Advertisement

Diagnostic findings and long-term prognosis in children with anemia undergoing GI endoscopies

Published:January 06, 2020DOI:https://doi.org/10.1016/j.gie.2019.12.042

      Background and Aims

      Intestinal diseases are regarded as a common cause of anemia, but the diagnostic outcomes of children with anemia undergoing endoscopic investigations are unclear. We investigated this issue in a large cohort of children.

      Methods

      Indications for and findings of consecutive gastrointestinal (GI) endoscopies were collected. Clinical presentation and diagnostic outcomes were compared between anemic and nonanemic patients and between anemic patients with and without a diagnosis. Diagnoses received during follow-up were collected.

      Results

      Of 2395 consecutive endoscopies, 251 children with and 613 children without anemia had undergone either diagnostic esophagogastroduodenoscopy (EGD) (51.4% and 51.4%, respectively), colonoscopy (4.0% and 11.4%), or both (45.8% and 37.8%). Children with anemia more often received diagnoses (72.9% vs 39.3%; odds ratio [OR], 4.18; 95% confidence interval [CI], 3.03-5.77), particularly of celiac disease (26.3% vs 15.5%, P < .001) and of inflammatory bowel disease (31.1% vs 9.1%, P < .001), than did nonanemic children. The diagnosis in anemic patients was predicted by age 5 to 12 years (OR, 3.52; 95% CI, 1.27-9.75), presence of diarrhea (OR, 2.04; 95% CI, 1.07-3.90), melena/hematochezia (OR, 2.40; 95% CI, 1.17-4.92), poor growth (OR, 3.94; 95% CI, 1.70-9.15), positive celiac serology (OR, 11.81; 95% CI, 3.47-40.12), high calprotectin (OR, 12.86; 95% CI, 4.00-41.32), hypersedimentation (OR, 2.65; 95% CI, 1.29-5.44), and hypoalbuminemia (OR, 5.05; 95% CI, 1.56-16.34). Thirty children with anemia (12.0%) had no GI symptoms, and 22 of them (73.3%) were given diagnoses at the time of the endoscopies. All 22 had additional laboratory abnormalities, whereas these were present in only 2 of 8 undiagnosed children. None of them was diagnosed later in the follow-up of up to 11 years, in contrast to 4 (6.7%) of all anemic and 33 (8.9%) of all nonanemic patients.

      Conclusions

      Anemia increased the probability of being given a diagnosis, emphasizing its importance as an alarm symptom. However, endoscopies in anemic patients without additional symptoms or laboratory abnormalities seldom improved the diagnostic yield.

      Graphical abstract

      Abbreviations:

      CI (confidence interval), ESR (erythrocyte sedimentation rate), IBD (inflammatory bowel disease), MCV (mean corpuscular volume), OR (odds ratio), Rf (reference value), GI (gastrointestinal), EGD (esophagogastroduodenoscopy), GERD (gastroesophageal reflux disease), Hb (hemoglobin)
      Anemia is one of the most common chronic medical problems in children, affecting on average 16.7% of preschool children in Europe and as many as 64.6% in Africa.
      • McLean E.
      • Cogswell M.
      • Egli I.
      • et al.
      Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993-2005.
      Although in developing countries pediatric anemia is usually caused by inadequate nutrition and deficits of essential micronutrients, in developed countries the leading cause is iron deficiency, which may be either dietary or secondary to a gastrointestinal (GI) or hematologic disease or other chronic conditions.
      • Camascella C.
      Iron deficiency.
      Importantly, anemia is not a diagnosis as such but a clinical sign, the underlying cause of which should always be investigated.
      Alimentary tract diseases are regarded as a common reason for anemia. The development of anemia in these conditions is multifactorial, including, for example, insidious blood loss and inadequate iron absorption.
      • Ferrara M.
      • Coppola L.
      • Coppola A.
      • et al.
      Iron deficiency in childhood and adolescence: retrospective review.
      ,
      • Gasche C.
      Iron, anaemia, and inflammatory bowel diseases.
      It has been suggested that GI pathologies should always be ruled out in children and adolescents with unexplained anemia.
      • Thomson M.
      • Tringali A.
      • Dumonceau J.M.
      • et al.
      Paediatric gastrointestinal endoscopy: European Society for Paediatric Gastroenterology, Hepatology and Nutrition and European Society of Gastrointestinal Endoscopy guidelines.
      Unfortunately, a reliable examination of the intestine requires invasive methods, particularly esophagogastroduodenoscopy (EGD) and colonoscopy, which in children are usually performed with the patient under general anesthesia. Optimal targeting of these invasive procedures in children would be of particular importance, but currently the evidence on this issue is scarce.
      • Franciosi J.P.
      • Fiorino K.
      • Ruchelli E.
      • et al.
      Changing indications for upper endoscopy in children during a 20-year period.
      • Thomson M.
      • Sharma S.
      Diagnostic yield of upper and lower gastrointestinal endoscopies in children in a tertiary centre.
      • Sheiko M.A.
      • Feinstein J.A.
      • Capocelli K.E.
      • et al.
      Diagnostic yield of EGD in children : a retrospective single-center study of 1000 cases.
      • Wang S.
      • Younus O.
      • Rawat D.
      • et al.
      Clinical presentation and outcomes of diagnostic endoscopy in newly presenting children with gastrointestinal symptoms.

      Methods

      Study ethics

      The study design and data collection were duly approved by the Department of Pediatrics, Tampere University Hospital. All identifiable personal data were coded, and analyses were performed anonymously. According to the national guidelines in Finland, this registry-based study required no ethical approval.

      Patients and study design

      This retrospective cross-sectional study of a tertiary center was conducted at Tampere University and Tampere University Hospital. We have long maintained comprehensive medical records on all children undergoing GI endoscopies and performed systematic mucosal sampling irrespective of macroscopic findings. These practices afforded us an opportunity to investigate the diagnostic yield of endoscopic investigations and the long-term prognoses in children with unexplained anemia.
      The medical data on children (age <17 years) who had undergone consecutive EGDs and/or colonoscopies in 2007 to 2014 were recorded. Only patients who had undergone diagnostic endoscopy(ies) and for whom anemia/hemoglobin data were available were included for further analyses, whereas follow-up and other endoscopies (rectosigmoidoscopies) were excluded. These children were divided into anemic and nonanemic groups, and the anemia group were divided further into those presenting with and without additional GI symptoms (Fig. 1). All study data were compared between these groups. Further, patient characteristics predicting diagnosis were identified. In addition to medical information before and at the time of the endoscopic studies, follow-up data were collected from 2 to 11 years.
      Figure thumbnail gr1
      Figure 1A flowchart of the study. *A patient was considered anemic if the hemoglobin value was below the age- and sex-dependent reference at the time of endoscopy. †For up to 11 years of follow-up.

