Gastroesophageal reflux disease (GERD) is a condition that develops when the reflux of gastric content causes troublesome symptoms.¹ It can be further classified into esophageal and extra-esophageal syndromes according to the Montreal Definition and Classification of GERD.² Esophageal manifestations such as heartburn and regurgitation are common. Laryngeal symptoms, including chronic cough, hoarseness, globus sensation, and laryngitis,^3,4 are the most common extra-esophageal symptoms. An association between laryngopharyngeal reflux (LPR) and GERD has been established.²

Since most laryngeal symptoms in the absence of apparent pathologies on laryngopharyngeal local examination are frequently suspected to be reflux-related, a proton-pump inhibitor (PPI) trial plus diet and lifestyle modifications is generally recommended.^5,6 With erratic clinical response to therapy, patients with refractory laryngeal symptoms are increasingly recognized by general physicians, lung specialists, gastrointestinal specialists, and otolaryngologists.⁷ In non-responsive cases, after assessing treatment compliance, patients could benefit from a personalized therapeutic scheme, as assessed by additional gastrointestinal surveys, including esophagogastroduodenoscopy (EGD), multichannel intraluminal impedance and pH (MII-pH) monitoring and esophageal manometry.⁵ MII-pH monitoring assesses the presence, proximal extent, and consistency of the refluxate in cases of acid and non-acid reflux.⁸ High-resolution impedance manometry (HRIM) measures both bolus movement and esophageal contractile activity, providing better characterization of esophageal motility disorders.⁹

Meanwhile, psychogenic problems and stress are thought to trigger the globus sensation.¹⁰ Rome IV¹¹ introduced the concept of functional esophageal disorders, which include heartburn and reflux hypersensitivity, and usually contribute to troublesome GERD symptoms despite PPI therapy.¹² Treatment strategies involving psychological interventions have been suggested to provide more symptomatic control when given in combination with anti-reflux medications.¹³ These lines of evidence suggest that psychological comorbidities may be included as an important target in the multidisciplinary approach in the case of patients who have troublesome laryngeal symptoms but who have failed PPI therapy.

In cases of persistent laryngeal symptoms despite PPI treatment, EGD and further reflux monitoring test is recommended in the most recent guideline by the American College of Gastroenterology.¹⁴ The purpose of this study was to comprehensively investigate the clinical features of such patients based on a combination of EGD, MII-pH monitoring, HRIM, and validated symptom questionnaires. Comparisons of clinical characteristics between patients with or without concurrent GERD symptoms were also evaluated.

Study Design and Participants

From December 2014, consecutive patients who were referred to the motility laboratory at the National Taiwan University Hospital (NTUH) for evaluation of persistent laryngeal symptoms, including hoarseness, throat clearing, throat pain, globus sensation in the throat, or chronic cough, despite a standard dose of PPI treatment for at least 8 weeks, were eligible for enrollment. To be included, the participants had to be of ≥ 20 years old, and not have any apparent pathologies in the pharynx and larynx on ear, nose, and throat (ENT) evaluation with laryngoscopy. Exclusion criteria consisted of major esophageal motility disorders and history of prior upper gastrointestinal surgical treatment. Also, patients with proven GERD on EGD (Los Angeles classification grade C and D, peptic strictures and Barrett’s esophagus), or abnormal acid exposure (> 6%) and excessive reflux number (> 80 in 24 hours) evidenced by ambulatory pH or pH-impedance monitoring based on the updated Lyon consensus were excluded considering such patients should be managed with GERD treatment.^15,16 Patients who remained on PPIs on enrollment were asked to cease the drug 2 weeks prior to the study (wash out period). All patients provided basic demographic information and underwent a comprehensive diagnostic work-up, consisting of validated symptom questionnaires, EGD, 24-hour MII-pH monitoring, and esophageal HRIM. EGD was performed by an experienced gastroenterologist (P.H.T.) to inspect the oropharynx, hypopharynx, larynx, esophagus, stomach, and duodenum. We adopted the Los Angeles classification, the most commonly used and validated guideline, to diagnose and grade the severity of esophagitis.^17,18 Validated symptom questionnaires were completed on the day of enrollment. The presence of concurrent GERD symptoms was defined with a gastroesophageal reflux disease questionnaire (GerdQ) score ≥ 8.¹⁹

In addition, we investigated the prevalence of psychological morbidity and sleep disturbance between patients with PPI-unresponsive laryngeal symptoms and healthy asymptomatic individuals, who were recruited through advertisement or by word of mouth. This study was approved by the Institutional Ethics Committee of NTUH (No. 201505131RND). All participants provided written informed consent prior to the study.

