Secondary Logo

Journal Logo


Collagenous Gastritis in Children: Incidence, Disease Course, and Associations With Autoimmunity and Inflammatory Markers

Käppi, Timo MD1,2; Wanders, Alkwin MD, PhD3,4; Wolving, Mats MD5; Lingblom, Christine PhD6; Davidsson Bården, Birgitta MD7; Arkel, Rikard MD8; Hätting, Josefine MD9; Anderzén, Johan MD10; Wennerås, Christine MD, PhD6,11; Saalman, Robert MD, PhD1,2

Author Information
Clinical and Translational Gastroenterology: August 2020 - Volume 11 - Issue 8 - p e00219
doi: 10.14309/ctg.0000000000000219



Collagenous gastritis (CG) is a rare gastrointestinal disorder with fewer than 300 cases reported in the English-language literature (1–24). Of these cases, about one-third have been childhood-onset CG (15–45). The condition is characterized histologically by an increased subepithelial layer of collagen (conventionally defined as being >10 μm in thickness) in the gastric mucosa, together with an inflammatory cell infiltrate in the lamina propria (46,47). In most pediatric cases of CG, the collagenous mucosal inflammation is restricted to the stomach (22–24,41), whereas in adult-onset disease, concurrent involvement of the small bowel and/or colon is more common (1,23,24,46,48). Pediatric cases of CG generally present with severe iron deficiency anemia and/or recurrent abdominal pain (22–24,41), whereas diarrhea and malabsorption are the predominant symptoms in adults, presumably linked to the frequently observed concurrent intestinal involvement (32,46,49). Associated immune-mediated comorbidities, such as celiac disease and type 1 diabetes, have been reported in both pediatric (22–24,26,31) and adult (3–5,23,24,45,47,50) patients. CG is believed to be a chronic disease, and there is currently no effective treatment (46).

The current literature on childhood-onset CG consists mainly of case reports (15–20,27–37,42,43,51), apart from 6 small case series (21,22,38–41) and 3 histopathologic studies with limited clinical information and follow-up data (23,24,45). Consequently, the knowledge regarding the evolution of the clinical, endoscopic, and histologic features of the disease over time is sparse. Furthermore, although immune-mediated/autoimmune disease mechanisms have been hypothesized (47), convincing supporting evidence is lacking. In collagenous colitis, the human leukocyte antigen (HLA) DQ2.5 haplotype, encoded by the HLA DQ2/DQ8 genes, seems to be associated with increased disease susceptibility (52). However, no studies of the prevalence of the HLA DQ2/DQ8 haplotype in CG have been published. Moreover, the methodologies used in the previous studies of CG have not allowed estimations of the disease incidence or prevalence, although a possible female predominance has been noted (1,41).

The aims of this population-based cohort study, which combines longitudinal and cross-sectional approaches and involves 15 patients with childhood-onset CG, were to investigate: (i) the incidence and prevalence of CG in a pediatric population in western Sweden; (ii) the clinical, endoscopic, and histologic characteristics of childhood-onset CG through the course of the disease; and (iii) the frequencies of autoimmune comorbidities and heredity, as well as the prevalences of autoantibody development, increased blood inflammatory biomarkers, and the HLA DQ2/DQ8 haplotypes in patients with childhood-onset CG.


Study design

The study was designed as a population-based cohort study that comprised 15 persons of White ethnicity with childhood-onset CG (12 female and 3 male patients) and age range of 8.7–23 years (median age, 15 years) recruited from western Sweden. All cases of CG diagnosed before the age of 18 years in the counties of Halland, Jönköping, Värmland, and Västra Götaland in western Sweden during the period of January 2008 through June 2019 were identified. These 4 counties comprise 26% of the Swedish population and in 2019 had a pediatric population (aged <18 years) of approximately 568,000 (53). The Department of Pediatric Gastroenterology, Hepatology, and Nutrition at Queen Silvia Children's Hospital, Gothenburg, Sweden, which serves as a tertiary referral center for these counties, was involved in the diagnosis of all the cases. Furthermore, an established practice in Sweden is that all pediatric endoscopies are performed under general anesthesia in county hospitals or higher-level medical institutions within the public healthcare system. This enables the identification of all eligible cases within the geographic area covered by the study. The recruitment of subjects was carried out at a clinical follow-up visit during the period of May 2019 to November 2019, at which time point all the blood samples were drawn. All eligible patients accepted to participate in the study.

Diagnostic criteria

The diagnosis of CG was based on histologic demonstration of increased (>10 μm) subepithelial collagen deposition in at least 1 biopsy taken from gastric mucosa during endoscopy, in addition to other histologic, endoscopic, and clinical findings supporting the diagnosis (32,47). The day when the first endoscopic examination that met the diagnostic criteria was performed was defined as the date of diagnosis.

Epidemiological analyses

The incidence rate of childhood-onset CG for the period spanning January 2008 to June 2019 was calculated according to the recommended principles for dynamic populations (54). To define the population at risk, the data on the annual numbers of children aged younger than 18 years registered in the counties of Halland, Jönköping, Värmland, and Västra Götaland during the period 2008–2019 were retrieved from the Swedish Government agency Statistics Sweden. The prevalence of CG in the population aged younger than 18 years in the 4 counties was calculated as of June 2019.

Clinical, endoscopic, and histologic data

The complete medical records and endoscopic reports for all the patients were reviewed. Demographic data, existing comorbidities, clinical symptoms related to CG, and outcomes from tested treatment modalities were retrieved during the follow-up consultations. Furthermore, heredity for autoimmune and other diseases among first-degree relatives was systemically evaluated.

