Advertisement

Mortality and causes of death in different celiac disease phenotypes during long-term follow-up

  • Inka Koskinen
    Correspondence
    Correspondence to: Dr. Inka Koskinen, Celiac Disease Research Center, Faculty of Medicine and Health Technology, FIN-33014 Tampere University, Finland.
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland

    Department of Internal Medicine, Central Finland Central Hospital, Jyväskylä, Finland
    Search for articles by this author
  • Kaisa Hervonen
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland

    Department of Dermatology, Tampere University Hospital, Tampere, Finland
    Search for articles by this author
  • Heini Huhtala
    Affiliations
    Faculty of Social Sciences, Tampere University, Tampere, Finland
    Search for articles by this author
  • Camilla Pasternack
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland
    Search for articles by this author
  • Teea Salmi
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland

    Department of Dermatology, Tampere University Hospital, Tampere, Finland
    Search for articles by this author
  • Timo Reunala
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland

    Department of Dermatology, Tampere University Hospital, Tampere, Finland
    Search for articles by this author
  • Pekka Collin
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland

    Department of Gastroenterology and Alimentary Tract Surgery, Tampere University Hospital, Tampere, Finland
    Search for articles by this author
  • Katri Kaukinen
    Affiliations
    Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland

    Department of Internal Medicine, Tampere University Hospital, Tampere, Finland
    Search for articles by this author
Open AccessPublished:May 17, 2022DOI:https://doi.org/10.1016/j.dld.2022.04.016

      Abstract

      Background

      Celiac disease has been associated with increased mortality, but data on long-term mortality are scarce.

      Aims

      To determine long-term mortality in celiac disease.

      Methods

      The study cohort consisted of all celiac disease patients (n=1,392) diagnosed in Tampere University Hospital catchment area 1960 – 2000. Patients were categorized into subgroups based on demographic (age, gender, decade of diagnosis) and celiac disease characteristics (e.g., phenotype, severity of villous atrophy) collected from medical records. Overall and cause-specific mortality was compared to those of age-, sex-, and place of residence matched reference individuals (n=4,177) over time.

      Results

      During the 41 years of follow-up (median 26.5 years), 376 celiac disease patients and 1,155 reference individuals died. All-cause mortality was not increased (hazard ratio (HR) 0.96, 95% confidence intervals (CI) 0.85–1.08). Mortality from lymphoproliferative diseases and diseases of the central nervous system was increased (HR 2.42, 95% CI 1.38–4.24 and HR 2.14, 95% CI 1.05–4.36 respectively) while the risk from alcohol related diseases was decreased (HR 0.31, 95% CI 0.09–1.00). Examination of various celiac disease phenotypes revealed no significant differences in mortality

      Conclusions

      Overall mortality was not increased in any celiac disease phenotype during a very long-term follow-up.

      Keywords

      1. Introduction

      Celiac disease has been associated in numerous studies with 1.2- to 2-fold increased mortality [
      • Corrao G
      • Corazza GR
      • Bagnardi V
      • et al.
      Mortality in patients with coeliac disease and their relatives: a cohort study.
      ,
      • Tio M
      • Cox MR
      • Eslick GD.
      Meta-analysis: Coeliac disease and the risk of all-cause mortality, any malignancy and lymphoid malignancy.
      ,
      • Quarpong W
      • Card TR
      • West J
      • et al.
      Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
      ,
      • Holmes GKT
      • Muirhead A.
      Mortality in coeliac disease: a population-based cohort study from a single centre in Southern Derbyshire, UK.
      ,
      • Lebwohl B
      • Green PHR
      • Söderling J
      • et al.
      Association between celiac disease and mortality risk in a Swedish population.
      ], mainly caused by lymphoproliferative malignancies and cancers of the gastrointestinal tract [
      • West J
      • Logan RFA
      • Smith CJ
      • et al.
      Malignancy and mortality in people with coeliac disease: population based cohort study.
      ,
      • Catassi C
      • Fabiani E
      • Corrao G
      • et al.
      Risk of non-Hodgkin lymphoma in celiac disease.
      ]. On the other hand, risk estimates similar to those in general population have also been reported [
      • Koskinen I
      • Virta LJ
      • Huhtala H
      • et al.
      Overall and cause-specific mortality in adult celiac disease and dermatitis herpetiformis diagnosed in the 21st century.
      ,
      • Abdul Sultan A
      • Crooks CJ
      • Card T
      • et al.
      Causes of death in people with coeliac disease in England compared with the general population: a competing risk analysis.
      ,
      • Hervonen K
      • Alakoski A
      • Salmi TT
      • et al.
      Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
      ]. In most studies the mean follow-up has not exceeded 5 – 13 years and therefore persistence of increased mortality during prolonged follow-up after diagnosis is uncertain [
      • Corrao G
      • Corazza GR
      • Bagnardi V
      • et al.
      Mortality in patients with coeliac disease and their relatives: a cohort study.
      ,
      • Holmes GKT
      • Muirhead A.
      Mortality in coeliac disease: a population-based cohort study from a single centre in Southern Derbyshire, UK.
      ,
      • Lebwohl B
      • Green PHR
      • Söderling J
      • et al.
      Association between celiac disease and mortality risk in a Swedish population.
      ,
      • Koskinen I
      • Virta LJ
      • Huhtala H
      • et al.
      Overall and cause-specific mortality in adult celiac disease and dermatitis herpetiformis diagnosed in the 21st century.
      ,
      • Abdul Sultan A
      • Crooks CJ
      • Card T
      • et al.
      Causes of death in people with coeliac disease in England compared with the general population: a competing risk analysis.
      ]. The only study with a longer follow-up (mean 30 years) included 602 celiac disease patients [
      • Quarpong W
      • Card TR
      • West J
      • et al.
      Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
      ]. In that study, overall mortality initially seen to increase (SMR 1.4) compared to general population, but over time to decrease, the increase after 15 years from celiac disease diagnosis being no longer significant.
      Celiac disease can manifest with a multitude of symptoms arising from different organ systems and with varying severity. In addition to frequently encountered diarrhea and malabsorption, the disease may present with dermatitis herpetiformis (the skin involvement of celiac disease), arthralgia, osteoporosis, growth retardation, infertility, and neurological disorders. These distinct celiac disease presentations may result in different prognoses; lower mortality rates have been reported in dermatitis herpetiformis [
      • Hervonen K
      • Alakoski A
      • Salmi TT
      • et al.
      Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
      ,
      • Swerdlow AJ
      • Whittaker S
      • Carpenter LM
      • et al.
      Mortality and cancer incidence in patients with dermatitis herpetiformis: a cohort study.
      ,
      • Viljamaa M
      • Kaukinen K
      • Pukkala E
      • et al.
      Malignancies and mortality in patients with coeliac disease and dermatitis herpetiformis: 30-year population-based study.
      ] and screen-detected celiac disease [
      • Lohi S
      • Mäki M
      • Rissanen H
      • et al.
      Prognosis of unrecognized coeliac disease as regards mortality: a population-based cohort study.
      ,
      • Godfrey JD
      • Brantner TL
      • Brinjikji W
      • et al.
      Morbidity and mortality among older individuals with undiagnosed celiac disease.
      ,
      • Kårhus LL
      • Skaaby T
      • Petersen J
      • et al.
      Long-term consequences of undiagnosed celiac seropositivity.
      ] than in other celiac disease phenotypes. Moreover, variations in symptom severity, as well as gender, and age at celiac disease diagnosis have been suggested as possible factors confounding detected mortality risks [
      • Corrao G
      • Corazza GR
      • Bagnardi V
      • et al.
      Mortality in patients with coeliac disease and their relatives: a cohort study.
      ,
      • Quarpong W
      • Card TR
      • West J
      • et al.
      Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
      ,
      • Ludvigsson JF
      • Montgomery SM
      • Ekbom A
      • et al.
      Small-intestinal histopathology and mortality risk in celiac disease.
      ,
      • Ilus T
      • Kaukinen K
      • Virta LJ
      • et al.
      Refractory coeliac disease in a country with a high prevalence of clinically-diagnosed coeliac disease.
      ].
      The aim of this study was to ascertain celiac disease-associated mortality during a long follow-up of up to 41 years using a well-defined cohort of celiac disease patients including all celiac disease phenotypes. Special attention was paid to the mortality risks in different celiac disease phenotypes.

