QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Stevenson, D. D. Search for Related Content PubMed PubMed Citation Articles by Stevenson, D. D.

---

Supplement

Monosodium Glutamate and Asthma^{1 ,2}

Division of Allergy, Asthma and Immunology, Scripps Clinic and the Scripps Research Institute, La Jolla, CA 92037

	ABSTRACT

TOP ABSTRACT INTRODUCTION The beginnings of the... Subsequent studies to confirm... DISCUSSION REFERENCES

 
Allen et al. (1987) conducted oral monosodium glutamate (MSG) challenges with 32 asthmatic volunteers and reported that 14 reacted to MSG. Another study by Moneret-Vautrin (1987) also reported MSG-induced asthma attacks in 2 of 30 asthmatic patients. Four additional studies have been conducted and none has confirmed the results of the above authors. These studies, by Schwartzstein et al. (1987) , Germano (1991) , Woods et al. (1998) and Woessner et al. (1999) , challenged a total of 45 patients who gave a history of asthma attacks in oriental restaurants. None of these patients experienced asthmatic reactions after ingesting MSG (one-sided confidence interval of 0–0.066). Another 109 asthmatic patients, without a history of asthma in oriental restaurants, also did not react to ingestion of MSG (one-sided confidence interval of 0–0.027). With a confidence interval < 0.05 there is a >95% probability that MSG history–negative asthmatic patients are not sensitive to MSG. For the MSG history–positive asthmatics, 45 patients, in well-performed studies, underwent negative challenges to MSG, contrasting with two studies reporting positive challenges. Allen et al. (1987) and Moneret-Vautrin (1987) , who reported positive MSG challenge results, performed studies with the following characteristics: 1) single blinded, conducted after discontinuing essential antiasthma medications; 2) used effort-dependent peak expiratory flow rate measurement of lung function; 3) added AM bronchodilators in some patients; 4) ignored wandering baselines on the placebo challenge days; and 5) conducted some challenges in the AM and some at night. In summary, the existence of MSG-induced asthma, even in history-positive patients, has not been established conclusively.

KEY WORDS: • humans • asthma • monosodium glutamate

	INTRODUCTION

TOP ABSTRACT INTRODUCTION The beginnings of the... Subsequent studies to confirm... DISCUSSION REFERENCES

 
Asthma is a common disorder. On the basis of the study of a total community survey in a small town in Michigan, the cumulative prevalence of asthma for all ages was 6% in females and 7.2% in males (Broder et al. 1974 ). In males, examining the age group 5–9 y, the prevalence of asthma was 8.9%, and 10.7% in the 10- to 15-y-old group. It is well known that asthma has a worldwide distribution; a trend toward increasing incidence and death from asthma has also been well documented (Farber et al. 1997 , McFadden and Warren 1997 , Sly 1994 ). The severity of asthma ranges from mild and intermittent to severe and persistent (Guidelines for the Diagnosis and Management of Asthma 1997 ). Although the causes of asthma are complicated and vary from patient to patient, inflammation of the bronchial airways is the characteristic finding in the majority of asthmatic patients (O’Byrne 1997 ). In addition, multiple trigger or provoking factors activate asthma attacks in asthmatic patients already afflicted with inflammation of the bronchial tree. Provoking factors vary from patient to patient, but are important because avoidance of trigger factors, such as allergens, irritant smoke or viral colds, can substantially improve the quality of life of asthmatic individuals (Mathison 1998 ). All of these facts establish asthma as a major disease with serious consequences, including death.

Therefore, the discovery of a new provoking factor, which might trigger asthma attacks, frequently generates considerable interest and important implications, not only for the patients, who suffer from the disease, but also their consulting physicians, public health officials and government regulatory agencies.

	The beginnings of the monosodium glutamate-asthma connection

TOP ABSTRACT INTRODUCTION The beginnings of the... Subsequent studies to confirm... DISCUSSION REFERENCES

 
In a letter to the New England Journal of Medicine, Allen and Baker (1981) reported two patients who described asthma attacks 12 h after ingesting a meal in a Chinese restaurant. The authors reported that both asthmatic patients developed asthma attacks, with associated declines in peak expiratory flow rates (PEFR)³ 10 and 12 h after oral challenges with capsules containing 2.5 g of monosodium glutamate (MSG). One patient developed such severe asthma that she had to be intubated. The authors concluded that MSG was responsible for their bronchospasm. They wrote: "This long delay between eating foods containing MSG and the development of asthma is unlike any other reaction to a food additive and probably accounts for the lack of awareness of MSG-induced asthma." They went on to write that "Chinese-restaurant asthma can be life threatening and difficult to recognize. Unless patients and physicians are alerted to this unusual reaction, unnecessary deaths may occur."

This letter, in a respected journal, subsequently became the basis for presenting an entirely new provoking factor for asthma, i.e., given the pervasiveness of glutamate in both natural foods and as an additive in prepared foods, asthmatic patients are continuously exposed to glutamate in their diets. Therefore, a potential explanation for ongoing asthma might be the daily ingestion of glutamate, both in its natural form and as a food additive (MSG).