Regarding the histopathologic evaluation, all cases had initially been re-evaluated by a senior gastrointestinal pathologist (M.W.) at the Department of Pathology, Sahlgrenska University Hospital, Gothenburg, Sweden, to confirm the diagnosis. For the purpose of this study, the gastric mucosal biopsies from all gastrointestinal endoscopies performed on study subjects during the follow-up were reviewed independently by a different senior gastrointestinal pathologist (A.W.). In this evaluation, the specific histologic parameters recently proposed for CG, such as the location of collagen deposits, the grade of active and chronic inflammation, and the presence of eosinophil- or lymphocyte-predominant type of inflammation in the gastric mucosa, were applied (23). In addition, the maximum number of intraepithelial eosinophilic granulocytes per high-power field (size, 0.24 mm2) was determined. Moreover, the maximal thicknesses of the subepithelial collagen deposits in gastric mucosal biopsies were measured using an ocular micrometer in the microscope and collagen-specific stains (i.e., Masson trichrome, Van Gieson, or Sirius Red). Furthermore, mucosal biopsies acquired from other sites in the gastrointestinal tract were also reviewed, and any diagnostic abnormalities were recorded.

Laboratory parameters

All the blood laboratory parameters were analyzed on inclusion of the patient in the study, and they included 11 autoantibody specificities and 4 inflammatory markers. The following autoantibodies were measured in the serum samples: antinuclear antibody (ANA), antismooth muscle antibody, antimitochondrial antibody, antineutrophil cytoplasmic antibody, tissue transglutaminase IgA, antithyrotropin receptor antibodies, antithyroid peroxidase antibodies, antithyroglobulin antibodies, antiglutamic acid decarboxylase antibodies, antiintrinsic factor antibodies, and antiparietal cell antibodies. The blood inflammatory markers analyzed were plasma C-reactive protein (CRP), serum calprotectin, serum amyloid A (SAA), and serum orosomucoid. The values of all the inflammatory parameters were continuous, with the exception of SAA, which had a detection limit of 11 mg/L with the analysis method used.

In addition, genotyping of HLA DQ2/DQ8 was performed. Moreover, the serum levels of albumin, blood total leukocyte counts, and blood absolute neutrophil counts, as well as the total serum levels of immunoglobulins A, M, and G (including IgG subclasses), were measured. The laboratory methods used and the cutoff levels and reference intervals applied in the study are listed in Tables S1a and S1b (see Supplementary Digital Content 1, All blood samples were drawn after a 12-hour period of fasting and analyzed in accredited laboratories at Sahlgrenska University Hospital, Gothenburg, Sweden, or at affiliated county hospitals. The blood analysis laboratory data were complete for all the study participants, with the exception of a blood absolute neutrophil count that was missing for a single patient.


Written informed consent was obtained from all study participants and/or their parents. The study was approved by the Regional Ethical Review Board of Gothenburg, Gothenburg, Sweden (Dnr. 298-12).

Statistical methods

Owing to the limited sample size and non-normal distribution of the variables, the data are presented as medians with ranges. To calculate the confidence interval for the incidence rate ratio, a 2-sample z-test was applied. The Spearman test was used for determination of correlations between laboratory parameters. A 2-tailed P-value <0.05 was considered statistically significant. The z-test was performed using Microsoft Excel for Mac 2011 version 14.7.7 software (Microsoft, Redmond, WA). For all other statistical analyses, IBM SPSS Statistics version 22 software (IBM, New York, NY) was used.


Patient characteristics

The median age at the diagnosis of CG was 9.9 years (range, 4.4–17 years), whereas the median age at the initial presentation was 9.8 years (range, 4.3–16 years). The median duration of disease, measured from the initial presentation, was 6.2 years (range, 0.4–11 years) on inclusion in the study. The collagenous inflammation was restricted to the stomach in all, but 2 patients (13/15, 87%), and associated collagenous duodenitis and collagenous colitis were seen in 1 patient each. No patient had an established current or previous comorbidity involving other gastrointestinal, autoimmune, or inflammatory diseases. Allergic disease was seen in 2 (13%) patients, both of whom suffered from allergic rhinoconjunctivitis. Apart from 1 patient with concomitant collagenous colitis who was receiving budesonide therapy and 3 other patients who were being treated with proton pump inhibitors, the remaining 11 patients were not receiving any anti-inflammatory or immunosuppressive therapy.

Incidence rate and prevalence

The annual number of persons aged younger than 18 years at risk for CG in the study's geographic area was in the range of 505,414–565,425 (corresponding to approximately 20% of the total population) during the study period and consisted predominantly of persons of Caucasian ancestry. The total number of person-years of follow-up (PYFU) in this population-based cohort was 6,020,927, which translates to an incidence rate of childhood-onset CG of 0.25 cases per 100,000 PYFU during the study period. Categorized by sex, the incidence rate for CG was 0.41 cases per 100,000 PYFU for girls and 0.097 cases per 100,000 PYFU for boys, yielding an incidence rate ratio (girls to boys) of 4.2 (95% confidence interval, 1.2–15). In June 2019, the prevalence of CG in children aged younger than 18 years was 2.1/100,000 in the counties of western Sweden.

Clinical picture and disease course

All but 1 patient had iron deficiency anemia on initial presentation, and in most of the cases, this was severe (Table 1). The chief complaint that prompted the patient to seek medical attention was related to the anemia (e.g., fatigue and pallor) in 5/15 patients (33%) or to gastrointestinal symptoms in 3/15 patients (20%). In the remaining 7 cases (47%), the iron deficiency anemia that initiated the diagnostic workup was an incidental finding, discovered when the patients were seeking medical attention for reasons unrelated to anemia or gastrointestinal symptoms, such as a respiratory infection.

Table 1.
Table 1.:
Chief complaints, blood hemoglobin levels, erythrocyte indices, and serum iron parameters at presentation of 15 patients with childhood-onset collagenous gastritis

In a targeted anamnesis, recurrent abdominal pain was reported by 6/15 patients (45%) on presentation, although in most of the cases, the pain was not severe and did not affect the activities of daily living. However, in the sole patient with associated collagenous colitis, the presenting symptoms were dominated by watery diarrhea. None of the patients showed hematemesis, melena, or hematochezia. Nonetheless, testing for fecal hemoglobin was carried out in 12/15 patients (80%) as part of the initial diagnostic workup. The testing did not indicate occult gastrointestinal bleeding in any of these patients. In 10 of the 12 girls in the study, all of whom had iron deficiency anemia at presentation, the diagnosis of CG was made before menarche. The remaining 2 girls reported no menorrhagia. At presentation, all the patients were on a regular diet, and no one was taking any medication.

