Article Text

Download PDFPDF

Vitamin D deficiency is common in patients with RA and linked to disease activity, but circulating levels are unaffected by TNFα blockade: results from a prospective cohort study
  1. Paul Welsh1,
  2. Mike J L Peters2,
  3. Iain B McInnes1,
  4. Willem F Lems2,
  5. Paul T Lips2,
  6. Gayle McKellar3,
  7. Susan Knox4,
  8. A Michael Wallace4,,
  9. Ben A C Dijkmans2,5,
  10. Michael T Nurmohamed2,5,
  11. Naveed Sattar1
  1. 1BHF GCRC, University of Glasgow, Glasgow, UK
  2. 2VU University Medical Centre, Amsterdam, The Netherlands
  3. 3Centre for Rheumatic Diseases, Glasgow Royal Infirmary, Glasgow, UK
  4. 4Department of Clinical Biochemistry, Glasgow Royal Infirmary, Glasgow, UK
  5. 5Jan van Breemen Institute, Amsterdam, The Netherlands
  1. Correspondence to Dr Paul Welsh, British Heart Foundation Glasgow Cardiovascular Research Centre, 126 University Place, Faculty of Medicine, University of Glasgow, Glasgow G12 8TA, UK; p.welsh{at}

Statistics from

Preliminary data suggest an inverse relationship between serum 25-hydroxyvitamin D (25OHD) levels and disease activity score (DAS) in rheumatoid arthritis (RA).1 Furthermore, data for an apparent beneficial effect of statins on circulating 25OHD concentrations2 have been speculated to be attributable to statin ‘pleiotropic’ anti-inflammatory effects. Serum 25OHD concentrations also fall in the acute phase following surgery.3 Collectively, such data suggest that systemic inflammation may negatively modulate serum 25OHD concentrations. On this basis, we hypothesised that serum 25OHD levels in patients with RA would rise after short-term administration of a tumour necrosis factor α (TNFα) blocking agent, adalimumab.

Our work involved 170 consecutive patients with RA at the rheumatology outpatient clinic of the Jan van Breemen Institute, Amsterdam (latitude 52.3°N). Patients were either treated with adalimumab alone (40 mg every 2 weeks) or with adalimumab and concomitant drug-modifying antirheumatic drugs.4 Other antirheumatic treatment remained unchanged during the study. All patients were seen by a research physician and completed a medical questionnaire.5 Serum samples were collected at baseline before the first injection of adalimumab and after 16 weeks of treatment. Serum levels of 25OHD were determined (blinded) by an LC-MS/MS method which is currently in routine National Health Service use and calibrated against the National Institute of Standards and Technology standard SRM 972.6

Median baseline circulating levels of 25OHD were low (18.5 ng/ml (IQR 11–25)) and 25OHD insufficiency (15–25 ng/ml) and deficiency (<15 ng/ml) were common (37% and 39%, respectively). Pooled serum samples from both timepoints showed expected seasonal variation (figure 1).

Figure 1

Seasonal variation in vitamin D levels in patients with rheumatoid arthritis at baseline. Data are based on definitions of winter (December–February), spring (March–May), summer (June–August) and autumn (September–November) and are shown as medians (horizontal lines), IQR (boxes) and 95% CI (vertical lines). *p<0.05.

The clinical characteristics of the study population stratified for 25OHD status were broadly similar across the groups (table 1), with the exception of season sampling time and disease activity features. At multivariate analyses (comprising all variables with p for trend <0.20) adjusted for seasonal differences, only erythrocyte sedimentation rate (ESR; p=0.003) and body mass index (p=0.037) remained inversely associated with 25OHD.

Table 1

Clinical characteristics of patients with RA stratified for 25OHD status

Treatment with adalimumab led to an improved DAS28 score (5.1±1.2 vs 3.2±1.4, p<0.001). However, median circulating levels of 25OHD did not change after adalimumab treatment (19.0 ng/ml (IQR 12–25) at 16 weeks, p=0.67). The prevalence of 25OHD deficient and insufficient patients also did not change after treatment. Neither the change in DAS28 nor the change in ESR over the 16 weeks was associated with a change in 25OHD levels (p=0.534 and p=0.317, respectively).

Our data suggest that, despite an inverse baseline association of ESR with 25OHD levels, TNFα blockade therapy, at least over a period of 4 months, does not directly influence circulating 25OHD levels. Whether longer term biological therapy has any beneficial effect on circulating 25OHD concentrations requires further study, although any such effect may be attributable to increased sunlight exposure rather than decreasing inflammation. Our observations also weaken the possibility that TNFα blockers, which improve bone mineral density7 and potentially lower cardiovascular risk,8 do so via changes in 25OHD levels. Nevertheless, our data confirm a high prevalence of 25OHD deficiency among patients with RA (39%). In broad agreement with our findings, the Longitudinal Aging Study Amsterdam showed a serum level of <10 ng/ml in 8% of men and 14% of women aged 55–85 years.9

The strengths of the study are its prospective study design, large sample size and robust blinded 25OHD measurements. Observations of expected associations of 25OHD with DAS, season and body mass index lend external validity to the study. A placebo control group would have been useful but ethically difficult.

In conclusion, despite an inverse association between baseline measures of inflammatory status in RA and circulating 25OHD levels, short-term TNFα blockade treatment does not improve 25OHD levels. Further research is needed to address determinants of poor 25OHD status in RA.


In fond memory of Professor Mike Wallace.


View Abstract


  • PW and MJLP are joint first authors

  • MTN and NS are joint senior authors

  • Deceased

  • Funding This study was facilitated by the Clinical Research Bureau of the Jan van Breemen Institute which receives financial support from the Dutch Arthritis Association. MJP received a EULAR bursary and this research was conducted while he was an ARTICULUM Fellow.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the local ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.