankenberg@uke.de or at the Uni-
versity Heart and Vascular Center, Depart-
METHODS ment for Cardiology, Center of Population
Health Innovation, University Medical
We pooled and harmonized individual-level data from 112 cohort studies conducted Center Hamburg–Eppendorf, Martinistr.
in 34 countries and 8 geographic regions participating in the Global Cardiovascular 52, 20246 Hamburg, Germany.
Risk Consortium. We examined associations between the risk factors (body-mass A list of the investigators in the Global
index, systolic blood pressure, non–high-density lipoprotein cholesterol, current smok- Cardiovascular Risk Consortium is pro-
ing, and diabetes) and incident cardiovascular disease and death from any cause using vided in the Supplementary Appendix,
available at NEJM.org.
Cox regression analyses, stratified according to geographic region, age, and sex.
Population-attributable fractions were estimated for the 10-year incidence of cardio- Drs. Magnussen, Ojeda, and Leong con-
tributed equally to this article.
vascular disease and 10-year all-cause mortality.
This article was published on August 26,
RESULTS 2023, at NEJM.org.
Among 1,518,028 participants (54.1% of whom were women) with a median age of This is the New England Journal of Medi-
54.4 years, regional variations in the prevalence of the five modifiable risk factors cine version of record, which includes all
Journal editing and enhancements. The
were noted. Incident cardiovascular disease occurred in 80,596 participants during Author Accepted Manuscript, which is
a median follow-up of 7.3 years (maximum, 47.3), and 177,369 participants died the author’s version after external peer
during a median follow-up of 8.7 years (maximum, 47.6). For all five risk factors review and before publication in the Jour-
nal, is available at PubMed Central.
combined, the aggregate global population-attributable fraction of the 10-year inci-
dence of cardiovascular disease was 57.2% (95% confidence interval [CI], 52.4 to 62.1) N Engl J Med 2023;389:1273-85.
DOI: 10.1056/NEJMoa2206916
among women and 52.6% (95% CI, 49.0 to 56.1) among men, and the correspond- Copyright © 2023 Massachusetts Medical Society.
ing values for 10-year all-cause mortality were 22.2% (95% CI, 16.8 to 27.5) and 19.1%
(95% CI, 14.6 to 23.6).
CONCLUSIONS
Harmonized individual-level data from a global cohort showed that 57.2% and
52.6% of cases of incident cardiovascular disease among women and men, respec-
tively, and 22.2% and 19.1% of deaths from any cause among women and men,
respectively, may be attributable to five modifiable risk factors. (Funded by the
German Center for Cardiovascular Research (DZHK); ClinicalTrials.gov number,
NCT05466825.)
C
ardiovascular diseases are the Study Population
most common noncommunicable condi- We pooled and harmonized individual-level data
tions worldwide and account for approxi- from 1,518,028 participants in 112 cohort studies
mately one third of all deaths globally.1 Modifi- conducted in eight geographic regions (North
A Quick Take is
able risk factors such as body-mass index, systolic America, Latin America, Western Europe, Eastern
available at blood pressure, low-density lipoprotein cholesterol Europe and Russia, North Africa and the Middle
NEJM.org level, tobacco smoking, and diabetes account for East, sub-Saharan Africa, Asia, and Australia) par-
a percentage of the prevalence and incidence of ticipating in the Global Cardiovascular Risk
cardiovascular disease; however, the percentage Consortium. Data were harmonized by applying
varies according to the populations studied and the variable definitions used by the MORGAM
the methods used.2,3 These risk factors are used (MONICA [Multinational Monitoring of Trends
to derive contemporary risk scores4-6 for the esti- and Determinants in Cardiovascular Diseases]
mation of the 10-year risk of cardiovascular dis- Risk, Genetics, Archiving, and Monograph) proj-
ease, although they are given different weights. ect.8 Investigators in the studies that were not
These cardiovascular risk factors also have dif- part of the MORGAM project were sent a list of
ferent associations with cardiovascular and non- study-related variables with definitions and were
cardiovascular outcomes. Tobacco use is strongly asked to provide these data. A description of each
associated with premature death, whereas elevat- study cohort and information about the local eth-
ed blood pressure and non–high-density lipopro- ics committees and participant informed consent
tein (HDL) cholesterol level are more specifically are provided in the Supplementary Appendix. The
related to cardiovascular disease.7 cohorts that were included in the Global Cardio-
A tailored reduction in the burden of cardio- vascular Risk Consortium were selected on the
vascular disease and death from any cause for basis of literature review, existing collaborations
persons and populations can be achieved with a among investigators, and the availability of the
better understanding of the region- and sex-specif- variables of interest (Table S1). The flow of partici-
ic associations of these cardiovascular risk factors pants through the study is shown in Figure S1 in
with the development of cardiovascular disease. the Supplementary Appendix.
The Global Cardiovascular Risk Consortium ana-
lyzed a global harmonized individual-level data Cardiovascular Risk Factors and Outcome
set of population-based cohorts to overcome the Definitions
limitations of summary data and methodologic Five risk factors (body-mass index, systolic blood
heterogeneity. pressure, non-HDL cholesterol, current smoking,
and diabetes) and two outcomes (cardiovascular
disease and death from any cause) were assessed
Me thods
in the study because of the heterogeneity of the
Study Design and Oversight effects of the risk factors on outcomes and the
The study was designed by the Global Cardiovas- widespread availability of these data in the popu-
cular Risk Consortium Management Group, whose lation. Further, these risk factors can be modi-
members are listed in the Supplementary Ap- fied with interventions. Information on these five
pendix, available with the full text of this article modifiable risk factors was collected at baseline
at NEJM.org. Data were collected by the Hamburg according to the protocols of the respective stud-
Data Center. Analyses were performed by the ies included in the Global Cardiovascular Risk
second author and reviewed by the Global Car- Consortium. The standardized definitions that
diovascular Risk Consortium Statistical Working were used to classify cardiovascular disease events
Group. The first draft of the manuscript was pre- are provided in Table S2, and the representative-
pared by the first, second, and last authors and ness of the study population is shown in Table S3.