All 14 patients who had iron deficiency anemia at presentation responded to iron supplementation with normalization of blood hemoglobin levels and iron parameters. After initial treatment, iron deficiency with or without associated anemia recurred within a median time of 1 year (range, 1–2 years) in 13 of the patients, most of whom continued to require repeated iron substitution subsequently during the follow-up. Orally delivered iron supplementation (ferrous asparto glycinate plus polysaccharide iron complex or iron sulphate) was the primary choice of therapy, and intravenous administration of iron (ferric carboxymaltose) was used primarily when oral intake of iron created bothersome gastrointestinal side effects, such as abdominal pain. Apart from the recurrence of iron deficiency, most of the patients remained asymptomatic during the median follow-up of 6.2 years (range, 1.0–11 years). Only 1 patient experienced recurrent abdominal pain that affected daily activities, whereas another patient who had concurrent collagen colitis experienced occasional bouts of watery diarrhea despite maintenance therapy with budesonide.

Treatment outcomes

Efforts were made to treat CG with proton pump inhibitors (n = 5) and dietary modifications (n = 7), such as cow's milk-free diet (n = 4), gluten-free diet (n = 2), and diet free of cow's milk, soy, egg, and wheat (n = 1). However, none of the patients showed endoscopic or histologic improvements during the treatment period (see Table S2, Supplementary Digital Content 2, Apart from the patient with concurrent collagenous colitis who was on maintenance therapy with budesonide, no steroids or immunomodulators were administered to any of the remaining 14 patients.

Endoscopic and histologic features

The endoscopic and histologic findings in the gastric mucosa in the patient cohort at the time of diagnosis are listed in Table 2. The endoscopic appearance of the gastric mucosa was judged to be abnormal in all, but 1 case. However, this patient had aberrant macroscopic findings in the stomach in follow-up endoscopies. The most commonly observed endoscopic features were mucosal nodularity, hypertrophic gastric folds, and erythema. Only 1 patient demonstrated endoscopic signs of mucosal bleeding from the stomach. Biopsy-based tests showed no signs of concurrent Helicobacter pylori gastritis in any patient.

Table 2.
Table 2.:
Endoscopic and histologic findings at the time of diagnosis of collagenous gastritis for the patients in the study cohort

Histologically, the maximal thickness of the subepithelial collagen deposits was in the range of 15–100 μm at diagnosis. Moreover, most (13/15, 87%) of the patients exhibited an eosinophil-rich (≥30 eosinophils/high-power field) inflammation of the gastric mucosa, with up to 150 eosinophils/high-power field. By contrast, intraepithelial lymphocytosis (>25 surface intraepithelial lymphocytes/100 epithelial cells) was seen in only 1 patient.

Duodenal mucosal biopsies had been acquired from all patients on diagnosis of CG, and 3 patients showed an associated subtotal or total villous atrophy with or without intraepithelial lymphocytosis (Table 2). One of these patients had collagen deposits in the duodenal bulb consistent with collagenous gastroduodenitis and tested negative for celiac disease in serologic tests. In the other 2 cases, no collagen deposits were detected in the duodenal biopsies (including follow-up endoscopies), and the serologic tests did not support a finding of celiac disease (see Table S2, Supplementary Digital Content 2, However, at the time of this study, a weak increase in the levels of tissue transglutaminase IgA antibodies had developed in one of these 2 patients, indicating that celiac disease was under development, thereby necessitating additional diagnostic work-up.

Follow-up gastroscopies with gastric and duodenal mucosal biopsies had been performed for 11/15 patients (73%), with the median time from the diagnosis to the latest gastroscopy being 4.4 years (range, 1.4–7.4 years). None of these patients had evidence of endoscopic or histologic normalization of the gastric mucosal pathology, including persistence of >10 μm-thick subepithelial collagen deposits in gastric mucosal biopsies (see Table S2, Supplementary Digital Content 2, However, there was no worsening of the mucosal inflammation over time, and no patient developed intestinal metaplasia or advanced mucosal atrophy in the gastric corpus. Furthermore, 11/15 patients (73%) had been evaluated with at least 1 colonoscopy with multiple mucosal biopsies during the follow-up period. With the exception of 1 patient with associated collagenous colitis, no additional pathologies were noted in the colonic or ileal biopsies of the remaining patients.

Autoimmune predisposition

In total, 9/15 patients (60%) had predisposition to autoimmune disease, defined as the presence of autoimmunity-related hereditary factors and/or autoantibodies in the circulation (Table 3). Approximately half (7/15, 47%) of the study subjects had hereditary factors for autoimmune diseases among their first-degree relatives. Moreover, 6/15 patients (40%) tested positive for autoantibodies. The most common autoantibody specificity detected was ANA, which was detected in a fifth (3/15, 20%) of the cohort. One patient tested positive for antithyroid peroxidase antibodies but had normal levels of thyroid hormones. None of the patients tested positive for antimitochondrial antibody, antithyroglobulin antibodies, antithyrotropin receptor antibodies, antiglutamic acid decarboxylase antibodies, anti-intrinsic factor antibodies, or antiparietal cell antibodies. More than half of the patients (8/15, 53%) carried the HLA DQ2 or DQ8 haplotypes. The most common haplotype was DQ2.5, which was detected in 5/15 (33%) patients (Table 3).