      Clinical and laboratory data

      The information recorded included demographic parameters, clinical presentation, duration of symptoms, presence of poor growth, chronic conditions, and family history of GI diseases. Other possibly conducted relevant investigations of the GI tract, such as wireless capsule endoscopy, magnetic resonance enterography, or esophageal pH monitoring, were also recorded, as well as all diagnoses set in either the primary or follow-up investigations. Children with minor unspecific and apparently clinically insignificant endoscopic or histologic abnormalities, such as lymphonodular hyperplasia common in children or inactive chronic gastritis, and those with only functional abdominal symptoms were assigned to the no-diagnosis group.
      The results of the following laboratory tests were collected as available: blood hemoglobin and mean corpuscular volume (MCV; reference values [Rfs]
      Fimlab-ohjekirja
      Perusverenkuva ja trombosyytit.
      presented in Supplementary Table 1, available online at www.giejournal.org), plasma ferritin (Rf > 6 μg/L), plasma transferrin receptor (Rf for children age <1 years, 1.6-7.0 mg/L; 1-3 years, 2.7-5.4 mg/L; and 4-6 years, 2.4-6.3 mg/L; for boys ages 7-12 years, 2.4-5.7 mg/L; girls 7-12 years, 2.0-5.1 mg/L; boys 13-17 years, 2.0-6.8 mg/L; and girls 13-17 years 1.6-5.2 mg/L), erythrocyte sedimentation rate (ESR; Rf < 15 mm/h), plasma albumin (Rf from 35-46 to 37-51 g/L), plasma alanine aminotransferase (Rf < 40 U/L), serum antiendomysial antibodies (Rf titer 1: <5) and IgA-class antibodies against transglutaminase 2 (Rf < 7.0 U/L), and fecal calprotectin (Rf < 100 μg/g). To calculate the possible effect of the severity of anemia on the likelihood of a subsequent diagnosis, hemoglobin values at the time of endoscopy were further subclassified using the World Health Organization criteria.
      WHO
      Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. Vitamin and Mineral Nutrition Information System.
      Failure to thrive was defined as reduced growth velocity in the longitudinal evaluation of growth charts and/or impaired height or weight development when compared with the gender- and age-dependent Rfs or to target height calculated using midparental height.
      • Saari A.
      • Harju S.
      • Mäkitie Outi
      • et al.
      Systematic growth monitoring for the early detection of celiac disease in children.

      Histology

      Endoscopic abnormalities were reported systemically and photographed if considered potentially significant. During EGD, at least 2 mucosal biopsy specimens were taken systemically, regardless of the preceding clinical and laboratory findings, from the lower esophagus, antrum, and gastric corpus and antrum and at least 4 specimens from the duodenum. Since 2012, biopsy specimens were also obtained from the duodenal bulb and middle part of the esophagus. During colonoscopy, biopsy specimens were taken from the rectum, sigmoid/descending colon, ascending colon/cecum, and terminal ileum. In both EGD and colonoscopy, additional specimens were taken as clinically indicated. Endoscopies were considered adequate when biopsy specimens were taken systemically from each abovementioned bowel section.
      • Thomson M.
      • Sharma S.
      Diagnostic yield of upper and lower gastrointestinal endoscopies in children in a tertiary centre.
      Specimens were cut, stained, and evaluated by standard histopathology methods. If needed, special stainings were performed. Only representative and correctly oriented and cut histologic specimens were accepted for quantitative morphometric analyses of duodenal samples.
      • Taavela J.
      • Koskinen O.
      • Huhtala H.
      • et al.
      Validation of morphometric analyses of small-intestinal biopsy readouts in celiac disease.
      All abnormalities reported by a pathologist were regarded as histologic findings even if not diagnostic, including, for example, unspecified duodenal inflammation, mildly increased number of eosinophils in the esophagus,
      • Furuta G.T.
      • Liacouras C.A.
      • Collins M.H.
      • et al.
      Eosinophilic esophagitis in children and adults: a systematic review and consensus recommendations for diagnosis and treatment.
      and inactive chronic gastritis.
      Individual diagnoses were set according to international consensus
      • Furuta G.T.
      • Liacouras C.A.
      • Collins M.H.
      • et al.
      Eosinophilic esophagitis in children and adults: a systematic review and consensus recommendations for diagnosis and treatment.
      • Walker-Smith J.A.
      • Guandalini S.
      • Schmizt J.
      • et al.
      Revised criteria for diagnosis of coeliac disease. Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition.
      • Rudolph C.D.
      • Mazur L.J.
      • Baker R.D.
      • et al.
      Guidelines for evaluation and treatment of gastroesophageal reflux in infants and children: recommendations of the North American Society for Pediatric Gastroenterology and Nutrition.
      IBD Working Group of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition
      Inflammatory bowel disease in children and adolescents: recommendations for diagnosis—the Porto criteria.
      • Valent P.
      • Sperr W.R.
      • Schwartz L.B.
      • et al.
      Diagnosis and classification of mast cell proliferative disorders: delineation from immunologic diseases and non-mast cell hematopoietic neoplasms.
      • Chapman J.A.
      • Berstein I.L.
      • Rufus E.L.
      • et al.
      Food allergy: a practice parameter.
      • Drumm B.
      • Koletzko S.
      • Oderda G.
      Helicobacter pylori infection in children: a consensus statement. European Paediatric Task Force on Helicobacter pylori.
      or, if no specific guidelines existed, based on previous literature and clinical experience. During the study period the European guidelines on celiac disease were revised in 2012 but the new criteria
      • Husby S.
      • Koletzko S.
      • Korponay-Szabó I.R.
      • et al.
      European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease.
      were not implemented in our clinical practice until 2015. Additionally, the guidelines on food allergy
      • Sackeyfio A.
      • Senthinathan A.
      • Kandaswamy P.
      • et al.
      Diagnosis and assessment of food allergy in children and young people: summary of NICE guidance.
      and Helicobacter pylori
      • Koletzko S.
      • Jones N.L.
      • Goodman K.J.
      • et al.
      Evidence-based guidelines from ESPGHAN and NASPGHAN for Helicobacter pylori infection in children.
      were both updated in 2011 and the guidelines of reflux disease
      • Vandenplas Y.
      • Rudolph C.D.
      • Di Lorenzo C.
      • et al.
      Pediatric gastroesophageal reflux clinical practice guidelines: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology.
      in 2009, and the new criteria were subsequently used to establish the diagnoses. The diagnosis of GI food allergy was set by experienced clinicians based on a combination of clinical and histologic findings, allergy testing, and elimination diet followed by food challenge.
      • Chapman J.A.
      • Berstein I.L.
      • Rufus E.L.
      • et al.
      Food allergy: a practice parameter.
      ,
      • Sackeyfio A.
      • Senthinathan A.
      • Kandaswamy P.
      • et al.
      Diagnosis and assessment of food allergy in children and young people: summary of NICE guidance.

      Statistical analysis

      Clinical characteristics, abnormal laboratory parameters, and endoscopic and histologic findings are presented as percentage distributions. Statistical significances of differences in qualitative parameters were calculated using the χ2 test or the Fisher exact test. Most quantitative variables were found to be skewed by the Shapiro-Wilk method and included outliers. The variables were thus analyzed by nonparametric Mann-Whitney test and expressed as medians and quartiles.
      • Altman D.G.
      Practical statistics for medical research.
      The associations between symptoms/clinical signs and diagnoses were calculated using binary logistic regression analysis with 95% confidence intervals (CIs). Odds ratios (ORs) were calculated comparing boys with girls, age groups with those younger than 1 year, and patients with autoimmune disease, intestinal disease in relatives, different symptoms, or abnormal laboratory parameters with those without such findings. A P < .05 was considered statistically significant. All statistical analyses were performed using SPSS Statistics version 23 (IBM Corp, Armonk, NY, USA).

