Incidence, Prevalence, Etiology, and Prognosis of First-Time Chronic Pancreatitis in Young Patients: A Nationwide Cohort Study
- First Online:
- Cite this article as:
- Joergensen, M., Brusgaard, K., Crüger, D.G. et al. Dig Dis Sci (2010) 55: 2988. doi:10.1007/s10620-009-1118-4
- 254 Views
Publications on etiology of chronic pancreatitis (CP) are infrequent. Etiologies today encompass genetic disorders. We wanted to describe etiologies of today and identify patients with genetic disorders like hereditary pancreatitis (HP), mutations in Serine Protease Inhibitor Kazal type1 (SPINK1), and the Cystic Fibrosis Transmembrane Conductance Regulator gene (CFTR) among patients formerly considered to have idiopathic CP.
Data on patients diagnosed with first-time CP < 30 years of age in Denmark identified in the Danish National Registry of Patients were retrieved. Patients previously considered to have idiopathic pancreatitis were offered genetic counseling and evaluation for HP, SPINK1, and CFTR mutations.
In the period 1980–2004, 580 patients < 30 years of age presented with CP, the standardized prevalence ratio of CP increased from 11.7 per 100,000 person years in 1980–1984 to 17.0 per 100,000 in 2000–2004 (p < 0.001). The odds ratio (OR) having gallstone-related CP increased in the latter time period, especially in women, that of alcohol-induced CP decreased over time. OR having idiopathic CP increased in the latter period; 50% of patients with idiopathic pancreatitis accepted genetic reevaluation; 28 patients had a genetic mutation that totally or partly could explain their pancreatitis, nine of these had two, and 11 patients had HP.
The prevalence of CP, especially in women, increased over time. Genetic causes that partly or totally could explain the CP were found in 54.90% (95% CI (40.45–68.62)) of those with idiopathic CP, as a minimum estimation 1.9% (95% CI (1.00–3.47)) of the total cohort had HP.
KeywordsChronic pancreatitisPrevalenceEtiologyHereditary pancreatitisCFTR mutationsSPINK1 mutations
Diagnosis of chronic pancreatitis may be difficult and comparative studies on epidemiology of chronic pancreatitis troublesome, partly because the diagnostic criteria are not unified and partly because methods for diagnosis differ greatly especially over time. In addition, a geographical variation which may reflect differences in etiology has been demonstrated .
In 1981 the incidence rate and prevalence of chronic pancreatitis in Denmark were reported to be, respectively, 4.0 and 13.0 per year per 100,000 citizens aged 20 years or more . Recent population-based epidemiological studies on chronic pancreatitis are sparse and time trends in incidence and prevalence rates are often lacking. A few reports show a crude incidence rate of chronic pancreatitis ranging from 5.9 to 7.9 per 100,000 inhabitants and one study—including all age groups–has reported a crude prevalence rate of 26.4 per 100,000 inhabitants [3–5].
In 2002, Banks concluded that the two main etiologies of chronic pancreatitis were alcohol and tropical pancreatitis . Two recent studies from 2004 and 2008 have reported that etiological features of chronic pancreatitis have changed. In particular, the frequency of alcohol-induced chronic pancreatitis is lower, obstruction of the pancreatic duct accounts for a higher percentage of chronic pancreatitis (approximately 10%) and etiologies as autoimmune and especially hereditary/genetic pancreatitis are reported with increasing frequency [7, 8].
Since the mutations in the gene for cationic trypsinogen (PRSS1) causing hereditary pancreatitis (HP) were not identified until 1996 , these patients have to be found after reevaluation of former diagnosis in retrospective studies, especially among patients diagnosed with idiopathic pancreatitis. Genetic etiologies of pancreatitis now also include Serine Protease Inhibitor Kazal type1 (SPINK1) and Cystic Fibrosis Transmembrane Conductance Regulator gene (CFTR) mutations [10, 11].
