Left ventricular hypertrophy (LVH) diagnosed by Cornell product in electrocardiogram predicted future stroke events at lower thresholds in Japanese general populations than those in western countries; therefore, we thought to evaluate whether the gender-adjustment and the levels of Cornell product to detect echo LVH was appropriate in Japanese subjects.Design and Method:
We consecutively enrolled 345 hypertensive patients who were performed echocardiography for the screening of LVH. We evaluated Cornell voltage (S in leads V3 + R in leads aVL), Cornell product [(Cornell voltage, added 0.6 mV if females) × QRS duration] and left ventricular mass index (LVMI) (echo LVH; LVMI 116 > g/m2 in males and LVMI 96 > g/m2 in females.Results:
The mean age was 63.8 ± 12.5 years (males/females 174/172). There were 39 patients with echo LVH in males (22.7 %) and 64 patients in females (37.2 %). In linear regression model, the equations to estimate LVMI from Cornell voltage were as follows: (1) LVMI = 14.5 × Cornell voltage + 78.9 in males and (2) 21.5 × Cornell voltage + 61.5 in females. The cutoff levels to estimate echo LVH were 2.6 mV in males and 1.6 mV in females; the levels were lower than levels in the current guideline. In both gender, the area under curve (AUC) to detect echo LVH was increased from the gender-adjusted Cornell product (AUC = 0.690) to the gender-unadjusted Cornell product (AUC = 0.738). The equations to estimate LVMI from the gender-adjusted Cornell product were as follows: (1) LVMI = 0.15 × Cornell product + 68.8. In the receiver operating curve, the levels of Cornell product to estimated echo LVH were 170 mV × msec (sensitivity 0.74, specificity 0.60); these levels were also lower than the levels in the current guideline.Conclusions:
The gender adjustment of Cornell product seems to be appropriate even in Japanese patients, but the levels detecting echo LVH may be lower than the levels in western countries.