Presented in part at the 104th Annual Meeting of the American Gastroenterological Association, Orlando, Florida, May 17–22, 2003 (Gastroenterology 2003;124(4 Suppl 1):A36.)
Update on the incidence and prevalence of Crohn's disease and ulcerative colitis in Olmsted County, Minnesota, 1940–2000†
Article first published online: 19 DEC 2006
Copyright © 2006 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 13, Issue 3, pages 254–261, March 2007
How to Cite
Loftus, C. G., Loftus, E. V., Harmsen, W. S., Zinsmeister, A. R., Tremaine, W. J., Melton, L. J. and Sandborn, W. J. (2007), Update on the incidence and prevalence of Crohn's disease and ulcerative colitis in Olmsted County, Minnesota, 1940–2000. Inflamm Bowel Dis, 13: 254–261. doi: 10.1002/ibd.20029
- Issue published online: 16 FEB 2007
- Article first published online: 19 DEC 2006
- Manuscript Accepted: 18 SEP 2006
- Manuscript Received: 11 SEP 2006
- Mayo Foundation for Medical Education and Research
- National Institutes of Health. Grant Number: AR30582
- Crohn's disease;
- ulcerative colitis;
Background: We previously reported that the prevalence of Crohn's disease (CD) and ulcerative colitis (UC) in Olmsted County, Minnesota, had risen significantly between 1940 and 1993. We sought to update the incidence and prevalence of these conditions in our region through 2000.
Methods: The Rochester Epidemiology Project allows population-based studies of disease in county residents. CD and UC were defined by previously used criteria. County residents newly diagnosed between 1990 and 2000 were identified as incidence cases, and persons with these conditions alive and residing in the county on January 1, 2001, were identified as prevalence cases. All rates were adjusted to 2000 US Census figures for whites.
Results: In 1990–2000 the adjusted annual incidence rates for UC and CD were 8.8 cases per 100,000 (95% confidence interval [CI], 7.2–10.5) and 7.9 per 100,000 (95% CI, 6.3–9.5), respectively, not significantly different from rates observed in 1970–1979. On January 1, 2001, there were 220 residents with CD, for an adjusted prevalence of 174 per 100,000 (95% CI, 151–197), and 269 residents with UC, for an adjusted prevalence of 214 per 100,000 (95% CI, 188–240).
Conclusion: Although incidence rates of CD and UC increased after 1940, they have remained stable over the past 30 years. Since 1991 the prevalence of UC decreased by 7%, and the prevalence of CD increased about 31%. Extrapolating these figures to US Census data, there were ≈1.1 million people with inflammatory bowel disease in the US in 2000.
(Inflamm Bowel Dis 2007)
Although investigators have made significant strides toward a better understanding of the pathophysiology of inflammatory bowel disease (IBD) in recent years, this group of conditions remains idiopathic.1 Genetic, environmental, and immunological mechanisms of etiopathogenesis continue to be explored.1 In addition, epidemiological studies are often performed in patients with chronic conditions such as Crohn's disease (CD) and ulcerative colitis (UC) to provide important information about the natural history, health care burden, and causal mechanisms of the disease. In particular, many centers have described increasing incidence rates of CD2–17 and UC16–35 over the past 5 decades. More recently, however, the incidence of CD36 and UC37 appears to be stabilizing. Despite this, CD and UC continue to grow more prevalent as a result of the early age of onset and low mortality (albeit substantial morbidity) of these conditions.2, 3, 6, 7, 12–14, 16, 18, 20, 23, 26, 35–39 It has been estimated that there may be as many as 1.3 million persons in the US and Canada with IBD.40
In contrast to studies performed on patients seen at referral centers, population-based epidemiological studies are more likely to reflect the true spectrum of illness.41 In the US the nonunified structure of health care delivery systems makes such studies difficult. It is particularly difficult to identify all cases of a given disease in a defined geographic area, especially if such patients do not routinely require hospitalization for the disease or its complications. In Olmsted County, Minnesota, however, the small number of institutions providing health care share a linked diagnostic index, allowing for identification of all recognized cases of a particular disease, thus providing a suitable framework in which population-based studies on IBD can be performed.36, 37
A cohort of Olmsted County residents diagnosed with IBD between 1935 and 1993 has been studied at several time points.3, 6, 18, 36–38 Between the 1950s and 1970s, the incidence of CD in Olmsted County increased rapidly,3 but stabilized thereafter at ≈7 cases per 100,000 person-years.36 The prevalence of CD in Olmsted County rose from 91 cases per 100,000 persons in 19836 to 144 cases per 100,000 persons in 1991.36 Likewise, the incidence of UC also increased rapidly in the postwar period,18 peaking in the early 1970s, with a stabilization in incidence thereafter.37 The prevalence of UC in Rochester, Minnesota (the urban center of Olmsted County) increased from 117 cases per 100,000 in 196518 to 268 cases per 100,000 in 1991.37 Olmsted County investigators have also reported that the incidence rate of CD was greater in women6 and in the urban part of the county,3, 6 while a male predominance of incident cases has been noted in patients with UC.18, 37, 38 Whether the rise in incidence of CD and UC represented a worsening environmental risk profile or improved diagnostic techniques and case ascertainment remains unclear.