The aim of this study was (1) to describe the incidence, prevalence, etiology, time trends, and prognosis of first-time chronic pancreatitis in patients younger than 30 years of age at diagnosis in the period 1977–2004 and (2) to test patients previously considered to have chronic idiopathic pancreatitis for PRSS1 mutations and/or fulfillment of the criteria complex for hereditary pancreatitis, SPINK1 and CFTR mutations. The present study is a retrospective population-based cohort study.
Materials and Methods
The cohort comprised the entire Danish population < 30 years of age. We retrieved data from the Danish National Registry of Patients  and the Civil Registration System. Data were linked by the unique civil personal number (CPR-number). Retrieved data encompassed; data on first-time diagnosis of chronic pancreatitis (from 1977–1993 identified by the codes 577.1 and 577.9 according to the World health Organization (WHO) international classification of diseases (ICD-8) and K 86.0, K 86.1, K 86.8, and K 86.9 according to ICD-10 from 1994), number and date of hospital admissions and discharges because of chronic pancreatitis from 1 January 1977 until 31 December 2004, date of birth, migration in and out of the country, date of death, and cause of death. Patient records, autopsy reports, and death certificates were scrutinized.
Criteria for the diagnosis chronic pancreatitis
Histology (chronic inflammation, fibrosis, calcifications)
ERCP (typical duct pathology, stones)
Exocrine insufficiency (Lundh test, secretine-CCK test, fecal elastase)
Clinical symptoms (chronic upper abdominal pain, attacks of acute pancreatitis)
Endocrine insufficiency (diabetes mellitus)
4 points: Chronic Pancreatitis
The diagnosis of acute pancreatitis was accepted if the patients had abdominal pain and a > 3 times increase in serum or urine amylase level and/or typical signs of acute pancreatitis by radiological examination, laparoscopy, laparotomy, or autopsy. If acute pancreatitis had been diagnosed, the date was established via the patients’ records and the Danish National Registry of Patients.
A first-time chronic pancreatitis was present if no evidence of a previous diagnosis was found in clinical records or in the Danish National Registry of Patients. Using the time period from 1977–1979 as a run-in period, we excluded prevalent cases of chronic pancreatitis when calculating the incidence. Patients diagnosed in this period were included in the calculation of prevalence.
Patients were subdivided into nine main categories of etiology
Acute pancreatitis only
Patient records not available
Trauma, post-ERCP-pancreatitis, drugs, virus (mononucleosis, hepatitis), bacteria (Versinia enterocolitica), helminthes (Ascaris lumbricoides), cancer radiotherapy
Crohn's disease, ulcerative colitis, hyperlipidemia, cystic fibrosis, primary sclerosing cholangitis, nesidioblastosis, Shwachman-Diamonds syndrome, Systemic lupus erythematosus, porphyria, hypercalcaemia, renal dysfunction, hypoparathyroidism, sarcoidosis, myxedema
Patients with idiopathic pancreatitis were offered genetic counseling before giving consent to participate in the study. In each case a pedigree was drawn. We did not contact minors (<15 years of age), unless they still had contact to the hospital system, in which case parents gave informed consent on behalf of their children. In order to exclude other causes of pancreatitis patients were tested for hyperlipidemia, hypercalcemia, hemochromatosis, and renal dysfunction. Since only approximately 80% of patients with hereditary pancreatitis have a PRSS1 mutation , the diagnosis is established on the basis of a criteria complex in some of the patients. The international diagnostic criteria for hereditary pancreatitis are: two first-degree relatives or three or more second-degree relatives in two or more generations with recurrent acute pancreatitis, and/or chronic pancreatitis for which there were no known precipitating factors .
Blood was drawn into tubes with EDTA and stored at –20 degrees until analysis. Genomic DNA was extracted using the Maxwell (Promega, Ramcon Denmark) DNA purification robot. The samples were tested for small deletions, insertions, or point mutations in all exons and exon-intron boundaries of the PRSS1 and SPINK1 genes using DHPLC (WAVE 3500HT High Sensitivity System; Transgenomic Inc, Elancourt, France). Variants in the PRSS1 gene were categorized as polymorphisms if they were identified in control samples with a comparable frequency. Samples with deviating chromatographic profiles were sequenced in both directions using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and analyzed on an automated ABI PRISM 3100 (Applied Biosystems).
