Thesis
Reference
Reproductive factors as predictors of disease development and disease progression in rheumatoid arthritis patients
ALPIZAR RODRIGUEZ, Deshire
Abstract
A ce jour, de nombreuses controverses existent sur le rôle des facteurs hormonaux féminins dans la PR. Nous avons investigué le rôle des facteurs hormonaux féminins dans une population à haut risque, à savoir les parents au premier degré de patients atteints de PR.
Nous avons montré que chez les personnes à risque, la présence d'auto-immunité associée à la PR est associée au sexe féminin, au tabagisme et à l'âge, ainsi que à la ménopause. Nous avons étudié le rôle des facteurs hormonaux dans la progression de la maladie en comparant les femmes atteintes de PR pré- et post-ménopausées. Nos résultats suggèrent que la période post-ménopausique est associée à une progression plus rapide de l'incapacité fonctionnelle. En résumé, les résultats des études présentées dans cette thèse ont montré le rôle important de la période post-ménopausique pour le développement et la progression de la PR.
ALPIZAR RODRIGUEZ, Deshire. Reproductive factors as predictors of disease development and disease progression in rheumatoid arthritis patients. Thèse de doctorat : Univ. Genève, 2018, no. Sc. Méd. 30
URN : urn:nbn:ch:unige-1115106
DOI : 10.13097/archive-ouverte/unige:111510
Available at:
http://archive-ouverte.unige.ch/unige:111510
Disclaimer: layout of this document may differ from the published version.
Section de médecine Clinique Département de médecine interne Service de Rhumatologie
Thèse préparée sous la direction du Professeur Axel Finckh
Reproductive Factors as Predictors of Disease Development and Disease Progression in Rheumatoid Arthritis Patients
Thèse
présentée à la Faculté de Médecine de l'Université de Genève
pour obtenir le grade de Docteur en Sciences médicales « MD-PhD » par
Deshiré Alpízar Rodríguez du
Mexique
Thèse n° 30 Genève
2018
Geneva University Hospitals Prof. Axel Finckh
Reproductive Factors as Predictors of Disease Development and Disease Progression in Rheumatoid Arthritis Patients
Thesis
Presented at the Faculty of Medicine of University of Geneva for the MD-PhD Doctorate in Medical Sciences
by
Deshiré Alpízar Rodríguez
from Mexico
Prof. Axel Finckh - Geneva University Hospitals (Thesis Director) Prof. Cem Gabay – Geneva University Hospitals (Thesis Committee) Prof. Thomas Perneger – Geneva University Hospitals (Thesis Committee)
Acknowledgments page 5
Summary page 6
Résumé page 7
1. Rheumatoid Arthritis
1.1. Introduction to rheumatoid arthritis page 8
2. Female hormonal factors role on rheumatoid arthritis development
2.1. Pre-clinical phases of disease development page 9
2.2. Identification of high risk individuals and prediction rules page 20
2.3. Rheumatoid arthritis prevention page 21
2.4. Risk factors to develop RA and pre-clinical phases in first-degree relatives of patients with RA
page 23
2.4.1. Published article: The prevalence of anti-citrullinated protein antibodies increases with age in healthy individuals at risk for rheumatoid arthritis
page 26
2.5 The role of reproductive factors in the development of rheumatoid arthritis page 32
2.5.1 Introduction page 32
2.5.2 Published article: Female hormonal factors and the development of anticitrullinated protein antibodies in women at risk of rheumatoid arthritis (RA)
page 50
3. Female hormonal factors role on rheumatoid arthritis progression
3.1. Disease activity, disability and progression page 61 3.2. Prognostic risk factors for disease progression page 62 3.3. The role of reproductive factors in the progression of rheumatoid arthritis page 65
3.3.1. Introduction page 65
3.3.2. Submitted article: Role of reproductive and menopausal factors in functional and structural progression of rheumatoid arthritis. Results from the SCQM cohort.
Page 68
4. Conclusions page 97
Abbreviations page 98
Appendix
Appendix A. Published review: Environmental Factors and Hormones in the Development of Rheumatoid Arthritis.
page 113
Appendix B. Published review: The Role of Female Hormonal Factors in the Development of Rheumatoid Arthritis.
page 122
Appendix C. Questionnaire SCREEN-RA page 132
Appendix D. Women’s questionnaire SCQM-RA page 134
Appendix E. Abstracts presented in international scientific meetings page 141
Acknowledgments
I thank my supervisor, Prof. Axel Finckh for all his advice, his kindness, guide and inspiration
I thank Prof. Cem Gabay for his support and precious advice
I thank my colleagues and collaborators who have taught me and helped me in this project Dr. Delphine Courvoisier, PhD, Statistician, University Hospitals of Geneva
Dr. Almut Scherer, PhD, SCQM Foundation Dr. Frauke F Förger, MD, PhD, University of Bern
Clinical Research Team of the Department of Rheumatology of the University Hospitals of Geneva
Swiss Clinical Quality Management (SCQM) Foundation
I thank my former mentors Dr. Juanita Romero Diaz, Dr. María del Carmen Cravioto, Dr. Jorge Sánchez Guerrero and Dr. Armando Seuc for their advice and teaching
Specially I thank God, my family and friends for their love and support My husband for his love and encouragement
My parents for support me and loving me always
Summary
Rheumatoid arthritis (RA) has a higher prevalence, worse activity and severity in women, which suggests that female reproductive factors play a role in the development and progression of RA.
Studies have reported many controversial results. We demonstrated the role of female hormonal factors in the development of RA in a population at high risk of developing RA, namely first- degree relatives of patients with RA. First, we showed that in individuals at risk for RA, the presence of autoimmunity associated with RA is associated with female sex, tobacco smoking and age. We found a stronger association with specific auto-antibodies of the disease around the perimenopausal period in women, but not in men. In a second study, we confirmed that being a post-menopausal woman, particularly during the first years after menopause, is associated with the autoimmunity associated with RA in first degree relatives of RA patients. We found no association of other reproductive factors. In a third study we investigated the role of reproductive factors on the progression of the disease. Our findings suggest that the post-menopausal period is associated with a significant more rapid progression of functional disability in women with RA, particularly women with an earlier age at menopause. However, radiographic joint damage progression did not differ between pre- and post-menopausal women. In summary, the results of the studies presented in this thesis underscore the important role of the postmenopausal period for the development and progression of RA.
