Skip to content
View commentaries on this research

Please note that this summary was posted more than 5 years ago. More recent research findings may have been published.

This is a plain English summary of an original research article. The views expressed are those of the author(s) and reviewer(s) at the time of publication.

Metformin worked best at keeping blood sugar levels under control and remains a good first choice as single therapy. Overall, the nine classes of blood sugar-lowering drugs had similar effect on risk of death from cardiovascular causes and overall mortality, though estimates are imprecise because so few people died in these studies. Risk of other side effects, such as weight gain or risk of low blood sugar did vary between treatments.

This very large systematic review of 301 trials compared the drugs given alone or in combination to treat type 2 diabetes against each other. The researchers analysed data from direct comparisons and from studies with dummy treatment arms together. Until now it has been unclear if there are reasons to use one treatment regimen over another based on different rates of heart disease or death. These findings suggest that the drugs are so similar in this respect that the research could not tell them apart.

Metformin remains the drug of first-choice when starting to treat type 2 diabetes. Further recommended blood sugar lowering treatments can be tailored to the control of sugar levels and individual patients’ priorities, such as avoiding weight gain or very low blood sugar.

Why was this study needed?

Diabetes care accounts for up to 10% of the NHS budget. In 2010, the direct costs associated with type 2 diabetes were £13.8 billion, which cover medical costs of treatment of disease and associated complications. Additional costs associated with days of work missed, early retirement and social benefits amounted to £15.4 billion. About 6% of adults had diabetes in 2013, mostly type 2 diabetes.

There have been many reviews assessing how well different treatment combinations work to lower the blood sugar of people with type 2 diabetes. However, it’s been uncertain whether the range of treatments has different effectiveness and safety when it comes to preventing deaths from cardiovascular disease.

Such evidence could help clinicians choose the best drug or drug combination for each individual patient to extend life expectancy and reduce the risk of complications.

What did this study do?

This systematic review identified 301 randomised controlled trials comparing blood sugar lowering drugs in 56,598 adults with type 2 diabetes.

The main outcome was risk of death from cardiovascular disease. Other outcomes included death from any cause, long-term blood sugar control (measured by glycated haemoglobin, HbA1c), treatment failure, low blood sugar and body weight.

Researchers performed a network meta-analysis, comparing treatments directly in head-to-head trials, and comparing them indirectly across placebo controlled trials. They used statistical techniques to take account of study differences, including different patient characteristics.

Common sources of bias across trials included method of treatment allocation, assessors being aware of treatment given, incomplete outcomes reported, and drug sponsors having potential role in authoring. There were relatively few deaths from cardiovascular causes. For example, 65 cardiovascular deaths in the 25 trials of monotherapy and this reduces the certainty with which the researchers can conclude there is no difference between drugs. Although the risk of bias was high or unclear in many trials, these would mostly have acted to increase the mortality difference between the drug classes and so would have had little effect on the main outcome in the study.

What did it find?

  • There was no difference in risk of death from cardiovascular causes between any blood sugar-lowering drugs, given alone or in dual or triple combinations, when compared with each other or with placebo/no treatment. However, estimates were imprecise with wide confidence intervals, because deaths were rare. For example, metformin against placebo had an Odds Ratio 1.38 (95% confidence interval [CI] 0.41 to 4.72). There was also a lack of data assessing the combination of metformin with basal insulin or α-glucosidase inhibitors.
  • When using a single drug, metformin showed the best control of long term blood sugar levels. HbA1c was higher with a sulfonylurea (standardised mean difference [SMD] 0.18, 95% confidence interval [CI] 0.01 to 0.34), thiazolidinedione (SMD 0.16, 95% CI 0.00 to 0.31), DPP-4 inhibitor (SMD 0.33, 95% CI 0.13 to 0.52) and α-glucosidase inhibitor (SMD 0.35, 95% CI 0.12 to 0.58). There was no difference in HbA1c between metformin and single drug treatment with SGLT-2 inhibitor, meglitinides, GLP-1 receptor agonists or basal insulin. All drugs showed similar ability to reduce blood sugar when added to metformin.
  • Risk of low blood sugar was highest with a sulfonylurea (odds ratio [OR] 3.13, 95% CI 2.39 to 4.12) or basal insulin (OR 17.9, 95% CI 1.97 to 162). The combination of a sulfonylurea and metformin carried highest risk, while metformin with an SGLT-2 inhibitor was ranked best to avoid low blood sugar.
  • As single therapy, risk of weight gain was higher with a sulfonylurea or thiazolidinedione than with metformin, but metformin carried higher risk than GLP-1 receptor agonists. Similarly the combination of a sulfonylurea and metformin carried highest risk of weight gain, with risk lower for other drug combinations.