Symptom Evaluation

All participants were assessed using validated symptom questionnaires. Reflux symptom index (RSI) was used to evaluate and document the severity of laryngeal symotoms.²⁰ GerdQ involves 4 positive (heartburn, regurgitation, sleep disturbance due to reflux symptoms, and use of over-the-counter medication) and 2 negative (epigastric pain and nausea) predictors of GERD.¹⁹ The patient assessment of upper gastrointestinal symptom severity index (PAGI-SYM) is composed of 20 items and 6 subscales.²¹ In addition, 5-item brief symptom rating scale (BSRS-5) screens for common psychiatric morbidities such as anxiety, depression, and related disorders; a BSRS-5 score ≥ 6 indicates psychiatric morbidity.²² The Pittsburgh sleep quality index (PSQI) is a 7-item self-rated questionnaire assessing sleep disturbance over a 1-month period.²³ A global PSQI score > 5 indicates poor sleep quality and sleep disturbance.

Multichannel Intraluminal Impedance and pH Monitoring

The catheter contained a single pH and 8 ring electrodes (Medical Measurement Systems, Enschede, Netherlands). It was introduced transnasally and the pH sensor was placed about 5 cm above the lower esophagus sphincter (LES), which was confirmed with the esophageal HRIM. The 6 impedance segments defined the 6 zones along the esophagus, Z1 to Z6, which were centered at 17, 15, 9, 7, 5, and 3 cm above the LES, respectively. All impedance and pH data were recorded over the following 24 hours. Each MII tracing was automatically analyzed using the package analytic software (Medical Measurement Systems), and was then reviewed manually and interpreted by the same experienced gastroenterologist (P.H.T.).^24,25

The parameters obtained from pH monitoring included the total number of reflux episodes, the symptom index and symptom association probability. In addition, an acid exposure time (AET; the percentage of time that esophageal acid exposure was below a pH of 4.0) was also obtained. An AET < 4.0% was considered normal.¹⁵ An episode of acid reflux was defined as a drop in intraesophageal pH ≤ 4 for 5 seconds or more.

High-resolution Impedance Manometry

After an ≥ 8-hour fast, all patients underwent an HRIM study performed by an experienced gastroenterologist (P.H.T.).²⁶ We used a 4.2 mm diameter HRIM catheter, with 22 closely spaced water-perfused pressure sensors, and 12 impedance channels at 2-cm intervals (Medical Measurement Systems). After calibration, the catheter was transnasally introduced into the esophagus in an upright position. The upper esophageal sphincter and LES were identified based on the presence of upper and lower high-pressure zones, respectively. After 5 minutes of accommodation in supine position, basal upper esophageal sphincter and LES pressures were obtained, and participants were then asked to perform 10 liquid swallows of 5 mL saline at 30-second intervals. Data were analyzed using the package analysis software (Solar GI HRIM water-perfused system, Medical Measurement Systems) and assessed by P.H.T. LES relaxation and esophageal body contraction were evaluated, and esophageal motility disorders were diagnosed using the criteria stipulated by the Chicago Classification version 3.0.²⁷

Statistical Methods

Quantitative data are summarized as mean ± standard deviation (SD) for continuous data and were compared using independent sample t tests; categorical variables are summarized as percentages and were compared using chi-square tests. Pearson correlation coefficients were used to evaluate the relationship between RSI scores and scores of other symptom questionnaires. P < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 26 (SPSS, Chicago, IL, USA).

Participant Demographics and Clinical Characteristics

In the pilot exploratory study, we initially planned to enroll 100 participants to investigate the clinical features of patients with persistent laryngeal symptoms despite PPI treatment. However, non-urgent and non-time sensitive endoscopies and esophageal motility studies were put on pause during the severe Coronavirus Disease 19 outbreak in Taiwan in 2021 and case enrollment of the present study was thus terminated early. As a result, a total of 97 patients (56 women; mean age, 49 years; range, 18-85 years) were included in the final analysis. Of all the patients, 5.2% had been on lansoprazole 30 mg once daily, 6.2% on rabeprazole 20 mg once daily, 4.1% on pantoprazole 40 mg once daily, 1.0% on omeprazole 20 mg once daily, 25.8% on esomeprazole 40 mg once daily and 57.8% on dexlansoprazole 60 mg once daily. Fifty-eight patients (59.8%) had concurrent GERD symptoms (Laryngeal-GERD symptomatic). There was no significant difference between the subgroups of patients with laryngeal symptoms only and Laryngeal-GERD symptoms in terms of their age, sex, body mass index, tobacco use, and alcohol consumption (Table 1).

