Glycated hemoglobin and cardiovascular risk in diabetes mellitus: Evolving evidence beyond glycemic control

Abstract

Glycated hemoglobin (HbA1c) remains the cornerstone biomarker for long-term glycemic control in diabetes mellitus, reflecting average plasma glucose over approximately three months. Beyond its diagnostic and monitoring utility, HbA1c has emerged as a robust predictor of cardiovascular disease, the leading cause of morbidity and mortality in individuals with diabetes. Epidemiological studies, including the United Kingdom Prospective Diabetes Study (UKPDS) and the Diabetes Control and Complications Trial, have consistently demonstrated a graded, near-linear relationship between HbA1c levels and macrovascular risk. However, the translation of glycemic control into cardiovascular benefit has proven complex. While early intensive glycemic control confers long-term cardiovascular protection, a phenomenon termed “metabolic memory” later trials such as ACCORD, ADVANCE, and VADT have highlighted the limitations and potential risks of aggressive HbA1c lowering, particularly in high-risk populations. Emerging evidence suggests that HbA1c variability, rather than mean levels alone, may independently contribute to cardiovascular outcomes through mechanisms involving oxidative stress, endothelial dysfunction, and inflammation. Furthermore, contemporary glucose-lowering therapies, including sodium-glucose transporter 2 inhibitors and glucagon-like peptide-1 receptor agonists, reduce cardiovascular events largely independent of HbA1c reduction, challenging the primacy of HbA1c as a surrogate endpoint. This review critically appraises the evolving role of HbA1c in cardiovascular risk stratification, emphasizing the need for individualized glycemic targets and a broader, multifactorial approach to cardiovascular risk reduction in diabetes.

Key Words: Glycated hemoglobin; Cardiovascular disease; Diabetes mellitus; Glycemic variability; Cardiovascular risk

Core Tip: Glycated hemoglobin (HbA1c) is a well-established marker of chronic glycemia and a predictor of cardiovascular risk in diabetes, yet its role as a therapeutic target for reducing cardiovascular events is increasingly nuanced. Evidence from landmark trials highlights both the benefits of early glycemic control and the limitations of intensive HbA1c reduction in advanced disease. Glycemic variability and emerging cardioprotective therapies further challenge HbA1c-centric paradigms. Clinicians should interpret HbA1c within a broader cardiometabolic framework, integrating patient-specific factors and adjunctive risk modifiers to optimize outcomes. A shift from glucose-centric to outcome-driven strategies is essential for contemporary diabetes care.

INTRODUCTION

Glycated hemoglobin (HbA1c) holds a distinctive place in diabetes care because it is simultaneously diagnostic, therapeutic, epidemiological, and prognostic. It reflects average glycaemic exposure over roughly the preceding 2-3 months, is deeply embedded in treatment algorithms, and is one of the few biomarkers familiar to both specialist and generalist clinicians[1,2]. This prominence has inevitably encouraged its use as a surrogate for cardiovascular risk. The biological logic is strong: Chronic hyperglycaemia promotes advanced glycation, oxidative stress, endothelial dysfunction, lipoprotein modification, vascular inflammation, myocardial fibrosis, and ultimately atherosclerotic and heart-failure phenotypes. Yet the clinical meaning of HbA1c is more complicated than its convenience suggests. Observational studies consistently link higher HbA1c with myocardial infarction, stroke, heart failure (HF), and death, but intervention trials have shown that lowering HbA1c does not translate linearly into cardiovascular benefit, particularly in patients with long-standing disease or established vascular injury[3-6]. At the same time, cardiovascular outcome trials of sodium-glucose cotransporter-2 inhibitors and glucagon-like peptide-1 receptor agonists have shown substantial reductions in major cardiovascular, heart-failure, and kidney outcomes that are disproportionate to HbA1c lowering alone, shifting current management away from a purely “glucocentric” model[7-11]. For a clinical gastroenterology/hepatology or general medical readership, the key question is therefore not whether HbA1c matters, but how far it should drive cardiovascular risk estimation and treatment decisions in contemporary diabetes care.

