Esophageal Intraluminal pH Recording in the Assessment of Gastroesophageal Reflux

and Its Consequences

MICHAEL ATKINSON, MD, and A. VAN GELDER, MD

lntraluminal pH in the lower esophagus has been recorded during a 3-hr period following a light meal and a consecutive 12-hr nocturnalperiod in 20 patients with typical symptoms and radiological evidence of gastroesophageal reflux and in 10 patients without such signs of reflux. Evidence of acid reflux was obtained in 3 of the patients without reflux during the postcibal period but in only one during the 12-hr nocturnal period. In contrast all except one of the 20 patients who had evidence o f reflux showed spells o f high acidity both in the postcibal and nocturnal periods. There was no clear correlation between the frequency o f pain from gastroesophageal reflux over the preceding two weeks and the duration of high acidity in the nocturnal period. Those patients with endoscopic evidence of severe esopha- g'itis showed a significantly longer duration of high esophageal acidity in the nocturnal period. We conclude that nocturnal exposure of the esophageal mucosa to acid is a major fcwtor in the causation of reflux esophagitis.

Heartburn is experienced occasionally by most of the population, yet its significance in relation to gas- troesophageal reflux and esophagitis remains far from clear. Most practicing clinicians have encoun- tered, on the one hand, patients with disabling heart- burn and little endoscopic evidence of esophagitis and, on the other, those with esophagitis severe enough to cause hematemesis or malena who give no history of preceding chest pain.

Quantitation of the exposure of the esophageal mucosa to acid over a period of time is provided by intraesoplhageal pH recording. However, the evi- dence this method has provided to date is not easy to evaluate. Pattrick (1) could find little difference in esophageal pH levels between patients with and

From Worcester Royal Infirmary and the General Hospital, Nottingham, England.

Dr. Van Gelder's present address is Juliana Ziekenhuis, Apeldoorn, Netherlands.

Address for reprint requests: Dr. M. Atkinson, Consultant Physician, General Hospital, Nottingham NG1 6HA, England.

those without reflux during the day, and even at night a prominent difference was only obtained when the patient was lying on his right side. Yet Spencer (2) found a much closer correlation be- tween reflux symptoms and changes in esophageal pH during an overnight recording, irrespective of which lateral position was adopted.

The purpose of the present study was to examine further the relationship between pain, esophageal acidity, and esophagitis in groups of patients with and without symptoms of gastroesophageal reflux.

MATERIALS AND METHODS

Patients Studied

Twenty patients (11 males and 9 females) with typical symptoms and radiological evidence of gastroesophageal reflux, 17 of whom had a demonstrable hiatal hernia, were compared with 10 patients (7 males and 3 females) with- out clinical or radiological signs of esophagogastric dis- ease, most of whom were initially suspected to be suffer-

Digestive Diseases, Vol. 22, No. 4 (April 1977) 365

ATKINSON AND VAN GELDER

i i i . . . . . . . . . . . ' ! O 1 :

Fig 1. Recordings of overnight esophageal pH from 4 representative patients. The top tracing, from a patient without symptoms of reflux or evidence of alimentary disease, shows no acid reflux. The second tracing, from a patient with symptoms of reflux and moderate esophagitis, shows spikes of acidity in the postcibal period with two periods or more of prolonged acidity before midnight. Further reflux occurred after waking at 6:00 AM. The third tracing, from a patient with symptoms of reflux and mild esophagitis, shows reflux in both the postcibal and nocturnal periods. The bottom tracing, from a patient with severe esophagitis, shows almost continuous acid reflux throughout the period of recording. This patient had chest pain at 8:00 PM, between 12 midnight and 12 AM, and again at 6:00 AM.

ing from duodenal ulcer although in 7 no endoscopic ab- normali ty of the stomach or duodenum was present . Informed consent was obtained from all patients studied.