Résumé
La polyarthrite rhumatoïde (PR) a une plus grande prévalence, une activité et une sévérité plus élevée chez les femmes, suggérant que des facteurs hormonaux féminins jouent un rôle dans le développement et la sévérité de la maladie. A ce jour, de nombreuses controverses existent sur le rôle des facteurs hormonaux dans la PR. Nous avons investigué le rôle des facteurs hormonaux féminins dans le développement de la PR dans une population à haut risque, à savoir les parents au premier degré de patients atteints de PR. D'abord, nous avons montré que chez les personnes à risque de PR, la présence d'auto-immunité associée à la PR est associée au sexe féminin, au tabagisme et à l'âge. Nous avons trouvé une association plus forte entre la présence d’auto- anticorps spécifiques avec la tranche d'âge correspondant à la péri-ménopause chez les femmes.
Dans une deuxième étude, nous avons confirmé que la ménopause, en particulier pendant les premières années suivant la ménopause, est associée à l'auto-immunité de la PR. Nous n'avons trouvé aucune association avec d'autres facteurs hormonaux féminins. Dans une troisième étude, nous avons étudié le rôle des facteurs hormonaux dans la progression de la maladie en comparant les femmes pré- et post-ménopausées. Nos résultats suggèrent que la période post-ménopausique est associée à une progression plus rapide de l'incapacité fonctionnelle chez les femmes atteintes de PR, en particulier chez les femmes ayant eu une ménopause précoce. Cependant, la progression radiographique des lésions articulaires ne diffère pas entre les femmes pré- et post-ménopausées.
En résumé, les résultats des études présentées dans cette thèse ont montré le rôle important de la période post-ménopausique pour le développement et la progression de la PR.
1. Rheumatoid arthritis
1.1. Introduction to Rheumatoid Arthritis
Rheumatoid arthritis (RA) is the most prevalent systemic autoimmune inflammatory disease affecting between 0.5 and 1% of the adult population worldwide [1]. RA is more prevalent in women, with a female to male ratio above 4 before 50 years old and below 2 after the age of 60 [2-6]. The peak of RA incidence takes place during the fifth decade of life [6, 7], around the age at menopause in women [8], however in approximately 50% of patients the disease may start during reproductive years [9].
The etiopathogenesis of RA is partially understood. Our current understanding is that in genetically susceptible individuals, environmental factors induce specific post-translational modifications, which in turn initiate a pathologic activation of the immune system that eventually leads to the clinical onset of the disease [3, 10]. Monozygotic twin studies indicate that genetic predisposition alone does not explain the development of RA,suggesting that environmental factors play an important role in the etiology of RA [11, 12]. In addition, RA is characterized by the high prevalence of autoantibodies such as rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPAs), that could be present long time before RA clinical onset. Interactions between genetic factors, environmental factors and the presence of autoantibodies have been demonstrated to increase the risk of RA onset. The increased prevalence of RA in women suggests that female hormonal factors play a role in the development of the disease, however many controversies exist about the role of reproductive and menopausal factors, such as parity, menopause or use of hormonal contraception.
RA is characterized by chronic joint inflammation, with a variable clinical presentation. Patients may present varying levels of disease activity, unpredictable responses to drug treatment, leading to a wide range of progressions in disability and erratic impact on mortality [1, 3, 13]. Disability progression is very interesting to study as it shows the long-term effect impact of the disease on patients. Disability progression is influenced by factors, such as disease activity, treatments and presence of autoantibodies. Disease severity may be influenced by socio-economic factors, genetic factors, environmental exposures, and treatment-related effects [14-16]. However, disease progression tends to be more severe in women compared to men [17]. Regarding the role of reproductive factors, such as parity, menopause and hormonal therapy on the development and the progression of the disease many controversies remain.
The aims of this thesis project were to analyze the role of reproductive factors on RA disease development and on disease progression in female RA patients.
2. Female hormonal factors role on rheumatoid arthritis development 2.1 Pre-clinical phases of disease development
The European League Against Rheumatism (EULAR) has proposed a terminology for RA preclinical phases: A) Genetic risk factors for RA, B) Environmental risk factors for RA, C) Systemic autoimmunity associated with RA D) Arthralgias without clinical arthritis, E) Unclassified arthritis [18, 19] (Figure 1).
Figure 1. Phases of RA development. Phase A: Genetic Risk Factors, Phase B: Environmental Risk Factors, Phase D: Symptoms Without Arthritis, Phase E: Undifferentiated Arthritis (UA), Phase F: Rheumatoid Arthritis (RA). Figure from - Mankia K, et al. Arthritis Rheumatol. 2016 Apr;68(4):779-88 [19].
The six phases of RA development are not necessarily consecutives or mutually exclusive. I will outline each phase and information about their association with the development of the disease as it is important for the first aim of this thesis.
Phase A to C, are the pre-symptomatic phases.
Phase A is defined by the presence of genetic risk factors for RA. A genetic contribution to the development of RA has been suspected for a long time due to the increased risk of the disease in first-degree relatives of patients with RA (FDR-RA) [20, 21].Twin studies have estimated 12 to 68 % of the global causality of RA to be heritable [11, 12].
The risk to develop RA is particularly associated with the human leukocyte antigen (HLA) class II molecule-encoding locus in the chromosomal position 6p21.31. The ‘shared epitope (SE)’ is a specific common amino acid sequence in residues 70 to 74, encoded by some alleles of the HLA-
antigen D related (DR) locus, especially the third hypervariable region of the DRβ1 chain. The SE has been associated with 3-fold increase in risk of developing RA [22-24]. Single nucleotide polymorphisms associations within the major histocompatibility complex (MHC) region but outside HLA- DRβ1 have been found and more than 100 loci across the genome have been described, but the SE has the strongest association [23, 25, 26].
Familial aggregation may arise both from genetic and from environmental similarities. The odds ratio of being FDR-RA is 3 to 5 [21]. In a large population-based cohort, familial aggregation was not modified by sex, but a higher familial association was found for RA with younger disease onset (<60 years) [21, 27, 28] . The risk increases also if other autoimmune diseases are present in the family, such as Sjögren syndrome with a standardized index ratio (SIR) of 2.25 or systemic lupus erythematosus with a SIR of 2.1 for RA development [29]. However, as demonstrated in homozygotic twin studies, genetic factors are not sufficient to explain the susceptibility to develop RA, suggesting that other factors, such as environmental factors play an important role, sometimes in association with genetic markers, in the etiology of RA [11, 12, 30].
Phase B is defined by environmental risk factors for RA, the most studied risk factor is tobacco smoking. A meta-analysis of 16 studies estimated that ever smoking increases the risk of developing RA by 40% (odds ratio (OR) 1.40 (95% CI: 1.3 – 1.6)) [31]. The precise mechanism by which tobacco induces the development of RA is still unclear. However, it has been demonstrated that tobacco smoking may induce protein citrullination, particularly in the lungs [10, 32]. The interaction between tobacco-smoking and the presence of shared epitope has been demonstrated and will be commented in detail in pages 16 and 17 of this thesis. Silica has been reported associated with development of RA, particularly in workers with high exposure [33, 34].