      Results

      Altogether, 2395 consecutive GI endoscopies were conducted during the study period. Of these, 1117 were follow-up or other endoscopies, and 399 children lacked hemoglobin values and were therefore excluded. Of the remaining 864 children with hemoglobin and anemia data available, altogether 251 (29.1%) had anemia (Fig. 1). Of these, 128 (51.0%) had undergone EGD alone, 10 (4.0%) had undergone colonoscopy alone, and the other 113 (45.1%) had undergone both EGD and colonoscopy. The corresponding figures for nonanemic children were 311 (50.7%), 70 (11.4%), and 232 (37.8%), respectively. The duodenum was reached in 98.8% and 99.1% and the ileum in 65.3% and 67.1% of the anemic and nonanemic children, respectively. Subsequently, combined with missing samples in some cases, respectively, 83.0% and 85.1% of EGDs and 62.1% and 67.2% of colonoscopies of anemic and nonanemic children were considered adequate according to our strict definition. Additional alimentary tract investigations were conducted in 93 (37.1%) anemic and 214 (35.0%) nonanemic children, including wireless capsule endoscopy in 22 (8.8%) and 37 (6.0%).
      Children with anemia were significantly older (median 10.8 vs 8.3 years, P = .012) and more often had concomitant autoimmune disease, positive calprotectin, and low MCV than did those without anemia (Table 1). They also presented with higher plasma transferrin receptor and ESR and lower ferritin and albumin values (Table 1), whereas there were no differences between the groups in gender distribution, presence of other diseases, familial history, celiac antibody, or alanine aminotransferase levels. Failure to thrive was significantly more common and abdominal pain, reflux, and dysphagia less common in children with anemia (Fig. 2). Children with anemia also had more often endoscopic findings in both EGD (60.3% vs 47.9%, P = .001) and colonoscopy (72.1% vs 48.3%, P < .001) as well as histologic abnormalities at each biopsy site excluding the esophagus (Table 2).
      Table 1Characteristics of 864 children with and without anemia at the time of their first GI endoscopies
      Categorical variablesAnemia (n = 251)No anemia (n = 613)P value
      N%N%
      Girls13553.832653.2.872
      Autoimmune disease
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      208.0274.4.036
      P ≤ .050.
      Asthma, allergy, or atopy4417.511618.9.632
      Other chronic disease3212.7589.5.151
      Intestinal disease in relatives
      Celiac disease or inflammatory bowel disease.
      4819.115325.0.065
      Positive EmA or TGA-IgA70
      Data from 177 anemic and 419 nonanemic patients.
      39.5132
      Data from 177 anemic and 419 nonanemic patients.
      31.5.058
      Fecal calprotectin > 100 μg/g65
      Data from 94 anemic and 254 nonanemic patients.
      69.184
      Data from 94 anemic and 254 nonanemic patients.
      33.1<.001
      P ≤ .050.
      Low MCV91
      Data from 215 anemic and 557 nonanemic patients.
      42.338
      Data from 215 anemic and 557 nonanemic patients.
      6.8<.001
      P ≤ .050.
      Continuous variablesNMedian (quartiles)NMedian (quartiles)
      Age, y (range)25110.8 (.03, 16.7)6138.3 (.02, 17.7).012
      P ≤ .050.
      Ferritin, μg/L627.5 (5.0, 15.3)7020.5 (13.8, 31.0)<.001
      P ≤ .050.
      Transferrin receptor 1, mg/L587.2 (5.0, 14.5)374.3 (3.7, 5.2)<.001
      P ≤ .050.
      ESR, mm/h16812.0 (5.0, 25.75)4347.0 (5.0, 11.3)<.001
      P ≤ .050.
      Alanine aminotransferase, U/L17219.5 (14.0, 29.0)39917.0 (14.0, 24.0).189
      Albumin, g/L10236.0 (33.0, 40.0)30540.0 (37.0, 43.0)<.001
      P ≤ .050.
      EmA, Antiendomysial antibodies; ESR, erythrocyte sedimentation rate; MCV, mean corpuscular volume; TGA-IgA, IgA-antibodies against transglutaminase 2.
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      P ≤ .050.
      Celiac disease or inflammatory bowel disease.
      § Data from 177 anemic and 419 nonanemic patients.
      Data from 94 anemic and 254 nonanemic patients.
      ǁ Data from 215 anemic and 557 nonanemic patients.
      Figure thumbnail gr2
      Figure 2Comparison of the prevalence of different symptoms in children with and without anemia. Other symptoms included, for example, fatigue, persistent fever, and dermatologic symptoms.
      Table 2Distribution of histologic abnormalities in the endoscopic biopsy samples of patients with and without anemia
      Biopsy sample locationAnemia (n = 251)No anemia (n = 613)P value
      No. of samplesAbnormalities
      Subjects could have histologic abnormalities in 1 or more sampling site.
      (%)
      No. of samplesAbnormalities
      Subjects could have histologic abnormalities in 1 or more sampling site.
      (%)
      Total EGD24272.754351.6<.001
      P ≤ .050.
       Esophagus23622.553819.7.382
       Stomach24149.054226.8<.001
      P ≤ .050.
       Duodenum23936.853822.1<.001
      P ≤ .050.
      Total colonoscopy12471.830033.0<.001
      P ≤ .050.
       Ileum8128.420813.5.003
      P ≤ .050.
       Ascending colon11060.027420.4<.001
      P ≤ .050.
       Descending/sigmoid colon12262.329721.5<.001
      P ≤ .050.
       Rectum11657.827720.9<.001
      P ≤ .050.
      Subjects could have histologic abnormalities in 1 or more sampling site.
      P ≤ .050.
      Altogether, 183 (73.0%) anemic and 241 (39.3%) non-anemic patients received a diagnosis during the primary investigations (OR, 4.18; 95% CI, 3.03-5.77) (Fig. 1). Eight of 10 children (80.0%) with severe anemia, 91 of 111 children (82.0%) with moderate anemia, and 77 of 119 children (64.7%) with mild anemia received a diagnosis (P = .011). Altogether, anemia had 43.2% sensitivity, 84.6% specificity, 72.9% positive predictive value, 60.7% negative predictive value, and 64.2% overall accuracy for the subsequent diagnosis. Seven children with anemia (2.8%) with normal histologic findings in EGD and/or colonoscopy received a diagnosis in other concomitant investigations, including 1 with Meckel’s diverticulum on technetium scan and 1 with Burkitt’s lymphoma in magnetic resonance enterography. No endoscopic abnormalities were reported in 24 anemic (13.1%) and 59 nonanemic (24.5%) children who received a diagnosis.
