Cornell Product for Left Ventricular Hypertrophy
The Cornell product is calculated by multiplying the Cornell voltage (S wave in V3 + R wave in aVL, with +0.6 mV added for females) by the QRS duration in milliseconds, with a diagnostic threshold of >2440 mm·ms indicating LVH. 1
Calculation Formula
- Cornell voltage = S wave in V3 + R wave in aVL 2
- For females, add 0.6 mV (6 mm) to the Cornell voltage before multiplying 3
- Cornell product = Cornell voltage (in mm) × QRS duration (in ms) 1
- Diagnostic threshold: >2440 mm·ms indicates LVH 1, 4
Performance Characteristics
- The Cornell product has high specificity (85-90%) but relatively low sensitivity (<50%) for detecting anatomical LVH 1
- The Cornell product shows stronger correlation with left ventricular mass index (r=0.392) compared to Sokolow-Lyon voltage (r=0.315) 3
- The Cornell product is independently associated with global longitudinal strain and regional wall motion abnormalities, even after adjusting for other ECG criteria 3
Clinical Significance
- The Cornell product predicts cardiovascular and stroke mortality, with particular predictive value even in normotensive and prehypertensive populations 5
- In Japanese populations, cardiovascular risk increases significantly when Cornell product exceeds 158.7-193.4 mm·ms (4th quintile), which is substantially lower than the Western diagnostic threshold 6
- The Cornell product reflects not only LV mass but also LV geometry and diastolic function 7
- Approximately 23-24% of hypertensive patients in clinical practice have LVH detected by Cornell product 4
Important Caveats
- Only 11.2% of patients meeting Cornell voltage criteria also meet Sokolow-Lyon criteria, indicating these criteria identify different patient populations 2
- The Cornell product performs better in females than males for predicting cardiovascular outcomes 6
- QRS voltages decline with increasing age, and standard criteria apply best to adults over 35 years 2
- Obesity increases left ventricular mass but paradoxically decreases QRS voltage, potentially reducing sensitivity 2