Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
A recent and enigmatic epidemiologic observation is that persons who have had an appendectomy are less likely to develop ulcerative colitis (UC).1–17 This finding has been reproduced in nearly all studies that have addressed the question, but the mechanisms underlying the association remain unknown.
It has been suggested that the association might be explained by an unknown factor that both predisposes to appendectomy and protects against ulcerative colitis—possibly a genetic trait or a factor related to childhood hygiene.14,18 This mechanism would imply that an appendectomy would have no effect on a person's risk of developing ulcerative colitis. If such a factor existed, we would observe a low incidence of ulcerative colitis after appendectomy, as reported. However, there would also be a similarly low incidence of ulcerative colitis before appendectomy, because those who eventually have an appendectomy are protected against ulcerative colitis by the same traits that promote the appendectomy. Thus, the relative ulcerative colitis incidence would be symmetrically reduced before and after appendectomy.
Another proposition is that the appendix itself may promote the development of ulcerative colitis, perhaps by some antigenic processing of luminal bacteria.19,20 If this were true, there would be a low incidence of ulcerative colitis after appendectomy, as described previously, but the incidence would not be reduced before appendectomy. Our main study objective was to compare the incidence of ulcerative colitis before and after appendectomy. If it is not symmetric, then the association cannot be explained by genes or other confounders that are stable over time, and appendectomy may confer a genuine protective effect.
The Danish National Registry of Patients
Since 1977, the Danish National Health Board has recorded data on all individual hospital discharges in the Danish National Registry of Patients.21 Each record contains a personal registration number (a unique identifier of all Danish citizens) and selected medical data, including discharge diagnoses and operative procedures. Diagnoses have been recorded by the International Classification of Diseases, 8th edition (ICD-8) from 1977 through 1993 and 10th edition (ICD-10) since 1994. Because virtually all inpatient medical care in Denmark is furnished by the public health authorities, this data resource allows true population-based studies, covering all 5.2 million inhabitants of Denmark.
The Study Population
We included all 234,559 persons who had an appendectomy during the period January 1977 through December 1999. Data on such procedures were extracted from the Danish National Registry of Patients.
To obtain data on incident cases of ulcerative colitis, all patients discharged with a diagnosis of ulcerative colitis (ICD-8, 563.19; ICD-10, K51) or hemorrhagic proctitis (ICD-8, 569.04; ICD-10, K51.2) during the period January 1977 through December 1999 were retrieved from the National Register and were considered for inclusion in the study (n = 21,769). Because we were interested in studying only incident cases, we subsequently excluded those who had their first admission with a recorded ulcerative colitis diagnosis during the years 1977 through 1980 (n = 5482). These were likely to be a mixture of incident and nonincident cases in whom the ulcerative colitis onset could not validly be determined. The appropriateness of this 4-year run-in period was confirmed by inspection of the annual incidence of ulcerative colitis without known prior hospital admission (Fig. 1). This distribution presumably is contaminated in the early years by “recurrent” admissions for ulcerative colitis, in which the previous admissions occurred before the start of the follow-up period. The figure shows a descending segment of approximately 3 years, indicating that readmissions would usually occur within 3 years and that first admissions after a run-in period of 3 years or more would represent incident admissions.22
We also excluded those who ever had a modifier code of “suspected” or “not confirmed” ulcerative colitis diagnoses (n = 3191) or ever had a diagnosis of Crohn's disease (ICD-8, 563.00–563.09; ICD-10, K50) (n = 1166). The remaining 11,930 patients were 5844 men and 6086 women. For members of this cohort, we defined the ulcerative colitis onset as the date of the first hospitalization with a diagnosis of ulcerative colitis. By definition, these first admissions all occurred during the period January 1981 through December 1999. The median age at first admission among this cohort was 43.8 years.
For the purpose of censoring, we retrieved data from the Danish Central Person Registry on the date of death, emigration, or immigration for all subjects who had had an appendectomy. We also obtained data on all codings of colectomy for the cohort during the period January 1981 through December 1999 from the Danish National Registry of Patients. Data on all diagnoses, surgical procedures, migrations, and deaths were linked by use of the personal identification numbers.