The samples were also tested for 33 CFTR mutations: 394delTT, R553X, 621 + 1G > T, R1162X, 1717-1G > A, 3659delC, G542X, 2183A > G, W1282X, 1078delT, 711 + 1G > T, F508del, S549 N, I507del, S549R, 2184delA, G551D, G85E, N1303 K, R560T, R117H, R347H, R347P, R334 W, 2789 + 5G > A, 3849 + 10kbC > T, A445E, 3120 + 1G > A, V520F, 1898 + 1G > A, 3876delA, 3905insT, and IVS8-5T.
DNA was amplified by multiplex PCR (Hybaid 4 A62). Samples were mixed with an Oligonucleotide Ligation Assay (Abbott), the ligation products were separated by electrophoresis and analyzed by ABI 3130 Genetic Analyzer (Applied Biosystems).
Data were non-parametric and followed a Poisson distribution. Pair-wise binary outcomes were analyzed using Fisher’s exact test and groups of categorical variables were compared using the Chi-squared test. Continuous variables were tested using the Mann–Whitney test or Kruskall Wallis test. Direct standardization was used calculating the standardized incidence rates (SIR) and the standardized prevalence rates (SPR) and their differences. Incidence rate ratios were calculated when appropriate. Logistic regression analysis was used to estimate the association between hospital admissions and sex, age, age at first-time diagnosis of acute pancreatitis, and etiology of acute pancreatitis. The number of deaths during follow-up was reported by standard mortality rates (SMR) standardized by sex in 10-year age strata. The hazard ratios for the various etiologies were assessed by multivariate Cox’s regression analysis including sex and age. All analyses were performed by Stata 9.2 (Stata Corporation, Texas USA). P-values below 0.05 were considered as statistically significant. The study was approved by the Regional Scientific Ethics Committee of Southern Denmark and by the Danish Data Protection Agency.
Between 1980 and 2004, 849 Danish citizens were for the first time considered to have chronic pancreatitis before 30 years of age. A total of 83 of the 849 patients had acute pancreatitis only (36 women and 47 men), 56 patients (28 women and 28 men) were misdiagnosed and the records of 130 patients (32 woman and 98 men) were unobtainable. These 269 patients (96 women and 173 men) encompassed 31.68% of the 849 patients and were excluded from further analyses. This left a cohort amounting to 580 patients, 211 (36.38%) women, and 369 (63.62%) men. The female:male ratio of the group of excluded patients was similar to that of included patients (p = 0.88), but median age was 1.1 years older, 26.5 years (95% CI (25.8–26.9)) versus 25.5 years (95% CI (24.8–26.0)), (p = 0.009).
Incidence and Prevalence
SIR for men showed a falling trend (from 1.31 to 1.19 per 100,000 person years) when the period 1980–1984 was compared to the period 2000–2004. An increase in SIR from 0.66 per 100,000 person years in the period 1980–84 to 0.89 per 100,000 person years in 2000–2004 was found in women. Neither of the differences were statistically significant. The incidence of chronic pancreatitis of women surpassed that of men in 2003.
SIR for chronic pancreatitis increased by age and was significantly higher in the age group 20–29 than in the age groups 0–9 and 10–19, both in 1980–1984 (2.14 vs. 0.5 and 0.26 per 100,000, respectively) and in 2000–2004 (2.14 vs. 0.26 and 0.66 per 100,000, respectively). The incidence did not increase over time in the age group 20–29, whereas a statistically significant increase was found in the age group 10–19. SIR in age group 0–9 did not increase.
The standardized prevalence rate (SPR) increased over time in all age groups but in the youngest group changes were not within the limits of statistical significance (age group 0–9 increased from 4.39 to 5.45, age group 10–19 from 3.84 to 8.26 and age group 20–29 from 8.9 to 11.85 per 100,000).