      Anemic patients with diagnoses presented more often with positive celiac serology, positive calprotectin, and higher ESR and lower albumin values than those without diagnoses (Supplementary Table 2, available online at www.giejournal.org). Predictors for initial diagnoses in anemic patients were age from 5 to 12 years, presence of diarrhea, melena/hematochezia, poor weight gain, positive celiac antibodies, high calprotectin or ESR values, and hypoalbuminemia, whereas children with reflux symptoms were less likely to receive a diagnosis (Table 3). In nonanemic patients, symptoms significantly increasing the likelihood of diagnosis were comparable with those in anemic patients excluding diarrhea and hypoalbuminemia (data not shown). In a multivariate binary regression analysis, anemia, melena/hematochezia, and failure to thrive predicted diagnosis independently of other GI symptoms (Supplementary Table 3, available online at www.giejournal.org).
      Table 3Relationships between clinical features, presence of endoscopic and histologic findings, and diagnosis received in initial investigations in 251 patients with anemia
      Abnormal findingsDiagnosis
      No. of casesEndoscopic (%)Histologic (%)%Odds ratio
      Calculated by binary regression analysis with 95% confidence intervals comparing boys with girls, age groups with those younger than 1 year, and patients with autoimmune disease, intestinal disease in relatives, different symptoms, or abnormal laboratory parameters with those without such findings.
      95% Confidence interval
      Sex
       Female13572.680.770.4
       Male11680.282.875.91.32.75-2.32
      Age
       <1 y2161.961.952.4
       1-4 y5273.182.773.12.47.86-7.07
       5-12 y7878.287.279.53.521.27-9.75
       13-17 y10079.081.072.02.34.89-6.11
      Autoimmune disease
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      4488.675.077.30.86.32-2.33
      Intestinal disease in relatives
      Inflammatory bowel disease or celiac disease.
      4881.385.479.21.52.71-3.25
      Symptoms
       Abdominal pain13573.383.771.9.89.51-1.56
       Diarrhea8284.192.781.72.041.07-3.90
       Melena/hematochezia6989.989.984.12.401.17-4.92
       Poor weight gain/growth6484.492.289.13.941.70-9.15
       Constipation3871.176.365.8.29.32-1.40
       Vomiting2673.173.169.2.82.34-1.98
       Reflux2352.243.539.1.20.08-.49
       Other10180.282.274.31.12.63-1.99
      Laboratory parameters
       Positive celiac serology
      Antiendomysial antibodies or IgA antibodies against transglutaminase 2.
      ,
      Data from 177 patients.
      7080.097.195.711.813.47-40.12
       Fecal calprotectin > 100 μg/g
      Data from 94 patients.
      2987.795.492.312.864.00-41.32
       Hypersedimentation
      Data from 168 patients.
      7884.688.582.12.651.29-5.44
       Hypoalbuminemia
      Data from 102 patients.
      4883.395.891.75.051.56-16.34
      —, Reference group in binary regression analysis.
      Calculated by binary regression analysis with 95% confidence intervals comparing boys with girls, age groups with those younger than 1 year, and patients with autoimmune disease, intestinal disease in relatives, different symptoms, or abnormal laboratory parameters with those without such findings.
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      Inflammatory bowel disease or celiac disease.
      § Antiendomysial antibodies or IgA antibodies against transglutaminase 2.
      Data from 177 patients.
      ǁ Data from 94 patients.
      ∗∗ Data from 168 patients.
      †† Data from 102 patients.
      An additional 4 anemic (1.6%) and 33 nonanemic (5.4%) patients received diagnoses later in the follow-up (Fig. 1). Follow-up diagnoses in anemic children were Crohn’s disease, GERD, angiodysplasia, and Imerslund-Gräsbeck syndrome, whereas in nonanemic patients the most common diagnoses were celiac disease (n = 9), Crohn’s disease (n = 6), gastroesophageal reflux disease (GERD) (n = 5), and GI food allergy (n = 4).
      The most common initial diagnoses in the anemia group were celiac disease and inflammatory bowel disease (IBD), which were also more common in anemic than in nonanemic patients (Fig. 3). Other GI diagnoses found in more than 1 anemic child were H pylori gastritis (n = 11), GERD (n = 7), GI food allergy (n = 5), gastric or duodenal ulcer (n = 3), juvenile polyps (n = 2), and undefined colitis (n = 2), whereas GERD (n = 30), H pylori gastritis (n = 12), GI food allergy (n = 14), esophagitis (n = 6), mastocytosis (n = 5), pinworms (n = 5), juvenile polyps (n = 3), polyposis syndrome (n = 2), gastric/duodenal ulcer (n = 3), and rectal prolapse (n = 2) were found in nonanemic children.
      Figure thumbnail gr3
      Figure 3Comparison of the diagnoses received in the initial investigations between patients with or without anemia. H. pylori, Helicobacter pylori; IBD inflammatory bowel disease.
      Of the anemic children, 45 (17.9%) presented with only anemia-related laboratory abnormalities. Of these, 23 (51.1%) had undergone EGD alone, 2 (4.4%) colonoscopy alone, and 20 (44.4%) both EGD and colonoscopy. Fifteen (33.3%) of them received a diagnosis in the initial investigations, including 6 IBD, 1 fungal esophagitis, 3 reflux esophagitis, 1 GERD diagnosed based on esophageal pH monitoring, 2 gastric/duodenal ulcers, 1 Burkitt’s lymphoma, and 1 eating disorder. Two children received a diagnosis during subsequent follow-up (Crohn’s disease and angiodysplasia).
      Thirty anemic children had no additional GI symptoms. All 30 underwent EGD and 7 also colonoscopy. Initial endoscopies revealed a diagnosis in 22 cases (Fig. 1), including 13 celiac disease, 4 H pylori gastritis, 2 IBD, 1 GI stromal tumor, 1 autoimmune gastritis, and 1 Epstein-Barr virus infection of the colon. Additional laboratory abnormalities besides anemia and/or low MCV were present in all 22 (Supplementary Table 4, available online at www.giejournal.org). None of the remaining 8 children underwent repeated endoscopies or received a diagnosis during a follow-up of up to 11 years (Fig. 1). After endoscopy, a wireless capsule endoscopy was conducted in 4 children and a technetium scan in 2, with normal findings. One of them had elevated ESR (21 mm/h) and 1 elevated calprotectin (2160 μg/g) that later normalized. In 6 of these 8 children, hemoglobin values normalized during iron supplementation.
      Forty-four nonanemic children (7.2%) had signs of iron deficiency based on low MCV and/or ferritin value and/or high plasma transferrin receptor value. Twenty of these 44 (45.5%) received a diagnosis compared with 39% of the remaining nonanemic children (P = .570).

