Retrospective analysis of the role of serum vitamin D in early rheumatic disease SIR, The role of vitamin D in rheumatic disease is of increasing interest. Although in vitro work implies an immunomodulatory function, its value as a therapy in autoimmune disease remains unproved [1]. Observational studies in established RA suggest an inverse association between vitamin D and composite DAS-28 [2–4], but data in early arthritis are lacking. We therefore investigated the relationship between serum vitamin D levels and clinical phenotype in patients with recent onset of joint pain. This work conforms to standards currently applied in the country of origin and was performed with the eth-ical approval of the Newcastle upon Tyne NHS Foundation Trust; all the patients provided written in-formed consent. Consecutive, consenting patients with a range of rheumatological diagnoses, nave to immunomodulatory drugs, were recruited from an early arthritis clinic in Newcastle upon Tyne, UK, between October 2007 and March 2009. Serum 25-hydroxy vitamin D [25(OH)D] levels (DiaSorin, Stillwater, MN automated immunoassay), CRP level and ESR were measured at inception, along with symptom duration, tender joint count, swollen joint count and patient-reported general global health score (GH-VAS). Statistical analyses were performed using IBM SPSS, Statistics 21, employing a = 5% as the signifi-cance threshold throughout. Diagnostic categories (validated at the 1-year consultant rheumatologist follow-up) for 344 patients in the study cohort comprised RA, undifferentiated arthritis, other inflammatory arthritis (psoriatic, enteropathic), reactive arthritis, crystal arthritis, OA, and non-inflammatory arthral-gia. Demographics and 25(OH)D measurements for these groups are summarized in Table 1. Significant differences in age, sex and symptom duration, but not 25(OH)D levels, were seen between diag-nostic groups (Kruskal–Wallis and 2 test). There was no significant association between serum 25(OH)D and any of the clinical or laboratory parameters measured in the cohort as a whole, after correction for age, sex and symp-tom duration (linear regression). Within individual diagnos-tic categories, however, a clear inverse association between GH-VAS and 25(OH)D was observed in early OA patients (P < 0.001 after correction for age, sex and symp-tom duration), which was not seen in early RA patients or any other diagnostic subgroup. This is the first study to attempt to dissect the effects of 25(OH)D in early arthritis. Our observation of an inverse association between vitamin D status and perceived health only in OA patients is intriguing; it contrasts with published findings in established RA, where low vitamin D levels predicted high disease activity [2, 3]. Of the four components of DAS28, GH-VAS appears to be the most sensitive to vitamin D status in established RA [5] and may disproportionately influence the reported associations with DAS28. Degenerative joint disease is a frequent con-sequence of advancing RA and thus could affect TABLE 1 Demographic data of the diagnostic categories