Table 3.
Table 3.:
Family history of autoimmune diseases, presence of autoantibodies in serum, and HLA DQ2/D8 haplotypes in the study cohort

Inflammatory parameters and immunoglobulins

The plasma level of CRP was discretely increased (i.e., 9 mg/mL; reference, <5 mg/mL) in only 1 patient. By contrast, the serum levels of calprotectin and amyloid A were increased in 10/15 (67%) and 5/15 (33%) patients, respectively (Figure 1a, b). For patients with increased values, the median level of serum calprotectin was 4,550 ng/mL (range, 3,600–11,000 ng/mL; reference, <2,300 ng/mL) and that of amyloid A was 22 mg/L (range, 12–55 mg/L; reference, <11 mg/L), respectively. In 4 of the 5 patients with elevated SAA, serum calprotectin was concurrently increased (Figure 1c); one of these patients also had an increased level of CRP. Moreover, serum calprotectin showed a strong positive correlation with the absolute neutrophil counts in the blood (Figure 1d), although the blood total leukocyte and absolute neutrophil counts were not above the reference range in any patient. The serum level of albumin was slightly decreased (i.e., 32 g/L) in 1 patient, and no patient showed increased serum levels of orosomucoid. A retrospective chart review revealed that erythrocyte sedimentation rate (ESR) had been analyzed in 14/15 (93%) and CRP in 13/15 (87%) patients during the disease course, in many cases repeatedly, and normal values were detected in all the patients.

Figure 1.
Figure 1.:
Levels of (a) serum calprotectin and (b) serum amyloid A in the cohort of 15 patients with childhood-onset collagenous gastritis. The cutoffs (indicated by dashed lines) applied were ≥2,300 ng/mL for serum calprotectin and ≥11 mg/L for serum amyloid A, respectively. The analysis method for serum amyloid A had a detection limit of 11 mg/L and the values below that level are given as zero. Relationship between serum levels of calprotectin and (c) serum amyloid A and (d) blood absolute neutrophil counts (n = 14). The correlation coefficient (r) with 95% confidence interval (CI) was determined using the Spearman test. The open triangle indicates the single patient in the cohort with concurrent collagen colitis.

No patient had total serum levels of immunoglobulins consistent with hypogammaglobulinemia or selective IgA deficiency, although slightly decreased levels of IgG and IgA were seen in 3 patients (see Table S3, Supplementary Digital Content 3, None of the patients in the cohort had high levels of IgG4 in sera indicative of IgG4-related disease.


In this population-based study, we have investigated the clinical, endoscopic, and histologic characteristics of childhood-onset CG in a cohort of 15 patients during a median follow-up period of 6.2 years from the initial presentation. The results of this study support the view that CG in children follows a chronic disease course and has a skewed sex distribution, with female predominance (41). Furthermore, although more than half of the patients had hereditary factors for autoimmune diseases and/or had developed autoantibodies, the development of autoimmune comorbidities early in the disease course of CG was not seen. A novel finding was that despite normal levels of plasma CRP, the level of serum calprotectin was increased in two-thirds of the patients and the level of SAA was increased in one-third of the patients, indicative of systemic inflammation.

To the best of our knowledge, this is the largest cohort of CG with longitudinal follow-up reported to date, and this is the first study of such a cohort with a population-based methodology that allows calculations of incidence and prevalence values. The incidence rate of childhood-onset CG is 0.25/100,000 person-years of follow-up, and the prevalence is 2.1/100,000 children aged younger than 18 years in western Sweden, which substantiates the idea that this is a rare disease. Furthermore, the incidence rate of childhood-onset CG was approximately 4-fold higher in female patients than in male patients, supporting the notion that there is female predominance in the childhood-onset type of CG. The skewed sex distribution has previously been suggested by aggregated data from published reports of CG for both the pediatric age group (41) and the whole (i.e., pediatric and adult combined) population (1). For the associated condition of collagenous colitis, female predominance is well documented in population-based studies in adults, reporting female-to-male ratios of up to 9:1 (55–58).

Approximately half of the patients in our cohort exhibited heredity for autoimmune diseases among their first-degree relatives, and 40% had developed autoantibodies. These findings support the view of an autoimmune/immune-mediated mechanism underlying the disease process, as previously indicated mainly by the frequent association with autoimmune comorbidities, such as celiac disease, in adults with CG (1,32). The frequency of heredity for autoimmune diseases observed in the present study is high, considering the estimated prevalence of autoimmune diseases in the Scandinavian general population of <10% (59–61). Similarly, the prevalences of the 2 most frequently observed autoantibody specificities in our study, ANA (20%) and antismooth muscle antibody (13%), can be compared with reported frequencies of <5% for the same autoantibodies in healthy children (62,63). Interestingly, the rate of autoantibody development is comparable with that reported for collagenous colitis in adult studies, e.g., ANA positivity in 10%–26% of the patients (64,65).

Although the frequencies of heredity for autoimmune diseases and autoantibody development in our cohort are relatively high, none of the patients developed any autoimmune disease during the follow-up period, which extended up to 11 years from the initial presentation. To our knowledge, in the English-language, peer-reviewed literature to date, 10 cases with development of concomitant autoimmune disease (22–24,26,31) among a total of 79 pediatric cases of CG have been reported (15–24,26–37,39–41,43–45). Therefore, despite the low rate of autoimmune comorbidities observed in the short term in our study, childhood-onset CG may still be associated with an increased long-term risk of developing these diseases.

Half of the patients in our cohort were HLA DQ2/DQ8-positive, which is similar to the prevalence of DQ2/DQ8 haplotypes in the general Scandinavian population (66,67). However, the DQ2.5 haplotype was detected in 33% of the patients, which can be compared with the reported frequency for this haplotype of 20%–25% in healthy controls in Scandinavian studies (52,67,68). There are no previous studies on the prevalence of the HLA DQ2/DQ8 in patients with CG. Notably, however, the frequency of the DQ2.5 haplotype in adult patients with collagenous colitis has been reported as 32%–45% (52,69,70), with 1 report providing data to support an association between the DQ2.5 haplotype and an increased risk of collagenous colitis (52).

