- Email: [email protected]
Crohn's disease: Influence of age at diagnosis on site and clinical type of disease
GASTROENTEROLOGY 1996;111:580–586
Crohn’s Disease: Influence of Age at Diagnosis on Site and Clinical Type of Disease JOSEPH M. POLITO II,*,† BARTON CHILDS,§ E. DAVID MELLITS‡,x,¶ AARON Z. TOKAYER,*,† MARY L. HARRIS,*,† and THEODORE M. BAYLESS*,† *Gastroenterology Division, Department of Medicine, and Departments of §Pediatrics and xNeurology, The Johns Hopkins University School of Medicine; ¶Department of Biostatistics, The Johns Hopkins University School of Hygiene and Public Health; and †Meyerhoff Center for Digestive Diseases and Inflammatory Bowel Disease, The Johns Hopkins Hospital, Baltimore, Maryland
See editorial on page 813. Background & Aims: Crohn’s disease has a bimodal age distribution of disease onset diagnosis. The peaks (20 and 50 years) may represent different phenotypes or different genetic and/or environmental influences between younger- and older-onset individuals. The aim of this study was to examine the influences of age at diagnosis of Crohn’s disease on disease site, type, and course. Methods: Records of 552 consecutive patients with Crohn’s disease were reviewed retrospectively. Results: Younger age at diagnosis (younger than 20 years), compared with an older age (40 years or older), was associated with a greater prevalence of a family history of Crohn’s disease (29.9% vs. 13.6%), greater small bowel involvement (88.7% vs. 57.5%), more stricturing disease (45.8% vs. 28.8%), and a higher frequency of surgery (70.6% vs. 55.3%). Older age at diagnosis was associated with a greater prevalence of colonic disease (84.8% vs. 71.2%) and the inflammatory subtype (54.5% vs. 34.4%). A conditional logistic regression analysis confirmed an independent effect of age at diagnosis on ileal disease and surgery for intractable disease. Conclusions: In Crohn’s disease, early age at diagnosis is associated with more complicated disease and a greater likelihood of having affected relatives. Stratification of Crohn’s disease by age at diagnosis provides support for the concept of genetic heterogeneity.
sent different groups or phenotypes or may indicate different genetic and/or environmental influences between younger-onset patients and older-onset individuals.8,9 The purpose of this study was to examine the association, if any, between age at diagnosis of Crohn’s disease and the disease site and type and family history in 552 consecutive patients. Differences in frequency of a family history of Crohn’s disease; separate sites of disease involvement, i.e., ileal, ileocolonic, colonic, and jejunoileitis; and clinical types, i.e., inflammatory, fistulizing and/or perforating, and stricturing, were compared in groups of patients based on age at diagnosis.
Materials and Methods Patient Population The records of 552 consecutive patients with documented Crohn’s disease examined by two of the investigators (T.M.B. and M.L.H.) at The Johns Hopkins Hospital between January 1985 and December 1991 were reviewed retrospectively. Three hundred patients (54%) were women, and 252 patients (46%) were men. Five hundred forty-two patients were white, 10 were black, and none were Asian. The diagnosis of Crohn’s disease was based on accepted clinical, radiological, and histological criteria.10,11 This is the same population used in our study of the influence of concordance within families on site and clinical type of Crohn’s disease.12 Two patients, both with negative family histories of IBD, who were in the original study were excluded because of lack of data of age at diagnosis or onset.
Family History
C
rohn’s disease is a form of idiopathic inflammatory bowel disease (IBD) with a positive family history in at least one sixth of patients.1 – 3 Having a sibling with Crohn’s disease increases the risk of an individual developing the same illness by a factor of 30 times compared with the general population.4,5 There is a bimodal age distribution of disease onset and diagnosis in population studies of Crohn’s disease with a peak incidence in the third decade and a smaller peak in the eighth decade.6,7 These two peaks may repre/ 5E11$$0008
08-16-96 06:58:23
gasa
The presence or absence of a family history of idiopathic IBD as well as age at diagnosis of disease were determined from the patient and any relatives in attendance by use of a standardized form at the time of the initial visit to The Johns Hopkins Hospital for all patients. Among these 552 consecutive patients with Crohn’s disease, 95 (17%) were con-
‡Deceased. q 1996 by the American Gastroenterological Association 0016-5085/96/$3.00
WBS-Gastro
September 1996
CROHN’S DISEASE: INFLUENCE OF AGE AT DIAGNOSIS 581
firmed to have a positive family history of Crohn’s disease. Family history was confirmed by contacting, with permission, the affected relative and the relative’s physician. Family history status was reconfirmed in all patients at least 1 year before closure of the study. Seven patients reported a family history of both ulcerative colitis and Crohn’s disease. Sixty-eight patients had first-degree relatives with Crohn’s disease. These 95 patients with a family history of Crohn’s disease were similar in sex and distribution and type of disease to the 432 patients without a family history of IBD (Table 1). The sites of disease are the same as in Table 1 of our report on concordance of site and type of disease within families, except that 2 patients without a family history of IBD and with colonic disease of an inflammatory type were excluded for lack of data of age at diagnosis.12 Information on religion and quantitative smoking history was not available in most patients.
Classification All patients were classified at the time of the latest visit by site of disease involvement (small bowel [jejunoileitis, ileum, or ileocecal], colon, and other [usually esophagogastroduodenal]) and by segregating them into one of three subtypes of disease (inflammatory, stricturing, and fistulizing). Inflammatory disease was defined as episodes of active inflammation with or without the need for resection for intractability or bleeding.10 – 14 With time, the inflammatory process could evolve into either fistulizing or stricturing disease. Stricturing disease was defined by the presence of clinically symptomatic fixed narrowing or fibrostenosis of the bowel.14 – 17 Fistulizing disease was defined as transmurally aggressive disease with fistula, abscess formation or perforation as a primary event, not secondary to stricture or obstruction.13 – 15 Surgical history was classified as to the need for surgery of any type for Crohn’s disease, surgery for fistulizing or perforating disease, surgery for intractable disease, and the need for an ostomy.
Age at Diagnosis Patients were also segregated by age at diagnosis rather than age at symptom or disease onset because of the lack of specific data in many patients. The age groups were as follows: younger than 20 years, 20–39 years, and 40 years or older. The presence or absence of a family history of Crohn’s disease was also used.
Statistical Analysis As an exploratory analysis, data were compared using the Student’s t test, Fisher’s Exact Test for 2 1 2 tables, or x2 analysis with Yates’ correction where appropriate. To determine the independent effect of age at diagnosis on disease course, a conditional logistic regression analysis was used.18,19 Age at diagnosis was tested conditionally, while controlling for site of disease, clinical type of disease, sex, family history of Crohn’s disease, and duration of disease. The significance of the effect of age at diagnosis on each of the other variables was calculated and expressed as P values.
