This site is intended for health professionals only

Rimonabant and type 2 diabetes: results of the RIO study

Rimonabant is a cannabinoid receptor antagonist that has been licensed as an antiobesity drug. While more long-term studies are needed before definitive recommendations can be made for its use, it has been thought that it might have  a role in obese or overweight people with type 2 diabetes. The RIO-Diabetes study aimed to assess just that.
Peter Burrill
BPharm(Hons),  MRPharmS DipPresSci FCPP FFPMM

Specialist Pharmaceutical Adviser for Public Health
Derbyshire County PCT

Peter is a pharmaceutical adviser of many years standing and provides strategic input into the prescribing and clinical effectiveness agenda for the Derbyshire Public Health Network. He is involved with the appraisal of evidence, evaluation of new drugs and guidelines development, and facilitates the Derbyshire Clinical Effectiveness, Prescribing and Prioritisation Committee. Peter also trains health professionals in the principles of critical appraisal and evidence-based medicine.

This article is one of many that Peter will be contributing to NiP over the coming issues. In each Peter will summarise the results of a recent drug study relevant to primary care.

Rimonabant is an oral selective cannabinoid CB1 receptor antagonist that is licensed as an adjunct to diet and exercise for the treatment of obese patients, or overweight patients with associated risk factor(s), such as type 2 diabetes or dyslipidaemia. A recent Cochrane review found that compared with placebo, rimonabant 20 mg produced a 4.9 kg greater reduction in bodyweight in trials with one-year results.(1) However, rimonabant caused significantly more adverse effects and attrition rates were approximately 40% at the end of one year.
The Cochrane review concluded that studies with longer follow-ups after the end of treatment and of more rigorous quality should be conducted before definitive recommendations can be made regarding the role of this new medication in the management of overweight or obese patients.
A possible priority target group for the use of rimonabant are people with type 2 diabetes. The RIO-Diabetes study was published in 2006.(2) The aim of this study was to assess the efficacy and safety of rimonabant in overweight or obese patients with type 2 diabetes who were inadequately controlled by metformin or sulphonylureas.

Patients aged 18-70 years with type 2 diabetes who had been treated with metformin or sulphonylurea monotherapy for at least six months (stable dose for at least three months), but who remained inadequately controlled, were recruited.

  • nclusion criteria were BMI of 27-40 kg/m2, HbA1c level of 6.5-10%, and a fasting glucose concentration of 5.55-15.04 mmol/l.
  • Exclusion criteria included unstable bodyweight (defined as more than 5 kg variation within the past three months), any clinically significant disorder (including severe microvascular or macrovascular complications of diabetes), systolic BP greater than 160 mmHg or diastolic BP greater than 95 mmHg, and use of antiobesity drugs within the past three months or use of any drugs known to affect bodyweight.
  • A two-week screening period preceded a four-week placebo run-in period, followed by randomisation to double-blind treatment. Patients were randomised to placebo, 5 mg/day rimonabant, or 20 mg/day rimonabant.
  • All patients were put on a mild hypocalorific diet and were advised on increased physical activity during the run-in period and until the end of the study.
  •  The primary endpoint was weight change from baseline at the last observation carried forward (LOCF). Patients were classified as having a response of a 5% or 10% weight loss if they had a reduction in bodyweight from baseline at the LOCF of at least 5% or 10%.
  • Secondary endpoints included changes in HbA1c, high-density lipoprotein-cholesterol (HDL-C), triglyceride, fasting glucose, fasting insulin, waist circumference and blood pressure.
  • Analyses were done on a modified intention-to-treat basis.
  • RIO-Diabetes was sponsored by Sanofi Synthelabo Research.