EPIDEMIOLOGICAL EVIDENCE

The epidemiological case for HbA1c as a cardiovascular biomarker is substantial. A meta-analysis of observational studies found that each 1-percentage-point increase in glycosylated haemoglobin was associated with an 18% higher risk of cardiovascular disease (CVD) in type 2 diabetes[12]. Subsequent cohorts have broadly reinforced this association. In the SCOUT trial population of overweight or obese adults with type 2 diabetes at high cardiovascular risk, each 1% higher HbA1c was associated with a 17% higher hazard of the primary cardiovascular endpoint and a stronger mortality gradient in women[13]. However, the relationship is not invariably linear. In a secondary analysis of TECOS involving patients with established atherosclerotic disease, associations between baseline or time-varying HbA1c and heart-failure hospitalisation, all-cause death, and composite cardiovascular outcomes were U-shaped, with the lowest risk at about 7%[14]. A broad meta-analysis of 46 studies reached a similar conceptual conclusion at the population level, reporting the lowest cardiovascular and all-cause mortality in diabetes at HbA1c values around 6.0%-8.0%, with excess risk emerging at both higher and lower levels[15-17]. These data support HbA1c as a meaningful risk gradient, but they also argue against a simplistic “lower is always better” interpretation.

HF has been increasingly recognised as a first cardiovascular presentation in many adults with type 2 diabetes. A meta-analysis concluded that higher HbA1c levels were significantly associated with heart-failure risk in observational studies of people with diabetes[18]. In a Japanese cohort, HbA1c predicted future heart-failure hospitalisation independently of baseline B-type natriuretic peptide and echocardiographic measures[19]. More recently, a United Kingdom Biobank analysis found that heart-failure risk rose with longer diabetes duration and poorer glycaemic control, with particularly elevated risk among those with HbA1c ≥ 8.0% and diabetes duration ≥ 15 years[20]. Importantly, average HbA1c is only part of the story. Visit-to-visit variability appears to add prognostic information beyond mean control. In a 147811-person Hong Kong cohort, HbA1c variability was strongly associated with CVD and mortality across age groups, even after adjustment for usual HbA1c; in younger adults, each 1% increase in variability was linked to a 28% higher risk of the composite outcome of CVD and death[21]. A 2022 meta-analysis likewise found that top-vs-bottom quartiles of HbA1c variability were associated with higher CVD risk, whereas fasting glucose variability showed weaker and less consistent associations with CVD[22]. Thus, mean HbA1c reflects chronic exposure, whereas variability may better capture biological instability and treatment-related volatility.

Randomised evidence is more nuanced and more clinically instructive. In newly diagnosed diabetes, UKPDS 33 showed that intensive glucose control lowered mean HbA1c from 7.9% to 7.0% and reduced any diabetes-related endpoint by 12%, driven mainly by a 25% reduction in microvascular complications, without a clear macrovascular benefit during the trial itself[23]. UKPDS 34, however, suggested that drug choice and timing matter: In overweight patients, intensive treatment with metformin achieved a median HbA1c of 7.4% vs 8.0% with conventional therapy and reduced any diabetes-related endpoint by 32%, diabetes-related death by 42%, and all-cause mortality by 36%[24]. The 10-year UKPDS follow-up then introduced the crucial concept of glycaemic legacy: Despite early loss of between-group HbA1c separation, prior intensive therapy was associated with emergent reductions in myocardial infarction (15%) and all-cause death (13%), while the metformin subgroup retained even larger reductions in myocardial infarction (33%) and death (27%)[25]. For cardiovascular prevention, this remains the strongest argument that safe, early, sustained glycaemic improvement has lasting value[21].