Methods Esophageal pH. This was measured using a small glass

pH electrode attached by a coaxial cable to a pH meter

and pen-writing recorder. The reference electrode was connected by a bridge tube containing a saturated solu- tion of potassium chloride running alongside the coaxial cable and opening adjacent to the glass bulb. It was cali- brated before and after use in buffers of pH 4 and pH 9. In the late afternoon the electrode was passed through the nose and its position adjusted under x-ray control until

rain. 90.

60"

~0"

- � 9 1 4 9 1 4 9

NO REFLUX

o�9149

�9 . � 9

� 9 1 4 9

REFLUX

Fig 2. Duration of high esophageal acidity (pH < 4) during the 3- hr postcibal period (6:00-9:00 PM) in patients with and without symptoms of reflux.

rain. 180" 260

120" ee

%�9 60'

o � 9

- - � 9 1 4 9 1 4 9 1 4 9 ~ 0 o o 0

NO REFLUX REFLUX

Fig 3. Duration of high esophageal acidity (pH < 4) during the nocturnal period (9:00 PM--9:00 AM) in patients with and without symptoms of reflux.

366 Digestive Diseases, Vol. 22, No. 4 (April 1977)

ESOPHAGEAL INTRALUMINAL pH RECORDING

the glass bulb lay 5 cm above the position of the cardia estimated from x-ray films, and pH recording was com- menced. At 6:00 eN the patient took a light meal of soup and sandwiches and remained sitting in bed until 9:00 eN when he was placed in the recumbent position with one pillow. The recording was continued for another 12 hrs until 9:00 AM, and the patient was asked to signal if he had pain. He was allowed to lie on his back or on either side and to change from one position to another at will. The position of the electrode was checked radiologically at the end of the recording period.

In 12 of the patients with symptoms of reflux, a 2-hr pH recording was made with simultaneous intraluminal pres- sure measurement from the level of the pH electrode and from a point 5 cm higher in the esophagus. Pressure was recorded through water-filled polyethylene tubes (1.5 mm bore) perfused with water at a rate of 30 ml/hour.

Pain Frequency. This was assessed by asking the patient to keep a record of the number of day and night periods in which pain was felt over the two consecutive weeks preceding esophageal acidity recording. Thus the maximum score possible was 28.

Degree of Esophagitis. The presence was assessed by fiberoptic endoscopy as mild (hyperemia only), moderate (hyperemia and increased friability to touch), and severe (spontaneous bleeding and ulceration). In none was a stricture present. Adequate mucosal biopsies were ob- tained from many but not all patients, and these were not therefore used in the study.

RESULTS

Esophageal pH. For purposes of analysis the trac- ings were divided into the 3-hr postcibal period (6:00-9:00 PM) and the 12-hr nocturnal period (9:00 eM-9:00 AM). Fou r r ep resen ta t ive t racings are shown in Figure 1.

Patients without Reflux Symptoms Postcibal Period. Of the 10 patients in this group,

7 showed no evidence of reflux, and the esophageal pH remained between 5 and 7 throughout the period (Figure 2). Two showed a series of spikes in the pH tracing rising transiently to between pH 4 and 2. The 10th patient had spikes followed by longer peri- ods of high acidity with a pH of less than 4 for 32 of the 180 miin following the meal.

Noeturnal Period. During this period esophageal acidity varied little, and in only 2 patients did it ever fall below pH 4 in very transient spikes, each lasting for not more than 30 sec, interspaced at long inter- vals of time (Figure 3).

Patients with Symptomatic Reflux Postcibal Period. Esophageal pH fell below 4 in all

except one of the 20 patients in this group during the postcibal period. In 19 patients transient spikes of

high acidity, which often followed each other in rap- id succession, were noted. In 12 patients acidity re- mained high for periods of several minutes at a time. In 4 of the 20 patients a pH of less than 4 pre- vailed for 60 rain or more in the 180-rain postcibal period. The mean duration of acidity of pH 4 or less was 29 rain for the group as a whole.