Subsequently, Allen et al. (1987) expanded their investigation of MSG-induced asthmatic reactions by including an additional 30 patients, to make a total of 32 asthmatic patients in their 1987 report. Of these 32, 14 gave a history of wheezing attacks after ingesting an oriental meal, and 18 were recruited because they had "unstable asthma, usually with sudden, severe, unexplained attacks" and "sensitivity to other chemicals (aspirin, benzoic acid, tartrazine and sulfites)." All 32 asthmatic subjects were admitted to hospital, underwent single-blind oral challenges with MSG, followed by PEFR measurements for 12 h. In summary, Allen et al. (1987)<,zharvx> reported that 14 of 32 patients experienced asthma attacks after ingesting 1.5 g MSG (one patient) or 2.5 g MSG (13 patients), with elapse times from ingestion of MSG to onset of a 20% decline in PEFR ranging from 1 to 12 h. One of 14 "reactors" was recorded as having an elapse time, from MSG ingestion to reaction, of 12 h, one had an elapse of 10 h, seven had elapse times of 3–6 h and four had elapse times of 1–2 h. None of the patients was said to react to MSG in h 1 after ingestion, the time when circulating glutamate concentrations would most likely have been elevated.

The study by Allen et al. (1987) has been criticized for a number of reasons. First, effort PEFR were used, instead of the more reliable flow/volume measurement. Second, placebo challenges were always carried out on d 1, and theophylline was discontinued just before this first placebo challenge day. By eliminating a bronchodilator at the beginning of the challenge sequence, deprivation was minimal during the placebo day, but marked during the second and third challenge days, when MSG was administered. Such experimental circumstances make meaningful comparisons between placebo and active treatment days impossible (Stevenson, 1988 ). Third, the authors stated that in some patients "an inhaled beta agonist bronchodilator was administered once at 6 AM, 3 h before first challenge." Such a practice is associated with the probability that at least some unstable asthmatics experienced initial bronchodilation after ß-agonist treatment, followed by declining lung function values 6 h later, as the effects of the bronchodilator disappeared (Stevenson 1988 ). One might have the erroneous impression that ingestion of MSG, not the clearance of the bronchodilator, was responsible for the decline in lung function values. Fourth, curious and lax criteria for interpreting PEFR values during placebo challenges also existed. For example, declines in PEFR of 20% were accepted as normal "baseline" or "no significant change." Fifth, the authors accepted morning baseline PEFR values as stable, even though these values changed significantly from day to day, and low values were improved by inhaling ß-agonists. Sixth, some patients were challenged with MSG at night and presumably were awakened hourly or allowed to sleep all night (in the case of patients #1 and #2). Other patients were challenged during the day with MSG; all placebo challenges were conducted during the day. Spontaneous nocturnal asthma is a known consequence of poorly controlled asthma. Thus, whether or not MSG-induced asthma or spontaneous nocturnal asthma occurred during these nighttime challenges cannot be determined in this study. It is unclear why the authors decided to start the MSG challenges either in the AM or the evening and combine the data as if the test environments were the same. Furthermore, the reader is not informed concerning which patients received their challenge doses in the AM or evening (except patients #1 and #2). Seventh, patient # 3 was challenged with MSG doses of 0.5, 1.5 and 2.5 g and experienced a 20% or greater decline in PEFR with each of the three doses. The other 11 (of thirteen) "reactor " patients "reacted" only to the 2.5-g dose of MSG and one patient "reacted" only to 1.5 g of MSG. Therefore, repeated challenges were performed in only one patient and in that patient, the challenges were not double blind. The timing of the decline in PEFR values for patient #3 was always 3 h after MSG. We are not told whether or not this patient received ß-agonists 6 h before this drop in PEFR values on each of the three MSG challenge days. Furthermore, because placebo challenges were always on d 1, the normal lung function values on the placebo day could be explained by the continued presence of theophylline on d 1 only. Rather than a dose-response challenge sequence, with reactions to larger doses of MSG, we could be observing the consequences of unstable airways in a patient whose theophylline was discontinued and was therefore treated with ß-agonists 6 h before reactions.

	Subsequent studies to confirm the validity of these claims

TOP ABSTRACT INTRODUCTION The beginnings of the... Subsequent studies to confirm... DISCUSSION REFERENCES

 
Five additional studies attempted to clarify the issue of whether MSG induces bronchospasm in asthmatics (Table 1 ). A study by Moneret-Vautrin (1987) reported 2 of 30 asthmatic patients undergoing oral challenges with MSG (2.5 g) developed asthmatic reactions. The author studied patients in a single-blind, placebo-controlled protocol and used declines in PEFR determinations as evidence for asthma attacks. The patients were observed, and hourly PEFR determinations were obtained for 12 h. The two "reactors" were not rechallenged in a double-blind protocol. Furthermore, both patients exhibited wandering baseline PEFR values during their placebo challenges, such that the differences between placebo and MSG PEFR determinations were difficult to detect. One of the two "reactors" was alleged to be sulfite sensitive but "ate frequently in Chinese restaurants without incident." It is not stated whether any other patients had or had not previously experienced asthma attacks in oriental restaurants. The study protocol was unique in that corticosteroid therapy was discontinued 21 d before and theophylline therapy was terminated 3 d before challenges. Because 10 study subjects were said to be intolerant of aspirin and nonsteroidal anti-inflammatory drugs, and another 7 were allergic to house dust, discontinuing antiasthmatic therapy could have led to airway instability. Therefore, the chances of false-positive challenge results were high.