By contrast, trials in longer-duration, higher-risk diabetes exposed the limits of intensive HbA1c lowering. ACCORD, which targeted near-normal glycaemia, was stopped early because intensive therapy increased mortality, even though the primary composite cardiovascular outcome was not reduced convincingly[4]. Nine-year follow-up showed a neutral long-term effect on death and the primary cardiovascular outcome[26]. ADVANCE produced a more moderate message: Intensive control reduced some microvascular outcomes, but extended follow-up found no long-term mortality or macrovascular benefit[5,27]. VADT likewise showed no significant reduction in major cardiovascular events at 5.6 years despite a median HbA1c of 6.9% vs 8.4% with standard therapy, although later follow-up did show a delayed reduction in major cardiovascular events during the period of post-trial separation in glycaemic exposure[3,28]. Meta-analyses reconcile these apparently conflicting data. The 2009 collaborative meta-analysis concluded that more-intensive lowering modestly reduced major macrovascular events but increased major hypoglycaemia, and an updated 2024 systematic review similarly found reduced non-fatal myocardial infarction and several microvascular outcomes without compelling evidence that intensive HbA1c lowering alone broadly reduces mortality[29,30]. Taken together, the trial literature suggests that HbA1c is a meaningful cardiovascular risk marker, but that aggressive late-stage HbA1c normalisation is a blunt and sometimes hazardous prevention strategy (Table 1)[3-5,12,14,23,24,27,28,31].

Table 1 Summary of various trials evaluating glycated hemoglobin and cardiovascular outcome.

Ref.	Design and population	HbA1c exposure or contrast	Cardiovascular outcome	Main finding
Selvin et al[12]	Meta-analysis of observational studies in diabetes	Per 1% higher HbA1c	Composite CVD	Pooled relative risk 1.18 for CVD in type 2 diabetes
UKPDS 33[23]	RCT; newly diagnosed type 2 diabetes	Intensive 7.0% vs conventional 7.9%	Diabetes-related and vascular endpoints	12% lower any diabetes-related endpoint; 25% lower microvascular risk; no clear macrovascular benefit during trial
UKPDS 34[24]	RCT; overweight, newly diagnosed type 2 diabetes	Metformin 7.4% vs conventional 8.0%	Diabetes-related death and mortality	32% lower any diabetes-related endpoint; 42% lower diabetes-related death; 36% lower all-cause mortality
UKPDS 80[31]	Post-trial follow-up	Prior intensive vs prior conventional control	MI and death	Emergent legacy benefit: 15% lower MI and 13% lower all-cause death; metformin subgroup retained larger benefits
ACCORD[4]	RCT; long-standing, high-risk type 2 diabetes	Target < 6.0% vs standard 7.0%-7.9%	Composite major adverse cardiovascular events and death	No convincing primary cardiovascular gain; mortality increased with intensive strategy
ADVANCE/ADVANCE-ON[5,27]	RCT and observational follow-up	Intensive strategy targeting ≤ 6.5%	Vascular events, mortality	Reduced microvascular events; no durable macrovascular or mortality benefit in follow-up
VADT/15-year follow-up[3,28]	RCT and extended follow-up; veterans with long-standing type 2 diabetes	Median 6.9% vs 8.4%	Major cardiovascular events	Neutral at trial end; delayed reduction in major CVD events on longer follow-up
TECOS secondary analysis[14]	Secondary analysis; type 2 diabetes with established ASCVD	Baseline and time-varying HbA1c	HF hospitalisation, death, non-HF cardiovascular events	U-shaped association; lowest risk around HbA1c 7%

CVD: Cardiovascular disease; HbA1c: Glycated hemoglobin; HF: Heart failure; MI: Myocardial infarction; RCT: Randomised controlled trial; UKPDS: United Kingdom Prospective Diabetes Study.

MECHANISTIC LINKS AND ALTERNATIVE GLYCAEMIC MARKERS

The pathophysiological link between HbA1c and atherosclerotic disease is biologically convincing, even if it is not perfectly specific. Chronic hyperglycaemia drives the formation of advanced glycation end products (AGEs), increases receptor for AGE signalling, amplifies oxidative stress, promotes vascular inflammation, stiffens the extracellular matrix, and alters endothelial nitric oxide biology[32-34]. Hyperglycaemia also modifies lipoproteins: Glycated low density lipoprotein becomes more atherogenic and more readily retained in the arterial wall, while glycated high density lipoprotein may lose protective functions, thereby offering a plausible bridge between a higher HbA1c and accelerated plaque formation[35,36]. In that sense, HbA1c is not merely a statistical correlate; it is a proxy for multiple glyco-oxidative injuries that converge on the vessel wall (Figure 1).