was reviewed and edited by all the authors. The
authors jointly agreed to submit the manuscript Statistical Analysis
for publication and vouch for the accuracy and Missing data were imputed by means of multiple
completeness of the data and for the fidelity of imputation with chained equations (Table S4).9
the study to the protocol. Both crude and age- and sex-standardized base-
line characteristics were calculated according to intervals have not been adjusted for multiple com-
region with the use of direct standardization, with parisons and should not be used in place of hy-
the age and sex distribution of the Global Cardio- pothesis testing. All statistical analyses were per-
vascular Risk Consortium data set as the standard. formed with the use of R software, version 4.1.3.13
Age-standardized event rates stratified according
to region were also estimated and reported per R e sult s
1000 person-years. Cumulative incidence curves
were generated for cardiovascular disease and Participant Characteristics and Prevalence
death from any cause. Associations between risk of Risk Factors
factors and outcome events were evaluated with The baseline examination for all study cohorts
the use of a two-stage, multivariate, random- included in the Global Cardiovascular Risk Con-
effects meta-analysis of individual participant sortium took place between 1963 and 2020. In
data.10 Sex-specific Cox models, with age as the the age- and sex-standardized analysis of data
time scale,11 were computed for each study, and from 1,518,028 participants (54.1% of whom
coefficients were pooled across studies according were women) with a median age of 54.4 years,
to region as well as globally. Covariates (body-mass the median body-mass index (the weight in kilo-
index, systolic blood pressure, non-HDL choles- grams divided by the square of the height in
terol level, current smoking, diabetes, and the use meters) was 26.4 (interquartile range, 23.7 to 29.7),
of antihypertensive medications) were included the median systolic blood pressure 130 mm Hg
simultaneously in the models. Both linear and re- (interquartile range, 118 to 144), and the median
stricted cubic spline models for continuous covari- non-HDL cholesterol level 156.9 mg per deciliter
ates and models that allowed for time-varying ef- (4.06 mmol per liter; interquartile range, 128.8 to
fects were performed. Models that included receipt 187.9 mg per deciliter [3.33 to 4.86 mmol per
of lipid-lowering medications as an additional liter]); 21.6% were current smokers and 8.3%
covariate were also computed with the use of had diabetes. The age- and sex-standardized
data from the studies in which this information prevalence of the five risk factors and the use of
was available (these data were missing for ap- antihypertensive and lipid-lowering medications
proximately 20% of participants). across geographic regions are shown in Table 1
For the five modifiable risk factors, region- and Table S5. Baseline characteristics without
and sex-specific population-attributable fractions age and sex standardization are shown in Table
were estimated for the 10-year incidence of cardio- S6, and the distributions of the risk factors ac-
vascular disease and 10-year all-cause mortality cording to sex are shown in Tables S7 and S8.
(see the Supplementary Appendix). Population- The prevalence of modifiable risk factors across
attributable fraction is an estimate of the pro- contemporary national health examination sur-
portion of an outcome that could be prevented if veys, which were used in the population-attrib-
the value of a risk factor were replaced by a hypo- utable fraction analyses, is shown in Tables S9,
thetical, ideal value. The approach used by Laak- S10, and S11.
sonen and colleagues,12 which takes into account
the time-to-event nature of the data, was applied Cardiovascular Disease and Death
to calculate the population-attributable fractions. from Any Cause
Weibull models were used in the estimation, and The median duration of follow-up among partici-
their distributional assumptions were assessed pants was 7.3 years (interquartile range, 5.9 to
graphically. Reference categories for the risk fac- 11.8; maximum, 47.3) for incident cardiovascu-
tors are provided in the Supplementary Appendix. lar disease and 8.7 years (interquartile range,
All the models that were used in the analyses 7.0 to 15.9; maximum, 47.6) for death from any
of associations and population-attributable frac- cause. The follow-up times for each of the indi-
tions were run with the exclusion of first-year vidual cohorts are provided in Table S12. A total
follow-up data (1-year landmark analysis). Two- of 80,596 cardiovascular disease events (30,033
year landmark analyses that excluded data from in women and 50,563 in men) and 177,369 deaths
the first 2 years of follow-up were performed as from any cause (78,608 in women and 98,761 in
sensitivity analyses. The widths of the confidence men) were observed during the follow-up period
North Africa
North Latin Western Eastern Europe and the Sub-Saharan
Characteristic Global America America Europe and Russia Middle East Africa Asia Australia
Cohort studies
Cohort studies — no. 112 11 10 58 16 5 2 4 6
Participants — no. 1,518,028 65,182 191,244 907,760 51,133 185,608 10,390 59,802 46,909
Range of survey years† 1963–2020 1971–2011 1990–2013 1970–2015 1983–2014 1963–2020 2011–2017 1988–2015 1983–2007
Participants
The
Median age (IQR) — yr‡ 54.4 54.0 54.0 54.6 54.1 54.0 54.0 54.0 54.6
(4.2–63.0) (45.0–63.0) (45.0–63.0) (45.5–63.0) (45.5–63.0) (45.0–62.6) (45.0–63.0) (45.0–63.0) (45.5–63.0)
Male sex — %‡ 45.9 45.9 45.9 45.9 45.9 45.9 45.9 45.9 45.9
Median BMI (IQR) 26.4 27.2 28.2 26.1 27.2 27.0 21.0 22.8 26.4
(23.7–29.7) (24.1–31.0) (25.4–31.5) (23.6–29.2) (24.3–30.6) (24.0–30.3) (19.0–23.4) (20.5–25.2) (23.7–29.5)
Median SBP (IQR) — 130.0 122.0 126.7 134.0 132.0 115.0 125.0 123.5 127.0
mm Hg (118.0–144.0) (111.0–136.0) (118.0–138.7) (122.0–148.0) (120.0–148.0) (105.0–130.0) (113.0–140.0) (112.0–136.0) (116.5–139.0)
nejm.org
terol (IQR) — mg/dl (128.8–187.9) (123.0–179.4) (131.1–185.2) (134.8–193.8) (135.0–191.8) (115.3–167.8) (77.3–154.7) (117.6–167.0) (124.5–181.0)
n e w e ng l a n d j o u r na l
Current smoking — % 21.6 22.5 30.8 20.9 29.2 14.2 18.6 23.5 14.3
Diabetes — % 8.3 13.0 15.3 4.8 9.0 18.3 2.0 5.1 4.8
Antihypertensive medica- 19.4 27.5 19.3 17.9 28.8 24.7 18.5 11.6 13.7
tions — %
october 5, 2023
Lipid-lowering medica- 9.6 8.0 2.3 11.5 8.8 11.6 NA 4.4 4.1
tions — %
History of CVD — % 5.6 7.2 3.6 5.6 11.2 5.6 0 6.3 7.2
* Percentages are presented for binary variables. Percentages, medians, and interquartile ranges (IQRs) are based on data from the available number of cases per variable. Percentages,
medians, and IQRs per geographic region were computed with the use of direct standardization according to age (≤40 years, >40 to ≤45 years, >45 to ≤50 years, >50 to ≤55 years, >55
to ≤60 years, >60 to ≤65 years, >65 to ≤70 years, and >70 years) and sex distribution in the Global Cardiovascular Risk Consortium data set. To convert the values for non–high-density
lipoprotein (HDL) cholesterol to millimoles per liter, multiply by 0.02586. BMI denotes body-mass index, CVD cardiovascular disease, DBP diastolic blood pressure, NA data not avail-
able, and SBP systolic blood pressure.