A third study by Germano et al. (1991) reported that 1 of 30 asthmatics, during single-blind screening oral challenges with MSG (up to 6 g; 7.5 g cumulative dose over 2 h), experienced a significant reduction in FEV1 values. However, when the one preliminary "reactor" was rechallenged with the same dose (6 g of MSG), under double-blind, placebo-controlled conditions, the response to MSG challenge was negative. This study was criticized because only 2 of 30 asthmatic patients gave a history of chest tightness or wheezing after ingestion of a Chinese meal. Furthermore, the study was reported only in abstract form.

The fourth study by Woods et al. (1998) was conducted on an outpatient basis on 12 asthmatic subjects, who reported that MSG caused them to have asthma attacks. Elaborate controls were instituted, including strict dietary avoidance of MSG, home spirometry (PEFR measurements) before and after the challenges, as well as a double-blind, placebo-controlled challenge protocol. Usual medications were continued. In addition, methacholine inhalation challenge was performed before and after MSG challenges to determine whether nonspecific bronchial hyperactivity occurred as a result of MSG exposure. Markers for inflammation were obtained before and after MSG challenges [serum eosinophilic cationic protein (ECP) and tryptase]. The actual challenge protocol consisted of a standard breakfast (pears, cereal, toast, eggs and decaffeinated coffee), consumed 30 min before challenge, followed by ingestion of capsules (placebo, MSG 1 and 5 g). Patients were monitored in the outpatient laboratory for 8 h with observation and FEV1 determinations and then sent home with a PEFR meter for the next 4 h. The results of the study were completely negative. One patient drifted down to 15% in her FEV1 values during MSG challenge days but the same change occurred during placebo days. The areas under the curve, comparing FEV1 values for placebo and MSG, were the same for the study group. Furthermore, there was no increase in bronchial nonspecific hyperactivity or changes in tryptase or ECP serum levels.

This study has received only three minor criticisms. First, because it was an outpatient study, reliability of the dietary program could not be supervised directly. For the same reason, patients could not be monitored for a full 12 h after ingesting placebo or MSG. During the last 4 h, the patients were at home performing unsupervised PEFR determinations. Finally, the number of subjects participating in the study was small. Given the elaborate controls and procedures, all of which required extensive investigator time and resources, the use of a small number of subjects had both theoretical and practical merit. However, the 12 patients represented a high risk group for MSG sensitivity, according to the criteria of Allen et al. (1987) [Allen reported that 10 of 14 (71%) of his MSG history–positive asthmatics reacted to MSG during challenges with 2.5 g of MSG in capsules]. We would expect between 8 and 9 of the patients in the study of Woods et al. (1998) also to have reacted to MSG. Indeed, Woods recruited her patients from the same Australian population used by Allen.

The fifth study, Woessner et al. (1999) , was performed in our inpatient General Clinical Research Center (GCRC) of The Scripps Clinic, Green Hospital and The Scripps Research Institute. Using newspaper advertisements and referrals, two groups of patients were recruited for this study. Group A consisted of 30 asthmatic patients, who had experienced asthmatic attacks in oriental restaurants. These volunteers believed that MSG induced their asthmatic attacks and were attempting to avoid ingestion of MSG. Group B consisted of 70 asthmatic patients, referred to the Scripps Clinic for acetylsalicylic acid (ASA; aspirin) challenges and desensitization (Pleskow et al. 1984 ). Group B patients had never experienced asthma attacks in oriental restaurants. According to Allen at al (1987) , tartrazine-, sulfite- and ASA-sensitive asthmatic patients are at high risk for MSG-induced asthmatic attacks.

Before admission, usual medications for maintenance of asthma remissions were continued (inhaled and systemic corticosteroids and theophylline). However, antihistamines, ß-agonists and leukotriene-modifying drugs were discontinued. If a patient was experiencing an exacerbation of asthma from sinusitis, viral respiratory illness or other conditions, such events were treated and cleared before admission to GCRC. If FEV1 values were <70% of predicted, a burst of prednisone (30–60 mg/d) was initiated and the patient was reevaluated several weeks later for potential admission to the MSG oral challenge studies. If FEV1 was >70% of predicted, without requiring ß-agonist by inhalation to achieve this result, the patient was scheduled for admission to GCRC for oral challenge studies without prednisone bursts. Patients were instructed to ingest a low MSG diet and to avoid adding MSG to their food during the week before admission.

Patients were admitted to the GCRC, underwent complete history and physical exam, signed consents forms approved by Human Subjects Committee, verified and continued their maintenance medications and continued a low MSG diet. They underwent serial lung function studies to establish consistency of FEV1 values.