Results Family History There was a significant correlation between age at diagnosis and the presence of a positive family history of Crohn’s disease. The 95 patients with Crohn’s disease and a family history of Crohn’s disease were, as a group, significantly younger at the time of diagnosis than the 432 patients with a negative family history of IBD (mean, 21.8 years vs. 26.7 years; P õ 0.0005) (Table 1). A greater proportion of patients with an younger age at diagnosis (younger than 20 years) was found to have a positive family history of IBD compared with patients with an older age at diagnosis: 29.9% of patients diagnosed before the age of 20 years compared with 18.8%
Table 1. Characteristics of Total Crohn’s Disease Population IBD
Total population No. Types (%) Inflammatory Stricturing Fistulizing Sex (%) Male Female Age at diagnosis ( yr ) Mean Range SD
Negative family history
Positive family history
UC, positive family history
552
432
120
95
32
235 (43) 204 (37) 113 (20)
179 (41) 165 (38) 88 (20)
56 (47) 39 (33) 25 (21)
42 (44) 31 (33) 22 (23)
18 (56) 10 (31) 4 (13)
252 (46) 300 (54)
199 (46) 233 (54)
53 (44) 67 (56)
44 (46) 51 (54)
12 (38) 20 (62)
25.8 6–76 11.94
26.7a 7–76 12.36
22.5 6–65 9.58
21.8a 6–65 9.93
24.6 14–41 6.96
a
P õ 0.0005. CD, Crohn’s disease; UC, ulcerative colitis.
/ 5E11$$0008
CD, positive family history
08-16-96 06:58:23
gasa
WBS-Gastro
582 POLITO ET AL.
GASTROENTEROLOGY Vol. 111, No. 3
Figure 1. Crohn’s disease. Proportion of patients with family history of IBD decreases with age at diagnosis. **P õ 0.005.
of patients diagnosed between the ages of 20 and 39 years and 13.6% of patients diagnosed at the age of 40 years or older. These differences were significant (P õ 0.005) (Figure 1). Distribution of Disease Distribution of disease differed significantly for patients categorized by age at diagnosis. As age at diagnosis increased, the proportion of patients with small bowel involvement decreased from 88.7% for patients diagnosed before the age of 20 years to 77.7% for patients diagnosed between the ages of 20 and 39 years (P õ 0.005) and to 57.5% for patients diagnosed at the age of 40 years or older (P õ 0.0005). Conversely, colonic involvement was significantly more common in the 40 years and older group (84.8%) compared with the younger age groups (P õ 0.05) (Figure 2). Esophagogastroduodenal involvement was more common in the youngest age group than in the oldest group (P õ 0.05).
Figure 2. Crohn’s disease. Small bowel distribution of disease more frequent in patients with younger age at diagnosis compared with the greater frequency of colonic disease in patients diagnosed after the age of 40 years. EGD, esophagus/gastric/duodenal. *P õ 0.05; **P õ 0.005; ***P õ 0.0005. j, Younger than 20 years old; , 20–39 years old; h, 40 years of age or older.
age of 20 years, 44.7% among those diagnosed between 20 and 39 years (P õ 0.05), and 54.5% among those diagnosed at the age of 40 years or older (P õ 0.005). Fistulizing disease was present in all three age groups, and there were no significant differences in frequency.14 Surgical History The frequency of the need for surgery for any indication for Crohn’s disease, for fistulizing or perforat-
Disease Subtypes Disease subtypes also differed for patients by age at diagnosis (Figure 3). As age at diagnosis increased, the proportion of patients with stricturing-type Crohn’s disease decreased significantly from 45.8% for patients diagnosed before the age of 20 years to 33.7% (P õ 0.05) for patients diagnosed between the ages of 20 and 39 years and 28.8% (P õ 0.005) for patients diagnosed at the age of 40 years or older. Conversely, the proportion of patients with inflammatory-type disease increased with increasing age at diagnosis: 34.4% among those diagnosed before the / 5E11$$0008
08-16-96 06:58:23
gasa
Figure 3. Crohn’s disease. Inflammatory subtype of disease relatively more frequent in patients diagnosed at an older age, whereas stricturing disease is relatively less frequent. *P õ 0.05; **P õ 0.005. j, Younger than 20 years old; , 20–39 years old; h, 40 years of age or older.
WBS-Gastro
September 1996
CROHN’S DISEASE: INFLUENCE OF AGE AT DIAGNOSIS 583
ing disease, for an ostomy, and for surgery for intractable disease was determined in two groups of patients: age at diagnosis of younger than 20 years and age at diagnosis of 20 years or older. A history of surgery for Crohn’s disease was significantly more common in those patients with a younger age at diagnosis (younger than 20 years) compared with those patients with an older age at diagnosis (20 years or older) (70.6% vs. 55.3%; P õ 0.001; odds ratio, 1.93). Patients with a younger age at diagnosis were also more likely to have had surgery for intractable disease (21.6% vs. 12.5%; P õ 0.01; odds ratio, 1.91). Mean duration of illness from diagnosis to the conclusion of the study was 1.95 years longer in the younger-age-at-diagnosis group (14.85 years) compared with the older-age-at-diagnosis group (12.90 years). Although this difference is significant (P õ 0.05), the conditional logistic regression model (discussed below) was controlled for any influence on disease site or type of complications. Correlation of Surgical History With Family History A positive family history for Crohn’s disease compared with no family history of Crohn’s disease was associated significantly with a history of surgery for abscess or perforation (26.3% vs. 16.6%; P õ 0.05; odds ratio, 1.79). This was not explained by disease distribution because there were no significant differences in disease distribution among the patients with or without a family history of Crohn’s disease. The mean duration of illness for patients with a positive family history (14.1 years) was not significantly different than for patients without a positive family history (13.3 years). Conditional Logistic Regression Analysis The conditional logistic regression analysis indicated that there was an independent effect of age at
Figure 4. Crohn’s disease. Sites of disease in five series of children and adolescents totaling 252 patients compared with seven series of predominantly adult populations totaling 2740 patients (see Table 2 for references and details). Jejunoileitis was more common in children and adolescents, and colitis was more common in adult populations. s, Children/adolescents; ●, adults only.
/ 5E11$$0008
08-16-96 06:58:23
gasa
Table 2. Crohn’s Disease: Differences in Age at Onset or Diagnosis in Patients With or Without a Family History of IBD Positive family history
Negative family history
Study
n
Average age (yr )
n
Average age (yr )
Baltimorea Stockholm2 France19
95 98 160
21.8 25 22.6
432 950 NA
26.8 33 26.5
NA, not available. a Current study.
diagnosis on Crohn’s disease site and need for surgery above and beyond the effect of the other covariables included in the analysis. This model was controlled for site of disease, type of disease, sex, age at diagnosis, family history of Crohn’s disease, and duration of disease. A younger age at diagnosis was found to be predictive of ileal site of disease (P õ 0.0001) and of surgery for intractable disease (P õ 0.05).