      2. Materials and methods

      2.1 Study population

      All patients (n=1,460) with biopsy-proven celiac disease including dermatitis herpetiformis diagnosed in the Tampere University Hospital catchment area between January 1960 and December 2000 were prospectively collected to form the celiac disease cohort [
      • Koskinen I
      • Hervonen K
      • Pukkala E
      • et al.
      Cancer incidence and factors associated with malignancies in coeliac disease during long-term follow-up.
      ]. Sixty-eight patients were excluded from further analysis: two patients had died before the start of follow-up (January 1, 1978) and 66 patients had inadequate baseline data preventing selection of reference individuals. Thus, the study population eventually comprised 1,392 celiac disease patients. The diagnoses were based on approved diagnostic guidelines i.e., duodenal villous atrophy and crypt hyperplasia in celiac disease, and granular IgA deposits of the dermal papillae of the skin in dermatitis herpetiformis. In celiac disease the date of duodenal biopsy was set as the date of diagnosis. In dermatitis herpetiformis the date of diagnosis was assigned to be the date of immunofluorescence examination, except for patients diagnosed before the test became available in clinical practice in the 1980s. In such cases the date of diagnosis was taken to be the date of clinical diagnosis rather than the date of confirmatory immunofluorescence test performed at a later timepoint. Eighty-seven percent of dermatitis herpetiformis patients also had duodenal biopsy results available and villous atrophy was detected in 71% of them while 29% had normal villous architecture.

      2.2 Reference group

      Three reference individuals matched for age, gender, and place of residence at the time of diagnosis of celiac disease index case were selected for each celiac disease patient from the Finnish Population Information System (PIS) maintained by the Digital and Population Data Services Agency (n = 4,182). After excluding the five individuals who died before initiation of follow-up, the reference group eventually comprised 4,177 persons. The PIS contains basic information on all Finnish and foreign citizens permanently resident in Finland, each of whom can be identified by a unique personal identification code.

      2.3 Review of medical records and stratification

      The medical records of the celiac cohort were reviewed in 2015. All data available on patient and celiac disease characteristics and on adherence to gluten-free diet were collected for each patient. Patient records of 88 deceased patients were no longer available.
      The cohort was stratified into subgroups according to the following characteristics: gender; age at diagnosis (< 20, 20 – 39, 40 – 59 and ≥ 60 years); decade of diagnosis (≤ 1980, 1981 – 1990, 1991 – 2000), and duration of follow-up since diagnosis (< 2, 2 – 4.9, 5 – 14.9, 15 – 29.9, ≥ 30 years). In addition, patients were divided into the following five celiac disease phenotype categories based on the main presenting symptom: 1) gastrointestinal symptoms, including diarrhea, abdominal pain, reflux, obstipation, 2) malabsorption, including weight loss, anemia or deficiencies of iron, cobalamin, folic acid or calcium either alone or in combination, 3) dermatitis herpetiformis including patients with itchy and blistering rash at typical locations, 4) other extraintestinal symptoms including osteoporosis, infertility, neurological manifestations, arthralgia and childhood growth retardation, 5) screen detected patients including individuals with few or no symptoms who were screened due to having either an autoimmune disease e.g. type 1 diabetes, autoimmune thyroiditis, arthritis or IgA nephropathy or celiac disease in the family. Moreover, all celiac disease patients with small bowel biopsy results available were categorized into the following three subgroups according to the degree of small bowel mucosal damage: total or subtotal villous atrophy corresponding to Marsh-Oberhuber 3b-c [
      • Marsh MN.
      Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity ('celiac sprue’).
      ,
      • Oberhuber G.
      Histopathology of celiac disease.
      ], partial villous atrophy corresponding to Marsh-Oberhuber 3a, normal villous architecture corresponding to Marsh 0-2. Finally, patients were divided to two subgroups according to response to gluten-free diet. Response was considered to have been achieved if villous atrophy appeared alleviated in follow-up biopsies. The no-response group included patients with no histological villous improvement and those not adhering to gluten-free diet.

