Article Text
PDF
Abstract
Objectives To determine whether there is seasonal variation in disease activity of patients with psoriatic arthritis (PsA).
Methods The authors identified the first available set of consecutive summer and winter visits for every patient from the prospective cohort of PsA. Comparison between summer and winter visits, and comparison of the repeated summer/winter visits, from 1978 until 2011 for the same identified patients was conducted for demographics, disease activity outcomes, laboratory results and treatment. The authors categorised disease activity into high, moderate and low states, and improvement versus flare/worsening. Descriptive statistics were computed and multivariate analyses using logistic regression and generalised estimating equations were conducted.
Results The first available summer and winter visit were identified for 253 patients, and 1789 observations were analysed. There was no statistically significant difference in patients' demographics, disease activity outcomes, laboratory results and treatment between summer and winter visits. Bath Ankylosing Spondylitis Disease Activity Index scores were greater for summer visits and patients graded their disease as being worse in winter as compared with summer in the univariate analysis, but this difference did not hold in the multivariate analysis.
Conclusions The change in season does not affect PsA patients' characteristics and disease activity outcomes as determined by the physicians and patients.
Statistics from Altmetric.com
Background
Seasonal variation occurs in the immune system and studies have shown a trend for worsening of cutaneous psoriasisin the winter and improvement in the summer.1,–,5 Patients with psoriatic arthritis (PsA) tend to report more joint pain during the winter time.5 ,6 In this study, we determined (1) whether there is seasonal variation in disease activity of patients with PsA and (2) whether seasonal variation predicts the occurrence of clinically important change in disease activity.
Methods
Patient selection
This is a retrospective analysis conducted on patients from the PsA cohort at the Toronto Western Hospital. All patients fulfilled the CASPAR (Classification Criteria for Psoriatic Arthritis) criteria for PsA.7
Demographic data
Patients with adult PsA are evaluated in the PsA clinic by rheumatologists according to a standard protocol.7 Demographic data for all patients are collected during the visit.
Clinical assessment
PsA activity (peripheral and axial) and psoriasis activity
At each visit, the number of actively inflamed and damaged joints was recorded.8 All patients completed a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) form on each visit since 2007.9 Psoriasis severity is assessed by the Psoriasis Area and Severity Index (PASI).10
Patient global assessment (PGA)
The patients rated their disease activity at the time of the assessment (PGA) (data available since 2008) (online supplementary figure S1).
Laboratory and radiological assessment
On each visit, information on drugs (non-steroidal anti-inflammatory drugs (NSAIDs), disease-modifying antirheumatic drugs (DMARDs) and biological drugs) and ultraviolet therapy for psoriasis was determined along with C reactive protein (CRP) and erythrocyte sedimentation rate (ESR).
Study design
Demographic statistical analysis and disease activity by season
We identified the first available set of consecutive summer and winter visits for every patient in the PsA cohort since 1978 (within 6 months of one another). The summer and winter seasons were defined as follows: June 22 to September 21 and December 22 to March 21, respectively.
Baseline descriptive statistics were computed with continuous variables summarised by their means and SD, while categorical variables were summarised by their numbers and proportions. Paired t test and χ2 test were used for comparison between summer and winter visits in continuous and categorical variables, respectively.
Comparison between summer and winter visits longitudinally
In patients with the first available set of consecutive summer and winter visits, comparisons between the repeated summer and winter visits from 1978 until 2011 were conducted for demographic data, disease activity outcomes, PGA, ESR, CRP and treatment.
We further categorised PsA disease activity into remission-minimal/mild-moderate/severe for active joints, minimal-moderate/moderate-severe for BASDAI and remission-minimal/mild-moderate/severe for PASI scores (online supplementary figure S1).
All potential predictive factors were assessed for association with each outcome (demographic and disease activity measures) in the univariate and the multivariate analyses. In the multivariate analysis, the generalised estimating equations were used to account for within-patient correlation for repeated summer and winter visits. Since the outcome consisted of more than two discrete categories for joint activity and PASI scores, a multinomial model was fit for the following outcomes: joint activity, PASI scores and PGA. The follow-up time per patient was modelled as an offset in all models. All the models included the same following covariates: season (winter vs summer), age (1-year increase), sex (males vs females), duration of PsA at first visit (1-year increase), calendar time (one decade increase), biologics (yes vs no), DMARDs (yes vs no), NSAIDs (yes vs no) and ultraviolet therapy (yes vs no).
Association between summer and winter visits for improvement versus flare/worsening
Improvement
(1) PASI: decrease by ≥50% as compared with previous assessment, (2) active joints: decrease by ≥40% in the total active joints count as compared with previous assessment, (3) BASDAI: 50% relative change or absolute change of 2 (decrease) (BASDAI≥4 means active disease) and (4) PGA: decrease by 2 grades (patient global score was scored as follows: very good, good, fair, poor and very poor).11,–,14
Flare/worsening
(1) PASI: increase by ≥50% as compared with previous assessment, (2) active joints: increase by ≥40% in the total active joints count as compared with previous assessment, (3) BASDAI: a new score ≥4 (for instance, if the previous BASDAI was ≤3) or increase in BASDAI by ≥50% and (4) PGA: increase by 2 grades.13,–,15
Logistic regression was used to study improvement versus flare/worsening between the first available summer and winter visits. This analysis was conducted on active joint count and PASI scores. The covariates were: order of seasons (1=summer, 2=winter), age, sex, duration of PsA at first visit, calendar time and treatment.
