Left ventricular hypertrophy: is it reversible, and by what exactly

Left ventricular hypertrophy is neither a diagnosis nor a symptom. It is a description of what the machine saw: the wall of the heart muscle is thicker than expected, and the calculated myocardial mass is above a threshold. The phrase lands in an echocardiography report or an ECG readout, produces no sensation of its own, and is almost always found while looking for something else.

The question that follows is not "what is it" but "is this permanent". Medicine has an answer, and it differs by cause: some forms are reversible and their regression has been measured, some are not reversible in principle, and some are not disease at all. Here is what is known, in whom it was shown, and by what design.

One word, several different things

"Hypertrophy" describes a shape, not a cause. The same wall thickness can sit on top of conditions with different natures and different behaviour.

• Hypertensive, from pressure overload. The commonest form in adults: the muscle works against raised resistance and thickens, the way any muscle thickens under load. This is where the question of reversibility is meaningful.

• Athlete's heart. In endurance athletes, myocardial mass and wall thickness exceed population values, and this is adaptation rather than damage.

• Hypertrophic cardiomyopathy. Here the thickening is primary rather than a response to load. It is defined by phenotype — unexplained wall thickness ≥15 mm, or ≥13 mm with a family history (ESC cardiomyopathy guidelines 2023, PMID 37622657); the phenotype is estimated at roughly 1 in 500. 🔴 It is often called "a disease of the sarcomere", but pathogenic sarcomeric variants are not found in everyone: across series, in roughly 30–60% of clinical cases.

• Infiltrative disease, above all cardiac amyloidosis. The wall looks thickened, but what thickens it is not muscle — it is protein deposited between the fibres. On echocardiography it resembles hypertrophy; in substance it is not.

• Valve disease. Aortic stenosis creates the same pressure overload as hypertension, except the source sits inside the heart.

How far this is from a theoretical distinction shows in the share of amyloidosis among people whose wall is thickened "unexplainedly": about 5% hereditary transthyretin amyloidosis in an unselected cardiology sample with increased thickness (PMID 26537620), about 9% among those referred with a diagnosis of hypertrophic cardiomyopathy after the age of forty (PMID 31748190), and about 13% among patients aged 60+ hospitalised with heart failure with preserved ejection fraction and a wall ≥12 mm (PMID 26264527). There is no single figure here: the share depends entirely on who was selected.

Telling these apart is a clinician's job, and everything downstream depends on it: in some the cause can be removed, in others the underlying disease is treated, in others there is nothing to treat.

What measures it, and why the numbers disagree

ECG judges hypertrophy indirectly, from wave amplitude (Sokolow–Lyon and Cornell criteria). It is cheap and universal, and insensitive: in a systematic review of 21 studies in 5608 patients with hypertension, median sensitivity was 21% for Sokolow–Lyon and about 15% for Cornell voltage, at specificities of roughly 89% and 93–97% respectively (PMID 17726091). 🔴 So a normal ECG does not exclude hypertrophy — it simply fails to see it most of the time. The ECG picture nonetheless carries prognostic information of its own, and the largest trial in this field recruited people precisely on it.

Echocardiography calculates myocardial mass from linear measurements — interventricular septum, posterior wall and end-diastolic dimension — using the Devereux formula, validated against autopsy (PMID 2936235). Mass is divided by body surface area to give the left ventricular mass index. Current ASE/EACVI chamber-quantification thresholds: above 115 g/m² in men and above 95 g/m² in women for the linear method, with relative wall thickness above 0.42 separating concentric from eccentric geometry (PMID 25559473). The 2025 documents on strain and report standardisation do not replace these thresholds.

Cardiac magnetic resonance is the reference method for reproducibility: it does not lean on the geometric assumptions of a formula. In a direct comparison, the coefficient of variability for mass was 2.8–4.8% for CMR against 11.6–15.7% for echocardiography (60 subjects, PMID 12142111). Studies that need precision in a change of mass use it.

The practical consequence is plain: only numbers obtained by the same method — ideally on the same machine — are worth comparing. A difference between two echocardiograms from two different rooms can be explained entirely by measurement rather than by the heart. Same principle as with any other tracked value: a series beats a single point.

Why anyone looks at this at all

Wall thickening is not a cosmetic finding. In the Framingham cohort, where echocardiographic myocardial mass was measured in 3220 people over 40 without clinically overt heart disease and followed for four years, each additional 50 g/m of height-indexed mass corresponded to a relative risk of cardiovascular events of 1.49 in men and 1.57 in women — after adjustment for age, blood pressure, treatment for hypertension and other known factors (PMID 2139921). Cohorts of people with essential hypertension specifically show the same: 140 men over a mean 4.8 years (PMID 2942070), and 280 patients of whom 253 were followed a mean 10.2 years — there, the only independent predictors left in the multivariable model were age and myocardial mass, not the blood pressure level (PMID 1825164).