      2.4 Mortality data

      The Registry of Causes of death is maintained by Statistics Finland and covers more than 99% of all deceased residents of Finland since 1936 [

      Statistics Finland - Quality Description:Causes of death 2016. (https://tilastokeskus.fi/til/ksyyt/2016/ksyyt_2016_2017-12-29_laa_001_en.html) accessed November 5, 2021.

      ]. Immediate, underlying, and contributary causes of death are derived from death certificates [

      Statistics Finland - Quality Description:Causes of death 2016. (https://tilastokeskus.fi/til/ksyyt/2016/ksyyt_2016_2017-12-29_laa_001_en.html) accessed November 5, 2021.

      ]. The causes of death are classified into 54 groups according to the underlying cause of death [

      Statistics Finland: Causes of death, time series classification with 54 categories. (https://www.tilastokeskus.fi/til/ksyyt/ksyyt_2021-10-29_luo_001_en.pdf) accessed January 24, 2022.

      ], and these groups were used when analyzing cause-specific mortality. Alcohol-related causes of death include all alcohol-related codes in all the different disease categories of the Tenth Revision of the International Classification of Diseases (ICD-10), e.g. alcohol associated liver diseases (K70) or alcohol poisoning (X45), but not deaths caused by accidents and violence [

      Statistics Finland: Causes of death, time series classification with 54 categories. (https://www.tilastokeskus.fi/til/ksyyt/ksyyt_2021-10-29_luo_001_en.pdf) accessed January 24, 2022.

      ]. The personal identification codes of celiac disease patients and reference individuals were linked with the Registry. The dates of death were acquired until August 10, 2020, whereas causes of death were available until the end of 2018.

      2.5 Statistical analysis

      Follow-up was set to begin from January 1, 1978 or from the date of diagnosis of celiac disease or dermatitis herpetiformis (and corresponding dates in reference individuals), whichever was later. Follow-up continued until death, emigration, or August 10, 2020, whichever came first. Categorical variables were reported as frequencies and percentages and continuous variables as medians and ranges. To determine celiac disease-associated mortality overall and for different causes of death, crude mortality rates per 10,000 person-years at risk were calculated. Hazard ratios (HR) with 95% confidence intervals (CI) were calculated to estimate the risk of death using Cox proportional hazards model. Potential risk factors for stratified analyses were chosen according to the literature. All-cause mortality in each stratified subgroup was assessed comparing celiac disease patients to their designated reference individuals. Particular attention in stratified analyses was paid to duration of follow-up and to the various celiac disease phenotypes. In addition, a separate analysis was performed to ascertain whether the previously reported discrepancy in mortality risks between dermatitis herpetiformis and all the other celiac disease phenotypes combined persisted. All analyses were performed using SPSS (IBM SPSS Statistics for Windows, Version 27.0, Armonk, NY).

      2.6 Ethics

      Data permits were granted by Statistics Finland (TK-53-679-20), Digital and Population Data Services Agency (DVV/4482/2020-2), and Tampere University Hospital Science Center (R17554). All patient data were pseudonymized prior to analyses. According to Finnish legislation, no informed consent or approval by an ethics committee is required for registry-based studies when the study participants are not contacted.

