Global Longitudinal Strain: Normal Values Explained

Global longitudinal strain (GLS) in a healthy adult heart typically falls between about −16% and −24%, with most studies reporting a mean near −20%. The negative sign indicates the heart muscle is shortening as it squeezes blood out, so a “more negative” number actually reflects stronger contraction. Understanding where your GLS result sits within that range, and what can nudge it higher or lower without actual heart disease, turns out to be more complicated than a single cutoff can capture.

What the Numbers Actually Mean

GLS measures how much the heart’s left ventricle shortens from base to apex during each beat. It tracks tiny acoustic markers (“speckles”) frame by frame on an ultrasound image and calculates the percentage change in muscle length. A GLS of −20% means the muscle fibers shortened by 20% at peak contraction. Because the muscle gets shorter, the value is negative by convention.

A large meta-analysis pooling data from multiple studies found that reported normal GLS values ranged from −15.9% to −22.1%, with an overall mean of −19.7%.¹ A more recent guideline-directed study established reference ranges for left ventricular GLS of −24% to −16%.² Most clinicians treat a GLS value less negative than about −16% as abnormal, though context matters enormously.

The spread between −16% and −24% is not noise. It reflects genuine biological variability across the population, shaped by age, sex, blood pressure, body size, and the equipment used to measure it. An individual GLS of −17% might be slightly abnormal for a 30-year-old woman but borderline normal for a 70-year-old man. A single number line cannot tell the whole story.

Why GLS Gets Compared to Ejection Fraction

Ejection fraction (EF) has been the workhorse measurement for decades. It estimates the percentage of blood the left ventricle pumps out with each beat, and anything above roughly 50–55% is considered normal. The problem is that EF can stay normal even when heart muscle is already quietly failing. The subendocardial muscle fibers, the ones running lengthwise from base to apex, are the first to suffer from reduced blood supply or increased wall stress. EF does not register damage in those fibers until the problem is well advanced.

GLS directly tracks the shortening of those longitudinal fibers. Over the past decade, evidence has shown GLS to be more sensitive to early dysfunction than EF and to carry stronger prognostic information.³ A systematic review and meta-analysis found that the risk reduction in mortality associated with each standard-deviation improvement in GLS was about 1.6 times greater than the equivalent improvement in EF.⁴ In practical terms, two patients can share the same “normal” EF of 55% while their GLS values differ dramatically, and the one with the worse GLS faces a meaningfully higher risk of hospitalization and death.

This gap becomes especially visible in a condition called heart failure with preserved ejection fraction, where the EF looks reassuring yet the patient is clearly symptomatic. In the RELAX trial, about two-thirds of these patients had abnormal GLS (less negative than −16%) despite having a preserved EF.⁵ Compared with healthy controls, these patients had substantially lower longitudinal strain.⁶ GLS essentially exposes a layer of dysfunction that EF hides.

How Sex and Age Shift Normal Values

Women consistently show more negative (stronger) GLS than men by roughly 1.5 to 2 percentage points. Data from the Framingham Heart Study found women had approximately 1.7% greater longitudinal strain than men, and the upper limits of normal differed accordingly: −14.4% to −17.1% for women versus −14.4% to −15.2% for men across age groups.⁷ This matters clinically because applying a single sex-neutral cutoff can mislabel a healthy young woman as abnormal or miss early disease in an older man.

Age behaves differently between the sexes, too. A study of over a thousand participants found that increasing age was linked to reduced GLS in women but not in men, with the sharpest decline beginning around age 66 or 67.⁸ The reasons probably involve hormonal changes after menopause, differences in vascular stiffening, and changes in cardiac geometry over time. A post-menopausal woman whose GLS has drifted from −21% to −18% may still be physiologically normal for her age, but the trajectory matters and could warrant closer monitoring.

Blood Pressure and Other Factors That Affect Your Result

Your blood pressure at the moment the ultrasound is done can shift GLS readings noticeably. Higher afterload, the resistance the heart pumps against, makes the muscle shorten less, producing a less negative GLS that can mimic dysfunction. Population data from the Tromsø Study showed that for every 10 mm Hg increase in systolic blood pressure, GLS decreased by about 0.2%.⁹ Both systolic and diastolic pressure are inversely correlated with GLS.¹⁰ This load-dependence is one reason clinicians sometimes record blood pressure alongside the strain measurement, so the number can be interpreted in context.

