Context: Several reports have suggested the usefulness of plasma brain natriuretic peptide (BNP) as a screening test for left ventricular hypertrophy (LVH) and systolic dysfunction (LVSD). Prior studies were limited by small sample sizes and selection bias and none compared the diagnostic performance of these peptides in men and women.

Objectives: To examine the usefulness of natriuretic peptides for screening for elevated LV mass and LVSD in the community.

Design, setting, and participants: Community-based prospective cohort study of 3177 participants (1707 women) from the Framingham Study who attended a routine examination in 1995-1998.

Main outcome measures: Receiver operating characteristic (ROC) curves, test sensitivity, specificity, positive and negative predictive values, and likelihood ratios for identifying elevated LV mass (sex-specific 90th percentile or higher of LV mass/[height](2)), LVSD (ejection fraction < or = 50% and/or fractional shortening <29%), and moderate to severe LVSD (ejection fraction < or = 40% and/or fractional shortening <22%) at different discrimination limits of plasma BNP and N-terminal proatrial natriuretic peptide (NT-ANP), with echocardiography as the criterion standard.

Results: The areas under the ROC curves for elevated LV mass or LVSD were at or below 0.75 for both peptides, were higher for men compared with women, and were similar for BNP and NT-ANP. The diagnostic performance of natriuretic peptides for LVSD improved in women but not in men when select high-risk subgroups were targeted. Discrimination limits based on high specificity (0.95) yielded better positive predictive values and likelihood ratios compared with age- and sex-specific reference limits yet only identified less than one third of participants who had elevated LV mass or LVSD.

Conclusion: In our large community-based sample, the performance of BNP and NT-ANP for detection of elevated LV mass and LVSD was suboptimal, suggesting limited usefulness of natriuretic peptides as mass screening tools.