      3. Results

      The study cohorts included 1,392 celiac disease patients and 4,177 reference individuals diagnosed at a median age of 39 years (range 0.8 – 84.0) (Table 1). Follow-up extended up to 41 years (median 26.5 years) producing 38,183 and 113,360 person-years of follow-up for celiac disease patients and reference individuals respectively. Throughout follow-up, altogether 376 celiac disease patients and 1,155 reference individuals died, thereby yielding respective mortality rates of 98.5 and 101.9 per 10,000 person-years (Table 2). Thus, overall mortality in the celiac cohort was equal to that among the reference individuals (HR 0.96, 95% CI 0.85 – 1.08). When examining overall mortality in the subgroups based on celiac disease phenotype and other clinical and demographic characteristics, no significant differences in relative risks were observed in comparison to reference individuals (Table 3). Overall mortality in dermatitis herpetiformis (HR 0.87, 95% CI 0.70 – 1.01) did not differ significantly from that observed in all the other celiac disease phenotypes combined (HR 0.99, 95% CI 0.87 – 1.14).
      Table 1Characteristics of the study cohort and their designated matched reference individuals.
      CharacteristicCeliac disease patients n = 1,392Reference individuals n = 4,177
      n%n%
      Gender
      Female88063.22,63963.2
      Male51236.81,53836.8
      Age at diagnosis
      < 20 years26419.079219.0
      20 – 39 years45032.31,35132.3
      40 – 59 years49235.31,47735.4
      ≥ 60 years18613.455713.3
      Decade of diagnosis
      – 198018113.054413.0
      1981 – 199060243.21,80643.2
      1991 – 200060943.81,82743.7
      Follow–up from celiac disease diagnosis
      < 2 years161.1571.3
      2 – 4.9 years271.9791.9
      5 – 14.9 years1148.23358.0
      15 – 29.9 years74553.52,28754.8
      ≥ 30 years49035.21,41934.0
      Follow–up time
      Median, years (range)26.5 (0.1 – 41)25.7 (0.0 – 41)
      Person-years at risk, years38,183113,360
      Median age at diagnosis, years (range)39.3 (0.8 – 84.8)39.3 (0.6 – 84.8)
      Median age at the end of follow-up, (range)65.7 (8.9 – 99.0)65.6 (17.5 – 102.5)
      Table 2Mortality rates per 10,000 person-years and hazard ratios (HR) for overall and cause-specific mortality
      Follow-up for overall mortality until August 10, 2020 and for cause-specific mortality until December 31, 2018
      in patients with celiac disease and in the reference group during up to 41 years of follow-up (median 26 years).
      Cause of death
      • Oberhuber G.
      Histopathology of celiac disease.
      Celiac cohort
      Includes all celiac disease phenotypes including dermatitis herpetiformis
      (n=1,392)
      Reference group (n=4,177)
      n%
      Percentage of all deaths in the group
      Rate
      Mortality rate per 10,000 person-years
      n%
      Percentage of all deaths in the group
      Rate
      Mortality rate per 10,000 person-years
      HR95% CI
      Overall mortality
      Follow-up for overall mortality until August 10, 2020 and for cause-specific mortality until December 31, 2018
      3762.998.511553.0101.90.960.86 – 1.08
      Malignancies overall8823.424.431026.828.90.840.66 – 1.06
      Gastrointestinal
      Including cancers of the liver, pancreas, and gastrointestinal tract
      266.97.2796.87.40.970.62 – 1.52
      Lymphoproliferative225.96.1272.32.52.421.38 – 4.24
      NHL143.73.990.80.84.612.00 – 10.65
      T-cell NHL51.31.420.20.27.461.45 – 38.46
      Breast
      Analyzed only for females
      83.93.5345.64.90.700.33 – 1.52
      Gynecological
      Analyzed only for females
      21.00.9182.92.60.330.08 – 1.44
      Prostate
      Analyzed only for males
      21.21.5173.14.40.340.08 – 1.47
      Central nervous system diseases133.53.6181.61.72.141.05 – 4.36
      Alzheimer's disease308.08.41059.10.80.830.56 – 1.25
      Digestive diseases184.85.0353.03.31.520.86 – 2.68
      Cardiovascular diseases10527.929.130526.428.41.020.82 – 1.27
      Ischemic heart disease7219.119.921218.419.81.010.77 – 1.32
      Cerebrovascular diseases215.65.8907.88.40.690.43 – 1.11
      Respiratory diseases164.34.4585.05.40.820.47 – 1.42
      Infectious diseases30.80.8151.31.40.600.17 – 2.06
      Accidents and violence256.66.9736.36.81.020.65 – 1.60
      Alcohol related diseases30.80.8292.52.70.310.09 – 1.00
      CI, confidence interval; HR, hazard ratio
      a Follow-up for overall mortality until August 10, 2020 and for cause-specific mortality until December 31, 2018
      b Includes all celiac disease phenotypes including dermatitis herpetiformis
      c Percentage of all deaths in the group
      d Mortality rate per 10,000 person-years
      e Including cancers of the liver, pancreas, and gastrointestinal tract
      f Analyzed only for females
      g Analyzed only for males
      Table 3All-cause mortality in celiac disease subgroups and compared to reference individuals.
      GroupNumberNumber of events (%)
      Celiac cohortReference individualsCeliac cohortReference individualsHR (95% CI)
      All1,3924,177378 (27.2)1,159 (27.7)0.96 (0.85 – 1.08)
