Abnormal 24-Hour pH-Impedance Testing Does Not Predict Reduced Quality of Life in Children with Reflux Symptoms

Abstract

Objectives:

Symptoms of gastroesophageal reflux (GER) have been associated with reduced health-related (QOL) in children, though it is unclear whether reflux burden on objective diagnostic testing such as 24-hour multichannel intraluminal impedance with pH (pH-MII) is associated with impaired QOL in children. Our aim was to identify predictors of QOL impairment based on the results of esophageal reflux testing.

Methods:

In this prospective, cross-sectional study of children undergoing pH-MII and esophagogastroduodenoscopy (EGD) for evaluation of suspected gastroesophageal reflux disease, we collected validated questionnaires (Pediatric Gastroesophageal Symptom and Quality of Life Questionnaire [PGSQ], Pediatric Quality of Life Inventory 4.0 [PedsQL] and PedsQL Gastrointestinal Symptoms Module [GI PedsQL]) to determine the relationship between objective measures of GER, patient reported symptoms and QOL outcomes.

Results:

Of the 82 subjects, 38% of children had an abnormal pH-MII study. There were no significant differences in QOL scores on any questionnaire between patients with normal and abnormal pH-MII studies (P > 0.11). There was no correlation between individual pH-MII parameters and QOL scores. Subjects with gross esophagitis on EGD reported significantly worse QOL with lower total PedsQL (P = 0.002) and GI PedsQL (P = 0.03) scores. Microscopic esophagitis was not associated with differences in QOL scores (P > 0.32).

Conclusions:

There was no relationship between abnormalities on pH-MII testing and reduced QOL in children. These findings highlight the importance of considering other diagnoses beyond GERD as they may be significant drivers for QOL impairment.

Introduction

Gastroesophageal reflux disease (GERD) is often implicated as a cause for symptoms such as chest pain, regurgitation, heartburn and epigastric abdominal pain in children, adolescents and young adults. These symptoms result in frequent visits to healthcare providers, multiple medication trials and the performance of diagnostic testing to clarify the etiology behind the symptoms. These symptoms have also been associated with reduced health-related quality of life (QOL) in children ¹. Two of the most common diagnostic tests ordered to evaluate for GERD are the 24-hour multichannel intraluminal impedance with pH test (pH-MII) and esophagogastroduodenoscopy (EGD). Despite these commonly recommended tests, little data exists to determine if the presence of reflux by either test is associated with impairments in health-related QOL. Furthermore, we have previously shown that the majority of children who present with symptoms of reflux do not have abnormal diagnostic testing (i.e. have functional heartburn or reflux hypersensitivity), though remain quite symptomatic ². Therefore, we hypothesized that symptomatic patients with normal testing may have worse QOL impairment and that there is a discrepancy between test results and impairment. The aim of this study was to identify predictors of QOL impairment based on the results of esophageal reflux testing with pH-MII or EGD.

Methods

Study Design and Participants

Patients ages 5-20 were recruited between January 2008 - December 2017 if they were undergoing pH-MII to evaluate for suspected GERD. Most patients also underwent an EGD at the time of pH-MII testing. Subjects were referred for diagnostic testing by their primary gastroenterologist due to persistent bothersome symptoms. Patients completed validated pediatric QOL questionnaires at the time of testing. Demographic and medication use data were collected at enrollment. Response to previous medication trials was obtained from review of the medical record. Subjects were excluded from analysis if they had a history of thoracic or abdominal surgery, enteral tube, esophageal motility disorder, a known diagnosis of eosinophilic esophagitis prior to testing, did not report symptoms during pH-MII testing or if they did not complete any QOL questionnaires. The protocol was approved by the Boston Children’s Hospital Institutional Review Board and all participants provided written informed consent.

Quality of Life Questionnaires

Patients or their guardian, depending on age, were asked to complete the Pediatric Gastroesophageal Symptom and Quality of Life Questionnaire (PGSQ), the Pediatric Quality of Life Inventory 4.0 (PedsQL) and the PedsQL Gastrointestinal Symptoms Module (GI PedsQL). All questionnaires were scored by two blinded reviewers.

