The Epidemiology of IBD

During the past 50 years, inflammatory bowel disease has become one of the major gastroenterologic problems in the westernized world. During this period, clinicians (gastroenterologists as well as surgeons) and epidemiologists have been fascinated by these disease entities (ie, ulcerative colitis and Crohn's disease) and there is an abundance of descriptive data both cross sectional and over time in different populations. It is, however, somewhat disheartening that these efforts so far have yielded so little knowledge about the underlying biologic mechanisms. The overruling goal of this talk is to provide benchmarks, which can be used to test the different hypotheses that presently are floating around or will emerge in the future. Hopefully, this talk will also provide some ideas on how to proceed in the future and to avoid inventing the wheel once more. There is a strong correlation in the occurrence of ulcerative colitis and Crohn's disease. Areas or populations with a high incidence or mortality attributable to ulcerative colitis also have a high incidence or mortality due to Crohn's disease and vice versa. 1 There are common genetic susceptibilities for both diseases. Relatives to patients with Crohn's disease have an increased risk of ulcerative colitis and vice versa. 2 Although misclassification of either disease could lead to a false correlation, the consistency of these findings in different settings argues convincingly against such a bias. With regards to the temporal trends for ulcerative colitis and Crohn's disease, there is a strikingly consistent pattern. An increasing incidence of ulcerative colitis precedes an increase in Crohn's disease with a time lag of approximately 15 to 20 years. This has been shown in Sweden, Iceland, Denmark, the Faeroe Islands, and the US. In Figure 1, this is illustrated by the incidence figures of ulcerative colitis and Crohn's disease between 1945 and 1983 in Uppsala county, Sweden, where an increase in ulcerative colitis starting in the late 1940s was followed by an increase of Crohn's disease in around 1960. 3,4 The correlation between the diseases, both in incidence figures and genetics, leads us to two different hypotheses: 1. Ulcerative colitis and Crohn's disease represents the opposite ends of a continuous spectrum of one disease entity, but with different clinical characteristics, or 2. There are some shared genetic and/or environmental risk factors for ulcerative colitis and Crohn's disease. Annual incidence of ulcerative colitis and Crohn's disease in Uppsala County, Sweden, 1945–1983. Neither ulcerative colitis nor Crohn's disease are new clinical entities. There were case series of patients with ulcerative colitis published already during the 19th century. A case series published by Kenneth Dalziel in 1913 5 constitutes the first case series of Crohn's disease 20 years before Dr. Burrill B. Crohn's original description. 6 After the Second World War, there was an increase in incidence in both ulcerative colitis and Crohn's disease, especially in the United States, Great Britain, and Scandinavia. In most instances, the incidence figures of ulcerative colitis are higher than for Crohn's disease. In the case of Crohn's disease, the incidence seems to level off around 6.0 per 100,000, and for ulcerative colitis, around 15 to 20 per 100,000, although higher incidence figures have been reported for both disease entities. 3 This corresponds to a lifetime risk for either one of the two diseases between 0.5% and 1%. There are, however, exceptions. Reports from both Belgium and France indicate that in those areas, Crohn's disease is more common than ulcerative colitis. Another consistent pattern in all studies is that increasing incidence will be most pronounced in the age group 20 to 40 years for both ulcerative colitis and Crohn's disease. Some authors also report a second peak in the older age group (60+), but this remains controversial. It has been argued that this represents a delayed diagnosis made when the disease relapses. Another emerging phenomenon is that, since the beginning of the 1990s, there is an increase in juvenile Crohn's disease reported from both the United Kingdom and Sweden. During the 1960s and 1970s both in Europe and northern America, the results from different incidence studies suggested a north-south gradient. 7 The results from a major undertaking in Europe in the early 1990s to test this hypothesis included 20 different European centers 8 indicate, however, that the north–south gradient is a historic phenomenon. Another interesting feature of inflammatory bowel disease is the uniformity in incidence within countries and that these figures seem to follow not natural boundaries but national borders. This is well illustrated in the temporal trends in Malmö and Copenhagen (Figure 2), where it is obvious that in spite of the proximity of the two centers (10 miles of water) the increase in incidence in Crohn's disease occurred much earlier in Malmö compared with Copenhagen. 9,10 The figures in Malmö are similar to other incidence figures in other parts of Sweden. Annual incidences of Crohn's disease in Copenhagen County during the period 1962 to 1987 and Malmö during the period 1958 to 1973. A consistent finding in early observational studies for both ulcerative colitis and Crohn's disease was the association between high socio-economic status and an increased risk. However, recent studies in high incidence areas have failed to demonstrate such an association. 11 In this context, it is worth pointing out that socio-economic status is not a biologic relevant exposure as such. It stands for differences in diet, crowding and hygiene, exposures that change over time in a society. Ethnicity is to some extent a parallel phenomenon. In Crohn's original description, all 14 patients were Jewish. 6 Ever since then, there has been a perceived association between Jewish ethnicity and an increased risk of IBD. This notion has been further substantiated by the results in studies from the US Army, where the risk of ulcerative colitis in 1944 was higher among those with Jewish ethnicity compared with other groups. 