What does current guidance say on this issue?

The NICE guideline for treating type 2 diabetes in adults, last updated in July 2016, recommends metformin (standard release) as initial drug treatment. If metformin is not appropriate, initial treatment should be with a dipeptidyl peptidase-4 (DPP-4) inhibitor, a sulfonylurea or pioglitazone (a thiazolidinedione). Individual needs and preferences should be taken into account. If single drug treatment is ineffective, the first intensification step is dual therapy with a combination of two non-insulin drugs. The next step is either triple non-insulin drug combination, or any combination containing insulin.

NICE note a lack of evidence on the effects of treatment, particularly newer drugs, on long-term cardiovascular outcomes.

What are the implications?

This study broadly supports current recommended practice to start type 2 diabetes treatment with metformin where possible. It also provides support for basing choice of further treatment on individual patient goals, for example avoiding sulphonylurea drugs or insulin if the person is at risk from low blood sugar episodes, or choosing GLP-1 receptor agonists if a person needs to reduce their weight.

The evidence is not comprehensive for all drugs or combinations and may not be able to rule out the possibility that different drug classes or individual drugs carry small differences in cardiovascular risk. The risk of bias in these studies reduces our confidence in the size of any benefit of second line drugs compared to metformin. This means that monitoring an individual’s blood sugar responses to a change in therapy remains important.


Citation and Funding

Palmer SC, Mavridis D, Nicolucci A, et al. Comparison of clinical outcomes and adverse events associated with glucose-lowering drugs in patients with type 2 diabetes: A meta-analysis. JAMA. 2016;316(3):313-24.

This study was funded by the Royal Society of New Zealand (grant RDF-UOO1302).



NHS Choices. Type 2 diabetes - Treatment. London: Department of Health; 2016.

NICE. Type 2 diabetes in adults: management (NG28). London: National Institute for Health and Care Excellence; 2015.

Kanavos P, van den Aardweg S, Schurer W. Diabetes expenditure, burden of disease and management in 5 EU countries. London: London School of Economics and Political Science; 2012.

Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre

NIHR Evidence is covered by the creative commons, CC-BY licence. Written content and infographics may be freely reproduced provided that suitable acknowledgement is made. Note, this license excludes comments and images made by third parties, audiovisual content, and linked content on other websites.

  • Share via:
  • Print article

Type 2 diabetes is a chronic condition in which the body cannot control the levels of sugar, or glucose, in the blood. In the long term, raised blood sugar can have widespread damaging effects on the body, affecting both large blood vessels, so causing heart disease and strokes, and affecting smaller blood vessels in the eyes, kidneys and nerves. Keeping blood sugar to near normal levels through healthy diet and medication helps protect against these complications. HbA1c shows how well blood sugar levels were controlled in the previous two to three months.

The different types of blood glucose lowering drugs in this study were:

  • Basal insulin (long-acting insulin detemir or insulin glargine)
  • Metformin (a biguanide)
  • Sulphonylureas (e.g. glibenclamide, glipizide)
  • Thiazolidinedione (pioglitazone is the only authorised drug)
  • DPP-4 inhibitor (gliptins, e.g. sitagliptin)
  • SGLT-2 inhibitor (e.g. dapaglifozin)
  • GLP-1 receptor agonist (incretin mimetics, e.g. exenatide)
  • α-glucosidase inhibitor (acarbose)
  • Meglitinides (prandial glucose regulators, or glinides, e.g. repaglinide or nateglinide)
Back to top