October 27th, 2014
Traditional CVD Risk Factors Mediate Between Lifestyle Factors and Cardiovascular Risk in Women
CardioExchange’s Harlan M. Krumholz interviews Nina P. Paynter about her research group’s case-cohort study comparing lifestyle-based with traditional cardiovascular disease prediction among women in the Women’s Health Initiative Observational Study. The article, published in Circulation, includes complete lists of the lifestyle-based and traditional risk factors.
Krumholz: This is a very interesting study. Please describe your main results for our readers.
Paynter: In a multiethnic cohort of nonsmoking postmenopausal women, when lifestyle factors were added to traditional risk factors in the “pooled cohort” risk score (from the most recent ACC/AHA guidelines) or the Reynolds risk score, only recreational physical activity remained independently associated with 10-year cardiovascular disease (CVD) risk. A greater number of healthy lifestyle factors reduced the risk for CVD — and adding lifestyle factors to traditional factors did improve the overall model fit but had little effect on risk-stratification measures.
Krumholz: So healthy lifestyle factors did not add to predictive models. Does that mean that lifestyle has little effect on risk, beyond how it affects traditional risk factors?
Paynter: Our results suggest that knowing someone’s lifestyle today, in addition to his or her traditional risk factors today, does not change the category of 10-year CVD risk for most people (i.e., whether they are high, intermediate, or low risk today). However, physical activity had an independent effect on risk, even after controlling for traditional risk factors.
Krumholz: Does this mean that if someone has his or her traditional CVD risk factors under control, lifestyle contributes little to the person’s future risk?
Paynter: Much of the effect of lifestyle factors is through the traditional risk factors, so a healthy lifestyle may explain the control of the traditional risk factors. Although we did not directly study this issue, control of risk factors via lifestyle may be more beneficial than medical treatment is.
Krumholz: You focused on whether risk models should incorporate lifestyle factors, but isn’t the finding more profound for what we should recommend to patients? It seems that, in the end, these factors contribute relatively little to future risk, beyond traditional factors.
Paynter: Our analysis (and assessment of prediction in general) is not set up to look at causal contributions. One factor might be a great predictor and not at all causal, whereas another factor might be causal and not a great predictor. Risk prediction, though an important tool, is only one part of decision making about treatment. In this case, there is a wide consensus that lifestyle factors significantly affect health, and although they may not dramatically change predictions of 10-year risk, they do matter for treatment recommendations. As mentioned above, the traditional risk factors are mediators between lifestyle behaviors and CVD risk — by improving their lifestyle, patients may reduce their risk without the need for medication.
JOIN THE DISCUSSION
How will Dr. Paynter’s findings affect how you advise patients about lifestyle-based and traditional CVD risk factors?