† The approximate number of observations according to categorized examination year were 198,517 (13.1%) between 1963 and 1989, 227,002 (15.0%) between 1990 and 1999, 746,074
Downloaded from nejm.org by Percival Camargo on November 8, 2023. For personal use only. No other uses without permission.
(49.1%) between 2000 and 2009, and 342,887 (22.8%) between 2010 and 2020.
‡ Similar values for the characteristics of age and sex across the geographic regions are due to the age and sex standardization.
Table 2. Age-Standardized Outcomes According to Geographic Region and Sex.*
CVD Events CVD Events Deaths from Any Cause Deaths from Any Cause
Region among Women among Men among Women among Men
nejm.org
North Africa and 1,146 6.4 4.0 1,805 9.4 6.8 1,650 8.1 6.1 8,615 11.9 16.2
the Middle East (5.6–7.2) (3.7–4.3) (8.5–10.2) (6.5–7.2) (7.2–9.0) (5.7–6.5) (11.3–12.5) (15.8–16.6)
Sub-Saharan 3 — — 1 — — 431 27.2 14.1 456 34.6 26.7
Africa‡ (16.0–36.9) (12.7–15.6) (25.1–42.9) (24.1–29.6)
Asia 311 2.5 2.5 353 4.2 5.1 6,399 11.0 7.6 4,751 16.7 12.0
* Computations were performed with the use of data from 1,088,670 participants in the analysis of cardiovascular disease and from 1,419,699 participants in the analysis of all-cause
Downloaded from nejm.org by Percival Camargo on November 8, 2023. For personal use only. No other uses without permission.
1277
The n e w e ng l a n d j o u r na l of m e dic i n e
(Table 2). The age-standardized 10-year incidence according to geographic region and sex, as cal-
of cardiovascular disease was 7.4% for women culated with the exclusion of data from the first
and 13.7% for men in North America, 6.4% for year of follow-up (1-year landmark analysis), are
women and 9.4% for men in North Africa and shown in Table S13 and Figures S3 through S7.
the Middle East, 5.7% for women and 9.9% for Subdistribution hazard ratios for cardiovascular
men in Eastern Europe and Russia, 3.7% for disease were calculated with death from noncar-
women and 7.3% for men in Western Europe, diovascular causes as the competing event, and
and 2.5% for women and 4.2% for men in Asia. the results were similar to those in the 1-year
The global 10-year incidence of cardiovascular dis-
landmark analysis (Table S14). The associations
ease was 4.0% among women and 7.8% among between the risk factors and cardiovascular dis-
men (Table 2). Cardiovascular disease appeared ease and death from any cause in the models
to develop in women at older ages than men that used continuous risk factors and allowed for
(Fig. S2). The age-standardized 10-year all-cause nonlinear effects are shown in Figure 1 and Fig-
mortality was 27.2% for women and 34.6% for ure S4. In a 2-year landmark analysis that ex-
men in sub-Saharan Africa, 10.1% for women cluded data from the first 2 years of follow-up,
and 17.9% for men in Eastern Europe and Russia, the observed associations appeared to be similar
11.0% for women and 16.7% for men in Asia, and to those in the 1-year landmark analysis (Fig. S5),
4.7% for women and 7.2% for men in Australia as were the results of other sensitivity analyses
(Table 2). (2-year landmark Cox regression models with a
linear exposure–effect assumption, models re-
Effects of Modifiable Risk Factors stricted to cohorts starting in the year 2000 or
The risk factor–associated hazard ratios for car- later, models restricted to participants with data
diovascular disease and death from any cause on the use of lipid-lowering medications, and
Women Men
Hazard Ratio for Cardiovascular
Figure 1. Associations of Continuous Risk Factors with Cardiovascular Disease and Death from Any Cause.
Shown are the results of a global 1-year landmark analysis that allowed for nonlinear effects. Participants with cardiovascular disease at
baseline were excluded. Age was used as the time scale. All five risk factors, together with the use of antihypertensive medications, were
included as covariates in the models. The widths of the 95% confidence intervals (shaded areas) have not been adjusted for multiplicity
and should not be used in place of hypothesis testing. To convert the values for non–high-density lipoprotein (HDL) cholesterol to milli-
moles per liter, multiply by 0.02586.
Women Men
Per 5-Unit Increase Per 20–mm Hg Increase Per 38.67–mg/dl Increase
in BMI in SBP in Non-HDL Cholesterol Current Smoking Diabetes Mellitus
Hazard Ratio for Cardiovascular
50 60 70 80 50 60 70 80 50 60 70 80 50 60 70 80 50 60 70 80
50 60 70 80 50 60 70 80 50 60 70 80 50 60 70 80 50 60 70 80
Age (yr)
Figure 2. Associations of Risk Factors with Cardiovascular Disease and Death from Any Cause.
Shown are the results of a global 1-year landmark analysis that allowed for effects to change with age. Participants with cardiovascular
disease at baseline were excluded. Age was used as the time scale. All five risk factors, together with the use of antihypertensive medica-
tions, were included as covariates in the models. The widths of the 95% confidence intervals (shaded areas) have not been adjusted for
multiplicity and should not be used in place of hypothesis testing. BMI denotes body-mass index, and SBP systolic blood pressure.
64.2% (95% CI, 59.8 to 68.6) among women and among those in Australia; and 15.7% (95% CI, 9.3
60.5% (95% CI, 57.2 to 63.9) among men in North to 22.0) and 2.1% (95% CI, −4.3 to 8.6), respec-
Africa and the Middle East and 55.5% (95% CI, tively, among those in Western Europe (Fig. 3).
