Stress Hyperglycemia Ratio and Left Atrial Thrombus

Author Ünsal Vural

Description In acute ischemic stroke, the stress hyperglycemia ratio (SHR) relates admission glucose to background glycemia estimated from glycated hemoglobin. An association with thrombus found in the left atrium or its appendage has prompted interest in using SHR to select patients for transesophageal echocardiography. Atrial thrombus arises from several rhythm, structural, hemodynamic and systemic conditions; glucose can also rise in response to the stroke itself. Existing evidence does not establish whether SHR reflects a thrombotic cause, an acute consequence or both, or whether it improves imaging decisions.

Keywords stress hyperglycemia ratio; acute ischemic stroke; left atrial thrombus; left atrial appendage; transesophageal echocardiography; reverse causation; risk stratification

1 Introduction and the clinical question

A negative transthoracic echocardiogram does not exclude a left atrial or left atrial appendage thrombus. Transesophageal echocardiography (TEE) can reveal a cardiac source missed by transthoracic echocardiography (TTE), with possible consequences for secondary stroke prevention [2,12]. Yet TEE is not equally informative for every patient after ischemic stroke. The practical challenge is deciding whom to investigate further when the first cardiac assessment is unrevealing.

A recent report links higher SHR to left atrial or appendage thrombus found by TEE after ischemic stroke [1]. The question is more demanding than whether two measurements are associated: does a glucose-derived signal reveal pre-existing atrial thrombogenicity, or does it largely register the physiological response to a stroke caused by that thrombus? The answer requires the established causes of atrial thrombus, the timing of the glucose measurement and the incremental value of SHR over information already available before TEE.

2 Atrial thrombus has more than one cause

Left atrial thrombus develops where blood stasis, atrial tissue abnormalities and prothrombotic conditions intersect. Atrial fibrillation is a major contributor because it disrupts effective atrial contraction and promotes slow flow in the appendage. However, a sinus-rhythm tracing at admission cannot exclude intermittent atrial fibrillation or a pre-existing atrial substrate. Atrial enlargement and fibrosis, impaired appendage emptying, mitral stenosis or a prosthetic mitral valve, cardiac dysfunction and systemic inflammatory or coagulation abnormalities may also affect risk [1,3–5].

These factors are related but are not interchangeable. Mitral valve disease may promote atrial enlargement and flow stagnation; an atrial cardiomyopathy can coexist with or precede recognized atrial fibrillation. Reduced appendage emptying velocity and spontaneous echo contrast provide evidence of sluggish flow, but they are commonly characterized during TEE [2,4,5]. They may help explain a thrombus after imaging, whereas they cannot ordinarily be used to choose patients for the TEE that reveals them.

Anatomy and rhythm should anchor interpretation before glucose. A metabolic marker should be judged against the atrial and valvular setting in which the clot develops. An association between glucose and thrombus is not evidence that glucose alone generated the thrombus.

In the relevant stroke cohort, patients were in sinus rhythm on admission, but 24–72-hour monitoring subsequently identified paroxysmal atrial fibrillation in some of them; known persistent or permanent atrial fibrillation had been excluded [1]. Initial sinus rhythm therefore did not mean the atrial rhythm substrate was absent. A short monitoring period can also leave uncertainty about an intermittent arrhythmia. This matters when an acute blood marker is evaluated alongside rhythm findings obtained at different times.

3 What SHR can measure

SHR compares acute glucose with an estimate of usual glycemia derived from glycated hemoglobin (HbA1c). A common formulation divides admission glucose by estimated average glucose, expressed in the same units; in mg/dL, estimated average glucose = 28.7 × HbA1c (%) − 46.7 [6]. The thrombus study used the equivalent mmol/L formulation: [admission glucose (mg/dL) / 18] / [1.59 × HbA1c (%) − 2.59] [1]. Higher SHR describes a greater acute glucose elevation relative to background glycemia, rather than directly measuring catecholamines or thrombogenicity.

SHR is sensitive to when blood was drawn and to acute infection, treatment, nutrition and other stressors. It is less interpretable when HbA1c fails to represent earlier glycemia. Studies using other numerators or denominators do not produce interchangeable SHR cutoffs [6,7].

