Inflammation, obesity and rheumatic disease: common mechanistic links. A narrative review

Abstract

Obesity represents a rising global health concern, linked to significant social, psychological and physical burden to the individual affected, people around them and the society as a whole. Obesity has been described as a low-grade inflammatory condition, associated with increased production of pro-inflammatory mediators like tumor necrosis factor alpha or interleukin 6 and altered expression of adipokines. Adipokines, mainly produced by adipose tissue, have mixed pro- and anti-inflammatory properties. Obesity rarely exists on its own; instead, it tends to coexist with (often multiple) other comorbidities, including metabolic, cardiovascular, and rheumatic and musculoskeletal diseases (RMDs). In the case of RMDs, evidence is rapidly accumulating on common mechanistic pathways implicated in the inflammatory states seen between RMDs and obesity. Although there remain unanswered questions on the exact mechanisms of inflammation that link obesity to RMDs, what is becoming increasingly known is the association between obesity and adverse clinical outcomes in RMDs. This narrative review discusses insights into mechanisms of inflammation linking obesity and RMDs and evidence on the impact of obesity on treatment response and important disease outcomes. We highlight the importance of targeting obesity, a common and modifiable comorbidity, as part of the routine care of people with RMDs.

Introduction

Obesity represents an important health concern and an escalating global epidemic according to data from the World Health Organization.¹ Aside from the social and psychological dimensions associated with obesity, the physical burden it carries with it is considerable. The latter can be in more global terms, for example, obesity affecting mobility, but also in more specific terms, with direct consequences on the development and progression of other comorbidities; rheumatic and musculoskeletal diseases (RMDs) being no exception. The immunomodulatory properties of adipose tissue and the links with inflammation and autoimmunity have suggested obesity to be a low-grade, chronic inflammatory condition.² It is therefore not surprising that obesity has been implicated as a risk factor for the development of rheumatoid arthritis (RA),^3–6 psoriatic arthritis (PsA)^7–9 and other immune-mediated diseases.¹⁰ In RA, where most of the evidence exists, obesity has been identified as an increasingly prevalent comorbidity even on first presentation of RA.¹¹ This narrative review provides an overview of the association between obesity and RMDs, starting from shared mechanistic pathways through to disease pathogenesis, and the impact on disease outcomes.

Inflammatory pathways in obesity and rheumatic diseases

Adipose tissue and the role in inflammation

Adipose tissue consists not only of adipocytes (fat cells) and preadipocytes (precursor cells) but also of immune cells, like T cells (including invariant natural killer cells), eosinophils, B-regulatory cells (Bregs)¹² and macrophages.^13,14 In the lean state, the immune cells are thought to have a homeostatic role, interacting with adipocytes¹⁵ and maintaining an anti-inflammatory state, producing interleukin 10 (IL-10), IL-4 and IL-13.¹³ In contrast, in obesity, there are heightened T-helper 1 (Th1) and Th17 responses; Bregs, T-regulatory cells and Th2 cells are depleted, and macrophages are polarized to the inflammatory macrophage 1 (M1) phenotype.^13,16,17 These changes result in the production of inflammatory cytokines like tumor necrosis factor alpha (TNFa) and interleukin 6 (IL-6), fueling an inflammatory state. Furthermore, obesity leads to alterations in the expression of the so-called adipokines. The ‘family’ of adipokines consists of more than 50 members, mainly expressed by the adipose tissue and implicated in metabolic and immune processes characterized by both pro- and anti-inflammatory properties.¹⁶ The most well-described adipokines are leptin, adiponectin, resistin and visfatin, all recognized as having a role in inflammation, obesity and autoimmunity (see below). This suggests that adipose tissue is not an ‘innocent bystander’ but an active player in the inflammatory processes,¹⁸ contributing to a local or systemic ‘low grade’ inflammation.