Figure 1 Pathophysiological cascade linking chronic hyperglycemia to cardiovascular and renal outcomes in diabetes mellitus. Chronic hyperglycemia promotes the formation of advanced glycation end products and activation of their receptors, leading to oxidative stress, endothelial dysfunction, inflammation, and modification of circulating lipoproteins. These interconnected mechanisms drive vascular injury and atherogenesis. Downstream effects include endothelial activation, arterial stiffness, left ventricular remodeling, and microvascular damage, ultimately contributing to major clinical outcomes such as coronary artery disease, heart failure, stroke, chronic kidney disease, peripheral arterial disease, and increased mortality. This framework highlights the continuum from metabolic disturbance to organ damage and underscores the importance of early and comprehensive risk modification in diabetes care. LV: Left ventricular; ROS: Reactive oxygen species.

The heart-failure signal is equally plausible. Experimental and translational reviews implicate hyperglycaemia and AGE accumulation in myocardial fibrosis, impaired calcium handling, mitochondrial dysfunction, microvascular rarefaction, and impaired relaxation, all of which are central to diabetic cardiomyopathy[37-39]. This helps explain why HbA1c tracks heart-failure risk in cohort studies, but it also clarifies why HbA1c is imperfect: It captures mean exposure, not postprandial spikes, nocturnal hypoglycaemia, or visit-to-visit volatility. It can also mislead when erythrocyte turnover is abnormal or haemoglobin biology is altered, under which circumstances fructosamine or glycated albumin may be more reliable[1,40]. Thus, HbA1c is mechanistically relevant, but it is not a complete portrait of dysglycaemia.

This is why comparisons with other glycaemic markers matter. Fasting plasma glucose is cheaper and physiologically intuitive, but it represents only one moment in a highly variable metabolic day. Head-to-head prognostic data are mixed: In one type 2 diabetes cohort, fasting glucose and the triglyceride-glucose index outperformed HbA1c for cardiovascular event prediction, whereas in other settings HbA1c has shown stronger or more stable associations than fasting glucose with long-term adverse outcomes[41-43]. Fructosamine and glycated albumin, which reflect approximately 2-3 weeks of glycaemia, are increasingly attractive where HbA1c is unreliable[44]. In Atherosclerosis Risk in Communities study, both were associated with coronary heart disease, ischaemic stroke, HF, and mortality, with associations broadly similar to those seen for HbA1c, and they provided prognostic value comparable to HbA1c for incident diabetes complications[12,45]. Continuous glucose monitoring adds another dimension again. Time-in-range (TIR) captures glycaemic exposure and variability that HbA1c averages away; lower TIR has been linked with carotid intima-media thickness and adverse carotid wall characteristics, and post hoc analyses suggest that greater derived TIR is associated with lower major adverse cardiovascular events[46-48]. Yet for hard cardiovascular end points, TIR still has a shallower evidence base than HbA1c. The most defensible conclusion is therefore complementary, not competitive: HbA1c remains the anchor marker, but fructosamine/glycated albumin and CGM metrics are increasingly useful adjuncts in discordant or high-risk phenotypes.

CLINICAL IMPLICATIONS, GUIDELINES, AND FUTURE DIRECTIONS

For practical cardiovascular risk stratification, HbA1c should be used neither too narrowly nor too reverently. It is informative because it captures cumulative glycaemic exposure and repeatedly emerges as a major predictor in multivariable models. In a large Swedish analysis, glycohaemoglobin ranked among the strongest predictors of death and cardiovascular outcomes, alongside albuminuria, diabetes duration, systolic blood pressure, and low density lipoprotein cholesterol[49]. The problem is not relevance but over-interpretation. A raised HbA1c, particularly above about 8.0%-8.5%, usually signals important residual risk, but a very low HbA1c in older or intensively treated patients may also signal frailty, treatment burden, or hypoglycaemia exposure rather than vascular safety. The U-shaped findings from TECOS and mortality meta-analyses make this clinically hard to ignore[14,15]. Moreover, therapeutic delay matters: A real-world analysis showed that delaying treatment intensification by one year in people with poor glycaemic control increased the risks of myocardial infarction, HF, stroke, and composite cardiovascular events[50]. The implication is straightforward: HbA1c is best used as one element within a clustered-risk assessment, not as a solitary triage tool.