Results
Patient demographics
Two hundred and fifty-three PsA patients with the first available summer and winter visits were identified. In this group of patients, 917 summer and 872 winter visits were analysed from 1978 until 2011. The characteristics of the patients are represented in table 1. There was no statistically significant difference in the patients' demographics, disease activity, laboratory results and treatment in the first available set of summer and winter visits (data not shown) and between all repeated summer and winter visits (table 1).
Demographic characteristics of the 253 patients for a total of 1789 observations by season.
Clinical assessment
PsA activity and PGA
Table 1 shows disease activity in both summer and winter visits for all variables. There was no statistically significant difference in the first available set of summer and winter visits (results not shown) and for all repetitive summer and winter visits for peripheral/axial PsA, PASI and PGA.
BASDAI scores were worse in summer as compared with winter (p=0.047). Disease activity was categorised as severe by 43.3% and 32.7% in summer and winter, respectively. The percentage of patients with mild-moderate disease activity was greater in winter as compared with summer, with 67.3% and 56.6%, respectively. Overall, PGA differs significantly between the summer and winter visits (p=0.046). Patients graded their disease as worse (fair, poor and very poor) in the winter visits as compared with the summer visits, and better (good and very good) in summer as compared with winter (table 2).
Categorised disease characteristics (n=1789 observations) by season for all summer and winter visits for 253 patients
Multivariate analysis results
Although BASDAI and PGA scores were different between the summer and the winter visits, this difference did not hold in the multivariate analysis.
BASDAI: In the first model, BASDAImoderate-severe versus BASDAIminimal-moderate was evaluated and the results were not statistically significant with OR=0.85; 95% CI 0.34 to 2.13, p=0.73. In the second model, BASDAI was evaluated as a continuous variable (1 unit increase) and the results were not statistically significant with OR=1.05; 95% CI 0.44 to 2.49, p=0.91.
PGAfair-very poor versus PGAgood-very good was evaluated and the results were not statistically significant with OR=1.39; 95% CI 0.99 to 1.94, p=0.56.
As in the univariate analysis, the multivariate analysis showed no statistically significant difference for joint severity and PASI scores (table 3).
Association between summer and winter visits for improvement versus flare/worsening
Summary of the results of all fitted models evaluating the association of seasons (winter vs summer) with the outcomes while controlling for different covariates
An improvement in active joints count was documented in 15.3% and worsening/flare in 84.7% of the patients between summer and winter visits. An improvement in PASI scores occurred in 28.8% while worsening occurred in 71.2% of the patients between summer and winter visits.
No association between summer versus winter and clinically significant change in disease activity was found. There was no association between improvement or flare/worsening and seasons (active joints count: OR=1.49, 95% CI 0.27 to 8.14, p=0.65; PASI: OR=0.50, 95% CI 0.10 to 2.54, p=0.41) (results not shown).
For the BASDAI and PGA, the model could not be fitted due to the non-availability of data.
To evaluate the effect of variation in treatment between visits, all patients who had a change in their treatment (start/stop/change in the dose of DMARDs or stop/start of biological drugs) were identified and excluded from the analysis. In this subgroup analysis, we also found no statistically significant change in disease activity between winter and summer visits (data not shown).
Discussion
This analysis shows that seasonal variation does not affect PsA disease activity outcomes. The association of seasonal variation with BASDAI and PGA did not hold after adjusting for other factors. Seasonal variation was also not associated with flare/worsening and improvement.
A study on the Étude et Suivi des Polyarthrites Indifférenciées Récentes cohort showed that the onset of arthritis symptoms during winter or spring is associated with greater radiographic progression at 6 months for patients with early arthritis.16 Our study benefited from the availability of a large sample size of patients with repeated visits over time. Nevertheless, this study also has limitations. First, the analysis used the data available from a single cohort, so the results may not be generalisable. Second, the analysis was based on the selection of two seasons in a group of patients where the majority of the patients are from Toronto. However, this limitation was minimised by choosing the two extreme conditions of the climate—winter and summer.
A recent study has confirmed that vitamin D insufficiency is common in PsA, but no seasonal variation in vitamin D levels was found and, more importantly, no causality between vitamin D levels and disease activity outcomes was established.6 In the current study, we have confirmed that the seasonal variation is not an independent factor associated with clinical outcomes in PsA patients.
Acknowledgments
ZT is the recipient of the Geoff Carr Lupus Fellowship and Arthritis Centre of Excellence Fellowship. VC is supported by the Canadian Institutes of Health Research and the Krembil Foundation. The Psoriatic Arthritis Program is supported by the Krembil Foundation and The Arthritis Society through the SPARCC National Research Initiative.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Competing interests None.
Ethics approval Research Ethics Board, University Health Network, University of Toronto.
Provenance and peer review Not commissioned; externally peer reviewed.