Table 1 . Demographics and Clinical Characteristics

Characteristics	Total (n = 97)	Laryngeal symptoms alone (n = 39)	Concurrent laryngeal and GERD symptoms (n = 58)	P-value
Age (yr)	49.30 ± 13.48	46.62 ± 14.14	51.10 ± 12.82	0.109
20-40	25 (25.8%)	15 (38.5%)	10 (17.2%)
40-60	52 (53.6%)	16 (41.3%)	36 (62.1%)
> 60	20 (20.6%)	8 (20.5%)	12 (20.7%)
Female	56 (57.7%)	21 (53.8%)	35 (60.3%)	0.538
BMI (kg/m2)	22.13 ± 3.39	22.04 ± 3.48	22.20 ± 3.36	0.824
Waist (cm)	77.97 ± 9.79	77.22 ± 10.76	78.48 ± 9.15	0.535
Smoking	12 (12.4%)	5 (12.8%)	7 (12.1%)	> 0.999
Drinking	13 (13.4%)	4 (10.3%)	9 (15.5%)	0.553
RSI total score	16.28 ± 6.86	15.65 ± 7.12	16.71 ± 6.72	0.456
RSI total score ≥ 13	64 (66.0)	24 (61.5)	40 (67.0)	0.515
Hoarseness or other voice problemsa	40 (41.2)	15 (38.5)	25 (43.1)	0.679
Clearing throata	56 (57.7)	21 (53.8)	35 (60.3)	0.538
Excess throat mucus or postnasal dripa	58 (59.8)	27 (69.2)	31 (53.4)	0.143
Difficulty swallowing food, liquid, or pillsa	17 (17.5)	6 (15.4)	11 (19.0)	0.788
Coughing after eating or after lying downa	18 (18.6)	8 (20.5)	10 (17.2)	0.791
Breathing difficulties or choking episodesa	10 (10.3)	5 (12.8)	5 (8.6)	0.517
Troublesome or annoying cougha	19 (16)	7 (17.9)	12 (20.7)	0.799
Sensations of something sticking in throat or lump in throata	73 (75.3)	26 (66.7)	47 (81.0)	0.150
Heartburn, chest pain, indigestion, or stomach acid coming upa	64 (66.0)	13 (33.3)	41 (70.7)	< 0.001

^aScore ≥ 3 (moderate symptom) on the item.

Data are presented as mean ± SD or n (%).

The most common laryngeal symptoms, based on the RSI, were “sensations of something sticking in throat or lump in throat” (75.3%), “heartburn, chest pain, indigestion, or stomach acid coming up” (66.0%), and “excess throat mucus” (59.8%) for subjects across the whole study. On the other hand, “excess throat mucus” (69.2%), “sensations of something sticking in throat or lump in throat” (66.7%), and “clearing throat” (53.8%) were most common in patients with laryngeal symptoms alone, while “lump in throat” (81.0%), “heartburn, chest pain, indigestion, or stomach acid coming up” (70.7%), and “clearing throat” (60.3%) were most common in Laryngeal-GERD symptomatic patients .

Gastrointestinal Symptoms and Objective Evaluation of Gastroesophageal Characteristics

Patients reported a GerdQ score of 9.16 ± 2.68, with “nausea” and “pain in the middle of the upper stomach area” being the 2 most prominent symptoms. The total score on PAGI-SYM was 1.13 ± 0.74, and the parameters with the 2 highest scores were “bloating” (1.69 ± 1.37) and “heartburn/regurgitation” (1.64 ± 1.10) (Table 2).

Table 2 . Associated Gastrointestinal Symptoms

Characteristics	Total (n = 97)	Laryngeal symptoms alone (n = 39)	Concurrent laryngeal and GERD symptoms (n = 58)	P-value
GerdQ
Burning feeling behind the breastbone (heartburn)	1.31 ± 1.31	0.44 ± 0.82	1.90 ± 1.25	< 0.001
Stomach contents moving up to the throat or mouth (regurgitation)	1.79 ± 1.18	1.10 ± 1.02	2.26 ± 1.05	< 0.001
Pain in the middle of the upper stomach area	2.10 ± 1.15	2.23 ± 1.11	2.02 ± 1.18	0.373
Nausea	2.44 ± 0.97	2.36 ± 1.01	2.50 ± 0.94	0.485
Trouble getting a good night’s sleep because of heartburn or regurgitation	1.01 ± 1.24	0.33 ± 0.66	1.47 ± 1.33	< 0.001
Need for over-the-counter medicine for heartburn or regurgitation, in addition to the medicine your doctor prescribed	0.51 ± 1.07	0.10 ± 0.50	0.78 ± 1.26	0.002
Total score	9.16 ± 2.68	6.56 ± 1.29	10.91 ± 1.81	< 0.001
PAGI-SYM
Heartburn/regurgitation	1.64 ± 1.10	1.46 ± 1.14	1.77 ± 1.07	0.176
Nausea/vomiting	0.65 ± 0.87	0.76 ± 0.96	0.58 ± 0.80	0.320
Postprandial fullness/early satiety	0.95 ± 0.93	1.04 ± 1.09	0.88 ± 0.81	0.405
Bloating	1.69 ± 1.37	1.51 ± 1.26	1.81 ± 1.44	0.297
Upper abdominal pain	1.34 ± 1.31	1.44 ± 1.49	1.28 ± 1.18	0.558
Lower abdominal pain	0.48 ± 0.93	0.32 ± 0.66	0.59 ± 1.06	0.169
Total score	1.13 ± 0.74	1.11 ± 0.80	1.15 ± 0.70	0.753

There are statistically significant differences (P < 0.05) between laryngeal symptoms and Laryngeal-gastroesophageal reflux disease (GERD) symptoms groups.