On the next day, if the FEV1 AM baseline value was 70% of the predicted value, single-blind, placebo challenges were performed. Regular maintenance medications were given at 0600 h. A low MSG breakfast was served between 0600 and 0700 h. Between 0700 and 0800 h, five red capsules containing sucrose (placebo) were ingested by the patient. Spirometry (best of 3 expirations) was performed before and after (hourly) initial capsule ingestion for 12 h, and again at 24 h. The patient was under continuous monitoring and observation by a research nurse, who recorded the appearance of any symptoms or signs. A low MSG lunch was served between 1200 and 1300 h and five more placebo capsules were given between 1300 and 1400 h.

On the next day, single-blind challenges were continued if, during the baseline placebo challenge, FEV1 values had varied by <10% during the previous day’s placebo challenges. Furthermore, to ensure day-to-day stability, baseline and 24-h FEV1 values were required to be within ± 5% of each other. If these criteria were not met, the patient was classified as having an unstable airway and did not undergo MSG challenges. For asthmatics with stable bronchial airways, using the above criteria, MSG challenges were started in the same manner as for the placebo day, except for the substitution of MSG 2.5 g in five red capsules (500 mg of MSG/capsule) given between 0700 and 0800 h. FEV1 values were obtained every hour for 12 h and again at 24 h. All symptoms were recorded in the same manner as described for the placebo day. At the 6-h interval, five placebo capsules were given to conform to the challenge sequence on the preceding placebo challenge day.

The criterion for a presumptive or preliminary (single-blind) MSG-induced asthmatic attack was a 20% decline in FEV1 values. Other symptoms, such as wheezing, chest tightness, cough, headache, flush or rash, were recorded and compared with those symptoms occurring on the placebo day. If there were no significant changes in FEV1 values (i.e., <20% decline from baseline), the patient was either discharged from the GCRC or proceeded to additional oral challenge studies on the d 4 (usually with aspirin).

If the FEV1 values fell by 20% (presumptive screening positive), the patient underwent two double-blind, placebo-controlled challenges. An unblinded nurse kept the challenge sequence sequestered until the blinded nurse completed both double-blind challenges.

A total of 142 potential subjects responded to advertisements or were referred for studies. Eight patients, with a history of oriental restaurant asthma attacks, could not arrange their schedules to be admitted for challenge studies. One of the eight patients called to complain to the nurse coordinator about the life-threatening danger of MSG challenges and declined the nurse’s invitation to participate in the study.

A total of 134 patients underwent screening for the presence of asthma and agreed to participate in the challenge studies. However, 34 patients did not successfully meet the placebo challenge criteria for the following reasons: 1) they produced low baseline FEV1 values; 2) they experienced a > 10% change in FEV1 values during the 12-h placebo day; 3) they produced significant variations in AM baseline FEV1 values from placebo to challenge day; or 4) they relied upon ß-agonists to maintain airway stability. Because we were unable to obtain airway stability in these subjects, they were not included in the MSG challenge phase of the study. The remaining 100 patients, who met all study criteria, as previously listed, underwent challenges with MSG (2.5 g) as follows.

Group A.

This group (MSG history positive) consisted of 30 subjects. Only one experienced a 20% decline in FEV1 values during the single-blind screening challenge. This patient was asymptomatic when her spirometry recorded a 20% drop in FEV1. The nurse conducting the study asked the patient to expire into the spirometer for only one expiratory blow, which was decreased by 20%. Inexplicably, the nurse did not follow the protocol and conduct the additional two expirations, a standard procedure used to produce consistency. In this patient, serum tryptase baseline was <1 ng/mL before and after the 20% decline in FEV1 value. Two double-blind, placebo-controlled challenges with MSG (2.5 g) were then conducted in this patient, and the FEV1 values varied by < 1%. In the studies by Allen et al. (1987) and Moneret-Vautrin (1987) , this patient would have been included as being MSG sensitive because neither author included a double-blind confirmatory challenge, with the provoking dose of MSG, after a single-blind positive response.

In these 30 patients, there were many candidates, other than MSG, that could act as precipitants for the asthma experienced in oriental restaurants, i.e., gastroesophageal reflux, cigarette smoke sensitivity, food allergy, unstable asthma, aspirin sensitivity and anxiety/depression. In most cases, a friend or relative had suggested that MSG was responsible for the original oriental restaurant asthma attacks. All 30 patients were relieved to hear at the conclusion of our study that they were not MSG sensitive. Table 2 lists the symptoms experienced by a minority of the patients during the placebo and MSG challenge days. Headaches predominated and were slightly more common during placebo challenge days. None of the patients experienced any symptoms suggesting asthma.

This group (MSG history negative) consisted of 70 asthmatic patients who were referred to the Scripps Clinic for aspirin challenges, followed by aspirin desensitization. None experienced any symptoms suggesting asthma in oriental restaurants and their FEV1 values did not change significantly during placebo or MSG challenge days. None of these 70 patients experienced chest tightness, wheezing dyspnea or cough during either the placebo or MSG challenge days. Thus the incidence of asthmatic reactions to MSG was 0 of 70. The one-sided confidence interval for 0 of 70 is 0–0.04, which is < 0.05, or a 95% confidence that the event (MSG asthmatic sensitivity) does not occur in this subpopulation of asthmatics. Recorded symptoms on the placebo and MSG challenge days are listed in Table 2 . Headaches were more common on the placebo days, but a scattering of other symptoms occurred more frequently on MSG challenge days.