The PGSQ is a validated questionnaire which includes a parent/caregiver report for children 2-8 and a patient report for children 9-17 ³. The PGSQ provides subscales for symptom burden, impact on everyday life, school impact (if applicable) and a total score. Higher PGSQ subscale scores represent more frequent symptoms and more severe impact on QOL. The total score is calculated as the sum of the symptom and everyday life impact subscales divided by the number of questions answered for these sections. School impact is not included in the total score, as not all patients completing the survey attend school. If respondents answered less than 50% of the survey questions, that questionnaire was not scored. Higher PGSQ scores are indicative of worse QOL.

The PedsQL is a 23-item self-report or parent/proxy-report with subscales for physical health, psychosocial health and a total score and has been validated in healthy controls and a number of acute and chronic medical conditions ^4,5. The GI PedsQL is a 74-item questionnaire which includes GI-specific domains with subscales for individual GI symptoms and worry about symptoms ⁶. The total score is calculated as the sum of all subscales divided by the number of items answered. If more than 50% of the items in the survey were unanswered, that questionnaire was not scored. For both the PedsQL and GI PedsQL, higher scores represent better QOL. Previously published cut-off values for the PedsQL were used to define patients at risk for clinically significant QOL impairments ^7,8.

Multichannel Intraluminal Impedance with pH Testing

All patients underwent pH-MII testing. Testing was considered to be done “off” proton pump inhibitor (PPI) therapy if the PPI was stopped for a minimum of 48 hours prior to testing, otherwise testing was considered “on” PPI therapy ⁹. The pH-MII studies were analyzed and pH-MII parameters (acid, nonacid, pH-only and full column reflux episodes) were defined as previously reported ². The pH portion of the study was considered abnormal if the pH was < 4 for ≥ 6% of the study for patients on and off PPI ¹⁰. The MII portion of the study was considered abnormal if there were ≥ 73 reflux episodes ¹¹. Symptoms reported within a 2-minute window of a reflux event were considered to be associated with that particular reflux event. The symptom index (SI) was used as the measure of symptom association and was calculated by dividing the number of reported symptoms associated with reflux by the total number of reported symptoms and multiplying by 100. A SI ≥ 50% was considered positive ¹². We elected to use SI over other symptom association measures (such as the symptom sensitivity index, SSI or the symptom association probability, SAP) because there is no evidence of superiority of any one measure and recent evidence that the SAP has poor association with clinical outcomes and response to therapy ¹³. Typical symptoms were defined as reports of heartburn, chest pain, nonspecific pain, abdominal pain, regurgitation, vomiting, gagging or “reflux” during pH-MII testing. Reports of cough, throat pain, throat clearing, dyspnea or hoarseness during the pH-MII study were considered atypical symptoms.

Endoscopy and Biopsies

The majority of subjects (N = 76, 93%) underwent EGD within 6 months of pH-MII testing. Patients were considered to have gross esophagitis if esophageal erosions, ulcerations, exudate, erythema or pallor were seen during endoscopy. Esophageal biopsies were obtained in all patients. Those with eosinophils on high power examination were classified as having microscopic esophagitis.

Statistical Methods

QOL measures were assessed for normality by histogram, quantile-quantile plots and the Shapiro-Wilk test. Many QOL scores were not normally distributed; however, parametric and nonparametric group comparisons yielded similar results and only the former are presented. Continuous data are summarized as mean ± standard error (SE) unless otherwise noted, and categorical data as frequency and percentage.

Cross-tabulations of dichotomous variables often contained expected cell counts <5, rendering the Pearson chi-square test unsuitable. Fisher’s exact test was used instead for all such comparisons. Student’s t-test was used for two-group comparisons and analysis of variance (ANOVA) for three or more groups. A general linear model was used to investigate differences in QOL scores between patients with and without normal pH-MII, adjusted for age and PPI response. The association between QOL measures and each pH or impedance parameter was assessed by Spearman rank correlation. Ninety-five percent confidence limits were determined by Fisher’s z transformation and transformed back to original correlation units using the tangent function.