Percentages of etiologies, odds ratio, and significance level for first-time chronic pancreatitis in women and men comparing the periods 1980–1998 and 1999–2004
1980–1998 (% in the period)
1999–2004 (% in the period)
Odds ratio (95% CI)
Idiopathic pancreatitis was significantly more common after 1999 than before. In the period 1999–2004 it constituted 27.2% of the causes of chronic pancreatitis. Alcoholic chronic pancreatitis was significantly more prevalent in the period 1980–1998 compared to the period 1999–2004. The prevalence of chronic pancreatitis due to gallstones increased in the latter period. The prevalence of chronic pancreatitis because of predisposing illness also increased significantly in the latter time period.
Sex differences in percentages of etiology of first-time chronic pancreatitis in the period 1980–1998
Odds ratio (95% CI)
Sex differences in percentages of etiology of for first-time chronic pancreatitis in the period 1999–2004
Odds ratio (95% CI)
The Diagnosis Chronic Pancreatitis
Median age of onset of chronic pancreatitis in the periods 1980–1998 and 1999–2004 by etiology
Median age of onset of chronic pancreatitis 1980–1998 (95% CI)
Median age of onset of chronic pancreatitis 1999–2004 (95% CI)
In total, the median age at first-time diagnosed chronic pancreatitis was 1.59 years lower (p = 0.01) in the latter time period. In the time period 1980–1998 the median age at first-time chronic pancreatitis was 25.69 years (95% CI (25.13–26.3)) and in 1999–2004 it was 24.01 years (95% CI (22.86–25.25)).
The age at first-time diagnosed chronic pancreatitis in both periods also depended on etiology (1980–1998: p = 0.0001, d.f. = 5 and 1999–2004: p = 0.02, d.f. = 5) (Table 6). In the period 1980–1998 patients who developed chronic pancreatitis because of an exogenous factor were youngest and patients with alcohol-related first-time chronic pancreatitis were oldest. In the period 1999–2004, patients with first-time chronic pancreatitis and pancreas divisum were youngest and again patients with alcohol-related pancreatitis were oldest.
Concomitant Chronic and Acute Pancreatitis
A total of 348 (60%) of the 580 patients also had a diagnosis of acute pancreatitis, 116 (33.3%) of these were women.
Approximately 2/3 of those who had both acute and chronic pancreatitis had alcohol as etiology whereas alcohol only accounted for 1/3 of those who had chronic pancreatitis only. Idiopathic pancreatitis was diagnosed in approximately 1/5 and 1/6, respectively. More patients with chronic pancreatitis only had predisposing illness or provoking factors as etiology.
Both men and women were admitted to hospital with similar frequencies (p = 0.18). Admission to hospital for chronic pancreatitis was more common in the time period 1980–1998 than in the period 1999–2004 (p = 0.02). The IRR for admissions in the later time period was 0.59 (95% CI (0.38–0.91)) compared to the first time period.
Patients with Idiopathic Pancreatitis
Some 113 patients (19.5%) had idiopathic pancreatitis, one of these had died, five were minors and three had emigrated. We thus contacted 104 (92.0%) of the 113 patients with idiopathic pancreatitis and offered genetic evaluation; 51 (49.0%) accepted genetic evaluation.
Female:male ratio and age at diagnosis of chronic pancreatitis were similar in participants and non-participants, but age at diagnosis of acute pancreatitis was slightly higher in participants than in non-participants (23.6 years (95% CI (20.6–24.7)) versus 20.3 years (95% CI (16.7–22.4))).
Eight of the 51 tested had a PRSS1 mutation and three fulfilled the criteria complex of having hereditary pancreatitis (HP), i.e., 11 or 21.57% (95% CI (11.57–35.70)) of the tested had HP. One woman and one man had the A16 V mutation, two women and three men had the R122H mutation, one man had the novel mutation K23_I24insIDK (c.61_73dup) and three women fulfilled the criteria complex. Three of the women and two men had also a SPINK1 mutation (N34S). In addition, one woman had a CFTR mutation (F508del).
A total of 16 (31.37%, 95% CI (19.52–46.03)) had a SPINK1 mutation; six women and ten men had a N34S mutation. Five had also PRSS1 mutations as mentioned above, three a CFTR mutation, two of these were women (1 F508del and 1IVS8-5T) and one was a man (1 R117H).