Noeturnal Period. Spikes of high acidity were seen in the pH records of each of the 20 patients with symptoms of reflux. In 6, these were transient and no periods of sustained high acidity were en- countered. In the remaining 14 patients acidity re- mained high for periods of 30 sec to 30 min at a time, and in one a pH of less than 4 was recorded for more than 4 of the 12 hr. Taking the group as a whole, the mean duration of an acidity of pH 4 or less was 68 min.

Relationship Between Esophageal Pain and Acidity Pain during the period of pH recording was expe-

rienced in 8 of the 20 patients with symptoms of re- flux and in none of the asymptomatic group. Each of the 8 had pain during the nocturnal period. In each it coincided with a period of high acidity (pH less than 4), and isolated spikes of acidity were not usually painful. In the individual patient pain did not seem to occur during every period in which acid- ity was raised, and often patients slept during such periods. However , it did appear that the greater the duration of the period of high acidity, the more like- ly was it to be accompanied by pain. Of the 8 patients, 3 also had pain in the initial postcibal peri- od and again this coincided with a period of sus- tained acidity. Equally high levels of acidity were re- corded from many of the other 12 patients who had no pain during the test. Fur thermore the pH thresh- old for pain was not closely dependent upon the de- gree of esophagitis, and during the test 2 patients with severe esophagitis tolerated pH levels of 2-3 for periods of several minutes without complaining of pain.

The records of symptoms kept by the patient for the 2 weeks before the study showed no close corre- lation between the pain score and the mean duration of high esophageal acidity (pH less than 4), although the group with the highest pain scores did have the longest mean duration of high acidity (Table 1).

The pattern of pain from reflux does vary from patient to patient, and while in most of those stud- ied nocturnal pain predominated, in a minority pain during the day was the principal complaint. Of the 20 patients with reflux symptoms, 3 had pain fie-

Digestive Dis,eases, Vol. 22, No. 4 (April 1977) 367

ATKINSON AND VAN GELDER

TABLE 1.

Pain score No. of patients Postcibal Nocturnal

TABLE 2.

Mean duration of pH < 4 Mean duration of pH < 4

(rain) Degree of (min)

esophagitis No. of patients Postcibal Nocturnal

5 - 10 4 21 22 11 - 15 8 20 72 16 - 20 5 35 60 20 - 28 3 42 100

Nil 3 26 26 Mild 8 18 35 Moderate 4 22 54 Severe 5 73* 144t

*P < 0.01. t0.01 < P < 0.05.

quently during the day and seldom at night. How- ever, no difference in their patterns of esophageal acidity could be detected, and all showed acid re- flux during the night as well as during the postcibal period. Taking the group as a whole the duration of high acidity in the postcibal period correlated direct- ly with that in the nocturnal period (Figure 4), and we could find no evidence to support the suggestion that certain patients get reflux predominantly during the day and others predominantly at night.

Relationship Between Esophageal Acidity and Esophagitis

In 3 patients the esophageal mucosa appeared normal at endoscopy, and mucosal biopsies ob- tained in 2 did not show the changes of esophagitis. Endoscopy indicated esophagitis of mild degree in 8, of moderate degree in 4, and of severe degree in 5. The mean duration of high acidity (pH less than 4) in relationship to the degree of esophagitis is shown in Table 2, which indicates that acidity was significantly greater in the 5 patients with severe

rain

120"

2~

6 0 �84

O0 �9

Oo �9 �9

0 .~0 6"0 rain POST-CIBAL

F i g 4. The relationship between the duration of high acidity (pH < 4) in the postcibal and nocturnal periods in 20 patients with symptoms of reflux. A direct relationship is shown.

esophagitis in both the postcibal and nocturnal peri- ods than in the other patients.

