NPC Archive Item: CONTROL blood glucose in type 2 diabetes — but individualise care

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

6th April 2011

The CONTROL meta-analysis of four large RCTs found that more-intensive blood glucose lowering reduced the absolute risk of major cardiovascular (CV) events by 0.9% over 4.4 years (mainly due to a reduction in myocardial infarction [MI]), but did not reduce all-cause mortality, CV mortality or stroke. However, more-intensive control was associated with an absolute risk increase of major hypoglycaemia of 4.6% over 4.4 years.  

Level of evidence:
Level 1 (good quality patient-oriented evidence) according to the SORT criteria.

There is no argument for poor blood glucose control in people with type 2 diabetes. However, the balance of risks and benefits of more-intensive blood glucose control need to be carefully considered on an individual patient basis. NICE guidance on type 2 diabetes should be followed and individual targets for HbA1c should be agreed with each individual patient, taking into account the patient’s own preferences for care and the balance of likely benefits and harms. On a population basis, there may be greater benefits from a holistic approach, deploying maximally, lifestyle interventions (stopping smoking, losing weight, and taking more exercise as appropriate), controlling blood pressure, taking a statin, and taking metformin, than from concentrating more on intensive versus good blood glucose control.

What is the background to this?
Four key RCTs have assessed the benefits and harms of conventional blood glucose control, compared with intensive blood glucose control (UKPDS, ACCORD, ADVANCE and VADT). There have also been a number of meta-analyses, including Ray KK, et al and Kelly TN, et al, which have been discussed in previous MeReC Rapid Reviews. There has been some criticism, particularly relating to the selection of studies in previous meta-analyses.1,2 This meta-analysis (CONTROL) resulted from a collaboration of the UKPDS, ACCORD, ADVANCE and VADT investigators with the aim of providing a more accurate estimate of the effects of more-intensive blood glucose lowering on major CV events.

What does this study claim?
The CONTROL meta-analysis of the four key RCTs (n=27,049) in people with type 2 diabetes found that more-intensive blood glucose control reduced the relative risk of major CV events (CV death, non-fatal stroke, or non-fatal MI) by 9% (hazard ratio [HR] 0.91, 95% confidence interval [CI] 0.84 to 0.99, absolute risk reduction [ARR] 0.9%, over 4.4 years), compared with less-intensive control. This was mainly due to a reduced risk of MI (HR 0.85, 95%CI 0.76 to 0.94). However, there was no significant difference in all-cause mortality (HR 1.04, 95%CI 0.90 to 1.20), stroke (fatal or non-fatal: HR 0.96, 95%CI 0.83 to 1.10) or cardiovascular mortality (HR 1.10, 95%CI 0.84 to 1.42). There were significantly more major hypoglycaemic events in the more-intensive group (HR 2.48, 95%CI 1.91 to 3.21, absolute risk increase [ARI] 4.6%, over 4.4 years. The authors conclude that more-intensive blood glucose control (mean HbA1c 0.88 percentage points reduction) affords a modest but significant CV benefit in the short to medium term, but all-cause and CV mortality are not benefited.

How does this relate to other studies?
Both the Ray KK, et al and Kelly TN, et al meta-analyses found similar findings to the CONTROL meta-analysis — some CV benefit, no reduction in all-cause mortality or stroke, and an increase in severe hypoglycaemia. Furthermore, in the ACCORD RCT (n=10,251) there was an increased risk of mortality in the more-intensive group (5.01% vs. 3.96%, HR 1.22, 95%CI 1.01 to 1.46, number needed to harm [NNH] 95 over an average of four years).

So what?
Poor glucose control is associated with an increased risk of microvascular complications and mortality. Therefore, good control of blood glucose is an important priority, but pursuing intensive control at the expense of other priorities (e.g. lifestyle, smoking cessation, managing BP, managing lipids) would seem inappropriate. However, individualisation of care is needed and clinicians must also take account of physical, psychological and social needs, including the patient’s own preferences for care.