4 Biological plausibility and reverse causation

Hyperglycemia can be accompanied by endothelial dysfunction, oxidative stress, platelet activation and impaired fibrinolysis. These processes offer plausible links between acute metabolic disturbance and thrombosis [7]. Biological plausibility, however, does not establish the direction of causation. Observing SHR after stroke leaves at least two time sequences possible.

In one sequence, metabolic dysregulation contributes to a prothrombotic environment before the atrial thrombus and stroke. In another, an atrial thrombus embolizes to the brain; the resulting injury and sympathetic, endocrine and inflammatory responses raise glucose after the event. The latter sequence would make SHR a marker of the stroke response rather than of prior thrombus formation. Shared factors such as inflammation or diabetes could also contribute to both findings [3,7].

A single post-onset glucose result cannot distinguish these explanations. Stroke severity, infarct distribution, the interval from symptom onset to blood sampling, infection and treatments must be considered. If the association weakens after accounting for these variables, SHR could remain prognostic while having limited value for identifying a pre-existing atrial source. Prognosis and diagnosis are separate questions [8,9].

The temporal distinction has practical implications for research. A stable metabolic association would have to remain informative after considering stroke severity and early inflammatory illness. A signal concentrated in severe strokes, later blood samples or patients with pronounced systemic inflammation would be more compatible with a post-event stress response. Neither pattern alone would prove a causal pathway; both identify which explanation requires further testing.

5 What the available clinical evidence shows

In the retrospective study by Cicek and colleagues, 486 patients with ischemic stroke underwent TEE after a TTE without visible thrombus. The study excluded isolated left ventricular and right atrial thrombi and defined its outcome as a mass in the left atrium or appendage. TEE detected thrombus in 64 patients (13.2%). Mean SHR was 0.99 ± 0.42 in the thrombus group and 0.84 ± 0.27 in the group without thrombus; the adjusted odds ratio for SHR was 2.39 (95% confidence interval 1.11–5.17) [1].

The same cohort also demonstrates why a single metabolic explanation would be incomplete. Paroxysmal atrial fibrillation was reported in 62.5% of those with thrombus and 6.9% of those without; mitral valve replacement in 23.4% and 8.3%, respectively. Ejection fraction was lower; admission glucose was 134 versus 118 mg/dL and C-reactive protein 63.4 versus 27.2 mg/L. HbA1c did not differ significantly between groups [1]. The glucose and inflammatory differences raise the possibility that the acute response contributes to the SHR signal, but they cannot determine its cause. The coexistence of rhythm, valvular, cardiac and inflammatory signals makes the biological interpretation of SHR less straightforward.

The reported area under the curve of 0.796 describes the multivariable model, not SHR used alone. The analysis involved only patients selected for TEE, and its measurements cannot show whether elevated SHR preceded thrombus or followed cerebral injury [1]. Among patients who were not referred for TEE, the proportion of missed atrial thrombi and the behavior of the SHR signal remain unknown.

Prognostic findings answer another question. Roberts and colleagues compared SHR, glucose and the glycemic gap in 300 patients with ischemic stroke [8]. In a separate analysis of 4,515 patients with mild stroke or high-risk transient ischemic attack, the highest SHR quartile had a greater adjusted 90-day recurrent-stroke risk than the lowest quartile (hazard ratio 1.84; 95% confidence interval 1.30–2.61) [9]. These observations concern outcome after stroke, not the presence of an atrial thrombus at TEE.

6 When could SHR change a TEE decision

The clinical test is incremental value. Before adding SHR, a TEE-selection model should use information known at the decision point: atrial fibrillation history and rhythm monitoring, mitral valve disease or prosthesis, available TTE findings including left atrial size and ventricular function, and stroke imaging characteristics [2,3,12]. Appendage flow velocity and spontaneous echo contrast, when known only from the planned TEE, cannot enter this pre-TEE selection model.

SHR, admission glucose and HbA1c should be compared as alternative additions to a credible clinical model. The question is whether SHR changes predictions beyond glucose or the existing clinical assessment, particularly for patients whose need for TEE is uncertain. A significant odds ratio does not answer that question. Nor can an AUC from a combined model be attributed to its glucose-derived component [10].

A patient with established atrial fibrillation or a high-risk mitral valve condition may already warrant detailed cardiac assessment on clinical grounds. SHR is more likely to be informative if it improves discrimination among patients whose need for TEE remains uncertain after rhythm and structural assessment. Even in that group, any proposed SHR threshold must be tested in a population that includes patients not selected for imaging on the basis of the marker.