  • A total of 513 men and 532 women (mean age 55.5 years) were randomised to double-blind treatment but only 66.2% completed the one-year follow-up.
  • At baseline, mean weight was 96.3 kg, mean waist circumference 109 cm, mean HbA1c 7.3%, and mean total cholesterol 5 mmol/l. Metformin was the antidiabetic treatment in 65% and sulphonylurea in 35%.
  • I am only reporting the 20 mg/day results as this is the licensed dose for rimonabant. The placebo-corrected weight loss overall after one year was 3.9 kg (p 
  • 49.4% of those allocated rimonabant 20 mg/day achieved ≥ 5% weight loss compared with 14.5% of those on placebo (p 
  • 16.4% of those allocated rimonabant 20 mg/day achieved ≥ 10% weight loss compared with 2.0% of those on placebo (p 
  • Waist circumference was significantly lower with rimonabant 20 mg/day compared with placebo, -5.2 cm vs -1.9 cm (p 
  • HbA1c levels were lower with rimonabant 20 mg/day compared with placebo, -0.6% vs +0.1% (p 
  • There was no difference in total cholesterol (TC) or low-density lipoprotein-cholesterol (LDL-C) but there was a small change in the TC/HDL-C ratio, -0.51 vs -0.16 (p 
  • Patients in the rimonabant 20 mg/day group reported having less appetite (p 
  • Discontinuations due to adverse events were more frequent in the rimonabant 20 mg/day group, 15% vs 5% (no p-value quoted). This is a NNH (number needed to harm) of 10.
  • The most common adverse events that led to premature study discontinuation in the rimonabant 20 mg/day group were depressed mood disorders, nausea and dizziness. Depressed mood disorders leading to study discontinuation occurred in 3% of those on rimonabant 20 mg/day and 0.9% of those on placebo (no p-value quoted). This is a NNH of 48.

The authors conclude that rimonabant 20 mg/day for one year significantly reduced weight, waist circumference, and HbA1c levels and improved a number of cardiovascular and metabolic risk factors in overweight or obese patients with type 2 diabetes inadequately controlled by metformin or sulphonylureas. It would have been interesting to see how rimonabant would have compared with trying to achieve control by increasing the metformin dose or adding metformin to the sulphonylurea.
The authors argue that about half the effect of rimonabant 20 mg/day on HbA1c levels is independent of weight loss and consistent with the direct peripheral metabolic effects of the drug. However, the accompanying editorial suggests that this claim merits closer scrutiny.(3)
A weight loss of 10% is usually regarded as the minimum required to show a clinically meaningful change and this was achieved by one in seven of those treated with rimonabant 20 mg/day compared with placebo. However, this study was only of one-year duration and long-term studies are needed to assess the effect on diabetes-related complications, especially cardiovascular outcomes.
The retention rate of about 66% in this study might be considered as rather low.
The suggestion that rimonabant increases depression is of concern. The editorial points out that the study recruits had relatively low depression scores, perhaps somewhat lower than scores generally characteristic of a population with type 2 diabetes.
Rimonabant is more expensive than orlistat and sibutramine so cost-benefit analyses are required to aid decision-making.
Orlistat and sibutramine are covered by NICE guidance and there are no head-to-head studies of rimonabant against these drugs. It might be sensible to consider rimonabant as a thirdline drug until NICE pronounces on its place in therapy.
The Cochrane review comments that efforts focusing on the prevention of obesity in nonobese people and nonpharmacological management in obese people should remain the cornerstone of obesity therapy.(1)
Rimonabant has recently been reviewed by the Drug and Therapeutics Bulletin, which concluded "On current  evidence, we do not believe that rimonabant represents a significant advance for patients with obesity."(4) As  part of its continuous monitoring of medicines, the European Medicines Agency (EMEA) has reviewed the available information on the safety of rimonabant,  particularly on the medicine's psychiatric safety.(5) It has concluded that rimonabant's benefits continue to outweigh its risks, but that patients with ongoing major  depression or who are receiving antidepressant treatment should not be  prescribed rimonabant and that some of the warnings in the medicine's prescribing information should be  strengthened.


  1. Rimonabant for overweight or obesity. Cochrane Database Systematic Rev 2006;Issue 4.
  2. Scheen AJ, Finer N, Hollander P, Jensen MD, Van Gaal LF; RIO-Diabetes Study Group. Efficacy and tolerability of rimonabant in overweight or obese patients with type 2 diabetes: a randomised controlled study (RIO-Diabetes). Lancet 2006;368:1660-72.
  3. Cleland SJ, Sattar N. Does rimonabant pull its weight for type 2 diabetes? [Editorial] Lancet 2006;368:1632-4.
  4. Less weight or more hype with rimonabant? Drug Ther Bull 2007;45:41-3.
  5. EMEA Press Release, London, 19 July 2007. Available from: