Clinical Review Clinical evidence

Glycaemic control in diabetes

BMJ 1999; 319 doi: (Published 10 July 1999) Cite this as: BMJ 1999;319:104
  1. William H Herman, associate professor of medicine (wherman{at} on behalf of Clinical Evidence
  1. Division of Endocrinology and Metabolism, University of Michigan Medical Center, Ann Arbor, MI 48109-0354, USA

    This review is one of over 60 chapters included in the first issue of Clinical Evidence, a new information resource for clinicians available from 28 June 1999. The compendium will be updated and expanded every six months. Future issues of Clinical Evidence will summarise the relevant results of the UK prospective diabetes study.

    Key messages

    • We found strong evidence that intensive compared with conventional treatment reduces the development and progression of microvascular and neuropathic complications in both type 1 and type 2 diabetes

    • RCTs have found that intensive treatment causes hypoglycaemia and weight gain without adverse impact on neuropsychological function or quality of life

    • We found no evidence that intensive treatment reduces adverse cardiovascular outcomes

    • Large RCTs have found that diabetic complications increase with HbA1c concentrations above the non-diabetic range


    Definition: Diabetes mellitus is a group of metabolic diseases characterised by hyperglycaemia (fasting plasma glucose ≥7.0 mmol/l, or two hour post 75g oral glucose load plasma glucose ≥11.1 mmol/l, on two or more occasions). Intensive treatment is designed to achieve blood glucose values as close to the non-diabetic range as possible. The essential components of such treatment are education, counselling, monitoring, self management, and pharmacological treatment with insulin or oral antidiabetic agents, to achieve specific glycaemic goals.

    Incidence/prevalence: Diabetes is diagnosed in around 5% of adults aged 20 years or over in the United States.1 A further 2.7% have undiagnosed diabetes on the basis of fasting glucose. The prevalence is similar in men and women, but diabetes is more common in many ethnic groups. The prevalence in people aged 40-74 has increased over the past decade.

    Aetiology: Diabetes results from deficient insulin secretion, decreased insulin action, or both. Many processes can be involved, ranging from autoimmune destruction of the β cells of the pancreas to incompletely understood abnormalities that result in resistance to insulin action. Genetic factors are involved in both mechanisms. In type 1 diabetes there is an absolute deficiency of insulin. In type 2 diabetes, insulin resistance and an inability of the pancreas to compensate are involved. Hyperglycaemia sufficient to cause tissue damage can be present without clinical symptoms for many years before diagnosis.

    Prognosis: Severe hyperglycaemia causes numerous symptoms, including polyuria, polydipsia, weight loss, and blurred vision. Acute, life threatening consequences of diabetes are hyperglycaemia with ketoacidosis or the non-ketotic hyperosmolar syndrome. There is increased susceptibility to certain infections. Long term complications of diabetes include retinopathy (with potential loss of vision), nephropathy (leading to renal failure), peripheral neuropathy (increased risk of foot ulcers, amputation, and Charcot joints), autonomic neuropathy (gastrointestinal, sexual, and bladder dysfunction), and greatly increased risk of atheroma affecting large vessels (macrovascular complications of myocardial infarction, stroke, or peripheral vascular disease). The physical, emotional, and social impact of diabetes and the demands of intensive treatment can also create problems for people with diabetes and their families.

    Interventions for glycaemic control in diabetes

    • Intensive control of hyperglycaemia in people aged 13-70 years old

    Trade off between benefits and harms:
    • Intensive control of hyperglycaemia in people with frequent severe hypoglycaemia

    Aims: To slow the development and progression of the microvascular and neuropathic complications of diabetes while minimising adverse effects of treatment (hypoglycaemia and weight gain) and maximising quality of life.

    Outcomes: Quality of life; short term burden of treatment; long term clinical complications; risks and benefits of treatment. Both the development of complications in people who have previously been free of them, and the progression of complications, are used as outcomes. Scales of severity are used to detect disease progression—for example, 19 step scales of diabetic retinopathy; normoalbuminuria, microalbuminuria, and albuminuria for nephropathy; absence or presence of clinical neuropathy.


    We searched Medline and Embase in July 1998 for systematic reviews assessing the impact of control of blood glucose in people with diabetes which measured clinical outcomes of the type described above. Subsequent relevant randomised controlled trials (RCTs) were also reviewed.

    Effects of intensive versus conventional glycaemic control

    One systematic review and subsequent RCTs in people with type 1 or type 2 diabetes found that intensive treatment compared with conventional treatment reduces the development and progression of microvascular and neuropathic complications. Intensive treatment causes hypoglycaemia and weight gain but does not seem to adversely affect neuropsychological function or quality of life. These trials, and prospective observational studies, have found no evidence that intensive treatment reduces adverse cardiovascular outcomes, but the confidence intervals were large and a clinically important effect might still exist.


    Microvascular and neuropathic complications: One systematic review published in 1993 of 16 small RCTs compared intensive with conventional treatment in people with type 1 diabetes, with follow up ranging from 8 to 60 months.2 Two subsequent longer term RCTs have repeated and extended these findings. (The recent results of the UK prospective diabetes study have not been included in this review.) The diabetes control and complications trial (DCCT) compared intensive with conventional treatment over 6.5 years in 1441 people with type 1 diabetes.3 About half had no retinopathy and half had mild retinopathy. The Kumamoto study compared intensive versus standard treatment in 110 people with insulin treated type 2 diabetes over six years.4 Results of these studies are shown in the table. The relative risks of retinopathy, nephropathy, and neuropathy were all reduced.