GerdQ, gastroesophageal reflux disease questionnaire; PAGI-SYM, patient assessment of upper gastrointestinal symptom severity index.

Data are presented as mean ± SD.

On endoscopy, 23 (23.7%) patients had erosive esophagitis (EE) and 1 (1.0%) patient had hiatal hernia. On pH-impedance monitoring, abnormal pH testing was present in 16 (16.5%) patients. 19 (19.6%) patients had an abnormal AET in the upright position. On impedance monitoring, the average number of reflux episodes was 38.13 ± 26.80. Of these, 20.55 ± 18.38 were acidic, 15.91 ± 14.05 were weakly acidic, and 1.86 ± 4.18 were non-acidic. 14.78 ± 16.62% and 25.50 ± 20.18% of all reflux episodes reached the proximal extent of 17 cm and 15 cm above the LES, respectively. On HRIM, the LES 4-second integrated relaxation pressures (IRP-4s) was 7.81 ± 5.61 mmHg. Twenty-three (23.7%) patients had ineffective esophageal motility and 12 (12.4%) had hiatal hernia as defined by manometry (Table 3).

Table 3 . Objective Evaluation of Gastroesophageal Characteristics

Characteristics	Total (n = 97)	Laryngeal symptoms alone (n = 39)	Concurrent laryngeal and GERD symptoms (n = 58)	P-value
Endoscopic findings
Erosive esophagitis	23 (23.7)	9 (23.1)	14 (24.1)	0.701
LA Grade A	20 (20.6)	8 (20.5)	12 (20.7)
LA Grade B	3 (3.1)	1 (2.6)	2 (3.4)
Hiatal hernia	1 (1.0)	0 (0.0)	1 (1.7)
24-hour pH-impedance monitoring
Esophageal acid exposure (%)	3.04 ± 9.00	3.76 ± 13.48	2.56 ± 3.83	0.522
Abnormal acid exposure > 4%	16 (16.5)	5 (12.8)	11 (19.0)	0.579
Abnormal acid exposure > 6%	9 (9.3)	4 (10.3%)	5 (8.6%)	> 0.999
Upright reflux time (%)	4.02 ± 8.95	4.21 ± 12.13	3.89 ± 6.06	0.862
Abnormal acid exposurea	19(19.6)	6 (15.4)	13 (22.4)	0.445
Supine reflux time (%)	1.69 ± 9.98	3.07 ± 15.51	0.75 ± 2.23	0.264
Abnormal acid exposureb	15 (15.5)	6 (15.4)	9 (15.5)	> 0.999
DeMeester score	10.10 ± 29.92	12.82 ± 45.17	8.27 ± 11.78	0.466
Number of symptoms	5.77 ± 7.97	4.28 ± 6.13	6.78 ± 8.91	0.131
Positive SI	22 (22.7)	8 (20.5)	14 (24.1)	0.806
Positive SAP	32 (33.0)	11 (28.2)	21 (36.2)	0.510
Number of reflux episodes on impedance	38.13 ± 26.80	35.74 ± 29.15	39.74 ± 25.24	0.474
Acidic reflux	20.55 ± 18.38	19.18 ± 20.34	21.47 ± 17.07	0.551
Weakly acidic reflux	15.91 ± 14.05	13.59 ± 12.71	17.47 ± 14.79	0.184
Non-acidic reflux	1.86 ± 4.18	2.64 ± 5.66	1.33 ± 2.70	0.129
Proximal extent (%)
Extent 17 cm	14.78 ± 16.62	13.39 ± 18.10	15.69 ± 15.68	0.511
Extent 15 cm	25.50 ± 20.18	22.74 ± 20.74	27.31 ± 19.77	0.280
Extent 9 cm	62.86 ± 22.70	61.11 ± 25.08	64.02 ± 21.14	0.542
Extent 7 cm	74.34 ± 22.63	74.53 ± 25.33	74.22 ± 20.89	0.949
Extent 5 cm	94.90 ± 21.90	94.74 ± 22.63	95.00 ± 21.62	0.954
Extent 3 cm	95.83 ± 20.09	94.74 ± 22.63	96.55 ± 18.41	0.667
High-resolution impedance manometry
LES resting pressure (mmHg)	21.25 ± 11.37	21.05 ± 10.16	21.39 ± 12.19	0.889
LES IRP-4s (mmHg)	7.81 ± 5.61	8.28 ± 5.30	7.49 ± 5.83	0.502
Intact peristalsis (%)	66.29 ± 37.45	69.49 ± 34.86	64.14 ± 39.25	0.493
Weak peristalsis (%)	15.93 ± 23.50	16.03 ± 22.48	15.86 ± 24.35	0.973
Failed peristalsis (%)	17.78 ± 31.25	14.49 ± 29.22	20.00 ± 32.61	0.397
DCI, mmHg∙s∙cm	1015.12 ± 853.59	1055.50 ± 778.84	989.61 ± 903.42	0.719
Break size (cm)	3.93 ± 5.67	3.44 ± 3.78	4.24 ± 6.60	0.513
Ineffective esophageal motility	23 (23.7)	9 (23.1)	14 (24.1)	0.381
Hiatal hernia	12 (12.4)	4 (10.3)	8 (13.8)	0.757

