A simple strategy for screening for glucose intolerance, using glycated haemoglobin, in individuals admitted with acute coronary syndrome

Authors


Dr Nitin Gholap, Department of Health Sciences, University of Leicester, Leicester, UK. E-mail: ng94@le.ac.uk

Abstract

Diabet. Med. 29, 838–843 (2012)

Abstract

Glucose intolerance is common but often remains undiagnosed and untreated in people with acute coronary syndrome. The best approach to screening for glucose intolerance post-acute coronary syndrome remains debated. The World Health Organization has recently advocated the use of HbA1c in diagnosing Type 2 diabetes. A screening strategy using HbA1c as the preferred test would be pragmatic and improve early detection and management of glucose intolerance in acute coronary care practice. In this commentary, we discuss the relevant literature and guidelines in this area and propose a simple and pragmatic algorithm based on the use of HbA1c to screen for glucose intolerance during and after admission with acute coronary syndrome.

Introduction

Previously undiagnosed glucose intolerance, including Type 2 diabetes and impaired glucose regulation (pre-diabetes, impaired glucose tolerance and/or impaired fasting glucose defined using World Health Organization 1999 criteria), are highly prevalent and adversely impact prognosis in individuals with acute coronary syndrome. Systematic screening on admission with acute coronary syndrome would enable early detection and aggressive management of these glucose abnormalities, potentially improving outcomes. Current screening methods using either fasting plasma glucose or the oral glucose tolerance test have limitations in the acute care setting, hampering routine screening for glucose intolerance following acute coronary syndrome. The World Health Organization recently approved glycated haemoglobin (HbA1c) as an additional diagnostic test for Type 2 diabetes, bringing an opportunity to improve screening post-acute coronary syndrome. In this commentary, we propose a simple screening algorithm based on the new diagnostic criteria and advocate HbA1c as a preferred diagnostic test over fasting plasma glucose or the oral glucose tolerance test in individuals with acute coronary syndrome.

Undiagnosed glucose intolerance in people with acute coronary syndrome: current evidence

Type 2 diabetes and impaired glucose regulation are risk factors for development of coronary heart disease [1]. These conditions are seen undiagnosed in 40–45% of individuals admitted with acute coronary syndrome, with an additional 20% already diagnosed with diabetes [2–4]. Furthermore, acutely elevated blood glucose levels, irrespective of underlying diabetes status is common following admission with acute coronary syndrome [5]. While admission hyperglycaemia may indicate a stress-related epiphenomenon, in some patients it represents undiagnosed glucose intolerance [3]. In the Glucose tolerance in patients with Acute Myocardial Infarction (GAMI) study, patients without diabetes admitted for acute coronary syndrome underwent an oral glucose tolerance test [3]. Only 34% showed normal glucose tolerance, with Type 2 diabetes in 31% and impaired glucose regulation in 35%. Similar results were observed in other studies; however, higher prevalence is reported in certain ethnic groups such as South Asians [6]. Furthermore, hyperglycaemia is more commonly detected following a glucose challenge during an oral glucose tolerance test compared with a fasting test in individuals with acute coronary syndrome [2,7]. This suggests that an oral glucose tolerance test may be essential for accurate diagnosis of glucose intolerance following acute coronary syndrome.

Prognostic importance of glucose intolerance and the case for screening post-acute coronary syndrome

People with established Type 2 diabetes have a worse prognosis following acute coronary syndrome. Similarly, people newly diagnosed with Type 2 diabetes post-acute coronary syndrome also have adverse outcomes [8]. The excess risk associated with diabetes can be improved using evidence-based interventions, including early reperfusion and revascularization, and secondary prevention measures such as statins, anti-platelet agents, renin–angiotensin blockers and beta-blockers [9]. Indeed, the absolute benefits from such interventions are greater in people with diabetes [9]. In the Euro Heart Survey, the numbers needed to treat with secondary prevention measures to save one life (n = 24 vs. 1826) or prevent a cardiovascular event (n = 32 vs.141) were lower in people with diabetes compared with those without diabetes [9]. A similar trend was observed for revascularization therapies [9]. However, evidence-based therapies may be underused in people with diabetes following acute coronary syndrome, and this could explain the reported adverse outcomes [10]. Early diagnosis of Type 2 diabetes would enable aggressive management of cardiovascular risk factors with evidence-based therapies. Furthermore, impaired glucose regulation carries a significant risk of progression to Type 2 diabetes and increased cardiovascular morbidity and mortality [11]. Early diagnosis of impaired glucose regulation allows initiation of lifestyle interventions, including diet and exercise, to prevent Type 2 diabetes and associated complications.