    Risk (odds ratio) for development or progression of microvascular and neuropathic complications with intensive versus conventional treatment. Odds ratios (OR), number needed to treat (NNT), and confidence intervals (CI) were all calculated from data in papers

    View this table:

    Cardiovascular outcomes: In the DCCT, intensive treatment reduced the relative risk of any major macrovascular event from 0.8 to 0.5 events per 100 patient years (relative risk reduction 41%, 95% confidence interval 10% to 68%).3 In the Kumamoto study, the number of major cerebrovascular, cardiovascular, and peripheral vascular events in the intensive treatment group was half that of the conventional treatment group (0.6 v 1.3 events per 100 patient years), but the event rates in this small trial were low and the results were not significant.4


    Hypoglycaemia: The systematic review found no consistent change in the risk of severe hypoglycaemic reactions in six RCTs,2 but the subsequent RCTs found increased risk of hypoglycaemia with intensive treatment.3 4 In the DCCT the rates of severe hypoglycaemic episodes (requiring help from someone else) were 0.6 episodes per patient year in the intensive treatment group and 0.2 episodes per patient year in the conventional treatment group.3 On average, one extra severe hypoglycaemic event occurred for every two people treated intensively for one year. In the Kumamoto study the rate of hypoglycaemia was lower and there was no significant difference between groups.4 Over six years, six people in the intensive treatment group and four in the conventional treatment group had one or more mild hypoglycaemic reactions. There were no episodes of coma, seizure, or severe hypoglycaemia that required the help of another person.

    Neuropsychological impairment: The systematic review did not comment specifically on neuropsychological impairment.2 However, one of the trials included in the review, the Stockholm diabetes intervention study (SDIS), assessed cognitive impairment while comparing intensified with standard treatment over 7.5 years in 102 people with type 1 diabetes, non-proliferative retinopathy, normal serum creatinine concentrations, and unsatisfactory blood glucose control.5 After three years it found no cognitive impairment associated with hypoglycaemia.6 In the DCCT, intensive treatment did not affect neuropsychological performance.7 In addition, people who had repeated episodes of hypoglycaemia did not perform differently from people who did not have repeated episodes.

    Weight gain: Three RCTs found more weight increase with intensive treatment than with standard treatment. In the SDIS, weight remained stable in the conventional treatment group but body mass index increased by 5.8% in the intensive treatment group (95% confidence interval not presented, P<0.01).8 In the DCCT, intensive treatment was associated with a 33% increase in the risk of developing a body weight more than 120% above the ideal (12.7 cases per 100 patient years with intensive treatment v 9.3 cases per 100 patient years with conventional treatment). At five years, people treated intensively gained 4.6 kg more than people treated conventionally (confidence interval not presented for weight data).3 In the Kumamoto study, the increase in body mass index from baseline to six years was not significant in either group (intensive treatment group 20.5 to 21.2 kg/m2; conventional treatment group 20.3 to 21.9 kg/m2).4

    Quality of life: Only the DCCT looked at quality of life. It found that people undergoing intensive treatment did not experience deterioration in the quality of their lives, even while the rigour of their diabetes care was increased.9 The occurrence of severe hypoglycaemia was not consistently associated with a subsequent increase in distress due to symptoms or decline in diabetes related quality of life. However, in the primary prevention intensive treatment group, people who had repeated severe hypoglycaemia (three or more events resulting in coma or seizure) tended to be at increased risk of measurable distress due to symptoms.


    None of the trials was designed to assess the impact of treatment on macrovascular disease. Major cardiovascular events were infrequent because the trials included relatively young people and follow up was relatively short. As a result, the 95% confidence interval for the effect on macrovascular events was very wide.

    Optimum target blood glucose

    Large RCTs in people with type 1 and type 2 diabetes have found that the risk of the development or progression of complications increases progressively as glycated haemoglobin increases above the non-diabetic range.


    The DCCT found no glycated haemoglobin threshold below which there was no increased risk of the development or progression of complications.10


    In the intensive treatment group of the DCCT, the absolute risk of severe hypoglycaemia increased as HbA1c decreased.10


    As concentrations of glycated haemoglobin were reduced, there were continuing reductions in the relative risk of complications, but there was a slower rate of increase in the risk of hypoglycaemia.10 The balance between benefits and harms of intensive treatment may be less favourable in children under 13 years or adults over 70 and in people with repeated severe hypoglycaemia or unawareness of hypoglycaemia. The benefit of intensive treatment is limited by complications of advanced diabetes (such as blindness, end stage renal disease, or cardiovascular disease), major comorbidity, and reduced life expectancy. The risk of intensive treatment is increased by a history of severe hypoglycaemia or unawareness of hypoglycaemia, advanced autonomic neuropathy or cardiovascular disease, and impaired ability to detect or treat hypoglycaemia (such as altered mental state, immobility, or lack of social support). For people likely to have limited benefit or increased risk with intensive treatment, less intensive goals for glycaemic management should be negotiated, reflecting the individual's self determined goals of care and willingness to make lifestyle modifications.

    Clinical Evidence is published by BMJ Publishing Group and American College of Physicians-American Society of Internal Medicine. The first issue is available now and will be updated and expanded every six months. Individual subscription rate £45, institutional rate £120(2 issues). For more information including how to subscribe please visit the Clinical Evidence website at


    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    View Abstract

    Sign in

    Log in through your institution