🔴 The caveat that makes the statement honest: this is an observational association. It says people with greater myocardial mass have more events; it does not say the thickening itself is the cause. Mechanisms are plausible — fibrosis, impaired relaxation, electrical heterogeneity — but a cohort is not what confirms a mechanism.

Is it reversible, and what has been shown

For the hypertensive form the answer is yes at the level of the measured value: when blood pressure falls, myocardial mass decreases in randomised trials.

The main summary is a meta-analysis of double-blind randomised trials in essential hypertension: 80 trials, 146 active arms, 3767 patients. After adjustment for treatment duration and change in diastolic pressure, left ventricular mass index fell by 13% on angiotensin receptor blockers, 11% on calcium antagonists, 10% on ACE inhibitors, 8% on diuretics and 6% on beta-blockers (PMID 12867233).

🔴 It is easy to read more into this than it says. In pairwise comparisons corrected for multiplicity, three classes were significantly better than beta-blockers — ARBs, calcium antagonists and ACE inhibitors; the differences among those three were not significant. A later meta-analysis of 84 pairwise comparisons in 6001 patients gave the same picture and put it plainly: what is convincing is not the superiority of any one class but the shortfall of beta-blockers (PMID 19770405). And all of this concerns myocardial mass, not survival.

A second line is not about pills:

• Weight loss. Randomised data with MRI exist here: in a trial of two hypocaloric diets, left ventricular mass fell by about 5 g over six months with 6–8% weight loss, equally on both diets (90 participants with complete MRI, PMID 22371331). In a one-year trial in 383 adults aged 18–55 with about 10% weight loss, mass fell significantly only in the diet-only group — by roughly 3 g; where physical activity was added to diet, mass was preserved (PMID 35112748). After bariatric surgery the reduction is larger — across 80 studies in 3332 patients, about 31 g of absolute mass and 12% of the index — but those are before-and-after series, not randomisation.

• Sodium restriction. The evidence here is weaker than commonly assumed. Twenty-four-hour sodium excretion correlates with wall thickness independently of diastolic pressure, but that is a cross-sectional study, an association rather than an effect (PMID 2971473). The closest thing to an intervention is a one-year trial of a non-pharmacological programme in 76 previously untreated patients, where the mass index fell by 4.7%; but weight fell along with salt, so the separate contribution of sodium cannot be isolated from that work (PMID 8124786). Trials changing sodium alone and measuring myocardial mass essentially do not exist.

• Removing the cause when it is structural. After aortic valve replacement for severe stenosis, mass falls by roughly 15–20%, mostly within the first six months: a review of 27 studies in 1546 patients (PMID 15223407) and an MRI meta-analysis after transcatheter replacement with a median of −15.1 g/m² (PMID 32487183). Here the source of the overload itself has been removed.

• Sleep apnoea — an important negative example. Obstructive sleep apnoea itself is associated with hypertrophy in observational data: across 39 studies, an odds ratio of about 1.70 (PMID 32371769). But continuous positive airway pressure does not undo it: in a pooled estimate across 385 patients the change in left ventricular mass was −0.22 standard deviations with a confidence interval from −0.56 to 0.12 — that is, no significant effect was shown, even though strain measures did improve. Association with a condition and reversibility on treating it are different questions, and here the answers diverge.

🔴 What is absent from this list: no supplement, vitamin or herbal preparation has a randomised trial in which it reduced myocardial mass in people with hypertrophy. This was checked deliberately. Omega-3s were studied after myocardial infarction with end-systolic volume as the primary endpoint, not mass (PMID 27482002); in an ancillary analysis of a large prevention trial, neither vitamin D nor omega-3s changed left ventricular structure over two years. Coenzyme Q10, taurine, L-carnitine, magnesium and hawthorn were studied against other endpoints — ejection fraction, symptoms, exercise tolerance — and in other populations. The "cardioprotector" shelf exists not because something on it worked against this endpoint, but because a frightening, comprehensible line in a report creates demand.

A separate story is hypertrophic cardiomyopathy, where the logic of "remove the load" does not apply at all, yet pharmacological reduction of thickness does exist: in the MRI substudy of the mavacamten trial in symptomatic obstructive disease (35 patients, 30 weeks), mass index fell by 17.4 g/m² against 1.6 g/m² on placebo, and maximal wall thickness by 2.4 mm; fibrosis on late gadolinium enhancement did not change over that time (PMID 33550811). That is a different disease and a different treatment, and nothing here transfers to hypertensive hypertrophy.

