Article Text
PDF
Abstract
Objectives Obesity is a known risk factor for developing rheumatoid arthritis (RA). However, it is unclear whether obesity exerts its risk effect during the asymptomatic or the symptomatic clinically suspect arthralgia (CSA) phase of risk. To improve understanding of the effect of obesity on RA development, we aimed to (1) compare body mass index (BMI) at CSA onset to BMI of the general population and (2) study within CSA patients if obesity increases the risk for progression to RA.
Methods 1107 symptomatic persons at risk for RA from four cohorts (CSA Leiden, CSA Rotterdam, SONAR and TREAT EARLIER placebo arm) were studied. For the first aim, baseline BMI was compared with age-matched/sex-matched BMI of the general population. Patients were stratified for anticitrullinated protein antibody (ACPA) status. Regarding the second aim, the association between BMI and inflammatory arthritis (IA) development during 2 years was studied with Cox regression analysis within each cohort and via meta-analysis in all cohorts.
Results CSA patients of all cohorts were more often obese than the general population (respectively 21.9% vs 14.0%, 25.7% vs 14.5%, 26.7% vs 14.5% and 33.3% vs 14.9%, in CSA Leiden, CSA Rotterdam, SONAR, TREAT EARLIER placebo arm). Both ACPA-positive and ACPA-negative CSA patients had a higher frequency of obesity. Within CSA, obesity was not associated with IA development compared to normal weight (pooled effect in meta-analysis of four cohorts HR 1.01 (95% CI 0.93 to 1.08)).
Conclusions Obesity is not associated with RA development within CSA patients but BMI has already increased in CSA compared to the general population. Obesity, therefore, presumably exerts its risk effect at an early asymptomatic phase of RA development, rather than being associated with the disease processes that ultimately result in clinical arthritis.
- Rheumatoid Arthritis
- Risk Factors
- Prevalence
Data availability statement
Data are available on reasonable request. Data can be obtained from the corresponding author on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Multiple case–control studies have identified obesity as a risk factor for the development of rheumatoid arthritis (RA); however, it is unknown whether obesity primarily exerts its effect when the systemic autoimmune response develops and matures, or whether it exerts its effect in the symptomatic at-risk stage when clinical arthritis develops.
WHAT THIS STUDY ADDS
Patients with arthralgia at risk for progression to RA are more often obese compared to the general population.
Combining existing data revealed that obesity in clinically suspect arthralgia (CSA) patients is not associated with RA development.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The combination of results suggests that the risk effect of the well-known environmental risk factor obesity is unrelated to the processes driving the progression of arthralgia to clinical arthritis. The risk effect, therefore, presumably takes place before onset of CSA, when systemic autoimmune responses develop and mature.
From a pathophysiological point of view, this study aids to the comprehension of obesity as risk factor for RA development, and from a clinical perspective, efforts to reduce weight for the purpose of RA prevention might be more effective before CSA onset than when applied thereafter.
Introduction
Obesity is an environmental risk factor for the development of rheumatoid arthritis (RA). Various case–control studies reported a higher frequency of obesity in patients with RA compared to controls.1–3 Subsequent population-based studies confirmed the association of body mass index (BMI) and the development of RA.4–6 However, how and when obesity exerts its effect on the trajectory of RA development is unknown.
The trajectory of the development of RA consists of an asymptomatic and a symptomatic stage. Patients with clinically suspect arthralgia (CSA) are considered to reside in this symptomatic risk stage. The differentiation of these two phases (asymptomatic followed by symptomatic) forms an opportunity to study the trajectory of disease development.7 8 The aetiology of RA is multifactorial and multiple factors of influence are considered necessary before RA (identified with clinically apparent arthritis) emerges. From a more pathophysiological perspective, a European Alliance of Associations for Rheumatology (EULAR) study group subdivided the trajectory of RA development into genetic factors, environmental factors, systemic autoimmunity and symptoms without clinical arthritis.9 However, the timing of different genetic and environmental risk factors remains to be determined. For example, genetic factors, although present since birth, may not be the first hit of RA development. The well-known HLA-shared epitope alleles are recently described not to be important for the development of anticitrullinated protein antibodies (ACPA), while they are important for the broadening of this response, for the development of CSA symptoms and for progression to RA.10 Another example is the extensively studied environmental risk factor ‘smoking’, which does confer risk for developing autoantibodies. However, within the symptomatic CSA-phase, smoking does not mediate progression to RA.10 Knowledge on the timing of the risk effects of risk factors increases our pathophysiological understanding of RA development. In addition, these insights may provide direction to when possible lifestyle interventions, regarding environmental risk factors, could be beneficial.