      Gender
      Male5121,538172 (33.6)544 (35.4)0.92 (0.77 – 1.09)
      Female8802,639204 (23.2)611 (23.2)0.99 (0.85 – 1.16)
      Age at celiac disease diagnosis (years)
      < 2026479210 (3.8)19 (2.4)1.55 (0.72 – 3.34)
      20 – 394501,35138 (8.4)142 (10.5)0.78 (0.55 – 1.12)
      40 – 594921,477169 (34.3)523 (35.4)0.92 (0.77 – 1.09)
      ≥ 60186557159 (85.5)471 (84.6)1.02 (0.86 – 1.23)
      Decade of diagnosis
      – 198018154459 (32.6)299 (36.8)0.82 (0.61 – 1.10)
      1981 – 19906021,806173 (28.7)541 (30.0)0.96 (0.81 – 1.13)
      1991 – 20006091,827144 (23.6)414 (22.7)1.05 (0.96 – 1.26)
      Presenting phenotype at celiac disease diagnosis
      HR for a combination of celiac disease phenotypes excluding dermatitis herpetiformis 0.99, 95% CI 0.87 – 1.14
      Gastrointestinal4431,32995 (21.4)318 (23.9)0.86 (0.69 – 1.08)
      Malabsorption22166355 (24.9)168 (25.3)0.97 (0.72 – 1.32)
      Dermatitis herpetiformis299897107 (35.8)344 (38.4)0.87 (0.70 – 1.09)
      Other extraintestinal18254638 (20.9)109 (20.0)1.04 (0.82 – 1.51)
      Screen detected15345931 (20.3)74 (16.1)1.30 (0.85 – 1.97)
      Degree of villous atrophy at celiac disease diagnosis
      Normal8625839 (45.3)121 (46.9)0.83 (0.58 – 1.20)
      Partial villous atrophy32296481 (25.2)261 (27.1)0.90 (0.70 – 1.15)
      Subtotal or total villous atrophy8612,583186 (21.6)575 (22.3)0.95 (0.80 – 1.12)
      Response to gluten-free diet
      Yes8242,472196 (23.8)609 (24.6)0.94 (0.80 – 1.10)
      No7923625 (31.6)69 (29.2)1.04 (0.66 – 1.64)
      Follow-up from celiac disease diagnosis
      < 2 years1,3924,17714 (1.0)54 (1.3)0.97 (0.53 – 1.76)
      2 – 4.9 years1,3764,12027 (2.0)73 (1.7)1.18 (0.75 – 1.85)
      5 – 14.9 years1,3494,041113 (8.1)328 (7.9)1.13 (0.91 – 1.40)
      15 – 29.9 years1,2353,706173 (12.4)542 (13.0)0.96 (0.81 – 1.14)
      ≥ 30 years4901,41949 (3.5)158 (3.8)0.87 (0.63 – 1.20)
      CI, confidence interval; HR, hazard ratio
      a HR for a combination of celiac disease phenotypes excluding dermatitis herpetiformis 0.99, 95% CI 0.87 – 1.14
      Assessment of specific causes of death throughout follow-up revealed an increase in mortality from lymphoproliferative diseases (HR 2.42, 95% CI 1.38 – 4.24), and especially from non-Hodgkins lymphoma (HR 4.61, 95% CI 1.20 – 10.65) and T-cell non-Hodgkins lymphoma (HR 7.46, 95% CI 1.45 – 38.46). The increased risk was apparent particularly in patients diagnosed with celiac disease after 40 years of age (Table 4), while no cases of lymphoproliferative diseases were detected among patients diagnosed with celiac disease before 30 years of age. Mortality from diseases affecting the central nervous system (not including Alzheimer's disease) was increased in celiac disease (HR 2.14, 95% CI 1.05 – 4.36) due to various causes; 6 Parkinson's disease, 2 Amyotrophic lateral sclerosis, 2 familial spinocerebellar ataxia (siblings), 2 other specified neurodegenerative diseases, and 1 Lewy body dementia. Conversely, risk of dying from alcohol related diseases was decreased in the celiac cohort (HR 0.31, 95% CI 0.09 – 1.00). Mortality from malignancies overall and from gastrointestinal malignancies, as well as from cardiovascular diseases and digestive diseases was not increased among celiac disease patients.
      Table 4Specific relative mortality risks in celiac disease patients
      Involves all celiac disease phenotypes including dermatitis herpetiformis.
      compared to reference individuals by time since and by age at celiac disease diagnosis
      Malignancies overallLymphoproliferative diseasesCentral nervous system diseases
      N CD / RefHR (95% CI)N CD / RefHR (95% CI)N CD / RefHR (95% CI)
      Time from diagnosis
      < 2 years6 / 171.10 (0.42 – 2.86)3 / 19.36 (0.92 – 94.84)0 / 1-
      2 – 4.9 years8 / 181.34 (0.58 – 3.11)5 / 114.50 (1.69 – 124.70)0 / 1-
      5 – 14.9 years25 / 930.87 (0.56 – 1.36)7 / 92.27 (0.85 – 6.10)6 / 212.91 (2.49 – 67.07)
      15 – 29.9 years40 / 1450.83 (0.59 – 1.18)7 / 111.92 (0.75 – 4.96)7 / 131.62 (0.65 – 4.07)
      ≥ 30 years9 / 370.69 (0.33 – 1.42)0 / 5-0 / 1-
      Age at diagnosis
      < 20 years1 / 40.72 (0.08 – 6.48)0 / 3-0 / 1-
      20 – 39 years11 / 470.69 (0.36 – 1.32)3 / 81.10 (0.29 – 4.14)3 / 32.97 (0.60 – 14.71)
      40 – 59 years43 / 1530.82 (0.58 – 1.15)7 / 45.20 (1.52 – 17.75)7 / 121.68 (0.66 – 4.28)
      > 60 years33 / 1060.96 (0.65 – 1.41)12 / 123.05 (1.37 – 6.79)3 / 24.62 (0.77 – 27.66)
      CI, confidence interval; HR, hazard ratio
      a Involves all celiac disease phenotypes including dermatitis herpetiformis.
      In the stratified analysis assessing the development of trend in mortality over time after celiac disease diagnosis, mortality from lymphoproliferative diseases increased significantly 2 – 5 years after diagnosis and from diseases of the central nervous system 5 – 15 years after diagnosis and diminished in both thereafter (Table 4). No alterations were noted in mortality over time from malignancies overall (Table 4) nor from gastrointestinal malignancies or cardiovascular diseases (data not shown).