The study was approved by the Danish Data Protection Agency.
We analyzed the data as a cohort study in which most subjects accrued both preappendectomy and postappendectomy person-time at risk. Follow up began January 1, 1981, after which most “first” cases of ulcerative colitis were indeed incident. Subjects contributed with person-time from the start of the follow-up period, their immigration, or their birth, whichever event was latest. Follow up ended when the person died, had a diagnosis of ulcerative colitis, a colectomy, or emigrated from Denmark, or the end of the study period (December 1999), whichever event came first. By these definitions, persons who had their first ulcerative colitis diagnosis during the period 1977 through 1980 did not contribute any person-time, and persons with an appendectomy during this period contributed all their person-time (1981–1999) as “postappendectomy.” Mean follow up was 17.5 years (standard deviation [SD], 3.8 years). All person-time was categorized by age in 10-year strata, sex, calendar year, and appendectomy status (pre- vs. postappendectomy).
We calculated the age- and sex-standardized incidence rate of ulcerative colitis for person-time before and after appendectomy using indirect standardization. The standard population was the midperiod age and sex distribution of the Danish population in 1990. The expected incidence rate was calculated simply as the average crude incidence rate of ulcerative colitis in Denmark during the period 1981 through 1999. The main results were expressed as the ratio of the observed standardized incidence rate to the expected incidence rate or the standardized incidence rate ratio (SIR). Confidence intervals for the standardized incidences were calculated under the assumption that the observations conformed to a Poisson distribution. For the ratio of post- to preappendectomy SIR, we used the methods described by Ederer and Mantel.23 Twenty-seven persons had their first diagnosis of ulcerative colitis at the admission during which the appendectomy was performed. These were not counted in either of the estimates of ulcerative colitis incidence rates.
To confirm our analyses, we also performed a direct standardization in which the age- and sex-specific incidence rates of the preappendectomy period were transferred to the age and sex distribution of the postappendectomy person-time, thereby calculating an expected number of incident ulcerative colitis cases after appendectomy. Finally, we performed a multivariate Poisson regression. The model included age, calendar year, sex, and pre- and postappendectomy status. Categorization in calendar year and age was performed posthoc, after some exploratory analyses of the age- and calendar-year distribution of appendectomies in the Danish population.
We identified 234,559 persons who had an appendectomy during the period 1977 through 1999, corresponding to an average incidence rate of 239.0 per 100,000 person-years. Persons were followed 1981–1999 for a mean of 6.3 years (SD, 5.1 years) before the appendectomy and 11.3 years (SD, 6.3 years) after, thus contributing 1,466,938 and 2,643,572 person-years of observation, respectfully. Fifty-eight percent of the population was female, and the median age at appendectomy was 25.8 years. The age distribution of the subjects is shown in Table 1.
There were 202 incident cases of ulcerative colitis in the preappendectomy period, yielding a crude incidence rate of 13.8 per 100,000 person-years. The corresponding figures for the postappendectomy period were 330 cases and an incidence rate of 12.5 per 100,000 person-years. The standardized incidence rates of ulcerative colitis before and after appendectomy were 17.6 and 13.0 per 100,000 person-years, respectively. The ratio of the standardized ulcerative colitis incidence rate after appendectomy to the analogous rate before appendectomy was 0.74 (95% confidence interval [CI] = 0.62–0.88). We also estimated this ratio in a number of subgroups. All point estimates except one (appendectomy later than 1990) were below unity (Table 2). A particularly low SIR was observed for subjects below 15 years of age at start of follow up (0.13; 0.09–0.19).
Using direct standardization, we arrived at 496 expected ulcerative colitis cases in the postappendectomy period. Compared with the 330 observed cases, this yields a standardized incidence rate ratio of 0.67 (95% CI = 0.56–0.79).