Continued study of the epidemiology of CD and UC may shed light on the etiopathogenesis of these conditions. In addition, an updated population-based inception cohort of IBD patients may serve as a platform for further detailed epidemiologic studies and also provide valuable information with regard to disease burden in a defined population. In the current study the Olmsted County inception cohort of patients with CD and UC was updated through 2000; previously identified cases were reviewed to assure consistent diagnostic criteria; and the prevalence of CD and UC was calculated on January 1, 2001. Temporal trends in the incidence of CD and UC were examined for the entire period, 1940–2000.
PATIENTS AND METHODS
Olmsted County is located in southeastern Minnesota and had a population of ≈124,000 residents in the 2000 US Census. Rochester is the urban center of Olmsted County, comprised of ≈86,000 residents in 2000. The remainder of the county is predominantly rural. In 2000, 90% of the population of Olmsted County was white and 4% of the population was African American. Although 25% of Olmsted County residents are employed in health care services (8% nationwide), and the level of education is somewhat higher, with 30% having completed college (21% nationwide), residents of the county are otherwise similar to the US white population from the socioeconomic standpoint.41
Rochester Epidemiology Project
The Rochester Epidemiology Project is a unique medical records-linkage system that encompasses the care delivered to residents of Rochester and Olmsted County, Minnesota.41 This records-linkage system exploits the fact that virtually all health care for the local residents is provided either by Mayo Medical Center, comprised of the Mayo Clinic and its two affiliated hospitals, or Olmsted Medical Center, comprised of a smaller subspecialty group and its affiliated hospital. Each year, over half of the Olmsted County population is examined at one of the Mayo facilities, and during any given 3-year period greater than 90% of local residents are examined at either one of the two health care systems.41 The central diagnostic index of the Rochester Epidemiology Project comprises all diagnoses generated from outpatient evaluations, hospitalizations, emergency room evaluations, nursing home visits, surgical procedures, autopsy reports, and death certificates. It is therefore possible to identify all cases of a disease for which patients sought medical attention over a particular period of time.41 Through the resources of this medical records linkage system, we identified all permanent residents of Olmsted County who were diagnosed with CD or UC between January 1, 1990, and December 31, 2000.
This study was approved by the Institutional Review Boards of Mayo Clinic and Olmsted Medical Center. All potential new cases of CD and UC were identified through the central diagnostic index. A diagnosis of CD was confirmed if at least two of the following criteria were satisfied, on two occasions, separated by at least 2 months: 1) clinical history of abdominal pain, diarrhea, weight loss, malaise, and/or rectal bleeding; 2) endoscopic findings of linear ulceration, mucosal cobblestoning, skip areas, or perianal disease; 3) radiologic findings of fistula, stricture, mucosal cobblestoning, or ulceration; 4) laparotomy appearance of “creeping fat,” bowel wall induration, and mesenteric lymphadenopathy; or 5) histologic findings of transmural inflammation and/or epithelioid granulomas. These criteria were identical to those used in two previous studies of CD in Olmsted County.6, 36 A diagnosis of UC was made if a potential case satisfied the following criteria, on two occasions, separated by at least 6 months: 1) diffusely granular or friable colonic mucosa on endoscopy; and 2) continuous mucosal involvement as observed by endoscopy or barium studies. These criteria were identical to those used in two previous studies of UC in Olmsted County.37, 38
The medical records of all Olmsted County residents in previous studies of CD3, 6, 36 or UC18, 37, 38 (1940–1993) were reexamined to confirm accurate diagnoses. The date of onset of symptoms, date of diagnosis, residency on January 1, 2001, date of last follow-up, and vital status at last follow-up were recorded.
