Objective To determine the risk of fracture and osteoporosis among patients with psoriatic arthritis (PsA) and psoriasis, compared with the general population and patients with rheumatoid arthritis (RA).
Methods A population-based cohort study was performed in The Health Improvement Network in the UK using data from 1994 to 2014. Patients aged 18–89 years with PsA or psoriasis and up to five unexposed controls matched by practice and start date within that practice were included. Patients with RA and matched controls were included for comparison. Severe psoriasis was defined by a code for psoriasis and either phototherapy or a systemic medication for psoriasis. Incidence and adjusted HRs (aHR) for fracture (all, hip, vertebral) were calculated.
Results Patients with PsA (n=9788), psoriasis (n=158 323) and controls (n=821 834) were identified. Patients with PsA had an elevated risk of all fracture aHR 1.26 (1.06 to 1.27). Patients with mild psoriasis had elevated risk of all fractures, vertebral and hip fracture: aHR 1.07 (1.05 to 1.10), 1.17 (1.03 to 1.33) and 1.13 (1.04 to 1.22). Patients with severe psoriasis had significantly elevated risk of all fracture and vertebral fracture: aHR 1.26 (1.15 to 1.39) and 2.23 (1.54 to 3.22).
Conclusions PsA and psoriasis are associated with an elevated risk for fracture.
- Psoriatic Arthritis
- Rheumatoid Arthritis
Statistics from Altmetric.com
Osteoporosis (OP) is one of the most common and costly diseases: one in every two women and one in five men will experience a fracture after the age of 50 years.1 Hospitalisations for OP fracture are more common than hospitalisations for myocardial infarction and stroke combined2 and fractures result in pain, immobility, nursing home placement, isolation and depression, in addition to other health problems.1 Rheumatoid arthritis (RA) is a known risk factor for OP.3 Ankylosing spondylitis (AS), despite its association with new bone formation and syndesmophytes, is also associated with vertebral OP and fractures.
While psoriatic arthritis (PsA) and psoriasis demonstrate a similar Th1-driven and Th17-driven inflammation to RA and a pathophysiological link to AS, few studies have addressed the risk of OP in these patients4 ,5 Two studies have reported an increased prevalence of osteopenia or OP among patients with psoriasis in Taiwan and Israel.6 ,7 Additional studies examined bone mineral density in patients with PsA compared with healthy controls, though with conflicting results.8–12 These studies have been limited by cross-sectional designs and lack of adjustment for obesity, smoking or other risk factors for OP. To our knowledge, no studies have evaluated the risk of incident fracture in PsA or psoriasis. Therefore, the objective of this study was to examine the incidence of fracture in patients with PsA and psoriasis and compare this with matched controls from the general population and patients with RA.
We performed a longitudinal cohort study to examine the risk of incident fracture among patients with psoriasis and PsA compared with patients from the general population and patients with RA.
Data from The Health Improvement Network (THIN) in the UK between 1994 and January 2014 were used.35
All patients with PsA or psoriasis between the ages of 18 and 89 at the start date were included if they had observation time in THIN after Vision software implementation. Patients were excluded if they died or transferred out of the practice prior to the implementation of Vision software. Patients with a history of fracture or OP or a history of bisphosphonate prescriptions were excluded. Patients with psoriasis, PsA and RA were matched to up to five unexposed controls from the general population (matching is described in the online supplementary methods).
Exposure and outcome definitions
PsA, psoriasis and RA were defined by the presence of at least one read code consistent with these diseases using previously validated codes.13–17 Severe psoriasis was defined as a code for psoriasis plus a code for either phototherapy or a systemic medication for psoriasis. The outcomes of interest were fractures (all fractures, hip fracture and vertebral fracture).18–21 Disease-modifying antirheumatic drugs (DMARDs) and covariates are listed in the online supplementary methods.
Cohort time started at the latest of the following: diagnosis with psoriasis, PsA or RA (diagnosis date for unexposed controls was the encounter date within 6 months of the matched patient's diagnosis date), 180 days after registration in the practice or Vision date (software implementation in the practice). Cohort time ended at earliest of development of the outcome, transfer out of the practice, practice stops contributing to THIN, death or the end of the study. All covariates of interest were measured prior to cohort entry.
Descriptive statistics were used to examine age, sex, person-time and covariate distribution between patients with PsA, psoriasis and unexposed controls. The number of events and cumulative incidence of fracture were calculated for each group. Cox proportional hazards models were used to calculate unadjusted and adjusted HRs with 95% CIs. A purposeful selection modelling approach was used to determine in the most biologically plausible and parsimonious model.22 The proportional hazards assumption was assessed using log–log plots. Sensitivity analyses are described in the online supplementary data.
This study was approved by the University of Pennsylvania Institutional Review Board and the Cegedim Scientific Review Committee. This paper was prepared according to STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.23
After applying inclusion and exclusion criteria, 9788 patients with PsA, 158 323 patients with psoriasis and 821 834 matched controls were identified. Baseline demographics are shown in table 1 and online supplementary table S1. Time in the cohort was similar among the groups. Approximately 5% of patients with psoriasis had been prescribed a DMARD or received phototherapy. Oral corticosteroids were prescribed for 17% of patients with PsA, 21% of patients with severe 9% of patients with mild psoriasis and controls. Proton pump inhibitors were commonly prescribed for patients with PsA and severe psoriasis (31% each) than controls and those with mild psoriasis (both 15%). Patients with PsA and psoriasis were also more likely to have been prescribed an antidepressant in the baseline period and had a higher prevalence of diabetes than controls. Those with mild or severe psoriasis were more likely to be current smokers, while patients with PsA and severe psoriasis had higher rates of heavy alcohol use.