A novel finding of the present study is that the serum levels of calprotectin and amyloid A are increased in two-thirds and one-third of the patients, respectively. This suggests that serum calprotectin in particular, but also SAA, is more sensitive than CRP for estimating the severity of inflammatory activity in patients with CG. To our knowledge, serum calprotectin and SAA have not previously been investigated in relation to CG. Although a few previous reports have provided data on inflammatory markers in childhood-onset CG, those studies have limited themselves to examining CRP and ESR (16,31–34,39,41). In keeping with our results, those previous studies have demonstrated normal levels of CRP and/or ESR in children with isolated CG.

Serum calprotectin and SAA have been proposed as complementary blood biomarkers to CRP and ESR for monitoring disease activity in other inflammatory conditions, such as rheumatoid arthritis (71–75) and inflammatory bowel disease (71,76–79). Calprotectin is produced by activated neutrophils and monocytes both locally at the site of inflammation and in the circulation (80), but the acute-phase protein SAA is synthesized by the hepatocytes in the liver (81). Although the increased serum levels of calprotectin noted in the present study may reflect in part local mucosal inflammation, the increased levels of SAA are indicative of systemic inflammation. Moreover, the positive association observed between serum calprotectin levels and blood absolute neutrophil counts has previously been reported for patients with inflammatory bowel disease (79) and cardiovascular disease (82).

In our cohort, the collagenous inflammation was confined to the stomach in all but 2 patients who had concurrent collagenous duodenitis and colitis, respectively. This rarity of concurrent collagenous inflammation in other parts of the digestive tract is consistent with the compiled results of previous pediatric reports. Overall, 79 pediatric patients with CG have, to our knowledge, been reported in the peer-reviewed, English-language literature to date (15–24,26–37,39–41,43–45). Of these, only 7 patients have concurrent collagenous colitis (22–24,26,40) and 4 patients have concurrent collagenous inflammation in the small bowel (17,23,24,39). In 3 additional reported cases, the collagenous inflammation affects all 3 locations in the gastrointestinal tract (24,30,35). This contrasts with adult patients, in whom CG is predominantly seen in combination with collagenous inflammation engaging the small bowel and/or colon (1,23,24).

The pathologic endoscopic and histologic features of CG persisted in all the study participants for whom follow-up endoscopies were performed, indicating a chronic disease course without spontaneous improvement. The endoscopic picture was dominated by mucosal nodularity, hypertrophic gastric folds, and erythema, which have been reported also by others as the most common findings in cases of childhood-onset CG (22–24,41). Histologically, a variation in the thickness of the subepithelial collagen bands in the gastric mucosa was noticed between different patients and longitudinally in the same patient. This might be attributable to sampling bias and the patchy distribution of subepithelial collagen bands in the gastric mucosa. The same explanation might be forwarded for the spontaneous “resolution” of collagen deposits previously reported for a few patients with CG (23).

Regarding the inflammatory cell infiltrate in the gastric mucosa, the histologic profiles of most of the patients were characterized by an eosinophil-rich inflammation, whereas increased intraepithelial lymphocytosis was seen in only 1 patient at diagnosis. This is in line with previous histopathologic studies, in which an eosinophil-rich inflammatory infiltrate was apparent in 50%–62% of the children with CG, but a lymphocytic gastritis-like pattern was seen in <10% of the cases (23,24). The opposite has been noted in adults with CG, in whom the eosinophil-rich inflammation seems to be less prevalent, whereas the lymphocytic gastritis-like pattern may be more common, as compared with children with CG (23,24).

Iron deficiency anemia was the dominant feature in the clinical picture of childhood-onset CG in this study. It was observed in approximately 90% of the patients at the initial presentation. Furthermore, iron deficiency with or without associated anemia showed a high rate of recurrence. Our results are in accordance with the few small case series of childhood-onset CG published to date, which have reported a variable prevalence of anemia (up to 92%) at presentation (22–24). The recurrence of iron deficiency has also been reported to be common (21,22). It is plausible that the high prevalence of iron deficiency observed in our cohort is attributable to CG per se rather than the result of other potentially biasing factors, such as dietary iron deficiency or menstrual blood loss.

The mechanism(s) leading to iron deficiency in (childhood-onset) CG remain unclear. Although understanding this mechanism was not the focus of our study, our data indicate neither overt nor occult gastrointestinal bleeding as underlying causes in most patients. Nonetheless, it has been previously suggested that iron deficiency associated with CG is because of chronic bleeding from dilated capillaries entrapped in the collagen layer of the gastric mucosa (27). Although clinical and/or endoscopic evidence of gastric bleeding has been noted in some cases (17,22–24,28,34), most children with CG seem to lack such signs. Instead of occult bleeding from the gastric mucosa, one can speculate that the iron deficiency in these patients results from decreased iron absorption due to gastric hypochlorhydria or other mechanisms.

Although a few cases of CG and associated common variable immune deficiency (4,24) or selective IgA deficiency (5) have previously been reported, such cases were not detected in our cohort. Moreover, we found no evidence of underlying IgG4-related disease in any of our patients because none of them had increased serum levels of IgG4. An indication of an IgG4-related disease mechanism was previously found in 1 patient with CG who had IgG4-positive plasma cells in the gastric mucosa (23).

One of the main strengths of the present study is the comprehensive clinical, endoscopic, and histologic characterization of the patients with childhood-onset CG. Furthermore, all eligible patients in a population-based setting were included. In addition, detailed follow-up data were available for all patients regarding the clinical course and endoscopies, with the longest median follow-up time reported for a cohort of patients with CG. A potential limitation is the relatively small sample size, which is mitigated by the rarity of the disease.

The main clinical significance of the present study is that it highlights the chronic nature of CG in children, thereby motivating long-term monitoring and follow-up of these patients. Awareness must be heightened regarding the potentially increased long-term risk for autoimmune/immune-related diseases. Furthermore, CG should be suspected in all children with unexplained iron deficiency anemia, and an upper endoscopy needs to be considered, even in an otherwise asymptomatic patient.