Discussion The age at diagnosis of Crohn’s disease, an admittedly imperfect surrogate for disease onset, allows one to further categorize patients in terms of factors that may influence site and clinical type of disease. Presumably, disease onset occurs earlier with certain genes rather than there being a gene for age at onset. Patients with an inherited disorder that has an early onset in life are likely to express the disease more severely than those patients with a later onset of disease and are more likely to have similarly affected relatives.8 This is presumably because of a proportionally greater contribution of more disadaptive genes to the heritability of the disease in question. Patients diagnosed before the age of 20 years were more likely to have a positive family history of Crohn’s disease (30%),20 suggesting that they may be distinguished genetically from patients who are diagnosed later (positive family history in 13%) and who presumably have developed the disease at a later age. In addition, patients with a positive family history, in our series as well as in studies in Sweden and France, are younger at the time of diagnosis than those with a negative family history (Table 2).1,21 This same trend of seemingly less genetic influences in older patients is notable in studies, such as one study in Vermont of 24 elderly patients with Crohn’s disease. None of the 24 elderly patients (mean age at onset, 57 years) had a family history of IBD compared with 5 of 24 of younger patients (21%) (mean WBS-Gastro
584 POLITO ET AL.
GASTROENTEROLOGY Vol. 111, No. 3
age at onset, 23 years) matched for sex and duration of disease.22 Determining the age at onset of a subtle and often subclinical disease such as Crohn’s disease can be difficult. Objectives measures that preceded or accompanied the onset of clinical symptoms, such as a decrease in growth velocity, were helpful in dating the onset of disease in ú80% of prepubescent adolescents.23 Observational bias by families and physicians, as well as differences in local diagnostic approaches and management practices, can also influence the results of studies of disease natural history. The similarity of data showing a younger age at onset and/or diagnosis with a positive family history in Sweden, France, and the United States1,21 (Table 2) helps argue, to some degree, against regional differences and also against a consistent observational bias with families with an illness being more alert to the illness in ensuing children.1,8,21 The concept of genetic anticipation, in which the affected offspring manifest a disease at a younger age and often more severely than the affected parent, usually a man, was also attributed to observational bias22 until a molecular mechanism for the phenomenon was identified. An inverse correlation was found between the length of genomic triplet repeats and the age at onset in the offspring in parent-child pairs with Huntington’s disease.23 Similar triplet-repeat explanations for anticipation in other monogenic neurological disorders has strengthened the validity of this phenomenon. The parent-child data in our series of patients with Crohn’s disease fit with the concept of anticipation.24 Twenty-four of 27 children were younger than their affected parent at the age at diagnosis. The average differ-
ence in age at diagnosis, using a marginal regression model with the generalized estimating equations approach that was adjusted for index case, sex, and generation, was 10.8 years (P õ 0.05).24 The age at diagnosis was associated with the site of involvement with Crohn’s disease, regardless of the presence or absence of a positive family history. A younger age at diagnosis (younger than 20 years) correlated significantly with an increased frequency of ileal disease compared with patients diagnosed at an older age in whom colonic disease was more common. The literature strongly supports the concept of more frequent small bowel disease in childhood and adolescent series and more colonic disease in older-onset populations11,16,25 – 32 (Table 3 and Figure 4). An exception was the Olmsted County population–based study, in which there were no significant differences in age-specific incidence rates in the patients grouped by disease site. However, the youngeronset patients were not well represented in that population because only 6 of the 103 incidence cases were diagnosed before the age of 15 years.33 When analyzing our 27 parent-child pairs for extent of disease,24 the disease was more extensive in 15 children and equally extensive in 7 other children compared with the parent. When the parent was a man, more extensive disease in the child was more likely (13 of 16 pairs) than when the parent was a woman (2 of 11 pairs). We concluded that the data suggest genetic anticipation in Crohn’s disease.24 Because anatomic involvement at the time of diagnosis of Crohn’s disease is predictive of clinical course, type of complications, and prognosis,34 – 38 it is not unexpected
Table 3. Crohn’s Disease: Differences in Predominant Site of Disease in Children and Adolescents Compared With AdultOnset Predominant Populations Percent of population according to predominant site of bowel involvement (%) Study Children and adolescents Baltimore22 New Haven24 Hamilton25 Baltimore21 Boston26 Adult predominant populations Cleveland13 National Crohn’s disease study15 Stockholm27 Oxford28 Olmsted County29 Tronso, Norway30 Baltimorea
n
Jejeunoileitis
Ileitis
Ileocolitis
Colitis
58 86 58 50
26 10 24 18 20
38 19 22 34
21 52 40 42 52
14 9 14 6 9
615 569 826
3
29 30 41 30 35 31 38
41 55
27 14 17 26 36 29 21
1 2
103 75 552
1
a
Current study.
/ 5E11$$0008
08-16-96 06:58:23
gasa
WBS-Gastro
42 29 37 39
September 1996
CROHN’S DISEASE: INFLUENCE OF AGE AT DIAGNOSIS 585
that we found that younger age of onset, being associated with small bowel disease, was also associated with more stricturing disease and a greater need for surgery. The literature supports the concept of more frequent use of surgery and more frequent postresection recurrences in children and adolescents with Crohn’s disease compared with older adults, again presumably as a result of having more small bowel disease and its associated complications. There is one as yet unconfirmed report of a small group of patients suggesting that adenocarcinoma of the colon occurs with a greater frequency in patients with Crohn’s colitis that was diagnosed before the age of 25 years compared with those patients diagnosed after the age of 25 years.39 We assume that the younger-age-at-diagnosis patients represent a separate and often more severe phenotype of Crohn’s disease and have a different genetic component to their disease compared with patients who develop Crohn’s disease later in life. We believe that younger age at diagnosis provides a valid form of stratification supporting the concept of heterogeneity in Crohn’s disease.
13.
14.
15.
16.
17.
18.
19.
20.
21.
References 1. Monsen U, Burnell O, Johansson C, Hellers G. Prevalence of inflammatory bowel disease among relatives of patients with Crohn’s disease. Scand J Gastroenterol 1991;26:302–306. 2. Lashner BA, Evans AA, Kirsner JB, Hanauer SB. Prevalence and incidence of inflammatory bowel disease in family members. Gastroenterology 1986;91:1396–1400. 3. Farmer RG, Michener WM, Mortimer EA. Studies of family history among patients with inflammatory bowel disease. Clin Gastroenterol 1980;9:271–277. 4. Mayberry JF, Rhodes J, Newcombe RG. Familial prevalence of inflammatory bowel disease in relatives of patients with Crohn’s disease (abstr). BMJ 1980;280:84. 5. Fielding JF. The relative risk of inflammatory bowel disease among patients and siblings of Crohn’s disease patients. J Clin Gastroenterol 1986;8:655–657. 6. Rose JDR, Roberts GM, Williams G, Mayberry JF, Rhodes J. Cardiff Crohn’s jubilee. The incidence over 50 years. Gut 1988; 29:346–351. 7. Haug K, Schrumpf E, Halvorsen JF, Fluge G, Hamre E, Hamre T, Skjollingstad and the Study Group of Inflammatory Bowel Disease in Western Norway. Epidemiology of Crohn’s disease in Western Norway. Scand J Gastroenterol 1989;24:1271–1275. 8. Childs B, Scriver CR. Age at onset and causes of disease. Part I. Pers Biol Med 1986;29:437–460. 9. Binder V, Both H, Hansen PK, Hendriksen C, Kreiner S, TorpPedersen K. Incidence and prevalence of ulcerative colitis and Crohn’s disease in the county of Copenhagen 1962–1978. Gastroenterology 1982;83:562–568. 10. Podolsky DK. Inflammatory bowel disease. N Engl J Med 1991; 325:928–937;1008–1016. 11. Farmer RG. Clinical features, laboratory findings and course of Crohn’s disease. In: Kirsner JB, Shorter RB, eds. Inflammatory bowel disease. 3rd ed. Philadelphia: Lea & Febiger, 1988:175– 183. 12. Bayless TM, Tokayer AZ, Polito JM, Quaskey SA, Mellits ED, Harris ML. Crohn’s disease: concordance for site and clinical type
/ 5E11$$0008
08-16-96 06:58:23
gasa
22. 23.