Esophageal Motor Changes in Relationship to Esophageal Acidity

Simultaneous recording of esophageal pH and motility for periods of 2 hr during recumbency in 12 of the patients with symptoms of reflux revealed no evidence of reflux in 4, in one of whom in- coordinated motor activity was present for brief pe- riods. In the remaining 5 patients prolonged periods of high esophageal acidity were recorded, and in each this was associated with much incoordinated activity which was virtually c~ntinuous in 2. In 3 of these patients the assumption of the erect posture resulted in a fall in esophageal acidity and a diminu- tion of incoordinated motor activity although this only ceased completely in one.

DISCUSSION

Measurement of the intraluminal pH of the lower esophagus provides an accurate means of detecting acid reflux and quantitating the duration of esopha- geal mucosal exposure to acid. Our finding that oc- casional patients, who had neither clinical nor radio- logical evidence of gastroesophageal reflux, show transient spells of high esophageal acidity after meals but that none show this when recumbent at night is in general agreement with the conclusions reached by Spencer (2) and Pattrick (1). Kantro- witz et al (3), using a similar method of esophageal pH recording, found evidence of minimal reflux with changes of posture, deep breathing, or during the valsalva maneuvre in 7 of 15 patients and volun- teers without clinical or radiological evidence of re- flux. It would appear, therefore, that while the nor- mal cardia allows occasional transient reflux during

368 Digestive Diseases, Vol. 22, No. 4 (April 1977)

ESOPHAGEAL INTRALUMINAL pH RECORDING

the day, it effectively prevents this at night. In con- trast, patients with symptoms of reflux nearly all showed periods of esophageal acidity during the night, and it seems likely that it is then that the ma- jor part of the damage to the esophageal mucosa oc- curs. The close correlation between the presence of clinical symptoms and that of reflux detected by means of nocturnal esophageal pH recording in- dicates that this method is a very reliable means of detecting reflux.

The position of the electrode in relation to the car- dia is of critical importance: If it is too low there is a risk that it may become displaced into the stomach, and if it is too high it may fail to detect acid in the lower-most part of the esophagus. The endoscopic and histologic signs of es0phagitis are occasionally confined to the lower-most few centimeters of the esophagus, and furthermore, withdrawing a pH electrode from the stomach into the esophagus re- veals that refluxed acid is often confined to its termi- nal few centimeters (Tuttle and Grossman (4). It is therefore obviously important to place the pH elec- trode carefully, using radiological control, and dur- ing and after the period of recording to make further checks to determine whether displacement has oc- curred.

The relationship of symptoms of reflux to the du- ration of ihigh esophageal acidity is complex. Wood- ward (5) could show no clear relationship between the severity of symptoms and the duration of high esophageal acidity in overnight recordings, and our findings support his conclusions. In the individual, however, we found that pain only occurred at times when esophageal acidity was high, although the pH threshold varied a good deal from patient to patient and many withstood fairly prolonged periods of high acidity without experiencing pain. Transient spikes of acidity were not associated with pain. That the mechanism causing pain is rather more complex than a simple summation of the height and ~luration of intraluminal acidity is suggested by the fact that not all periods of high acidity are accompa- nied by pain. Other factors such as reflux of bile and pancreatic juice may be of importance in this re- spect. Since the abolition of all motor activity by an anticholinergic drug does not generally diminish the severity of pain induced by esophageal acid per- fusion (6), it seems unlikely that motor factors are important in determining the pH threshold for pain.

Incoordinated motor activity commonly accom- panies periods of high esophageal acidity and may increase their duration by impairing esophageal