	DISCUSSION

TOP ABSTRACT INTRODUCTION The beginnings of the... Subsequent studies to confirm... DISCUSSION REFERENCES

 
The following discussion is divided into sections relating to the validity of the study of Allen et al. (1987) and the question whether MSG induces asthma attacks.

Medication support during challenge studies in asthmatics

A dilemma exists between the need to support bronchial patency with antiasthmatic therapy during challenges, thus preventing spontaneous bronchoconstriction in the middle of a challenge, and the need to allow bronchoconstriction to occur if stimulated by the challenge substance. Ideally, all asthmatics should be challenged with test substances when their asthma is in complete remission and they are not taking any antiasthmatic medications. Such an ideal state requires the recruitment of asthmatics with the acute intermittent form of the disease. Asthmatic populations, such as those with aspirin sensitivity, are usually moderately or severely persistent asthmatics. According to Allen et al. (1987) , these are the asthmatics most at risk for reactions to a variety of ingested substances. Allen et a1. (1987) and Moneret-Vautrin (1987) preferentially selected study populations of asthmatic patients whose asthma was unstable (irritable airways) and therefore most likely to require daily medication to support bronchial patency.

Withdrawal of theophylline has been shown to uncover asthma activity and lead to spontaneous decline in lung function tests in the subpopulation of asthmatic patients that Allen studied (Stevenson 1988 , Weber et al. 1979 ). After ingesting a potential asthma-producing substance, simultaneous decline in lung function values might be erroneously ascribed to the test substance, rather than withdrawal of an essential antiasthmatic medication, as was true in the study by Weber et al. (1979) . If the investigator then compounds the problem of declining lung function tests by adding an inhaled bronchodilator, an additional error is created. After inhalation of a short-acting ß-agonist, lung function tends to improve by 20% or more to a new baseline. However, because this baseline is maintained by drug effect, lung function returns to the true baseline 4 to 6 h later, when the drug effect wears off (Stevenson 1988 ). In the Allen (1987) study, both of the above principles were violated. Theophylline was discontinued at the beginning of d 1 of challenges and ß-agonists were added only to those patients with low AM PEFR determinations.

In the study of Moneret-Vautrin (1987) , theophylline was discontinued 3 d before challenges but, more importantly, corticosteroids were discontinued 21 d before challenges in a population of asthmatics with a high prevalence of steroid dependency. Uncontrolled asthma was essentially guaranteed in this study. Evidence for this statement can be observed in the highly variable placebo-associated PEFR values recorded in individual patient PEFR time grafts.

We have shown that oral ASA routinely induces bronchoconstriction in ASA sensitive asthmatics whose airway patency is maintained with corticosteroids and theophylline (Pleskow et al. 1984 ). In fact, it is impossible to challenge unstable asthmatics without medication support because false-positive results, regardless of what "test substance" is ingested, will be the consequence of discontinuing medications. Furthermore, of what importance is an oral challenge substance that induces such weak asthma attacks that it cannot overcome the maintenance bronchodilator effects of corticosteroids and theophylline? Would a patient ingesting MSG in an oriental restaurant even know they were reacting to MSG if they were taking their usual bronchodilators? If, for the sake of argument, the patient omitted their usual bronchodilators before eating in an oriental restaurant, would an asthma attack be due to the omission of medications or the ingestion of MSG? In dealing with these important questions, we maintain that continuation of baseline, long-acting antiasthmatic medications in most asthmatic subjects should be part of any serious protocol. In this manner, the investigator has a greater chance of recorded stable lung function during the placebo challenge days, particularly in asthmatics such as those recruited in the studies by Allen et al. (1987) and Moneret-Vautrin (1987) .

This conundrum is further complicated by a recent study by Szczeklik et al. (1998) . These investigators demonstrated that the long-acting bronchodilator, salmeterol, given before inhalation aspirin challenges, completely blocked ASA-induced reactions to the prior threshold dose of ASA-lysine. Furthermore, salmeterol significantly attenuated release of leukotrienes during ASA inhalation challenges. Although oral ASA challenges have induced bronchospasm in patients receiving salmeterol in our clinic, this issue remains controversial. Therefore, to eliminate any possibility that we were preventing MSG-induced reactions by pretreating with salmeterol or sustained-release albuterol tablets, we eliminated these patients from inclusion in our study group.

Criteria for positive challenge results

Flow/volume measurements integrate expiratory flow rates with the volume of air expelled, producing a flow/volume curve. The configuration of the expiratory curve tells the operator whether or not a maximum effort has been expended. Furthermore, each curve can be compared to previous tracings stored in computer memory. FEV1 measures a volume of air expelled in the 1st s of expiration. FEV1 measurements reflect constriction of bronchial airways and are reproducible from hour to hour (Fitzgerald et al. 1973 , Kory et al. 1963 ). With the use of a flow/volume recording device, such as a wedge spirometer, accurate and reproducible measurement of lung function can be recorded. There are two reasons for this. Even though lung function always has an element of effort dependency, the shape of the expiratory curve can be visualized on the recording screen and, if amputated or distorted by poor effort, can be readily detected and multiple repeat expiratory efforts initiated. Three recordings per measurement are routinely employed and the best of three selected. Submaximal effort can be detected with such a system. In addition, vocal chord dysfunction can be identified because the inspiratory curve of the flow/volume loop shows a characteristic flat and notched pattern.

