Article Text

Download PDFPDF

Will the increasing prevalence of atopy have a favourable impact on rheumatoid arthritis?
  1. University of Cambridge

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    See article on page 275

    Changes in the prevalence or incidence of certain diseases over time are likely to provide clues to pathogenesis. While we rheumatologists worry from time to time about the possible disappearance of rheumatoid arthritis (RA), one of our major reasons for getting up in the morning,1 2 our colleagues in the chest and allergy clinics have no such anxieties, presiding as they do over a substantial increase in the prevalence of asthma and other atopic disorders.3 Particularly striking have been the changes, over comparatively short periods of time, in the prevalence of atopic disorders in countries of the former Eastern bloc, particularly East Germany, where relative genetic homogeneity of the populations in the former East Germany and West Germany makes comparisons particularly telling. Thus despite heavy levels of atmospheric pollution in the east, which would have been thought likely to exacerbate respiratory diseases, the prevalence of asthma was found to be considerably higher in the less polluted west.4 A more recent study shows that only four years after the fall of communism, the prevalence of hay fever and atopy were measurably increased in Leipzig schoolchildren.5 The slogan that has emerged from such observations is that, with respect to atopic disorders, “civilisation is bad for you”, civilisation being a synonym for western lifestyle.6 What are the likely reasons for this, and why would they be of interest to rheumatologists?

    To explain the rheumatology interest requires a little immunological background. One of the main advances in immunology in the past decade has been the realisation that recognition of antigen by T lymphocyte can lead to quite different outcomes, depending on the set of cytokines made by the responding T cell.7 Two such sets have been defined, and the cells making them termed Th1 and Th2 cells (Th referring to CD4+ T helper cells, although a similar sub-division of CD8+ cytotoxic T cells has been demonstrated). The division is clearest for murine T cells, but is also generally applicable to human T cells.8 While postulating only two categories of T helper cell is rather simplistic,9 the Th1/Th2 division has nevertheless been a useful concept. Th2 T cells are required for atopic responses, with an absolute requirement for IL4, the signature Th2 cytokine, in generating IgE responses. Th2 cells also make IL5, which is important for eosinophil activation. In contrast, Th1 T cells, via their production of interferon γ, are critical for protection against intracellular pathogens such as mycobacteria. They also activate monocytes to produce TNFα and IL1, which are known to be important in synovitis, and Th1 cells can themselves make TNFβ. It has also become clear that these kinds of immune response are mutually antagonistic, as interferon γ down regulates IL4 production, while IL4 down regulates interferon γ production, as does IL10, which is a Th2 cytokine in the mouse. Analysis of rheumatic diseases by applying the Th1/Th2 paradigm is not always straightforward, but with respect to synovitis, particularly in RA but also in reactive arthritis and Lyme disease, there is general agreement that T cells in the joint have a predominantly Th1 phenotype.10-16 Thus, the idea has emerged that in some sense atopy and RA might be at opposite ends of an immunological spectrum as far as T cell responses are concerned. Following on from this, it might be predicted that atopy could protect against the development of RA, or conversely, that RA would decrease atopic diseases.

    These predictions have now been tested. In a paper in this issue, Verhoef and colleagues demonstrated that the prevalence of hay fever in RA patients was around half of that in appropriate controls (8%v 15%), and the diagnosis was confirmed in each case by skin prick testing with inhaled allergens, or demonstration of allergen specific IgE. Furthermore, there was evidence that in the small number of RA patients who did suffer from hay fever, the arthritis was less severe than in the non-hay fever RA patients. As expected, in the light of what is known about Th1 and Th2 cells, these RA patients with hay fever had higher levels of serum IgE and peripheral blood eosinophils than other RA patients, and T cells that produced lower levels of interferon γ after a maximal in vitro stimulus. The levels of IL4 produced were not significantly different, but the ratio of IL4 to interferon γ, which indicates the Th1:Th2 balance, was significantly changed. During the hay fever season, these changes were more marked, and although no statistically significant alleviation of disease occurred at this time, there was a trend towards less active disease even in this rather small group of patients. Similar observations have been reported in another study, where the lifetime prevalence of asthma, hay fever, and eczema (taken together) was 7.5% in RA and 18.8% in controls; the difference in point prevalence was even more striking (3.5% v 16.2%).17 This study relied on questionnaires without any confirmation of diagnosis by challenge. Before these recent studies there was a report that the prevalence of RA in an allergy clinic (two of 266) was the same as in the normal population, and that RA patients and controls had a similar prevalence of positive skin prick tests (five of 40 v nine of 40) or atopic disease (14 in each group).18 However, these numbers are too small to resolve the issue, and the difference in prevalence of positive skin prick tests is suggestive of a real difference that would emerge in a larger cohort of RA patients. There is another negative study in the literature when 100 RA patients showed no difference in the prevalence of positive skin prick tests compared with patients with degenerative disease,19 so further confirmation of the findings of Verhoef et al may be required.