The average incidence rate of ulcerative colitis for the entire Danish population during the period 1981 through 1999 was 12.2 per 100,000 person-years. Comparing the study population to the general population, the standardized incidence rate ratio of ulcerative colitis was 1.44 (95% CI = 1.26–1.66) before appendectomy and 1.06 (95% CI = 0.95–1.19) after.
We performed a supplementary analysis with the end point restricted to a diagnosis of ulcerative colitis, excluding all subjects who had ever had a diagnosis of hemorrhagic proctitis. Only 399 such cases emerged, of which 128 and 251 occurred before and after the appendectomy, and 20 cases were diagnosed at the admission during which the appendectomy was performed. The standardized incidence rates of ulcerative colitis after and before appendectomy were 9.7 and 11.0 per 100,000 person-years, respectively, with a ratio of 0.88 (0.72–1.10).
In the multivariate Poisson regression analysis (Table 3), postappendectomy status was a marker that suggests a lower risk of ulcerative colitis, with an incidence rate ratio of 0.69 (0.57–0.85). A secondary finding was an increasing incidence of ulcerative colitis with age.
The main finding of this study is that the incidence of ulcerative colitis is lower after appendectomy than before. Therefore, the known inverse association between appendectomy and ulcerative colitis cannot be explained by a mutual genetic determinant, childhood hygiene, or some other factor that is stable over time.19 Although this points toward the possibility of lowering a person's risk of ulcerative colitis by performing an appendectomy, the clinical implications are questionable. Even first-degree relatives of ulcerative colitis patients have only a 1% to 5% prevalence of ulcerative colitis.24–26 Thus, it would be difficult to identify subjects with a risk of developing ulcerative colitis that by itself would justify an appendectomy.
Our findings are unlikely to be explained by selection bias. We used a true population-based data resource, and our data on age- and sex-specific appendectomy or ulcerative colitis incidence rates correspond well with other estimates.27–34 It is unlikely that knowledge of the ulcerative colitis–appendectomy association would affect the decision to refer a patient with suspected appendicitis or ulcerative colitis. Furthermore, the inverse association was not widely publicized until late in the study period.
There are some sources of information bias in our study that may affect the result. We did not have data on outpatient care. Many ulcerative colitis patients have a long outpatient course before they are hospitalized for the first time. This clinical pattern implies that our first ulcerative colitis diagnosis is delayed relative to the true onset of disease, and, furthermore, that some of the ulcerative colitis cases with apparent onset after the appendectomy did in fact begin before it. In the present context, this bias is conservative and tends to lower the contrast between pre- and postappendectomy incidence rate of ulcerative colitis. We conducted supplementary analyses restricted to ulcerative colitis cases only, without hemorrhagic proctitis diagnoses, presuming that ulcerative colitis without hemorrhagic proctitis—being a more troublesome condition—would have less delay between onset of disease and first admission. We also shifted the timeframe to mimic diagnostic delay (not shown). Neither of these had any substantial effect on the symmetry. The strong asymmetry among subjects who were under 15 years of age at start of follow up may indicate less delay in first diagnoses in this group. Ulcerative colitis rarely has its clinical debut before this age, and new cases may be more readily admitted.
