Epidemiology of Lupus
IntroductionLupus (Latin for ‘wolf’), was first coined as a medical term in the 18th Century to describe a variety of skin conditions. In the early 21st century, systemic lupus erythematosus (SLE) is recognised as a chronic, often-severe autoimmune rheumatic disease that affects people worldwide.With the evolution of immunological understanding, including the recognition of the antinuclear factor, and the development of American College of Rheumatology criteria for the classification of lupus, more studies have been carried out allowing a deeper understanding of occurrence, epidemiology and risk factors.
Prevalence and IncidenceLupus appears to be a relatively uncommon disease. The prevalence has been estimated in several different countries mostly, however, in the developed world, using different techniques of case ascertainment. The authors of one metaanalysis (including those studies in Europe and North America) suggested an overall weighted mean prevalence of 24/100,000 population.Three English studies have produced prevalence estimates of: 12/100,000, 25/100,000 and 28/100,000 and the only study in N Ireland estimated a rate of 254/100,000. Studies in countries which include predominantly white populations have resulted in lower prevalence estimates (e.g. England) when compared with studies among populations with a significant proportion of Afro-Caribbeans, Asians and Hispanics.
It is more difficult to estimate incidence for a rare disease but studies in both North America and Europe have produced estimates that are similar (approximately 1 - 8 cases per 100,000 persons per year).The lowest rates of incidence were seen among Caucasian Americans, Canadians and Spaniards and the highest rates among Asian (10.0 cases /100,000) and Afro-Caribbean (21.9 cases /100,000) residents of the UK. In an average UK practice list of 3000 patients, therefore, a GP would not expect to see a new case of lupus more often than every 7-10 years.
GenderLupus occurs primarily in women, in particular during childbearing years, when approximately 9 of 10 cases of lupus are female. Female gender is the single strongest host factor in predicting occurrence of lupus. Although still evident in childhood and late-onset lupus, the female predominance rises with puberty, peaking in young adulthood and then declines after the female menopause. The reasons for this gender imbalance are not currently known, but genetic, hormonal and differential exposure to occupational and environmental factors may be relevant.
AgeAlthough the age-adjusted prevalence and incidence rates obtained in the epidemiological studies are not homogeneous, the peak age of onset among women seems to be during childbearing years, between 15 and 40 years. Importantly however, the median age at diagnosis among women is 37-50 years across these studies, reflecting that this is frequently a delayed diagnosis. Lupus with age at onset > 50 years has been shown to have slightly lower disease activity compared with onset below 50 years of age.
EthnicityWhilst lupus is found worldwide, it is more commonly found in some countries, and within a country certain ethnic groups appear to be more susceptible to developing lupus than others. Additionally, the course and presentation of lupus appears to vary between patients of different ethnicities.
In the UK, lupus was more prevalent among people of Afro-Caribbean origin, followed by Asians and then Caucasians. Similarly, studies among populations in Australia and New Zealand found higher rates of occurrence among aboriginals when compared with Caucasians and studies in China, the Philippines and Japan all find higher rates among these populations than Caucasians. Interestingly, where lupus has been investigated in West African countries (the origin of the ancestors of many of the European and North American immigrant populations), rates of occurrence appear to be low. It is noteworthy, however, that in most of these studies the non-Caucasians are ethnic minorities within a country and notdistinct ethnic groups. For example, Asians within the UK are different groups of individuals representing much biological, environmental and genetic heterogeneity. In many parts of the world, particularly North America, minority population groups are often socioeconomically disadvantaged. Socioeconomic factors affect access to healthcare, quality of care and compliance with care and it may be that differences between minority and majority ethnicities may be exaggerated by these factors.
Ethnicity does not seem to purely affect the rate of occurrence of disease but also the clinical presentation and autoantibody profile, such that black patients with lupus seem to be more commonly affected by discoid skin lesions, cellular casts in the urine and serositis. Immunologically, black patients are more likely to have a pattern of Sm and RNP antibodies than white patients and anti-Ro antibodies are seen more frequently in Southern Chinese and North Africans. There is evidence that lupus nephritis is more prevalent in African and Hispanic Americans as well as Chinese and other Asians. The mortality attributed to lupus in the US shows a less favourable outcome for African-American patients and this has been mirrored in British studies. Rates of survival have markedly increased in the developed world in the past few decades but poorer survival continues to be recorded in India and amongst Black Caribbean patients. Again, it is unclear whether this reflects heterogeneous disease presentation or other factors such as inequity of care and socioeconomic factors.
Genetic factorsLupus has long been considered to have a genetic predisposition. Cases of lupus cluster within families such that a female first-degree relative of a lupus patient may carry a risk increased by up to 6-fold of developing the disease herself. Surveys among twins have demonstrated a high concordance for the diagnosis of lupus among monozygotic twins (24-69%) compared with 10% concordance among dizygotic twins. Additionally, one study showed a striking concordance for the clinical and immunological manifestations of lupus among monozygotic twins when compared with non-twin sib pairs. The authors did, however, reflect that ‘behaviour as well as biology’ could explain the short interval in diagnosis between twin pairs – 50% of concordant pairs had sought care because of the recent diagnosis in their twin.
The results of many genetic studies suggest a polygenic mode of inheritance for lupus, with possibly as many as 100 genes involved in susceptibility, but perhaps involving 3-4 dominant alleles. Attention has focussed on the HLA (Human Leukocyte antigen) for detailed analysis of genetic epidemiology. These studies suggest that it is Class II antigens (HLA-DR2 and DR3) which are most important among lupus patients but with different strengths of association in different racial groups. One group has investigated an ‘autoimmunity’ gene among multi-case families and have proposed an autosomal dominant pattern of inheritance of this gene, with its expression modified by age and gender (higher penetrance in females than males). Polymorphisms in genes for interferon regulatory factor 5 and protein tyrosine phosphatise N22 have been found to be associated with increased risk of lupus.