Incidence and Prevalence Calculations
Annual incidence rates for UC and, separately, CD, were estimated by dividing the number of events by the population at risk, with direct standardization to the age and sex distribution of the 2000 US white population. The population at risk was estimated by linear interpolation of decennial census data for Olmsted County assuming the entire population of the county to be at risk.42 The 95% confidence intervals (95% CI) were estimated assuming a Poisson distribution for the number of age, sex, and calendar year-specific cases. A Poisson regression analysis was used to assess the association of age, gender, and calendar year with the crude incidence rates. Prevalence rates as of January 1, 2001, were estimated in the same manner as for incidence rates.
For the entire period, 1940–2000, 308 Olmsted County residents were diagnosed with CD and 372 with UC. One hundred fifty-seven CD patients were female (51%), while 211 of those diagnosed with UC were male (57%). The diagnosis of both CD and UC was most common in the third and fourth decades of life. The median age at diagnosis of CD was 29 years (range, 4–91), and the median age at diagnosis of UC was 33 years (range, 1–88). The median age at diagnosis of Crohn's appeared to decrease until 1970–1979 (26 years) and then increased slightly (median, 34 years) in 1990–2000. In UC, the median age decreased until 1970–1979 (31 years) and increased as well (median, 35 years) in 1990–2000. Among the CD patients, 89 had small bowel involvement (29%), 91 had colonic involvement (30%), and 128 had both small bowel and colonic involvement (41%). Among the UC patients, 63 had proctitis (17%), 134 had left-sided colitis (36%), and 174 had extensive colitis (47%).
|Age Group, Years||1940–49||1950–59||1960–69||1970–79||1980–89||1990–2000|
|Subtotal, age-adjusted||3.2 (0.7, 5.7)||4.7 (2.1, 7.4)||7.0 (4.2, 9.8)||9.0 (6.0, 11.9)||6.1 (3.9 8.3)||7.0 (5.0, 9.0)|
|Subtotal, age-adjusted||3.0 (0.3, 5.7)||3.6 (1.2, 6.0)||12.0 (7.9, 16.1)||11.2 (7.7, 14.7)||11.9 (8.6, 15.3)||10.8 (8.1, 13.6)|
|Total, age- and sex-adjusted||3.1 (1.3, 4.9)||4.2 (2.4, 6.0)||9.4 (7.0, 11.9)||10.1 (7.8, 12.4)||8.9 (7.0, 10.9)||8.8 (7.2, 10.5)|
|Age Group, Years||1940–49||1950–59||1960–69||1970–79||1980–89||1990–2000|
|Subtotal, age-adjusted||1.0 (0.0, 2.2)||3.2 (0.9, 5.4)||6.5 (3.8, 9.3)||9.1 (6.3, 11.9)||5.6 (3.6, 7.5)||7.2 (5.2, 9.3)|
|Subtotal, age-adjusted||3.8 (0.7, 6.8)||1.9 (0.0, 3.9)||6.4 (3.6, 9.3)||6.6 (4.0, 9.2)||8.1 (5.5, 10.8)||8.7 (6.3, 11.2)|
|Total age- and sex-adjusted||2.3 (0.8, 3.9)||2.6 (1.1, 4.1)||6.5 (4.5, 8.5)||7.9 (6.0, 9.8)||6.8 (5.2, 8.4)||7.9 (6.3, 9.5)|
There was a significant increase in the incidence of UC over the entire study period (P < 0.0001) (Figure 1). The rates in the first 2 decades, 3.1 and 4.2 cases per 100,000 person-years, were significantly lower than the annual incidence in 1990–2000 (8.8 per 100,000). The incidence rates in decades from the 1960s forward were relatively stable (Table 1, Figure 1). The overall age-adjusted annual incidence among males (9.8 per 100,000; 95% CI, 8.4–11.2), was significantly higher than among females (6.5 per 100,000; 95% CI, 5.5–7.6) (Table 3, Figure 2). Incidence rates among males and females were similar for the first 2 decades, but the rates for males were consistently higher from the 1960s forward. A significant association with age was also observed (P < 0.0001), as the highest incidence rate (13.9 per 100,000 person-years) was seen in the 20–39-year-old group, and the lowest incidence was observed among those individuals 0–19 years of age (2.4 per 100,000 person-years) (Table 3). No significant interactions among gender, age group, and decade of diagnosis were detected.