      Discussion

      We found abnormal endoscopic and histologic findings at almost every biopsy sampling location and the number of final diagnoses to be more frequent in anemic than in nonanemic children. In addition, the anemia group presented significantly more often with growth problems (P < .001) and previously diagnosed autoimmune diseases (P = .036) and, of the laboratory parameters, hypersedimentation, hypoalbuminemia, increased fecal calprotectin (P < .001 each), and positive celiac antibodies with borderline significance (P = .058). Co-occurrences of these symptoms and markers, alongside diarrhea and melena/hematochezia, also further increased the likelihood of diagnoses in children with anemia. Although classically considered an “alarm signal,” data on the actual significance of anemia when considering pediatric GI endoscopies have been limited.
      • Thomson M.
      • Tringali A.
      • Dumonceau J.M.
      • et al.
      Paediatric gastrointestinal endoscopy: European Society for Paediatric Gastroenterology, Hepatology and Nutrition and European Society of Gastrointestinal Endoscopy guidelines.
      More evidence on this issue is particularly important in children because even if permanent adverse effects are relatively rare, these invasive and expensive procedures should be carried out only after careful assessment of the benefits and risks involved.
      • Thomson M.
      • Tringali A.
      • Dumonceau J.M.
      • et al.
      Paediatric gastrointestinal endoscopy: European Society for Paediatric Gastroenterology, Hepatology and Nutrition and European Society of Gastrointestinal Endoscopy guidelines.
      ,
      • Tringali A.
      • Balassone V.
      • De Angelis P.
      • et al.
      Complications in pediatric endoscopy.
      Although studies with a similar design are lacking, our results can, to some extent, be compared with studies investigating the yield of GI endoscopies at a more general level. In a small study by Wang et al,
      • Wang S.
      • Younus O.
      • Rawat D.
      • et al.
      Clinical presentation and outcomes of diagnostic endoscopy in newly presenting children with gastrointestinal symptoms.
      anemia was the most common laboratory finding in children undergoing EGD and/or colonoscopy and up to 75% of the 20 subjects with anemia were found to have abnormal histology. Nevertheless, neither the prevalence of anemia in patients undergoing EGD nor histologic changes in nonanemic subjects were presented. Noble and colleagues
      • Noble A.
      • Drouin E.
      • Tamblyn R.
      Design of predictive models for positive outcomes of upper and lower gastrointestinal endoscopies in children and adolescents.
      reported anemia as well as rectal bleeding, hypoalbuminemia, and hypersedimentation to predict endoscopic and/or histologic findings in colonoscopy, whereas age ≥13 years, hematemesis, dys/odynophagia, vomiting, and hypoalbuminemia, but not anemia, were associated with EGD abnormalities. Vomiting and dysphagia did not increase the likelihood of diagnoses here, but this may be because of the different study designs and the low frequency of these symptoms in our cohort. Moreover, in our setting systematic sampling from the middle part of the esophagus was not done before 2012. We also found children with anemia between 5 and 12 years old to be more likely to be given a diagnosis than were older children. This is at least partially explained by the fact that the median age of patients with anemia with celiac disease, our most common diagnosis, was 7.1 years. The constantly increasing number of pediatric IBD diagnoses may change the age distribution in the future.
      Celiac disease and IBD were the most frequent cause for anemia in our study. This differs from reports from Turkey,
      • Gulen H.
      • Kasirga E.
      • Yildirim S.A.
      • et al.
      Diagnostic yield of upper gastrointestinal endoscopy in the evaluation of iron deficiency anemia in older children and adolescents.
      Taiwan,
      • Huang S.
      • Yang Y.
      • Cheng C.
      • et al.
      The etiology and treatment of iron deficiency and iron deficiency anemia in children.
      Egypt,
      • Fayed S.B.
      • Aref M.I.
      • Fathy H.M.
      • et al.
      Prevalence of celiac disease, Helicobacter pylori and gastroesophageal reflux in patients with refractory iron deficiency anemia.
      and Italy,
      • Ferrara M.
      • Coppola L.
      • Coppola A.
      • et al.
      Iron deficiency in childhood and adolescence: retrospective review.
      where H pylori gastritis (11%-43%) and/or inadequate iron intake (16%-39%) were the leading causes, whereas celiac disease was common (19%) only in the Italian study
      • Ferrara M.
      • Coppola L.
      • Coppola A.
      • et al.
      Iron deficiency in childhood and adolescence: retrospective review.
      and IBD reported only in 1 Taiwanese child.
      • Huang S.
      • Yang Y.
      • Cheng C.
      • et al.
      The etiology and treatment of iron deficiency and iron deficiency anemia in children.
      Here dietary cause was rare, and H pylori was present in only 4.4% of children, which likely reflects improved nutrition, rapid decrease of H pylori infection,
      • Elitsur Y.
      • Dementieva Y.
      • Rewalt M.
      • et al.
      Helicobacter pylori infection rate decreases in symptomatic children: a retrospective analysis of 13 years (1993-2005) from a gastroenterology clinic in West Virginia.
      • Kawakami E.
      • Machado R.S.
      • Ogata S.K.
      • et al.
      Decrease in prevalence of Helicobacter pylori infection during a 10-year period in Brazilian children.
      • Oona M.
      • Utt M.
      • Nilsson I.
      • et al.
      Helicobacter pylori infection in children in Estonia: decreasing seroprevalence during the 11-year period of profound socioeconomic changes.
      and concurrent worldwide increase in immune-mediated diseases.
      • Burisch J.
      • Pedersen N.
      • Cukovic-Cavka S.
      • et al.
      Incidence and initial disease course of inflammatory bowel diseases in 2011 in Europe and Australia: results of the 2011 ECCO-EpiCom inception cohort.
      The differences observed between countries regarding the cause of pediatric anemia may thus reflect temporal variation in these changes. Interestingly, up to 16% of Italian children with anemia were diagnosed with cow’s milk allergy, the severe enteropathic form,
      • Kuitunen P.
      • Visakorpi J.K.
      • Savilahti E.
      • et al.
      Malabsorption syndrome with cow’s milk intolerance: clinical findings and course in 54 cases.
      which we currently find exceptional.
      The decision to proceed to endoscopy in children with suspicion of celiac disease or IBD is usually quite straightforward based on, respectively, either serology or specific symptoms and high blood and fecal inflammatory markers. In fact, the symptoms and markers we found to best predict diagnoses in children with anemia are the classic signs of these 2 conditions.
      • Husby S.
      • Koletzko S.
      • Korponay-Szabó I.R.
      • et al.
      European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease.
      ,
      • Levine Y.Y.
      • Koletzko S.
      • Turner D.
      • et al.
      ESPGHAN revised Porto criteria for the diagnosis of inflammatory bowel disease in children and adolescents.
