Determinants and prognostic significance of electrocardiographic left ventricular hypertrophy criteria in chronic kidney disease.

BACKGROUND AND OBJECTIVES The diagnosis of left ventricular hypertrophy (LVH) has prognostic value in the general population. However, among those with chronic kidney disease (CKD), the determinants of electrocardiographic (EKG) LVH and its prognostic value are not clear. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS A cross-sectional study was performed among 387 consenting consecutive patients from a veterans hospital with a longitudinal follow-up. RESULTS The overall prevalence of EKG-LVH by the Sokolow-Lyon criteria was 8% and by the Cornell voltage-duration product was 11%. Compared with non-CKD controls, CKD patients had unadjusted odds ratio (OR) for LVH by Cornell criteria of 2.52 (95% CI 1.18 to 5.42). Significance was lost after adjustment. The unadjusted OR for LVH by Sokolow-Lyon criteria was 2.24 (95% CI 0.95 to 5.33). This OR remained statistically insignificant after multivariate adjustment. Anemia, proteinuria, and 24-hour ambulatory systolic BP were associated with EKG-LVH regardless of diagnostic criteria. After a 7.5-year median follow-up, the hazard ratio for all-cause mortality was not associated with EKG-LVH diagnosed by the Sokolow-Lyon criteria; however, multivariable adjustments made EKG-LVH significant. A statistically significant relationship was seen between mortality and Cornell criteria; however, multivariable adjustments made EKG-LVH nonsignificant. CONCLUSIONS The Sokolow-Lyon and Cornell EKG-LVH criteria cannot be used interchangeably to diagnose LVH or determine prognosis. Among those with CKD, ambulatory systolic BP predicts all-cause mortality. Moreover, the duration and severity of BP elevation presumably reflected in EKG-LVH diagnosed by Sokolow-Lyon criteria is also of prognostic significance; the Cornell criteria do not carry independent prognostic information.