Key ‘inflammatory players’ in adipose tissue

Leptin, a 16 KDa protein, is the first adipokine identified; it acts as a satiety factor, also stimulating energy expenditure.^16,18,19 Although secretion from tissues like synovial cells, chondrocytes and others has been reported, leptin is mainly produced by adipose tissue and is correlated with the adipose mass. Leptin is a pro-inflammatory molecule, acting in both innate and adaptive immunity pathways.¹⁸ In animal arthritis models, leptin-deficient mice developed less severe arthritis with decreased T-cell proliferative response and production of inflammatory cytokines.²⁰ From a clinical point of view (Table 1), leptin seems to be increased in RA patients^21,22 and is associated with increased body mass index (BMI)^22,23 and disease activity,²⁴ while treatment with TNF inhibitors does not appear to reduce its serum levels.^22,23,25 Furthermore, no association was evident between leptin serum levels and radiological progression.^26–28 In ankylosing spondylitis (AS), most studies, including two recent meta-analyses, agree that leptin levels are comparable to those of healthy individuals^29–31 while some investigators have reported decreased leptin serum levels in male patients.³² In contrast to what is observed in RA, leptin levels in AS do not seem to correlate with disease activity³³ and leptin levels at the time of the diagnosis offer protection for subsequent radiologic progression.³⁴ Data for PsA are very limited, with Xue and colleagues³⁵ reporting that leptin levels are increased in these patients and associated with disease activity. In support of the latter, Eder and colleagues reported on PsA patients displaying higher leptin serum levels compared with those with skin psoriasis.³⁶ In connective tissue diseases, most evidence (although still sparse) is seen in systemic lupus erythematosus (SLE), where a recent meta-analysis showed much higher leptin levels in these patients.³⁷

Table 1.

Adipokine serum levels in autoimmune rheumatic diseases.

Adipokine	RA	PsA	AS	SLE
	Trend Association with disease features
Leptin	↑21,22	↑35	↓ (Males)32	↑37
	BMI, disease activity23,34	Disease activity35	Inversely with radiographic progression34	NR
Adiponectin	↑38	NA35,36,39	Unchanged40–42	↑43,44
	Radiographic progression26,38,45	NA35,36,39	HMW inversely correlated with radiographic progression34	Renal involvement (possibly)44,46
Resistin	NA21,47–49	↑39	↑40,41,50	↑51,52
		NR	NA40,41,50	Disease activity, disease damage51
Visfatin	↑21,38,53	↑39	↑41	NA54–56
	Radiographic progression (possibly)28,53	NR	Radiographic progression (possibly)34,41	NR

AS, ankylosing spondylitis; BMI, body mass index; HMW, high molecular weight; NA, existing data are contradictory or not enough to conclude; NR, not reported; PsA, psoriatic arthritis; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus.

Adiponectin is one of the most well-studied adipokines. It is decreased in obesity, being an insulin-sensitizing protein,⁵⁷ and although initially thought to exhibit solely anti-inflammatory properties,¹⁹ it seems that it has both an anti- and pro-inflammatory profile, possibly attributed to its different isoforms.^18,19 In RA, it has been found to be highly present in inflamed RA synovium and synovial fluid⁵⁸ and also to act as a mediator triggering inflammatory responses from RA effector cells, like synovial fibroblasts, chondrocytes and others.^59,60 It has also been implicated in the development of bone erosions, inducing expression of osteopontin which mediates osteoclast recruitment.⁶¹ From a clinical point of view (Table 1), a recent meta-analysis showed that circulating adiponectin levels are much higher in RA patients compared with healthy individuals.³⁸ Furthermore, and as suggested above, it is likely that adiponectin is strongly associated with early radiographic disease⁴⁵ and radiographic progression in RA,²⁶ irrespective of the metabolic status of the patient.^38,45 In AS, adiponectin serum levels seem to be comparable between patients and healthy individuals^40–42 and do not correlate with disease activity. It has been suggested that high-molecular-weight adiponectin might offer protection from radiographic progression in these patients.³⁴ Regarding PsA, data are contradictory.^35,36,39 In SLE, although data are scarce, it has been shown and also confirmed by a recent meta-analysis⁴³ that adiponectin serum levels are increased.⁴⁴ Some authors have suggested that this is more pronounced in those who have renal involvement.^44,46 As mentioned above, the different isoforms of adiponectin might prove challenging in the interpretation of the results, especially for RA.⁶²