Current guidelines on screening for glucose intolerance in people admitted with acute coronary syndrome

Most guidelines acknowledge the importance of routine screening for Type 2 diabetes and impaired glucose regulation post-acute coronary syndrome [1,13–15]. Currently, fasting plasma glucose, random plasma glucose or the oral glucose tolerance test are available to diagnose these glucose abnormalities (Table 1) [12]. However, these tests have limitations and therefore the best screening approach in the acute setting of acute coronary syndrome is debated. European guidelines advocate performing an oral glucose tolerance test before or soon after discharge in people with ST-elevation acute myocardial infarction [1,13]. The Joint British Society Guidelines (JBS-2) recommend using either fasting plasma glucose or an oral glucose tolerance test during admission with an acute cardiovascular event [14]. While the oral glucose tolerance test is considered the standard diagnostic test, it has disadvantages during an acute admission. Requirements of regular carbohydrate diet and physical activity levels prior to the test, together with severity of myocardial damage and timing of the test in relation to an index event, can affect reproducibility of results [16]. An oral glucose tolerance test performed within 4–5 days of acute coronary syndrome may be less reliable, especially if there has been extensive myocardial infarction [16]. Furthermore, the test is resource intensive and involves less convenient overnight patient fasting. In the Euro Heart Survey, the recommended oral glucose tolerance test was performed in only 56% of the patients, reflecting these various issues [7]. The alternative test used by many clinicians on pragmatic grounds is fasting plasma glucose. However, fasting plasma glucose alone lacks sensitivity for Type 2 diabetes, even when lower diagnostic thresholds are used, and may not detect glucose intolerance in over 40% of acute coronary syndrome cases [7].

Table 1.   The current World Health Organization criteria for diagnosis of Type 2 diabetes, impaired glucose tolerance and impaired fasting glucose [12]
 DiabetesImpaired glucose toleranceImpaired fasting glucose
  1. *Conditions precluding accurate measurement of HbA1c need to be excluded (see Fig. 1).

  2. †In the presence of classical symptoms, one abnormal value is diagnostic of diabetes. In the absence of symptoms, two abnormal values are required to confirm the diagnosis of diabetes.

  3. Any of the above criteria are suitable for diagnosis of diabetes.

  4. In the absence of unequivocal hyperglycaemia, repeat testing is required to confirm the diagnosis.

  5. Plasma glucose and HbA1c values are for venous plasma samples.

  6. Fasting is defined as no caloric intake for at least 8 h.

  7. ‡The oral glucose tolerance test should be conducted using a glucose load containing the equivalent of 75 g anhydrous glucose dissolved in water. A fasting and 2-h post-glucose loading sample is required.

HbA1c*≥ 48 mmol/mol (≥ 6.5%)†
Fasting plasma glucose≥ 7.0 mmol/l†≥ 6.1 mmol/l and < 7.0 mmol/l
Random plasma glucose≥ 11.1 mmol/l†
Oral glucose tolerance test‡2-h plasma glucose ≥ 11.1 mmol/l†2-h plasma glucose ≥ 7.8 mmol/l and < 11.1 mmol/l 

Recently, National Institute for Health and Clinical Excellence (NICE) guidelines on the management of hyperglycaemia in acute coronary syndrome have advocated any hyperglycaemia (blood glucose > 11.0 mmol/l) without known diabetes, be followed up with an HbA1c measurement before discharge and fasting plasma glucose test 4 days after the onset of acute coronary syndrome [15]. NICE recommend against routine use of the oral glucose tolerance test in patients with acute coronary syndrome with fasting plasma glucose and HbA1c in the normal range [15]. However, guidance on categorization of glycaemic status of those with elevated HbA1c and fasting plasma glucose, as well as screening for diabetes in those without hyperglycaemia, is less clear. The lack of simple strategy for early identification of glucose intolerance in acute coronary syndrome is potentially leaving many people undiagnosed and under-treated.

