For definition of risk categories, see footnotes to Table 1.
Daviglus ML, Stamler J, Pirzada A, Yan LL, Garside DB, Liu K, Wang R, Dyer AR, Lloyd-Jones DM, Greenland P. Favorable Cardiovascular Risk Profile in Young Women and Long-term Risk of Cardiovascular and All-Cause Mortality. JAMA. 2004;292(13):1588-1592. doi:10.1001/jama.292.13.1588
Author Affiliations: Department of Preventive
Medicine (Drs Daviglus, Stamler, Pirzada, Yan, Liu, Wang, Dyer, Lloyd-Jones,
and Greenland and Mr Garside); Department of Medicine, Division of Geriatrics
(Drs Daviglus and Liu); and Department of Medicine, Division of Cardiology
(Drs Greenland and Lloyd-Jones), Northwestern University, Feinberg School
of Medicine, Chicago, Ill.
Context For women, impact of cardiovascular risk factors measured in young adulthood,
particularly favorable (low-risk) profile, on mortality has been difficult
to assess due to low short-term death rates.
Objective To assess the relationship of baseline coronary risk factor status to
mortality from coronary heart disease (CHD), cardiovascular diseases (CVDs),
and all causes in young women.
Design Prospective cohort study.
Setting and Participants A total of 7302 women aged 18 to 39 years without prior CHD or major
electrocardiographic abnormalities screened between 1967 and 1973 for the
Chicago Heart Association Detection Project in Industry. Risk groups were
defined using national guidelines for values of systolic and diastolic blood
pressure, serum cholesterol level, body mass index, presence of diabetes,
and smoking status. Participants were divided into 4 groups: low risk, 0 risk
factors high but 1 or more unfavorable, 1 only risk factor high, and 2 or
more risk factors high.
Main Outcome Measures All-cause mortality, CHD mortality, and CVD mortality; hazard ratio
of outcome measures comparing low-risk group with other groups.
Results Only 20% met low-risk criteria; 59% had high levels of 1 or more risk
factors. During an average follow-up of 31 years, there were 47 CHD deaths,
94 CVD deaths, and 469 deaths from all causes. The age-adjusted CVD death
rate per 10 000 person-years was lowest for low-risk women and increased
with the number of risk factors, ie, 1.5, 1.7, 5.0, and 9.1 for low-risk;
0, 1, and 2 or more risk factors high, respectively. Multivariate-adjusted
CVD mortality hazard ratio for low-risk women was 0.19 (95% confidence interval,
0.08-0.45) compared with women with 2 or more risk factors high. Similar patterns
were observed for CHD and all-cause mortality and for both blacks and whites.
Conclusion For women with favorable levels for all 5 major risk factors at younger
ages, CHD and CVD are rare; long-term and all-cause mortality are much lower
compared with others.
Young adult men and middle-aged men and women with favorable levels
of all major cardiovascular risk factors, ie, low-risk status, have much lower
age-specific risks for cardiovascular disease (CVD) and all-cause mortality
than those with adverse levels of 1 or more risk factors.1 However,
the impact of a favorable cardiovascular risk profile in young women on subsequent
mortality has only been estimated statistically using coronary heart disease
(CHD) risk prediction models.2 In this study,
we examined the relationship of low CHD/CVD risk and individual risk factors
in young women to long-term CHD, CVD, and all-cause mortality.
From 1967 to 1973, the Chicago Heart Association (CHA) Detection Project
in Industry Study screened 39 522 employed Chicago-area men and women
aged 18 years and older. Standardized examination methods, follow-up procedures,
and death certificate coding were used.3- 7 Vital
status was ascertained through 2001, with an average (SD) follow-up of 31
(1.3) years. Informed consent was obtained from each study participant. The
study has been periodically approved by the Northwestern University Institutional
Of 7748 women (baseline ages 18-39 years), 446 were excluded for the
following reasons: baseline CHD (n = 7); major electrocardiographic
(ECG) abnormality (n = 403); missing data on smoking, blood pressure
(BP), serum cholesterol level, body mass index (BMI), diabetes, or education
(n = 36).