24.
25.
26.
27. 28.
29. 30.
31.
32. 33.
34.
in affected family members—potential hereditary influences. Gastroenterology 1996;111:573–579. Farmer RG, Hawk WA, Turnbull RB. Clinical patterns in Crohn’s disease. A statistical study of 615 cases. Gastroenterology 1975;68:627–635. Sachar DB, Andrews H, Farmer RG, Pallone F, Pena S, Prantera C, Rutgeerts P. Proposed classification of patient subgroups in Crohn’s disease. Working team report 4. Gastroenterol Int 1992; 3:141–154. Greenstein AJ, Lachman P, Sachar DB. Perforating and nonperforating indications for repeated operations in Crohn’s disease. Evidence for 2 clinical forms. Gut 1988;29:588–592. Mekhijian HS, Swit DM, Melnyk CS, Rankin GB, Brooks RK. Clinical features and natural history of Crohn’s disease. Gastroenterology 1979;77:898–906. Kelly JK, Sutherland LR. The chronological sequence in the pathology of Crohn’s disease. J Clin Gastroenterol 1988;10:28– 33. Jenrich RI, Moore RH. Maximum likelihood estimation by means of non-linear least squares. Proc Stat Comput Sect Am Stat Assoc 1975:57–65. Cantor RM, Rotter JI. Analysis of genetic data: methods and interpretation. The genetic basis of common diseases. Oxford, England: Oxford University, 1992:49–70. Calenda KA, Grand RJ. Clinical manifestations of pediatric inflammatory bowel disease. In: Kirsner JB, Shorter RG, Inflammatory bowel disease. 4th ed. Baltimore: Williams & Wilkins, 1995: 380–389. Gower-Rousseau C, Legat S, Evard JP, Grandbastien B, Hugot JP, Dupas JL, Gendre JP, Modigliani R, Belaiche J, Hostein J, Vortot A, Colombel JF. Familiality in clinical characteristics of Crohn’s disease in 69 French families (abstr). Gastroenterology 1995;108:A826. Penrose LS. The problem of anticipation in pedigrees of dystrophia myotonica. Ann Eugenics 1948;14:125–132. Snell R, MacMillan T, Cheadle JP. Relationship between trinucleotide repeat expansion and phenotypic variation in Huntington’s disease. Nat Genet 1993;4:393–397. Polito JM II, Rees RC, Childs B, Harris ML, Bayless TM. Crohn’s disease: evidence for genetic ‘‘anticipation.’’ Lancet 1996;347: 798–800. Harper PC, McAuliffe TL, Beeken WL. Crohn’s disease in the elderly. A statistical comparison with younger patients matched for sex and duration of disease. Arch Intern Med 1986;146: 753–755. Kanof M, Lake A, Bayless TM. Decreased height velocity in children and adolescents before the diagnosis of Crohn’s disease. Gastroenterology 1988;95:1523–1527. Burbige E, Huang SS, Bayless TM. Clinical manifestations of Crohn’s disease in adolescence. Pediatrics 1975;55:866–871. Whittington P, Barnes H, Bayless TM. Medical management of Crohn’s disease in adolescence. Gastroenterology 1977;72: 1338–1344. Gryboski JD, Spiro HM. Prognosis in children with Crohn’s disease. Gastroenterology 1978;74:807–817. Hamilton JR, Bruce GA, Abdourhaman M, Gall DG. Inflammatory bowel disease in children and adolescents. Adv Pediatr 1979; 26:311–341. Hellers G. Crohn’s disease in Stockholm County 1955–1974. A study of epidemiology, results of surgical treatment and longterm prognosis. Acta Chir Scand 1979;490(suppl):1–84. Truelove SC, Pena AS. Course and prognosis of Crohn’s disease. Gut 1976;17:192–201. Gollop JH, Phillips SF, Melton LJ III, Zinsmeister AR. Epidemiologic aspects of Crohn’s disease: a population based study in Olmsted County, Minnesota, 1943–1982. Gut 1988;29:49–56. Kildebo S, Breckan K, Norgaard K, Burhol PG, Jorde R, Members
WBS-Gastro
586 POLITO ET AL.
35. 36.
37. 38.
39.
GASTROENTEROLOGY Vol. 111, No. 3
of the Northern Norway Gastroenterology Society. The incidence of Crohn’s disease in northern Norway from 1983 to 1986. Scand J Gastroenterol 1989;24:1265–1270. Stalnikowicz R, Eliakin R, Diab R, Rachmilewitz D. Crohn’s disease in the elderly. J Clin Gastroenterol 1989;11:411–415. Farmer RG, Michener WM. Prognosis of Crohn’s disease with onset in children or adolescents. Dig Dis Sci 1979;24:752– 757. Booth IW, Harries IT. Inflammatory bowel disease in childhood. Gut 1984;25:188–202. Farmer RG, Whelan G, Fazio VW. Long-term follow up of patients with Crohn’s disease. Relationship between the clinical pattern and prognosis. Gastroenterology 1985;88:1818–1825. Gillen CD, Walmsley RS, Prior P, Andrews HA, Allen RN. Ulcerative colitis and Crohn’s disease. A comparison of the colorectal cancer risk in extensive colitis. Gut 1994;35:1590–1592.
/ 5E11$$0008
08-16-96 06:58:23
gasa
Received July 19, 1995. Accepted March 18, 1996. Address requests for reprints to: Theodore M. Bayless, M.D., Gastroenterology Division, Blalock 461, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, Maryland 21287. Fax: (410) 6141280. Supported by the Allen Guerrieri family, Pat and John Rosenwald Philanthropic Fund, Cynthia and Peter Rosenwald Fund, Samuel Himmelrich, Sr., family, and Mr. and Mrs. Sherlock Hibbs. Presented in part at the 1995 annual meeting of the American Gastroenterological Association and published in abstract form (Gastroenterology 1995;108:A895). The authors thank Dr. Aaron Z. Tokayer for his important role in caring for some of these patients and in classifying the patients as to site and type of disease, respectively; Shirley A. Quasky for performing the logistical regression analyses; and Donna Rode for secretarial support.