Athlete's heart regresses, but not completely. Among 947 elite athletes, wall thickness exceeded 12 mm in only 16 — 1.7%, all men in endurance disciplines (PMID 1827661). When load is reduced, thickness goes within weeks to months: in a small series of six Olympians the septum fell from 13.8 to 10.5 mm over about 13 weeks (PMID 8435237). The cavity does worse: in 40 athletes 1–13 years after leaving sport, thickness had fallen 15% and normalised in all of them, while cavity size fell only 7% and remained 60 mm or more in 22% (PMID 11864953).

And is regression a better prognosis, or just a nicer number?

A separate question from the previous one. Moving a measurement and improving an outcome are not the same thing; the history of medicine is full of surrogate endpoints that moved without moving anything else.

The data are encouraging, but their design has to be named precisely. In the LIFE trial, 9193 patients aged 55–80 with hypertension and ECG evidence of hypertrophy received losartan or atenolol for a mean 4.8 years; with almost identical blood pressure reduction, the composite of cardiovascular death, myocardial infarction or stroke was less frequent on losartan — relative risk 0.87 (0.77–0.98), with the benefit falling mostly on stroke (PMID 11937178).

Separately, under a prespecified analysis plan, the severity of ECG hypertrophy was entered into the model as a time-varying value. Each one-standard-deviation reduction corresponded to a hazard ratio of 0.86 for the Cornell product and 0.83 for Sokolow–Lyon, and significance survived adjustment for achieved blood pressure and baseline severity (PMID 15547161). The echocardiographic substudy in 941 patients gives the same picture: 0.78 per 25.3 g/m² reduction in mass index (PMID 15547160).

🔴 And still, this is an analysis by state attained during treatment, not randomisation to it. Patients were not allocated at random into "regression" and "no regression" groups — the split happened on its own, and it could have happened for reasons themselves linked to prognosis: adherence, disease severity, comorbidity. So the correct wording is "regression is accompanied by fewer events", not "regression reduces risk". What LIFE randomised was the drug, not a hypertrophy target.

Trials allocating people at random to a strategy of "achieve regression" and counting death or major cardiovascular events have not been run. The nearest thing is small work with surrogate endpoints such as interstitial volume on MRI.

What the data do not say

• They do not say how long it takes. Remeasurement intervals in studies are months and years. Promising a specific timeline to a specific person is not possible.

• They do not say the regression is complete. The fall in mass comes first of all from the muscle cells themselves. Fibrosis is a separate process, and ordinary echocardiography does not see it at all: it is shown by late gadolinium enhancement and T1 mapping on MRI. After valve replacement, cellular hypertrophy regresses while focal mid-wall fibrosis stays (PMID 29471937). The diffuse interstitial component can partly decrease: in a small randomised trial in 35 patients with myocardial biopsy, collagen fraction fell on lisinopril and did not fall on hydrochlorothiazide at comparable blood pressure (PMID 10993857).

• They do not say a normalised number ends follow-up. The cause that produced the hypertrophy has not gone anywhere and needs the same monitoring it needed before.

What to do with these numbers

Take nothing on the strength of an article. The practical meaning comes down to a few things, each a conversation with a doctor rather than a solo action.

1. Establish which hypertrophy it is. The first question for a cardiologist is not "how do I shrink it" but "what is it from". Everything else follows from the answer, including whether extra tests are needed to rule out amyloidosis or cardiomyopathy.

2. Know your numbers, not the phrase. "LV hypertrophy" in a conclusion is a word; the work is done with numbers — septal and posterior wall thickness, end-diastolic dimension, calculated mass, mass index, relative wall thickness. Worth copying out of the report, because next time there will be nothing to compare against.

3. Understand that comparison needs a series. One echocardiogram is a point. The question "is it changing" is answered only by a second one, done comparably. In a personal record such values live as series rather than as separate PDFs — precisely so that a difference shows up as a number.

4. Treat blood pressure as the main modifiable input. It is the same value discussed in the piece on blood pressure and longevity: load on the myocardium is its direct continuation.

5. Do not forget the other inputs. Body mass and where it sits, sleep, cardiorespiratory fitness and the lipid number are neighbouring stories by mechanism, not separate topics.

This material is informational and does not replace a medical consultation. Interpreting an echocardiography report, choosing therapy and deciding on follow-up are matters for the treating physician, on the whole clinical picture.