Currently, no data are available on when obesity exerts its risk effect. Therefore, our overall goal was to increase the comprehension of the timing of the risk effect of obesity during the development of RA. We questioned whether obesity exerts its effect in the asymptomatic at-risk stage, the symptomatic phase, or both. A unique set of symptomatic at-risk individuals from four different cohorts was available to this end. Our first aim was to compare the BMI of CSA patients of these cohorts with BMI of the general population, in order to explore obesity as risk factor exerting an effect in the asymptomatic phase before CSA onset. Our second aim was to determine in symptomatic persons at risk, whether obesity conferred risk for progression from CSA to clinical arthritis and RA.
Methods
Patients
Four longitudinal Dutch cohorts of symptomatic persons at risk for progression to RA were studied.
Firstly, 724 patients consecutively included between April 2012 and August 2022 in the Leiden Clinically Suspect Arthralgia cohort (CSA Leiden) were studied. The inclusion criterium was recent onset of arthralgia (<1 year) of the small joints (metacarpophalangeal (MCP), proximal interphalangeal (PIP), wrist or metatarsophalangeal (MTP) joints) that, according to the expert opinion of the rheumatologist, was considered clinically suspect for development of RA. Exclusion criteria were clinical arthritis at physical examination or another more likely explanation for the arthralgia (e.g., osteoarthritis or fibromyalgia). This cohort is described previously in more detail.7 After exclusion of 13 patients with missing BMI data, 711 patients were studied.
Secondly, SONAR was a multicentre, observational inflammatory arthralgia cohort that included patients at risk for development of inflammatory arthritis (IA) between 2011 and 2015. This Dutch cohort was extensively described previously.11 Of a total of 178, 148 disease-modifying antirheumatic drug (DMARD)-naïve patients with complete BMI data were studied. Included and excluded SONAR patients were similar regarding baseline characteristics (online supplemental table 1).
Supplemental material
Thirdly, the CSA Rotterdam cohort included 132 patients between May 2017 and December 2022. The study protocol, inclusion and exclusion criteria were identical to the CSA Leiden cohort, as described above. One patient was excluded due to missing BMI data, so 131 patients were studied.
Finally, the TREAT EARLIER trial was a randomised placebo-controlled trial that included patients with CSA (similar to the definition of abovementioned CSA Leiden) and the presence of MRI-detected subclinical joint inflammation between April 2015 and September 2019. This clinical trial studied if intervention with a 52-week course of methotrexate induces permanent disease modification.12 In the current study, only patients assigned to the placebo arm were studied (n=117) who all had complete BMI data.
All patients provided written informed consent. More extensive methods, and inclusion and exclusion criteria of the cohorts are described in online supplemental file 1 and summarised in online supplemental table 2.
Body mass index
BMI (kg/m2) was calculated from baseline weight and height using the formula weight (kg)/(height(m)×height(m)). BMI was categorised following the WHO definition (and equally in all cohorts) into normal weight patients (BMI<25.0 kg/m2), overweight patients (BMI 25.0–29.9 kg/m2) and obese patients (BMI≥30 kg/m2).13 Since a relatively low number of patients was underweight (BMI<18.5 kg/m2, n=14/1107), these patients were analysed together with the normal weight patients.