The number of events, unadjusted incidence per 10 000 person-years and HRs for fracture are presented in table 2. Results for RA are included in the table for comparison. After adjusting for OP risk factors, patients with PsA and psoriasis had elevated risk for incident fracture: PsA 1.16 (95% CI 1.06 to 1.27), mild psoriasis 1.07 (1.05 to 1.10) and severe psoriasis HR 1.26 (1.15 to 1.39). Patients with mild psoriasis had an elevated risk for hip fracture 1.13 (1.04 to 1.22) and vertebral fracture (HR 1.17, 1.03 to 1.33). Patients with severe psoriasis had a substantially elevated risk for vertebral fracture: HR 2.23 (1.54 to 3.22). These results were robust to several sensitivity analyses (see online supplementary tables S3 and S4). Defining fracture by a code for fracture followed by a bisphosphonate prescription resulted in slight increases in the HR. The results did not substantially change when patients with a history of fracture were included.
In this study, we found patients with PsA and psoriasis had an increased prevalence of risk factors for OP and fracture (eg, diabetes, alcohol abuse, smoking, depression, antidepressant use, corticosteroids, methotrexate and ciclosporin).24–29 Additionally, patients with PsA and psoriasis had an increased incidence of fracture compared with the general population by 7%–26%. The incidence of vertebral fracture was also increased in patients with severe psoriasis and while hip fracture was elevated in both psoriasis groups, it was only statistically significant in patients with mild psoriasis relative to matched controls after adjusting for risk factors for OP. We found that the risk for any fracture in patients with PsA and severe psoriasis was similar to RA. To our knowledge, these are the first population-based estimates of the risk for incident fracture and OP in patients with psoriasis and/or PsA and the first longitudinal cohort study to address this issue.
Strengths of this study include a large cohort of patients with an average of 6 years of follow-up; the use of THIN in which the exposures definitions (codes for PsA, RA, psoriasis) have been validated and fractures have been previously examined13–17; and the ability to adjust for other measured risk factors for OP, including concomitant medications, body mass index and smoking. Additionally, inclusion of a cohort of patients with RA for internal comparison provides validity to the results as our estimates for RA were similar to previous studies.30–32 Similarly, the incidence of hip fracture among controls in our study was similar to population statistics in the UK (10.7 vs 10.3 per 10 000 person-years), further supporting the validity of our results.33 Finally, the HRs were robust to numerous sensitivity analyses.
Our study also has limitations. There is a risk for misclassification of the outcome when using diagnosis codes to define an event rather than imaging. We addressed this through sensitivity analyses in which we changed the outcome definition; this did not significantly change the results. We also used a secondary definition for fracture in which we required a therapy for OP to address osteoporotic fractures. Vertebral fracture may be underdiagnosed and thus under recorded.24 We conducted a sensitivity analysis to examine whether observation bias affected these results and found no difference when we only included patients in the study followed at least once yearly. Next, disease manifestations, disease activity and use of biological DMARDs are not available in THIN, and therefore we were unable to directly examine their effects on risk of OP and fracture. We were also unable to account for some lifestyle factors such as degree of immobility or laboratory parameters such as vitamin D. Finally, the relatively small number of patients with PsA and/or severe psoriasis may have resulted in insufficient power for some of the outcomes, resulting in wide CIs that include 1.0 despite elevated point estimates (eg, for hip fracture among patients with severe psoriasis).34
In conclusion, fractures, in particular osteoporotic fractures, are a major health problem that results in poor outcomes and OP is largely underdiagnosed. We found that similar to PsA and psoriasis (both mild and severe) were associated with an increased risk for fractures. Screening and management of OP should still be considered for patients with psoriasis and PsA using guidelines available for the general population.5 ,32
The authors thank Yihui Jiang for administrative support.
Handling editor Tore K Kvien
AO and LH contributed equally.
Contributors AO and LH designed the study, performed the statistical analysis, wrote the first draft of the paper and integrated feedback. DS assisted in assembling the analytic dataset. All of the coauthors assisted in interpretation of the data and provided feedback on the manuscript draft. All authors have approved the final version of the manuscript.
Funding AO is supported by NIH by K23 AR063764. JT is supported by NIAMS K23AR068433. JFB is supported by a Veterans Affairs Clinical Science Research & Development Career Development Award (IK2 CX000955). JMG is supported by NIH/NIAMS K24 AR064310. Funds to extract the data relevant for this cohort study were provided by the Rheumatology Research Foundation Preceptorship Award.
Competing interests AO has received support from an investigator-initiated research grant from Pfizer and has consulted for Novartis and Pfizer and has received payment for continuing medical education work related to psoriatic arthritis. JT has an investigator-initiated research grant from Pfizer and has received payment for continuing medical education work related to psoriasis. JMG served as a consultant for Abbvie, Astrazeneca, Celgene, Coherus, Eli Lilly, Janssen Biologics (formerly Centocor), Sanofi, Merck, Novartis, Endo, Valeant and Pfizer, receiving honoraria; and received research grants (to the Trustees of the University of Pennsylvania) from Abbvie, Amgen, Eli Lilly, Janssen, Novartis, Regeneron and Pfizer; and received payment for continuing medical education work related to psoriasis.
Ethics approval This study used de-identified patient information, and was therefore eligible for institutional review board exemption. The Cegedim Scientific Review Committee also reviewed and approved the study protocol.
Provenance and peer review Not commissioned; externally peer reviewed.
Transparency AO affirms that the manuscript is an honest, accurate and transparent account of the study being reported; no aspects of the study have been omitted. There were no deviations from the original study plan.
Data sharing statement While we are unable to share the datasets, code lists are available on request.
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.