WBS-Gastro
Crohn’s Disease: Influence of Age at Diagnosis on Site and Clinical Type of Disease JOSEPH M. POLITO II,*,† BARTON CHILDS,§ E. DAVID MELLITS‡,x,¶ AARON Z. TOKAYER,*,† MARY L. HARRIS,*,† and THEODORE M. BAYLESS*,† *Gastroenterology Division, Department of Medicine, and Departments of §Pediatrics and xNeurology, The Johns Hopkins University School of Medicine; ¶Department of Biostatistics, The Johns Hopkins University School of Hygiene and Public Health; and †Meyerhoff Center for Digestive Diseases and Inflammatory Bowel Disease, The Johns Hopkins Hospital, Baltimore, Maryland
See editorial on page 813. Background & Aims: Crohn’s disease has a bimodal age distribution of disease onset diagnosis. The peaks (20 and 50 years) may represent different phenotypes or different genetic and/or environmental influences between younger- and older-onset individuals. The aim of this study was to examine the influences of age at diagnosis of Crohn’s disease on disease site, type, and course. Methods: Records of 552 consecutive patients with Crohn’s disease were reviewed retrospectively. Results: Younger age at diagnosis (younger than 20 years), compared with an older age (40 years or older), was associated with a greater prevalence of a family history of Crohn’s disease (29.9% vs. 13.6%), greater small bowel involvement (88.7% vs. 57.5%), more stricturing disease (45.8% vs. 28.8%), and a higher frequency of surgery (70.6% vs. 55.3%). Older age at diagnosis was associated with a greater prevalence of colonic disease (84.8% vs. 71.2%) and the inflammatory subtype (54.5% vs. 34.4%). A conditional logistic regression analysis confirmed an independent effect of age at diagnosis on ileal disease and surgery for intractable disease. Conclusions: In Crohn’s disease, early age at diagnosis is associated with more complicated disease and a greater likelihood of having affected relatives. Stratification of Crohn’s disease by age at diagnosis provides support for the concept of genetic heterogeneity.
sent different groups or phenotypes or may indicate different genetic and/or environmental influences between younger-onset patients and older-onset individuals.8,9 The purpose of this study was to examine the association, if any, between age at diagnosis of Crohn’s disease and the disease site and type and family history in 552 consecutive patients. Differences in frequency of a family history of Crohn’s disease; separate sites of disease involvement, i.e., ileal, ileocolonic, colonic, and jejunoileitis; and clinical types, i.e., inflammatory, fistulizing and/or perforating, and stricturing, were compared in groups of patients based on age at diagnosis.
Materials and Methods Patient Population The records of 552 consecutive patients with documented Crohn’s disease examined by two of the investigators (T.M.B. and M.L.H.) at The Johns Hopkins Hospital between January 1985 and December 1991 were reviewed retrospectively. Three hundred patients (54%) were women, and 252 patients (46%) were men. Five hundred forty-two patients were white, 10 were black, and none were Asian. The diagnosis of Crohn’s disease was based on accepted clinical, radiological, and histological criteria.10,11 This is the same population used in our study of the influence of concordance within families on site and clinical type of Crohn’s disease.12 Two patients, both with negative family histories of IBD, who were in the original study were excluded because of lack of data of age at diagnosis or onset.
Family History
C
rohn’s disease is a form of idiopathic inflammatory bowel disease (IBD) with a positive family history in at least one sixth of patients.1 – 3 Having a sibling with Crohn’s disease increases the risk of an individual developing the same illness by a factor of 30 times compared with the general population.4,5 There is a bimodal age distribution of disease onset and diagnosis in population studies of Crohn’s disease with a peak incidence in the third decade and a smaller peak in the eighth decade.6,7 These two peaks may repre/ 5E11$$0008
08-16-96 06:58:23
gasa
The presence or absence of a family history of idiopathic IBD as well as age at diagnosis of disease were determined from the patient and any relatives in attendance by use of a standardized form at the time of the initial visit to The Johns Hopkins Hospital for all patients. Among these 552 consecutive patients with Crohn’s disease, 95 (17%) were con-
‡Deceased. q 1996 by the American Gastroenterological Association 0016-5085/96/$3.00
WBS-Gastro
September 1996
CROHN’S DISEASE: INFLUENCE OF AGE AT DIAGNOSIS 581
firmed to have a positive family history of Crohn’s disease. Family history was confirmed by contacting, with permission, the affected relative and the relative’s physician. Family history status was reconfirmed in all patients at least 1 year before closure of the study. Seven patients reported a family history of both ulcerative colitis and Crohn’s disease. Sixty-eight patients had first-degree relatives with Crohn’s disease. These 95 patients with a family history of Crohn’s disease were similar in sex and distribution and type of disease to the 432 patients without a family history of IBD (Table 1). The sites of disease are the same as in Table 1 of our report on concordance of site and type of disease within families, except that 2 patients without a family history of IBD and with colonic disease of an inflammatory type were excluded for lack of data of age at diagnosis.12 Information on religion and quantitative smoking history was not available in most patients.
Classification All patients were classified at the time of the latest visit by site of disease involvement (small bowel [jejunoileitis, ileum, or ileocecal], colon, and other [usually esophagogastroduodenal]) and by segregating them into one of three subtypes of disease (inflammatory, stricturing, and fistulizing). Inflammatory disease was defined as episodes of active inflammation with or without the need for resection for intractability or bleeding.10 – 14 With time, the inflammatory process could evolve into either fistulizing or stricturing disease. Stricturing disease was defined by the presence of clinically symptomatic fixed narrowing or fibrostenosis of the bowel.14 – 17 Fistulizing disease was defined as transmurally aggressive disease with fistula, abscess formation or perforation as a primary event, not secondary to stricture or obstruction.13 – 15 Surgical history was classified as to the need for surgery of any type for Crohn’s disease, surgery for fistulizing or perforating disease, surgery for intractable disease, and the need for an ostomy.
Age at Diagnosis Patients were also segregated by age at diagnosis rather than age at symptom or disease onset because of the lack of specific data in many patients. The age groups were as follows: younger than 20 years, 20–39 years, and 40 years or older. The presence or absence of a family history of Crohn’s disease was also used.
Statistical Analysis As an exploratory analysis, data were compared using the Student’s t test, Fisher’s Exact Test for 2 1 2 tables, or x2 analysis with Yates’ correction where appropriate. To determine the independent effect of age at diagnosis on disease course, a conditional logistic regression analysis was used.18,19 Age at diagnosis was tested conditionally, while controlling for site of disease, clinical type of disease, sex, family history of Crohn’s disease, and duration of disease. The significance of the effect of age at diagnosis on each of the other variables was calculated and expressed as P values.