^aAbnormal esophageal acid exposure in the upright positions was defined as a pH < 4 for more than 6% of the time.

^bAbnormal esophageal acid exposure in the supine positions was defined as a pH < 4 for more than 1% of the time.

There are statistically significant differences (P < 0.05) between laryngeal symptomatic and Laryngeal-gastroesophageal reflux disease (GERD) asymptomatic groups.

LA, Los Angeles classification of the severity of reflux esophagitis; SI, symptom index; SAP, symptom association probability; IRP-4s, 4-second integrated relaxation pressure; DCI, distal contractile integral.

Data are presented as mean ± SD or n (%).

Psychological Characteristics

The patient group reported a significantly higher BSRS-5 score (6.52 ± 4.55 vs 1.33 ± 2.00, P < 0.001) and PSQI (10.41 ± 4.57 vs 4.60 ± 3.18, P < 0.001) than healthy volunteers (Table 4). The prevalence of psychiatric morbidity (52.6% vs 2.1%, P < 0.001) and sleep disturbance (82.5% vs 37.5%, P < 0.001) was also significantly higher. For individual items on questionnaires, the patients reported significantly higher scores on all items except for “feeling inferior to others” on BSRS-5 and “sleep duration” on PSQI. Furthermore, in the patient group, 42 out of 97 (43.3%) reported use of sleep medication (Question 6: During the past month, how often have you taken medicine, prescribed or over-the-counter, to help you sleep? In PSQI). Among them, 9.5% used it less than once a week, 16.7% reported once to twice a week, and 73.8% reported 3 or more times a week. While only 2 (4.2%) in healthy volunteer group reported medication use, with 1 less than once a week and another 3 or more times a week.

Table 4 . Comparison of Psychological Characteristics and Sleep Quality Between Patients With Laryngeal Symptoms and Healthy Volunteers

Characteristics	Patients (n = 97)	Healthy volunteer (n = 48)	P-value
Age (yr)	49.30 ± 13.48	45.56 ± 11.14	0.100
Female	56 (57.7)	22 (45.8)	0.216
BSRS-5
Trouble falling asleep	1.92 ± 1.32	0.25 ± 0.53	< 0.001
Feeling tense or keyed up	1.52 ± 1.29	0.35 ± 0.64	< 0.001
Feeling easily annoyed or irritated	1.40 ± 1.31	0.33 ± 0.60	< 0.001
Feeling depressed	1.25 ± 1.20	0.17 ± 0.43	< 0.001
Feeling inferior to others	0.43 ± 0.91	0.23 ± 0.47	0.148
Total scores	6.52 ± 4.55	1.33 ± 2.00	< 0.001
Psychiatric comorbidity	51 (52.6)	1 (2.1)	< 0.001
PSQI
Sleep duration	1.14 ± 1.57	0.83 ± 0.75	0.197
Sleep latency	1.38 ± 1.08	0.40 ± 0.61	< 0.001
Habitual sleep efficiency	1.34 ± 1.06	0.46 ± 0.68	< 0.001
Sleep disturbances	2.45 ± 0.97	1.46 ± 1.17	< 0.001
Subjective sleep quality	1.87 ± 0.73	0.83 ± 0.60	< 0.001
Use of sleeping medication	1.14 ± 1.38	0.08 ± 0.45	< 0.001
Daytime dysfunction	1.08 ± 1.21	0.54 ± 0.87	< 0.001
Total scores	10.41 ± 4.57	4.60 ± 3.18	< 0.001
Sleep disturbance	80 (82.5)	18 (37.5)	< 0.001

Psychiatric comorbidity indicates a 5-item brief symptom rating scale (BSRS-5) total score ≥ 6.