In conclusion, this study shows that childhood-onset CG is rare, follows a chronic disease course, and has a skewed gender distribution with female predominance. Although signs of autoimmune predisposition were frequent, early development of autoimmune comorbidities seems to be a rare phenomenon in childhood-onset CG. Serum calprotectin and amyloid A may represent novel markers of inflammatory activity in CG.


Guarantor of the article: Robert Saalman, MD, PhD.

Specific author contributions: T.K. and R.S.: planning and conducting the study, data acquisition, analysis and interpretation, and drafting the manuscript. C.W.: planning the study, data interpretation, and drafting the manuscript. C.L.: planning the study, data interpretation, and critical review of the manuscript. A.W. and M.W.: data acquisition and interpretation, and critical review of the manuscript. B.D.B., R.A., J.H., and J.A.: data acquisition and critical review of the manuscript. All the authors approved the final draft.

Financial support: The study was financed by grants from the Swedish state under the agreement between the Swedish Government and the County Councils, the ALF-agreement, The Västra Götaland Research and Development Fund (80,830), and Queen Silvia Children's Hospital Research Fund, Gothenburg, Sweden (The Märtha and Gustaf Ågren Foundation).

Potential competing interests: None to report.

Study Highlights


  • ✓ CG is a rare gastrointestinal disorder characterized histologically by inflammation of the gastric mucosa together with increased subepithelial deposition of collagen.
  • ✓ The clinical picture seems to differ between childhood-onset and adult-onset disease.


  • ✓ The incidence rate of childhood-onset CG in western Sweden was 0.25/100,000 person-years of follow-up and significantly higher for girls than for boys, whereas the disease prevalence was 2.1/100,000 in children younger than 18 years old.
  • ✓ Childhood-onset CG seems to have a chronic disease course without endoscopic or histologic improvement and is associated with a high frequency of recurrent iron deficiency.
  • ✓ Although heredity for autoimmune disorders and autoantibodies were commonly detected in the CG cohort, early development of autoimmune comorbidities was rarely observed.
  • ✓ No clear evidence of an association between childhood-onset CG and HLA DQ2/DQ8 haplotypes was found.
  • ✓ Increased serum levels of calprotectin and amyloid A were commonly observed and could be useful as markers for monitoring disease activity.


We thank Eva Karlsson, RN, and Birgitta Svensson, RN, for coordination of the clinical visits and expert logistical assistance.