50.7 to 60.3) and 50.3% (95% CI, 46.8 to 53.8), The population-attributable fractions calcu-
respectively, among those in North America. The lated for the individual five modifiable risk factors
aggregate population-attributable fraction of 10- are shown in Figure S9. The population-attribut-
year all-cause mortality was 34.3% (95% CI, 29.7 able fraction of the 10-year incidence of cardio-
to 38.9) among women and 43.2% (95% CI, 39.8 to vascular disease associated with systolic blood
46.6) among men in Asia; 13.7% (95% CI, 7.1 to pressure was 29.3% (95% CI, 25.4 to 33.2) among
20.3) and 2.9% (95% CI, −3.7 to 9.5), respectively, women and 21.6% (95% CI, 18.7 to 24.5) among
Women Men
0 25 50 75 100 0 25 50 75 100
Percentage
B Population-Attributable Fractions According to Different Regions for Five Modifiable Risk Factors Combined
North America Eastern Europe
CVD Death and Russia
55.5% 21.2% CVD Death
50.3% 13.7% 60.2% 22.4%
Western Europe 57.6% 27.7%
CVD Death
55.8% 15.7%
53.2% 2.1%
Latin America
CVD Death
57.9% 21.7%
51.9% 13.9%
Figure 3. Population-Attributable Fractions of the 10-Year Incidence of Cardiovascular Disease (CVD) and 10-Year All-Cause Mortality.
Population-attributable fractions were estimated with the use of Weibull models. For the five modifiable risk factors combined, the ag-
gregate global population-attributable fractions are shown in Panel A and the aggregate population-attributable fractions according to
geographic region are shown in Panel B. Population-attributable fractions for the individual risk factors according to geographic region
are provided in Figure S9 in the Supplementary Appendix.
men; the corresponding values were 15.4% (95% region. Elevated systolic blood pressure ap-
CI, 10.9 to 19.8) and 16.6% (95% CI, 12.6 to peared to be the largest contributor to the pop-
20.6) for cardiovascular disease associated with ulation-attributable fraction of incident cardio-
non-HDL cholesterol level and 15.2% (95% CI, vascular disease events in all regions.
13.3 to 17.1) and 10.2% (95% CI, 9.2 to 11.2) for The Global Cardiovascular Risk Consortium
cardiovascular disease associated with diabetes. and other studies14-16 confirmed apparent differ-
The population-attributable fraction of the 10-year ences in cardiovascular risk factor profile and
incidence of cardiovascular disease among wom- event rates between women and men, irrespec-
en was 6.7% (95% CI, 5.8 to 7.6) for current smok- tive of geographic region. Differences in risk
ing and 7.6% (95% CI, 5.1 to 10.1) for body-mass factor level have been shown to translate into
index; the corresponding values among men were lifetime risk of cardiovascular disease17 but not
10.7% (95% CI, 9.6 to 11.7) and 7.6% (95% CI, 5.6 necessarily to affect other fatal outcomes. Car-
to 9.7). Among women, the population-attribut- diovascular risk factors are known to increase
able fraction of 10-year all-cause mortality as- cardiovascular disease risk differently across
sociated with diabetes was 12.2% (95% CI, 11.1 various geographic regions.2,18 Among them,
to 13.3), whereas among men, the population- high blood pressure is associated with up to
attributable fraction of 10-year all-cause mortal- 13.5% of all deaths annually worldwide and is
ity associated with current smoking was 14.4% considered to be the leading risk factor for car-
(95% CI 13.3 to 15.4). The population-attribut- diovascular disease.19 Strict blood-pressure con-
able fractions of the 10-year incidence of cardio- trol to a systolic blood pressure of less than 120
vascular disease and 10-year all-cause mortality, mm Hg has been associated with lower rates of
according to modifiable risk factor level or sta- cardiovascular events and all-cause mortality.20
tus, are shown in Tables S20 and S21. Our data corroborate this observation; of the
five risk factors studied, systolic blood pressure
may offer the greatest potential for the preven-
Discussion
tion of cardiovascular disease. Although there is
The Global Cardiovascular Risk Consortium har- a strong continuous association between non-
monized individual-level data from 1,518,028 par- HDL cholesterol level and incident cardiovascular
ticipants who participated in 112 cohort studies disease,21 we and others3,22,23 observed an inverted
conducted in 34 countries in North America, J-shaped association of non-HDL cholesterol level
Latin America, Western Europe, Eastern Europe with all-cause mortality. Although very low levels
and Russia, North Africa and the Middle East, of non-HDL cholesterol are related to a reduction
sub-Saharan Africa, Asia, and Australia to assess in cardiovascular disease events,24,25 some obser-
the effect of five modifiable risk factors on the vations point toward higher all-cause mortality
incidence of cardiovascular disease and all-cause among participants with very low levels, at least
mortality. The study showed that the prevalence in longer-term follow-up.26 In contrast to what was
of the five modifiable risk factors and the inci- previously reported,3 body-mass index and current
dence of cardiovascular disease and all-cause mor- smoking (at least in some parts of the world) were
tality varied across geographic regions worldwide, associated with comparatively modest population-
and women had consistently lower event rates attributable fractions of cardiovascular disease
than men. The association between individual events in the populations participating in the
modifiable risk factors and both incident cardio- Global Cardiovascular Risk Consortium. These
vascular disease and death from any cause also associations may be related to underlying differ-
varied across regions. The five modifiable risk ences in population characteristics, risk-factor
factors accounted for an aggregate population- definition and prevalence, or methods used to
attributable fraction of the 10-year incidence of estimate population-attributable fractions.
cardiovascular disease of 57.2% among women Case–control studies such as INTERHEART
and 52.6% among men, and the corresponding may have overestimated the population-attribut-
values for 10-year all-cause mortality were 22.2% able fraction of the incidence of cardiovascular
and 19.1%. Population-attributable fractions of disease subtypes by attributing 90% of the risk
the incidence of cardiovascular disease and all- of myocardial infarction to nine targetable risk
cause mortality varied according to geographic factors.2 Data from 155,722 participants who were
evaluated prospectively in the Prospective Urban was used to reduce variation, and sensitivity
Rural Epidemiology (PURE) study suggested that analyses provided results similar to those for the
71% of cardiovascular disease cases are attribut- overall study population. Standardized event rates
able to 14 potentially modifiable metabolic and should be interpreted as descriptive measures
behavioral risk factors, a result that is consistent rather than actual incidences in a population. To
with our findings.3 Our study focused on five overcome bias resulting from deaths from non-
modifiable risk factors for which strict control cardiovascular diseases that were present at the
could potentially prevent 57.2% of all cases of time of the baseline examination, analyses were
cardiovascular disease in women and 52.6% of all performed with the exclusion of first-year follow-
cases in men globally. The varying effect of in- up data. Information about modifiable risk factors
dividual risk factors across different regions could was available from the baseline examination, and
enable ranking and prioritization of risk-factor the effect of changes in exposure over time are
control for public health action within those re- not known; the analyses were not corrected for
gions. However, there is substantial scope for a regression dilution bias. Residual confounding
more complete characterization of the risk of car- cannot be completely excluded. The effects of
diovascular disease. Environmental and exposure- overweight and obesity may be mediated by hy-
related factors such as physical activity,2 alcohol perlipidemia, hypertension, and diabetes.35 Mod-
consumption,27 air pollution,28 climate and noise,29 els that included body-mass index, systolic blood
educational level,3 or psychosocial risk factors, pressure, and diabetes attribute the share of the
including depression,30 have an effect on the risk effect of body-mass index to systolic blood pres-
of cardiovascular disease. Biomarkers31,32 and ge- sure and diabetes, even if overweight or obesity
netic variants most likely would add to the pre- is the real underlying cause. The definition of cur-
diction of cardiovascular disease risk. rent smoking may not capture the entire spectrum
The analysis by the Global Cardiovascular Risk and dose of tobacco exposure, and smoking ces-
Consortium differs from other global initiatives sation during follow-up might have led to an un-
that combine different data sources such as reg- derestimation of tobacco smoking as a risk factor.
istries, population surveys, and health system It was also assumed that risk factor effects and
administrative data to produce meta-analytic prevalence within a region are homogeneous;
summaries.33,34 The Global Cardiovascular Risk however, intraregional differences may exist.
Consortium maintains a large and comprehensive Information about ethnic group is not provided,
database of harmonized, observational, individ- because definitions differed among the cohorts
ual-level, prospectively collected data. This data- or collection of the variable was incomplete or not
base allows for multiple prespecified statistical available to a comparable standard. The categori-
analyses on large-scale individual-level data. This zation of geographic regions by the World Health
study relates major modifiable cardiovascular risk Organization and United Nations was adapted to
factors to the incidence of cardiovascular disease accommodate cohort size and representativeness
and all-cause mortality. The inclusion of cohorts of a geographic region, so a different categoriza-
with a large spectrum of follow-up times en- tion of regions may produce different results.
abled robust sex-specific analyses and the evalu- In our study, harmonized individual-level data
ation of differences across geographic regions. from the Global Cardiovascular Risk Consortium
Our study has several limitations. The Global showed that 57.2% and 52.6% of cases of incident
Cardiovascular Risk Consortium database includes cardiovascular disease among women and men,
cohorts with varying representativeness, data qual- respectively, and 22.2% and 19.1% of deaths from
ity and quantity, dates of baseline assessments, any cause among women and men, respectively,
follow-up times, end-point definitions, and use may be attributable to five modifiable risk factors.
of clinical interventions. Variation in the adjudi- The prevalence and effect of these risk factors on
cation of causes of death or surrogates of non- the incidence of cardiovascular disease and all-
fatal myocardial infarction is plausible across cause mortality vary according to sex and geo-
regions, but an analysis that included the use of graphic region.
a secondary definition of cardiovascular disease
that excluded unclassifiable death, unstable Supported by the German Center for Cardiovascular Research
(DZHK).
angina, and coronary revascularization did not Disclosure forms provided by the authors are available with
change the results. Structured harmonization the full text of this article at NEJM.org.
Appendix
The authors’ full names and academic degrees are as follows: Christina Magnussen, M.D., Francisco M. Ojeda, Ph.D., Darryl P. Leong,
M.B., B.S., Ph.D., Jesus Alegre‑Diaz, M.D., Philippe Amouyel, M.D., Ph.D., Larissa Aviles‑Santa, M.D., Dirk De Bacquer, Ph.D., Chris-
tie M. Ballantyne, M.D., Antonio Bernabé‑Ortiz, M.D., Ph.D., Martin Bobak, Ph.D., Hermann Brenner, M.D., Rodrigo M. Carrillo‑Larco,
M.D., Ph.D., James de Lemos, M.D., Annette Dobson, Ph.D., Marcus Dörr, M.D., Chiara Donfrancesco, Ph.D., Wojciech Drygas, M.D.,
Ph.D., Robin P. Dullaart, M.D., Ph.D., Gunnar Engström, M.D., Ph.D., Marco M. Ferrario, M.D., Ph.D., Jean Ferrières, M.D., Ph.D.,
Giovanni de Gaetano, M.D., Ph.D., Uri Goldbourt, Ph.D., Clicerio Gonzalez, M.D., Guido Grassi, M.D., Allison M. Hodge, Ph.D.,
Kristian Hveem, M.D., Ph.D., Licia Iacoviello, M.D., Ph.D., M. Kamran Ikram, M.D., Ph.D., Vilma Irazola, M.D., Ph.D., Modou Jobe, M.D.,
Pekka Jousilahti, M.D., Ph.D., Pontiano Kaleebu, M.D., Ph.D., Maryam Kavousi, M.D., Ph.D., Frank Kee, M.D., Davood Khalili,
M.D., Ph.D., Wolfgang Koenig, M.D., Anna Kontsevaya, M.D., Ph.D., Kari Kuulasmaa, Ph.D., Karl J. Lackner, M.D., David M. Leistner,
M.D., Lars Lind, M.D., Ph.D., Allan Linneberg, M.D., Ph.D., Thiess Lorenz, M.A., Magnus Nakrem Lyngbakken, M.D., Ph.D., Reza
Malekzadeh, M.D., Sofia Malyutina, M.D., Ph.D., Ellisiv B. Mathiesen, M.D., Ph.D., Olle Melander, M.D., Ph.D., Andres Metspalu,
M.D., Ph.D., J. Jaime Miranda, M.D., Ph.D., Marie Moitry, M.D., Joseph Mugisha, Ph.D., Mahdi Nalini, M.D., Ph.D., Vijay Nambi, M.D.,
Ph.D., Toshiharu Ninomiya, M.D., Ph.D., Karen Oppermann, M.D., Ph.D., Eleonora d’Orsi, Ph.D., Andrzej Pająk, M.D., Ph.D., Luigi
Palmieri, Ph.D., Demosthenes Panagiotakos, M.D., Ph.D., Arokiasamy Perianayagam, Ph.D., Annette Peters, Ph.D., Hossein Poustchi,
M.D., Ph.D., Andrew M. Prentice, Ph.D., Eva Prescott, M.D., Ulf Risérus, M.D., Ph.D., Veikko Salomaa, M.D., Ph.D., Susana Sans, M.D.,
Ph.D., Satoko Sakata, M.D., Ph.D., Ben Schöttker, Ph.D., Aletta E. Schutte, Ph.D., Sadaf G. Sepanlou, Ph.D., Sanjib Kumar Sharma,
M.D., Jonathan E. Shaw, M.D., Leon A. Simons, M.D., Stefan Söderberg, M.D., Ph.D., Abdonas Tamosiunas, M.D., Barbara Thorand,
Ph.D., Hugh Tunstall‑Pedoe, M.D., Raphael Twerenbold, M.D., Diego Vanuzzo, M.D., Giovanni Veronesi, Ph.D., Julia Waibel, M.Sc.,
S. Goya Wannamethee, Ph.D., Masafumi Watanabe, M.D., Ph.D., Philipp S. Wild, M.D., Yao Yao, M.D., Ph.D., Yi Zeng, M.D., Andreas
Ziegler, Ph.D., and Stefan Blankenberg, M.D.