    What explanations can be offered for these findings? It is possible that having RA produces a cytokine milieu that is not conducive to developing atopic disease. The data showing a reduced lifetime prevalence of atopy do not support this idea, as atopic disease will commonly present in childhood, long before RA develops. Conversely, does the atopic state make it less likely that RA will develop? If so, what are the factors that determine whether atopy develops, and why is it getting more frequent? Both genetic and environmental factors control atopy; some of the genetic influences are now being recognised (such as a mutation in the IL4 receptor20), and others will no doubt emerge from whole genome surveys.21 As the genome surveys proceed in RA, it may turn out that quite different genes are linked to susceptibility to the two diseases, and that the population at risk of developing RA is separate from the potentially atopic population. More interestingly, it might be that different alleles of the same gene are associated with each disease, which would mean that the diseases had some degree of mutual exclusion at a genetic level. However, the rapid changes in the prevalence of atopy observed at the same time as changes in environment/lifestyle strongly argue that genetic influences do not predominate in determining atopy, and that environmental influences are critical; of these the most important identified thus far are infectious agents.

    Atopy is less common after measles or hepatitis A infection3 and, in a recent study in Japan, was negatively correlated with responsiveness to mycobacterial antigens after immunisation or natural subclinical infection.22 The differentiation of T cells into effectors with a Th1 pattern of cytokine release depends principally on the cytokine IL12, which is secreted by dendritic cells and macrophages, particularly when they are infected with intracellular pathogens such as mycobacteria.23 24 In an “infection-poor” Westernised environment (as recently discussed25), there may be less infection, clinical and subclinical, and hence less IL12 to drive the differentiation of Th1 T cells. Conversely, exposure to the high concentrations of allergens that are common (for example, house dust mite faeces in centrally heated houses), along with a cytokine milieu that does not favour development of Th1 cells, may lead to a Th2 response to allergens with IL4 and IL5 secretion, and consequent IgE production plus eosinophilia. Under normal non-Western conditions, this Th2 capability is reserved for dealing with extracellular parasites, particularly worms, where IgE can assist in their expulsion from the gastrointestinal tract.26 In the face of the current shortage of worms, the Th2 response, uncontrolled by feedback from ongoing Th1 responses, becomes directed against environmental allergens.

    If these are plausible reasons for a “population shift” away from Th1-type immune responses and towards a Th2 pattern of response, what are the implications for RA? Firstly, a predominant Th2 milieu may make the occurrence of RA less likely when susceptible people (with the right HLA-DR type, and additional susceptibility genes) encounter the antigen that triggers the arthritis, irrespective of whether the antigen is exogenous or an autoantigen. Although a Th2 pattern of cytokine production can be associated with autoimmune disease in experimental models,27 it is more often the case that the Th2 response is protective.28-30 In fact, even an ongoing Th2 response to an irrelevant antigen can protect against the development of autoimmune disease after challenge with autoantigen.31 Secondly, RA might still be induced but be less severe—as suggested by the observations on the minority of RA patients with hay fever. Again, in experimental models, joint inflammation and cartilage destruction can be lessened by treatment with the Th2 cytokine IL4 or IL10,32 or with IL13,33 a close cousin of IL4.

    In conclusion, the “Th2 immune deviation” that is typical of an increasingly westernised European population may have very significant effects on the incidence or severity of RA. An important research priority is to ensure that this possibility is thoroughly investigated, particularly in places where lifestyle changes are occurring rapidly, so as to gain as much insight as possible into the influence of cytokine balance in rheumatoid disease. If it is too late to do a study in Leipzig because westernisation there is already at an advanced stage, perhaps we should begin collaborations with our colleagues in the Baltic states. Time for a “Norfolk”34 Arthritis Register in Vilnius, Riga, and Tallinn?

    See article on page 275


    View Abstract

    Linked Articles