Editorials

Insulin infusion in diabetic patients with acute myocardial infarction

BMJ 1996; 313 doi: https://doi.org/10.1136/bmj.313.7058.639 (Published 14 September 1996) Cite this as: BMJ 1996;313:639
  1. Gail Davey, Clinical research fellow,
  2. Paul Mckeigue, Senior lecturer
  1. Department of Epidemiology and Population Sciences, London School of Hygiene and Tropical Medicine, London WC1E 7HT

    Effective in diabetes, but patients with glucose intolerance may also benefit

    Despite the improvements brought about by thrombolysis, β blockade, and treatment with angiotensin converting enzyme inhibitors, the mortality in diabetic patients admitted to hospital for acute myocardial infarction remains up to twice that in non-diabetic patients matched for age and sex.1 This increased mortality is apparent both during hospital stay and at longer term follow up in a range of populations, including white Americans,2 Mexican-Americans,3 and people from the Asian subcontinent.4

    The excess mortality is shared by those with diabetes already diagnosed and those with previously undiagnosed diabetes.3 5 The higher mortality of diabetic patients compared with non-diabetic patients is not explained by differences in the extent or severity of coronary artery disease or by differences in infarct size.6 7 8 In patients with longstanding diabetes, complications such as microvascular disease and autonomic neuropathy could predispose to poorer outcome. A more specific reason for poor outcome is the high incidence of cardiac failure in diabetic patients with myocardial infarction, which is also not explained by differences in infarct size.7 8 The occurrence of cardiac failure in diabetes, often associated with structural changes in the myocardium, has been termed diabetic cardiomyopathy.9 Similar structural changes are seen in animals with experimentally induced diabetes: defects in myocardial contractile protein, increased collagen synthesis, and myocardial fibrosis.

    Myocardial function is further impaired in diabetic patients by the metabolic changes that occur in the early stages of myocardial infarction: insulin resistance and hyperglycaemia are induced by release of catecholamines, cortisol, glucagon, and growth hormone.10 At the same time, secretion of insulin by the pancreatic islets is reduced,11 which impairs the ability to compensate for this state of insulin resistance. The combination of low insulin concentrations and elevated catecholamine concentrations increases release of non-esterified fatty acids, which augment myocardial oxygen requirements and depress mechanical performance.12

    It is thus plausible that controlling plasma glucose and non-esterified fatty acid concentrations from an early stage by continuous insulin infusion could help to preserve myocardial function and reduce the mortality of diabetic patients with myocardial infarction. Until recently no large randomised study of the effect of this intervention has been reported. In the 1980s two trials were reported: one showed reduction in one year mortality with insulin infusion,13 but the other showed no alteration in death rate.14 Both of these studies were non-randomised and used historical controls. Difficulties in recruiting enough patients for large trials,15 the requirement for central venous access to reduce the risk of phlebitis, and the advent of thrombolytic treatments lessened interest in the use of insulin infusions.

    Last year, the results of a multicentre trial of insulin-glucose infusion in Swedish diabetic patients with acute myocardial infarction were reported.16 Over 600 patients were randomised: 542 with previously diagnosed diabetes and 78 with plasma glucose concentration at admission greater than 11 mmol/l. In those patients randomised to receive insulin-glucose infusion, blood glucose concentration was kept within a target range of 7-10.9 mmol/l, with hourly or two hourly monitoring of glucose concentrations, for at least 24 hours after admission. Although hypoglycaemic episodes occurred in 15% of patients randomised to the insulin-glucose infusion, these episodes did not lead to arrhythmias or other untoward events. Mortality at one year was 29% lower in the intervention group than in the control group (95% confidence intervals for the reduction 4% to 51%), with 57 deaths in the intervention group compared with 82 deaths in the controls (P = 0.03). More than half the reduction in mortality was accounted for by deaths after the first three months, suggesting that the intervention had a lasting effect on myocardial function.

    Why did this trial show reduction of plasma glucose concentrations and mortality when others failed? As usual, the key strength of the study was that its sample size was large enough to have adequate statistical power: more than 1200 patients had to be identified over three years to obtain 620 who were able and willing to participate. The infusion protocol may also have helped to control plasma glucose more rapidly than protocols used in previous studies.14 15

    What relevance might the findings have to people without a history of diabetes who present with myocardial infarction and a plasma glucose concentration at admission of 8-11 mmol/l? There is evidence that mortality in hospital is increased in this group even when patients with a previous diagnosis of diabetes are excluded.4 17 During the 1960s and 1970s, combined infusions of glucose, insulin, and potassium were widely used in the acute management of myocardial infarction even in non-diabetic patients, especially in Italy. In a small randomised study, this seemed to increase ejection fraction and to reduce pulmonary artery end diastolic pressure.18

    Despite the wide confidence intervals for the size of effect, the Swedish trial is impressive evidence that rigorous control of plasma glucose by insulin infusion improves the long term outcome of myocardial infarction for patients with established diabetes or plasma glucose concentration greater than 11 mmol/l on admission. The possibility that glucose and insulin infusions could benefit patients with lesser degrees of hyperglycaemia is one worth investigating in future trials.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    View Abstract

    Sign in

    Log in through your institution

    Free trial

    Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
    Sign up for a free trial

    Subscribe