An approach using HbA1c as a preferred diagnostic tool

The World Health Organization has recently endorsed use of HbA1c as an additional diagnostic test for Type 2 diabetes (Table 1) [17]. An HbA1c test has several advantages over fasting plasma glucose or an oral glucose tolerance test in an acute setting. The test can be performed in the non-fasting state and reflects average glucose concentration over the preceding 2–3 months. Furthermore, HbA1c is a strong and independent predictor of incident Type 2 diabetes and a stronger predictor of cardiovascular disease and mortality in people without diabetes and separately in people with acute coronary syndrome [2,18,19]. In the GAMI study, HbA1c independently predicted glucose intolerance [odds ratio (95% CI): 2.58 (1.17–6.09, P = 0.024) in people with acute coronary syndrome without known diabetes, correlating closely with the 2-h plasma glucose in an oral glucose tolerance test (r = 0.39, P < 0.0001) [2]. Furthermore, an HbA1c≥ 30 mmol/mol (4.9%) had sensitivity and specificity of 79 and 49% for detecting undiagnosed diabetes, with the area under curve of 0.685 (P = 0.001) [2]. In an observational study (n = 4176) of people with ST-elevation myocardial infarction without diabetes, increasing quartiles of HbA1c levels, even below the 48 mmol/mol (6.5%) threshold, identified people with higher cardiovascular risk factors at baseline and independently predicted risk of mortality (hazard ratio: 1.2; 95% CI 1.0–1.3 per interquartile range) [19]. Moreover, HbA1c is less affected by stress-related acute changes in blood glucose levels, making it a reliable test.

An approach to screening for Type 2 diabetes and impaired glucose regulation following acute coronary syndrome, using HbA1c as the preferred initial diagnostic test, therefore seems pragmatic. Glycaemic categorization using the oral glucose tolerance test is still important in acute coronary syndrome as post-challenge hyperglycaemia may be the predominant abnormality [7]. However, as the test is performed infrequently in the acute setting [4], an approach reducing the number of oral glucose tolerance tests is likely to be more acceptable.

We propose a simple screening algorithm based on an HbA1c test performed before discharge from hospital and to consider an oral glucose tolerance test according to HbA1c categories of ≥ 48 mmol/mol (6.5%), between 42 and < 48 mmol/mol (6.0 and 6.4%) and < 42 mmol/mol (< 6.0%) (Fig. 1). We suggest specific HbA1c cut points based on pragmatic grounds, as studies in this area of acute coronary syndrome are limited.

Figure 1.

 Algorithm for screening for Type 2 diabetes and impaired glucose regulation in patients with acute coronary syndrome.

Above HbA1c≥ 48 mmol/mol (≥ 6.5%), individuals should be assessed for symptoms of diabetes (Fig. 1), ruling out other causes, for example polyuria attributable to diuretic therapy. In those with unequivocal symptoms the diagnosis is confirmed [17]. Conversely, those with ambiguous or absent symptoms should undergo a confirmatory HbA1c measurement 4–8 weeks post-discharge for consistency and to counteract any potential laboratory errors on the first occasion. The time interval of 4–8 weeks is based on logistical grounds and can be shortened to 2–4 weeks if feasible. If repeat HbA1c is also ≥ 48 mmol/mol (≥ 6.5%), the diagnosis of Type 2 diabetes is confirmed [17]. Conversely, those with repeat HbA1c between 42 and < 48 mmol/mol (6.0 and 6.4%) should undergo an oral glucose tolerance test.

Studies suggest that, in individuals with an HbA1c of < 48 mmol/mol (6.5%) with acute coronary syndrome, many have undiagnosed diabetes defined on an oral glucose tolerance test [2,7]. This highlights a discrepancy in Type 2 diabetes diagnosed by HbA1c and the oral glucose tolerance test [17]. HbA1c levels < 48 mmol/mol (< 6.5%) do not rule out Type 2 diabetes defined by World Health Organization glucose criteria [17]. Furthermore, an international expert committee report described an HbA1c between 42 and < 48 mmol/mol (6.0 and 6.4%) as ‘high risk’ and the equivalent of impaired glucose regulation [20]. However, the World Health Organization has not recommended use of HbA1c for diagnosis of impaired glucose regulation and an oral glucose tolerance test is still required for this diagnosis [17].

HbA1c is a continuous risk factor for mortality, as demonstrated in an epidemiological study where the majority of deaths (75%) over 6 years follow-up occurred in people without diabetes and moderately elevated HbA1c levels between 37 and 52 mmol/mol (5.5 and 6.9%) [21]. Similarly, in patients with ST-elevation myocardial infarction without diabetes, HbA1c levels below the diagnostic threshold of 48 mmol/mol (6.5%) demonstrate an incremental association with increased mortality [19]. On this background, we recommend performing an oral glucose tolerance test 4–8 weeks post-discharge in those with HbA1c between 42 and < 48 mmol/mol (6.0 and 6.4%) for accurate categorization of glucose intolerance (Fig. 1).