Sleep disturbance indicates a Pittsburgh Sleep Quality Index (PSQI) score > 5.

Data are presented as mean ± SD or n (%).

Correlation Between Psychiatric Morbidity and Sleep Disturbance With Laryngeal Symptoms and Reflux Symptom Burden

Correlation analysis between RSI and BSRS-5 scores, and between RSI and PSQI scores, showed significant correlation (Pearson correlation coefficient r = 0.26, P = 0.010, and r = 0.29, P = 0.004, respectively) (Figure 1). Correlation analysis showed significant correlation between GerdQ and PSQI scores (Pearson correlation coefficient r = 0.34, P = 0.001) but not between GerdQ and BSRS-5 scores (Pearson correlation coefficient r = 0.12, P = 0.260).

Figure 1. Correlation between psychiatric morbidity and sleep disturbance with laryngeal symptoms and reflux symptoms burden. The score of Reflux Symptom Index (RSI) was significantly correlated with the scores of 5-item brief symptom rating scale (BSRS-5) (A) and Pittsburgh sleep quality index (PSQI) (B) (Pearson correlation coefficient r = 0.26, P = 0.010 and r = 0.29, P = 0.004, respectively). The score of gastroesophageal reflux disease questionnaire (GerdQ) was significantly correlated with the scores of PSQI (C) (Pearson correlation coefficient r = 0.34, P = 0.001) but not with BSRS-5 (D) (Pearson correlation coefficient r = 0.12, P = 0.260).

Comparison Between Patients With Proton Pump Inhibitor-refractory Laryngeal Symptoms Alone and Patients With Concurrent Gastroesophageal Reflux Disease Symptoms

Gastrointestinal symptoms

The total score on the GerdQ was 10.91 ± 1.81 in the Laryngeal-GERD symptomatic patient group, which was significantly higher for the laryngeal symptoms alone group (6.56 ± 1.29, P = 0.000). The Laryngeal-GERD symptomatic patients reported higher symptom burden on items including “heartburn”, “regurgitation,” “trouble sleeping”, and “need for over-the-counter medicine”. There were no significant differences in either total score or on any of the subscales on the PAGI-SYM.

Objective evaluation of gastroesophageal characteristics

Endoscopy showed no significant differences in the prevalence of EE and hiatal hernia (P = 0.701) between the 2 groups (Table 3). On pH-impedance monitoring, none of the parameters including AET (P = 0.522), acid reflux (P = 0.551), and non-acidic reflux (P = 0.129) differed between the 2 groups. The prevalence of abnormal AET overall and in upright position were similar. On HRIM, there were also no significant differences in key parameters including LES IRP-4s, failed peristalsis, and ineffective esophageal motility.

Psychological characteristics and sleep quality

Overall, the scores of BSRS-5 and PSQI were high in all patients (Table 4). The total BSRS-5 score and the prevalence of psychiatric comorbidity were similar between the laryngeal symptoms alone and the Laryngeal-GERD symptomatic patient groups (Table 5). However, Laryngeal-GERD symptomatic patients reported a significantly higher score on “trouble falling asleep” on the BSRS-5 (2.17 ± 1.35 vs 1.54 ± 1.19, P = 0.020). Moreover, Laryngeal-GERD symptomatic patients reported an even higher total PSQI score and higher prevalence of sleep disturbance (11.50 ± 4.63 vs 8.79 ± 4.01, P < 0.001 and 89.7% vs 71.8%, P < 0.001, respectively).

Table 5 . Comparison of Psychological Characteristics and Sleep Quality Between Patients With Laryngeal Symptoms Alone and Those With Concurrent Gastroesophageal Reflux Disease Symptoms

Characteristics	Laryngeal symptoms alone (n = 39)	Concurrent laryngeal and GERD symptoms (n = 58)	P-value
Age (yr)	46.62 ± 14.14	51.10 ± 12.82	0.109
Female	21 (53.8)	35 (60.3)	0.538
BSRS-5
Trouble falling asleep	1.54 ± 1.19	2.17 ± 1.35	0.020
Feeling tense or keyed up	1.56 ± 1.12	1.48 ± 1.41	0.763
Feeling easily annoyed or irritated	1.38 ± 1.25	1.41 ± 1.36	0.915
Feeling depressed	1.18 ± 1.07	1.29 ± 1.28	0.650
Feeling inferior to others	0.38 ± 0.78	0.47 ± 1.00	0.671
Total scores	6.05 ± 4.08	6.83 ± 4.85	0.413
Psychiatric comorbidity	19 (48.7)	32 (55.2)	0.543
PSQI
Sleep duration	1.03 ± 0.78	1.22 ± 1.94	0.545
Sleep latency	1.10 ± 1.05	1.57 ± 1.08	0.037
Habitual sleep efficiency	1.05 ± 0.94	1.53 ± 1.10	0.027
Sleep disturbances	2.13 ± 1.17	2.67 ± 0.74	0.006
Subjective sleep quality	1.72 ± 0.69	1.97 ± 0.75	0.102
Use of sleeping medication	0.87 ± 1.34	1.33 ± 1.39	0.112
Daytime dysfunction	0.90 ± 1.10	1.21 ± 1.27	0.217
Total scores	8.79 ± 4.01	11.50 ± 4.63	0.004
Sleep disturbance	28 (71.8)	52 (89.7)	0.030