Another potential limitation is possible miscoding of discharge diagnoses, a well-known problem in registry-based studies. If the magnitude of miscoding is constant over time, our comparison of pre- and postappendectomy incidence rate of ulcerative colitis will remain valid. In addition, a validation study has shown a fairly high validity of coding for inflammatory bowel disease in our setting.35
The finding of nonsymmetric incidence of ulcerative colitis is unlikely to be explained by confounding. Apart from age and appendectomy, the only known determinants of ulcerative colitis development of major importance are smoking, which is inversely associated,36 and genetic disposition toward ulcerative colitis.24–26 Neither of these are known to be associated with appendicitis and would hardly produce a nonsymmetric incidence rate of ulcerative colitis around the time of appendectomy. The symmetry design used here effectively controls confounders that are stable over time.37
The existing studies have shown without reasonable doubt that the incidence rate of ulcerative colitis after appendectomy is lower than in the background population. With only one other exception,38 all studies have found an inverse association between appendectomy and ulcerative colitis. Most studies1–6,8,9,14 reported a strong association, with relative risks as low as 0.02.1 One other study based on a smaller dataset from the Danish National Registry of Patients reported a relative risk of 0.87 (95% CI = 0.69–1.07).17 In our study, the postappendectomy SIR (comparing the rate in those who had an appendectomy with that of the background population) was 1.06 (95% CI = 0.95–1.19). One of the major limitations of the study is the short follow up, on average only 11.3 years after appendectomy. It has been reported that the inverse association between appendectomy and ulcerative colitis is strongest for appendectomies performed in childhood.2 Possibly, we would have seen a low postappendectomy incidence rate of ulcerative colitis with a longer follow up. As shown in Table 1, the strongest asymmetry was found among those who had their appendectomy early in the study period and who were under 15 years of age at the start of follow up. It has been reported that the inverse appendectomy–ulcerative colitis association is confined to subjects who have appendicitis at appendectomy.16 This was also reflected by a strong asymmetry in the subgroup analysis.
The mechanism of the inverse association between appendectomy and ulcerative colitis is not established. Some intriguing results have emerged from a study by Mizoguchi et al.19 of T-cell receptor-alpha mutant mice who spontaneously develop colitis resembling ulcerative colitis. Mice whose appendices are removed at 3–5 weeks of age have a marked reduction of mesenteric lymph nodes at 6–7 months compared with sham-operated mice. In addition, only 3% of the mice whose appendices were removed developed colitis, in contrast to 80% among the sham-operated mice. In another study of transgenic rats, also with early spontaneous colitis resembling ulcerative colitis, the cecum was excluded from the fecal stream in some rodents, whereas others had a sham operation. The bacterial load in the cecal loop strongly influenced the development of colitis in remote regions, and this development could be manipulated with antibiotics. The investigators suggested that antigen processing lymphoid tissue in the appendix might mediate disease that is related to gut bacteria.20
We conclude that the inverse association between appendectomy and ulcerative colitis cannot be explained by confounders that are stable over time, and that a genuine prophylactic effect of appendectomy against ulcerative colitis cannot be dismissed.
We are indebted to John Baron for revising the paper.
1. Rutgeerts P, D'Haens G, Hiele M, et al. Appendectomy protects against ulcerative colitis. Gastroenterology
2. Duggan AE, Usmani I, Neal KR, et al. Appendicectomy, childhood hygiene, Helicobacter pylori
status, and risk of inflammatory bowel disease: a case control study. Gut
3. Gilat T, Hacohen D, Lilos P, et al. Childhood factors in ulcerative colitis and Crohn's disease. An international cooperative study. Scand J Gastoenterol
4. Gent AE, Hellier MD, Grace RH, et al. Inflammatory bowel disease and domestic hygiene in infancy. Lancet
5. Marion JF, Bodian CA, Janowitz HD. Appendectomy, appendicitis, and inflammatory bowel disease. Gastroenterology