All statistical analysis was conducted with SAS version 9.4 (Cary, NC). Plots were created with SAS and Microsoft Excel 2016. Comparisons were 2-sided with P<0.05 indicating statistical significance.

Results

Eighty-two patients were included in this analysis (55% male, mean age 11, range 5 – 20). Subject characteristics are shown in Table 1. Sixty-seven patients were off PPI therapy during pH-MII testing and 14 were tested on PPI. Only one patient was taking a prokinetic, erythromycin.

Table 1.

Subject characteristics

	pH-MII OFF PPI	pH-MII ON PPI
	N = 67	N = 14	P
Demographics
Age, years	10.8 ± 0.5	10.4 ± 1.2	0.76
Female sex	25 (37%)	11 (79%)	0.007
Acid suppression trial
Previous PPI trial, n=72	58 (88%)	14 (100%)	0.34
Number PPIs trialed, n=72			1.00
1	47 (81%)	11 (79%)
2	7 (12%)	2 (14%)
3	4 (7%)	1 (7%)
Dose escalation of PPI during trial, n=72	22 (38%)	7 (50%)	0.55
Max dose during trial (mg/kg), n=72	0.92 ± 0.06	1.04 ± 0.13	0.36
No response to PPI trial, n=72	26 (45%)	6 (43%)	1.00
Endoscopy
EGD done within 6 months of pH-MII	62 (93%)	14 (100%)	0.58
Gross esophagitis, n=76	13 (21%)	5 (36%)	0.30
Microscopic esophagitis, n=76	14 (23%)	2 (14%)	0.72
Completion of QOL surveys
PGSQ	56 (84%)	14 (100%)	0.20
PedsQL	59 (88%)	12 (86%)	0.68
GI PedsQL	31 (46%)	10 (71%)	0.14

Shown are mean ± standard error or N (%)

P value from Student’s t-test and Fisher exact test, respectively

pH-MII = multichannel intraluminal impedance with pH; PPI=proton pump inhibitor; QOL = quality of life; EGD=esophagogastroduodenoscopy; PGSQ=Pediatric Gastroesophageal Symptom and Quality of Life Questionnaire; PedsQL=Pediatric Quality of Life Inventory 4.0; GI PedsQL=PedsQL Gastrointestinal Symptoms Module

Impact of pH-MII reflux parameters on QOL

Thirty-one (38%) patients had an abnormal pH-MII study based on either an abnormal impedance portion (n=7), abnormal pH portion (n=16) or both portions (n=8). Figure 1 shows the total and subscores among those with normal and abnormal pH-MII studies for the PGSQ (panel A), PedsQL (panel B) and GI Peds QL (panel C). There were no significant differences in QOL scores on any scale between patients with normal and abnormal pH-MII studies (P > 0.11 for all comparisons).

Figure 1.

Comparison of quality of life (QOL) measures by pH-MII results. Mean scores (plus 1 standard error) are shown for totals and subscores of the Pediatric Gastroesophageal Symptom and Quality of Life Questionnaire (PGSQ; n=71), the Pediatric Quality of Life Inventory 4.0 (PedsQL; n=71) and the PedsQL Gastrointestinal Symptoms Module (GI PedsQL; n=42). The school impact subscore of the PGSQ was completed for 57 subjects. Except for the PGSQ, higher scores indicate better QOL. There are no differences in any QOL measure by pH-MII result (P > 0.11 by Student’s t-test).

Associations between QOL score and pH-MII results are shown in Table 2. With or without adjustment for age and PPI response, there were no significant differences in mean total QOL score on the PGSQ, PedsQL or GI PedsQL in patients with normal pH-MII studies compared to those with an abnormal pH or abnormal reflux burden. These findings did not change when excluding subjects with pH-MII testing done on PPI therapy (P > 0.30). Regardless of PPI status, there were no significant correlations between mean total QOL scores on any scale and the total number of reflux episodes, % time pH < 4 or the bolus clearance time (BCT) (Spearman rank correlations < ∣0.36∣; P > 0.05)).

Table 2.