1. Nielsen OH, Riis LB, Danese S, et al. Proximal collagenous gastroenteritides: Clinical management. A systematic review. Ann Med 2014;46:311–7.
2. Soeda A, Mamiya T, Hiroshima Y, et al. Collagenous gastroduodenitis coexisting repeated Dieulafoy ulcer: A case report and review of collagenous gastritis and gastroduodenitis without colonic involvement. Clin J Gastroenterol 2014;7:402–9.
3. Al-Kandari A, Al-Alardati H, Sayadi H, et al. An unusual case of collagenous gastritis in a middle-aged woman with systemic lupus erythromatosis: A case report. J Med Case Rep 2014;8:278.
4. Mandaliya R, Burkart AL, DiMarino AJ, et al. Association between common variable immunodeficiency and collagenous infiltrative disorders of the gastrointestinal tract: A series of four patients. Indian J Gastroenterol 2016;35:133–8.
5. Anwar MS, Aamar A, Marhaba A, et al. Collagenous gastritis in a young female with IgA deficiency. Gastroenterol Res 2017;10:126–7.
6. Zamani F, Boghratian A, Zare Mehrjardi A, et al. Collagenous gastritis, a rare cause of dyspepsia resistant to treatment: A case report. Middle East J Dig Dis 2018;10:263–6.
7. Akkari I, Skandrani K, Abdelkader AB, et al. Anemia revealing a collagenous gastritis in a young Tunisian man. Pan Afr Med J 2018;30:231.
8. Kawasaki K, Fujita Y, Uesugi N, et al. Collagenous gastritis observed by magnifying narrow-band imaging endoscopy. J Gastrointestin Liver Dis 2018;27:7.
9. Lim HW, Wong BY, Elkowitz D, et al. An elderly patient's complete response to steroid therapy for collagenous gastritis. Ther Adv Chronic Dis 2018;9:143–6.
10. Limaiem F, Mzabi S. Collagenous gastritis: A rare entity. J Interdiscip Histopathol 2015;3:68–70.
11. Anderson B, Ravi K. Chronic abdominal pain caused by collagenous gastritis. Clin Gastroenterol Hepatol 2019;17:e144.
12. Hayashi S, Nishida T, Adachi S, et al. Collagenous gastritis: A rare disease with distinctive endoscopic findings. Gastrointest Endosc 2018;88:186–7.
13. Singh S, Loo LEJ, Watters C, et al. Collagenous panenteritis: A rare cause of chronic diarrhoea. Frontline Gastroenterol 2017;8:232–5.
14. Sandhu DS, Bellizzi AM, El Abiad R. An elderly woman with upper abdominal pain and a 60-pound weight loss. Gastroenterology 2018;154:e14–e15.
15. Bajwa RU, Joshi A, Heikenen JB. Successful treatment of collagenous gastritis in a child with a gluten-free diet. WMJ 2015;114:271–3.
16. Rosell-Camps A, Riera-Llodra JM, Colom-Segui M, et al. Collagenous gastritis in the pediatric age. Rev Esp Enferm Dig 2015;107:313–5.
17. Koide T, Mochizuki T, Kawai N, et al. Collagenous gastroduodenitis with recurrent gastric ulcer in 12-year-old girl. Pediatr Int 2015;57:754–7.
18. Appelman MH, de Meij TG, Neefjes-Borst EA, et al. Spontaneous gastric perforation in a case of collagenous gastritis. APSP J Case Rep 2016;7:7.
19. Vinsard DG, Mejia Perez LK, Nassar A, et al. Collagenous gastritis and Helicobacter pylori infection: A mere coincidence? ACG Case Rep J 2017;4:e69.
20. Eke CB, Brown RA, De Lacy RJ, et al. Collagenous gastritis: An unusual cause of generalized oedema in a child. J Trop Pediatr 2019;65:305–8.
21. Lee YJ, Lee M, Kim DJ, et al. Three case reports of collagenous gastritis in children: Lessons for an endoscopic and histologic approach to mucosal nodularity of the stomach. Medicine 2019;98:e14870.
22. Matta J, Alex G, Cameron DJS, et al. Pediatric collagenous gastritis and colitis: A case series and review of the literature. J Pediatr Gastroenterol Nutr 2018;67:328–34.
23. Arnason T, Brown IS, Goldsmith JD, et al. Collagenous gastritis: A morphologic and immunohistochemical study of 40 patients. Mod Pathol 2015;28:533–44.
24. Ma C, Park JY, Montgomery EA, et al. A comparative clinicopathologic study of collagenous gastritis in children and adults: The same disorder with associated immune-mediated diseases. Am J Surg Pathol 2015;39:802–12.
25. Colletti RB, Cameron DJ, Hassall EG, et al. Collagenous gastritis: An international puzzle. J Pediatr Gastroenterol Nutr 1998;26:540.
26. Camarero C, Leon F, Colino E, et al. Collagenous colitis in children: Clinicopathologic, microbiologic, and immunologic features. J Pediatr Gastroenterol Nutr 2003;37:508–13.
27. Cote JF, Hankard GF, Faure C, et al. Collagenous gastritis revealed by severe anemia in a child. Hum Pathol 1998;29:883–6.
28. Park S, Kim DH, Choe YH, et al. Collagenous gastritis in a Korean child: A case report. J Korean Med Sci 2005;20:146–9.
29. Ravikumara M, Ramani P, Spray CH. Collagenous gastritis: A case report and review. Eur J Pediatr 2007;166:769–73.
30. Leiby A, Khan S, Corao D. Clinical challenges and images in GI. Collagenous gastroduodenocolitis. Gastroenterology 2008;135:17–9.
31. Wilson C, Thompson K, Hunter C. Nodular collagenous gastritis. J Pediatr Gastroenterol Nutr 2009;49:157.
32. Brain O, Rajaguru C, Warren B, et al. Collagenous gastritis: Reports and systematic review. Eur J Gastroenterol Hepatol 2009;21:1419–24.
33. Vella C, Pullicino E, Fearne C, et al. Collagenous gastritis: A rare cause of anaemia in childhood. Malta Med J 2011;22:34.
34. Colletti RB, Trainer TD. Collagenous gastritis. Gastroenterology 1989;97:1552–5.
35. Billiemaz K, Robles-Medranda C, Le Gall C, et al. A first report of collagenous gastritis, sprue, and colitis in a 9-month-old infant: 14 years of clinical, endoscopic, and histologic follow-up. Endoscopy 2009;41(Suppl 2):E233–4.
36. Kuo P, Pieterse S, Harley HA. The stomach that “cracked” under pressure. Gastroenterology 2010;138:44–5.
37. Jin X, Koike T, Chiba T, et al. Collagenous gastritis. Dig Endosc 2013;25:547–9.
38. Meunier S, Villard F, Bouvier R, et al. Collagen gastritis, an unusual cause of anemia in children. Report of 2 cases. Arch Pediatr 2001;8:47–50.
39. Kori M, Cohen S, Levine A, et al. Collagenous gastritis: A rare cause of abdominal pain and iron-deficiency anemia. J Pediatr Gastroenterol Nutr 2007;45:603–6.
40. Suskind D, Wahbeh G, Murray K, et al. Collagenous gastritis, a new spectrum of disease in pediatric patients: Two case reports. Cases J 2009;2:7511.
41. Hijaz NM, Septer SS, Degaetano J, et al. Clinical outcome of pediatric collagenous gastritis: Case series and review of literature. World J Gastroenterol 2013;19:1478–84.
42. Camarero Salces C, Enes Romero P, Redondo C, et al. Collagenous colitis and collagenous gastritis in a 9 year old girl: A case report and review of the literature. Acta Gastroenterol Belg 2011;74:468–74.
43. Dray X, Reignier S, Vahedi K, et al. Collagenous gastritis. Endoscopy 2007;39(Suppl 1):E292–3.
