Abstract 15715: The Added Value of Female-Specific Risk Factors on Top of the Framingham Risk Score for the Prediction of Cardiovascular Disease in Women: Data From EPIC-NL
Introduction: Besides the traditional risk factors for CVD, women have female-specific risk factors, related to their hormonal and reproductive status. Whether these female-specific risk factors have added value in risk prediction on top of the traditional risk factors has not been evaluated yet.
Hypothesis: We assessed the hypothesis that female-specific risk factors will have added value on top of the Framingham risk score (FRS) in predicting CVD.
Methods: We used data from 25,513 women between 30-74 years of age (mean age 53.5±9.0 years) without prevalent CVD in the EPIC-NL study, a population-based cohort study from the Netherlands. Outcome was 10-year risk of fatal or non-fatal CVD. The FRS variables were first refitted using a Cox proportional hazards model. Subsequently, number of children, gestational hypertension or diabetes, age at menarche and menopause, estrogen use and miscarriages/stillbirths were added. Stepwise backward selection using Akaike’s Information Criterion was used to evaluate which female-specific risk factors were retained in the model. Discrimination, calibration and reclassification (NRI) were determined.
Results: During a median follow-up of 11.7 years 2,305 CVD events occurred. In the study population 71.5% were at low (<10%), 26.4% at intermediate (≥10-<20%), and 2.1% at high (≥20%) 10-year CVD risk. In univariable analysis, most female-specific risk factors were statistically significantly associated with CVD risk. The c-statistic of the refitted FRS was 0.69 (95% CI 0.67-0.71). Age at menarche, menopausal status and estrogen use were retained in the model, but discrimination and calibration did not improve. There was no net reclassification (NRI = 0). The most likely explanation is that the presence of female-specific risk factors is already expressed in the traditional risk factors such as high cholesterol and blood pressure and the presence of diabetes.
Conclusion: In conclusion, although female-specific risk factors are associated with CVD risk they have no added value on top of the FRS for the prediction of 10-year risk of CVD in women.
Author Disclosures: M.G. van der Meer: None. Y. van der Graaf: None. E. Schuit: None. W.M. Verschuren: None. J.M. Boer: None. L.M. Peelen: None. K.G. Moons: None. H.M. Nathoe: None. Y. Appelman: None. Y.T. van der Schouw: None.
- © 2014 by American Heart Association, Inc.