Associations between QOL score and pH-MII results

Abnormal vs. normal pH-metry	N (abnormal, normal)	Normal pH	Abnormal pH	P t	P GLM
PGSQ total score	51, 20	0.92 ± 0.08	1.08 ± 0.19	0.36	0.67
PedsQL total score	51, 20	75.2 ± 2.6	73.5 ± 4.3	0.74	0.97
GI PedsQL total score	33, 9	75.3 ± 2.2	66.4 ± 4.3	0.07	0.11
Abnormal vs. normal MII	N (abnormal, normal)	Normal MII	Abnormal MII	Pt	PGLM
PGSQ total score	61, 10	0.93 ± 0.08	1.15 ± 0.22	0.35	0.55
PedsQL total score	58, 13	74.7 ± 2.4	74.9 ± 5.3	0.96	0.87
GI PedsQL total score	35, 7	74.0 ± 2.1	70.8 ± 6.0	0.56	0.66
Abnormal vs. normal pH or MII	N (abnormal, normal)	Normal pH and MII	Abnormal pH or MII	Pt	PGLM
PGSQ total score	47, 24	0.92 ± 0.09	1.06 ± 0.16	0.41	0.72
PedsQL total score	44, 27	75.4 ± 2.9	73.5 ± 3.5	0.68	0.98
GI PedsQL total score	30, 12	74.4 ± 2.3	70.9 ± 4.2	0.43	0.57

Shown are mean ± standard error

QOL = quality of life; MII = multichannel intraluminal impedance

P_t: P value from Student’s t-test.

P_GLM: P value from a general linear model, adjusted for age and proton pump inhibitor response.

Impedance parameters in patients with and without impaired QOL based on previously published cut-off values ⁸ is shown in Table 3. These findings did not change when subjects tested on PPI therapy were excluded from analysis (P > 0.20; data not shown). There were no significant differences in the proportion of acid reflux episodes in patients with and without impaired QOL (p = 0.45).

Table 3.

Impedance parameters based on PedsQL cut-off total score

	Impaired QOL (PedsQL score <70) n=20	Nonimpaired QOL (PedsQL score ≥ 70) n=51	P
On PPI for pH-MII	4 (20%)	8 (16%)	0.73
No. total reflux episodes	62 ± 15	44 ± 3	0.26
No. acid reflux episodes	22 ± 4	24 ± 2	0.73
No. nonacid reflux episodes	40 ± 15	21 ± 3	0.23
No. pH only episodes	9 ± 3	9 ± 1	0.82
% pH < 4	6 ± 2	4 ± 1	0.51
% full column reflux	42 ± 5	42 ± 3	0.99

Shown are N (%) or mean ± standard error.

P value from Fisher exact test and Student’s t-test, respectively.

PedsQL = Pediatric Quality of Life Inventory 4.0; QOL = quality of life; PPI=proton pump inhibitor; pH-MII = multichannel intraluminal impedance with pH

Impact of typical and atypical symptoms on QOL

Fifty-five (67%) patients reported typical symptoms of GER during the pH-MII study. There were no significant differences in total PGSQ, PedsQL or GI PedsQL scores in patients with a positive SI for typical symptoms (P > 0.34).

Fifty-four (66%) patients reported atypical symptoms during pH-MII testing. Patients with a SI ≥ 50% for atypical symptoms did not have a worse QOL compared to patients with a SI < 50%, on any scale (P > 0.06).

Twenty-eight (34%) patients reported typical symptoms only, 27 (33%) reported atypical symptoms only, and 27 (33%) reported both typical and atypical symptoms. There were no differences in total QOL scores on any scale based on type of reported symptom (P > 0.40 by ANOVA). There were no significant differences in the mean number of full column reflux episodes between patients with typical and atypical symptoms (P = 0.12). There also were no differences in any impedance parameter when comparing subjects by symptom type (typical only, atypical only, or both) (P > 0.13 by ANOVA).

Impact of prior PPI use on QOL

Of the 72 subjects who trialed a PPI prior to diagnostic testing, 40 (56%) experienced at least some symptomatic improvement with the PPI. Patients who experienced symptomatic improvement had a higher QOL on the PedsQL compared to non-responders (79.6 ± 2.6 vs 68.0 ± 4.1, respectively; P = 0.02). There were no differences in mean total PGSQ (P = 0.69) or GI PedsQL scores (P = 0.93) between PPI responders and nonresponders.