The remaining category is patients with an HbA1c below 42 mmol/mol (6.0%), who we suggest should not undergo an oral glucose tolerance test, but annual monitoring with HbA1c (Fig. 1). This is based on pragmatic grounds with limited evidence available. In a recent study of acutely ill patients, including cardiovascular events, optimal HbA1c screening cut points for predicting diabetes diagnosed using fasting plasma glucose or the oral glucose tolerance test after recovery from acute illness was 42 mmo/mol (6.0%) (sensitivity 76.9%, specificity 87.3% and area under the curve of 0.868) [22]. Long-term prognostic data suggest, in people without known diabetes, an HbA1c cut point of > 42 mmol/mol (> 6.0%) may be a useful marker for identifying people at risk of diabetes, cardiovascular disease and mortality [18]. Furthermore, our preference for a threshold of < 42 mmol/mol (< 6.0%) for not performing an oral glucose tolerance test is in line with NICE recommendations [15].

However, we differ with the NICE approach of screening for diabetes post-acute coronary syndrome by not recommending fasting plasma glucose in addition to HbA1c [15]. HbA1c alone is a strong predictor of undiagnosed glucose intolerance in people with acute coronary syndrome [2]. Furthermore, higher HbA1c levels are associated with increased risk of developing, diabetes, cardiovascular disease and mortality, independent of baseline fasting plasma glucose [18]. In addition, fasting plasma glucose can be acutely elevated and therefore unreliable in the first 2 days of an acute event and in a large myocardial infarction [16]. NICE has suggested fasting plasma glucose testing should not be conducted within the first 4 days of the acute event [15]. However, in the current era of early reperfusion therapies, many patients with acute coronary syndrome are discharged earlier. Moreover, the logistical benefits of using a non-fasting HbA1c test can be offset by the inconvenience of an additional fasting test, consequently affecting the screening uptake. Nevertheless, we have suggested an alternative algorithm using fasting plasma glucose alone (Fig. 1). Further validation of this algorithm in prospective studies may be required in the contemporary UK population to resolve some of these uncertainties around HbA1c screening cut points and the use of fasting plasma glucose.

Measuring HbA1c assumes International Federation of Clinical Chemistry (IFCC) standardized laboratory assays are used. Furthermore, conditions precluding accurate measurement of HbA1c concentration for diagnosis are excluded, including abnormalities of red cell turnover, chronic renal or liver failure and chronic use of certain medications (Fig. 1). In addition, plans to repeat HbA1c measurements or the oral glucose tolerance test following discharge should be communicated to the patients and their general practitioners. Finally, people with acute coronary syndrome and normal glucose tolerance should undergo annual surveillance with HbA1c as incident impaired glucose regulation and diabetes is higher compared with the general population [23].

Conclusions

Glucose intolerance is common, but frequently remains undiagnosed and untreated in people with acute coronary syndrome. Newly diagnosed glucose intolerance is an independent predictor of poor outcomes post-acute coronary syndrome. While strategies of early detection and intensive management of Type 2 diabetes and impaired glucose regulation post-acute coronary syndrome have potential to improve outcomes, a lack of a simple and practical screening approach suitable in an acute coronary care practice is a significant barrier. A screening approach based on HbA1c as a preferred test seems pragmatic following admission with acute coronary syndrome and would improve early diagnosis and subsequent care of people at risk of glucose intolerance.

Competing interests

All authors declare that they have no support from companies for the submitted work. NG, KK and MJD are currently conducting studies as a part of the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) on early detection of diabetes in people admitted with acute coronary syndrome. NG, MJD, SAM, IS and KK have no non-financial interests that may be relevant to the submitted work. KK and MJD are advisors to the Department of Health National Health Service (NHS) Health Checks programme and are members (KK Chair) of the NICE programme development group on ‘Preventing Type 2 diabetes: risk identifications and interventions for individuals at high risk’.

Acknowledgments

The Department of Health Sciences and Cardiovascular Sciences, University of Leicester is grateful for support from the Collaboration for Leadership in Applied Health Research and Care (CLAHRC) Scheme and the NIHR Biomedical Research Centre scheme. NG is funded by the NIHR CLARHC PhD studentship. SAM has received a fellowship from Novo Nordisk Research Foundation UK for a PhD.

Ancillary