6. Smithson JE, Radford-Smith G, Jewell DP. Appendectomy, tonsillectomy and inflammatory bowel disease. Gut
7. Russel MG, Dorant E, Brummer R-JM, et al. Appendectomy and the risk of developing ulcerative colitis or Crohn's disease: results of a large case-control study. Gastroenterology
8. Parello T, Pavia M, Angelillo I, et al. Appendectomy is an independent protective factor for ulcerative colitis (UC); results of a multicenter case-control study. Gastroenterology
9. Tasiopoulos J, Kavin H, Goldman J, et al. Appendectomy is protective against the development of ulcerative colitis. Gastroenterology
10. Derby LE, Jick H. Appendectomy protects against ulcerative colitis. Epidemiology
11. Kubba AK, Price RF, Smith G, et al. Appendicectomy and ulcerative colitis. J R Coll Surg Edinb
12. Scott IS, Sheaff M, Coumbe A, et al. Appendiceal inflammation in ulcerative colitis. Histopathology
13. Minocha A, Raczkowski CA. Role of appendectomy and tonsillectomy in the pathogenesis of ulcerative colitis. Dig Dis Sci
14. Wurzelmann JI, Lyles CM, Sandler RS. Childhood infections and the risk of inflammatory bowel disease. Dig Dis Sci
15. Breslin NP, McDonell C, O'Morain C. Surgical and smoking history in inflammatory bowel disease: a case-control study. Inflamm Bowel Dis
16. Andersson RE, Olaison G, Tysk C, et al. Appendectomy and protection against ulcerative colitis. N Engl J Med
17. Frisch M, Johansen C, Mellemkjær L, et al. Appendectomy and subsequent risk of inflammatory bowel disease. Surgery
18. Ekbom A. Appendicectomy and childhood hygiene: different sides of the same coin? Gut
19. Mizoguchi A, Mizoguchi E, Chiba C, et al. Role of appendix in the development of inflammatory bowel disease in TCR-mutant mice. J Exp Med
20. Rath HC, Ikeda JS, Linde H-J, et al. Varying cecal bacterial loads influences colitis and gastritis in HLA-B27 transgenic rats. Gastroenterology
21. Andersen TF, Madsen M, Jorgensen J, et al. The Danish National Hospital Register. A valuable source of data for modern health sciences. Dan Med Bull
22. Hallas J, Gaist D, Bjerrum L. The waiting-time distribution for prescription drugs. Epidemiology
23. Ederer F, Mantel N. Confidence limits for the ratio of two Poisson variables. Am J Epidemiol
24. Orholm M, Munkholm P, Langholz E, et al. Familial occurrence of inflammatory bowel disease. N Engl J Med
25. Probert CSJ, Jayanthi V, Hughes AO, et al. Prevalence and family risk of ulcerative colitis and Crohn's disease: an epidemiological study among Europeans and south Asians in Leicestershire. Gut
26. Orholm M, Fonager K, Sørensen HT. Risk of ulcerative colitis and Crohn's disease among offspring of patients with chronic inflammatory bowel disease. Am J Gastroenterol
27. Primatesta P, Goldacre MJ. Appendectomy for acute appendicitis and for other conditions: an epidemiological study. Int J Epidemiol
28. Addiss DG, Shaffer N, Fowler BS, et al. The epidemiology of appendicitis and appendectomy in the United States. Am J Epidemiol
29. Blomqvist P, Ljung H, Nyren O, et al. Appendectomy in Sweden 1989–1993 assessed from the inpatient registry. J Clin Epidemiol
30. Fonager K, Sorensen HT, Olsen J. Change in incidence of Crohn's disease and ulcerative colitis in Denmark. A study based on the National Registry of Patients, 1981–1992. Int J Epidemiol
31. Tysk C, Järnerot G. Ulcerative proctocolitis in Örebro, Sweden. A retrospective epidemiological study, 1963–1987. Scand J Gastroenterol
32. Langholz E, Munkholm P, Nielsen OH, et al. Incidence and prevalence of ulcerative colitis in Copenhagen county from 1962 to 1987. Scand J Gastroenterol
33. Ekbom A, Helmick C, Zack M, et al. The epidemiology of inflammatory bowel disease: a large population-based study in Sweden. Gastroenterology
34. Kildebo S, Nordgaard K, Aaronsen O, et al. The incidence of ulcerative colitis in Northern Norway from 1983–1986. The Northern Norwegian Gastroenterology Society. Scand J Gastroenterol
35. Fonager K, Sorensen HT, Rasmussen SN, et al. Assessment of the diagnoses of Crohn's disease and ulcerative colitis in a Danish hospital information system. Scand J Gastroenterol
36. Koutroubakis I, Manousos ON, Meuwissen SG, et al. Environmental risk factors in inflammatory bowel disease. Hepatogastroenterology
37. Hallas J. Depression provoked by cardiovascular medication. A prescription sequence symmetry analysis. Epidemiology
38. Woods BL, Steinberg EN, Hornung CA, et al. Does appendectomy really protect against ulcerative colitis. Gastroenterology