BMI of general population
Data of the Dutch population were exported from statline (Statistics the Netherlands, CBS).14 Yearly questionnaires (2012–2020) provided information on sex, age and mean BMI of all responders, who were considered a representative sample of the general population. These questionnaires subsequently outlined the mean BMI per age group and for males/females separately. Furthermore, percentages of males/females per age group in different BMI categories (normal weight/overweight/obese) were provided.
Outcome inflammatory arthritis
All four cohorts followed patients prospectively for the development of clinical arthritis, determined similarly in all cohorts as clinically apparent arthritis at physical examination by the rheumatologist. CSA Leiden and CSA Rotterdam followed patients for 2 years or until development of clinical arthritis, with scheduled visits at 4, 12 and 24 months and additional visits in between in the case of increasing symptoms. SONAR patients were followed for 1 year with scheduled visits at 0, 6 and 12 months; no study visits occurred in between. Placebo-treated patients from the TREAT EARLIER trial were followed for 2 years with scheduled 4 monthly visits and an additional visit on increasing symptoms. Importantly, patients were not treated with DMARDs or corticosteroids during follow-up. As secondary outcome the development of RA was studied, defined as clinical arthritis plus fulfilment of the 2010 and/or the 1987 classification criteria for RA or clinical diagnosis of RA with DMARD-start.15 16
Statistical analyses
BMI of our study population was compared with BMI of the general population by performing a one-sample t-test; mean BMI was compared with the age-matched/sex-matched reference mean BMI of the general population (more extensive description in online supplemental file 2). Age-matched/sex-matched percentages of obese patients were subsequently compared. As subanalysis, these comparisons were repeated after stratifying for ACPA status. To determine if BMI of CSA patients predicts progression to IA, a Kaplan-Meier graph was plotted and Cox regression analyses were performed. As a sensitivity analysis, we repeated the Cox regression analyses with RA as outcome. All regression analyses were corrected for age and sex. Additionally, multivariable Cox regression analysis was performed by adjusting for ‘currently smoking’. Missingness was <1% for covariates age and sex, and 0%–18% over all cohorts for smoking. This missingness was assumed as missing completely at random and mean BMI was similar in patients with and without missing smoking data. An inverse probability-weighted random effects meta-analysis was performed to estimate a pooled effect hazard ratio (HR) between all cohorts. Additionally, a meta-analysis was performed that included all data from this study as well as data of previously published cohort studies in arthralgia patients at risk for RA development that studied obesity as risk factor.17 18 Time-to-event data were not available in these previous studies, therefore, odds ratios (ORs) were calculated in order to estimate the pooled effect. Analyses were conducted using SPSS (V.25) and STATA (V.16.1). Two-sided p values of <0.05 were considered statistically significant.
Results
Study population
Baseline characteristics are presented in table 1. Female patients were predominantly present in all cohorts (CSA Leiden, SONAR, CSA Rotterdam, TREAT EARLIER placebo arm), respectively, 78%, 83%, 77% and 68%. Mean age was respectively 44, 45, 45 and 47 years. Median tender joint count (TJC) was between 3 and 5 in all cohorts. The cohorts, respectively, consisted of 13%, 19%, 25% and 20% ACPA-positive patients. There were no clinically relevant differences between normal weight, overweight and obese CSA patients (online supplemental table 2).
Baseline characteristics of study populations
CSA patients have higher BMI than the general population
Mean BMI was, respectively, 26.6, 27.3, 28.1 and 28.0 kg/m2 in CSA Leiden, SONAR, CSA Rotterdam and TREAT EARLIER patients. BMI of CSA patients was compared separately per cohort with its age-/sex-matched BMI of the general Dutch population. Because ratios of males/females and age groups differed slightly per cohort, the mean BMI or percentage of obese patients of the general population that CSA patients were compared with, was different per cohort. The mean BMI of CSA patients was higher than that of the general population, for above-mentioned cohorts, respectively, 26.6 vs 25.3, 27.3 vs 25.4, 28.1 vs 25.4 and 28.0 vs 25.7 kg/m2 (all p values<0.001). Similarly, in all cohorts patients were more often obese (BMI≥30) compared to the general population (21.9% vs 14.0%, 25.7% vs 14.5%, 26.7% vs 14.5%, 33.3% vs 14.9%, figure 1).