Interestingly, the increase risk to develop RA was reported in rescue-recovery workers after the World Trade Center attack of 2001, who were exposed to large amount of silica and dust [35].
Other inhaled pollutants exposures, such as environmental pollution [36], and textile dust [37]
have been associated with the development of RA, particularly among smokers. Other potential exposure risk factors, such as pesticides have been studied but the results have been inconsistent [38].
Several infectious agents have been proposed as risk factors for RA; however, findings have not been conclusive. Some viral infections, such as parvovirus B19, Chikungunya and hepatitis C, are known to cause acute and chronic arthralgias and persistent arthritis, however the association with the development of classifiable RA is not clear [39-41]. Overall, the most prevalent chronic infection consistently associated with RA is periodontitis. RA patients have a higher prevalence of chronic periodontitis and tooth loss compared to patients without RA [42, 43]. A longitudinal population study based on the First National Health and Examination Survey (NHANES I) suggested that subjects with periodontal disease and missing teeth had an increased risk of developing RA (p=0.05). Porphyromonas gingivalis, a common periodontal pathogen associated with chronic periodontitis, is known to express an endogenous peptidyl-arginine deiminase (PAD) that is able to citrullinate different human proteins [44]. Citrullinated proteins are present in serum, synovial and periodontal tissues of RA patients and DNA of P gingivalis has been found in the synovial tissue [45, 46]. However, factors that contribute to the loss of tolerance to the citrullinated proteins are still unclear and studies have not found an association between the presence of antibodies to periodontal bacteria and seropositivity [47]. Another periodontitis related bacteria, Agregatibacter actinomycetemcomitans was associated with ACPA-positivity (p=0.01) in a sample of 196 patients with RA [48]. The American group that described this association provided
mechanistic insights into how Agregatibacter promoted hypercitrullination and neutrophil cell death. They identified a pore forming toxin, leukotoxin A (LtxA) as the molecular mechanism by which this bacterium triggers a dysregulated activation of citrullinating enzymes in neutrophils, mimicking membranolytic pathways that generate autoantigen citrullination in the RA joint.
However, another research group replicated the analysis of the association between the presence of Agregatibacter and ACPA positivity without finding significant results [49].
A dysbiotic microbiota has been observed in RA patients. Particularly an increased prevalence of Prevotella species have been shown in patients newly diagnosed with RA [50-53]. Dysbiotic microbiota in RA patients is partially resolved after disease-modifying antirheumatic drugs (DMARD) treatment [51]. The studies which have reported dysbiotic microbiota are association studies, then causality is difficult to establish with the information available so far [32].
Obesity has been described as a risk factor in women but as a protective factor in men [54-56].
Alcohol moderate consumption (defined as 5.0-9.9 gm/day) has been described as a protective factor among women. A meta-analysis of 9 observational studies found a significant protective effect of alcohol on the development of RA with an OR of 0.78 (95% CI: 0.63-0.96), which was more pronounced in ACPA-positive RA (OR 0.52 (95% CI: 0.36 - 0.76)) [57]. Olive and fish oil have been studied as protective factors, however these associations have not been confirmed. An explanation for the inconsistent findings regarding environmental factors, - taking as example dietary factors -, may be that a single type of food or nutrient may confer only a modest beneficial or detrimental effect, which may be difficult to demonstrate unless groups of dietary factors are measured together and jointly demonstrate stronger effects.
A review was published to discuss environmental factors in the development of RA. Alpízar- Rodríguez D, Finckh A. Environmental factors and hormones in the development of rheumatoid arthritis. Semin Immunopathol (2017) 39:461–468 (See Appendix A) [38]. Table 1 displays environmental factors associated with the development of RA in different populations and the range of published effect sizes.
Table 1. Environmental factors association with the development of RA in general population, focusing on the effect size in women, when available. Ranges of published Hazard or Odds ratio from selected publications are shown.
Environmental factors for RA Hazard ratio or OR
95% (CI)
Risk factors Curent Tobacco smoking [31] 1.4 (1.3-1.6)
Silica dust among male smokers [33] 3.6 (1.8-7.5)
Salt consumption in smokers [58] 2.3 (1.1-4.8)
Periodontal disease [42] 1.8 (1.0-3.2)
Women Current tobacco smoking 1.4 (1.2-1.8)
Posttraumatic syndrome [59] 1.8 (1.2-2.7)
Traffic pollution (living < 50 m distance to a major road) in nonsmokers [36]
1.4 (1.0-2.5)
Exposure to textile dust [37] 2.8 (1.6-5.2)
Obesity [60] 1.3 (1.1-1.7)
Sugar sweetened sodas (>1 daily servinga) [61] 1.6 (1.2-2.3) Protective
factors
Women High UV-B exposureb [62] 0.8 (0.7-0.9)
Moderate alcohol consumption 5.0-9.9 gm/day [63]
0.7 (0.5-0.9)
a 1 standard serving, cup, glass, can, or bottle.
b Low: ≤109 Robertson–Berger units×10−4 (R-B), medium: >109 to ≤117 R-B, high: >117 R-B. 440 R-B units over one half hour produce slight redness in untanned Caucasian skin.UV-B values were assigned by geographic regions.
Phase C is the systemic autoimmunity associated with RA, which is considered the immune onset of the disease and the initial pathologic changes of RA [19]. Autoimmunity associated with RA can be demonstrated long before the first clinical symptoms with the presence of autoantibodies, such as RF or ACPA [64-67]. IgM-RF and ACPAs can be found in the serum of blood donors up to 20 years before the onset of the disease. ACPAs seem to develop before RF and the IgM and IgA subtypes of RF seem to appear before IgG-RF, however the precise timing appears to be variable [66]. ACPAs are very specific for RA (> 95%), though isolated ACPA positivity leads only to a 5% increased risk of developing RA within the next five years. RF is not as specific, and its prevalence increases with age in the general population, independently of gender, reaching up to 25% in individuals older than 85 years. Isolated RF are not associated with an increased risk of RA in the general population [68].
In a high-risk population of individuals who have at least two first-degree relatives with RA (FDR- RA), the presence of RF or ACPAs increases the risk of developing RA by up to 38% and 69%
respectively. ACPAs can be directed towards several different citrullinated protein targets (ACPA fine specificities), which are combined in the commercially available ACPA tests (anti-CCP tests).