Resistin, also known as adipocyte-secreted factor (ADSF) or FIZZ3 (found in inflammatory zone 3) is a 12.5 kDA protein¹⁸ mainly produced by circulating and fat-tissue-resident monocytes and macrophages.^16,18,63 Adipocytes also contribute to its production.^19,64 Resistin, which is increased in obese patients, seems to play an important role in insulin resistance⁶⁵ and like leptin, is an inflammatory mediator. In fact, it is upregulated by inflammatory cytokines and in turn, enhances their production, creating a positive-feedback loop.^16,63 There are many data suggesting that this adipokine could be an important player in autoimmune diseases. In animal models, intra-articular injection of resistin has been found to cause histological and clinical arthritis.⁶⁶ Resistin, has also been found increased in RA synovial fluid, compared with that from non-inflammatory arthritis.⁶⁶ Furthermore, expression of this molecule and its receptor adenylate-cyclase-associated protein 1 (CAP1) has also been found increased in RA synovial tissue, inducing expression of inflammatory cytokines and chemokines from fibroblast-like synoviocytes.⁶⁷ Despite the locally increased resistin levels, data about the levels of this molecule in the serum of RA patients, compared with healthy donors, are contradictory^21,47–49 (Table 1).

Although not supported by all studies,³² resistin seems to be increased in serum from AS patients and associated with radiological progression of the disease.^40,41,50 It is rather unclear if there is any correlation between its levels and disease activity. A study in PsA patients suggested that resistin is also increased in these patients.³⁹ Resistin was also found to be increased in SLE patients and was associated with disease activity and damage,⁵¹ while it has also been found increased in serum and urine of SLE–nephritis patients.⁵²

Visfatin, also known as pre-B-cell colony-enhancing factor or nicotinamide phosphoribosyltransferase, is a 52 kDA protein produced mainly by adipose tissue and leukocytes.¹⁸ Visfatin has pro-inflammatory properties inducing the secretion of inflammatory cytokines and the expression of costimulatory molecules by human monocytes⁶⁸ and inducing the expression of matrix metalloproteinases and chemokines by human fibroblasts⁶⁹ In RA, visfatin was found to induce the production of inflammatory cytokines form synovial fibroblasts and monocytes.⁷⁰ Visfatin is found increased in RA serum,^21,38,53 possibly associated with radiographic progression^28,53 (Table 1). Additionally, it has been suggested that decreased serum levels of visfatin in early RA patients are combined with better disease outcome at month 12.⁷¹ Treatment with infliximab does not seem to affect serum levels of visfatin.⁷² Visfatin has been found increased in the serum of AS patients⁴¹ and possibly associated with radiographic progression^34,41 but not with disease activity, at least in patients who are on treatment with TNF inhibitors.³³ Data for PsA are less, although increased serum visfatin levels have been reported in PsA, too.³⁹ In patients with SLE, data are limited.^54–56

Association between obesity, disease onset and outcomes

Rheumatoid arthritis

Aside from being implicated as a risk factor for developing the disease,^3–6 obesity has also been shown to be a prevalent comorbidity in RA, even on first disease presentation.¹¹ A predisposition to central (trunk) obesity in RA and the link with cardiovascular comorbidities including hypertension, atherosclerosis, as well as impaired glucose level and metabolic syndrome, necessitate that lifestyle and nutritional factors are better addressed. Obesity is a key risk factor for insulin resistance⁷³ and along with the increased use of glucocorticoids as part of the management of RA, it requires targeted advice and management.⁷⁴ The association between obesity and cardiovascular and metabolic syndrome comorbidities among other, which in turn add to RMD burden and negatively influence disease outcomes, further necessitates that obesity is carefully addressed.