Heart rate plays a smaller but real role. A very fast rate reduces the time available for full contraction, which can slightly suppress GLS. Body habitus affects image quality, particularly in people with larger body size, where the ultrasound windows may be suboptimal. And hydration status shifts preload (how much blood fills the ventricle before a beat), which in turn nudges strain up or down. None of these factors invalidates the test, but they help explain why a single GLS reading should never be interpreted in isolation.

The Vendor Problem and Why It Is Getting Better

For years, one of the biggest headaches with GLS was that different ultrasound machines from different manufacturers could give meaningfully different numbers on the same heart. A standardization task force reported that average segmental strain values varied by up to 4.5% between vendors.¹¹ That difference is big enough to cross clinical thresholds, meaning the same patient could appear normal on one machine and abnormal on another.

A decade-long industry standardization effort has substantially narrowed that gap. An updated analysis showed that the latest semi-automated software from major companies now produces GLS values within a very tight range, with the maximum difference in mean GLS less than 1 percentage point.¹² The practical takeaway is that if your lab uses recent-generation equipment, vendor differences are becoming less of a concern for global measurements. But some older platforms and certain specialized clinical scenarios, particularly cardiomyopathies with thickened walls, still show meaningful vendor-to-vendor discrepancies.¹³ The general recommendation remains: when tracking GLS over time, try to use the same machine and software for serial measurements.

Image quality also matters. The tracking algorithm works best when the entire heart wall is visible in every frame of the clip. Anything that degrades the image, a poor acoustic window, respiratory motion, an uncooperative patient, reduces tracking accuracy.¹⁴ The good news is that GLS has high reproducibility between different observers. Even relatively inexperienced readers achieve very good agreement, and training improves precision for both global and segmental measurements.¹⁵

GLS in Children

Pediatric reference values for GLS are broadly similar to adult values but carry their own quirks. A systematic review of over 2,300 children found mean normal GLS ranged from −16.7% to −23.6%, with an overall mean of −20.2%.¹⁶ A separate meta-analysis reported a similar mean of −20.5%.¹⁷ Strain magnitude appears to peak around age four or five, then gradually levels off.¹⁸

Because children’s hearts are growing, simply borrowing adult cutoffs is problematic. Pediatric Z-scores, which express a measurement in terms of standard deviations from the age-matched mean, are increasingly available and let clinicians flag an abnormal GLS relative to what is expected for that child’s developmental stage.

Where Abnormal GLS Changes Clinical Decisions

Three areas stand out where GLS values directly alter what doctors do for patients.

Cancer therapy monitoring is arguably the best-known application. Many chemotherapy drugs and some targeted therapies damage the heart muscle, and catching that damage early can allow cardiologists to start protective medications before serious dysfunction sets in. A systematic review found that alterations in GLS consistently precede drops in ejection fraction during chemotherapy, and that a 10% to 15% relative reduction in GLS appears to be the best predictor of later cardiotoxicity.¹⁹ One study of breast cancer patients found that a GLS threshold of −16% or less negative at three months into chemotherapy predicted later systolic dysfunction with 80% sensitivity and 90% specificity.²⁰

Aortic valve disease is another area where GLS earns its keep. In patients with severe aortic stenosis who have not yet developed symptoms, GLS can help predict who is heading toward trouble. One study found that patients whose baseline GLS was less negative than −18.2% were more likely to develop symptoms and need valve intervention compared with those who had better strain values.²¹ A meta-analysis confirmed that impaired GLS in asymptomatic aortic stenosis is associated with higher rates of major adverse cardiac events and all-cause mortality.²²

GLS also serves as a powerful prognostic marker in heart failure. In a study of over 4,000 patients hospitalized with acute heart failure, each 1% worsening in GLS was linked to a 5% increase in five-year mortality. GLS stratified risk far better than ejection fraction in this population: patients with severely reduced strain had substantially higher mortality, while EF categories showed only modest differences.²³