Hormonal factorsDue to the strong link between gender and lupus, hormonal factors are strongly implicated in the aetiology of the disease. Investigators have explored the role of endogenous and exogenous hormonal factors.Age at menarche may be a marker of estrogen exposure and has been examined as a risk factor in several studies. Although the results are in some cases conflicting, these studies provide some evidence of an increased risk of incident lupus, such that women with a menarche aged 10 years had an approximate doubling in their relative risk of incident lupus. Menstrual irregularity (long and short cycles) was associated with lupus in two studies but this association was not found in other studies. Age at first intercourse has not been found to be associated. History of miscarriage was associated with lupus but this may reflect the relationship between lupus and the antiphospholipid antibody syndrome. Studies have produced conflicting results in exploring the role of oral contraceptives and lupus. The one study which demonstrated a positive association between the oral contraceptive and lupus involved exposure to relatively higher does of estrogen than those currently utilised and one study suggested a decreased risk of lupus amongst those who used the progesterone-only pill. There is an increased risk of lupus seen among postmenopausal women who had a surgical menopause, with an effect almost entirely explained by use of HRT.The findings of one recent study have suggested that Hormone Replacement Therapy (HRT) can be safely used among patients with known lupus without an increased risk of disease flares.
Socioeconomic factorsSocioeconomic status could have a major impact on lupus disease manifestations and mortality, independent of ethnicity. Access to healthcare, quality of care and compliance with care are all affected by these factors. Few studies have examined socioeconomic status in sufficient detail to truly understand its role in disease development or causation and this is an area requiring further research.
Dietary factorsNutrition has had little study as an aetiological factor in lupus. Better studied have been the effects of dietary factors among patients with known lupus on disease control. Omega-3 fatty acids are thought to provide a beneficial effect with some anti-inflammatory efficacy.Anti-oxidants (vitamins A, C E and beta-carotene) may also have a positive effect on disease activity. Lower levels of vitamin D have been observed in lupus patients, but it is unclear whether this is because of disease activity. Although nutritional factors may have been poorly studied, this has not prevented considerable coverage in the popular press and on the internet, most of the claims being unsubstantiated.
Cigarette smokingCigarette smoking has been causally linked to the development of multiple autoimmune diseases, including lupus. There have been a number of heterogeneous studies of cigarette smoking and lupus which suggest that the risk is modestly increased among current cigarette smokers, as compared with patients who have never smoked. In one study, current smoking was associated with a four-fold increased risk of positive dsDNA antibodies among lupus patients. Smoking may cause DNA damage, promoting the formation of anti-dsDNA antibodies, with a risk that reduces after cessation of smoking.
Alcohol consumptionConflicting results have been produced with some studies suggesting a protective effect of increased alcohol consumption and others the opposite and yet more suggesting no association at all.These results may well be explained by differences in study design, selection bias, recall bias or uncontrolled confounding.
VirologyLupus is recognised as being caused by a complex immune autoactivation state, interaction of genetics, selected autoantibodies and other factors. Although concordance is high among monozygotic twins, it is not 100% so, clearly, environmental factor(s) play a significant role in disease development. Despite significant research over several decades, no consistent relationship has been seen between any one virus, or group of viruses, and the occurrence of lupus. Some work with the Epstein Barr Virus (EBV) have shown an immunological structural relationship between Epstein-Barr nuclear antigen-1 and Ro, suggesting that EBV infection might be required before some patients can develop lupus. Serological studies among children and adults support a possible association and experiments have shown that the blood-born viral load is 10 - 100-fold higher in lupus patients when compared with controls. Although consistent with an association, these studies do not specifically implicate EBV in the aetiology or development of lupus. Other studies have explored the role of Cytomegalovirus (CMV), Herpes Zoster and Human Retroviruses (HIV and HTLV-1) and have shown some results to implicate these viral agents in the aetiology but results are currently inconclusive. It may be that infection with one or more viral agents is necessary, but perhaps not sufficient, to cause lupus and that it is the host response or the variation in viral strains that contributes to the risk of the disease.
ImmunisationsThere are numerous case reports of lupus exacerbation and onset with vaccination against pneumococcus, tetanus, Haemophilus B, hepatitis B and influenza. It is, however, unclear whether this is publication bias or a true association.
DrugsThe phenomenon of drug-induced lupus has been recognised as a lupus-like illness induced by environmental conditions. It has been recognised in association with more than 80 drugs (e.g. minocycline, hydralazine, procainamide, isoniazid, quinidine, chlorpromazine, methyldopa) but behaves differently from SLE.The age of onset is usually older, with men affected equally to women and Caucasians more than other ethnicities. Risk factors have been identified for the development of drug-induced lupus: slow acetylator status, female gender and HLA haplotypes.The prognosis is generally favourable.
Occupational exposuresEpidemiological studies and experimental research have suggested a role for some occupational exposures in lupus. Exposure to silica, solvents, metals and pesticides have all been studied. Silica has been shown to result in increased production of pro-inflammatory cytokines in animal studies and there is a strong body of epidemiological research suggesting increased risk of lupus among those exposed to silica. Work on solvents e.g. aromatic amines, pesticides and metals have not produced clear evidence for or against an association. The relevance of chemicals and drugs to the aetiology of most cases of lupus remains unknown.
Dr Karen Walker-Bone
Senior Lecturer (Honorary Consultant) in Rheumatology)
Brighton & Sussex Medical School
University of Sussex
East Sussex, BN1 9PX