ECG voltage criteria caught only 64% of real left ventricular hypertrophy

A routine ECG printing possible left ventricular hypertrophy is a common event, and it lands harder on people who lift heavy. But when that finding is checked directly against echocardiography, the ECG turns out to be a poor instrument for the question.

A sub-analysis of the coArtHA trial ran 12-lead ECG and focused echocardiography on 1,125 adults with untreated, uncomplicated hypertension in rural Tanzania and Lesotho. Hypertrophy was defined by left ventricular mass index above 95 g/m² in women and 115 g/m² in men. Only 56 participants (5%) actually met it.

How accurate were the ECG criteria?

The most consistent performer across metrics was the Cornell voltage product family. Its numbers still stopped here:

• AUROC ≈ 0.76 — discriminating, but not diagnostic
• Sensitivity 64% — more than one real case in three is missed
• Specificity 80% — one in five people without hypertrophy screens positive
• With a +0.8 mV adjustment in women, correlation with mass index was rho 0.37

Voltage-only criteria — Sokolow-Lyon and MESA-based indices — did worse in this cohort. The authors' conclusion matches: performance swings sharply with the cut-off chosen, so criteria cannot be transported between populations without local validation.

Why this is murkier for lifters

ECG voltage reflects not only muscle thickness but whatever sits between the heart and the electrodes. Lean people with thin chest walls read high; body fat damps the signal down. Add a heart genuinely enlarged by training and voltage criteria drift toward false positives — while, as the data above show, real hypertrophy often fails to cross the threshold at all.

One line on an ECG therefore cannot separate a heart enlarged by training from a wall thickened by blood pressure. How those two differ at the molecular level is covered in physiological versus pathological hypertrophy, and the pressure pathway itself in how blood pressure thickens the heart wall.

What to do when the finding appears

The order is simple. An ECG is a screen; the test that measures wall thickness and cardiac mass is an echocardiogram. A positive finding should be confirmed rather than self-interpreted — and this data set gives no grounds for treating a negative ECG as reassurance either. How ECG hypertrophy findings connect to outcomes is covered separately in ECG left ventricular hypertrophy and sudden death risk.

Should training change?

Neither stopping over an unconfirmed finding nor ignoring it and chasing a max is a good answer. Until the echocardiogram is back, the practical middle is holding normal training intensity while deferring max attempts that involve a long Valsalva. High blood pressure itself is a separate problem to address on its own terms; the training dose that moves it is in the exercise dose for blood pressure.