      In these circumstances, the presence of anemia rarely influences the decision to perform endoscopy. Nevertheless, anemia predicts more severe histopathology in celiac disease
      • Rajalahti T.
      • Repo M.
      • Kivelä L.
      • et al.
      Anemia in pediatric celiac disease: association with clinical and histological features and response to gluten-free diet.
      • Saukkonen J.
      • Kaukinen K.
      • Koivisto A.M.
      • et al.
      Clinical characteristics and the dietary response in celiac disease patients presenting with or without anemia.
      • Abu Daya H.
      • Lebwohl B.
      • Lewis S.K.
      • et al.
      Celiac disease patients presenting with anemia have more severe disease than those presenting with diarrhea.
      and could thus support EGD in cases with borderline positive serology. In case of IBD suspicion, the hemoglobin level may affect the urgency of endoscopies. Although good noninvasive antigen tests for H pylori are available
      • Sabbagh P.
      • Javanian M.
      • Koppolu V.
      • et al.
      Helicobacter pylori infection in children: an overview of diagnostic methods.
      and the infection is a possible cause of unexplained anemia, other causes of anemia should be ruled out before testing by noninvasive or invasive methods.
      • Jones N.L.
      • Koletzko S.
      • Goodman K.
      • et al.
      Joint ESPGHAN/NASPGHAN guidelines for the management of Helicobacter pylori in children and adolescents (update 2016).
      Then again, many other possible causes can only be ruled out by endoscopic studies, including colonoscopy. This brings us to the broader question of when to perform only EGD or colonoscopy or both on children with anemia. The guidelines by the European Society for Paediatric Gastroenterology, Hepatology and Nutrition consider unexplained anemia to be an indication for both, but the evidence is scarce.
      • Thomson M.
      • Tringali A.
      • Dumonceau J.M.
      • et al.
      Paediatric gastrointestinal endoscopy: European Society for Paediatric Gastroenterology, Hepatology and Nutrition and European Society of Gastrointestinal Endoscopy guidelines.
      Resolving this issue was not among the aims of our study, and further research on it is needed.
      The decision on endoscopy is more difficult in the presence of less specific symptoms, particularly when anemia is the sole presenting sign. We found the coexistence of constipation, abdominal pain, or vomiting with anemia did not affect the likelihood of diagnoses, and reflux symptoms even reduced the probability. Consequently, uncomplicated reflux symptoms, “functional GI disorders,” and constipation as such are not indications for endoscopy in the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) guidelines.
      • Thomson M.
      • Tringali A.
      • Dumonceau J.M.
      • et al.
      Paediatric gastrointestinal endoscopy: European Society for Paediatric Gastroenterology, Hepatology and Nutrition and European Society of Gastrointestinal Endoscopy guidelines.
      Evidence on the endoscopic yield in anemic children without any GI symptoms is almost nonexistent, and we thus deem our results important. Notably, all 22 children without GI symptoms who received a diagnosis presented with additional laboratory abnormalities, whereas these were (temporarily) present in only 2 nondiagnosed children. This suggests that in children with mild or moderate anemia as a sole clinical sign/laboratory finding, a short period of observation with possible iron supplementation could be a safe option, particularly because GI malignancies are extremely rare in this age group. Instead, when “red flag” GI symptoms such as melena/hematochezia or failure to thrive are present, an endoscopy and possible adjunct testing should be conducted even without anemia or other laboratory abnormalities.
      The main limitations of the study are the retrospective design and data collection from only 1 tertiary center. Although the retrospective design enabled a large study cohort and a long-term follow-up, it limited our opportunities to evaluate the significance of individual laboratory parameters separately, because in clinical practice they, with the exception of almost always measured hemoglobin, are usually taken selectively based on the clinical scenario. This also hampered the opportunity to evaluate the subtypes of anemia and children presenting with signs of iron deficiency despite normal hemoglobin. Another problem is that the pediatric reference values for iron-related parameters vary and remain controversial.
      • Domellöf M.
      • Dewey K.G.
      • Lönnerdal B.
      • et al.
      The diagnostic criteria for iron deficiency in infants should be reevaluated.
      WHO
      Serum ferritin concentrations for assessment of iron status and iron deficiency in populations.
      WHO/UNICEF/UNU
      Iron deficiency anaemia: assessment, prevention and control, a guide for programme managers.
      WHO/CDC
      Assessing the iron status of populations: report of a joint World Health Organization/Centers for Disease Control and Prevention technical consultation on the assessment of iron status at the population level.
      It is also important to realize that some laboratory parameters, such as anemia and ESR, may interact. These limitations were, however, counterbalanced by the large number of consecutive endoscopies with systematic biopsy sampling regardless of previous laboratory test results and endoscopic findings as well as by the availability of comprehensive medical information, including long-term follow-up data. The importance of systematic sampling is emphasized by the fairly high number of children receiving a diagnosis despite macroscopically normal endoscopy. The study population was also clinically and socioeconomically diverse and represented all pediatric age groups. Although several endoscopists and pathologists were involved during the study period, the diagnostic heterogeneity should be reduced by our unified clinical practices and regular discussion of problematic cases in multidisciplinary meetings. It is nevertheless possible that the higher number of EGDs conducted in children with anemia biased the results. Furthermore, some histologic abnormalities reported may not have any major clinical relevance, and the percentage of ileum intubation rate was unacceptably low. It must also be mentioned that we did not evaluate the significance of fecal occult blood testing, because the benefits of this approach in children remain controversial.
      • Goddard A.F.
      • James M.W.
      • Mcintyre A.S.
      • et al.
      Guidelines for the management of iron deficiency anaemia.
      Altogether, when considering the generalizability of our results, the substantial countrywide variation in the diagnostic approach to and etiology of anemia has to be kept in mind.
      • McLean E.
      • Cogswell M.
      • Egli I.
      • et al.
      Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993-2005.
      To conclude, anemia increased the likelihood of a diagnosis in pediatric GI endoscopies. However, in a subgroup of children with nonspecific symptoms and no additional laboratory abnormalities, especially with anemia as the sole presenting sign, a period of noninvasive surveillance with a possible treatment trial might be considered. Furthermore, subsequent diagnosis after endoscopy with negative results appears to be exceptional even in the long term.