Typically, the immune response starts with reactivity to only a few epitopes and then spreads over time to a variety of citrullinated peptides. So far, there is no evidence of a sequence of development in ACPA fine specificities, and there is no confirmation for a higher risk of RA with specific ACPAs. Other autoantibodies such as anti-carbamylated proteins (anti-CarP), anti- malondialdehyde-acetaldehyde, anti-peptidylarginine deiminase, anti-collagen type II and anti- IgG hinge are also often present during the preclinical stage, but their additive predictive value is still unclear [38, 69]. The transition from phase C to symptomatic phases may be influenced by environmental factors.
The phases D to F are symptomatic phases of pre-clinical RA.
Phase D is characterized by arthralgias without clinical arthritis. ACPA positive individuals experiencing arthralgias have a 6-fold increased risk of developing arthritis within 12 months [66].
Arthralgia is a non-specific symptom, self-reported and difficult to measure objectively. Arthralgia and other musculoskeletal symptoms are very prevalent in general population. However, it is usually one of the first symptoms prompting subjects to seek specialized care by a rheumatologist [70]. ‘Clinically suspect arthralgias’ (CSA) were defined by a group of researchers as arthralgias of hand or feet joints for less than 1 year, which were considered at risk for RA based on the clinical expertise of rheumatologists. Approximately 7% of patients with arthralgias referred to rheumatologists were identified as CSA. Among subjects with CSA, the odds for progression to RA were 55 times higher than the odds for patients with unexplained arthralgia [71]. European League Against Rheumatism (EULAR) proposed different clinical characteristics that could be used to define arthralgia suspicious for progression to RA in patients with new CSA referred to rheumatologists. To have a sensitive identification of arthralgias at risk for RA it is required the presence of at least three items [72].
Figure 2. European League Against Rheumatism (EULAR) definition of arthralgia at risk for RA.
RA, Rheumatoid Arthritis. MCP, metacarpophalangic. UA, Undifferentiated Arthritis. From - van Steenbergen HW, et al. Ann Rheum Dis 2017;76:491–6. [72]
Phase E. is also named unclassified arthritis and precedes established RA. Unclassified arthritis is an inflammatory oligo-/poly-arthritis without clinical characteristics that would allow a definite diagnosis. Patients with UA may progress towards RA or other chronic inflammatory rheumatic diseases, such as systemic lupus erythematosus, however, in some cases arthritis may completely resolve [73]. Unclassified arthritis associated with RA autoimmunity (phase C) appear to have a markedly increased risk of developing RA, with absolute risk of up to 50% at 12 months [66, 67, 74-76].
Phase F. Rheumatoid Arthritis is defined with the help of classification criteria, no diagnostic criteria exist [1]. It is important to recognize that most of the largest cohorts and with the longest
follow-up have used the 1987 American College of Rheumatology (ACR) criteria. Four out of seven criteria need to be met to establish RA. The first four criteria, including morning stiffness, arthritis of 3 or more joints, arthritis of hand joints and symmetric arthritis have to be present for at least 6 weeks. Rheumatoid factor was the only biomarker included. Chronic changes of the disease, such as rheumatoid nodules and radiographic changes, such erosions were also included, which delayed the ability of these criteria to classify patients as having RA [77].
The newest classification criteria published in 2010, are more sensitive and are able to classify individuals at an earlier stage of the disease. They must be applied to subjects presenting at least one clinically swollen joint that could not be explained by another disease. They include joint involvement and distribution, not requiring being symmetric; symptoms duration longer than six weeks; serology including ACPAs and RF as well as their titers, and acute-phase reactants. These criteria are being used for most of the newer clinical trials. However, some cohort studies tend to use other criteria, such as a diagnosis by a board-certified rheumatologist [78].
Interactions between genetic, environmental factors and autoimmunity
Interactions between genetic, environmental factors and autoimmunity have been described. In a cohort study of twins, the heritability for ACPA-positive RA was 41% and for ACPA without RA, 10% [12]. In a large population-based study, ACPA-positive heritability was 50% and ACPA- negative 20%. Between ACPA positive RA individuals, the relative risk with the presence of a single share epitope allele is 3.3 (95% CI: 1.8-5.9) and with two copies 5.4 (95%, CI:2.7-10.8) [21].
In epidemiologic studies, ACPA positive smokers without the presence of SE have only a modestly increased risk to develop RA (RR 1.5 (95% CI 0.8-2.6)), but this risk increases markedly if a single copy or two copies of the SE are present (RR 6.5 (95% CI:3.8-11.4) and 21 (95% CI:11.0-40.2), respectively) [10]. In the NHS cohort, population attributable risk (PAR) for RA was 41%
considering a group of environmental factors including smoking (>10 pack-years), low alcohol intake (<5 gm/day), overweight, nulliparity and breastfeeding (less than 12 months). In familial RA alone, defined as having at least one FDR-RA, the PAR was 21%. The findings of this study suggested as the twins studies previously commented, that the environmental factors are responsible for a large proportion of RA risk and different genetic and environmental factors may interact to increase the risk of disease [79].
Smoking is associated with increased levels of extracellular peptidyl-arginine deiminase (PAD) promoting protein citrullination, which may thereafter be presented to T-cells by antigen presenting cells. Individuals with HLA-DR-associated SE alleles have an increased affinity for citrullinated proteins, which may lead to an increased production of ACPA [80]. Another hypothesis is that PAD is the T cell target whose recognition triggers the production of ACPAs, according to a hapten/carrier model [81]. Regarding epigenetic modifications, DNA methylation levels are higher in smokers with the SE and ACPA positivity than in those without SE, but the difference was not observed in nonsmokers [82].
Once RA is established there is no definitive cure for the disease, which has driven the study of genetic and environmental factors that lead to the development of the disease. The identification of factors and the interactions between them have led to the development of prediction rules that could allow to identify high risk individuals even before the development of symptoms.
2.2 Identification of high risk individuals and prediction rules
Prediction rules estimating the individual risk of developing RA and identifying individuals at high risk of developing the disease have been developed. The EPIC-2-NOAR study proposed a risk score to identify individuals at higher risk of inflammatory polyarthritis based solely on easily ascertained lifestyle factors, such as pack-years of smoking, obesity, presence of diabetes and parity (≥2) as risk factors and as protective factors, alcohol consumption, occupational class (professional, manual, neither), and duration of breast-feeding in women. A simple summation of the risk factors and subtraction of the protective factors can identify individuals who are at up to six times higher risk of developing polyarthritis [83]. Other prediction rules have been proposed and it is currently unclear which prediction rule offers the best discrimination to predict future RA development [38].