Psychiatric comorbidity indicates a 5-item brief symptom rating scale (BSRS-5) total score ≥ 6.

Sleep disturbance indicates a Pittsburgh sleep quality index (PSQI) score > 5.

GERD, gastroesophageal reflux disease.

Data are presented as mean ± SD or n (%).

In the present study, patients with PPI-refractory laryngeal symptoms had markedly higher prevalence of psychological distress and sleep disturbance when compared to the healthy volunteers. The prevalence of EE and motility characteristics on HRIM were similar to the normal population.^26,28,29 Compared to patients with laryngeal symptoms alone, those with concurrent GERD symptoms had even more prominent sleep disturbances but presented with similar reflux profiles and esophageal motility.

Compared to normal pH-impedance monitoring data in the Chinese population, our laryngeal symptomatic patients had similar episodes of acid reflux, but their total and upright AETs were slightly higher.^30,31 Compared to healthy volunteers, however, patients in this study were observed to have a significantly higher prevalence of psychiatric morbidity and sleep disturbance. Bradley et al³² found that chronic reflux patients with chronic anxiety were more likely to perceive low-intensity esophageal stimuli as painful reflux symptoms. A recent study further compared individuals with non-erosive reflux disease, reflux hypersensitivity, and functional heartburn to healthy controls, and found that all 3 of these groups had increased sensation to visceral stimuli based on the visceral sensitivity index.³³ Similarly, many studies had reported the association between laryngeal symptoms and higher somatization symptoms,^34,35 and between laryngeal symptoms and psychiatric morbidity.^36-38 Furthermore, in cases of refractory laryngeal symptoms, 52.6% of our patients in the present study showed psychological symptoms on BSRS-5, while the prevalence of psychiatric morbidity was 27.2 % in a hospital-based general health screening.²² It seems that patients with PPI-refractory laryngeal symptoms presented with a high burden of psychological stress. Vertigan et al³⁹ found significant sensory impairments in patients with chronic refractory cough and globus pharyngeus using quantitative sensory testing. Visceral hypersensitivity and psychological factors may also play an important role in the development of laryngeal symptoms and symptoms that appear refractory to first-line treatment.^40,41

Extraesophageal reflux disease is usually multifactorial with GERD as one of the several potential aggravating cofactors.² In the present study, about 60% of the subjects had GERD symptoms, as defined with a GerdQ score ≥ 8. Recently, a Taiwanese multi-center observational study also showed that among 106 patients with predominant laryngopharyngeal reflux symptoms, 66 (62.2%) had concomitant typical reflux symptoms, who had similar PPI-responsiveness and similar distal esophageal acid exposure time when compared to patients with isolated laryngeal symptoms.⁴² In present study, the clinical features of patients with PPI-refractory laryngeal symptoms were further investigated and classified according to the presence or absence of concurrent GERD symptoms. We showed that the reflux profile and esophageal motility were similar. Although all patients in the present study had a higher prevalence of psychiatric morbidity and sleep disturbance when compared to healthy volunteers, patients with concurrent GERD symptoms reported even higher scores with regards to sleep disturbance on the BSRS-5 and PSQI scores than patients with laryngeal symptoms alone. Similarly, higher psychological comorbidities had also been identified in patients with overlapping laryngeal and GERD symptoms.⁴³ Poor sleep has been reported to be related to increased esophageal acid exposure and visceral hypersensitivity.^44,45 It is estimated that up to 50% of patients with GERD remained symptomatic on standard PPI therapy.^46,47 Poor medication compliance and adherence, differences in PPI metabolism, residual reflux (non-acidic, bile, or acidic), concomitant functional bowel disorders, psychological comorbidity, and delayed gastric emptying have been proposed as mechanisms of treatment failure.⁴⁸ A prospective cohort study demonstrated that reflux characteristics, measured using impedance pH, were not significantly different between patients who responded to standard PPI therapy when compared to those who failed to do so.¹³ Their findings also supported the important role of esophageal hypersensitivity in this patient population. The authors concluded that patients who do not respond to PPI therapy may benefit from adding a neuromodulator and potentially from psychological intervention. Hyperalgesia, hypervigilance, and heightened anxiety are thought to be the most plausible candidates for the dominant determinants of PPI-refractory symptoms.⁴⁹ Psychosocial factors influence physiological functioning of the GI tract via the brain–gut axis, are modulators of the patient’s pain experience and symptom behavior, and ultimately affect treatment selection and clinical outcome.⁵⁰

