Biochemical Manifestations of Gastroesophageal Reﬂux Disease Progression in Children: A Single Center Case-Control Study

: Gastroesophageal reﬂux disease (GERD) is a common digestive condition, representing one of the most frequent reasons for medical examination, especially in pediatric gastroenterology departments. GERD could be associated with biochemical alterations representing either its systemic manifestations or markers of complications. The aim of our paper was to evaluate biochemical parameters secondary to GERD in children. Two hundred and sixty-seven children of both genders aged between 1 month and 18 years who displayed suggestive symptoms for this condition were included in the study and were monitored for a period of 5 years. Depending on the range of symptoms and technical possibilities, the following procedures / investigations were performed: esophageal pH monitoring and imagistic or endoscopic examination, besides speciﬁc biochemical investigations. The cohort was sub-divided into two groups: one that included 213 children with conﬁrmed GERD who represented the study group and 54 healthy children where GERD had been excluded, the control group. Out of all the investigated children, 39.0% displayed low hemoglobin values, 43.7% displayed low values of erythrocyte indices (MCH), and 68.5% had increased erythrocyte sedimentation rate (ESR) values, while increased eosinophil levels were recorded in 46.9% of the cases. Such parameters were proven to be a biomarker of suspected eosinophilic esophagitis, whereas 32.9% of the cases displayed high blood glucose values that could be correlated with gastroesophageal reﬂux symptoms. Other measured parameters (such as magnesium, aminotransferases and proteins) remained within the normal limits, without statistically signiﬁcant di ﬀ erences compared to in the control group. This condition is diagnosed based on invasive investigations, which are often di ﬃ cult to accept by the patients’ parents. The biochemical modiﬁcations correlated to the clinical manifestations can anticipate the progression of the disease, thus limiting the necessity of performing invasive diagnosis tests.


Introduction
Based on the consensus of the two medical societies of the field-the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) and North American Society for Pediatric

Materials and Methods
The retro-prospective case-control study was performed on a cohort of 267 children of both genders, aged between 1 month and 18 years, from the area of Moldova, who were referred to "Sf. Maria" Children's Hospital of Iasi over a period of 5 years (January 2014-December 2018). They had clinical symptoms suggestive of reflux disease and were diagnosed with the available technical possibilities during their hospitalization; that is, with barium meals, upper digestive endoscopy or esophagus pH-monitoring. The cohort was divided into two groups of patients, one consisting of 213 children diagnosed with gastroesophageal reflux disease representing the study group, and the second, consisting of 54 healthy children, was the control group.
The main criteria for inclusion in the study were the age between one month and 18 years and suggestive symptoms such as heartburn, regurgitation, weight gain failure, abdominal or retrosternal chest pain, dysphagia and extra-digestive manifestations, namely respiratory symptoms such as coughing, laryngitis, wheezing and neurobehavioral symptoms. The exclusion criteria consisted of previously diagnosed gastroesophageal reflux that was under treatment at the time and the identification of anatomical abnormalities or a previous surgical intervention in the upper gastrointestinal tract.
In the two groups, we focused on the correlation between the disease and certain biochemical alterations such as aminotransferases, serum proteins, the blood glucose level, and magnesium. From an immune-chemical perspective, we analyzed immunoglobulin levels for a possible correlation with the immunologic statuses of the patients. Appropriate qualitative blood samples were taken from all the patients, as follows: • Five milliliters of blood for the full blood count and determining the erythrocyte sedimentation rate (ESR) in EDTA vacuum blood collection tubes, processed on the Mindray 500; • Five to ten milliliters of blood for biochemistry and an immunogram in simple vacuum blood collection tubes, without anticoagulant, processed with the Architect and BA400; • Five milliliters of blood for fibrinogen determination in Na citrate vacuum blood collection tubes, which were processed with the Sysmex XT 4000i.
All the parents or legal representatives of the patients included in the prospective study signed informed consent forms for participation. The study was approved by the Ethics Commission of "Sf. Maria" Children's Hospital, approval no. 15203 of 24.06.2016 in compliance with the ethical and deontological rules for medical and research practice. The study was conducted in accordance with the Helsinki Declaration and with several published principles [10][11][12][13][14][15].

The Analyzed Cohort
The analyzed cohort included 267 children, 174 boys (65.2%) and 93 girls (34.8%). The higher number of boys indicates the gender-based vulnerability to digestive symptoms. This aspect confirms the data provided by the literature in the field, according to which the degree of morbidity in male patients is generally higher than that in female patients in childhood (Pearson chi-square = 5.288, p = 0.021).
This group was further subdivided into two groups, one that included 213 children diagnosed with gastroesophageal reflux disease and a second group, used for reference, which included 54 children in which the investigation did not confirm the diagnosis. The group of children with GERD consisted of 68.5% male and 31.5% females, and in the control group, there were 51.9% boys and 48.1% girls.
Homogeneity according to age was assessed, with the all age categories being represented, the highest proportions (with almost similar percentages) belonging to the age categories of 1.1-3 years (21.3%) and 3.1-7 years (30.3%). The ages did not present a homogenous distribution; the minimum value was 1 month, and the maximum value was 17 years and 3 months (207 months). The average value in the study group was 65.05 months (about 5 years of age).