Impact of esophagitis on QOL

Seventy-six (93%) patients had an EGD within 6 months of pH-MII testing; 67 (88%) had the EGD and pH-MII performed on the same day, 4 (5%) had the EGD done after and 5 (7%) had the EGD done before the pH-MII study. In 1 subject the pH-MII study was done off PPI therapy but a PPI was started prior to the EGD done 10 days later. In the remainder of subjects who had the EGD and pH-MII done on separate days, both studies were done on the same therapy (i.e. either both “on” or both “off” PPI). Of those who had an EGD, 52 (68%) used a PPI within 1 month of the study. There were no significant differences in rates of gross (25% vs. 11%, P=0.32) or microscopic (23% vs. 22%, P=1.00) esophagitis in subjects who used a PPI within 1 month of their EGD and those who did not.

There were significant differences in QOL scores among 18 (24%) patients with gross esophagitis. Patients with gross esophagitis reported worse QOL, with lower total PedsQL (61.4 ± 5.0 vs. 78.3 ± 2.3; P = 0.002) and total GI PedsQL (64.6 ± 3.4 vs. 75.4 ± 2.4; P = 0.03) scores. There were no differences in total PGSQ (P = 0.34), PedsQL (P = 0.32) or GI PedsQL (P = 0.55) scores in patients with and without microscopic esophagitis on histology. There was no significant correlation between the number of eosinophils found on distal esophageal biopsy and total QOL scores on any scale (Spearman rank correlations < ∣0.24∣; P > 0.12).

Discussion

This is the first study to show that abnormalities on pH-MII testing are not associated with impaired QOL in children, adolescents and young adults. Similar findings have been reported in adult studies, where patient-reported QOL outcomes were assessed in the context of pH-MII testing. In a study by Kim et al., 45 adults with clinical symptoms of laryngopharyngeal reflux underwent pH-MII testing and completed validated QOL surveys ¹⁴. There were no significant correlations between the total reflux time or acid exposure time and QOL scores. Yadlapati et al. also found similar results in a study of 192 adults with PPI-unresponsive symptoms; QOL was impaired in patients who reported greater symptom severity but not in those with abnormal acid exposure time or positive symptom-reflux association on pH-MII testing ¹⁵. Finally, in a study of 225 adults with reflux symptoms, Kessing et al. found no significant differences in acid exposure time, number of symptoms or symptom reflux association, based on scores on a validated depression and anxiety questionnaire ¹⁶.

We also found no significant differences in QOL and pH-MII parameters when using a clinically significant cut-off value for the PedsQL. Huang et al. established clinically meaningful values for PedsQL by using information on health status, parent report of a child’s functional ability, healthcare service utilization and telephone interviews ⁸. Cut-offs varied by age and disease severity. We elected to use the most conservative cut-off identified by these investigators, a PedsQL score of 70, which was reported in children with the most severe chronic conditions (i.e those who would be expected to have the most impaired QOL). Even with this conservative cutoff, there were no differences in any pH-MII parameter.

Our pediatric study adds to the adult evidence in suggesting that symptoms can impact QOL independently of reflux burden. One hypothesis is that symptom severity, frequency or duration may be bigger drivers for QOL impairments in children than physiologic abnormalities, a notion supported by functional gastrointestinal disorders (FGID) literature. Varni et. al. found that children with FGIDs self-reported lower QOL scores for all dimensions on the PedsQL compared to patients with organic gastrointestinal disorders¹. For example, healthy controls had a mean (± standard deviation) total PedsQL score of 85.6 ± 11.9 compared to 70.2 ± 17.0 in FGIDs and 78.0 ± 14.6 in organic GI disorders. These findings have been echoed in a similar adult study ¹⁷. Our level of impairment in children with reflux symptoms, with a total PedsQL score of 74.7 ± 18.5, is on par with these other FGID studies and shows a similar degree of impairment to previously reported mean total PedsQL scores for a number of chronic medical conditions including cancer (68.5 ± 19.2), asthma (68.8 ± 19.3), obesity (75.0 ± 14.5), diabetes (76.6 ± 14.1) and cardiac disease (79.4 ± 16.5)⁵.