Assessment of patients with laryngeal symptoms referred for reflux evaluation has remained challenging and treatment response has been less than satisfactory.^6,51 Systemic assessment on clinical and physiologic characteristics of these patients has been demonstrated to be efficient in daily practice.⁵² Patients with proven pathological GERD would be the optimal candidates for an escalated anti-reflux therapy. On the other hand, patients with nonacid reflux events could be spared prolonged acid suppression and could be offered treatments addressed to associated mechanism such as alginates or antireflux surgery, which had been demonstrated to be more cost-saving.⁵³ Finally, for those presenting laryngeal symptoms with reflux hypersensitivity, neuromodulation may be particularly beneficial.⁵⁴

The strengths of the current study include its prospective study design and its inclusion of objective multidisciplinary evaluation modalities, as well as patient-reported questionnaires. There are however some limitations. Firstly, we used only 1 pH sensor at the lower esophagus and could not confirm the presence of a proximal esophageal acid reflux event reaching the hypopharynx. To deal with this limitation, we reported the proximal extent of reflux episodes on impedance monitoring, and focused on patients whose symptoms persisted after a PPI trial, with the MII-pH monitoring being performed during off PPI period. Incorporating hypopharyngeal-esophageal MII-pH may provide additional understanding of the nature of laryngeal symptoms in laryngopharyngeal reflux⁵⁵ and should be considered in future studies. On the other hand, pathological reflux to the hypopharynx could not solely explain the pathophysiology of laryngeal symptoms, as the neurally mediated vagal reflex has also been proposed as part of the pathogenesis of atypical GERD symptoms.⁵⁶ Secondly, there were no differences in the endoscopic findings and pH monitoring results between patients with laryngeal symptoms alone and concurrent GERD symptoms, except for significantly higher sleep disturbance in Laryngeal-GERD symptomatic patients. Although there were numerically higher percentage of patients having AET at upright, higher number of symptoms, positive symptom association probability, and higher number of reflux episode on impedance (especially acidic and weakly acidic reflux) in patients with concurrent GERD symptoms, none of these parameters achieves statistical significance. Future studies with a larger sample size could be helpful to clarify or confirm the clinical significance of these findings. Thirdly, in the employed study protocol, the prevalence and severity of psychological distress or sleep disturbance were investigated with validated questionnaires. However, the psychiatric medication or insomnia medication were not specifically recorded. In addition, we did not routinely refer our patients for psychiatric consultation. Although our study has demonstrated close association between psychological stress and these bothersome laryngeal symptoms, the cause-effect relationship remained undetermined. Therefore, considering the high prevalence of psychiatric comorbidities shown in patients with refractory laryngeal symptoms, future studies incorporating psychiatric specialists in the multidisciplinary team would help clarify the complex relationship between psychological distress and PPI-refractory laryngeal symptoms and tailor further treatment plans.

In conclusion, a multidisciplinary evaluation should be performed in patients whose troublesome laryngeal symptoms do not respond to a PPI trial.⁵ We investigated patients’ anatomical, physiological, and psychological characteristics using validated questionnaires, EGD, MII-pH monitoring, and HRIM. The present study confirmed the hypothesis that PPI-refractory laryngeal symptoms are more associated with psychological comorbidities and sleep disturbances rather than reflux-related. In addition to providing careful physical examination, physicians should also address these psychosocial factors. Our study suggests that meticulous recognition of psychological comorbidities appears to be clinically meaningful in the management of patients whose laryngeal symptoms persist after PPI therapy.

The authors would also like to thank Unit-Edit (www.uni-edit.net) for editing and proofreading this manuscript.

This study was supported by research grants from the National Taiwan University Hospital (NTUH 110-005025) and the Ministry of Science and Technology (109-2628-B-002-036 and 110-2628-B-002-048). The funding agencies had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

None.

Ping-Huei Tseng: guarantor of the article who was also in charge of study design, performing the study, data acquisition, interpretation, and manuscript revision; Wen-Hsuan Tseng: drafted manuscript, acquisition of data, and interpretation and analysis of data; Wei-Chung Hsu, Jia-Feng Wu, and Hui-Chuan Lee: study design, interpretation, and acquisition of data; and Tsung-Lin Yang, Tzu-Yu Hsiao, Hsiu-Po Wang, and Ming-Shiang Wu: study design and interpretation. All authors have approved the final version of the manuscript.