      4. Discussion

      Celiac disease-related long-term mortality showed no increase in this study, which was conducted in a well-defined large cohort of celiac disease patients within a very long follow-up. Furthermore, mortality risk remained largely the same during the entire follow-up. This is a reassuring finding, as over half of the celiac disease patients in our series were diagnosed before the age of 40 years, thus having a long life expectancy ahead of them. It could also be interpreted as an indication of the beneficial effects of a long-lasting gluten-free diet. Even though we were not able to evaluate the long-term adherence to gluten-free diet in our cohort, the 91% response rate observed in duodenal control biopsies suggests reasonable compliance with the diet. In addition, the response rate is also in line with those of earlier studies showing good long-term compliance in Tampere celiac disease (88-93%) and dermatitis herpetiformis (98%) cohorts [
      • Hervonen K
      • Alakoski A
      • Salmi TT
      • et al.
      Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
      ,
      • Viljamaa M
      • Collin P
      • Huhtala H
      • et al.
      Is coeliac disease screening in risk groups justified? A fourteen-year follow-up with special focus on compliance and quality of life.
      ,
      • Kurppa K
      • Lauronen O
      • Collin P
      • et al.
      Factors associated with dietary adherence in celiac disease: a nationwide study.
      ]. On the other hand, the beneficial effect of gluten-free diet on overall mortality is not self-evident as a strict diet is not nutritionally ideal. The diet itself may be associated with lower fiber, vitamin, and mineral intake as well as higher sugar and fat consumption [
      • See JA
      • Kaukinen K
      • Makharia GK
      • et al.
      Practical insights into gluten-free diets.
      ], which may in turn predispose to other comorbidities like obesity, fatty liver, cardiovascular diseases, and even cancer [
      • Ilus T
      • Kaukinen K
      • Virta LJ
      • et al.
      Incidence of malignancies in diagnosed celiac patients: a population-based estimate.
      ,
      • Rispo A
      • Imperatore N
      • Guarino M
      • et al.
      Metabolic-associated fatty liver disease (MAFLD) in coeliac disease.
      ].
      Our mortality result contradicts the higher mortality (1.3- to 2-fold) reported by many others [
      • Corrao G
      • Corazza GR
      • Bagnardi V
      • et al.
      Mortality in patients with coeliac disease and their relatives: a cohort study.
      ,
      • Quarpong W
      • Card TR
      • West J
      • et al.
      Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
      ,
      • Holmes GKT
      • Muirhead A.
      Mortality in coeliac disease: a population-based cohort study from a single centre in Southern Derbyshire, UK.
      ,
      • West J
      • Logan RFA
      • Smith CJ
      • et al.
      Malignancy and mortality in people with coeliac disease: population based cohort study.
      ,
      • Peters U
      • Askling J
      • Gridley G
      • et al.
      Causes of death in patients with celiac disease in a population-based Swedish cohort.
      ,
      • Schneider C.V.
      • Kleinjans M
      • Fromme M
      • et al.
      Phenome-wide association study in adult coeliac disease: role of HLA subtype.
      ] (Supplementary Table 1). In a recent large Swedish study the mortality risk observed was elevated (HR 1.2) [
      • Lebwohl B
      • Green PHR
      • Söderling J
      • et al.
      Association between celiac disease and mortality risk in a Swedish population.
      ], but the mortality rate detected among celiac disease patients (9.7%) was similar to ours (9.8%). However, mortality among Swedish reference individuals was lower than in our reference subjects (8.6% vs. 10.2% respectively) indicating some differences between the reference groups of these two studies despite similar matching parameters. Differences in celiac disease presentation and symptom severity could also offer one potential explanation for our lower risk estimate. Most studies investigating mortality in undetected, typically mild or asymptomatic celiac disease reported no excess risk while the risk was higher in patients with severe symptoms [
      • Corrao G
      • Corazza GR
      • Bagnardi V
      • et al.
      Mortality in patients with coeliac disease and their relatives: a cohort study.
      ,
      • Tio M
      • Cox MR
      • Eslick GD.
      Meta-analysis: Coeliac disease and the risk of all-cause mortality, any malignancy and lymphoid malignancy.
      ,
      • Lohi S
      • Mäki M
      • Rissanen H
      • et al.
      Prognosis of unrecognized coeliac disease as regards mortality: a population-based cohort study.
      ,
      • Godfrey JD
      • Brantner TL
      • Brinjikji W
      • et al.
      Morbidity and mortality among older individuals with undiagnosed celiac disease.
      ,
      • Kårhus LL
      • Skaaby T
      • Petersen J
      • et al.
      Long-term consequences of undiagnosed celiac seropositivity.
      ,
      • Biagi F
      • Schiepatti A
      • Maiorano G
      • et al.
      Risk of complications in coeliac patients depends on age at diagnosis and type of clinical presentation.
      ]. Lower mortality risks have also been reported in patients with dermatitis herpetiformis (SMR 0.5 – 0.9) [
      • Hervonen K
      • Alakoski A
      • Salmi TT
      • et al.
      Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
      ,
      • Swerdlow AJ
      • Whittaker S
      • Carpenter LM
      • et al.
      Mortality and cancer incidence in patients with dermatitis herpetiformis: a cohort study.
      ,
      • Viljamaa M
      • Kaukinen K
      • Pukkala E
      • et al.
      Malignancies and mortality in patients with coeliac disease and dermatitis herpetiformis: 30-year population-based study.
      ,
      • Lewis NR
      • Logan RFA
      • Hubbard RB
      • et al.
      No increase in risk of fracture, malignancy or mortality in dermatitis herpetiformis: a cohort study.
      ].
      Although overall cancer mortality was not shown to be increased in the present study, the risk of death from lymphoproliferative diseases and especially from non-Hodgkins lymphoma was elevated. Nevertheless, our risk estimates were lower than those reported earlier in patient cohorts from the same era [
      • Corrao G
      • Corazza GR
      • Bagnardi V
      • et al.
      Mortality in patients with coeliac disease and their relatives: a cohort study.
      ,
      • Quarpong W
      • Card TR
      • West J
      • et al.
      Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
      ,
      • Holmes GKT
      • Muirhead A.
      Mortality in coeliac disease: a population-based cohort study from a single centre in Southern Derbyshire, UK.
      ,
      • Viljamaa M
      • Kaukinen K
      • Pukkala E
      • et al.
      Malignancies and mortality in patients with coeliac disease and dermatitis herpetiformis: 30-year population-based study.
      ,
      • Peters U
      • Askling J
      • Gridley G
      • et al.
      Causes of death in patients with celiac disease in a population-based Swedish cohort.
      ]. Moreover, the risk was more pronounced particularly in patients diagnosed when over 40 years of age and within 2 – 5 years of diagnosis. In comparison, Quarpong et al. [
      • Quarpong W
      • Card TR
      • West J
      • et al.
      Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
      ] detected increased mortality from lymphoproliferative diseases both in childhood- and adulthood diagnosed celiac disease and despite a decreasing trend in risk over time it remained almost significantly increased even after 15 years of follow-up. Since the 1990s the widespread utilization of serological diagnostic tests [
      • Collin P
      • Hällström O
      • Mäki M
      • et al.
      Atypical coeliac disease found with serologic screening.
      ] together with investments in educating health care personnel about celiac disease have succeeded in reducing diagnostic delay [
      • Ukkola A
      • Mäki M
      • Kurppa K
      • et al.
      Diet improves perception of health and well-being in symptomatic, but not asymptomatic, patients with celiac disease.
      ] as well as a high prevalence of (0.7%) biopsy-proven celiac disease in Finland [
      • Virta LJ
      • Kaukinen K
      • Collin P
      Incidence and prevalence of diagnosed coeliac disease in Finland: results of effective case finding in adults.
      ] possibly influencing the mortality risks identified.
      The increase in mortality from central nervous system diseases was attributable to various diseases. A similar risk estimate was observed by Peters et al. in a Swedish celiac inpatient cohort [
      • Peters U
      • Askling J
      • Gridley G
      • et al.
      Causes of death in patients with celiac disease in a population-based Swedish cohort.
      ]. Neurological dysfunction has been shown in up to 67% of newly diagnosed celiac disease patients, the most frequent symptoms being headache, ataxia, and peripheral neuropathy [
      • Hadjivassiliou M
      • Croall ID
      • Grünewald RA
      • et al.
      Neurological evaluation of patients with newly diagnosed coeliac disease presenting to gastroenterologists: A 7-year follow-up study.
      ]. A possible link to neurological symptoms are the cross reacting transglutaminase antibodies in the brain tissue [
      • Hadjivassiliou M
      • Croall ID
      • Grünewald RA
      • et al.
      Neurological evaluation of patients with newly diagnosed coeliac disease presenting to gastroenterologists: A 7-year follow-up study.
      ,
      • Osman D
      • Umar S
      • Muhammad H
      • et al.
      Neurological manifestation of coeliac disease with particular emphasis on gluten ataxia and immunological injury: a review article.
      ]. In the 1990s, collaboration with the neurology clinic was intensive in our hospital, possibly improving detection of celiac disease patients with neurological symptoms [
      • Luostarinen L
      • Pirttilä T
      • Collin P
      Coeliac disease presenting with neurological disorders.
      ], which may have an impact on our increased mortality rate from these disorders.
      The decreased mortality from alcohol-related causes is a novel finding in celiac disease while in dermatitis herpetiformis it has already been noted (SMR 0.0) [
      • Hervonen K
      • Alakoski A
      • Salmi TT
      • et al.
      Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
      ]. The literature on alcohol consumption among celiac disease patients is scanty. Reilly et al. [
      • Reilly NR
      • Lebwohl B
      • Hultcrantz R
      • et al.
      Increased risk of non-alcoholic fatty liver disease after diagnosis of celiac disease.
      ] showed a smaller proportion of alcohol-related liver disease among celiac disease patients with non-alcoholic fatty liver disease (NAFLD) compared to controls with NAFLD (12% vs. 15%) and Hervonen et al. [
      • Hervonen K
      • Alakoski A
      • Salmi TT
      • et al.
      Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
      ] discovered a markedly reduced consumption of beer, but not of strong drinks, among dermatitis herpetiformis patients. In Finland, beer is one of the most widely used alcohol types and avoidance of gluten containing products could lead to diminished beer consumption and further to decreased mortality.
      A major strength of our study is the long follow-up of up to 41 years, which enabled us to assess the evolution of mortality in celiac disease in a wider time perspective and also for patients diagnosed in childhood. Another strength is the large and prospectively collected cohort of well-defined celiac disease patients with access to patient records for detailed patient and celiac disease-related information. This allowed a reasonable estimate of mortality in different celiac phenotypes and diagnostic age-groups. At the time our cohort was collected, the diagnostics of celiac disease and dermatitis herpetiformis were primarily centralized to Tampere University Hospital and thus the cohort includes all cases diagnosed during that period. We also had a matched reference group, enabling more precise assessment of relative mortality than comparison to general population [
      • Card TR
      • Solaymani-Dodaran M
      • Hubbard R
      • et al.
      Is an internal comparison better than using national data when estimating mortality in longitudinal studies?.
      ].
      Our study also suffered from some limitations. The information on celiac disease characteristics was collected from patient records and not all the data required were available anymore. We had no information on confounding factors like socioeconomic status, tobacco consumption or concomitant diseases possibly affecting mortality. Furthermore, despite the large cohort of celiac disease patients, the number of deaths remained low in many stratified groups and causes of death, and therefore many analyses were likely underpowered.

