Abstract
Objective Polymyalgia rheumatica (PMR) is an inflammatory disorder of the elderly characterized by girdle pain and stiffness. Obesity has an influence on disease activity and outcome in rheumatic diseases like osteoarthritis and rheumatoid arthritis. We aimed to investigate the relationship between high BMI and the severity and outcome of PMR, which is incompletely understood.
Methods In a post hoc analysis, 83 patients with recent-onset PMR were studied over 6 months using clinical examination, laboratory evaluation, and girdle ultrasound (US). The modified Health Assessment Questionnaire (mHAQ), 36-item Short Form Health Survey (SF-36), and PMR visual analog scale (VAS) scores, as well as prednisone therapy data, were recorded. Patients were grouped according to their BMI.
Results At baseline, the 12 patients with obesity had significantly more shoulder pain (P = 0.03), global pain (P = 0.03), PMR VAS (P < 0.01), and fatigue (P = 0.03); higher mHAQ (P = 0.01); and lower SF-36 physical component summary (P = 0.048) and SF-36 pain index (P < 0.001). The mean initial prednisone dose was similar among groups, but patients with obesity received a lower dose/kg (1.9 [SD 0.7] mg vs 2.2 [SD 0.7] mg; P < 0.01). At 6 months, patients with obesity were being treated with higher mean daily prednisone doses (8.5 [SD 3.2] mg/d vs 6.2 [SD 5.2] mg/d; P = 0.02), and 40% of them were receiving higher daily prednisone doses than the standard protocol compared with 14% patients without obesity (P = 0.048). Clinical features, laboratory results, and US results were similar between patients with and without obesity.
Conclusion Obesity affects both symptom severity and prednisone utilization in patients with PMR. The reason for this may relate to different subjective pain perception rather than increased inflammation in patients with obesity. BMI should be considered when interpreting symptoms in patients with PMR and deciding their prednisone doses.
In recent times, obesity has become epidemic in the developed world.1 The adipose tissue is an endocrine organ involved not only in metabolism, but also in immunity and inflammation through the production of adipokines and other cytokines.2 These functions have raised interest in the possible relationship of BMI (calculated as weight in kilograms divided by height in meters squared) to the clinical course of several inflammatory musculoskeletal conditions. Obesity, which is a well-known risk factor for osteoarthritis (OA),3 has also been shown to predispose patients to rheumatoid arthritis (RA).4 In patients with obesity, not only is RA incidence higher, but the response to treatment is often poorer.5 The interaction between obesity and inflammatory arthritis is multifaceted, as in the case of RA; patients with overweight or obesity with treatment-naïve RA show a higher degree of synovial membrane inflammation compared to patients with normal weight.6 However, they also have higher concentrations of plasma interleukin 1 receptor antagonist, an antiinflammatory cytokine.6 Accordingly, bone damage seems to be less severe in patients with overweight or obesity with RA.7 In addition, obesity is associated with pain and impaired mobility in patients with systemic lupus erythematosus.8
Polymyalgia rheumatica (PMR) is an inflammatory disorder of older persons that affects the articular and periarticular structures, resulting in girdle pain and stiffness, constitutional symptoms, and raised serological markers of inflammation such as erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP).9 In PMR, the effect of obesity is not completely known. In a previous study, obesity was not associated with an increased risk of developing PMR, nor was it predictive of duration of glucocorticoid (GC) treatment or development of giant cell arteritis (GCA).10 In another study, the risk for PMR was increased in patients who had been previously hospitalized for obesity.11 In addition, efficacy of GC treatment has been related to body weight,12 with leaner patients responding better to therapy. If high BMI is associated with the severity of PMR, then it should be factored in when deciding on disease-management strategies.
This prospective study is an evaluation of the influence of BMI on the clinical characteristics of patients with PMR at disease onset, their outcomes during short-term follow-up, and GC requirements to control disease activity.
