Article Text

AB0279 Incidence of diabetes and effect of etanercept and adalimumab on hba1c over 1 year: data from a randomised trial in patients with rheumatoid arthritis
  1. P. De Pablo1,2,
  2. F. Maggs2,
  3. D. Carruthers3,
  4. A. Faizal4,
  5. M. Pugh5,
  6. P. Jobanputra2
  1. 1Rheumatology, College of Medical & Dental Sciences, University of Birmingham
  2. 2Rheumatology, Queen Elizabeth Hospital, UHB NHS Trust
  3. 3Rheumatology, City Hospital, SWBH NHS Trust, Birmingham
  4. 4Rheumatology, Solihull Hospital, Heart of England NHS Foundation Trust, Solihull
  5. 5Rheumatology, St Mary’s Hospital, Newport, Isle of Wight, United Kingdom


Background Inflammation such as that which occurs in rheumatoid arthritis (RA) is associated with insulin resistance and risk of diabetes mellitus (DM). Some DMARDs including TNF inhibitors (TNFi) may improve insulin resistance and DM risk. However, it is unknown whether TNFi improve HbA1c in patients with RA with or without DM. Data on HbA1c was collected in a randomised trial comparing drug continuation rates for etanercept and adalimumab (1).

Objectives To estimate the incidence of DM from this data and study the impact of therapy on HbA1c.

Methods Participants with active RA, who had previously failed to respond to 2 non-biologic DMARDs including methotrexate, were randomised to etanercept or adalimumab and followed 3-monthly over 1 year. Data collected included co-morbidities, clinical and laboratory parameters, and medications at each visit. The primary endpoint was newly recorded diabetes defined as HbA1c ≥48mmol/mol at any time point. Predictors of HbA1c and HbA1c change were determined with univariate and multivariate analyses.

Results Of the 125 patients with active RA randomised to etanercept or adalimumab, 6 (4.8%) were known diabetics and 88% were RF/ACPA positive. Of the 119 without DM, 7 (5.9%) were diagnosed with DM (HbA1c ≥48mmol/mol) at baseline. 73 participants (73% female) completed 1 year of TNFi therapy: mean age 54 yrs (SD±12), mean BMI 27.8, mean HbA1c 38 mmol/mol. Of these, 4 (5%) patients had DM at baseline. A majority of patients were on methotrexate (67%) and 33% on prednisolone. Baseline characteristics were similar for patients’ allocated to adalimumab (52%) or etanercept (48%), except more patients on etanercept were on prednisolone (49% vs. 18%; p=0.006); and more patients on adalimumab were on hydroxycloroquine (24% vs. 3%; p=0.01).

After excluding those with known DM at baseline, among those completing one year of TNFi (n=69), 3 (4.4%) patients had an HbA1c ≥48 mmol/mol at baseline,1 (1.5%) at 3 months,1 (1.5%) at 6 months, and 2 (2.9%) at 12 months of follow-up. 2 (3%)cases had aHbA1c ≥48 mmol/mol at 2 follow-up visits. The incidence of DM was 29 new cases per 1000-person years (95% CI 3.51-105).

Those on adalimumab tended to have higher levels of HbA1c than those on etanercept but the differences between groups at each time point were non-significant. However, there was a significant rise in HbA1c levels after 1 year of adalimumab therapy (37.27 mmol/mol and 38.80 mmol/mol; p=0.01). Etanercept therapy did not influence HbA1c levels over time.

Conclusions Incidence of diabetes in patients entering a randomised trial of etanercept and adalimumab was considerably higher than other recent data. Treatment with a TNFi did not improve HbA1c levels with either agent in diabetics and non-diabetics. After excluding those with diabetes, those on adalimumab had higher mean HbA1c levels after 1 year of therapy.

  1. BMJ Open 2012;2:e001395.

Disclosure of Interest None Declared

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.