The authors’ affiliations are as follows: the Center for Population Health Innovation, University Heart and Vascular Center Hamburg
(C.M., F.M.O., T.L., R.T., J.W., A.Z., S.B.), University Medical Center Hamburg–Eppendorf (C.M., F.M.O., T.L., R.T., J.W., A.Z., S.B.),
the German Center for Cardiovascular Research (DZHK) Partner Site Hamburg–Kiel–Lübeck (C.M., T.L., R.T., J.W., S.B.), the Division
of Clinical Epidemiology and Aging Research, German Cancer Research Center, and Network Aging Research, Heidelberg University,
Heidelberg (H.B., B.S.), the Department of Internal Medicine B, University Medicine Greifswald, and DZHK Partner Site Greifswald,
Greifswald (M.D.), DZHK Partner Site Munich Heart Alliance (W.K., A. Peters), the German Heart Center, Technical University of
Munich (W.K.), and the Institute for Medical Information Processing, Biometry, and Epidemiology, Medical Faculty, Ludwig Maximil-
ians Universität München (A. Peters, B.T.), Munich, the Institute of Epidemiology and Medical Biometry, University of Ulm, Ulm (W.K.),
the Institute of Clinical Chemistry and Laboratory Medicine (K.J.L.), Preventive Cardiology and Preventive Medicine (P.S.W.), and
Clinical Epidemiology and Systems Medicine, Center for Thrombosis and Hemostasis (P.S.W.), University Medical Center of the Jo-
hannes Gutenberg–University Mainz, DZHK Partner Site Rhine–Main (K.J.L., P.S.W.), and Institute for Molecular Biology (P.S.W.),
Mainz, the University Heart and Vascular Center Frankfurt, DZHK Partner Site Rhine–Main, Frankfurt (D.M.L.), the Institute of Epide-
miology, Helmholtz Zentrum München, German Research Center for Environmental Health (A. Peters, B.T.), and German Center for
Diabetes Research (DZD) Partner Site Munich–Neuherberg (A. Peters, B.T.), Neuherberg — all in Germany; the Department of Medicine
(Cardiology), McMaster University, Hamilton, Canada (D.P.L.); the Experimental Medicine Research Unit, School of Medicine, Na-
tional Autonomous University of Mexico, Mexico City (J.A.-D.), and Centro de Estudios en Diabetes, Centro de Investigacion en Salud
Poblacional, Instituto Nacional de Salud Publica, Cuernavaca (C.G.) — both in Mexico; Université de Lille, INSERM, Centre Hospitalier
University de Lille, Institut Pasteur de Lille, UMR1167–RID-AGE–Risk Factors and Molecular Determinants of Aging-Related Diseases,
Epidemiology and Public Health Department, Lille (P.A.), the Department of Cardiology, INSERM UMR1295, Toulouse Rangueil Uni-
versity Hospital, Toulouse (J.F.), and the Department of Public Health, Strasbourg University Hospital, University of Strasbourg, Stras-
bourg (M.M.) — all in France; the Division of Clinical and Health Services Research, National Institute on Minority Health and Health
Disparities (L.A.-S.), and the Metabolic Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute
(M.N.), National Institutes of Health, Bethesda, MD; the Department of Public Health and Primary Care, Ghent University, Ghent,
Belgium (D.D.B.); the Department of Medicine, Baylor College of Medicine (C.M.B.), and Michael E. DeBakey Veterans Affairs Hospital
and Baylor College of Medicine (V.N.), Houston, and the Division of Cardiology, Department of Internal Medicine, University of Texas
Southwestern Medical Center, Dallas (J.L.) — all in Texas; the CRONICAS Center of Excellence in Chronic Diseases, Universidad Pe-
ruana Cayetano Heredia, Lima, Peru (A.B.-O., J.J.M.); the Department of Epidemiology and Public Health (M.B.) and the Research De-
partment of Primary Care and Population Health (S.G.W.), University College London, London, the Centre for Public Health, Queens
University Belfast, Belfast (F.K.), and the Cardiovascular Epidemiology Unit, Institute of Cardiovascular Research, University of Dundee,
Dundee (H.T.-P.) — all in the United Kingdom; the Emory Global Diabetes Research Center and Hubert Department of Global Health,
Rollins School of Public Health, Emory University, Atlanta (R.M.C.-L.); the School of Public Health, University of Queensland, Brisbane
(A.D.), the Division of Cancer Epidemiology, Cancer Council Victoria (A.M.H.), the Centre for Epidemiology and Biostatistics, Mel-
bourne School of Population and Global Health, the University of Melbourne (A.M.H.), and Baker Heart and Diabetes Institute (J.E.S.),
Melbourne, the Sydney School of Public Health, Faculty of Medicine and Health, University of Sydney (J.J.M.), and the George Institute
for Global Health (A.E.S.), Sydney, and the School of Population Health (A.E.S.), University of New South Wales, Kensington (L.A.S.)