METHODS
The data used in this study were derived from a cohort of 125 patients with newly diagnosed PMR. These patients formed the basis for the American College of Rheumatology (ACR)/European Alliance of Associations for Rheumatology (EULAR) classification criteria for PMR, and all 83 patients in our current study met these criteria.13 Briefly, the study population included a cohort of patients with new-onset PMR recruited at 21 community-based and academic rheumatology clinics in 10 European countries and the USA. Entry criteria for patients with PMR were age ≥ 50 years, recent-onset bilateral shoulder pain, and lack of GC treatment for any condition within the previous 12 weeks. Subjects met both the additional inclusion criteria (ie, morning stiffness > 45 minutes, raised CRP and/or ESR) and the exclusion criteria (ie, no infection, active cancer, GCA, or clinical features of common PMR mimics as defined by our previous report)14 at presentation. The diagnosis of PMR was confirmed upon completion of the study if an alternative explanation for a patient’s symptoms was not defined during the 26-week follow-up. Throughout the study, GC treatment for the majority of patients with PMR was administered according to the following predefined treatment protocol: a daily 15 mg oral dose of prednisone or prednisone equivalent (eg, prednisolone) during weeks 1-2, 12.5 mg daily dose during weeks 3-5, 10 mg daily dose during weeks 6-11, 10 mg/7.5 mg dose on alternate days during weeks 12-15, 7.5 mg daily dose during weeks 16-25, and a tapering dosage according to treatment response from week 26 onward. Deviations from the protocol could occur in the event of disease exacerbations.
All patients were evaluated at baseline, 1 week, 4 weeks, 12 weeks, and 26 weeks according to a standardized protocol. For this study, only baseline and week 26 data were considered. At each follow-up visit, clinical evaluation included girdle pain elicited by passive movements and upper limb elevation, as well as verification of alternative diagnoses. Patients whose PMR diagnosis was not confirmed at any time during follow-up were excluded from the study. Data were collected using standardized data collection forms and patient questionnaires translated into national languages. A 100-mm visual analog scale (VAS) was used for recording fatigue and global pain measures (shoulder pain, hip pain, global pain, and PMR VAS), with 0 indicating no pain or fatigue and 100 indicating worst pain or fatigue. Data on PMR VAS scale were recorded according to patients’ responses to the following question: “On a scale from no effect to maximum effect, how would you rate how your PMR affects you today?” Morning stiffness was assessed by directly questioning the patient on its duration in minutes in the previous 24 hours. Functional status and quality of life were assessed using the modified Health Assessment Questionnaire (mHAQ) and 36-item Short Form Health Survey (SF-36). The SF-36 yields physical component summary (PCS) and mental component summary (MCS) scores on a 0-100 VAS, with 0 indicating the least favorable score and 100 the most favorable score. Weight and height were measured at baseline and BMI was calculated. BMI was analyzed continuously using both linear and quadratic effects and was categorized into 4 groups: underweight (< 20), normal weight (20 to 24.99), overweight (25 to 29.99), and obese (> 30). In addition, in a separate analysis, patients with underweight and normal weight were compared to those with a BMI ≥ 25.
Ultrasound (US) evaluation, performed according to EULAR guidelines,15 was employed to assess the previously reported features associated with PMR, including bicipital tenosynovitis, subacromial and subdeltoid bursitis, trochanteric bursitis, as well as glenohumeral and hip effusion. A rheumatologist or a radiologist experienced in musculoskeletal US performed the US examination using linear probes with a frequency range of 6-10 MHz for shoulders and linear or curved array probes with the frequency range of 5-8 MHz for hips.
Statistical methods. Descriptive statistics (eg, percentages, means, medians) were used to summarize the data. Comparisons between groups were performed using chi-square and rank sum tests. Correlations between continuous variables were identified using Spearman methods. Analyses were performed using SAS version 9.4 (SAS Institute) and R version 3.6.2 (R Foundation for Statistical Computing).
RESULTS
Out of 83 patients with PMR, the mean BMI was 26.1 (SD 3.4); of these, 2 (2%) had underweight, 31 (37%) had normal weight, 38 (46%) had overweight, and 12 (15%) had obesity. Data from 79/83 patients (95%) were available for the follow-up evaluation. A first comparison was made at baseline between patients with and without obesity, as reported in Table 1. Patients with obesity with PMR reported significantly higher shoulder pain VAS (P = 0.03), global pain VAS (P = 0.03), PMR VAS (P < 0.01), and fatigue VAS (P = 0.03). They also had higher mHAQ (P < 0.01), lower SF-36 PCS (P = 0.048), and lower SF-36 pain index (P < 0.001) scores. The SF-36 social functioning index was marginally lower in patients with obesity with PMR (P = 0.05). The mean initial dose of prednisone was similar in patients without obesity (15.1 [SD 5.6] mg/d) and with obesity (16.6 [SD 7.0] mg/d). As expected, because prednisone dosing was not weight-based, patients with obesity received a lower dose/kg than those without obesity (1.9 [SD 0.7] mg/kg/d vs 2.2 [SD 0.7] mg/kg/d; P < 0.01). The same comparisons after 6 months of follow-up are shown in Table 2. There were no differences in clinical and laboratory findings, or in patient-reported outcomes (PROs). At 6 months, patients with obesity were treated with a higher mean daily dose of prednisone than those without obesity (8.5 [SD 3.2] mg/d vs 6.2 [SD 5.2] mg/d; P = 0.02). Out of 10 patients with obesity, 4 (40%) were being treated with a higher prednisone dose than the per-protocol scheme vs 8/57 patients (14%) without obesity (P = 0.048).
