Bone erosions are frequent in patients with psoriatic arthritis (PsA) but the prevalence of generalised bone loss in PsA is not well understood.1,–,6 To formally address this question, we analysed data from the Consortium of Rheumatology Researchers of North America registry.

Erosions on plain x-rays of hands and feet and T-scores at the lumbar spine, femoral neck and total hip were identified as the key outcome variables. Key independent variables included gender, age, weight, C reactive protein, functional status, Health Assessment Questionnaire, body mass index and medication use. Medications included methotrexate, other disease-modifying antirheumatic drugs (DMARDs), anti-tumour necrosis factor (anti-TNF) agents, prednisone use (yes or no) and prednisone dose (low dose <10 mg or high dose ≥10 mg). Multiple logistic regression models were constructed to assess the association between the two dependent variables—presence/absence of erosions and T-scores in patients with PsA. The data were adjusted for each of the independent variables. The analysis was restricted to patients with only complete records in all covariates.

A total of 2212 patients with PsA were included in the analyses, 889 without erosions (group 1) and 567 with erosions (group 2). Table 1 shows the differences between patients in groups 1 and 2 for demographics, clinical features and the independent risk factors for osteoporosis. Patients with PsA in group 2 were significantly more likely to be male (p=0.03) and also were younger than those in group 1 (p<0.01). The two groups showed no differences for the other independent variables. Patients in group 2 had a longer disease duration than patients in group 1 (11.9 vs 7.5 years; p<0.01). Corticosteroid use and dose were similar between the two groups. Patients in group 2 were also more likely to be on osteoporosis medications (p<0.01). Use of DMARDs and anti-TNF therapy was similar in the two groups.

The mean T-score for the lumbar spine for all patients with PsA was −0.33, but the mean T-score was significantly lower in patients in group 2 than in those in group 1 (−0.61 vs −0.11; p<0.01). A significant difference was noted between the two groups when logistic regression models were constructed to adjust for all covariates with steroid use (p<0.01) and steroid dose (p<0.01). The bone mineral density (BMD) measurements and other variables are shown in table 2.

The mean T-scores for the femoral neck and total hip for all patients with PsA were −0.98 and −0.42. No significant difference was noted in BMD for the femoral neck and total hip scores in the two cohorts of patients, even when the data were adjusted for all variables including steroid use and steroid dose.

In conclusion, we found that PsA patients with erosions had a significantly lower lumbar spine BMD than those without erosions even when adjusted for use of steroids, DMARDs, anti-TNF agents and osteoporosis medications. We propose that increased inflammatory cytokines in the circulation and in the joint combined with a marked upregulation of osteoclastic activity, common to both forms of bone loss, may explain this relationship.7,–,9