Hemoglobin
Lower values of hemoglobin (Hb) were recorded in GERD patients (39.0%) as compared to in the control group (16.7%). The differences, despite not being very high, are statistically significant (Mann-Whitney U = 2708.000, p = 0.000). Thirty-nine percent of the GERD patients had altered hemoglobin levels, whereas only 16.7% of the patients in the control group displayed lower hemoglobin levels.

Hematocrit
As far as the hematocrit (Ht) values are concerned, slightly lower levels were recorded in GERD patients. However, these differences were not statistically significant (Student's t = −0.680, p = 0.497). Moreover, the slightly lower number of red blood cells in the GERD patients as compared to that in the control group was also regarded as not statistically significant (Mann-Whitney U = 5273.500, p = 0.346).

Erythrocyte Indices
The mean corpuscular hemoglobin (MCH) had statistically significantly lower values in the GERD patients than in the control group (Mann-Whitney U = 3692.500, p = 0.000). On average, this parameter stayed within normal limits, rather towards the lower limit (23.0985 ± 3.39359 SD); we noted that a high percentage of the GERD patients (43.7%) had modified MCH values as compared to the control group (11.1%). In addition, statistically significantly lower values of the mean corpuscular hemoglobin concentration (MCHC) were noted in GERD patients as compared to in the control group (Mann-Whitney U = 4003.00, p = 0.001), although in the latter's case, the values remained within the normal variation limits (34.3258 ± 2.12420 SD). Moreover, the mean corpuscular volume (MCV) had statistically significant lower values in the GERD patients as compared to in the healthy patients (t-test t = −5.124, p = 0.000), whereas the average MCV value remained within normal limits (76.3646 ± 6.75707 SD).
The values for the white cell counts are slightly higher in the GERD group, yet the increase is not statistically significant (t-test t = 1.774, p = 0.079). The analysis of the eosinophils (Eo) in the GERD group showed statistically significant higher values than those in the control group (Mann-Whitney U = 3660.500, p = 0.000). Moreover, we noted that the average eosinophil value in the GERD patients was above the normal value limits (7.2638 ± 3.80711 SD). The data displayed in Table 1 indicate that there are no statistically significant differences between the platelet levels in GERD patients and those in the patients in the control group (Mann-Whitney U = 5581.000, p = 0.737).

Inflammatory Markers
The erythrocyte sedimentation rate (ESR) in the GERD patients had statistically significantly increased values (Mann-Whitney U = 2592.500, p = 0.000). Moreover, the average value of the ESR exceeds the maximum normal limit (21.243 ± 16.1342 SD). Table 2 indicates an average value of fibrinogen (Fg) above the normal limit (405.4225 ± 159.62580 SD), with the percentage of patients displaying this modified parameter regarded as statistically non-significant in the GERD group (33.3%) as compared to in the control group (46.9%).

Serum Iron
Serum iron was monitored in all patients that displayed modifications of the erythrocyte values, even if the values for hemoglobin and hematocrit were within the normal limits. The serum iron values in patients with gastroesophageal reflux disease are statistically significant lower than the values in the patients included in the control group (Mann-Whitney U = 3250.500, p = 0.000). The results indicate that a significant percentage of the GERD patients display modified serum iron values (12.2%), as compared to in the control group, where this parameter was within the normal limits in all the patients.

Aminotransferases
The average value for the alanine aminotransferase (ALT) parameter remained within the normal limits, with a very low percentage of patients displaying modification in both groups (9.3% in the control group, 5.6% in the active group). The data included in Table 3 indicate that the recorded difference is not statistically significant (t-test t = 1.341, p = 0.184). However, statistically significantly higher values were recorded for the aspartate aminotransferase (AST) parameter in GERD patients (Mann-Whitney U = 4640.000, p = 0.028), with an average value close to the upper limit yet within the normal range (33.5127 ± 11.55889 SD).

Proteins and Electrophoresis
Very similar values are observed for total proteins for both groups ( Table 4). The differences are not statistically significant (Mann-Whitney U = 5527.000, p = 0.659), whereas the average values for total proteins range towards the upper limits of the total values (60.3390 ± 13.31961 SD and 59.6593 ± 13.08089 SD, respectively). Slightly lower levels were recorded for albumin, yet without statistically significant differences (Mann-Whitney U = 5079.500, p = 0.185). The mean values of this parameter for both groups rest within the medium range of normal values (58.1653 ± 8.24658 SD and 59.9148 ± 8.38362 SD, respectively).