— all in Australia; the Department of Cardiovascular and Endocrine–Metabolic Diseases and Aging, Istituto Superiore di Sanità, Rome
(C.D., L.P.), the Research Center in Epidemiology and Preventive Medicine, Department of Medicine and Surgery, University of Insubria,
Varese (M.M.F., L.I., G.V.), the Department of Epidemiology and Prevention, IRCCS Neuromed, Pozzilli (G. de Gaetano, L.I.), Clinica
Medica, University of Milano–Bicocca, Milan (G. Grassi), and the MONICA (Monitoring Cardiovascular Diseases)–Friuli Study Group,
Udine (D.V.) — all in Italy; the Department of Epidemiology, Cardiovascular Disease Prevention and Health Promotion, National Insti-
tute of Cardiology, and Lazarski University, Warsaw (W.D.), and the Department of Epidemiology and Population Studies, Institute of
Public Health, Faculty of Health Sciences, Jagiellonian University Medical College, Krakow (A. Pajak) — all in Poland; the Department
of Endocrinology, University Medical Center Groningen, University of Groningen, Groningen (R.P.D.), and the Departments of Neurol-
ogy (M.K.I.) and Epidemiology (M.K.I., M.K.), Erasmus University Medical Center Rotterdam, Rotterdam — both in the Netherlands;
the Department of Clinical Sciences Malmö, Lund University, Malmö (G.E., O.M.), the Department of Public Health and Caring Sci-
ences, Clinical Nutrition and Metabolism, Uppsala University (U.R.), and the Department of Medical Sciences, Uppsala University,
Uppsala (L.L.), and the Department of Public Health and Clinical Medicine, University of Umea, Umea (S. Söderberg) — all in Sweden;
the Department of Epidemiology, Tel Aviv University School of Public Health, Tel Aviv, Israel (U.G.); HUNT (Trøndelag Health Study)
Research Center, Department of Public Health and Nursing, Norwegian University of Science and Technology, Levanger (K.H.), the
K.G. Jebsen Center for Genetic Epidemiology, Department of Public Health and Nursing, Trondheim(K.H.), the Department of Cardiol-
ogy, Division of Medicine, Akershus University Hospital, Lørenskog (M.N.L.), the K.G. Jebsen Center for Cardiac Biomarkers, Institute
of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo (M.N.L.), and the Department of Clinical Medicine, University of
Tromsø the Arctic University of Norway, Tromsø (E.B.M.) — all in Norway; the Department of Chronic Diseases, Institute for Clinical
Effectiveness and Health Policy, Buenos Aires (V.I.); Medical Research Council Unit The Gambia at London School of Hygiene and
Tropical Medicine, Banjul, Gambia (M.J., A.M.P.); the Department of Public Health and Welfare, Finnish Institute for Health and Wel-
fare, Helsinki (P.J., K.K., V.S.); Medical Research Council–Uganda Virus Research Institute and London School of Hygiene and Tropical
Medicine Uganda Research Unit, Entebbe, Uganda (P.K., J.M.); the Prevention of Metabolic Disorders Research Center, Research Insti-
tute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences (D.K.), the Liver and Pancreaticobiliary Disease Research
Center (R.M.), the Digestive Oncology Research Center (R.M., H.P.), and the Digestive Disease Research Center (R.M., M.N., H.P.,
S.G.S.), Digestive Disease Research Institute, Shariati Hospital, Tehran University of Medical Sciences — all in Tehran, Iran; the Na-
tional Research Center for Therapy and Preventive Medicine, Ministry of Healthcare of the Russian Federation, Moscow (A.K.), and the
Research Institute of Internal and Preventive Medicine–Branch of the Federal Research Center Institute of Cytology and Genetics, Sibe-
rian Branch of Russian Academy of Sciences, Novosibirsk (S.M.) — both in Russia; the Department of Clinical Medicine, Faculty of
Health and Medical Sciences (A.L.), and the Department of Cardiology, Bispebjerg Hospital (E.P.), University of Copenhagen, and the
Center for Clinical Research and Prevention, Bispebjerg and Frederiksberg Hospital (A.L.) — both in Copenhagen; the Estonian Ge-
nome Center, Institute of Genomics, University of Tartu, Tartu, Estonia (A.M.); the Department of Epidemiology and Public Health,
Graduate School of Medical Sciences, Kyushu University, Fukuoka (T.N., S. Sakata), and the Global Center of Excellence Program Study
Group, Yamagata University School of Medicine, Yamagata (M.W.) — both in Japan; the School of Medicine, University of Passo Fundo,
Passo Fundo (K.O.), and the Department of Public Health, Postgraduate Program in Public Health, Federal University of Santa Catarina,
Florianopolis (E.O.) — both in Brazil; the School of Health Sciences and Education, Harokopio University, Athens, Greece (D.P.); the
National Council of Applied Economic Research, Delhi (A. Perianayagam), and the International Institute for Population Sciences,
Mumbai (A. Perianayagam) — both in India; the Catalan Department of Health, Barcelona (S. Sans); the Hypertension in Africa Re-
search Team, South African Medical Research Council Unit for Hypertension and Cardiovascular Disease, North-West University,
Potchefstroom (A.E.S.), and the School of Mathematics, Statistics, and Computer Science, University of KwaZulu-Natal, Pietermaritz-
burg (A.Z.) — both in South Africa; the Department of Internal Medicine, B.P. Koirala Institute of Health Sciences, Dharan, Nepal
(S.K.S.); the Laboratory of Population Studies, Institute of Cardiology, and the Department of Preventive Medicine, Faculty of Public
Health, Lithuanian University of Health Sciences, Kaunas, Lithuania (A.T.); the China Center for Health Development Studies (Y.Y.) and
Center for Healthy Aging and Development Studies, National School of Development (Y.Z.), Peking University, and the Key Laboratory
of Epidemiology of Major Diseases, Peking University, Ministry of Education (Y.Y.) — both in Beijing; the Center for the Study of Aging
and Human Development and Geriatrics Division, Medical School of Duke University, Durham, NC (Y.Z.); and Cardio-CARE, Davos
(A.Z., S.B.), and the Swiss Institute of Bioinformatics, Lausanne (A.Z.) — both in Switzerland.
