What are the electrocardiographic criteria for diagnosing left ventricular hypertrophy?

ECG Criteria for Left Ventricular Hypertrophy

Primary Recommendation

Use multiple validated voltage criteria simultaneously—specifically the Sokolow-Lyon Index (SV1 + RV5/V6 ≥35 mm), Cornell Voltage (SV3 + RaVL >28 mm in men, >20 mm in women), and the Romhilt-Estes Point Score System—because only 11% of patients meeting one criterion will meet another, making their combined use additive for diagnostic accuracy. 1, 2

Core Voltage-Based Criteria

The following criteria have been validated by the American College of Cardiology and American Heart Association with high specificity (85-90%) but characteristically low sensitivity (<50%): 3, 1, 2

• Sokolow-Lyon Index: SV1 + RV5 or RV6 ≥35 mm 1, 2
• Cornell Voltage: SV3 + RaVL >28 mm (men) or >20 mm (women) 1, 2
• Cornell Product: Cornell voltage × QRS duration >2440 mm·ms 2, 4
• RaVL alone: >11 mm 1, 4

The Cornell product provides the highest overall diagnostic accuracy among ECG methods when QRS duration is incorporated. 1

Romhilt-Estes Point Score System

This scoring system integrates multiple ECG features, with ≥5 points indicating definite LVH and 4 points indicating probable LVH: 1, 2

• Voltage criteria (QRS amplitude abnormalities)
• ST-T wave abnormalities (secondary repolarization changes)
• Left axis deviation
• QRS duration prolongation
• Intrinsicoid deflection delay (QRS onset-to-peak time in V5/V6)
• P-wave abnormalities (left atrial enlargement)

Supporting ECG Features That Increase Diagnostic Confidence

• ST-T wave abnormalities (lateral ST depression with T wave inversion) strongly support LVH diagnosis and indicate larger left ventricular mass with higher cardiovascular risk 1, 2, 4
• Left atrial abnormality (P-wave changes) frequently accompanies LVH and may be the earliest ECG sign of hypertensive heart disease, even before voltage criteria are met 1, 2, 4
• QRS duration prolongation results from increased ventricular wall thickness and intramural fibrosis that prolongs transmural impulse propagation 3, 1, 2
• Left axis deviation may support the diagnosis but should never be used in isolation 2

Mandatory Demographic Adjustments

Age Considerations

Standard voltage criteria apply only to adults >35 years old; applying them to younger individuals (16-35 years) or highly trained athletes yields markedly reduced accuracy and high false-positive rates. 3, 1, 2

Gender-Specific Thresholds

Women have lower upper limits for QRS voltage, particularly in lead V3, requiring gender-adjusted cut-offs to avoid misclassification: 3, 1, 2

• Cornell voltage: >28 mm (men) vs >20 mm (women)
• Gender-specific thresholds improve both sensitivity and specificity

Body Habitus Impact

Obesity dramatically decreases ECG sensitivity for LVH despite actual increases in left ventricular mass by echocardiography, likely due to the insulating effect of adipose tissue increasing the distance between the heart and recording electrodes. 3, 1, 2

Practical Implementation Algorithm

Step 1: Measure all three primary voltage criteria (Sokolow-Lyon, Cornell, RaVL) and document which are met 1, 2

Step 2: Calculate the Cornell product (Cornell voltage × QRS duration) for highest diagnostic accuracy 1

Step 3: Apply the Romhilt-Estes point score incorporating voltage, ST-T changes, axis, QRS duration, and P-wave abnormalities 1, 2

Step 4: Apply gender-specific voltage thresholds to avoid sex-based misclassification 1, 2

Step 5: Assess supporting features including ST-T abnormalities, left atrial P-wave changes, and QRS prolongation 1, 2

Step 6: Adjust interpretation for age (<35 years: reduced reliability) and body habitus (obesity: reduced sensitivity) 3, 1, 2

Step 7: Use only validated criteria without ad-hoc modifications, and specify which criteria were used in documentation 1, 2

Special Circumstances: Bundle Branch Blocks

Complete Left Bundle Branch Block (LBBB)

ECG diagnosis of LVH should generally NOT be attempted in complete LBBB; if diagnosis is necessary, use QRS duration >155 ms combined with left atrial abnormality and precordial voltage criteria. 1, 2

Right Bundle Branch Block (RBBB)

Voltage criteria sensitivity is decreased, but the following can support diagnosis: 1, 2

• SV1 >2 mm
• RV5 or V6 >15 mm
• QRS axis left of -30° with SIII >largest R/S in precordial lead >30 mm

Left Anterior Fascicular Block (LAFB)

R wave amplitude in leads I and aVL are unreliable; instead use criteria incorporating S wave depth in left precordial leads (V5, V6). 1

Critical Pitfalls to Avoid

• Relying on a single criterion: Different voltage criteria capture distinct patient subsets; only 11.2% of patients meeting either Cornell or Sokolow-Lyon meet both criteria simultaneously 3, 1, 2
• Applying adult thresholds to patients <35 years: This yields unacceptably high false-positive rates, especially in athletes 3, 1, 2
• Neglecting sex-specific cut-offs: Using male thresholds in women reduces specificity and misses true LVH cases 3, 1, 2
• Ignoring obesity's effect: Obese patients have increased left ventricular mass but paradoxically decreased QRS voltage 3, 1, 2
• Modifying validated formulas: Only use ECG criteria exactly as validated in published studies without deviation 3, 1, 2

Diagnostic Performance Characteristics

All voltage-based criteria demonstrate: 3, 1, 2, 5

• High specificity: 85-90% (false positives are uncommon)
• Low sensitivity: Usually <50% (many true LVH cases are missed)
• Positive predictive value: Varies by population prevalence
• Negative predictive value: In hypertensive populations, most negative tests are false negatives

The low sensitivity means ECG is better suited for confirming rather than screening for LVH, particularly in low-complexity services without echocardiography access. 6

Prognostic Significance

LVH detected by ECG is an independent predictor of cardiovascular events in patients >55 years old, and regression of left ventricular mass on ECG predicts lower risk of major cardiovascular events. 4