By contrast, peak expiratory flow rates (PEFR) measure airflow past a recording device and are highly effort dependent. An individual can easily disguise a poor effort with a resulting decline in PEFR numbers. With maximum effort in a cooperative, unbiased subject with open vocal chords, PEFR can be accurate and reproducible. However, if the patient is biased and wishes to demonstrate "bronchial constriction," it is very simple to pretend to blow hard while constricting the neck muscles and producing a lower PEFR number. Vocal chord dysfunction can also produce a lowered peak flow rate measurement. Both the Allen et al. (1987) and Moneret-Vautrin (1987) studies relied upon effort-dependent PEFR determinations to prove bronchoconstriction. The other four studies, which did not demonstrate airway changes after ingesting MSG, used FEV1 measurements obtained with flow/volume spirometry recording systems.

Double-blind vs. single-blind challenge studies

Single-blind studies rely upon the concept that the patient is blinded to knowledge of the substance ingested, whereas the nurse has this information. Placebo controlled means that a similar time of challenge will occur, after ingestion of identical placebo capsules. Most single-blind, placebo-controlled studies conduct the placebo challenge first because of the time-saving element of discontinuing the challenge if airways are so irritable that the placebo challenge is positive. This is what we elected to do in the Woessner study. The main problem with single-blind studies is the ability of some patients to read concern or fear in the face and eyes of their study nurses. Double-blind studies are accurate in terms of blinding the patient because the patient’s individual nurse does not know whether she gave a placebo or a test substance. However, double-blind studies involve more time and personnel and are therefore less efficient.

A third strategy is to conduct a single-blind screening test first; if it is positive, the challenges are repeated using a double-blind protocol. If a change in lung function occurs on the MSG oral challenge days in a single-blind study, such a change could be due to investigator or patient bias. Therefore, any positive single-blind challenge result must be validated with several repeat double-blind, placebo-controlled challenges when the asthma is well controlled. This challenge sequence was used by Germano et a1. (1991) . In 1 of 30 single-blind challenges, one patient experienced a 20% decline in FEV1 values and became a potential reactor to MSG. However, when the challenge was repeated with the use of a double-blind protocol, the repeat challenge to MSG was negative. A similar protocol was used by Stevenson et a1. (1986) when studying presumed "tartrazine sensitivity" in ASA-sensitive asthmatics. Only 6 of 150 ASA-sensitive asthmatic patients experienced positive single-blind challenges to tartrazine. These six patients were then studied in more detail using double-blind challenges with tartrazine and placebo and all challenges were found to be negative. In the study by Woessner et al. (1999) , a single-blind challenge protocol was used. When one patient was recorded to have a 20% drop in FEV1 values, two double-blind challenges were initiated and <1% change in FEV1 values was recorded for both placebo and MSG challenges. If the patient were truly MSG sensitive, the same dose of MSG should reproduce the asthma attack every time it is given. In the Allen et al. (1987) and Moneret-Vautrin (1987) studies, the above patient would have been counted as being sensitive to MSG because both studies relied upon a single-blind challenge protocol without confirmatory double-blind challenges.

Finally, in the Woessner study, 34 patients were excluded from undergoing MSG challenges because of irritable airways (n = 21) or reliance on long acting ß-agonists (n = 13). In the Allen et al. (1987) and Moneret-Vautrin (1987) studies, our 34 patients with unstable airways would have been accepted for MSG challenges. A high probability of false-positive challenge results would have been expected and probably occurred.

MSG doses during oral challenge studies

Allen et al. (1987) used oral challenge doses of 0.5, 1.5, 2.5 and 5 g of MSG in gelatin capsules. One patient was alleged to have reacted to 0.5 g and another to 1.5 g. All remaining "reactors" (12 of 14) experienced changes in PFFR values after ingesting 2.5 of MSG. Although the 18 patients who did not "react" after ingesting 2.5 g of MSG were challenged the next day with 5 g of MSG, none "reacted" to this larger dose. Therefore, Allen et al. (1987) established 2.5 g of MSG in capsules as the largest dose, which they claimed induced bronchospasm. Because none of Allen’s patients reacted to 5 g of MSG, a logical challenge sequence would be placebo on one day and 2.5 g of MSG in capsule on d 2. In the Woessner et al. (1999) study, a challenge dose of 2.5 g of MSG in capsules was selected because it reproduced precisely the largest dosage of MSG used in the study of Allen et al. (1987) .

Patient selection

Before 1981, when Allen and Baker (1981) published their initial report, publicity associated with MSG as a threat to asthmatics was not an issue. However, in Australia, where Allen lives and works, a1996 general survey showed that asthmatic patients perceived that MSG was the most common food additive to induce asthma attacks (Woods et al. 1996 ).