Using a cohort of individuals with arthralgias and at least one autoantibody associated with RA, van de Stadt developed a prediction rule based on serologic markers and anamnestic factors to categorize individuals in low, intermediate or high risk, with a good discriminative ability for RA within the next 5 years [84]. The hazard ratio for RA (HR) was 14.8 (95% CI: 8.4-28.3) for individuals in the high-risk category, compared to people in the low-risk category.
Figure 3. Prediction rule is composed by 9 variables: familial history of RA, not alcohol use, duration of symptoms < 12 months, intermittent symptoms, arthralgias in upper and lower extremities, pain, morning stiffness >1hour, self-reported swollen joints and presence of autoantibodies. Combining these variables, individuals were categorized in low (0 to 4 points, dotted line), intermediate (5 to 6 points, dashed line) or high risk (7-12 points, continuous line).
From van de Stadt LA, et al. Ann rheum dis 2013, 72(12):1920-1926 [84].
The importance of identifying individuals at high risk lies in the possibility of prevention of RA development.
2.3 Rheumatoid arthritis prevention
Interventions in pre-clinical RA have focused on secondary prevention in patients with recent onset, undifferentiated arthritis. The “PROMPT-study”, has examined whether the development of RA could be prevented by treating patients with early undifferentiated arthritis with low dose methotrexate. [85] After 12 months, the study medication was tapered, and patients were followed for 3 years. The conclusion of this trial was that only ACPA positive patients benefitted from treatment with methotrexate, which delayed the development of RA. A post-hoc analysis of the
“PROMPT–study” has concluded that methotrexate was effective in preventing RA development only for patients with a high (>80%) 1-year predicted risk of progression to RA [86]. The SAVE
trial (Stop Arthritis Very Early) randomized 383 subjects with less than 16 weeks of inflammatory arthritis to a single dose of 120 mg of methylprednisolone versus placebo and found no difference between the groups in terms of progression to RA or drug-free clinical remission [87]. The ADJUST trial (Abatacept study to determine the effectiveness in preventing the development of RA in patients with Undifferentiated inflammatory arthritis) randomized 50 ACPA positive patients with undifferentiated arthritis, to abatacept or placebo for 6 months. After 1 year of follow-up, 46% vs 67% patients treated with abatacept or placebo progressed to RA, respectively [88]. No trial in patients with early undifferentiated arthritis was able to clearly demonstrate that a therapeutic intervention at this phase of disease development may prevent the occurrence of RA.
Thus, it has been hypothesized that earlier intervention, prior to the development of arthritis, would be needed to be curative[89].
Few studies have explored the impact of early pharmacotherapy in subjects with arthralgias, prior to clinical arthritis [89-92]. A trial randomized 83 ACPA positive subjects with arthralgias, but without inflammatory arthritis, to two doses of 100 mg of dexamethasone versus placebo.
Dexamethasone reduced the titers of autoantibodies but did not show a difference in the progression to inflammatory arthritis (20% vs 21%) [93]. The PRAIRI study (Prevention of clinically manifest RA by B-cell-directed therapy in the earliest phase of the disease) randomized individuals at high risk of RA to a single infusion of rituximab versus placebo. [94] High-risk individuals were defined by the presence of arthralgias and two autoantibodies associated with RA (IgM-RF and ACPA) and at least one of the following features: 1) CRP > 0.6 mg/l and 2) subclinical synovitis on ultrasound or on MRI. Preliminary results of this study have shown a delay of the development of arthritis. In total 81 subjects received treatment, after a follow-up of ~12 months 40% developed arthritis in the placebo group and 34% in the rituximab group and there
was a reduction of the risk in the rituximab group at 18 months HR 0.475 (95% CI: 0.190–1.191).
A NIH sponsored clinical trial, the “Strategy to Prevent the Onset of Clinically-Apparent Rheumatoid Arthritis (StopRA)” study will evaluate the efficacy of a 12-month course of hydroxychloroquine (HCQ) to prevent the development of clinically-RA [95]. The STOP-RA trial enrolls FDR-RA patients with elevation of ACPAs to the double of upper normal limit [38].
Prevention of RA may be more effective than treating the disease once developed, because it may be permanent. Prevention could also involve life-style changes, diet and other environmental factors. However, information about the use of non-pharmacological preventive strategies in high- risk populations is lacking.
2.4 Risk factors to develop RA and pre-clinical phases in first-degree relatives of patients with RA
One of the high-risk populations are the first-degree relatives of patients with RA (FDR-RA). The odds ratio of developing RA being FDR-RA is up to 5 times higher compared to the general population. In the case of twins, the risk increases to 6 and in a family with more than 2 cases the odds ratio is higher than 9 [21]. The genetic factors certainly contribute to the risk of developing RA in FDR-RA, however shared environmental factors may also contribute to the increase risk in a large proportion, therefore it is important to study the environmental risk factors contributing to the risk of RA in the FDR-RA population.
An important environmental risk factor is smoking that along with age are associated with both prevalent and incident arthritis in a high-risk cohort of FDR. FDRs younger than 50 years with
>10 pack-years had the highest risk of arthritis (OR 4.39 (95% CI:2.22-8.66)), compared to never
smokers younger than 50 years [96, 97]. Other environmental, diet or hormonal factors have not been studied in the development of inflammatory joint signs or RA in FDR-RA, but there are several ongoing studies [74, 98-100].
As commented previously, the systemic autoimmunity associated with RA is considered as the immune onset of the disease and is characterized by the presence of autoantibody positivity, particularly ACPAs. The prevalence of ACPA positivity has been reported in FDR-RA of 3 to 6% compared to 1% in the general population [101]. Studying the factors associated with the development of ACPAs in FDR-RA could help us identifying possible modifiable environmental factors, and eventually propose preventive measures.
In a cross-sectional study periodontal disease severity was higher in FDR-RA than among healthy controls (p=0.015) and p gingivalis-specific IgG2 was associated with ACPA-positivity (p=
0.049) [102]. In a cohort of FDR-RA, healthy individuals who developed ACPAs had used less omega-3 supplements (OR 0.14 (95% CI: 0.03 -0.68)) and had significantly lower concentrations of omega-3 fatty acids in red blood cell membranes [103]. In another study, the presence of Prevotella spp in stool samples was enriched in FDR-RA who developed ACPAs and/or symptoms, in comparison with FDR-RA without autoantibodies and/or symptoms [104].
Reproductive factors have not been studied in FDR-RA. Studying reproductive and menopausal factors in this population is the first aim of my thesis, in particular to understand their role in the development of autoimmunity associated with RA.
