Update on factor V Leiden association with venous thromboembolism in the LITE Study

In a previous article,¹  we reported population-based findings from the Longitudinal Investigation of Thromboembolism Etiology (LITE) on factor V Leiden and risk of venous thromboembolism (VTE). We used a nested case-control design (301 new VTE cases and 630 controls through 1998) from the prospective Atherosclerosis Risk in Communities (ARIC) Study and Cardiovascular Health Study (CHS). The odds ratio, overall, for risk of VTE for carriers of factor V Leiden was 3.67 (95% CI, 2.20-6.12).¹ 

We recently extended LITE to additional cases and controls through 2002. In the process, we discovered and corrected an error in the original selection of controls for ARIC that had caused us to oversample participants who had died into the control group for our earlier report. Compared with our published report, the updated sample of 502 cases and 1021 controls yielded an almost identical odds ratio for factor V Leiden of 3.46 (95% CI, 2.20-5.43). Genotype distributions and other odds ratios for factor V Leiden (Tables 1–2) are also similar to those in our original report's Tables 1–2.¹  The estimate for African Americans remains quite imprecise from this large prospective study because of the infrequency of factor V Leiden in African Americans.

We were unable to repeat other measures in our original report¹  (activated protein C resistance, factor V concentration, and HRZ haplotype) to confirm no influence of the control selection error. Nevertheless, these factor V Leiden results offer confidence of the accuracy of our original report.¹ 

Correspondence: Aaron R. Folsom, University of Minnesota, School of Public Health, 1300 South Second St, Ste 300, Minneapolis, MN 55454-1015; e-mail: folsom@epi.umn.edu.