Assuming that asthmatics who believe that MSG induces asthma attacks are also willing to undergo oral challenge studies with MSG, it should be relatively easy to recruit subjects for such studies. However, our experience (Woessner et al. 1999 ) was the opposite. Despite newspaper recruitment ads, letters and telephone calls to 200 referring physicians and constant questioning of the Scripps Clinic asthma population of southern California, we were able to identify only 38 asthmatic patients who perceived that MSG caused asthma attacks. Of the 38, eight subjects did not make themselves available for challenge studies. One patient believed that the study posed an extreme danger to the participants and declined our invitation to join the study. Therefore, it appears that a fearful population of believers in MSG adverse effects exists. Because these patients believe that MSG is dangerous, it is not surprising that they will not volunteer to be studied with MSG challenges. There is the additional probability, based upon the group we studied (Woessner et al. 1999 ), that a more casual group of MSG history–positive asthmatics also exists. It appears that only the casual group will submit to MSG challenges, leaving the fearful group unexplored.

It seems logical to assume that Allen et al. (1987) challenged at least some patients who perceived that MSG caused asthma because 14 agreed to participate in his studies. Of the 14 reported MSG reactors, 10 gave a history of asthma associated with a Chinese restaurant meal. Thus, Allen et al. (1987) claimed that patients with a history of asthma attacks in oriental restaurants will submit to oral MSG challenges.

Statistical calculations relative to the probability that MSG induces asthmatic attacks

In the Allen study, of the asthmatic patients providing a positive history of asthma attacks associated with Chinese meals, 10 of 14 (71%) were reported to react to MSG. Thus, in the Woessner et al. (1999) study, 21 of 30 (71%) MSG history–positive patients should have, but did not react to MSG. If one examines Table 1 , a combined total of 45 MSG history–positive patients were challenged with MSG in four studies [since the Allen et al. (1987) and Moneret-Vautrin (1987) studies]. Of these, none has reacted during MSG challenges. If Allen et al. (1987) had actually identified MSG-induced asthma, at least some of the 45 MSG history–positive patients, challenged in all subsequent studies, should have experienced a positive challenge to MSG (i.e., 71% of 45 is 32 patients). However, 45 patients is too small a sample to prove that MSG sensitivity does not exist. To provide statistically valid results for the entire asthma population (i.e., 95% confidence limit), at least 60 patients, with a history of asthma attacks occurring in oriental restaurants, would have to undergo properly controlled MSG challenges and have negative challenge results. In Table 1 , the one-sided confidence interval for 45 asthmatic patients who did not react to MSG was between 0 and 0.66. Recruitment of patients who believe that MSG causes asthma may be complicated by the fact that the association between MSG and asthma may not exist. However, at this point, all we can conclude is that either Allen’s study is flawed or the four follow-up studies are flawed. A difference between an incidence of 10 of 14 (71%) MSG challenge positive for the Allen et al. (1987) study and 0 of 45 (0%) for the other four studies cannot be reconciled if all five studies were conducted in the same manner and on the same population.

Allen et al. (1987) also claimed that 18 asthmatics, with unstable asthma and "sensitivity to other chemicals," were also at risk for MSG-induced asthma, even without a history of asthma attacks occurring in oriental restaurants. In Table 1 , it is apparent that 109 MSG history–negative asthmatics underwent MSG challenges in all four studies, and all challenges to MSG were negative. Thus, the statistical chances of any history-negative asthmatics reacting to MSG is < 0.05 (0–0.027) (95% one-sided confidence interval). Thus, there is a >95% probability that the event does not occur in the population of MSG history–negative asthmatic patients.

It is impossible to reconcile the six studies presented in this review. When one reviews all studies, the study procedures employed by Allen et al. (1987) and Moneret-Vautrin (1987) are not sufficiently credible to persuade me that the authors measured anything other than spontaneous asthma in patients deprived of their essential maintenance medications. Four additional studies have attempted to confirm the results presented above, and none of the asthmatic patients have experienced asthma attacks after ingesting MSG. During these four properly controlled oral challenge studies, I would have expected at least one positive response to MSG challenge, if MSG causes asthma attacks. Certainly, before any governmental restrictions on the use of MSG in humans are activated, well-designed studies will have to demonstrate that MSG can actually induce asthma attacks.

	FOOTNOTES

 
¹ Presented at the International Symposium on Glutamate, October 12–14, 1998 at the Clinical Center for Rare Diseases Aldo e Cele Daccó, Mario Negri Institute for Pharmacological Research, Bergamo, Italy. The symposium was sponsored jointly by the Baylor College of Medicine, the Center for Nutrition at the University of Pittsburgh School of Medicine, the Monell Chemical Senses Center, the International Union of Food Science and Technology, and the Center for Human Nutrition; financial support was provided by the International Glutamate Technical Committee. The proceedings of the symposium are published as a supplement to The Journal of Nutrition. Editors for the symposium publication were John D. Fernstrom, the University of Pittsburgh School of Medicine, and Silvio Garattini, the Mario Negri Institute for Pharmacological Research.

² Supported by grants from the International Glutamate Technical Committee and the General Clinical Research Unit of the Scripps Clinic, Green Hospital and the Scripps Research Institute (MO1R00833).

³ Abbreviations used: ASA, acetylsalicylic acid (aspirin); ECP, eosinophilic cationic protein; FEV, forced expiratory volume; GCRC, General Clinical Research Center; MSG, monosodium glutamate; PEFR, peak expiratory flow rates.