clearing (7). Stanciu and Bennett (8) found that af- ter instilling 0.1 N hydrochloric acid into the lower esophagus substant ia l ly more swal lows were needed to raise esophageal pH to 5.0 in subjects with symptomatic gastroesophageal reflux than in those without. This impairment of acid clearance correlated well with the mean duration of spon- taneous episodes of acid reflux during a 15-hr pH re- cording from the lower esophagus. It seems prob- able that such incoordination of motor activity lead- ing to impairment of acid c learance by the esophagus is of greatest importance at night when, in recumbency, gravity is no longer operative in ef- fecting clearance of the esophageal contents into the stomach. The interrelationships between pain from gastroesophageal reflux, esophagitis, and the duration of periods of high esophageal acidity are complex. The frequency of gross endoscopic abnor- mality of the esophageal mucosa in patients with typical symptoms of gastroesophageal reflux has been variously reported at 76.9% by Ward et al (9) and 50% by Sladen et al (10). It is now generally ac- cepted that many patients with typical symptoms of gastroesophageal reflux show little endoscopic ab- normality of the esophageal mucosa. This would be in keeping with the findings of Woodward (5) and the present study that there is no close correlation between pain severity or frequency on the one hand and the duration of periods of high esophageal acid- ity on the other. Indeed several of our patients who experienced pain during overnight esophageal pH recording had shown little or no evidence of esopha- gitis at endoscopy. This would be in keeping with the observation that esophageal acid perfusion may provoke pain in healthy subjects who have had no symptoms of gastroesophageal reflux (11). Skin- ner (12) found 20% of 226 patients with peptic stric- tures had not had to seek medical advice for pain be- fore the onset of dysphagia. Thus there appears to be no close correlation between the severity of esophagitis and the intensity or frequency of pain. In this study we showed a significant correlation be- tween the duration of nocturnal periods of high eso- phageal acidity and the severity of esophagitis. Thus it does seem that in general the greater the ex- posure of the esophageal mucosa to acid, the great- er the degree of resulting esophagitis. None of our patients had previously undergone surgery for pep- tic ulcer, and it may well be that because bile is a major irritant of the esophageal mucosa in this group, the relationship between esophageal acidity and esophagitis would be less apparent.

Digestive Dis . . . . . . Vol. 22, No. 4 (April 1977) 369

A T K I N S O N A N D V A N G E L D E R

REFERENCES

1. Pattrick FG: Investigation of gastro-oesophageal reflux in various positions with a two lumen pH electrode. Gut 1 t:659~67, 1970

2. Spencer J: Prolonged pH recording in the study of gastro- oesophageal reflux. Br J Surg 56:912-914, 1969

3. Kantrowitz PA, Corson JG, Fleischli DL, Skinner DB: Measurement of gastro-oesophageal reflux. Gastroenterolo- gy 56:666-674, 1969

4. Tuttle SG, Grossman MI: Detection of gastro-esophageal re- flux by simultaneous measurement of intra-luminal pressure and pH. Proc Soc Exp Biol Med 98:225-227, 1958

5. Woodward DAK: Response of the gullet to gastric reflux in patients with hiatus hernia and oesophagitis. Thorax 25:459- 464, 1970

6. Atkinson M: Patho-physiology of gastro-oesophageal reflux.

Topics in Gastroenterology, Vol. 4. SC Truelove, JA Ritchie (eds). Oxford, Blackwell Scientific Publications, 1976, pp 67-83

7. Booth DJ, Kemmerer WT, Skinner DB: Acid clearing for the distal oesophagus. Arch Surg 96:731-734, 1968

8. Stanciu C, Bennett JR: Oesophageal acid clearing: One fac- tor in the production of reflux oesophagitis. Gut 15:852-857, 1974

9. Ward AS, Wright DH, Collis JL: The assessment ofoesopha- gitis in hiatus hernia patients. Thorax 25:568-572, 1970

10. Sladen GE, Riddell RA, Willoughby, JMT: Oesophagos- copy, biopsy and acid perfusion test in diagnosis of "reflux oesophagitis." Br Med J 1:71-76, 1975

11. Bennett JR, Atkinson M: Oesophageal acid perfusion in the diagnosis of cardiac pain. Lancet 2:1150-1152, 1966

12. Skinner DB: Symptomatic esophageal reflux. Am J Dig Dis 11:771-779, 1966

370 Digestive Diseases, Vol. 22, No. 4 (April 1977)