GLS Predicts Outcomes Even in Healthy People

The prognostic power of GLS is not limited to people who are already sick. The Copenhagen City Heart Study followed a low-risk general population and found that lower GLS was independently associated with a higher risk of cardiovascular events, even after adjusting for conventional risk factors and ejection fraction. GLS added predictive information on top of established risk scoring systems.²⁴ This finding has fueled interest in using GLS as a screening tool, though it remains unclear whether population-wide screening would be cost-effective or change outcomes enough to justify it.

Regional Strain Patterns and Cardiac Amyloidosis

GLS is a global average, but the software also reports strain for each heart segment individually. One regional pattern in particular has gotten significant clinical attention: “apical sparing,” where the base and mid-wall of the heart are severely impaired but the apex contracts relatively normally. This signature has been associated with cardiac amyloidosis, a condition where abnormal protein deposits stiffen the heart muscle.

However, the pattern is less reliable than early enthusiasm suggested. A meta-analysis found that the relative apical sparing pattern detected cardiac amyloidosis with about 67% sensitivity and 85% specificity, numbers that shift depending on the exact cutoff used.²⁵ A multicenter study struck a more cautious note, finding that even with the best cutoff, apical sparing was only 72% sensitive and 66% specific for cardiac amyloidosis, and that roughly a third of control patients with similar cardiac conditions but no amyloidosis also showed some degree of apical sparing.²⁶ In another series, only about a third of patients referred for further evaluation based on the apical sparing pattern turned out to have confirmed or highly probable cardiac amyloidosis.²⁷ Apical sparing is a useful red flag that should prompt further testing, but it is not, by itself, diagnostic.

What Happens to GLS During Exercise

In a healthy person, GLS should improve during physical exertion because the heart is supposed to squeeze harder when demand increases. A study of healthy adults performing exercise stress echocardiography found that average GLS at peak exercise was about −25.4%, an increase of roughly 5 percentage points compared to resting values. This boost was consistent across age groups and between men and women.²⁸ When a patient’s GLS fails to increase normally with exercise, it may indicate reduced “contractile reserve,” a limited ability to recruit extra pumping power when needed, even if resting function looks acceptable.

2D Versus 3D Strain Measurement

Standard GLS is measured with two-dimensional ultrasound images, stitching together strain from several viewing angles. Three-dimensional speckle tracking captures the ventricle as a volume and could, in theory, produce more accurate strain because it follows the muscle’s actual movement through space rather than projecting it onto a flat plane.²⁹

In practice, 2D and 3D GLS correlate well at the global level, but the agreement breaks down when you look at individual wall segments.³⁰ Three-dimensional strain also has lower frame rates and needs good image quality from all angles simultaneously, which is harder to achieve in routine clinical scanning. For now, 2D GLS remains the standard, and the reference ranges cited throughout this article are based on 2D measurements. If you see a 3D GLS result on your report, the normal ranges may not be directly comparable to the familiar 2D values.

Cardiac MRI can also measure longitudinal strain using a technique called feature tracking, and its results correlate strongly with speckle-tracking echocardiography.³¹ MRI-derived strain is particularly useful when echo image quality is poor or when other MRI findings (like fibrosis mapping) are being evaluated alongside function. Training improves reproducibility for MRI-based strain, though software differences remain a source of variability, just as with ultrasound vendors.³²

Strain Measurements Beyond the Left Ventricle

Though left ventricular GLS gets the most attention, the same speckle-tracking approach works for the right ventricle and the atria. The guideline-directed study that set left ventricular GLS at −24% to −16% also reported right ventricular free-wall strain reference ranges of −35% to −17%.² Right ventricular strain is increasingly used in pulmonary hypertension and congenital heart disease, while left atrial strain is gaining traction as a marker for atrial fibrillation risk and diastolic dysfunction. These measurements are less standardized than left ventricular GLS and should be interpreted with extra caution, but the research trajectory suggests they will eventually become routine additions to the echocardiographic report.