Ten had a CFTR mutation (19.61%, 95% CI (10.29–33.55)); four had additional mutations that could cause pancreatitis. Two women and 1 man had a F508del mutation, 5 women an IVS8-5T mutation and 2 women and 1 man a R117H mutation. One woman both had a R117H and an IVS8-5T mutation.
Overall, 28 or 54.90%, 95% CI (40.45–68.62) of those tested with idiopathic chronic pancreatitis diagnosed for the first time before 30 years of age had a genetic mutation that totally or partly could explain their pancreatitis. One mutation was found in 19 patients, but in nine, an additional mutation was found. Overall, a minimum estimate of total or partly genetic explanation was found among 4.8%, 95% CI (3.29–6.99) of the total cohort. As a minimum estimation, 1.9% (95% CI (1.00–3.47)) were diagnosed with hereditary pancreatitis in the total cohort.
The cause of death was corrected in 23 (24.7%) of the patients after evaluation of files, autopsy reports, and death certificates. Corrections were made before further analyses.
Some 93 (16.0%) of the 580 patients died during follow-up until 15 May 2008, 21 (22.6%) of these were women. More men (19.5% of all men) than women (10% of all women) died during follow-up (p = 0.002). No difference was found in age at death (p = 0.98); women died at a median age of 38.1 years (95% CI (31.9–41.0)) and men at a median age of 36.4 (95% CI (34.1–39.4)). The duration from first-time diagnosis of chronic pancreatitis until death was 8.9 years (95% CI (5.5–13.7)) for women and 9.7 years (95% CI (6.9–11.8) for men.
A total of 17 patients (2.9%), five women and 12 men had emigrated during the follow-up period, 470 (81%) patients (185 women) remained when follow-up had finished.
Some 79 (85.0%) of those who died had alcohol-related pancreatitis, eight (8.6%) had predisposing illness, two (2.2%) had been exposed to an exogenous factor and for one (1.1%) no cause could be demonstrated (idiopathic). A significant difference in survival depending on etiology of the chronic pancreatitis was demonstrated (log rank test p = 0.0001). Controlling for age and sex the hazard ratio of death for the etiology alcohol was 23.3 (95% CI (3.2–169.9), p = 0.002)). The hazard ratio for predisposing illness was 15.4, 95% CI (1.9–123.8), p = 0.01).
Causes of death
Cause of death
Complications of diabetes mellitus
The second most frequent cause of death was infection, amounting to almost 1/5 of deaths. Approximately 1/10 died because of liver cirrhosis, including bleeding esophageal varices and in 1/10 the cause of death was not obvious. Only 1/20 died as a direct cause of chronic/acute pancreatitis, and a similar number died because of cancer and cardiopulmonary diseases. No significant relation was found between cause of death and sex (p = 0.61), etiology of the chronic pancreatitis (p = 0.10), age at diagnosis (p = 0.72), age at death (p = 0.44), or time difference from chronic pancreatitis (p = 0.68) or acute pancreatitis (p = 0.46) was diagnosed.
All-cause mortality rates per 1,000 person years and all-cause standard mortality ratio (SMR) for different age strata of patients with chronic pancreatitis
Number of deaths
MR (per 1,000 person-years)
SMR (95% CI)
MR (per 1,000 person-years)
SMR (95% CI)
MR (per 1,000 person-years)
SMR (95% CI)
We found an upward tendency for incidence of chronic pancreatitis in women and a decreasing incidence in men when comparing the periods 1980–1984 and 2000–2004. Differences were, however, not within limits of statistical significance. In total, the incidence did not rise over time, which is in alignment of the study of Worning from 1990  but in contrast to statements of others [1, 17, 18]. However, it should be stressed that our cohort only encompasses young patients with chronic pancreatitis. This should also be kept in mind when comparing the total incidence rate of 2.08 (women 0.89 and men 1.19) with previously reported incidence rates of 4.0 per 100,000 inhabitants in Denmark in patients older than 20 years of age . Like Worning, who reported a change in the male:female ratio from approximately 5 to 2–3 in the time period 1946–1984 [19, 20], we found a decrease from 1.98 in 1980–1985 to 1.34 in 2000–2004.