Results Family History There was a significant correlation between age at diagnosis and the presence of a positive family history of Crohn’s disease. The 95 patients with Crohn’s disease and a family history of Crohn’s disease were, as a group, significantly younger at the time of diagnosis than the 432 patients with a negative family history of IBD (mean, 21.8 years vs. 26.7 years; P õ 0.0005) (Table 1). A greater proportion of patients with an younger age at diagnosis (younger than 20 years) was found to have a positive family history of IBD compared with patients with an older age at diagnosis: 29.9% of patients diagnosed before the age of 20 years compared with 18.8%
Table 1. Characteristics of Total Crohn’s Disease Population IBD
Total population No. Types (%) Inflammatory Stricturing Fistulizing Sex (%) Male Female Age at diagnosis ( yr ) Mean Range SD
Negative family history
Positive family history
UC, positive family history
552
432
120
95
32
235 (43) 204 (37) 113 (20)
179 (41) 165 (38) 88 (20)
56 (47) 39 (33) 25 (21)
42 (44) 31 (33) 22 (23)
18 (56) 10 (31) 4 (13)
252 (46) 300 (54)
199 (46) 233 (54)
53 (44) 67 (56)
44 (46) 51 (54)
12 (38) 20 (62)
25.8 6–76 11.94
26.7a 7–76 12.36
22.5 6–65 9.58
21.8a 6–65 9.93
24.6 14–41 6.96
a
P õ 0.0005. CD, Crohn’s disease; UC, ulcerative colitis.
/ 5E11$$0008
CD, positive family history
08-16-96 06:58:23
gasa
WBS-Gastro
582 POLITO ET AL.
GASTROENTEROLOGY Vol. 111, No. 3
Figure 1. Crohn’s disease. Proportion of patients with family history of IBD decreases with age at diagnosis. **P õ 0.005.
of patients diagnosed between the ages of 20 and 39 years and 13.6% of patients diagnosed at the age of 40 years or older. These differences were significant (P õ 0.005) (Figure 1). Distribution of Disease Distribution of disease differed significantly for patients categorized by age at diagnosis. As age at diagnosis increased, the proportion of patients with small bowel involvement decreased from 88.7% for patients diagnosed before the age of 20 years to 77.7% for patients diagnosed between the ages of 20 and 39 years (P õ 0.005) and to 57.5% for patients diagnosed at the age of 40 years or older (P õ 0.0005). Conversely, colonic involvement was significantly more common in the 40 years and older group (84.8%) compared with the younger age groups (P õ 0.05) (Figure 2). Esophagogastroduodenal involvement was more common in the youngest age group than in the oldest group (P õ 0.05).
Figure 2. Crohn’s disease. Small bowel distribution of disease more frequent in patients with younger age at diagnosis compared with the greater frequency of colonic disease in patients diagnosed after the age of 40 years. EGD, esophagus/gastric/duodenal. *P õ 0.05; **P õ 0.005; ***P õ 0.0005. j, Younger than 20 years old; , 20–39 years old; h, 40 years of age or older.
age of 20 years, 44.7% among those diagnosed between 20 and 39 years (P õ 0.05), and 54.5% among those diagnosed at the age of 40 years or older (P õ 0.005). Fistulizing disease was present in all three age groups, and there were no significant differences in frequency.14 Surgical History The frequency of the need for surgery for any indication for Crohn’s disease, for fistulizing or perforat-
Disease Subtypes Disease subtypes also differed for patients by age at diagnosis (Figure 3). As age at diagnosis increased, the proportion of patients with stricturing-type Crohn’s disease decreased significantly from 45.8% for patients diagnosed before the age of 20 years to 33.7% (P õ 0.05) for patients diagnosed between the ages of 20 and 39 years and 28.8% (P õ 0.005) for patients diagnosed at the age of 40 years or older. Conversely, the proportion of patients with inflammatory-type disease increased with increasing age at diagnosis: 34.4% among those diagnosed before the / 5E11$$0008
08-16-96 06:58:23
gasa
Figure 3. Crohn’s disease. Inflammatory subtype of disease relatively more frequent in patients diagnosed at an older age, whereas stricturing disease is relatively less frequent. *P õ 0.05; **P õ 0.005. j, Younger than 20 years old; , 20–39 years old; h, 40 years of age or older.
WBS-Gastro
September 1996
CROHN’S DISEASE: INFLUENCE OF AGE AT DIAGNOSIS 583
ing disease, for an ostomy, and for surgery for intractable disease was determined in two groups of patients: age at diagnosis of younger than 20 years and age at diagnosis of 20 years or older. A history of surgery for Crohn’s disease was significantly more common in those patients with a younger age at diagnosis (younger than 20 years) compared with those patients with an older age at diagnosis (20 years or older) (70.6% vs. 55.3%; P õ 0.001; odds ratio, 1.93). Patients with a younger age at diagnosis were also more likely to have had surgery for intractable disease (21.6% vs. 12.5%; P õ 0.01; odds ratio, 1.91). Mean duration of illness from diagnosis to the conclusion of the study was 1.95 years longer in the younger-age-at-diagnosis group (14.85 years) compared with the older-age-at-diagnosis group (12.90 years). Although this difference is significant (P õ 0.05), the conditional logistic regression model (discussed below) was controlled for any influence on disease site or type of complications. Correlation of Surgical History With Family History A positive family history for Crohn’s disease compared with no family history of Crohn’s disease was associated significantly with a history of surgery for abscess or perforation (26.3% vs. 16.6%; P õ 0.05; odds ratio, 1.79). This was not explained by disease distribution because there were no significant differences in disease distribution among the patients with or without a family history of Crohn’s disease. The mean duration of illness for patients with a positive family history (14.1 years) was not significantly different than for patients without a positive family history (13.3 years). Conditional Logistic Regression Analysis The conditional logistic regression analysis indicated that there was an independent effect of age at
Figure 4. Crohn’s disease. Sites of disease in five series of children and adolescents totaling 252 patients compared with seven series of predominantly adult populations totaling 2740 patients (see Table 2 for references and details). Jejunoileitis was more common in children and adolescents, and colitis was more common in adult populations. s, Children/adolescents; ●, adults only.
/ 5E11$$0008
08-16-96 06:58:23
gasa
Table 2. Crohn’s Disease: Differences in Age at Onset or Diagnosis in Patients With or Without a Family History of IBD Positive family history
Negative family history
Study
n
Average age (yr )
n
Average age (yr )
Baltimorea Stockholm2 France19
95 98 160
21.8 25 22.6
432 950 NA
26.8 33 26.5
NA, not available. a Current study.
diagnosis on Crohn’s disease site and need for surgery above and beyond the effect of the other covariables included in the analysis. This model was controlled for site of disease, type of disease, sex, age at diagnosis, family history of Crohn’s disease, and duration of disease. A younger age at diagnosis was found to be predictive of ileal site of disease (P õ 0.0001) and of surgery for intractable disease (P õ 0.05).
