• arthritis
• low back pain
• spondyloarthritis

We read with interest the article on ‘Low incidence of vertebral fractures in early spondyloarthritis: 5-year prospective data of the DESIR cohort’ by Julie Sahuguet et al.1 The results are exciting regarding the significantly lower prevalence of vertebral fracture (VF) among patients with ‘early SpA’ and the crucial aspect of recognising vertebral defects as confounding factors. However, certain aspects of this study require further clarifications.

First, the external validity of the data remains questionable. The current study population has a higher number of female patients compared with the previous studies2 3 where male:female ratio was 4:1. The proportion of patients who were HLA B27 positive is also substantially lower in the current cohort compared with the previous one.3 So, the present cohort may not be the true representative of the spondyloarthropathy (SpA) population.

Second, the diagnostic utility of the low back pain criteria is much lower than expected as shown in a study by Poddubnyy et al. Specificity of the Calin criteria for diagnosing axial SpA is as low as 25%, and specificity for the Berlin criteria is 32%–44.8% depending on whether the physician was blinded or not. The specificity further reduced to 20% and 18%, respectively, if the patients applied the criteria.4 It will be interesting to know how many of the patients in the DEvenir des Spondylarthopathies Indifférenciées Récentes (DESIR) cohort subsequently satisfied the classification criteria for SpA.

Third, the authors have selected patients having low back pain for a duration of 3 months to 3 years. It will be interesting to know the total number of patients screened to achieve these number of patients to be included, as in general, the available data suggests that the median delay to diagnosis has remained stable at 5 years for the patients diagnosed between 1999–2003, 2004–2008 and 2009–2013.5 Could these patients be different from the usual axial SpA? Like having a higher incidence of peripheral arthritis that may be confounding factor for early case pickup.5

Fourth, the higher percentage of the female population in the DESIR cohort may be due to higher proportion having non-radiographic SpA, that may behave differently from radiographic SpA like accruing damage or osteopenia and VFs.

These informations would be crucial to the external validity—extrapolating the findings of the study.