We found a significant increase in the prevalence rate of chronic pancreatitis from 11.7 to 17.07. This prevalence rate is somewhat higher than the previously published prevalence rates of 13.0 per 100,000 inhabitants in Denmark (>20 years of age in 1981) . Very few have published about changes in prevalence. Otsuki found an increase in prevalence of chronic pancreatitis from 28.5 per 100,000 inhabitants in 1994 to 32.91 in 1999 including all age groups .
We have not found any other data on time trends in sex-specific prevalences of chronic pancreatitis. In this study we found a significant increase in prevalence in both men and women with a male:female ratio of 1.88 in the period 1985–1989 and 0.75 in the period 2000–2004.
We found that the odds ratio of having alcohol-related chronic pancreatitis in the period 1999–2004 was 0.35 compared to the period 1980–1998 and that the decrease was found in both men and women. In both periods the odds of alcohol-related pancreatitis were higher in men. In accordance with Ryu et al.  and Frulloni et al.  we found that alcohol as etiology in the total period constituted approximately 50%, but with a significant decrease from approximately 50% in 1980–1999 to 25% in 2000–2004. In many, but not all recent studies the proportion of chronic pancreatitis attributed to alcohol appears to decrease . Though Lankisch et al.  reported that alcohol abuse was the etiology in 72% of patients with chronic pancreatitis in 2002.
In this study the proportion of patients diagnosed with idiopathic pancreatitis increased from 17.4% in the early period to 27.2% of the etiologies in the later period. This increase in idiopathic pancreatitis was statistically significant in males. These percentages are similar to those found by Ryu et al.  and Frulloni et al. , Coyle et al. , Lankisch et al. , and Yusoff et al. . The increase might be explained by the increasing awareness of this disease entity.
The proportion of patients with chronic pancreatitis who were also diagnosed with acute pancreatitis corresponds to findings in young patients with acute pancreatitis who progressed to chronic pancreatitis in the County of Funen ; approximately 4% of those in whom the disease progressed from acute to chronic pancreatitis had gallstone-related acute pancreatitis. Approximately 10% of patients had chronic pancreatitis due to gallstones; gallstones were also found in 1/10 of the patients with chronic pancreatitis in the Copenhagen Pancreatitis Study . Gallstones are one of the most frequent causes of acute pancreatitis and an increasing incidence of gallstone-related acute pancreatitis is found in Denmark . One report shows that in about half of the patients with biliary acute recurring pancreatitis, chronic pancreatitis is demonstrated at ERCP. However, a retrospective study from 1995 found that biliary lithiasis in itself was not an etiological factor for chronic pancreatitis . The formerly referred studies of Ryu and Frulloni found similar percentages of obstruction (not only gallstones) causing chronic pancreatitis as we did, 8.6 and 17%, respectively [7, 8]. Another explanation for the increasing occurrence of chronic pancreatitis with stones in the pancreatic duct and gallstones might be the increasing sensitivity of the diagnostic methods used , but it is unsolved whether the two diseases are linked by a cause–effect relationship. The proportion of patients having chronic pancreatitis and pancreas divisum or exposure to provoking factors did not change over time. It is estimated that 1–7% of the patients undergoing ERCP has pancreas divisum, autopsy studies have shown that 5–14% of the patients had pancreas divisum and the relation to acute and chronic pancreatitis is debatable .
An increase in predisposing illnesses causing chronic pancreatitis might be due to an increasing time of living in these groups, exemplified by the increasing duration of life among patients with cystic fibrosis due to better treatments. Perhaps the increasing prevalence of chronic pancreatitis can also be explained by an increase in the duration of life among patients with chronic pancreatitis in general due to better treatments as we have shown that the incidence has not increased.
Increasing sensitivity of diagnostic tests and awareness of chronic pancreatitis as a cause of abdominal pain, diarrhea, and weight loss might also be the explanation for the significantly lower age at diagnosis in both sexes, though predominantly in men, in the period 1999–2004 compared to 1980–1998. Studies including patients from the whole age range report that the median age at diagnosis of chronic pancreatitis ranges from 33 to 59 years, one group stating that the median age of onset in men and women was approximately the same [3, 18, 28].