Eligible women (n = 7302) were classified into 4 risk groups
according to baseline CVD risk status. Low-risk status was defined as favorable
levels of all the following: systolic BP (SBP) 120 mm Hg or less and diastolic
BP (DBP) 80 mm Hg or less and not taking antihypertensive medication, serum
cholesterol level less than 200 mg/dL (<5.17 mmol/L) and not taking cholesterol-lowering
medication, BMI less than 25.0, no diabetes, and not smoking. Participants
not at low risk were classified as having 0 risk factors high but 1 or more
unfavorable, any 1 only, or 2 or more of the following: SBP 140 mm Hg or higher
or DBP 90 mm Hg or higher or taking antihypertensive medication, serum cholesterol
level 240 mg/dL or higher (≥6.21 mmol/L) or taking cholesterol-lowering
medication, BMI 30.0 or higher, diabetes, and cigarette smoking. Unfavorable
levels are defined as SBP 121 to 139 mm Hg and DBP 81 to 89 mm Hg and not
taking antihypertensive medication, serum cholesterol level 200 to 239 mg/dL
(5.17-6.18 mmol/L) and not taking cholesterol-lowering medication, and BMI
25.0 to 29.9. Race/ethnicity was assessed by interviewers to clarify reasons
for the higher CVD rates in blacks than whites, a major US problem requiring
additional research then and now.
Age-adjusted CHD, CVD, and all-cause mortality rates per 10 000
person-years of follow-up were computed by risk category. Cox proportional
hazards regression was used to calculate mortality hazard ratios (HRs) and
95% confidence intervals (CIs) by baseline risk category, adjusted for baseline
age only and for other risk factors. Tests of the proportional hazards assumption
showed no violation. Models were computed to estimate the HR and 95% CI (P .05) for individual risk factors and for the low-risk group compared
with other strata. Kaplan-Meier cumulative mortality curves were also plotted
for the 4 risk categories. All analyses were conducted with SAS statistical
software version 8.02 (SAS Institute Inc, Cary, NC).
Of the 7302 young women, 1469 (20.1%) were classified as low risk; a
majority (58.5%) of the cohort had 1 or more high-risk factors (Table 1). Low-risk women tended to be younger, white, and better
educated. During 31 years of follow-up, there were 47 CHD deaths, 94 CVD deaths,
and 469 deaths from all causes (Table 1).
All risk factors considered individually were related to CHD and CVD
death with cigarette smoking and BMI, significantly so in multivariate analyses
Age-adjusted CHD mortality rates were similar for low-risk women and
those with 0 risk factors high but 1 or more unfavorable and were much lower
than for women with 1 only risk factor or 2 or more risk factors high. Women
with 2 or more risk factors high had the highest CHD mortality rates (Table 3). Findings were similar for CVD and all-cause
mortality, eg, CVD mortality rate per 10 000 person-years for women with
2 or more risk factors high was 9.1, about 6 times that of low-risk women
(1.5). With adjustment for age, race, and minor ECG abnormalities, HRs for
31-year all-cause mortality were lowest for low-risk women and increased with
the number of risk factors. Results stratified by race were similar (data
not shown). Kaplan-Meier cumulative mortality curves depict similar results
Among young women without baseline major ECG abnormalities or prevalent
CHD at baseline, 31-year risks of CHD, CVD, and all-cause mortality were markedly
lower for those who were low-risk compared with others. The presence of high
levels of major risk factors was associated with much higher mortality risk.
Prospective population-based research involving large cohorts with long-term
follow-up into older age has delineated multiple favorable consequences of
baseline low-risk status. For the small percentage of young men and middle-aged
men and women who were low-risk at baseline compared with all others, CHD/CVD
is rare (endemic) not epidemic; CHD, CVD, and all-cause mortality rates are
remarkably lower.1,8 Persons at
low risk earlier in life experience higher quality of life, lower medication
use and prevalence of clinical diseases,9,10 less
subclinical coronary atherosclerosis,11 and
also substantially lower average annual health care costs in older age12 compared with others.