|Age Group, Years||Males||Females||Total|
Over the entire study period, 1940–2000, there was a significant increase in the incidence of CD (P < 0.0001) (Figure 3), although the association of incidence rates with time period depended on age group. Similarly, the association of age group with incidence depended on the decade of CD diagnosis (P = 0.03 for the interaction effect in the Poisson model). The 20–39 year age group typically had the highest rates (exceptions were males in the decades of 1940–1949 and 1990–2000), from a low of 0.0 per 100,000 person-years among males in the decade of 1940–1949 to a high of 18.3 among females in the decade of 1970–1979 (Table 2). There was no significant overall association of CD incidence with gender (P = 0.59) (Table 4, Figure 4). The age-adjusted rate for males was 6.7 per 100,000 person-years (95% CI, 5.6–7.8) compared with 6.1 per 100,000 (95% CI, 5.1–7.1) for females.
|Age Group, Years||Males||Females||Total|
On January 1, 2001, there were 220 Olmsted County residents alive with a diagnosis of CD and 269 residents with a diagnosis of UC. The age- and gender-adjusted prevalence rate was 214 cases per 100,000 (95% CI, 188–240) for UC (Table 5) and 174 cases per 100,000 (95% CI, 151–197) for CD (Table 6). The age- and gender-adjusted prevalence rate of CD in 2001 was 31% higher than that measured in 1991 (133 cases per 100,000; 95% CI, 111–155),36 while the prevalence rate of UC changed little from that measure in 1991 (229 cases per 100,000; 95% CI, 198–260).36
|Age Group, years||Males||Females||Total|
|Age- and sex-adjusted (95% CI)||243.5 (188.2–239.6)||184.6 (151.4–217.8)||213.9 (188.2–239.6)|
|Age Group, Years||Males||Females||Total|
|Age- and sex-adjusted (95% CI)||179.7 (145.8–213.7)||169.4 (137.5–201.2)||173.8 (150.7–196.9)|
The incidence of both CD and UC increased dramatically in Olmsted County, Minnesota, between 1940 and the early 1970s. Since then the rates have stabilized at ≈8 cases per 100,000 person-years for CD and 9 cases per 100,000 person-years for UC. The prevalence of CD increased by 31% between 1991 and 2001. In comparison, during the same time period the prevalence rate of UC decreased by ≈7%. Male gender and age and decade of diagnosis were associated with changing rates of UC, while age and decade of diagnosis were associated with the incidence of CD.
The main strength of this study is that ascertainment of IBD cases within a well-defined geographic area was close to 100%. The complete (inpatient and outpatient) community medical record for all potential cases was available for review, and strict yet consistent diagnostic criteria were applied. In addition, we were able to build on previous epidemiologic reports of CD3, 6, 36 and UC18, 37, 38 in Olmsted County, extending the period of study to 61 years, and allowing further analysis of temporal trends in the incidence and prevalence of these conditions. Therefore, this study provides true population-based information on the frequency of CD and UC in a small region of the US.