	REFERENCES

TOP ABSTRACT INTRODUCTION The beginnings of the... Subsequent studies to confirm... DISCUSSION REFERENCES

 

1. Allen D. H., Baker G. J. Chinese restaurant asthma. [Letter]. N. Engl. J. Med. 1981;278:796

2. Allen D. H., Delohery J., Baker G. Monosodium L-glutamate-induced asthma. J. Allergy Clin. Immunol. 1987;80:530-537[Medline]

3. Broder I., Higgins M. W., Mathews K. P., Keller J. B. Epidemiology of asthma and allergic rhinitis in a total community, Tecumseh, Michigan. J. Allergy Clin. Immunol 1974;53:127-138[Medline]

4. Farber H. J., Wattingney W., Berenson G. Trends in asthma prevalence. Ann. Allergy Asthma Immunol. 1997;78:265-269[Medline]

5. Fitzgerald M. X., Smith A. A., Gaensler E. A. Evaluation of electronic spirometers. N. Engl. J. Med. 1973;289:1283-1287

6. Germano P., Cohen S. G., Hahn B., Metcalfe D. D. An evaluation of clinical reactions to monosodium glutamate (MSG) in asthmatics, using a blinded placebo-controlled challenge. J. Allergy Clin. Immunol. 1991;87:177abs.

7. Guidelines for the Diagnosis and Management of Asthma Expert panel report II, National Heart, Lung and Blood Institute and the World Health Organization Global Initiative for Asthma, NIH publication no. 97–4051, Bethesda, MD 1997:

8. Kory R. C., Hamilton L. H. Evaluation of spirometers used in pulmonary function studies. Am. Rev. Respir. Dis. 1963;87:228-235[Medline]

9. Mathison D. A. Adult asthma. Middleton E. Reed C. E. Ellis E. Adkinson N. F. Yunginger J. W. Busse W. W. eds. Allergy Principles and Practice 1998;vol. II:901-926 C. V. Mosby St. Louis, MO.

10. McFadden E. R., Warren E. L. Observations on asthma mortality. Ann. Intern. Med. 1997;127:142-147[Abstract/Free Full Text]

11. Moneret-Vautrin D. A. Monosodium glutamate induced asthma: a study of the potential risk in 30 asthmatics and review of the literature. Allerg. Immunol. (Paris) 1987;19:29-35

12. O’Byrne P. M. Leukotrienes in the pathogenesis of asthma. Chest 1997;111:27S-34Ssuppl. l2[Abstract/Free Full Text]

13. Pleskow W. W., Stevenson D. D., Mathison D. A., Simon R. A., Schatz M., Zieger R. S. Aspirin desensitization in aspirin-sensitive asthmatic patients: clinical manifestations and characterization of the refractory period. 1984a;69:11-17

14. Pleskow W. W., Stevenson D. D., Mathison D. A., Simon R. A., Schatz M., Zieger R. S. Aspirin-sensitive rhinosinusitis and asthma: spectrum of adverse reactions to aspirin. J. Allergy Clin. Immunol. 1984b;71:574-579

15. Schwartzstein R. M., Kelleher M., Weingerger S. E., Weiss J. W., Drazen J. M. Airways effects of monosodium glutamate in subjects with chronic stable asthma. J. Asthma 1987;24:167-172[Medline]

16. Sly R. M. Changing asthma mortality. Ann. Allergy 1994;73:259-268[Medline]

17. Stevenson D. D. Oral challenges to detect aspirin and sulfite sensitivity in asthma. N. Engl. Reg. Allergy Proc. 1988;9:135-142[Medline]

18. Stevenson D. D, Simon R. A., Lumry W. R., Mathison D. A. Adverse reactions to tartrazine. J. Allergy Clin. Immunol. 1986;78:182-191[Medline]

19. Szczeklik A., Dworkski R., Mastalerz L., Prokop A., Sheller J. R., Nizankowska E., Cmiel A., Oates J. A. Salmeterol prevents aspirin-induced attacks of asthma and interferes with eicosanoid metabolism. Am. J. Respir. Crit. Care Med. 1998;158:1168-1172[Abstract/Free Full Text]

20. Weber R. W., Hoffman M., Rain D. A., Nelson H. S. Incidence of bronchoconstriction due to aspirin, azo dyes, non-azo dyes and preservatives in a population of perennial asthmatics. J. Allergy Clin. Immunol. 1979;64:32-37[Medline]

21. Woessner R. M., Simon R. A., Stevenson D. D. Monosodium glutamate (MSG) sensitivity in asthma. J. Allergy Clin. Immunol. 1999;104:305-310[Medline]

22. Woods R. K., Weiner J., Abramson M., Thein F., Walters E. H. Patients’ perceptions of food induced asthma. Aust. NZ J. Med. 1996;26:504-512[Medline]

23. Woods R. K., Weiner J. M., Thein F., Abramson M., Walters E. H. The effects of monosodium glutamate in adults with asthma who perceive themselves to be monosodium glutamate-intolerant. J. Allergy Clin Immunol. 1998;101:762-771[Medline]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Stevenson, D. D. Search for Related Content PubMed PubMed Citation Articles by Stevenson, D. D.