As a first step, we aimed to identify all possible risk factors for the development of autoimmunity associated with RA. We used the SCREEN-RA cohort, a Swiss cohort of FDR-RA. This ongoing cohort enrols individuals without clinical evidence of RA and follows them up yearly, clinically and biologically. We analyzed factors associated with the presence of ACPAs as a marker of autoimmunity associated with RA [101] (See attached paper, pp. 24-29). We found that the presence of ACPA are associated with female sex (OR 2.7 (95% CI: 1.1-6.5)), tobacco smoking (OR 1.8 (95% CI: 1.0-3.3)) and age, particularly 45-55 years old (OR 3.9 (95% CI: 1.6-9.2)). The strong association of ACPA positivity with the age around the perimenopausal period suggests that the decline in ovarian function may be associated with the development of RA autoimmunity and eventually contribute to the increased risk of the disease in women compared to men.
This is the first study to demonstrate the association of ACPA positivity with increasing age in healthy FDRs. The implications of this original research are that age and gender need to be accounted when screening for pre-clinical RA. A study published after our manuscript found that the age-specific prevalence of RA was significantly higher for FDR-RA than the age-specific prevalence in the general population, however authors did not examine the prevalence of ACPAs or stratified by sex [105].
The finding that ACPA positivity had a strong association with an age corresponding to the perimenopausal period suggests that the decline in ovarian function is associated with the development of autoimmunity. This observation is supportive of the role of female hormonal factors in the development of RA.
2.4.1 Published article: The prevalence of anticitrullinated protein antibodies increases with age in healthy individuals at risk for rheumatoid arthritis [101].
Author's personal copy
Clin Rheumatol (2017) 36:677–682 DOI 10.1007/s10067-017-3547-3
The prevalence of anticitrullinated protein antibodies increases with age in healthy individuals at risk for rheumatoid arthritis
D Alpizar-Rodriguez1& Laure Brulhart1& Ruediger B. Mueller2& Burkhard Möller3&
Jean Dudler4& Adrian Ciurea5& Ulrich A. Walker6& Ines Von Mühlenen6&
Diego Kyburz 6& Pascal Zufferey7& Michael Mahler8& Sylvette Bas 1& Danielle Gascon1&
Céline Lamacchia1& Pascale Roux-Lombard9& Kim Lauper1& Michael J. Nissen1&
Delphine S. Courvoisier1& Cem Gabay1& Axel Finckh1
Received: 15 December 2016 / Revised: 21 December 2016 / Accepted: 14 January 2017 / Published online: 21 January 2017
# International League of Associations for Rheumatology (ILAR) 2017
Abstract Transition from genetic risk to the development of systemic autoimmunity associated with rheumatoid arthritis (RA) is considered a key step for the development of RA and often referred to as the immune onset of the disease. The aim of this study is to identify predictors for the presence of anticitrullinated protein antibodies (ACPA) as a marker of systemic autoimmunity associated with RA in a high-risk pop- ulation, an ongoing cohort of first-degree relatives of patients with RA. We assessed the presence of ACPA in individuals without clinical evidence of RA. We examined characteristics associated with ACPA positivity using general estimation equations to account for multiple observations per individual.
A total of 1159 serum samples from 1025 subjects were ana- lyzed, 69 samples (6%) were ACPA-positive, and 227 (20%)
* D Alpizar-Rodriguez
Deshire.AlpizarRodriguez@hcuge.ch
1 Division of Rheumatology, Department of Internal Medicine Specialties, University Hospitals of Geneva, 26, Ave Beau-Sejour, CH 1211 Geneva, Switzerland
2 Division of Rheumatology, Immunology and Rehabilitation, Kantonsspital St. Gallen, St. Gallen, Switzerland
3 Inselspital, Bern, Switzerland
4 HFR, Fribourg, Switzerland
5 USZ, Zurich, Switzerland
6 USB, Basel, Switzerland
7 CHUV, Lausanne, Switzerland
8 Inova Diagnostics, San Diego, CA, USA
9 Division of Immunology and Allergy, Department of Internal Medicine Specialties, University Hospitals and Faculty of Medicine of Geneva, Geneva, Switzerland
positive for rheumatoid factor. Participants had a median age of 45 years (interquartile range (IQR): 33–55) at baseline and 76% were women. Overall, ACPA positivity increased with age (p < 0.001). Among women, ACPA positivity was partic- ularly associated with the age group 45 to 55 years (p
= 0.003), but not among men (p = 0.7). In multivariable adjusted analyses, age older than 45, female sex and tobacco smoking were independently associated with ACPA positivi- ty. In our cohort, the presence of ACPA was associated with older age and peaked in women around age 45 to 55 years, the perimenopausal period, suggesting that the development of ACPA may be favored by the decline in ovarian function.
Keywords Autoantibodies . Epidemiology . Observational studies . Rheumatoid arthritis
Introduction
The etiopathogenesis of rheumatoid arthritis (RA) is only par- tially understood and is thought to result from a multi-step process whereby environmental factors induce a pathological activation of the immune system in genetically susceptible individuals, leading to systemic autoimmunity and subse- quently to the clinical onset of the disease [1]. Specific pre- clinical phases of RA development have been proposed, including a stage of Bsystemic autoimmunity associated with RA,^ considered as the immunological onset of the disease [2]. It has been postulated that the risk factors driving the transition from one pre-clinical phase to another may differ.
Systemic autoimmunity is characterized by the presence of autoantibodies, such as rheumatoid factor (RF) and anticitrullinated protein autoantibodies (ACPA). Studies have shown that the presence of RF and ACPA precede the onset of BRIEF REPORT
Author's personal copy
678 Clin Rheumatol (2017) 36:677–682
RA by several years [3, 4]. Both ACPA and RF are compo- nents of RA classification criteria and define a more severe phenotype, characterized by a more rapid structural joint dam- age and functional impairment [4, 5]. Several ACPA and RF isotypes are more prevalent in the first-degree relatives of patients with RA (RA-FDRs) [6]. Among blood donors with a positive ACPA test, the risk of developing RAwithin 5 years was 5% in the general population, but 69% among individuals who had more than two first-degree relatives with RA [7].
Genetic and environmental factors interact to favor the devel- opment of ACPA in RA [3, 4]; however, the risk factors for the development of ACPA in a healthy population have not been well-established.
The aim of this study is to identify predictors for the pres- ence of ACPA, as a marker of systemic autoimmunity associ- ated with RA, in individuals without evidence of RA, but genetically at an increased risk.
Methods Study design
This is an ongoing Swiss multicenter cohort study (SCREEN- RA) of RA-FDRs [8], comprising subjects without a diagnosis of RA at inclusion. Study participants are enrolled in Geneva, Lausanne, Zurich, Bern, St. Gallen, Basel, and Fribourg. RA- FDRs answer a questionnaire about potential environmental risk factors and are examined by a rheumatologist or a spe- cialized study nurse to rule out the presence of RA, other autoimmune conditions, possible tender or swollen joints.