BMI is higher in CSA patients than in the general population. Histogram of percentages of patients within BMI categories. The general population age-matched/sex-matched reference percentages of patients in each BMI category (<25 kg/m2, 25–30 kg/m2 ≥30 kg/m2) are, respectively, the following for CSA Leiden 54.1%, 31.9%, 14.0%; for SONAR 53.9%, 31.6%, 14.5%; for CSA Rotterdam 53.0%, 32.6%, 14.5%; for TREAT EARLIER: 50.1%, 35.1%, 14.9%. Percentages of CSA patients within BMI categories are outlined in table 1. P values are the results of the one-sample t-test applied to compare mean BMI of the general population to mean BMI of the study cohort. BMI, body mass index; CSA, clinically suspect arthralgia.
Stratification for ACPA status revealed that both ACPA-positive and ACPA-negative CSA patients of all cohorts had a higher BMI, again both compared with the age-/sex-matched reference from the general population (table 2). Again, patients were more often obese than the general population (table 2).
In ACPA-positive and ACPA-negative CSA patients separately, BMI is higher compared with the general population
Obesity is not predictive for progression to IA
Subsequently, we studied the effect of high BMI on the progression of IA within CSA patients. Median follow-up duration was 102, 52, 56 and 104 weeks for, respectively, CSA Leiden, SONAR, CSA Rotterdam and TREAT EARLIER. Higher BMI did not entail a higher risk for progression to IA in CSA Leiden compared with lower BMI (HR 0.98, 95% CI 0.94 to 1.02, figure 2). Despite that high BMI entailed a slightly higher risk of developing IA in the CSA Rotterdam cohort, this association was also absent in the SONAR cohort and placebo arm of the TREAT EARLIER trial (table 3). Also when studying BMI categories, obese CSA patients of CSA Leiden, SONAR and TREAT EARLIER did not have a significantly different risk of progressing to IA compared with lower weight patients (table 3). After stratification, ACPA-positive and ACPA-negative patients had similar associations in all cohorts (table 3).
High BMI is not predictive for progression to inflammatory arthritis. Kaplan-Meier graph showing the development of inflammatory arthritis (IA) per BMI category (normal weight BMI<25.0, overweight BMI 25.0–29.9 and obese BMI≥30). BMI, body mass index.
High BMI is not predictive for progression to IA in CSA patients
Sensitivity analysis
As a sensitivity analysis, the Cox regression analyses were repeated with RA development as outcome (instead of IA development). High BMI was not associated with RA development (online supplemental table 2). Also when summarised via pooled analysis of all cohorts, high BMI did not confer risk for progression from CSA to RA (pooled effect HR 1.01 (95% CI 0.93 to 1.09)). After additional correcting for ‘currently smoking’, effect sizes did not change (online supplemental table 3).
Meta-analyses
In the meta-analysis of the four symptomatic at-risk cohorts studied here, BMI was not predictive for progression to IA (pooled effect HR 1.01, 95% CI 0.93 to 1.08, figure 3A). Thereafter, we performed a meta-analysis that included the four cohorts studied here and also data from previous studies of patients with arthralgia. Again, BMI was not associated with IA development (pooled effect OR 1.01 (95% CI 0.93 to 1.08), figure 3B).
High BMI is not predictive for progression to inflammatory arthritis in multiple ‘at-risk’ cohorts studied here (A) and additionally in those combined with previously published data (B). Forest plots showing the pooled effect of BMI on inflammatory arthritis development of four cohorts that were studied here (A) and in these four cohorts plus cohorts that were previously published (B).17 18 The pooled effect was determined via inverse probability-weighted meta-analysis. I2 depicts the heterogeneity statistic indicating the fraction of variance that can be attributed to heterogeneity between cohorts, for which the random effects model was applied. In addition to the HRs showed in figure A, figure B shows ORs because time-to-event data of the earlier published studies was not available. BMI, body mass index; CSA, clinically suspect arthralgia.