      5. Conclusions

      In conclusion, celiac disease-associated mortality showed no increase in long-term follow-up in a patient cohort covering the whole spectrum of celiac disease presentations. Nevertheless, mortality from lymphoproliferative diseases and non-Hodgkins lymphoma was increased and the risk was most prominent when celiac disease was diagnosed after 40 years of age. Our findings suggest that the prognosis is generally good for celiac disease patients adhering to gluten-free diet.

      Conflict of interest

      None declared.

      Funding

      This study was supported by grants to the Celiac Disease Study Group from the Sigrid Juselius Foundation, the Emil Aaltonen Foundation, the Academy of Finland, the Research Fund of the Finnish Coeliac Society, the Finnish Cultural Foundation and the Competitive State Research Financing of the Expert Responsibility Area of Tampere University Hospital.

      Appendix. Supplementary materials

      References

        • Corrao G
        • Corazza GR
        • Bagnardi V
        • et al.
        Mortality in patients with coeliac disease and their relatives: a cohort study.
        Lancet. 2001; 358: 356-361
        • Tio M
        • Cox MR
        • Eslick GD.
        Meta-analysis: Coeliac disease and the risk of all-cause mortality, any malignancy and lymphoid malignancy.
        Aliment Pharmacol Ther. 2012; 35: 540-551
        • Quarpong W
        • Card TR
        • West J
        • et al.
        Mortality in people with coeliac disease: Long-term follow-up from a Scottish cohort.
        United Eur Gastroenterol J. 2019; 7: 377-387
        • Holmes GKT
        • Muirhead A.
        Mortality in coeliac disease: a population-based cohort study from a single centre in Southern Derbyshire, UK.
        BMJ Open Gastroenterol. 2018; 5e000201
        • Lebwohl B
        • Green PHR
        • Söderling J
        • et al.
        Association between celiac disease and mortality risk in a Swedish population.
        JAMA. 2020; 323: 1277-1285
        • West J
        • Logan RFA
        • Smith CJ
        • et al.
        Malignancy and mortality in people with coeliac disease: population based cohort study.
        BMJ. 2004; 329: 716-718
        • Catassi C
        • Fabiani E
        • Corrao G
        • et al.
        Risk of non-Hodgkin lymphoma in celiac disease.
        J Am Med Assoc. 2002; 287: 1413-1419
        • Koskinen I
        • Virta LJ
        • Huhtala H
        • et al.
        Overall and cause-specific mortality in adult celiac disease and dermatitis herpetiformis diagnosed in the 21st century.
        Am J Gastroenterol. 2020; 115: 1117-1124
        • Abdul Sultan A
        • Crooks CJ
        • Card T
        • et al.
        Causes of death in people with coeliac disease in England compared with the general population: a competing risk analysis.
        Gut. 2015; 64: 1220-1226
        • Hervonen K
        • Alakoski A
        • Salmi TT
        • et al.
        Reduced mortality in dermatitis herpetiformis: a population-based study of 476 patients.
        Br J Dermatol. 2012; 167: 1331-1337
        • Swerdlow AJ
        • Whittaker S
        • Carpenter LM
        • et al.
        Mortality and cancer incidence in patients with dermatitis herpetiformis: a cohort study.
        Br J Dermatol. 1993; 129: 140-144
        • Viljamaa M
        • Kaukinen K
        • Pukkala E
        • et al.
        Malignancies and mortality in patients with coeliac disease and dermatitis herpetiformis: 30-year population-based study.
        Dig Liver Dis. 2006; 38: 374-380
        • Lohi S
        • Mäki M
        • Rissanen H
        • et al.
        Prognosis of unrecognized coeliac disease as regards mortality: a population-based cohort study.
        Ann Med. 2009; 41: 508-515
        • Godfrey JD
        • Brantner TL
        • Brinjikji W
        • et al.
        Morbidity and mortality among older individuals with undiagnosed celiac disease.
        Gastroenterology. 2010; 139: 763-769
        • Kårhus LL
        • Skaaby T
        • Petersen J
        • et al.
        Long-term consequences of undiagnosed celiac seropositivity.
        Am J Gastroenterol. 2020; 115: 1681-1688
        • Ludvigsson JF
        • Montgomery SM
        • Ekbom A
        • et al.
        Small-intestinal histopathology and mortality risk in celiac disease.
        JAMA. 2009; 302: 1171-1178
        • Ilus T
        • Kaukinen K
        • Virta LJ
        • et al.
        Refractory coeliac disease in a country with a high prevalence of clinically-diagnosed coeliac disease.
        Aliment Pharmacol Ther. 2014; 39: 418-425
        • Koskinen I
        • Hervonen K
        • Pukkala E
        • et al.
        Cancer incidence and factors associated with malignancies in coeliac disease during long-term follow-up.
        GastroHep. 2021; 3: 107-115
        • Marsh MN.
        Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity ('celiac sprue’).
        Gastroenterology. 1992; 102: 330
        • Oberhuber G.
        Histopathology of celiac disease.
        Biomed Pharmacother. 2000; 54: 368-372
      1. Statistics Finland - Quality Description:Causes of death 2016. (https://tilastokeskus.fi/til/ksyyt/2016/ksyyt_2016_2017-12-29_laa_001_en.html) accessed November 5, 2021.

