Landmark Biomarker Studies Are Likely to Overstate Findings

Results of frequently cited studies often differ markedly from those of meta-analyses or studies in larger populations.

Each year, a growing number of studies tout the effectiveness of various biomarkers to monitor health or predict disease outcomes. To investigate the possible effect of citation bias on biomarker research, investigators identified published biomarker studies that were cited >400 times in other published articles and compared the findings from the frequently cited studies with those of subsequent meta-analyses of multiple studies of the same biomarkers.

The demonstrated link between biomarker and disease was stronger in 29 of 35 frequently cited studies than in the subsequent meta-analyses. Furthermore, 30 of the 35 frequently cited studies reported a stronger association than was observed in the largest single study included in the meta-anal…

Reviewing Author

JoAnne M. Foody, MD

Disclosures

Consultant / Advisory boardAegerion; Amarin; Bristol-Myers Squibb; Janssen; Merck; Pfizer; Sanofi-Aventis; Boehringer Ingelheim

Leadership positions in professional societiesAmerican College of Cardiology, Cardiosmart (Co-chair)

JoAnne M. Foody, MD

Disclosures

Consultant / Advisory boardAegerion; Amarin; Bristol-Myers Squibb; Janssen; Merck; Pfizer; Sanofi-Aventis; Boehringer Ingelheim

Leadership positions in professional societiesAmerican College of Cardiology, Cardiosmart (Co-chair)

Landmark Biomarker Studies Are Likely to Overstate Findings

Topics

Reviewing Author