Discussion
Obesity is a risk factor for the development of RA, but the timing at which it exerts its risk effect during disease development was unknown. Therefore, we conducted studies within patients with CSA and compared their BMI to BMI of the general population. We demonstrated that BMI was significantly higher at CSA onset than in the general population. However, a high BMI in the CSA phase did not confer risk of progression to RA. Together these results imply that obesity does not influence the final (inflammatory) processes that result in the onset of clinical arthritis, but presumably exerts its risk effect in an earlier stage of disease development, thus before the onset of CSA (figure 4).
Summary of results of obesity as risk factor for the development of RA. BMI, body mass index; CSA, clinically suspect arthralgia; RA, rheumatoid arthritis.
Development of RA is a multifactorial process consisting of multiple subsequent factors, many of which are yet unidentified and of which the order is still unknown. Immunological processes underlying the development of RA can already evolve years before the onset of clinical arthritis. Autoantibody-response maturation mostly takes place before onset of CSA and does not change during progression from CSA to RA.19 Moreover, gene expression of cytokines, chemokines and related receptors are stable from CSA to clinical arthritis, suggesting that these immunological changes occurred mostly before CSA onset and were not related to the final hit or final process necessary for RA development.20 The well-known environmental risk factor ‘smoking’ shows a similar timing of risk effect, namely during the asymptomatic phase preceding CSA.10 Our results add to this knowledge of the trajectory of RA development by suggesting that a second environmental risk factor, obesity, also exerts its risk effect early and before the occurrence of joint symptoms. This is in line with the notion that obesity often forms a continuous exposure long before onset of symptoms, and would have exerted its risk effect by the time the symptoms arise.
The effect of obesity during RA development, based on the current data, is schematically depicted in figure 4. The pathophysiological mechanisms by which a high BMI promotes RA development remain to be elucidated. Since it is known that fat tissue can secrete proinflammatory cytokines, it is tempting to speculate that such mediators may promote developing an aberrant autoimmune response prior to the onset of joint symptoms.21 Translational research during the asymptomatic phase of RA is needed to identify relevant mediators underlying this risk effect of obesity.
Our data collected in patients at risk of RA development show no effect of obesity on progressing from CSA to RA. Only in the CSA Rotterdam cohort, high BMI was associated with a higher risk of IA development. Of note, the size of this cohort was relatively small, and it contained a relatively big proportion of patients with BMI≥40 kg/m2 (6% compared with 1%–3% in the other cohorts). When studying only CSA Rotterdam patients with BMI<40 kg/m2, no significant association was found anymore, in accordance with other cohorts (data not shown). In none of the other cohorts obesity conferred risk for RA development, and also after pooling in a meta-analysis the association was negative. We additionally combined previously published studies that evaluated BMI in symptomatic at-risk cohorts in a meta-analysis and this also provided a negative finding. Although negative findings in general can harbour the risk of false negativity, the absence of an effect in multiple studies and in meta-analyses lowers the probability of false negative findings.
The cross-sectional approach of the comparison of CSA patients to the general population (resembling a ‘case–control study’) can be considered a limitation. However, this approach is similar to various previous studies in which patients with RA were compared with healthy people.1–3 A longitudinal study in which healthy individuals are followed until the development of CSA would be optimal, although extremely labour-intensive due to the low incidence of CSA. Nevertheless, this could be a research topic for future population studies. For now, we used deductive reasoning. Since high BMI is a commonly known risk factor of RA,4–6 no risk effect of such was found within CSA patients and CSA patients were more often obese than the general population, it is likely that the timing of the risk effect of high BMI occurs before CSA onset. It can be speculated that obesity does not (only) increase the risk of CSA by inflammation-related processes but that obese patients are more often diagnosed with CSA because of an heightened pain sensitivity associated with their excess weight. Longitudinal research where healthy individuals are followed until the onset of CSA is, therefore, relevant for future research.