Although intensifying blood glucose control does have some CV benefit, it appears to be less effective at reducing CV disease than controlling either BP or cholesterol. Using data from the CONTROL meta-analysis, John Yudkin and colleagues estimated the effect of blood glucose lowering on improvements in CV outcomes. When compared with the estimated effects of BP lowering and cholesterol lowering, the number needed to treat (NNT) to prevent one CV event over 5 years was 34 for BP lowering, 44 for cholesterol lowering, and 119 for blood glucose lowering.

Furthermore, as observed in all the meta-analyses, more-intensive blood glucose control does not reduce mortality, and reducing HbA1c levels too low may actually increase mortality (particularly when treatment is intensified with insulin). A large observational cohort study by Currie CJ, et al (see MeReC Rapid Review 1017) identified a ‘U-shaped’ relationship between HbA1c levels and mortality in type 2 diabetes. It found that an HbA1c of about 7.5% (59mmol/mol) was associated with the lowest risk of all-cause mortality, with higher or lower levels associated with greater risk.

The CONTROL meta-analysis did not consider any potential microvascular benefits of more-intensive blood glucose control. The evidence is conflicting here as some trials have shown a reduction in some microvascular events, whereas other studies have not shown significant benefits, particularly with regard to advanced renal or eye complications. See MeReC Rapid Review 1956 for details. Yudkin and colleagues have also illustrated the effects of more-intensive blood glucose control on advanced microvascular outcomes.

In addition to any potential benefits, clinicians must also carefully consider the risks of more-intensive blood glucose control. This meta-analysis highlighted an increased risk of severe hypoglycaemia. Lower HbA1c targets may be relatively easy to achieve in a newly diagnosed patient, but HbA1c levels increase over time as the disease progresses. Pursuing more-intensive control may not confer additional benefit, it adds to the risks of adverse effects (e.g. hypoglycaemia, weight gain), increases costs, and requires careful consideration, especially in older patients with long-standing disease. Clinicians need to consider the law of cumulative benefits (or diminishing returns) and discuss with patients what additional improvement in outcomes might be gained from intensifying blood glucose treatment, and how this may affect their quality of life.


  1. Yudkin JS, Richter B. Letter: Concerns about meta-analysis of glucose control and cardiovascular disease in type 2 diabetes. Ann Intern Med 2010;152:63–4
  2. Lipska K, et al. Letter: Intensive glucose control and cardiovascular outcomes. Lancet 2009;374:522–3

Study details
Turnbull FM, et al. Intensive glucose control and macrovascular outcomes in type 2 diabetes. Diabetologia 2009;52:2288–98

Prospective group-level meta-analysis of large RCTs (>1000 person-years follow-up and minimum two years median post-randomisation follow-up) designed to assess directly the impact of lower vs. higher levels of glycaemia on CV outcomes.

27,049 adults (>16 years; mean age 62 years) with type 2 diabetes.

Intervention and comparison
More-intensive blood glucose control, compared with less-intensive blood glucose lowering (mean HbA1c difference 0.88 percentage points).

Primary outcome was a composite of major CV events — defined as death from CV causes (including sudden death), non-fatal MI and non-fatal stroke. Secondary outcomes were all-cause mortality, CV death, stroke (non-fatal and fatal), MI (non-fatal and fatal), heart failure resulting in hospitalisation or death, non-CV death, and major hypoglycaemia.


Outcome HR (95%CI) More-intensive vs. less-intensive blood glucose lowering
Major CV events 0.91 (0.84 to 0.99)
All-cause mortality 1.04 (0.90 to 1.20)
CV death 1.10 (0.84 to 1.42)
Stroke 0.96 (0.83 to 1.10)
MI 0.85 (0.76 to 0.94)
Heart failure resulting in hospitalisation or death 1.00 (0.86 to 1.16)
Non-CV death 1.02 (0.89 to 1.18)
Major hypoglycaemia 2.48 (1.91 to 3.21)

Sponsorship: Study funded by a Research Fellowship from the National Heart Foundation of Australia.

More information can be found in the NPC e-learning materials on type 2 diabetes.

Please comment on this blog in the NPC discussion rooms, or using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you