Intended for healthcare professionals


Impaired glucose tolerance

BMJ 1996; 312 doi: (Published 03 February 1996) Cite this as: BMJ 1996;312:264
  1. Melanie J Davies,
  2. I Peter Gray
  1. Consultant physician Department of Diabetes, Leicester Royal Infirmary, Leicester LE1 5WW
  2. Professor of chemical pathology University of Witwatersrand, Parktown, Johannesburg 2193, Republic of South Africa

    Detection and follow up should aim to reduce excess morbidity and mortality

    Impaired glucose tolerance is defined as a fasting plasma glucose concentration of less than 7.8 mmol/l and between 7.8 and 11.1 mmol/l two hours after a 75 g oral glucose load. This definition was first established in 1980 by the World Health Organisation, replacing terms such as “borderline” or “chemical” diabetes.1 It is based on long term prospective studies which conclude that individuals with lesser degrees of glucose intolerance are not at risk of microvascular complications such as retinopathy.2 The advent of health promotion clinics and screening programmes is likely to mean higher rates of detection. However, the clinical significance of impaired glucose tolerance remains unclear.3 Should people who are found to have impaired glucose tolerance be followed up, and what treatment, if any, should they receive?

    Impaired glucose tolerance is common; it affects about 11% of people aged 20-74 years in the United States and 17% of those aged 40-65 years in Britain.4 5 The pathogenesis is controversial, particularly the question of whether it is insulin resistance or insulin deficiency that predominates. (This may have implications for treatment since potential therapeutic agents have quite different mechanisms of action: sulphonylureas act by increasing insulin secretion, with some reports that they restore the early insulin release from pancreatic beta cells (the first phase insulin response); whereas newer agents such as the thiazolidinedione derivatives act by reducing insulin resistance.) Some studies have shown that people with impaired glucose tolerance have evidence of insulin resistance and hyperinsulinaemia.6 However, the first phase insulin response, thought to be a critical factor in determining overall glucose tolerance, has been shown to be reduced in these people, showing that the development of impaired glucose tolerance requires both insulin resistance and impaired insulin secretion.7 8 O'Rahilly et al showed that the normal pattern of pulsatile insulin secretion was lost in people with impaired glucose tolerance, and there may be qualitative as well as quantitative abnormalities of insulin secretion.9

    People with impaired glucose tolerance have increased mortality from cardiovascular disease: the Whitehall study found that impaired glucose tolerance doubled the risk of death from coronary artery disease among middle aged male civil servants.10 This has traditionally been ascribed to factors associated with glucose intolerance that exacerbate the atherogenic process. However, it has not been possible to explain the association between impaired glucose tolerance and cardiovascular disease by generally accepted risk factors such as diastolic hypertension, hypercholesterolaemia, or cigarette smoking; and epidemiological studies have found no association between coronary artery disease and either blood glucose concentration or duration of diabetes.11 12 Reports that lipoprotein(a) concentrations may be increased in people with impaired glucose tolerance have limited implications for treatment as there are no effective agents which significantly lower lipoprotein(a).13

    The natural history of impaired glucose tolerance is well documented. In a 10 year follow up study, 15% of people with impaired glucose tolerance went on to develop non-insulin dependent diabetes, 22% remained glucose intolerant, and the majority (53%) improved.14 Increasing age was an independent risk factor for developing diabetes, although other studies have found no such effect.15 Patients with transient impaired glucose tolerance tend to revert to normal within about six months, but they remain at increased long term risk of developing non-insulin dependent diabetes.16 17 By the time they develop diabetes, 50% will already have established complications, 16% coronary artery disease, and 30% retinopathy.18 This suggests that intervening at an earlier stage may be beneficial. But who might benefit from intervention, and what form should the intervention take?

    There are no clear biochemical markers that predict those at particular risk of progression to diabetes. In prospective studies the only consistent predictor was the initial degree of hyperglycaemia.14 15 19 Subjects who did not go on to develop diabetes had mean fasting plasma glucose levels of 5.2mmol/l, compared to mean levels of 5.7mmol/l in those who developed diabetes.15 Drug and dietary interventions have shown conflicting results regarding the progression of impaired glucose tolerance to non-insulin dependent diabetes. More recently, diet and exercise intervention programmes have shown some encouraging results and reduced the risk of developing non-insulin dependent diabetes in patients with impaired glucose tolerance. A formal programme aiming to increase levels of physical activity in people with impaired glucose tolerance gave a relative risk of developing diabetes of 0.37 compared with controls.15

    While many aspects of impaired glucose tolerance remain unclear, what is clear is that people with impaired glucose tolerance are susceptible to deteriorating glucose tolerance and macrovascular disease. Evidence suggests that once these patients are identified they should be advised about modifying their lifestyle, in particular to stop smoking, to lose weight if they are obese, and to maintain a reasonable level of physical fitness. They should be screened and treated for cardiovascular risk factors, particularly for hypertension and adverse fasting lipid profile (including high density lipoprotein and low density lipoprotein cholesterol and triglyceride estimation). Annual follow up examinations should include a test of fasting plasma glucose concentration, since this is the best predictor of risk of developing diabetes. A suggested cut off, based on prospective studies,14 19 20 is 5.5mmol/l, and indeed this figure is the entry criterion for two ongoing studies (the fasting hyperglycaemia study and the early diabetes intervention trial) which aim to examine the effects of dietary intervention, exercise, and treatment with agents such as sulphonylureas and acarbose.


    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.
    19. 19.
    20. 20.
    View Abstract