44. Lagorce-Pages C, Fabiani B, Bouvier R, et al. Collagenous gastritis: A report of six cases. Am J Surg Pathol 2001;25:1174–9.
45. Leung ST, Chandan VS, Murray JA, et al. Collagenous gastritis: Histopathologic features and association with other gastrointestinal diseases. Am J Surg Pathol 2009;33:788–98.
46. Kamimura K, Kobayashi M, Sato Y, et al. Collagenous gastritis: Review. World J Gastrointest Endosc 2015;7:265–73.
47. Vesoulis Z, Lozanski G, Ravichandran P, et al. Collagenous gastritis: A case report, morphologic evaluation, and review. Mod Pathol 2000;13:591–6.
48. Freeman HJ. Collagenous mucosal inflammatory diseases of the gastrointestinal tract. Gastroenterology 2005;129:338–50.
49. Gopal P, McKenna BJ. The collagenous gastroenteritides: Similarities and differences. Arch Pathol Lab Med 2010;134:1485–9.
50. Stancu M, De Petris G, Palumbo TP, et al. Collagenous gastritis associated with lymphocytic gastritis and celiac disease. Arch Pathol Lab Med 2001;125:1579–84.
51. Winslow JL, Trainer TD, Colletti RB. Collagenous gastritis: A long-term follow-up with the development of endocrine cell hyperplasia, intestinal metaplasia, and epithelial changes indeterminate for dysplasia. Am J Clin Pathol 2001;116:753–8.
52. Westerlind H, Mellander MR, Bresso F, et al. Dense genotyping of immune-related loci identifies HLA variants associated with increased risk of collagenous colitis. Gut 2017;66:421–8.
53. Population Statistics [database online]. Stockholm, Sweden: Statistics Sweden; 1968-2019. Updated February 20, 2020.
54. Vandenbroucke JP, Pearce N. Incidence rates in dynamic populations. Int J Epidemiol 2012;41:1472–9.
55. Bohr J, Tysk C, Eriksson S, et al. Collagenous colitis in Orebro, Sweden, an epidemiological study 1984-1993. Gut 1995;37:394–7.
56. Fernandez-Banares F, Salas A, Forne M, et al. Incidence of collagenous and lymphocytic colitis: A 5-year population-based study. Am J Gastroenterol 1999;94:418–23.
57. Williams JJ, Kaplan GG, Makhija S, et al. Microscopic colitis-defining incidence rates and risk factors: A population-based study. Clin Gastroenterol Hepatol 2008;6:35–40.
58. Pardi DS, Loftus EV Jr, Smyrk TC, et al. The epidemiology of microscopic colitis: A population based study in Olmsted County, Minnesota. Gut 2007;56:504–8.
59. Eaton WW, Rose NR, Kalaydjian A, et al. Epidemiology of autoimmune diseases in Denmark. J Autoimmun 2007;29:1–9.
60. Eaton WW, Pedersen MG, Atladóttir HO, et al. The prevalence of 30 ICD-10 autoimmune diseases in Denmark. Immunol Res 2010;47:228–31.
61. Cooper GS, Bynum MLK, Somers EC. Recent insights in the epidemiology of autoimmune diseases: Improved prevalence estimates and understanding of clustering of diseases. J Autoimmun 2009;33:197–207.
62. Martini A, Lorini R, Zanaboni D, et al. Frequency of autoantibodies in normal children. Am J Dis Child 1989;143:493–6.
63. Kasapçopur O, Ozbakir F, Arisoy N, et al. Frequency of antinuclear antibodies and rheumatoid factor in healthy Turkish children. Turk J Pediatr 1999;41:67–71.
64. Holstein A, Burmeister J, Plaschke A, et al. Autoantibody profiles in microscopic colitis. J Gastroenterol Hepatol 2006;21:1016–20.
65. Roth B, Gustafsson RJ, Ohlsson B. Auto-antibodies and their association with clinical findings in women diagnosed with microscopic colitis. PLoS One 2013;8:e66088.
66. Sandstrom O, Rosen A, Lagerqvist C, et al. Transglutaminase IgA antibodies in a celiac disease mass screening and the role of HLA-DQ genotyping and endomysial antibodies in sequential testing. J Pediatr Gastroenterol Nutr 2013;57:472–6.
67. Kårhus LL, Thuesen BH, Skaaby T, et al. The distribution of HLA DQ2 and DQ8 haplotypes and their association with health indicators in a general Danish population. United European Gastroenterol J 2018;6:866–78.
68. Ploski R, Ascher H, Sollid LM. HLA genotypes and the increased incidence of coeliac disease in Sweden. Scand J Gastroenterol 1996;31:1092–7.
69. Fernández-Bañares F, Esteve M, Farré C, et al. Predisposing HLA-DQ2 and HLA-DQ8 haplotypes of coeliac disease and associated enteropathy in microscopic colitis. Eur J Gastroenterol Hepatol 2005;17:1333–8.
70. Koskela RM, Karttunen TJ, Niemelä SE, et al. Human leucocyte antigen and TNFalpha polymorphism association in microscopic colitis. Eur J Gastroenterol Hepatol 2008;20:276–82.
71. De Beer FC, Mallya RK, Fagan EA, et al. Serum amyloid-A protein concentration in inflammatory diseases and its relationship to the incidence of reactive systemic amyloidosis. Lancet 1982;2:231–4.
72. Cunnane G, Grehan S, Geoghegan S, et al. Serum amyloid A in the assessment of early inflammatory arthritis. J Rheumatol 2000;27:58–63.
73. de Vries MK, van Eijk IC, van der Horst-Bruinsma IE, et al. Erythrocyte sedimentation rate, C-reactive protein level, and serum amyloid a protein for patient selection and monitoring of anti-tumor necrosis factor treatment in ankylosing spondylitis. Arthritis Rheum 2009;61:1484–90.
74. Abildtrup M, Kingsley GH, Scott DL. Calprotectin as a biomarker for rheumatoid arthritis: A systematic review. J Rheumatol 2015;42:760–70.
75. Hurnakova J, Hulejova H, Zavada J, et al. Serum calprotectin may reflect inflammatory activity in patients with active rheumatoid arthritis despite normal to low C-reactive protein. Clin Rheumatol 2018;37:2055–62.
76. Plevy S, Silverberg MS, Lockton S, et al. Combined serological, genetic, and inflammatory markers differentiate non-IBD, Crohn's disease, and ulcerative colitis patients. Inflamm Bowel Dis 2013;19:1139–48.
77. Bourgonje AR, von Martels JZH, Gabriels RY, et al. A combined set of four serum inflammatory biomarkers reliably predicts endoscopic disease activity in inflammatory bowel disease. Front Med 2019;6:251.
78. Carlsen K, Malham M, Hansen LF, et al. Serum calprotectin in adolescents with inflammatory bowel disease: A pilot investigation. J Pediatr Gastroenterol Nutr 2019;68:669–75.
79. Kalla R, Kennedy NA, Ventham NT, et al. Serum calprotectin: A novel diagnostic and prognostic marker in inflammatory bowel diseases. Am J Gastroenterol 2016;111:1796–805.
80. Wang S, Song R, Wang Z, et al. S100A8/A9 in inflammation. Front Immunol 2018;9:1298.
81. Sack GH Jr. Serum amyloid A: A review. Mol Med 2018;24:46.
82. Cotoi OS, Duner P, Ko N, et al. Plasma S100A8/A9 correlates with blood neutrophil counts, traditional risk factors, and cardiovascular disease in middle-aged healthy individuals. Arterioscler Thromb Vasc Biol 2014;34:202–10.
© 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of The American College of Gastroenterology