Discussion The age at diagnosis of Crohn’s disease, an admittedly imperfect surrogate for disease onset, allows one to further categorize patients in terms of factors that may influence site and clinical type of disease. Presumably, disease onset occurs earlier with certain genes rather than there being a gene for age at onset. Patients with an inherited disorder that has an early onset in life are likely to express the disease more severely than those patients with a later onset of disease and are more likely to have similarly affected relatives.8 This is presumably because of a proportionally greater contribution of more disadaptive genes to the heritability of the disease in question. Patients diagnosed before the age of 20 years were more likely to have a positive family history of Crohn’s disease (30%),20 suggesting that they may be distinguished genetically from patients who are diagnosed later (positive family history in 13%) and who presumably have developed the disease at a later age. In addition, patients with a positive family history, in our series as well as in studies in Sweden and France, are younger at the time of diagnosis than those with a negative family history (Table 2).1,21 This same trend of seemingly less genetic influences in older patients is notable in studies, such as one study in Vermont of 24 elderly patients with Crohn’s disease. None of the 24 elderly patients (mean age at onset, 57 years) had a family history of IBD compared with 5 of 24 of younger patients (21%) (mean WBS-Gastro
584 POLITO ET AL.
GASTROENTEROLOGY Vol. 111, No. 3
age at onset, 23 years) matched for sex and duration of disease.22 Determining the age at onset of a subtle and often subclinical disease such as Crohn’s disease can be difficult. Objectives measures that preceded or accompanied the onset of clinical symptoms, such as a decrease in growth velocity, were helpful in dating the onset of disease in ú80% of prepubescent adolescents.23 Observational bias by families and physicians, as well as differences in local diagnostic approaches and management practices, can also influence the results of studies of disease natural history. The similarity of data showing a younger age at onset and/or diagnosis with a positive family history in Sweden, France, and the United States1,21 (Table 2) helps argue, to some degree, against regional differences and also against a consistent observational bias with families with an illness being more alert to the illness in ensuing children.1,8,21 The concept of genetic anticipation, in which the affected offspring manifest a disease at a younger age and often more severely than the affected parent, usually a man, was also attributed to observational bias22 until a molecular mechanism for the phenomenon was identified. An inverse correlation was found between the length of genomic triplet repeats and the age at onset in the offspring in parent-child pairs with Huntington’s disease.23 Similar triplet-repeat explanations for anticipation in other monogenic neurological disorders has strengthened the validity of this phenomenon. The parent-child data in our series of patients with Crohn’s disease fit with the concept of anticipation.24 Twenty-four of 27 children were younger than their affected parent at the age at diagnosis. The average differ-
ence in age at diagnosis, using a marginal regression model with the generalized estimating equations approach that was adjusted for index case, sex, and generation, was 10.8 years (P õ 0.05).24 The age at diagnosis was associated with the site of involvement with Crohn’s disease, regardless of the presence or absence of a positive family history. A younger age at diagnosis (younger than 20 years) correlated significantly with an increased frequency of ileal disease compared with patients diagnosed at an older age in whom colonic disease was more common. The literature strongly supports the concept of more frequent small bowel disease in childhood and adolescent series and more colonic disease in older-onset populations11,16,25 – 32 (Table 3 and Figure 4). An exception was the Olmsted County population–based study, in which there were no significant differences in age-specific incidence rates in the patients grouped by disease site. However, the youngeronset patients were not well represented in that population because only 6 of the 103 incidence cases were diagnosed before the age of 15 years.33 When analyzing our 27 parent-child pairs for extent of disease,24 the disease was more extensive in 15 children and equally extensive in 7 other children compared with the parent. When the parent was a man, more extensive disease in the child was more likely (13 of 16 pairs) than when the parent was a woman (2 of 11 pairs). We concluded that the data suggest genetic anticipation in Crohn’s disease.24 Because anatomic involvement at the time of diagnosis of Crohn’s disease is predictive of clinical course, type of complications, and prognosis,34 – 38 it is not unexpected
Table 3. Crohn’s Disease: Differences in Predominant Site of Disease in Children and Adolescents Compared With AdultOnset Predominant Populations Percent of population according to predominant site of bowel involvement (%) Study Children and adolescents Baltimore22 New Haven24 Hamilton25 Baltimore21 Boston26 Adult predominant populations Cleveland13 National Crohn’s disease study15 Stockholm27 Oxford28 Olmsted County29 Tronso, Norway30 Baltimorea
n
Jejeunoileitis
Ileitis
Ileocolitis
Colitis
58 86 58 50
26 10 24 18 20
38 19 22 34
21 52 40 42 52
14 9 14 6 9
615 569 826
3
29 30 41 30 35 31 38
41 55
27 14 17 26 36 29 21
1 2
103 75 552
1
a
Current study.
/ 5E11$$0008
08-16-96 06:58:23
gasa
WBS-Gastro
42 29 37 39
September 1996
CROHN’S DISEASE: INFLUENCE OF AGE AT DIAGNOSIS 585
that we found that younger age of onset, being associated with small bowel disease, was also associated with more stricturing disease and a greater need for surgery. The literature supports the concept of more frequent use of surgery and more frequent postresection recurrences in children and adolescents with Crohn’s disease compared with older adults, again presumably as a result of having more small bowel disease and its associated complications. There is one as yet unconfirmed report of a small group of patients suggesting that adenocarcinoma of the colon occurs with a greater frequency in patients with Crohn’s colitis that was diagnosed before the age of 25 years compared with those patients diagnosed after the age of 25 years.39 We assume that the younger-age-at-diagnosis patients represent a separate and often more severe phenotype of Crohn’s disease and have a different genetic component to their disease compared with patients who develop Crohn’s disease later in life. We believe that younger age at diagnosis provides a valid form of stratification supporting the concept of heterogeneity in Crohn’s disease.
13.
14.
15.
16.
17.
18.
19.
20.
21.
References 1. Monsen U, Burnell O, Johansson C, Hellers G. Prevalence of inflammatory bowel disease among relatives of patients with Crohn’s disease. Scand J Gastroenterol 1991;26:302–306. 2. Lashner BA, Evans AA, Kirsner JB, Hanauer SB. Prevalence and incidence of inflammatory bowel disease in family members. Gastroenterology 1986;91:1396–1400. 3. Farmer RG, Michener WM, Mortimer EA. Studies of family history among patients with inflammatory bowel disease. Clin Gastroenterol 1980;9:271–277. 4. Mayberry JF, Rhodes J, Newcombe RG. Familial prevalence of inflammatory bowel disease in relatives of patients with Crohn’s disease (abstr). BMJ 1980;280:84. 5. Fielding JF. The relative risk of inflammatory bowel disease among patients and siblings of Crohn’s disease patients. J Clin Gastroenterol 1986;8:655–657. 6. Rose JDR, Roberts GM, Williams G, Mayberry JF, Rhodes J. Cardiff Crohn’s jubilee. The incidence over 50 years. Gut 1988; 29:346–351. 7. Haug K, Schrumpf E, Halvorsen JF, Fluge G, Hamre E, Hamre T, Skjollingstad and the Study Group of Inflammatory Bowel Disease in Western Norway. Epidemiology of Crohn’s disease in Western Norway. Scand J Gastroenterol 1989;24:1271–1275. 8. Childs B, Scriver CR. Age at onset and causes of disease. Part I. Pers Biol Med 1986;29:437–460. 9. Binder V, Both H, Hansen PK, Hendriksen C, Kreiner S, TorpPedersen K. Incidence and prevalence of ulcerative colitis and Crohn’s disease in the county of Copenhagen 1962–1978. Gastroenterology 1982;83:562–568. 10. Podolsky DK. Inflammatory bowel disease. N Engl J Med 1991; 325:928–937;1008–1016. 11. Farmer RG. Clinical features, laboratory findings and course of Crohn’s disease. In: Kirsner JB, Shorter RB, eds. Inflammatory bowel disease. 3rd ed. Philadelphia: Lea & Febiger, 1988:175– 183. 12. Bayless TM, Tokayer AZ, Polito JM, Quaskey SA, Mellits ED, Harris ML. Crohn’s disease: concordance for site and clinical type
/ 5E11$$0008
08-16-96 06:58:23
gasa
22. 23.