Our study sheds light onto the impact of upper tract FGID on QOL. We have previously shown that most children who present with typical reflux symptoms and have normal endoscopic evaluations actually have functional heartburn or reflux hypersensitivity rather than pathologic esophageal acid exposure, suggesting that they may be PPI unresponsive ². In this study, we found that PPI-unresponsive pediatric patients had worse QOL compared to PPI responders, a finding that merits additional investigation, particularly as many of these patients did not have abnormal reflux testing. Adult studies have reported conflicting results with variable associations between PPI response and QOL^18-22. Our findings underscore the importance of considering additional diagnoses such as psychosocial factors and visceral hypersensitivity in children with symptoms of GER and QOL impairment.

This study assessed the impact of both typical esophageal and atypical extraesophageal symptoms on QOL. This is particularly important in the pediatric population, where symptoms of GER are often nonspecific and vary by age ^23,24. Previous studies have included only patients with either esophageal ^15,16 or extraesophageal ¹⁴ symptoms. To our knowledge, only one other study assessed differences in QOL based on symptom type. Jung et al. found worse QOL in patients reporting typical (heartburn or regurgitation) versus extraesophageal symptoms of GER, though pH-MII testing was not used to objectively assess whether symptoms were truly reflux related ²⁵. We did not find any significant differences in QOL scores based on typical versus atypical reflux symptoms in our population.

Strengths of our study include a large pediatric sample size, use of several different QOL scales, using pH-MII as an objective test for reflux and including both patients with both typical and atypical reflux symptoms. Our study does have some limitations. First, while patients were prospectively enrolled in the study, information about PPI responsiveness was obtained from review of the medical record and relied on parent and physician recall which may have an impact on the true rates of PPI responsiveness ²⁶. Second, as patients in this study were referred for testing because of persistent symptoms, the proportion of PPI non-responsiveness and the rate of functional diagnoses may be higher than expected. However, we feel this population is typical of what a pediatric gastroenterologist encounters in clinical practice. Third, some subjects in this study were on acid suppression at the time of their pH-MII study (14/82) and endoscopy (52/76) and as a result rates of esophagitis on EGD and acid burden on pH-MII testing may be underestimated. Fourth, while we did find that gross esophagitis in general was associated with worse QOL scores, only a small subset of these patients had erosive esophagitis (2/18); the remainder had milder findings such as erythema, edema or nodular mucosa.

As there are no normal impedance values in children, our cutoff values were extrapolated from adult studies, which may be a limitation. Finally, because our study enrollment period spanned many years, different combinations of QOL questionnaires were used and not every patient completed all QOL scales.

In conclusion, we did not find significant associations between QOL impairments in pediatric patients with reflux symptoms and objective measurements of reflux. These results suggest that providers should consider other factors which could be contributing to reduced QOL in these patients.

What is Known

• Symptoms of gastroesophageal reflux (GER) are associated with reduced health-related quality of life (QOL) in children.

• Our aim was to determine if there was a correlation between QOL impairment and objective measures of reflux on diagnostic testing.

What is New

• Patients with abnormal pH-impedance testing did not have reduced QOL scores on validated questionnaires. Gross, but not microscopic, esophagitis was associated with QOL impairments.

• While reflux is frequently implicated as a cause for symptoms and worse QOL, this study shows that symptoms may impact QOL independently of reflux burden.

Abbreviations:

GERD

gastroesophageal reflux disease

GER

gastroesophageal reflux

QOL

quality of life

pH-MII

multichannel intraluminal impedance with pH

PGSQ

Pediatric Gastroesophageal Symptom and Quality of Life Questionnaire

PedsQL

Pediatric Quality of Life Inventory 4.0

GI PedsQL

Pediatric Quality of Life Inventory Gastrointestinal Symptoms Module

PPI

proton pump inhibitor

EGD

esophagogastroduodenoscopy

SI

symptom index

BCT

bolus clearance time