The corresponding limitations of this study include the fact that the population size of Olmsted County restricted the study to a relatively small number of cases. In addition, as has been seen in previous epidemiologic reports of this particular population, only limited follow-up was available in some cases, due to incidence cases having moved out of the county after diagnosis. Finally, our results may not necessarily be representative of the remainder of the US as a result of racial and ethnic differences.41 As of 2000, ≈89% of the residents of Olmsted County were non-Hispanic white, compared to 69% for the rest of the US. These differences in racial and ethnic distribution may be particularly important when extrapolating our CD incidence rates to the US as a whole, since it is believed that the incidence rate of CD may be significantly lower among Hispanics and Asian Americans.43
The adjusted annual incidence rates of CD and UC for the period 1990–2000 were 7.9 cases per 100,000 and 8.8 per 100,000, respectively. While these rates are historically high, even higher rates have been described from other centers in both North America and Europe during recent decades. The incidence of CD was reported as 10 cases per 100,000 in Alberta in 1981,7 while the incidence of UC in Northern California was reported as 10.9 per 100,000.44 When compared with crude rates obtained in 1991–1993 by the European Collaborative study of Inflammatory Bowel Disease (EC-IBD),45 the incidence rate of CD in Olmsted County was higher, while the incidence rate of UC was similar to the mean rate reported by the European centers. The incidence rates of CD and UC in the present study may be compatible with previously reported observations of a north–south gradient of inflammatory bowel disease within the US.46, 47
The incidence of CD and UC reported in recent decades is much higher than rates reported in the 1940s and 1950s. Although intuitively one could attribute the rise to the introduction of better diagnostic techniques for these conditions, we do not believe that this factor alone explains the precipitous rise in incidence. The two major modalities used to diagnose IBD over the past 60 years have been barium radiography and colonoscopy. Barium radiography and proctoscopy were widely used in Olmsted County prior to the rapid rise in incidence, while colonoscopy became available after this rise had taken place. In addition, one might expect a shorter interval between symptom onset and diagnosis if the rise in incidence were due to increased healthcare-seeking or better diagnostic techniques; while our group has previously described such a trend in the patients with UC,37 this has not been confirmed in patients with CD.36
In patients with UC, the distribution of age at diagnosis differed between the sexes, with a persistently higher incidence rate noted among men in the later decades of life (Fig. 2). In comparison, the distribution of age at diagnosis in CD was more equally distributed between females and males (Figure 4). Peak incidence for both CD and UC was noted in the third decade of life. While the incidence of UC in men demonstrated a bimodal distribution, this was not observed in women with UC or in patients with CD.
The adjusted prevalence of CD on January 1, 2001 (174 cases per 100,000 persons) is one of the highest reported and has increased by 31% over the past 10 years. While the adjusted prevalence of UC remains high (214 cases per 100,000 persons), this figure has decreased by 7% over the same 10-year period. The rising prevalence of CD is likely a reflection of prolonged survival due to the availability of more advanced medical therapies. If the results of this study are extrapolated to the US white population of 2000 (≈220 million), there may have been as many as 383,000 patients with CD and 471,000 patients with UC among US whites at the beginning of the 21st century. These figures are likely an underestimation of all cases in the entire US population since African Americans, who numbered ≈35 million in the 2000 census, may have an incidence approaching that of US whites.43, 44 It has previously been projected that the prevalence of CD in Olmsted County should eventually stabilize at roughly 250 cases per 100,000 persons.36 As the number of patients with IBD in the US continues to grow, so too will healthcare costs associated with this population. This must remain a concern for those caring for patients with IBD in this era of continually rising healthcare costs.
We thank Debra A. Jewell, RN, for assistance in data abstraction.
- 29Incidence of ulcerative colitis and indeterminate colitis in four counties of southeastern Norway, 1990–93. A prospective population-based study. The Inflammatory Bowel South-Eastern Norway (IBSEN) Study Group of Gastroenterologists. Scand J Gastroenterol. 1996; 31: 362–366., , , et al.
- 42Calculating incidence, prevalence, and mortality rates for Olmsted County, Minnesota residents: an update. Technical Report Series No 49. Rochester, Minnesota: Section of Biostatistics, Mayo Clinic; 1992., , , et al.