Serum samples are collected for genetic testing and autoanti- bodies (RF, ACPA) assessment. RA-FDRs are followed annu- ally to assess for the development of signs and symptoms of arthritis. To examine the risk factors for the development of ACPA before clinical apparent RA occurred, we further restricted the analysis to individuals who did not develop RA during follow-up nor had any symptoms suggestive of Bpossible RA,^ based on the connective tissue disease screen- ing questionnaire (CSQ) [9]. The protocol was approved by the local ethics committee and all participants signed an in- formed consent before enrolment.
Variables assessed in the analysis were demographic data, such as age, sex, and various putative environmental factors, such as ever tobacco smoking, heavy smoking (smoking >10 pack-years), current alcohol consumption and frequency of consumption (occasionally, every week, every day, or more than one glass by day), and overweight/obese status using the WHO definition (BMI ≥ 25 kg/m2). Age was categorized based in quartiles (<35, 35–45, 45–55, ≥55). Other variables examined were joint symptoms (self-reported and on exami- nation), more than one FDR with RA, the presence of shared epitope (SE ≥ 1 copy), tooth loss, and poor oral health defined
by the presence of any of the following variables: bleeding on brushing, gingivitis, loss of bone around teeth, or mobile teeth.
Study outcomes
The primary outcome of the study was systemic autoimmunity associated to RA, defined by ACPA positivity, operationally characterized by a positive result to any of the anti-cyclic citrullinated peptide antibodies tests (anti-CCP 2.0, 3.0, or 3.1). Autoantibodies were measured using standard, commer- cially available ELISA kits anti -CCP 2 (CCPlus®
Immunoscan, Eurodiagnostica), anti-CCP 3.1 (QUANTA Lite® CCP3.1 IgG/IgA, Inova Diagnostics), or anti-CCP 3 (QUANTA Lite® CCP3 IgG). A secondary outcome was the presence of RF, using the QUANTA Lite IgM and IgA®
ELISAs and QUANTA Flash® IgM, and IgA chemilumines- cent immunoassays (research use only, Inova Diagnostics).
ACPA and RF positivity was defined by a positive test accord- ing to the manufacturers’ cut-off values (Anti-CCP2 ≥ 25 U/
mL, anti-CCP3.1 and 3 ≥ 20 U/mL, RF QUANTA Lite ≥6 U/
mL, and RF QUANTA Flash ≥ 10344 RLU for IgM and
≥7425 for IgA) [10].
Statistical analysis
We used general estimation equations (GEE) analysis with a log link and robust variance estimates to assess the relative risk (RRs) of ACPA positivity associated with predictors of interest. We analyzed univariable and multivariable associa- tions, adjusting for potential confounders (see variables listed in Tables 1 and 2). We evaluated multiplicative interactions for SE-smoking and age-smoking, as previously published [11].
We further performed sub-analyses by type of ACPA tests (anti-CCP 2.0, 3.0, 3.1) and for RF tests. Finally, in order to explore the level of ACPA over time, we used mixed-effects linear regression for longitudinal data in individuals with at least two sequential ACPA assessments. P values less than 0.05 were considered statistically significant. Sporadically missing covariates data were managed using multiple imputa- tions. All analyses were performed with STATA 14.0 (Stata Corp LP, College Station, Tx, USA).
Results
Among the 1099 RA-FDRs available in the SCREEN-RA cohort, 34 did not have any ACPA assessment, 7 developed RA during follow-up, and 39 had symptoms related to possi- ble RA (by CSQ) and were excluded. A total of 1025 RA- FDRs, with a median age at baseline of 45 years (inter- quartile range (IQR): 33–55), contributing 1159 serum sam- ples were analyzed, of which 69 (6%) were ACPA-positive and 227(20%) were RF-positive.
Author's personal copy
Clin Rheumatol (2017) 36:677–682 679
Table 1 General characteristics
of participants by ACPA Characteristics ACPA-negative ACPA-positive Univariable analysis
positivitya n = 1090 (94%) n = 69 (6%) RR(95%CI)
Age, median (IQR) 45 (33–55) 52 (47–59) 1.0 (1.0–1.1)
Age groups, n (%)
<35 313 (29) 9 (13) Ref
35–45 237 (22) 5 (7) 0.7 (0.2–2.4)
45–55 275 (25) 33 (48) 3.8 (1.6–9.2)
≥55 265 (24) 22 (32) 2.7 (1.1–6.9)
Sex (female), n (%) 826 (76) 62 (90) 2.7 (1.1–6.5)
White European, n (%)b 1021 (94) 65 (94) 1.1 (0.3–3.6)
Ever smoking, n (%)c 510 (47) 43 (62) 1.8 (1.0–3.3)
Heavy smoking, n (%) 124 (11) 13 (19) 1.7 (0.9–3.4)
Current alcohol consumption, n (%)b 870 (80) 55 (80) 0.9 (0.5–2.1)
Never 220 (20) 14 (20) 1
Ocasionally/every week 774 (71) 47 (68) 0.9 (0.4–1.9)
Everyday/>1 glass by day 96 (9) 8 (2) 1.2 (0.5–3.2)
Overweight/obese, n (%)c 364 (33) 24 (35) 1.1 (0.6–2.1)
Lost teeth, n (%)b
<5 824 (76) 44 (64) Ref
5–9 170 (16) 14 (20) 1.5 (0.8–2.8)
>9 96 (9) 11 (16) 2.0 (0.9–4.5)
Poor oral health, n (%) 345 (32) 27 (39) 1.4 (0.8–2.4)
Tender joint ≥ 1 examination, n (%) 239 (22) 16 (23) 1.1 (0.6–1.9) Swollen joint ≥ 1 examination, n (%) 119 (11) 12 (17) 1.7 (0.9–3.2)
RF, n (%) 214 (20) 13 (19) 0.9 (0.5–1.8)
SE (≥1), n (%)b 566 (52) 41 (59) 1.3 (0.7–2.4)
>1 relative with RA, n (%)b 172 (16) 13 (19) 1.2 (0.6–2.6) ACPA anticitrullinated protein antibodies, RF rheumatoid factor, SE shared epitope
a 1159 serum collections from 1025 subjects. Relative risks were calculated by univariable GEE analysis
b Sporadically missing in 5–7%
c Sporadically missing in 2–3%
In univariable analyses (Table 1), ACPA-positive individ- uals were significantly older, predominantly female, and more often tobacco smokers. Among women, ACPA positivity was particularly associated with age 45 to 55 years (RR 4.5 (95%
confidence interval (CI): 1.7–12.1)), and to a lesser degree with age ≥ 55 years (RR 2.6 (95% CI: 0.9–7.6)), compared to women of less than 35 years. ACPA positivity was strongly associated with the age group 45 to 55 in women (p = 0.003), but not in men (p = 0.7). We observed a trend for increasing prevalence of RF positivity with age, but without sex differ- ence (Table 2, Fig. 1). Neither the interactions between tobac- co smoking (ever) and SE (p = 0.2), tobacco smoking and age (p = 0.1), neither tobacco smoking (heavy) nor SE (p = 0.4) were significant in our cohort.