To our knowledge, the relationship of low-risk status to subsequent
mortality has not been previously assessed in young women. The few reports
on women have presented estimates from statistical extrapolation,2 have defined low risk alternatively,13 or
did not include younger women.1 A report describing
CHD risk prediction models derived from Framingham data estimated that 10-year
incidence of fatal and nonfatal CHD was 1% or less for women ages 30 to 39
years who were low risk (ie, no diagnosed diabetes, nonsmoker, with optimal
BP and cholesterol levels).2 In the Nurses’
Health Study, low-risk status was defined as nonsmoking, moderate/vigorous
exercise 30 minutes or more per day, BMI less than 25, moderate alcohol consumption,
and a diet score in the top 2 quintiles. Low-risk women (only 3% of 84 129
women, baseline ages 30-55 years) had an 83% lower 14-year risk of fatal and
nonfatal coronary events compared with others.13
In a previous report among middle-aged women (baseline ages 40-59 years)
from the CHA study, CHD and CVD mortality rates for those at low risk were
lower by 79% and 73%, respectively, and all-cause mortality was lower by 40%
compared with those with adverse levels of 1 or more risk factors.1 With over 3 decades of follow-up, our results now
reveal the importance of a low-risk profile in young women. Of note, our study
incorporates BMI in the low-risk definition since overweight/obesity is a
major independent CVD risk factor that is increasingly prevalent among Americans.14- 16
Limitations of this study include measurement of risk factors once only
at baseline, which would likely lead to underestimation of the impact of risk
status, ie, regression dilution bias toward the null. Furthermore, information
on dietary habits and physical activity was not collected. Also, the use of
death certificates for ascertaining cause of death may result in overestimation
of CHD mortality.17 Nevertheless, it is unlikely
that this would differ systematically across risk strata.
Our findings show that for young women, a low cardiovascular risk profile
is associated with lower long-term CHD, CVD, and all-cause mortality—results
in concert with previous findings on young men and middle-aged men and women.1 They demonstrate that among persons at low risk earlier
in life, CHD and CVD cease to occur at epidemic rates. These data underscore
the importance of a national public policy priority emphasizing prevention
and control of all major CVD risk factors by lifestyle approaches from conception,
weaning, childhood, and youth on to increase proportions of the population
at low CVD risk.
Corresponding Author: Martha L. Daviglus,
MD, PhD, Department of Preventive Medicine, Northwestern University, Feinberg
School of Medicine, 680 N Lake Shore Dr, Suite 1102, Chicago, IL 60611 (email@example.com).
Author Contributions: Dr Daviglus had full
access to all of the data in the study and takes responsibility for the integrity
of the data and the accuracy of the data analysis.
Study concept and design: Daviglus, Stamler,
Liu, Dyer, Greenland.
Acquisition of data: Stamler, Garside, Greenland.
Analysis and interpretation of data: Daviglus,Stamler,
Pirzada, Yan, Wang, Dyer, Lloyd-Jones, Greenland.
Drafting of the manuscript: Daviglus, Stamler,Pirzada,
Critical revision of the manuscript for important
intellectual content: Daviglus, Stamler, Yan, Garside, Wang, Dyer,
Statistical analysis: Stamler, Garside, Liu,
Obtained funding: Stamler, Dyer, Greenland.
Administrative, technical, or material support:Stamler,
Study supervision: Daviglus, Stamler.
Funding/Support: This research was supported
by grants from the National Heart, Lung, and Blood Institute (R01 HL21010);
the Illinois Regional Medical Program; the Chicago Health Research Foundation;
and private donors.
Role of the Sponsor: The funding organizations
did not have input in the design or conduct or the study; the collection,
analysis, or interpretation of the data; or the preparation, review, or approval
of the manuscript.