Framingham, MA- Investigators from the longitudinal study bearing this city's name have devised a risk-factor scoring system that allows individual patients without atrial fibrillation (AF) to predict their risk of developing the arrhythmia over 10 years .
Based on follow-up data from >4000 members of the Framingham cohort, the scoring system could potentially help "individuals or their physicians understand what their specific risk is over 10 years using simple risk factors that are easily assessed in a primary-care office, things that don't cost a lot [to identify or measure]," observed Dr Emelia J Benjamin (Boston University, MA) for heartwire.
A number of individual risk factors for AF have been proposed, but Benjamin says this is probably the first time they've been combined into this kind of system for predicting a person's absolute risk.
She is senior author on the report outlining the system, which appears in the February 28, 2009 issue of the Lancet with lead author Dr Renate B Schnabel (Johannes Gutenberg-University, Mainz, Germany).
Benjamin acknowledges that the AF-risk scores are currently likely to be more useful in a research setting than in clinical practice; there aren't any accepted treatments for preventing AF that have randomized, controlled trial support.
"We know a tremendous amount about how to prevent coronary heart disease. But the lifetime risk of atrial fib is about one in four, similar to the lifetime risk of heart failure, and we really know very little about how to prevent the onset of atrial fibrillation," she said.
On the other hand, there is tantalizing observational evidence that patients on statins or ACE inhibitors or those with high intake of omega-3 fatty acids may be at reduced risk of developing AF. It might be reasonable, then, to explore those possible treatments in primary-prevention trials, which traditionally are conducted first in high-risk populations, Benjamin notes.
The scoring system could make a contribution there, she said. "At least now we can say who's at increased risk of atrial fibrillation, and now we can talk about individual risk rather than aggregate risk. I think this sets up those types of trials."
According to an accompanying editorial from Drs David B Brieger and S Ben Freedman (University of Sydney, Australia), "primary prevention of atrial fibrillation has not been on our radar; even secondary prevention of the arrhythmia has not been recommended," since randomized trials have failed to show it superior to rate control for most patients .
"Thus, we have accepted that, in many settings, atrial fibrillation is an unavoidable evil and we are preoccupied with preventing complications, such as heart failure and stroke"—although, they write, changes may be in store given recent results of the ATHENA trial , in which secondary prevention of AF with the drug dronedarone significantly improved clinical outcomes.
"That effective primary prevention for this condition should have the same effect seems intuitive," Brieger and Freedman write. The AF-risk prediction model from Schnabel et al "is the first step in that direction."
Of 4764 persons aged >45 and without AF examined as part of the longitudinal study between 1968 and 1987, 10% developed AF over 10 years of follow-up.
In multivariate analyses, baseline patient features significantly and most closely related to future AF included age, sex, body-mass index, systolic blood pressure, treatment for hypertension, electrocardiographic PR interval, clinically important cardiac murmur, and heart failure. Devised from those risk factors, the scoring system assigned, for example, more points with advancing age (according to different scales for men and women) and more points the younger patients are at diagnosis of heart failure or a clinically important murmur.
Score totals of 5 or 8 indicated 6% or 16% 10-year risks for AF, respectively, for example, and a score of >10 meant a >30% risk. According to the system, 1% of patients in the cohort <65>65 years had >15% risks for AF, the group reports.
They validated the prediction model in a later cohort of 5152 Framingham participants also followed for 10 years. Incorporation of M-mode echocardiographic parameters from this second cohort didn't greatly improve the overall risk-prediction model.
Echocardiography, Benjamin said, "didn't particularly help with the average patient, but that's not to say that it can't be helpful in a specific context. Of 18 clinical subgroups that we looked at, the only ones in which it looked helpful were patients with established valvular disease or heart failure."
There are ways the scoring system could conceivably be of help to patients now, she proposed. For example, "for a patient at increased risk, say one with heart failure or who is obese or older, it might be worthwhile teaching them how to take their pulse, for example. What you don't want is for them to go into atrial fib and have the first manifestation be stroke."
Artigo original: Schnabel RB, Sullivan LM, Levy D, et al. Development of a risk score for atrial fibrillation (Framingham Heart Study): a community-based cohort study. Lancet 2009; 373:739-745.