24.
25.
26.
27. 28.
29. 30.
31.
32. 33.
34.
in affected family members—potential hereditary influences. Gastroenterology 1996;111:573–579. Farmer RG, Hawk WA, Turnbull RB. Clinical patterns in Crohn’s disease. A statistical study of 615 cases. Gastroenterology 1975;68:627–635. Sachar DB, Andrews H, Farmer RG, Pallone F, Pena S, Prantera C, Rutgeerts P. Proposed classification of patient subgroups in Crohn’s disease. Working team report 4. Gastroenterol Int 1992; 3:141–154. Greenstein AJ, Lachman P, Sachar DB. Perforating and nonperforating indications for repeated operations in Crohn’s disease. Evidence for 2 clinical forms. Gut 1988;29:588–592. Mekhijian HS, Swit DM, Melnyk CS, Rankin GB, Brooks RK. Clinical features and natural history of Crohn’s disease. Gastroenterology 1979;77:898–906. Kelly JK, Sutherland LR. The chronological sequence in the pathology of Crohn’s disease. J Clin Gastroenterol 1988;10:28– 33. Jenrich RI, Moore RH. Maximum likelihood estimation by means of non-linear least squares. Proc Stat Comput Sect Am Stat Assoc 1975:57–65. Cantor RM, Rotter JI. Analysis of genetic data: methods and interpretation. The genetic basis of common diseases. Oxford, England: Oxford University, 1992:49–70. Calenda KA, Grand RJ. Clinical manifestations of pediatric inflammatory bowel disease. In: Kirsner JB, Shorter RG, Inflammatory bowel disease. 4th ed. Baltimore: Williams & Wilkins, 1995: 380–389. Gower-Rousseau C, Legat S, Evard JP, Grandbastien B, Hugot JP, Dupas JL, Gendre JP, Modigliani R, Belaiche J, Hostein J, Vortot A, Colombel JF. Familiality in clinical characteristics of Crohn’s disease in 69 French families (abstr). Gastroenterology 1995;108:A826. Penrose LS. The problem of anticipation in pedigrees of dystrophia myotonica. Ann Eugenics 1948;14:125–132. Snell R, MacMillan T, Cheadle JP. Relationship between trinucleotide repeat expansion and phenotypic variation in Huntington’s disease. Nat Genet 1993;4:393–397. Polito JM II, Rees RC, Childs B, Harris ML, Bayless TM. Crohn’s disease: evidence for genetic ‘‘anticipation.’’ Lancet 1996;347: 798–800. Harper PC, McAuliffe TL, Beeken WL. Crohn’s disease in the elderly. A statistical comparison with younger patients matched for sex and duration of disease. Arch Intern Med 1986;146: 753–755. Kanof M, Lake A, Bayless TM. Decreased height velocity in children and adolescents before the diagnosis of Crohn’s disease. Gastroenterology 1988;95:1523–1527. Burbige E, Huang SS, Bayless TM. Clinical manifestations of Crohn’s disease in adolescence. Pediatrics 1975;55:866–871. Whittington P, Barnes H, Bayless TM. Medical management of Crohn’s disease in adolescence. Gastroenterology 1977;72: 1338–1344. Gryboski JD, Spiro HM. Prognosis in children with Crohn’s disease. Gastroenterology 1978;74:807–817. Hamilton JR, Bruce GA, Abdourhaman M, Gall DG. Inflammatory bowel disease in children and adolescents. Adv Pediatr 1979; 26:311–341. Hellers G. Crohn’s disease in Stockholm County 1955–1974. A study of epidemiology, results of surgical treatment and longterm prognosis. Acta Chir Scand 1979;490(suppl):1–84. Truelove SC, Pena AS. Course and prognosis of Crohn’s disease. Gut 1976;17:192–201. Gollop JH, Phillips SF, Melton LJ III, Zinsmeister AR. Epidemiologic aspects of Crohn’s disease: a population based study in Olmsted County, Minnesota, 1943–1982. Gut 1988;29:49–56. Kildebo S, Breckan K, Norgaard K, Burhol PG, Jorde R, Members
WBS-Gastro
586 POLITO ET AL.
35. 36.
37. 38.
39.
GASTROENTEROLOGY Vol. 111, No. 3
of the Northern Norway Gastroenterology Society. The incidence of Crohn’s disease in northern Norway from 1983 to 1986. Scand J Gastroenterol 1989;24:1265–1270. Stalnikowicz R, Eliakin R, Diab R, Rachmilewitz D. Crohn’s disease in the elderly. J Clin Gastroenterol 1989;11:411–415. Farmer RG, Michener WM. Prognosis of Crohn’s disease with onset in children or adolescents. Dig Dis Sci 1979;24:752– 757. Booth IW, Harries IT. Inflammatory bowel disease in childhood. Gut 1984;25:188–202. Farmer RG, Whelan G, Fazio VW. Long-term follow up of patients with Crohn’s disease. Relationship between the clinical pattern and prognosis. Gastroenterology 1985;88:1818–1825. Gillen CD, Walmsley RS, Prior P, Andrews HA, Allen RN. Ulcerative colitis and Crohn’s disease. A comparison of the colorectal cancer risk in extensive colitis. Gut 1994;35:1590–1592.
/ 5E11$$0008
08-16-96 06:58:23
gasa
Received July 19, 1995. Accepted March 18, 1996. Address requests for reprints to: Theodore M. Bayless, M.D., Gastroenterology Division, Blalock 461, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, Maryland 21287. Fax: (410) 6141280. Supported by the Allen Guerrieri family, Pat and John Rosenwald Philanthropic Fund, Cynthia and Peter Rosenwald Fund, Samuel Himmelrich, Sr., family, and Mr. and Mrs. Sherlock Hibbs. Presented in part at the 1995 annual meeting of the American Gastroenterological Association and published in abstract form (Gastroenterology 1995;108:A895). The authors thank Dr. Aaron Z. Tokayer for his important role in caring for some of these patients and in classifying the patients as to site and type of disease, respectively; Shirley A. Quasky for performing the logistical regression analyses; and Donna Rode for secretarial support.
WBS-Gastro