References
1. Joseph P, Leong D, McKee M, et al. 7. Jaspers NEM, Blaha MJ, Matsushita K, the Framingham heart study. Circulation
Reducing the global burden of cardiovas- et al. Prediction of individualized lifetime 2009;119:3078-84.
cular disease, part 1: the epidemiology and benefit from lowering, blood pressure low- 15. Joseph P, Kutty VR, Mohan V, et al.
risk factors. Circ Res 2017;121:677-94. ering, antithrombotic therapy, and smok- Cardiovascular disease, mortality, and
2. Yusuf S, Hawken S, Ounpuu S, et al. ing cessation in apparently healthy people. their associations with modifiable risk
Effect of potentially modifiable risk fac- Eur Heart J 2020;41:1190-9. factors in a multi-national South Asia
tors associated with myocardial infarction 8. Evans A, Salomaa V, Kulathinal S, et al. cohort: a PURE substudy. Eur Heart J 2022;
in 52 countries (the INTERHEART study): MORGAM (an international pooling of 43:2831-40.
case-control study. Lancet 2004;364:937- cardiovascular cohorts). Int J Epidemiol 16. Walli-Attaei M, Joseph P, Rosengren
52. 2005;34:21-7. A, et al. Variations between women and
3. Yusuf S, Joseph P, Rangarajan S, et al. 9. van Buuren S, Groothuis-Oudshoorn K. men in risk factors, treatments, cardio-
Modifiable risk factors, cardiovascular dis- mice: Multivariate Imputation by Chained vascular disease incidence, and death in
ease, and mortality in 155 722 individuals Equations in R. J Stat Softw 2011;45:1-67. 27 high-income, middle-income, and low-
from 21 high-income, middle-income, and 10. Riley RD, Tierney JF, Stewart LA. Indi- income countries (PURE): a prospective
low-income countries (PURE): a prospec- vidual participant data meta-analysis: a cohort study. Lancet 2020;396:97-109.
tive cohort study. Lancet 2020;395:795-808. handbook for healthcare research. Med- 17. Berry JD, Dyer A, Cai X, et al. Lifetime
4. SCORE2 working group and ESC Car- ford, MA:Wiley, 2021. risks of cardiovascular disease. N Engl J
diovascular risk collaboration. SCORE2 11. Korn EL, Graubard BI, Midthune D. Med 2012;366:321-9.
risk prediction algorithms: new models to Time-to-event analysis of longitudinal 18. Ueshima H, Sekikawa A, Miura K, et al.
estimate 10-year risk of cardiovascular follow-up of a survey: choice of the time- Cardiovascular disease and risk factors in
disease in Europe. Eur Heart J 2021;42: scale. Am J Epidemiol 1997;145:72-80. Asia: a selected review. Circulation 2008;
2439-54. 12. Laaksonen MA, Virtala E, Knekt P, 118:2702-9.
5. Yadlowsky S, Hayward RA, Sussman Oja H, Härkänen T. SAS macros for calcu- 19. Lawes CM, Vander Hoorn S, Rodgers
JB, McClelland RL, Min YI, Basu S. Clini- lation of population attributable fraction A;International Society of Hypertension.
cal implications of revised pooled cohort in a cohort study design. J Stat Softw Global burden of blood-pressure-related
equations for estimating atherosclerotic 2011;43:1-25. disease, 2001. Lancet 2008;371:1513-8.
cardiovascular disease risk. Ann Intern 13. R Foundation for Statistical Comput- 20. SPRINT Research Group. A random-
Med 2018;169:20-9. ing. The R project for statistical comput- ized trial of intensive versus standard
6. World Health Organization. World ing (http://www.r-project.org/index.html). blood-pressure control. N Engl J Med 2015;
health statistics 2011 (https://web.archive 14. Pencina MJ, D’Agostino RB Sr, Larson 373:2103-16.
.org/web/20111116011154/http://www MG, Massaro JM, Vasan RS. Predicting 21. Brunner FJ, Waldeyer C, Ojeda F, et al.
.who.i nt/whosis/whostat/2011/en/). the 30-year risk of cardiovascular disease: Application of non-HDL cholesterol for
population-based cardiovascular risk strati- FOURIER. Am J Cardiovasc Drugs 2022; N, et al. Troponin I and cardiovascular
fication: results from the Multinational 22:5-8. risk prediction in the general population:
Cardiovascular Risk Consortium. Lancet 27. GBD 2020 Alcohol Collaborators. the BiomarCaRE consortium. Eur Heart J
2019;394:2173-83. Population-level risks of alcohol consump- 2016;37:2428-37.
22. Johannesen CDL, Langsted A, tion by amount, geography, age, sex, and 32. Neumann JT, Twerenbold R, Ojeda F,
Mortensen MB, Nordestgaard BG. Asso- year: a systematic analysis for the Global et al. Application of high-sensitivity tro-
ciation between low density lipoprotein Burden of Disease Study 2020. Lancet 2022; ponin in suspected myocardial infarction.
and all cause and cause specific mortality 400:185-235. N Engl J Med 2019;380:2529-40.
in Denmark: prospective cohort study. 28. Schraufnagel DE, Balmes JR, Cowl 33. Roth GA, Mensah GA, Johnson CO,
BMJ 2020;371:m4266. CT, et al. Air Pollution and noncommuni- et al. Global burden of cardiovascular dis-
23. Liu Y, Liu F, Zhang L, et al. Associa- cable diseases: a review by the Forum of eases and risk factors, 1990-2019: update
tion between low density lipoprotein cho- International Respiratory Societies’ Envi- from the GBD 2019 study. J Am Coll Car-
lesterol and all-cause mortality: results ronmental Committee, part 2: air pollu- diol 2020;76:2982-3021.
from the NHANES 1999-2014. Sci Rep tion and organ systems. Chest 2019;155: 34. Vaduganathan M, Mensah GA, Turco
2021;11:22111. 417-26. JV, Fuster V, Roth GA. The global burden
24. Sabatine MS, Giugliano RP, Keech AC, 29. Basner M, Babisch W, Davis A, et al. of cardiovascular diseases and risk: a com-
et al. Evolocumab and clinical outcomes Auditory and non-auditory effects of noise pass for future health. J Am Coll Cardiol
in patients with cardiovascular disease. on health. Lancet 2014;383:1325-32. 2022;80:2361-71.
N Engl J Med 2017;376:1713-22. 30. Walli-Attaei M, Rosengren A, Ranga- 35. Lu Y, Hajifathalian K, Ezzati M, Wood-
25. Robinson JG, Farnier M, Krempf M, rajan S, et al. Metabolic, behavioural, and ward M, Rimm EB, Danaei G. Metabolic
et al. Efficacy and safety of alirocumab in psychosocial risk factors and cardiovas- mediators of the effects of body-mass in-
reducing lipids and cardiovascular events. cular disease in women compared with dex, overweight, and obesity on coronary
N Engl J Med 2015;372:1489-99. men in 21 high-income, middle-income, heart disease and stroke: a pooled analy-
26. van Bruggen FH, Luijendijk HJ. Evo- and low-income countries: an analysis of sis of 97 prospective cohorts with 1·8 mil-
locumab’s long-term mortality risk un- the PURE study. Lancet 2022;400:811-21. lion participants. Lancet 2014;383:970-83.
clear due to shortened follow-up of 31. Blankenberg S, Salomaa V, Makarova Copyright © 2023 Massachusetts Medical Society.