Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Hyperglycaemia and outcomes after MI

Review
Results
Comment

A randomised controlled trial showed that insulin treatment of people with MI and concomitant hyperglycaemia lowered mortality ( Bandolier 48 ). This begs the question of what excess risk is associated with stress hyperglycaemia after a heart attack in people with and without diabetes. A systematic review [1] tells us that stress hyperglycaemia carries high additional risk for death and heart failure in diabetics and non-diabetics.

The reason or reasons why this may be so are not entirely clear, but the original observation that there was a high prevalence of glycosuria in people with a heart attack goes back 70 years.

Review


Researchers from McMaster and the Karolinska performed an extensive literature survey, but only English language reports were used. For inclusion a study had to:

Patients were regarded as diabetic if they had a reported history of diabetes. Hyperglycaemia was defined according to criteria set by the original studies.

Results


Fifteen cohort studies were eventually included in the review. Mean glucose concentrations were consistently higher in patients who died than those who did not.

Seven studies (Figure 1) reported on in-hospital mortality in patients without diabetes according to whether patients were hyperglycaemic or not (defined as blood glucose of greater than 6.0 to greater than 8.0 mmol/L). The average mortality was 25% in those with hyperglycaemia compared with 6% in those without hyperglycaemia (Table). For every five patients with hyperglycaemia after myocardial infarction one more died in hospital than would have done without hyperglycaemia.

Figure 1: In-hospital mortality in MI patients without diabetes

Table: Summary of results in MI patients without and with diabetes

Outcome Patients with hyperglycaemia number/total (%; 95% CI) without hyperglycaemia number/total (%; 95% CI) Relative risk number/total (95% CI) NNT (95% CI)
In-hospital death Non-diabetics 136/535 25 (22 to 29) 84/1321 6 (5 to 8) 3.9 (2.9 to 5.4) 5.2 (4.3 to 6.6)
  Diabetics 150/506 30 (26 to 34) 32/182 18 (12 to 23) 1.7 (1.2 to 2.4) 8.3 (5.3 to 19)
CHF or shock Non-diabetics 55/179 31 (24 to 37) 531/3704 9 (8 to 10) 3.3 (2.7 to 4.0) 4.6 (3.5 to 6.6)
  Diabetics 10/306 3 (1 to 5) 12/357 3 (1 to 5) 1.0 (0.4 to 2.2) N/A

 


Four studies (Figure 2) reported on in-hospital mortality in patients with diabetes according to whether patients were hyperglycaemic or not (defined as blood glucose of greater than 10 to greater than 11 mmol/L). The average mortality was 30% in those with hyperglycaemia compared with 18% in those without hyperglycaemia ( Table ). For every eight patients with hyperglycaemia after myocardial infarction one more died in hospital than would have done without hyperglycaemia.

Figure 2: In-hospital mortality in MI patients with diabetes


Four studies (Figure 3) reported on development of congestive heart failure or cardiogenic shock in patients without diabetes according to whether patients were hyperglycaemic or not (defined as blood glucose of greater than 8.0 to greater than 11 mmol/L). The average rate was 31% in those with hyperglycaemia compared with 9% in those without hyperglycaemia ( Table ). For every five patients with hyperglycaemia after myocardial infarction one more developed heart failure or shock than would have done without hyperglycaemia.

Figure 3: Congestive heart failure or cardiogenic shock in MI patients without diabetes (open circles) or with diabetes (filled circle)


One study (Figure 3) reported on development of congestive heart failure or cardiogenic shock in patients with diabetes according to whether patients were hyperglycaemic or not (defined as blood glucose of greater than 10 mmol/L). The rate was 3% in both those with hyperglycaemia and without hyperglycaemia ( Table ).

Comment


This is an interesting adjunct to the DIGAMI trial of intensive insulin therapy in diabetics with myocardial infarction and hyperglycaemia of more than 11 mmol/L ( Bandolier 48 ). That showed insulin therapy to reduce one-year mortality to 17% from 26% in controls. In this review of cohort studies the in-hospital death rate was 18% in diabetics without hyperglycaemia and 30% in those with hyperglycaemia.

The comparisons are obvious, but the review is about risks rather than about causation and treatments. The Table includes a calculation of NNT as an indicator of the absolute increased risk, which was an additional 1 in 5 risk of dying or having heart failure or shock for non-diabetics, and a 1 in 8 risk of dying for diabetics. These are big risks.

Do we understand what is going on, and can we do anything about it? The paper [1] gives a good discussion of the likely cause (excess fatty acids from relative insulin deficiency), though again it could be that hyperglycaemia is simply a marker of greater myocardial damage. The cardioprotective effects of insulin and beta-blockers are also discussed.

There's no simple answer, though. Blood glucose is clearly an important marker for morbidity and mortality. This is one occasion where the call for research to find out whether reversing the stress hyperglycaemia improves outcomes rings true. This looks increasingly important.

Reference:

  1. SE Capes et al. Stress hyperglycaemia and increased risk of death after myocardial infarction in patients with and without diabetes: a systematic review. Lancet 2000 355: 773-778.
previous or next story in this issue