Another limitation is that index event bias could be a contributing factor to the absent association of obesity and IA development in CSA patients. If this bias was the case, a seemingly protective effect of the exposure (obesity) on the outcome (IA) would occur in the CSA population due to presence of other risk factors of CSA onset. However, it is unlikely that this type of bias impacted our results, because after stratifying for ACPA status and correcting for ‘currently smoking’, effect sizes remained similar. This phenomenon is more extensively explained in online supplemental file 3.
Also, it should be recognised that BMI as weight/height ratio is a proxy that might only partially cover the inflammatory properties of adipose tissue. For instance, visceral fat is considered a source of inflammatory mediators such as adipokines.21 BMI as proxy could underestimate or overestimate the ‘true proinflammatory impact’, forming a limitation of this research. Also, potential changes of BMI over time were not accounted for. Nevertheless, BMI is a commonly assessed and reliably collected variable, which characteristics are useful for validation in different datasets.
Since validation in research is crucial for the reliability of scientific findings, our ability to have studied multiple cohorts is a strength of this study.22 The presence of consistent results enhances the robustness of our findings. Interestingly, the effect of high BMI seems to be disease stage dependent. While obesity does not predict progression from arthralgia to RA, obesity does affect the course of disease of RA after diagnosis. Different effects have been observed: several studies have revealed that high BMI seems beneficial for erosive progression,23–26 while a high BMI is also associated with a higher disease activity.27–30 The observations that the effect of obesity is disease phase dependent and also associated with the disease outcome, calls for translational studies to better understand these intriguing findings on obesity. For instance, an increased understanding of the obesity-related mediators or cytokines that underlie the risk effect is required.
The current findings may have implications for clinical trials of risk reduction. It has been suggested that a healthier lifestyle in people at risk for RA may be a valuable tool for RA prevention.31 32 The current results suggest that efforts to reduce weight, with a view to reduce the risk of RA, may be most valuable in the asymptomatic risk stage, and less efficacy might be expected if these efforts are applied at the stage where patients already developed the risk stage of CSA. Interestingly, previous research showed no association between weight loss through bariatric surgery and the incidence of RA.33
In summary, using data from different cohorts demonstrated that obesity appears to be more frequent in patients at CSA onset compared to the general population, while obesity within CSA is not a risk factor for further progression to RA. This enhances our understanding of the trajectory of RA development and may have implications for the timing at which weight reduction efforts may be considered in light of RA prevention.
Data availability statement
Data are available on reasonable request. Data can be obtained from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and all participants provided written informed consent before inclusion according to the Declaration of Helsinki. The research protocol for the Leiden CSA cohort (P11.210) was approved by the medical ethical committee of the Leiden University Medical Center (LUMC). The research protocol for the Rotterdam CSA cohort (MEC-2017-028) was approved by the local medical ethical committee of the Erasmus Medical Center Rotterdam (EMC). The research protocol of SONAR was approved by the medical ethics committee of EMC (MEC-2010-353) and was assessed for feasibility by the local ethical bodies of Maasstad Hospital and Vlietland Hospital. The research protocol of the TREAT EARLIER trial was approved by the medical ethical committee of the LUMC and the trial is registered with the Netherlands Trials Registry (NTR4853-trial-NL4599). Participants gave informed consent to participate in the study before taking part.
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.
Footnotes
Contributors QAD, HWvS and AHMvdH-vM designed the study. QAD, AMPB, DIK, AW, PHPdJ, EvM and HWvS collected the data. QAD analysed the data and acted as guarantor. All authors contributed to the interpretation of the data and the writing of the manuscript. AHMvdH-vM was the principal investigator. All authors approved the final version of the manuscript.
Funding This work was supported by a ZonMW grant (programma translationeel onderzoek), by an investigator-initiated grant from Pfizer bv., by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Starting grant, agreement No 714312) and the Dutch Arthritis Society.
Disclaimer The funder had no role in analysis and interpretation of the data, or writing of the manuscript.
Competing interests AvdH is an editorial board member for RMDopen. Otherwise none.
Patient and public involvement statement Patient partners were involved in the design of the TREAT EARLIER trial.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.