12 Analytical studies from other parts of the worlds such as Sweden, United Kingdom, and South Africa have given further credence to the existence of such specific ethnicity factors. However, there are reasons to question these results. In the two Swedish studies, the nominator (the number of subjects with Jewish ethnicity) was assessed through family name, but the denominator was assessed in a different manner. 10 A study from South Africa has similar drawbacks as the denominator was unknown and therefore ascertained from old census data, probably too low. These flaws would generate inflated incidence rates and the results from studies in Israel 13 where the incidence of prevalence rates do not differ from other high incidence areas such as the United Kingdom and Scandinavia, further strengthens question mark to what extent Jewish ethnicity is an independent risk factor. It is therefore not surprising, keeping the importance of socio-economic status in mind, that certain ethnic minorities (especially those with lower socio-economic status) often have a lower incidence compared with the rest of the population. However, these associations are not constant over time. For instance, in Great Britain even an opposite association has been reported in second generation immigrants from India, Pakistan, and Bangladesh. This indicates that environmental factors, not genetics with ethnicity as a proxy, are the major driving force for the increase in incidence over time. Smoking seems to be protective against ulcerative colitis in contrast to Crohn's disease where smoking increases the risk. In the case of ulcerative colitis, this is not a new observation. Since the first report from Australia in the 1950s, 14 there has been an abundance of studies showing ulcerative colitis as a protective factor and, sometimes, former smoking being associated with an increased risk. 15 The underlying biologic mechanism still remains unknown but for Crohn's disease, there is a reasonable biologic model for an increased risk if one assumes that part of the disease process is a multifocal gastrointestinal infarction, which could be aggravated by cigarette smoke. To what extent changes in smoking habits explain the time trends in IBD? In the case of ulcerative colitis, we should keep in mind that the majority of patients are never smokers, thus the introduction of cigarette smoking as a common feature in the population during the 20th century could not explain the emergence of this disease. Almost every new compound introduced in a population during the last 100 years has been proposed to be of importance for the etiology in inflammatory bowel disease. They include toothpaste, chewing gum, fast food, margarine, cornflakes, etc. The common characteristic of studies on such exposures is that they deal with small patient groups with data assembled retrospectively without taking into account methodologic problems such as recall bias and confounding by indication. 16 Oral contraceptive use is, to some extent, an exception where a positive association has been shown with Crohn's disease. 17 However, the attributable fraction is low and the results indicate that oral contraceptive use should be perceived as a promoter, not an initiator. The hypothesis of an early exposure of some agent as of major importance for the risk of inflammatory bowel disease have been around since 1913 when Dalziel proposed Mycobacterium paratuberculosis as the agent for “Crohn's disease.”5 Since then, early gastroenteritis, bottle feeding, and better hygiene during childhood have been implicated as risk factors for ulcerative colitis and Crohn's disease. 18,19 These findings are supported by the shift in incidence from high socio-economic strata in the population to the absence of such an association and the disappearance of a north–south gradient. An obvious line of future research is therefore to follow what happens in low incidence areas such as Asia or Eastern Europe. Already in the 1980s, it was demonstrated that patients with ulcerative colitis were less likely to have been subjected to an appendectomy compared with the background population. 11 There are two mutually exclusive interpretations of these findings. One is that the removal of the appendix is causally linked to a decrease of ulcerative colitis and animal models have given some credence to this. The other is that there is an association between appendicitis and ulcerative colitis (ie, patients who will get appendicitis are less likely to succumb to ulcerative colitis). There are data, especially from Sweden, which support the latter alternative as the authors could not find any protective effect against ulcerative colitis following the removal of a nondiseased appendix. 20 There was a raise in appendicitis in the late 19th century, which seemed to peak in the 1950s followed by a decline in the westernized world. Better hygiene during childhood has been hypothesized to be the underlying cause. One could therefore speculate that appendicitis, ulcerative colitis, and Crohn's disease are part of the same disease spectrum perhaps similar to other disease in which hygiene during childhood is of importance for the future risk such as allergy and asthma. Thus, a better understanding of the underlying causes of appendicitis could be a fruitful way forward. There has been an enormous amount of work being done to understand the underlying reasons for the epidemic in inflammatory bowel disease so far without success. However, the data from these efforts should be used to find ways to proceed. It is not unlikely that we during the next 10 years will have a unique opportunity in well-conducted epidemiologic studies in those populations in which we might expect a change in incidence, and there is also a need to broaden our understanding on the underlying causes of appendicitis, with a special focus on early exposures.