Global longitudinal strain
GLS (2D speckle tracking) · Three apical views (4-, 2- and 3-chamber), frame rate 50–80 fps, one vendor, stable ECG
Normal range & thresholds
Normal peak systolic GLS ≈ −20% (range −16% to −25%). |GLS| <16% is abnormal; 16–18% is borderline. Women have slightly higher magnitude than men, and magnitude falls modestly with age.[16,53,54]
Inter-vendor variability is up to 3.7 percentage points absolute — serial comparison is only valid on the same vendor and software version, and that must be stated in the report.[90]
Relative apical sparing ratio = apical GLS / (basal + mid GLS) >1.0 suggests cardiac amyloidosis.[30]
Inter-vendor variability is up to 3.7 percentage points absolute — serial comparison is only valid on the same vendor and software version, and that must be stated in the report.[90]
Relative apical sparing ratio = apical GLS / (basal + mid GLS) >1.0 suggests cardiac amyloidosis.[30]
Pathophysiology
Strain measures deformation, not motion, so it is immune to the translation and tethering that corrupt displacement and velocity indices. Longitudinal subendocardial fibres are the most load- and ischaemia-sensitive layer in the ventricle, so their deformation is the first thing lost in almost every myocardial disease — while circumferential and radial mechanics compensate and hold ejection fraction in the normal range. GLS is therefore the parameter that detects disease in the window between ‘normal EF’ and ‘heart failure’, and it out-predicts LVEF for mortality in heart failure and in the general population.[88,89,108]
Raised by
- (Magnitude increases, i.e. more negative) Athletic training; hyperdynamic states; after successful revascularisation, GDMT or valve intervention
- Note the sign convention: GLS is negative, so ‘increased GLS’ is ambiguous. Report magnitude, and say ‘less negative’ or ‘more negative’
Lowered by
- (Magnitude falls, i.e. less negative) Subclinical cardiotoxicity from anthracyclines, trastuzumab, checkpoint inhibitors — the earliest detectable abnormality[15,29,74]
- Ischaemia, prior infarction (regional bull’s-eye deficit in a coronary territory), hibernation
- Infiltrative and storage disease — amyloid (apical sparing), Fabry (basal inferolateral deficit), sarcoid (patchy, basal septal)
- HCM (reduced despite hyperdynamic LVEF), hypertensive heart disease, diabetes, obesity, CKD
- Aortic stenosis and chronic aortic/mitral regurgitation — falls before LVEF and before the dimensional thresholds are reached
- Myocarditis, chemotherapy-naive cancer cachexia, sepsis, HFpEF (GLS is abnormal in ≈half of HFpEF)
- Right-sided disease with septal involvement; LBBB and RV pacing (dyssynchrony, not necessarily disease)
Technique & pitfalls
- Non-foreshortened apical views, all three, with a clear ECG and a stable rhythm; the software must track the full myocardial thickness, not just the endocardial line.
- Frame rate 50–80 fps — too low and speckles decorrelate, too high and the signal degrades.
- Reject and re-trace any segment the software tracks badly; do not accept an automatic value with two or more excluded segments.
- Define aortic valve closure timing correctly — peak systolic strain versus end-systolic strain differ when there is post-systolic shortening.
- Read the bull’s-eye pattern, not only the single number: the pattern is what distinguishes amyloid, Fabry, sarcoid and coronary territories from each other.
- Same vendor, same version, ideally same operator for serial studies.[16,90]
Pseudo-change & artefact
- Vendor and software version change — the commonest artefactual ‘deterioration’.[90]
- Foreshortening, poor endocardial definition, and reverberation artefact in the lateral wall.
- Load: GLS falls with hypertension, rises after afterload reduction; it is not load-independent.
- Heart rate, arrhythmia, ectopy, atrial fibrillation (average multiple beats).
- Region-of-interest width set too wide, including pericardium, or too narrow, missing the epicardium.
- Sign confusion between reports (−18% vs 18%) — a persistent source of clinical error.
Treatment thresholds
- Cardio-oncology (the one strong, guideline-endorsed threshold): a relative reduction >15% from the patient’s own baseline defines mild asymptomatic cancer-therapy-related cardiac dysfunction and triggers cardioprotection (ACEi/ARB ± beta-blocker) and closer surveillance, without necessarily interrupting oncological treatment.[15,29] SUCCOUR showed strain-guided management produced a smaller LVEF fall than EF-guided management.[74]
- Amyloidosis: relative apical sparing >1.0 is one of the imaging red flags that mandates free light chains and bone scintigraphy.[30]
- Asymptomatic valve disease: impaired GLS (commonly cited as worse than −18% in AR/MR, −15% in AS) is used by Heart Teams to bring forward intervention. It is a supportive, not a class I, criterion in the 2025 ESC/EACTS and 2020 ACC/AHA documents — say so when you report it.[6,7]
- Heart failure with preserved EF: reduced GLS supports the diagnosis within the HFA-PEFF framework.[100,101]
Next step
- Next: read the bull’s-eye for pattern, then correlate with wall thickness, ECG voltage and the clinical context.
- Then: CMR — LGE, T1 and ECV convert a strain pattern into a tissue diagnosis.
- Then: disease-specific confirmation — light chains and bone scintigraphy (amyloid), α-galactosidase A (Fabry), ischaemia testing or angiography (coronary territory deficit), genetic testing (cardiomyopathy).
- In cardio-oncology: repeat with the same vendor before acting on a fall — and always compare to the patient’s own baseline, never to the population range.
Drugs
- Improve GLS: ACE inhibitors/ARBs and ARNI, beta-blockers, MRAs, SGLT2 inhibitors, statins (modest); revascularisation; CRT; valve intervention; mavacamten in obstructive HCM.[68,69,72]
- Worsen GLS: anthracyclines (dose-dependent, partly irreversible), trastuzumab (largely reversible), VEGF inhibitors and checkpoint inhibitors, radiotherapy.[15]
- Tafamidis slows deterioration in ATTR; enzyme replacement improves strain in early Fabry.[75]
- Acute afterload reduction improves GLS without any change in contractility — interpret peri-treatment improvement cautiously.
Reversibility
Trastuzumab-related GLS decline: >80% recovers within 3–6 months of interruption plus cardioprotection. Anthracycline-related: partially reversible, and recovery falls sharply as cumulative dose rises and as time to detection lengthens — the entire rationale for strain surveillance.[15,29,74]
Ischaemic: hibernating segments recover over weeks to months after revascularisation; scarred segments do not. HFrEF on GDMT: GLS improves in parallel with LVEF, typically by 2–4 absolute points over 6–12 months.
Aortic stenosis: GLS improves within months of AVR, but recovery is incomplete once replacement fibrosis is established — a further argument for earlier intervention.[86]
Amyloid: minimally reversible; stabilisation is the realistic goal.[75]
Ischaemic: hibernating segments recover over weeks to months after revascularisation; scarred segments do not. HFrEF on GDMT: GLS improves in parallel with LVEF, typically by 2–4 absolute points over 6–12 months.
Aortic stenosis: GLS improves within months of AVR, but recovery is incomplete once replacement fibrosis is established — a further argument for earlier intervention.[86]
Amyloid: minimally reversible; stabilisation is the realistic goal.[75]