Immunoglobulins
In GERD patients, we also noticed slightly higher, yet statistically non-significant, alpha 1 globulin (α1) values (Mann-Whitney U = 5178.500, p = 0.259), with an average value very close to the upper normal limit (3.9671 ± 1.43535 SD). As far as beta globulins are concerned, slightly higher, yet statistically non-significant, values were recorded in GERD patients; however, these values rest within the normal variation range in both groups (10.6202 ± 2.34302 SD in the GERD group, and 10.5907 ± 2.66962 SD in the control group). The gamma globulin (γ) values are statistically significantly lower in the GERD patients (Mann-Whitney U = 3584.500, p = 0.000), although the average value of this parameter remains within the normal limits in both groups (11.1042 ± 3.73209 SD in the GERD patients as compared to 13.5556 ± 3.43026 SD in the control group patients).
Regarding immunoglobulin G (IgG), the values are statistically significantly higher in the study group (Mann-Whitney U = 1048.000, p = 0.000), with the average value within the normal range for GERD patients (690.0845 ± 93.46931 SD), as compared to the lower average value in the patients included in the control group (553.1852 ± 48.41958 SD). Statistically significantly higher values were also recorded for immunoglobulin M (IgM) in the GERD group (Mann-Whitney U = 1741.500, p = 0.000), with an average value for this parameter within the normal variation range (110.0329 ± 40.64009 SD), whereas the average value is lower in the control group (64.5370 ± 18.90464 SD). The values for immunoglobulin E (Ig E) are, in turn, statistically significantly higher in the GERD patients (Mann-Whitney U = 1002.500, p = 0.000), with a mean average significantly exceeding the normal variation range (123.4178 ± 53.42804 SD), while the average value is normal in the control group patients (50.8333 ± 23.04364 SD).

Blood Glucose
There is no statistically significant difference between the blood glucose levels in the two groups (Mann-Whitney U = 5050.000, p = 0.166), although the blood glucose levels in the GERD patients are slightly higher. In both groups, the average blood glucose levels remain within the normal variation range. However, the percentage of patients displaying modified blood glucose levels is higher in the GERD group (32.9%) as compared to in the control group (24.1%).

Magnesium
The values recorded for magnesium (Mg) are closer to the upper limit of the normal variation range (2.1319 ± 0.44020 SD) in GERD patients as compared to the control group, where the average magnesium values remain closer to the lower limit of the normal variation range (1.9635 ± 0.40103 SD). The percentage of patients displaying modified values for this parameter is lower in the GERD group (26.8%) as compared to in the control group (35.2%); however, this difference is not statistically significant.

Correlation of Disease Progression According to Therapeutic Approach
In order to evaluate the progression/evolution of GERD by using biochemical markers, patients were reassessed at 3 months, 6 months and 9 months after recruitment in the study. Out of the 267 patients, only 65 met the follow-up conditions: initial evidence of specific endoscopic lesions, the provision of informed consent for upper digestive endoscopy, initial modifications of biochemical markers, compliance with therapy, and attendance at the scheduled follow up for endoscopic and biological investigations.
All the 65 patients were assessed by monitoring biochemical markers, which presented modified values (eosinophils, ESR and IgE revealed raised values, and Hb, Ht, MCH and MCHC had lower values), and by upper digestive endoscopy to identify specific endoscopic lesions. According to the therapeutic approach, the 65 patients were allotted to three subgroups: The first subgroup, made up of 42 children, were given a strict diet and lifestyle changes. The 3 month follow up revealed Grade II esophagitis lesions, and the biochemical markers did not have modified values, in comparison with those at the initial evaluation. PPI therapy was given, and the second evaluation revealed Grade I esophagitis lesions, with improved biochemical markers. The third assessment was normal; the values of the biochemical markers were also within the normal limits, except for the fact that in 29 cases, the values of Hb, Ht, MCH and MCHC remained low (Table 5).  The second subgroup, made up of 15 children, received a PPI therapy; the 3 month and 6 month evaluations identified biochemical markers within normal limits, except for in four children with lower values of Hb, Ht, MCH and MCHC. The endoscopic appearance showed normal esophageal mucosa ( Table 6).  In eight of the patients included in Subgroup 3, upper endoscopy revealed Grade I esophagitis. The 3 month follow up revealed alterations in both the endoscopic appearance and biochemical markers. PPI therapy was administered, and the second evaluation showed improvement. Both the endoscopic appearance and the biochemical markers were normal at the third evaluation (Table 7). Table 7. Correlation of disease progression in third subgroup, which did not receive any treatment in the first 3 months.