A recent meta-analysis indicates that the BMI category of obesity, but not overweight, reduces the chances of achieving minimal disease activity (MDA) in people with RA compared with those with normal BMI.⁷⁵ Despite higher BMI having been paradoxically linked to slower radiographic progression,^76,77 higher rates of total joint replacement are reported.^78,79 These contradicting findings raise questions around the negative impact of obesity on composite Disease Activity Score (DAS) in RA and whether this is driven by objective indicators of inflammation [e.g. erythrocyte sedimentation rate (ESR)] or the more subjective patient reported factors (e.g. tender joint count and patient global assessment).

We have recently addressed this question in our own group using data from two unique and large RA inception cohorts in the UK, the Early Rheumatoid Arthritis Study (ERAS) and the Early Rheumatoid Arthritis Network (ERAN), demonstrating a positive association between BMI and disease activity,⁸⁰ that is, higher BMI was associated with worse disease activity. More specifically, associations of obesity with DAS were replicated with laboratory measures of inflammation (ESR), suggesting direct effects on inflammatory mechanisms. In this study, further exploration of other DAS components (swollen and tender joint count and patient global assessment) and their association with obesity, showed no significant associations. Another study, assessing also the impact of obesity in outcomes in early RA patients, displayed similar findings, showing also that obesity was associated with increased scores of patients’ pain Visual Analogue Scale scores.⁸¹ The nonsignificant association between obesity and swollen joint count, in particular, also suggests that obesity did not bias the clinical examination and recording of this component. This has strong clinical relevance, since obesity can confound assessment of disease activity in inflammatory arthritis, through soft tissue (adiposity) around joints, or through increased pain sensitivity and central pain augmentation in obese patients.^82,83 The latter might explain why some studies have shown obesity to be linked with lower rates of DAS remission, but similar low rates of magnetic-resonance imaging detected inflammation in patients without obesity, signifying that obesity may bias composite disease activity measures.⁸⁴

Aside from the negative effect of obesity on disease activity in RA, it has been shown that increased BMI is also linked with increased odds of nonremission at 24 months and with worse functional ability and health-related quality of life in the short-term supported by our own,⁸⁰ and other data.^81,85–87

Psoriatic arthritis

PsA often coexists with other comorbidities, most of which fall in the metabolic spectrum, including diabetes mellitus, altered lipid profile, fatty liver disease and obesity.⁸⁸ Although obesity is not that frequent in pure axial forms of spondyloarthritis (SpA), its prevalence is increased in patients with peripheral forms of SpA such as PsA.⁸⁹ Interestingly, higher BMI has been reported in PsA patients compared with healthy individuals and patients with psoriasis or RA.⁷

A large UK study found that among psoriasis patients, PsA incidence rates increased with increasing BMI.⁹ It has also been suggested that in patients with psoriasis, BMI at age of 18, could serve as a predictive factor for the development of psoriatic arthritis.⁹⁰ Additionally, regardless of the occurrence of psoriasis, the risk of PsA development is increased in individuals with higher BMI.^8,9

Obesity has also been shown to influence treatment efficiency in PsA. Based on recent data from a Scandinavian registry, obesity was associated with an increased risk of treatment withdrawal (nonadverse event related) in PsA patients on TNF inhibitors. Furthermore, obese patients achieved a lower percentage of a good or moderate European League Against Rheumatism (EULAR) response.⁹¹ Similarly, in a prospective study, PsA patients starting treatment with TNF inhibitors were less likely to achieve MDA at month 12 compared with nonobese patients.⁹² From those who did achieve MDA, increased BMI was an adverse prognostic factor for maintaining MDA at month 24.⁹² Weight loss of ⩾5% from baseline levels was associated with higher rates of MDA at 6 months in obese PsA patients starting treatment with TNF inhibitors.⁹³ A plausible explanation could be that the adiposity increases the distribution volume.⁹⁴ On the other hand, a recent study showed that obese PsA patients were less likely to achieve sustained (more than 1 year) MDA, irrespective of treatment, further reinforcing the link between obesity and PsA course.⁹⁵