In the multivariable adjusted model, age older than 45, female sex and tobacco smoking were independently associ- ated with ACPA positivity, irrespective of the specific ACPA test. Being associated in the age group of 45 to 55 years was
significantly associated with anti-CCP2 and anti-CCP3.1 pos- itivity and tended to be associated with anti-CCP3 positivity, but did not reach significance (Table 3). In an analysis restrict- ed to subjects with sequential ACPA assessments (n = 125), we observed a slow increase of ACPA titers (anti-CCP3) over the years (p = 0.06).
Discussion
The present study focused on the identification of predictors for the development of ACPA, as a marker of systemic auto- immunity associated with RA in healthy RA-FDRs. We found a prevalence of 6% of ACPA positivity, in line with the prev- alence reported by other RA-FDR cohorts [11, 12]. The pres- ence of ACPA was associated with female sex, age, and to- bacco smoking.
The association with age was particularly strong in women around the perimenopausal period, with a
Author's personal copy
680 Clin Rheumatol (2017) 36:677–682
Table 2 General characteristics
of participants by RF positivitya RF negative
n = 929 (80%)
RF positive n = 227 (20%)
Univariable GEE RR(95%CI)
Age, median (IQR) 46 (34–55) 47 (33–57) 1.0 (0.9–1.0)
Age groups, n (%)
<35 258 (28) 64 (28) Ref
35–45 202 (22) 39 (17) 0.8 (0.6–1.2)
45–55 249 (27) 57 (25) 0.9 (0.7–1.3)
≥55 220 (24) 67 (29) 1.2 (0.8–1.6)
Gender (female), n (%) 710 (76) 175 (77) 1.0 (0.8–1.4)
White European, n (%) 873 (94) 209 (92) 0.8 (0.5–1.2)
Smoking ever, n (%) 453(49) 98 (43) 0.8 (0.6–1.1)
Alcohol, n (%) 339 (36) 76 (34) 0.9 (0.7–1.2)
Overweight/obese, n (%) 318 (34) 71 (31) 0.9 (0.7–1.2)
Lost teeth, n (%)
<5 693 (75) 170 (75) Ref
5–9 155 (17) 29 (13) 0.8 (0.5–1.2)
>9 81 (9) 28 (12) 1.3 (0.9–1.8)
Poor oral health, n (%) 289 (31) 82 (36) 1.2 (0.9–1.6)
Tender joint > 1 examination, n (%) 197 (21) 59 (26) 1.2 (0.9–1.6) Swollen joint > 1 examination, n (%) 108 (12) 27 (12) 1.0 (0.7–1.5)
SE (≥1), n (%) 491 (53) 112 (50) 0.9 (0.7–1.2)
>1 relative with RA, n (%) 145 (16) 39 (17) 1.1 (0.8–1.5)
ACPA anticitrullinate protein antibodies, RF rheumatoid factor, SE shared epitope
a 1156 serum collections from 1025 subjects. Relative risks were calculated by univariable GEE analysis
decrease in the prevalence of ACPA in older women, suggest- ing that the development of ACPA may be facilitated by ovar- ian aging. Regarding RF, we also observed a trend for an increasing prevalence with age, but did not find a difference by gender or peak in prevalence around the perimenopausal period.
The prevalence of RF positivity increases with age in the general population, irrespective of gender, reaching 25%
among individuals older than 85 years [13]. However, the effect of age on ACPA positivity has been less studied. The prevalence of ACPA has been examined in older populations than our cohort, with a median age between 60 to 70 years [14]. In a study in which samples collected from 3116 patients with RA and 15,542 participants with other rheumatic dis- eases were examined (non-RA patients), ACPA positivity was unrelated to age in both RA and non-RA patients [15].
However, in a Japanese study that included 9804 general pop- ulation volunteers without any autoimmune disease, age older than 70 years was associated with ACPA positivity, with an odds ratio (OR) of 2.5 (95% confidence interval [95% CI]:
1.5–4.2) [16]. In a prospective cohort of 966 FDRs without RA, the authors found an interaction between smoking, older age (≥50 years), and inflammatory joint signs [11]. To our knowledge, our study is the first to demonstrate the
association of ACPA positivity with an increasing age in healthy individuals at risk for rheumatoid arthritis.
We focused on an earlier stage of RA development, name- ly, autoimmunity associated with RA, defined by ACPA pos- itivity, in a population at a high risk. Interestingly, our findings regarding ACPA positivity are in line with the established difference in age and gender-ratio for RA, with a female to male ratio above four in patients younger than 50 years and below two in patients older than 60 years [17]. In our study, the effect of age on ACPA positivity was strong in women, but not significant in men. Nonetheless, the number of ACPA-positive men was too low to make a definite conclu- sion about a sex difference. Several studies have demonstrated the presence of increasing levels of different autoantibodies in elderly people, such as antinuclear autoantibodies or rheuma- toid factor [13, 18, 19]. The prevalence of antinuclear autoan- tibodies was 3% in young healthy blood donors and 12% in nonagenarians [19]. This phenomenon has been explained by the increased apoptotic activity related with the aging process and with inmunosenescence [20]. While we also found an increase prevalence of ACPA positivity with age, we did not observe a linear increase, but a peak of prevalence of ACPA positivity around 50 years in women (Fig. 1). The period of age between 45 and 55 years corresponds to the
Clin Rheumatol (2017) 36:677–682 681 Fig. 1 Percentage of a ACPA
positivity and b RF positivity by sex and age groups
perimenopausal period in women. The rapid decline in ovar- ian function and in circulating estrogens at menopause is as- sociated with an increase in pro-inflammatory cytokines and earlier age at postmenopause with an increase risk of RA [21, 22]. Alternative hypotheses for this observation could be immunosenescence, changes in diet, in physical activity or in other environmental exposures with aging. The association
of tobacco smoking and ACPA positivity, although only sig- nificant in univariable analyses, was in line with previous studies on RA development [11, 16].
In sensitivity analyses, we examined three different com- mercially available anti-CCP tests and found comparable re- sults between the different tests with the overall analysis, sug- gesting that our findings were not driven by a particular test.