After genetic evaluation of patients with idiopathic pancreatitis we found that 21.57%, 95% CI (11.57–35.70) of those tested, or as a minimum estimation 1.9%, 95% CI (1.00–3.47) of the total cohort, had hereditary pancreatitis (HP). Others have reported that 1% of all patients with chronic pancreatitis has HP. The higher occurrence in our cohort may be explained by the higher frequency of HP among patients younger than 30 years of age.
Some 31.37% (95% CI (19.52–46.03)) of those tested with idiopathic pancreatitis had SPINK1 mutations. This is in agreement with other studies in which SPINK1 mutations were present in 15–40% of the patients .
We found a CFTR mutation in 19.61% (95% CI (10.29–33.55)) of those who were tested. Others have stated that 20–30% of patients with idiopathic chronic pancreatitis should carry at least one CFTR mutation, but only a few were compound heterozygote [30–32]. Three of the patients with CFTR mutations had also a SPINK1 mutation and 1 a PRSS1 mutation.
In total, half of those with idiopathic chronic pancreatitis who were evaluated genetically and as a minimum estimation 5% of the total cohort, had a genetic disease that totally or partly could explain their chronic pancreatitis.
Lowenfels et al.  did not find any significant changes in hospitalization rates in the period 1983–2000. We found that patients with chronic pancreatitis were admitted to the hospital twice as often in the period 1980–1998 than in the period 1999–2004. This is in contrast with observations of doubled hospital admission rates for chronic pancreatitis in the period 1989–1990 to 1999–2000 by Tinti et al.  and Spanier et al. , but may reflect a better treatment as outward patients—which could also be in alignment of the increasing prevalence but stabile incidence of chronic pancreatitis in young people.
Mortality of patients with chronic pancreatitis has been reported to be 3.6 times higher than that of patients without pancreatitis . We found a much higher mortality rate of 6.25. Looking at the overall mortality we found that 16% of the cohort had died after median 13.61 years of follow-up compared to 55% found by Lowenfels in 1994 after 20 years of follow-up  and 21–35% after 6.3–9.8 years of observation time found by others [37–39]. The standard mortality rate (SMR) was higher (6.32, 95% CI (5.10–7.74)) than the SMR of 2.07 reported by Otsuki . This might be due to the inclusion of only young patients in our cohort. In agreement with Otsuki et al.  we found that the SMR of men was higher than for women and that the mortality of patients with alcoholic chronic pancreatitis was higher than for non-alcoholic chronic pancreatitis.
In the last five decades, beginning in 1948 , more than ten different classification systems of chronic pancreatitis have been proposed. The recent classification systems combine imaging, functional, and clinical criteria for the diagnosis of chronic pancreatitis. This must be kept in mind when results of epidemiological studies are compared, as studies may be based on different classification systems. Furthermore, the lists of possible etiological factors [41, 42] are not exclusive and the recorded cause–effect relationship may be spurious as a systematic search for all possible etiologies was not carried out in the majority of patients. We found rather many patients with chronic pancreatitis and gallstones, but the relation between gallstones and chronic pancreatitis may not be straightforward. The demonstration of gallstones may be an incidental finding combined with an idiopathic chronic pancreatitis. This is questionable as is also the importance of alcohol in inducing acute and chronic pancreatitis, as only 3–7% of the heavy drinkers develop chronic pancreatitis. Maybe alcohol should only be considered as a cofactor and only associated with pancreatitis in conjunction with other triggers. In this respect, genetic or environmental factors (especially nicotine consumption) seem to be of increasing importance [43, 44] and genetic evaluation of patients with acute and chronic pancreatitis, perhaps not restricted to patients with idiopathic pancreatitis, seems of increasing importance.
We thank the patients participating in the study and secretaries who have been helpful in scrutinizing hospital files. A special thanks to our laboratory technicians Henriette Hansen, Steen Gregersen, and Lene Byriel and molecular biologist Anders Aamann Rasmussen who have set up the CFTR analysis.