Current guidance from the American Diabetes Association, the European Association for the Study of Diabetes, and the European Society of Cardiology converges on a pragmatic position: Pursue safe glycaemic improvement early, individualise HbA1c targets thereafter, and do not allow HbA1c chasing to displace therapies that have proven cardiovascular or renal benefit. The American Diabetes Association’s current glycaemic goals retain an HbA1c target of < 7% for many non-pregnant adults, while explicitly allowing tighter or looser targets according to hypoglycaemia risk, life expectancy, and treatment burden[51]. The American Diabetes Association/European Association for the Study of Diabetes consensus report frames HbA1c within person-centred care rather than as an autonomous endpoint[52]. Contemporary summaries of the updated recommendations further note that in type 2 diabetes with high cardiovascular risk, sodium-glucose transporter 2 inhibitors or glucagon-like peptide-1 receptor agonists are recommended irrespective of HbA1c level[53]. This shift is evidence-based: Meta-analyses show that sodium-glucose transporter 2 inhibitors reduce major adverse cardiovascular events, cardiovascular death or heart-failure hospitalisation, and kidney outcomes, while glucagon-like peptide-1 receptor agonists reduce major cardiovascular events, stroke, cardiovascular mortality, and all-cause mortality[54-56]. HbA1c therefore remains necessary for risk framing and monitoring, but it is no longer sufficient for treatment selection in cardiovascular prevention. Table 2 shows various guidelines about HbA1c and cardiovascular risk[47,51,57].

Table 2 Various guidelines and their summaries about glycated hemoglobin and cardiovascular risk.

Guideline or consensus source	Glycaemic target statement	Cardiovascular implication
ADA standards 2025[51]	HbA1c < 7% is appropriate for many non-pregnant adults; targets should be individualised, with more stringent or less stringent goals according to safety and burden	HbA1c remains important, but glycaemic goals should not be pursued at the cost of severe hypoglycaemia or excessive treatment burden
ADA/European Association for the Study of Diabetes consensus 2022[47]	HbA1c remains central to monitoring, but target-setting should be person-centred and tailored to comorbidity, duration, patient preference, and treatment risk	In ASCVD, HF, or chronic kidney disease, drug selection should prioritise organ protection, not HbA1c lowering alone
European Society of Cardiology diabetes/CVD guideline 2023[57]	Glycaemic management should be personalised within comprehensive cardiovascular risk reduction	HbA1c should be interpreted within broader CVD prevention, with strong emphasis on therapies proven to reduce ASCVD, HF, and renal events

ADA: American Diabetes Association; CVD: Cardiovascular disease; HbA1c: Glycated hemoglobin; HF: Heart failure.

The main research gap is no longer whether HbA1c predicts cardiovascular risk, but how it should be integrated with richer phenotyping. Future work should compare average HbA1c, HbA1c variability, fructosamine/glycated albumin, and CGM metrics head-to-head for hard atherosclerotic and heart-failure end points rather than surrogate outcomes alone. HbA1c variability is particularly promising because it appears to add risk information beyond mean control, while emerging CGM studies suggest that TIR may converge more closely with vascular biology than HbA1c alone[22,58-60].

CONCLUSION

HbA1c remains a clinically indispensable biomarker of cardiovascular risk in diabetes because it captures cumulative glycaemic injury, predicts atherosclerotic and heart-failure outcomes in observational studies, and retains clear value for treatment monitoring. Yet the totality of evidence shows that it is an incomplete cardiovascular compass. Early, safe HbA1c improvement probably matters most, especially through legacy effects, whereas late aggressive normalisation in long-standing high-risk diabetes offers diminishing macrovascular returns and greater potential for harm. In modern practice, HbA1c should anchor cardiovascular risk assessment, but only alongside variability, renal and albuminuric burden, traditional risk factors, and preferential use of therapies with proven cardiovascular benefit.