A proposed rule should report calibration and demonstrate what its threshold means in practice: how many examinations are avoided, how many treatable thrombi are missed and whether the added findings alter care. Decision-curve analysis can compare the expected benefit of using the rule with existing approaches across plausible thresholds [11]. An association is not yet a TEE-selection policy.

7 Evidence needed and conclusion

Prospective studies should define a population with an unrevealing initial evaluation, record why each patient received TEE and specify whether the target is left atrial or appendage thrombus or a broader actionable cardiac source. Glucose timing, HbA1c, stroke severity, infarct pattern, atrial rhythm monitoring, mitral disease, left atrial size, ventricular function and acute inflammatory conditions should be captured before analysis. Internal and external validation should follow a prespecified comparison of clinical models with and without SHR [10,11].

The central distinction remains unresolved: elevated SHR may reflect a thrombogenic setting, the stress response to an embolic stroke, or both. Its association with TEE-detected atrial thrombus is a useful signal for research, but the atrial substrate and the timing of stroke must be addressed before it can guide imaging. The defensible question is not whether SHR correlates with thrombus, but whether it adds reliable information that changes a patient’s TEE decision [1–3,10,11].

References

1. Cicek V, Erdem A, Kozan Çıkırıkçı EH, Yılmaz İ, Günhan E, Uygun E, et al. Stress hyperglycemia ratio is associated with intracardiac thrombus in acute ischemic stroke. Turk Kardiyol Dern Ars. 2026;54:395–402. doi:10.5543/tkda.2026.50611.
2. Saric M, Armour AC, Arnaout MS, Chaudhry FA, Grimm RA, Kronzon I, et al. Guidelines for the use of echocardiography in the evaluation of a cardiac source of embolism. J Am Soc Echocardiogr. 2016;29:1–42. doi:10.1016/j.echo.2015.09.011.
3. Kamel H, Healey JS. Cardioembolic stroke. Circ Res. 2017;120:514–526. doi:10.1161/CIRCRESAHA.116.308407.
4. Akoum N, Fernandez G, Wilson B, McGann C, Kholmovski E, Marrouche N. Association of atrial fibrosis quantified using LGE-MRI with atrial appendage thrombus and spontaneous contrast on transesophageal echocardiography in patients with atrial fibrillation. J Cardiovasc Electrophysiol. 2013;24:1104–1109. doi:10.1111/jce.12199.
5. Wang X, Xu X, Wang W, et al. Risk factors associated with left atrial appendage thrombosis in patients with non-valvular atrial fibrillation by transesophageal echocardiography. Int J Cardiovasc Imaging. 2023;39:1263–1273. doi:10.1007/s10554-023-02841-x.
6. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ, et al. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008;31:1473–1478. doi:10.2337/dc08-0545.
7. Dungan KM, Braithwaite SS, Preiser JC. Stress hyperglycaemia. Lancet. 2009;373:1798–1807. doi:10.1016/S0140-6736(09)60553-5.
8. Roberts G, Sires J, Chen A, Thynne T, Sullivan C, Quinn S, et al. A comparison of the stress hyperglycemia ratio, glycemic gap, and glucose to assess the impact of stress-induced hyperglycemia on ischemic stroke outcome. J Diabetes. 2021;13:1034–1042. doi:10.1111/1753-0407.13223.
9. Li W, Yan H, Gao Y, Gao C, Li W, Pan Y, et al. Stress hyperglycemia ratio and adverse outcomes in acute mild ischemic stroke or high-risk transient ischemic attack: a secondary analysis of the INSPIRES trial. Stroke. 2026;57:349–361. doi:10.1161/STROKEAHA.125.052987.
10. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis: the TRIPOD statement. Ann Intern Med. 2015;162:55–63. doi:10.7326/M14-0697.
11. Vickers AJ, Elkin EB. Decision curve analysis: a novel method for evaluating prediction models. Med Decis Making. 2006;26:565–574. doi:10.1177/0272989X06295361.
12. Kleindorfer DO, Towfighi A, Chaturvedi S, Cockroft KM, Gutierrez J, Lombardi-Hill D, et al. 2021 guideline for the prevention of stroke in patients with stroke and transient ischemic attack. Stroke. 2021;52:e364–e467. doi:10.1161/STR.0000000000000375.