Mean
Std. Deviation Std. Error Mean p T0-3 m p T0-6 m p T0-9 m p 3-6 m p 3-9 m p 6-9 m  The decreases in the hemoglobin, hematocrit, MCH and MCHC values in both the first and the second sublots can be explained by the prolonged administration of the proton pump inhibitors.

Discussion
Regarding the hematological parameters, we point out the fact that the hemoglobin value is significantly lower in the GERD group as compared to in the control group (p = 0.017). Our study indicates that at least two erythrocyte indices values are significantly affected, namely the mean corpuscular hemoglobin concentration and the mean corpuscular volume, which can indicate a tendency towards iron deficiency anemia. The MCHC's behavior is similar to that of the hematocrit, with the same decreasing tendency, even if the decrease is not statistically significant.
In children diagnosed with GERD, anemia can be explained both by the bleeding occurring in peptic esophagitis and nutritional deficiencies, as well as by the toxicity induced by the aluminum content of the antacid medication [16,17]. In this case, the biological profile facilitates the diagnosis of a complicated evolution. Anemia can also be associated with the long-term administration of proton pump inhibitors, which represent the first-line therapy for GERD [18]. The study group included children diagnosed with GERD and no other conditions; therefore, the anemia was considered to be secondary to this disease.
The white cell values are slightly higher in the GERD patients, yet this increase is not statistically significant, whereas the eosinophils have statistically significant higher values in the GERD patients as compared to in the control group. The literature in the field speaks about the association between GERD and cow's milk protein allergy (CMPA), especially because there are clinical characteristics common to both disorders, such as regurgitation and vomiting. Up to 56% of the children diagnosed with GERD can also present CMPA symptoms, while the prevalence of food allergies in GERD children can reach 43% [19,20].
In this context, the higher eosinophil values indicated by our study can suggest an allergic component. At the same time, higher numbers of eosinophils can also represent a marker of eosinophilic esophagitis, especially when associated with other biomarkers, as suggested by relevant literature data [21,22]. To what extent the symptoms of GERD are determined by eosinophilic esophagitis is yet unknown. This is an important source of bias, as not all patients routinely benefit from endoscopy and biopsies to rule out eosinophilic esophagitis.
As indicated above, there were no statistically significant differences between the GERD group and the control group regarding the thrombocyte values. Similar studies do not quote significant differences between hemoglobin, the mean corpuscular volume and thrombocytes either in GERD children as compared to children who do not have this pathology or in children diagnosed with reflux that may or may be not associated with esophagitis [23].
The erythrocyte sedimentation rate was higher in the GERD patients, being one of the statistically significant modifications. Despite being slightly higher in GERD patients, fibrinogen values were not found to differ with statistical significance. On the other hand, anemia associated with the ESR can also suggest a systemic condition, and it thus becomes essential to exclude such a probability [24]. Comparative studies regarding the fibrinogen level in the middle ear fluid and in the blood of children diagnosed with pharyngo-laryngeal reflux and serous otitis media have been conducted, yet the results of these studies are not conclusive [25].
Regarding the serum iron values, the results seem more categorical, as the values recorded in the GERD patients are lower than the ones recorded in the control group and below the biological limit for serum concentrations, most probably due to anemia. Iron-deficiency anemia in GERD patients has also been described in other studies [26] and seems to be a common systemic manifestation of GERD in pediatric patients.
Within the study group, we did not detect any liver function test abnormalities. However, in children diagnosed with reflux, hepatitis and liver failure can occur as a result of the administration of H2 inhibitors [27].
In our study, the total protein values were not modified; nevertheless, albumin displayed slightly lower, yet statistically non-significantly so, values within both groups. However, the electrophoresis of proteins is a recommended test that increases the accuracy of diagnosis in immunodeficiency cases that can also occur in GERD [28]. Immunochemical tests indicated significant increases in all the immunoglobulins in the GERD patients, an aspect that could be interpreted as an increased antigenic stimulation induced by acid exposure in such patients. In particular, the Ig E increase could suggest the association with food allergies or asthma. However, according to some studies, food allergies with mainly gastrointestinal symptoms are non-IgE mediated [29].
The blood glucose levels do not show significant changes, as opposed to the magnesium levels, which tend towards the upper limit of the normal variation range in our study group patients. Recent studies have indicated that there may be statistically significant correlations between the severity of GERD symptoms and the basal blood glucose level, suggesting a genetic factor in the association of GERD with the metabolic syndrome [30].
Not least, GERD can also be associated with other less common manifestations, such as esophageal polyps [31]. Such cases were excluded from our cohort, as it would have been impossible to identify the real cause of the reflux symptoms, given the anatomical abnormalities.

Conclusions
Biochemical modifications in gastroesophageal reflux disease are correlated with disease progression. Our findings may improve the diagnosis of GERD complications, facilitating earlier effective therapy.
Further studies, performed on larger batches, would be necessary to confirm disease progression, using non-invasive tests, based on assessing different biochemical markers.