Characteristics of patients with PMR at baseline according to BMI.
Characteristics of patients with PMR at 6 months according to BMI.
On US examination, obtained at baseline in 80/83 patients (96%), 85% had at least 1 shoulder with signs of subdeltoid bursitis, biceps tenosynovitis, or glenohumeral synovitis; 59% had subdeltoid bursitis, biceps tenosynovitis, or glenohumeral synovitis in both shoulders; and 32% had ≥ 1 of PMR-associated US findings in at least 1 shoulder and at least 1 hip. All US measures improved significantly between baseline and the 6-month evaluation (P < 0.001 for all). However, US results were similar for patients with PMR with and without obesity, both at baseline and after 6 months (Table 3).
US results in patients with polymyalgia rheumatica at baseline and 6-month follow-up according to BMI.
Over 6 months, the 70% improvement in morning stiffness, PMR VAS, global pain VAS, fatigue VAS, shoulder pain VAS, and hip pain VAS, as well as in the composite response, was similar in patients with and without obesity (Supplementary Table S1, available from the authors upon request). Similar results were noted when comparing patients with underweight and normal weight to those with a BMI ≥ 25 (Supplementary Tables S2-S4).
The same features listed in Table 1 were evaluated according to the 4 BMI categories at baseline. There was a clear dose-effect relationship with increasing values from patients with normal weight to patients with obesity for PMR VAS (P = 0.02; Figure 1) and mHAQ (P = 0.02; Figure 2), whereas the values decreased for SF-36 pain VAS (P < 0.001) and SF-36 social functioning (P = 0.048; Figure 1). At 6-month follow-up, no differences were observed between BMI categories. The prednisone dose per weight also decreased continuously between BMI categories (P < 0.01; Figure 2).
Comparison of the baseline visual pain scores, PMR VAS (P = 0.02), SF-36 pain VAS (P < 0.001), and SF-36 social functioning VAS (P = 0.048), according to BMI categories. PMR: polymyalgia rheumatica; SF-36: 36-item Short Form Health Survey; VAS: visual analog scale.
Baseline mHAQ values (P = 0.02) and prednisone dose per 10 kg (P = 0.002) according to BMI categories. mHAQ: modified Health Assessment Questionnaire.
Analysis of the continuous data showed that BMI was negatively correlated with percent improvement of PMR VAS at 6 months (ρ −0.24; P = 0.046), an observation indicating that higher BMI was associated with lower percent improvement. There was no significant association between BMI and percent improvement in global pain VAS (ρ −0.13; P = 0.28), fatigue VAS (ρ −0.11; P = 0.35), shoulder pain VAS (ρ −0.06; P = 0.59), or hip pain VAS (ρ −0.13; P = 0.29).
DISCUSSION
A high percentage (60%) of patients in this study had overweight or obesity. This value was higher than expected; however, the geographical heterogeneity of our patients with PMR hampered predefinition of the expected rate for the different BMI categories. A previous retrospective study performed using data retrieved from medical records showed that the underweight category, defined as a BMI < 18.5, was significantly underrepresented in patients with PMR compared to controls.10 In a study from the Swedish Hospital Discharge Register, patients who were hospitalized for obesity were followed for several decades and every subsequent hospital admission for autoimmune and/or inflammatory diseases was recorded.11 The standardized incidence ratio for PMR in this cohort, compared with controls who were not hospitalized for obesity, was 1.65 (95% CI 1.22-2.19). The real value of these results is difficult to understand because most patients with PMR are not hospitalized. A satisfactory control population was not available for our study.