      2. Statistics Finland: Causes of death, time series classification with 54 categories. (https://www.tilastokeskus.fi/til/ksyyt/ksyyt_2021-10-29_luo_001_en.pdf) accessed January 24, 2022.

        • Viljamaa M
        • Collin P
        • Huhtala H
        • et al.
        Is coeliac disease screening in risk groups justified? A fourteen-year follow-up with special focus on compliance and quality of life.
        Aliment Pharmacol Ther. 2005; 22: 317-324
        • Kurppa K
        • Lauronen O
        • Collin P
        • et al.
        Factors associated with dietary adherence in celiac disease: a nationwide study.
        Digestion. 2013; 86: 309-314
        • See JA
        • Kaukinen K
        • Makharia GK
        • et al.
        Practical insights into gluten-free diets.
        Nat Rev Gastroenterol Hepatol. 2015; 12: 580-591
        • Ilus T
        • Kaukinen K
        • Virta LJ
        • et al.
        Incidence of malignancies in diagnosed celiac patients: a population-based estimate.
        Am J Gastroenterol. 2014; 109: 1471-1477
        • Rispo A
        • Imperatore N
        • Guarino M
        • et al.
        Metabolic-associated fatty liver disease (MAFLD) in coeliac disease.
        Liver Int. 2021; 41: 788-798
        • Peters U
        • Askling J
        • Gridley G
        • et al.
        Causes of death in patients with celiac disease in a population-based Swedish cohort.
        Arch Int Med. 2003; 163: 1566-1572
        • Schneider C.V.
        • Kleinjans M
        • Fromme M
        • et al.
        Phenome-wide association study in adult coeliac disease: role of HLA subtype.
        Aliment Pharmacol Ther. 2021; 53: 510-518
        • Biagi F
        • Schiepatti A
        • Maiorano G
        • et al.
        Risk of complications in coeliac patients depends on age at diagnosis and type of clinical presentation.
        Dig Liver Dis. 2018; 50: 549-552
        • Lewis NR
        • Logan RFA
        • Hubbard RB
        • et al.
        No increase in risk of fracture, malignancy or mortality in dermatitis herpetiformis: a cohort study.
        Aliment Pharmacol Ther. 2008; 27: 1140-1147
        • Collin P
        • Hällström O
        • Mäki M
        • et al.
        Atypical coeliac disease found with serologic screening.
        Scand J Gastroenterol. 1990; 25: 245-250
        • Ukkola A
        • Mäki M
        • Kurppa K
        • et al.
        Diet improves perception of health and well-being in symptomatic, but not asymptomatic, patients with celiac disease.
        Clin Gastroenterol Hepatol. 2011; 9 (e1): 118-123
        • Virta LJ
        • Kaukinen K
        • Collin P
        Incidence and prevalence of diagnosed coeliac disease in Finland: results of effective case finding in adults.
        Scand J Gastroenterol. 2009; 44: 933-938
        • Hadjivassiliou M
        • Croall ID
        • Grünewald RA
        • et al.
        Neurological evaluation of patients with newly diagnosed coeliac disease presenting to gastroenterologists: A 7-year follow-up study.
        Nutrients. 2021; 13
        • Osman D
        • Umar S
        • Muhammad H
        • et al.
        Neurological manifestation of coeliac disease with particular emphasis on gluten ataxia and immunological injury: a review article.
        Gastroenterol Hepatol Bed Bench. 2021; 14: 1-7
        • Luostarinen L
        • Pirttilä T
        • Collin P
        Coeliac disease presenting with neurological disorders.
        Eur Neurol. 1999; 42: 132-135
        • Reilly NR
        • Lebwohl B
        • Hultcrantz R
        • et al.
        Increased risk of non-alcoholic fatty liver disease after diagnosis of celiac disease.
        J Hepatol. 2015; 62: 1405-1411
        • Card TR
        • Solaymani-Dodaran M
        • Hubbard R
        • et al.
        Is an internal comparison better than using national data when estimating mortality in longitudinal studies?.
        J Epidemiol Community Health. 2006; 60: 819-821