      Appendix

      Supplementary Table 1Finnish age- and sex-dependent pediatric reference values
      Fimlab-ohjekirja
      Perusverenkuva ja trombosyytit.
      for hemoglobin and mean corpuscular volume
      Hemoglobin (g/L)Mean corpuscular volume (fl)
      1-7 days150-23088-126
      7-30 days100-20685-123
      1-2 months95-13080-103
      2-6 months95-14176-97
      6-12 months100-14172-87
      1-2 years100-14273-87
      2-4 years100-14273-87
      4-13 years110-15573-95
      Boys 13-17 years130-16076-98
      Girls 13-17 years125-16078-102
      Supplementary Table 2Baseline characteristics of 251 anemic children who received or did not receive diagnoses after their first GI endoscopies
      Categorical variablesDiagnosis (n = 183)No diagnosis (n = 68)P value
      N%N%
      Girls9551.94058.8.329
      Autoimmune disease
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      147.768.8.760
      Asthma, allergy, or atopy3418.61014.7.473
      Other chronic disease2212.01014.7.571
      Intestinal disease in relatives
      Celiac disease or inflammatory bowel disease.
      3820.81014.7.278
      Positive EmA or TGA-IgA67
      Data from 137 diagnosed and 40 nondiagnosed patients.
      48.93
      Data from 137 diagnosed and 40 nondiagnosed patients.
      7.5<.001
      P ≤ .050.
      Fecal calprotectin > 100 μg/g60
      Data from 74 diagnosed and 20 nondiagnosed patients.
      81.15
      Data from 74 diagnosed and 20 nondiagnosed patients.
      25.0<.001
      P ≤ .050.
      Low MCV74
      Data from 162 diagnosed and 53 nondiagnosed patients.
      45.717
      Data from 162 diagnosed and 53 nondiagnosed patients.
      32.1.082
      Continuous variablesNMedian (quartiles)NMedian (quartiles)
      Age, years, (range)18310.8 (.03, 16.4)689.8 (.04, 16.7).757
      ESR, mm/hr12116.0 (7.0, 28.5)478.0 (3.0, 15.0).002
      P ≤ .050.
      Alanine aminotransferase, U/l13318.0 (13.0, 28.5)3921.0 (14.0, 31.0).236
      Albumin, g/l8135 (33, 39)2139 (37, 42).008
      P ≤ .050.
      EmA, Antiendomysial antibodies; ESR, erythrocyte sedimentation rate; MCV, mean corpuscular volume; TGA-IgA, IgA antibodies against transglutaminase 2.
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      Celiac disease or inflammatory bowel disease.
      Data from 137 diagnosed and 40 nondiagnosed patients.
      § P ≤ .050.
      Data from 74 diagnosed and 20 nondiagnosed patients.
      ǁ Data from 162 diagnosed and 53 nondiagnosed patients.
      Supplementary Table 3Relationships between symptoms and diagnosis adjusted for presence of anemia in 251 children with and 613 without anemia
      Odds ratio
      Calculated by multivariate binary regression analysis with 95% confidence intervals.
      for the diagnosis
      95% Confidence intervalP value
      Anemia3.502.50-4.91<.001
      P ≤ .050.
      Abdominal pain.91.67-1.23.539
      Diarrhea.85.62-1.16.302
      Melena/hematochezia2.511.53-4.14<.001
      P ≤ .050.
      Poor weight gain/growth3.241.69-6.20<.001
      P ≤ .050.
      Constipation.74.49-1.10.137
      Vomiting.92.59-1.45.302
      Reflux.78.50-1.21.270
      Other
      Other symptoms included, for example, fatigue, persistent fever, and dermatologic symptoms.
      .82.61-1.10.177
      Calculated by multivariate binary regression analysis with 95% confidence intervals.
      P ≤ .050.
      Other symptoms included, for example, fatigue, persistent fever, and dermatologic symptoms.
      Supplementary Table 4Characteristics of 30 children who had anemia without GI symptoms and received or did not receive diagnoses after their first GI endoscopies
      Categorical variablesDiagnosis (n = 22)No diagnosis (n = 8)P value
      N%N%
      Girls1568.2337.5.210
      Autoimmune disease
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      313.6112.51.000
      Asthma, allergy, or atopy313.600.545
      Other chronic disease313.6112.51.000
      Intestinal disease in relatives
      Celiac disease or inflammatory bowel disease.
      29.1112.51.000
      Positive EmA or TGA-IgA13
      Data from 18 diagnosed and 5 nondiagnosed patients.
      72.20
      Data from 18 diagnosed and 5 nondiagnosed patients.
      0.007
      P ≤ .050.
      Fecal calprotectin > 100 μg/g5
      Data from 5 diagnosed and 3 nondiagnosed patients.
      1001
      Data from 5 diagnosed and 3 nondiagnosed patients.
      33.3.107
      Low MCV12
      Data from 16 diagnosed and 6 nondiagnosed patients.
      75.03
      Data from 16 diagnosed and 6 nondiagnosed patients.
      50.0<.001
      P ≤ .050.
      Continuous variablesNMedian (quartiles)NMedian (quartiles)
      Age, y (range)229.4 (1.4, 15.3)812.3 (.04, 15.1).963
      ESR, mm/h912.0 (5.5, 21.0)55.0 (3.0, 16.5).180
      Alanine aminotransferase, U/L1122.0 (17.0, 29.0)517.0 (11.0, 81.5).609
      Albumin, g/L728.0 (32.0, 42.0)231.5 (21.00, —).766
      EmA, Antiendomysial antibodies; ESR, erythrocyte sedimentation rate; MCV, mean corpuscular volume; TGA-IgA, IgA-antibodies against transglutaminase 2; —, upper quartile could not be calculated because of the low number of patients.
      Type 1 diabetes, rheumatic or thyroidal disease, celiac disease.
      Celiac disease or inflammatory bowel disease.
      Data from 18 diagnosed and 5 nondiagnosed patients.
      § P ≤ .050.
      Data from 5 diagnosed and 3 nondiagnosed patients.
      ǁ Data from 16 diagnosed and 6 nondiagnosed patients.