Our results support the view that the subjective clinical burden of PMR is higher in patients with obesity compared with those without. The biological basis for this suggestion may be related to the increased number of cytokines produced by the adipose tissue or, alternatively, to the poorer quality of life often experienced by individuals with obesity. The observation that ESR, CRP, clinical examination, and US results were similar along BMI categories does not support the view that patients with obesity had a higher degree of inflammation. It is possible that the study was not powered enough to find US differences among groups. In contrast, pain was significantly more intense in patients with obesity with a clear dose-effect response according to BMI. One possible explanation for this finding is that other concomitant musculoskeletal conditions in which pain is related to body weight were present.16 However, in a study of relatively healthy, community-dwelling individuals, those with abdominal obesity showed a 70% increased risk of experiencing pain even after adjustment for the presence of OA, peripheral neuropathy, and insulin resistance; CRP concentration; and nonsteroidal antiinflammatory drug use.17 Alternatively, this finding could be a result of the poor quality of life observed in patients with obesity or may reflect nociceptive patterns influenced by weight. In fact, people with obesity are more likely to be prescribed opioids in the USA.18
The outcome of PMR at the 6-month follow-up was not different in patients with vs without obesity, but the former required a significantly higher daily prednisone dose at the end of the follow-up period. Patients with obesity with PMR received a similar prednisone dose at baseline, but a lower dose/kg, as expected given the dosing protocol was not weight-adjusted. Further, a significantly lower number of patients with obesity were able to adhere to the scheduled tapering of GCs in comparison with patients with normal weight; as a result, the final daily prednisone dose was higher in patients with obesity. In addition, patients with a high BMI experienced a lower percent improvement of PMR VAS upon completion of the follow-up. The more intuitive interpretation of this finding is that the prednisone dose should be higher in patients with overweight or obesity because of the larger volume of distribution. The need to adjust the dose of GCs according to weight has been already shown in a previous study.12 The alternative explanation that obesity-associated inflammation calls for higher GC doses for disease control is not supported by our findings. Other mechanisms possibly include the number and type of GC receptors, which have a complex relationship with obesity.19
The main strength of our study is its prospective design in a cohort of newly diagnosed patients with PMR. All patients were examined following a standardized protocol that included a careful clinical examination and US assessment. The use of PROs provided additional value, as they have been demonstrated to be potent predictors of prognosis and response to treatment.14 This study has several limitations. First, BMI was chosen as a surrogate measure for adipose tissue mass, which was not directly measured. This could interfere with the results, since men have more muscle tissue than women, and inflammation has been associated with adiposity in women but not in men.20 Waist circumference or whole-body dual-energy x-ray absorptiometry are considered to be better indicators of fat tissue but were not performed. Second, as in all multicenter studies, variability in clinical examination, laboratory testing, and US practice was possible, despite our efforts to standardize them. Finally, the frequency of GC side effects was not recorded, and the number of patients with obesity was low.
In conclusion, obesity affects both severity of symptoms and GC treatment dosage. Considering that the data on which this paper is based are more than 12 years old and that the burden of obesity has since steadily increased, it is possible that the effects of obesity may now be even more pronounced. BMI should be considered by clinicians when interpreting symptoms in patients with PMR and deciding the most appropriate therapeutic strategy. Weight-based GC dosing or early use of steroid-sparing agents may be relevant for patients with obesity. However, pharmacokinetic studies on this topic are very few and a systematic review of these studies could not find conclusive evidence to support the view that patients with obesity need higher GC doses to reach appropriate plasma concentrations.21 To understand the reasons for the discrepancy between the results of our clinical study and the pharmacokinetic ones, further research is needed on other possible factors influencing GC requirements in patients with obesity.
ACKNOWLEDGMENT
We acknowledge the following members of the European Alliance of Associations for Rheumatology/American College of Rheumatology polymyalgia rheumatica classification criteria study group: Wolfgang A. Schmidt, Christina Duftner, Peter Mandl, Zsuzsa Schmidt, Annamaria Iagnocco, Carlotta Nannini, Pierluigi Macchioni, Nicolò Pipitone, Georgina Espígol-Frigolé, Maria C. Cid, Víctor M. Martínez-Taboada, Haner Direskeneli, Brian Hazleman, Richard J. Wakefield, Raashid Luqmani, and Andy Abril.
Footnotes
CONTRIBUTIONS
MAC and DC conceptualized and designed the study. MAC wrote the original manuscript, with input from all authors. CSC provided statistical analysis. BD, MS, CD, CS, and ELM made substantial contributions to the design and execution of the manuscript. All authors enrolled patients and made substantial contributions to the interpretation of data and editing of the manuscript, and all reviewed and approved the final version of the manuscript.
FUNDING
Funding for this study was provided by the European Alliance of Associations for Rheumatology, the American College of Rheumatology, and the Mayo Foundation, and supported further by the individual and uncompensated efforts of participating investigators and their staffs and institutions.
COMPETING INTERESTS
The authors declare no conflicts of interest relevant to this article.
ETHICS AND PATIENT CONSENT
Ethics board approval was obtained at all participating institutions before initiation of the study, and all participants gave written informed consent before enrollment.
- Accepted for publication October 17, 2024.
- Copyright © 2025 by the Journal of Rheumatology