Ankylosing Spondylitis

In AS, although fat mass does not seem to be increased compared with healthy subjects,³² adiposity has been linked to adverse disease outcomes. The use of TNF inhibitors in AS with obesity has been associated with significantly lower rates of response to treatment. More specifically, at year 1 of treatment, patients of normal weight, overweight, and obesity, achieved ASAS40 (Assessment in SpondyloArthritis International Society 40% response) in 44%, 34%, and 29%, respectively.⁹⁶ Similar results are seen with infliximab use in two retrospective studies, whereby obese AS patients were less likely to achieve BASDAI50 (improvement in Bath Ankylosing Spondylitis Disease Activity Index ⩾ 50%) after 6 or 12 months of treatment.^16,97 Lower probability of clinical response (as assessed by BASDAI) was also shown for AS patients treated with adalimumab.⁹⁸ Obesity was associated with worse patient perceptions about the benefits of exercise and some of the patient-reported outcomes like patient global score.⁹⁹ Findings on a potential association between obesity and worse disease activity (as assessed by BASDAI) and functionality (as assessed by BASFI, Bath Ankylosing Spondylitis Functional Index, and HAQ, Health Assessment questionnaire) in AS are contradicting.^37,99,100

Systemic lupus erythematosus and other immune-mediated rheumatic diseases

Data for other immune-mediated diseases are very limited. In SLE, obesity has been described in about one third of female patients.¹⁰¹ Whether increased BMI is a risk factor for SLE development remains an open question, as data from big registries are contradictory.^10,102,103 In SLE, increased BMI has been associated with coronary artery disease,¹⁰⁴ venous thromboembolic events,¹⁰⁵ hypertension,¹⁰⁶ decreased functional capacity,^101,107 presence of fibromyalgia,¹⁰⁶ worse quality of life.¹⁰⁸ Whether there is a true association between obesity and disease activity or disease damage remains to be established. Data are lacking on potential links between obesity and other connective tissue diseases, including primary Sjogren’s syndrome. Finally, there are some studies connecting obesity with increased risk for sarcoidosis.^109–111 Results from a large Danish register confirm these findings.¹⁰ Further studies are needed to better define the possible underlying pathogenetic mechanisms.

Conclusion

The inflammatory pathways implicated in obesity and that of RMDs suggest common mechanisms, some of which are yet to be clearly defined. There is accumulating evidence, however, on the negative impact of high BMI/obesity on important disease outcomes in RMDs such as disease activity, function, quality of life and impact on other coexisting conditions and overall prognosis, calling for attention to obesity in routine rheumatology practice. Despite, however, obesity being a prevalent and modifiable comorbidity, the truth is that it remains poorly addressed in routine clinical care. Nutrition and lifestyle advice are not, unfortunately, prioritized enough as part of the overall management of RMDs,⁷⁴ yet interventions to optimize BMI and prevent obesity may have implications in all aspects of disease, from disease activity to quality of life and overall prognosis. We advocate for addressing both generic lifestyle factors that are known to influence health in general (e.g. healthy diet, weight loss, physical exercise) as well as more specific lifestyle factors that have been studied in RMDs and which have been shown to influence outcomes. In line with this, strategies to encourage and support patients to lose weight at any stage of the disease should be addressed, as they could lead to immediate/short-term benefits as well as longer-term coexisting conditions such as cardiovascular disease and general health benefits. Indeed, emerging evidence demonstrates that weight loss either by diet and exercise or bariatric surgery can suppress disease activity in RA and other RMDs, reflected through decreased both serum inflammatory markers and use of medications.^112,113 We therefore wish to call attention, and impel immediate action to, this modifiable and preventable global health problem as part of the routine management of people with RMDs.