      References

        • McLean E.
        • Cogswell M.
        • Egli I.
        • et al.
        Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993-2005.
        Public Health Nutr. 2009; 12: 444-454
        • Camascella C.
        Iron deficiency.
        Blood. 2019; 133: 30-39
        • Ferrara M.
        • Coppola L.
        • Coppola A.
        • et al.
        Iron deficiency in childhood and adolescence: retrospective review.
        Hematology. 2006; 11: 183-186
        • Gasche C.
        Iron, anaemia, and inflammatory bowel diseases.
        Gut. 2004; 53: 1190-1197
        • Thomson M.
        • Tringali A.
        • Dumonceau J.M.
        • et al.
        Paediatric gastrointestinal endoscopy: European Society for Paediatric Gastroenterology, Hepatology and Nutrition and European Society of Gastrointestinal Endoscopy guidelines.
        J Pediatr Gastroenterol Nutr. 2017; 64: 133-153
        • Franciosi J.P.
        • Fiorino K.
        • Ruchelli E.
        • et al.
        Changing indications for upper endoscopy in children during a 20-year period.
        J Pediatr Gastroenterol Nutr. 2010; 51: 443-447
        • Thomson M.
        • Sharma S.
        Diagnostic yield of upper and lower gastrointestinal endoscopies in children in a tertiary centre.
        J Pediatr Gastroenterol Nutr. 2017; 64: 903-906
        • Sheiko M.A.
        • Feinstein J.A.
        • Capocelli K.E.
        • et al.
        Diagnostic yield of EGD in children : a retrospective single-center study of 1000 cases.
        Gastrointest Endosc. 2013; 78: 47-54
        • Wang S.
        • Younus O.
        • Rawat D.
        • et al.
        Clinical presentation and outcomes of diagnostic endoscopy in newly presenting children with gastrointestinal symptoms.
        J Pediatr Gastroenterol Nutr. 2018; 66: 876-881
        • Fimlab-ohjekirja
        Perusverenkuva ja trombosyytit.
        (Available at:) (setid=5935. Accessed November 5, 2019)
        • WHO
        Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. Vitamin and Mineral Nutrition Information System.
        World Health Organization, Geneva, Switzerland2011
        • Saari A.
        • Harju S.
        • Mäkitie Outi
        • et al.
        Systematic growth monitoring for the early detection of celiac disease in children.
        JAMA Pediatr. 2015; 169: e1525
        • Taavela J.
        • Koskinen O.
        • Huhtala H.
        • et al.
        Validation of morphometric analyses of small-intestinal biopsy readouts in celiac disease.
        PLoS One. 2013; 8e76163
        • Furuta G.T.
        • Liacouras C.A.
        • Collins M.H.
        • et al.
        Eosinophilic esophagitis in children and adults: a systematic review and consensus recommendations for diagnosis and treatment.
        Gastroenterology. 2007; 133: 1342-1363
        • Walker-Smith J.A.
        • Guandalini S.
        • Schmizt J.
        • et al.
        Revised criteria for diagnosis of coeliac disease. Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition.
        Arch Dis Child. 1990; 65: 909-911
        • Rudolph C.D.
        • Mazur L.J.
        • Baker R.D.
        • et al.
        Guidelines for evaluation and treatment of gastroesophageal reflux in infants and children: recommendations of the North American Society for Pediatric Gastroenterology and Nutrition.
        J Pediatr Gastroenterol Nutr. 2001; 32: S1-S31
        • IBD Working Group of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition
        Inflammatory bowel disease in children and adolescents: recommendations for diagnosis—the Porto criteria.
        J Pediatr Gastroenterol Nutr. 2005; 41: 1-7
        • Valent P.
        • Sperr W.R.
        • Schwartz L.B.
        • et al.
        Diagnosis and classification of mast cell proliferative disorders: delineation from immunologic diseases and non-mast cell hematopoietic neoplasms.
        J Allergy Clin Immunol. 2004; 114: 3-11
        • Chapman J.A.
        • Berstein I.L.
        • Rufus E.L.
        • et al.
        Food allergy: a practice parameter.
        Ann Allergy Asthma Immunol. 2006; 96: S1-S68
        • Drumm B.
        • Koletzko S.
        • Oderda G.
        Helicobacter pylori infection in children: a consensus statement. European Paediatric Task Force on Helicobacter pylori.
        J Pediatr Gastroenterol Nutr. 2000; 30: 207-213
        • Husby S.
        • Koletzko S.
        • Korponay-Szabó I.R.
        • et al.
        European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease.
        J Pediatr Gastroenterol Nutr. 2012; 54: 136-160
        • Sackeyfio A.
        • Senthinathan A.
        • Kandaswamy P.
        • et al.
        Diagnosis and assessment of food allergy in children and young people: summary of NICE guidance.
        BMJ. 2011; 342: 544-545
        • Koletzko S.
        • Jones N.L.
        • Goodman K.J.
        • et al.
        Evidence-based guidelines from ESPGHAN and NASPGHAN for Helicobacter pylori infection in children.
        J Pediatr Gastroenterol Nutr. 2011; 53: 230-243
        • Vandenplas Y.
        • Rudolph C.D.
        • Di Lorenzo C.
        • et al.
        Pediatric gastroesophageal reflux clinical practice guidelines: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology.
        Hepatology, and Nutrition (ESPGHAN). J Pediatr Gastroenterol Nutr. 2009; 49: 498-547
        • Altman D.G.
        Practical statistics for medical research.
        Chapman and Hall, London, UK1991
        • Tringali A.
        • Balassone V.
        • De Angelis P.
        • et al.
        Complications in pediatric endoscopy.
        Best Pract Res Clin Gastroenterol. 2016; 30: 825-839
        • Noble A.
        • Drouin E.
        • Tamblyn R.
        Design of predictive models for positive outcomes of upper and lower gastrointestinal endoscopies in children and adolescents.
        J Pediatr Gastroenterol Nutr. 2008; 46: 409-413
        • Gulen H.
        • Kasirga E.
        • Yildirim S.A.
        • et al.
        Diagnostic yield of upper gastrointestinal endoscopy in the evaluation of iron deficiency anemia in older children and adolescents.
        Pediatr Hematol Oncol. 2011; 28: 694-701
        • Huang S.
        • Yang Y.
        • Cheng C.
        • et al.
        The etiology and treatment of iron deficiency and iron deficiency anemia in children.
        J Pediatr Hematol Oncol. 2010; 32: 282-285
        • Fayed S.B.
        • Aref M.I.
        • Fathy H.M.
        • et al.
        Prevalence of celiac disease, Helicobacter pylori and gastroesophageal reflux in patients with refractory iron deficiency anemia.
        J Trop Pediatr. 2008; 54: 43-53
        • Elitsur Y.
        • Dementieva Y.
        • Rewalt M.
        • et al.
        Helicobacter pylori infection rate decreases in symptomatic children: a retrospective analysis of 13 years (1993-2005) from a gastroenterology clinic in West Virginia.
        J Clin Gastroenterol. 2009; 43: 147-151
        • Kawakami E.
        • Machado R.S.
        • Ogata S.K.
        • et al.
        Decrease in prevalence of Helicobacter pylori infection during a 10-year period in Brazilian children.
        Arq Gastroenterol. 2008; 45: 147-151
        • Oona M.
        • Utt M.
        • Nilsson I.
        • et al.
        Helicobacter pylori infection in children in Estonia: decreasing seroprevalence during the 11-year period of profound socioeconomic changes.
        Helicobacter. 2004; 9: 233-241
        • Burisch J.
        • Pedersen N.
        • Cukovic-Cavka S.
        • et al.
        Incidence and initial disease course of inflammatory bowel diseases in 2011 in Europe and Australia: results of the 2011 ECCO-EpiCom inception cohort.
        J Crohn’s Colitis. 2014; 8: 588-597
        • Kuitunen P.
        • Visakorpi J.K.
        • Savilahti E.
        • et al.
        Malabsorption syndrome with cow’s milk intolerance: clinical findings and course in 54 cases.
        Arch Dis Child. 1975; 50: 351-356
        • Levine Y.Y.
        • Koletzko S.
        • Turner D.
        • et al.
        ESPGHAN revised Porto criteria for the diagnosis of inflammatory bowel disease in children and adolescents.
        J Pediatr Gastroenterol Nutr. 2014; 58: 795-806
        • Rajalahti T.
        • Repo M.
        • Kivelä L.
        • et al.
        Anemia in pediatric celiac disease: association with clinical and histological features and response to gluten-free diet.
        J Pediatr Gastroenterol Nutr. 2017; 64: e1-e6
        • Saukkonen J.
        • Kaukinen K.
        • Koivisto A.M.
        • et al.
        Clinical characteristics and the dietary response in celiac disease patients presenting with or without anemia.
        J Clin Gastroenterol. 2017; 51: 412-416
        • Abu Daya H.
        • Lebwohl B.
        • Lewis S.K.
        • et al.
        Celiac disease patients presenting with anemia have more severe disease than those presenting with diarrhea.
        Clin Gastroenterol Hepatol. 2013; 11: 1472-1477
        • Sabbagh P.
        • Javanian M.
        • Koppolu V.
        • et al.
        Helicobacter pylori infection in children: an overview of diagnostic methods.
        Eur J Clin Microbiol Infect Dis. 2019; 38: 1035-1045
        • Jones N.L.
        • Koletzko S.
        • Goodman K.
        • et al.
        Joint ESPGHAN/NASPGHAN guidelines for the management of Helicobacter pylori in children and adolescents (update 2016).
        J Pediatr Gastroenterol Nutr. 2017; 64: 991-1003
        • Domellöf M.
        • Dewey K.G.
        • Lönnerdal B.
        • et al.
        The diagnostic criteria for iron deficiency in infants should be reevaluated.
        J Nutr. 2002; 132: 3680-3686
        • WHO
        Serum ferritin concentrations for assessment of iron status and iron deficiency in populations.
        (Available at:) (Accessed August 31, 2019)
        • WHO/UNICEF/UNU
        Iron deficiency anaemia: assessment, prevention and control, a guide for programme managers.
        World Health Organization, Geneva2001 (Available at:) (Accessed August 31, 2019)
        • WHO/CDC
        Assessing the iron status of populations: report of a joint World Health Organization/Centers for Disease Control and Prevention technical consultation on the assessment of iron status at the population level.
        World Health Organization, Geneva2005 (Available at:)
        • Goddard A.F.
        • James M.W.
        • Mcintyre A.S.
        • et al.
        Guidelines for the management of iron deficiency anaemia.
        Gut. 2011; (Accessed August 31, 2019): 1309-1316