- Click here for explanation of classification scheme.

To determine whether MSG would induce brochoconstriction in a group of adults with asthma who perceived that they were MSG sensitive.

• Documented history of reversible airway obstruction but clinically stable (no changes in regular asthma medication in the four weeks before starting the study)
• Must have previously completed a food and asthma questionnaire that investigated perceptions and practice in regard to potential contributions of dietary factors to asthma.

• A previous life-threatening attack of asthma (defined by respiratory arrest, endotracheal intubation, admission to an intensive care unit or pCO₂ less than 50mm Hg)
• Life-threatening anaphylaxis
• Females who were pregnant or lactating
• Those who baseline FEV₁ measurement was less than 60% predicted or less than 1.5L.

Design

Randomized, placebo-controlled, double-blind study.

Blinding Used

• The capsules used in the study (challenge dosages) comprised of 10 size 00 opaque capsules, which were manufactured with a capsule machine filler by a pharmacist not otherwise involved in the study
• All capsules were wiped clean after filling and rolled in lactose powder to prevent any MSG being detected on the outside of the capsules.

Intervention

•  Placebo: 5g lactose in opaque capsule
• 1g MSG
• 5g MSG.

Statistical Analysis

• Based on a summary measure:
  • Late asthmatic changes for FEV₁ and PEF quantified using the two- to 12-hour area decrement
  • Calculated as the area between a line extrapolated from the daily mean pre-challenge measurement and the actual measurements from two to 12 hours after the time of the challenges
  • Unpaired Student T-test was used to compare area decrement measurements on the three control days with the measurements after the active challenges. T>2.92 was regarded as significant at the 5% level.
• Based on serial hourly measurements of FEV₁ and PEF: 
  • The pooled standard deviation of the 11 hourly measurements was obtained from a one-way analysis of variance by time
    • By using this figure, each hourly measurement (for FEV and PEF) on a challenge day could be tested against the mean hourly measurement for that time point on the control days
    • This method is best represented graphically in the form of a lower boundary (for example, a time-response curve with a lower 95% CI)
  • A positive reaction was deemed to have occurred if the time-response curve for any of the active challenges crossed the lower boundary for two consecutive hours in the absence of such a change after placebo challenge
  • This method has a false-positive rate of 1%, although the sensitivity of this method has not been determined
• Descriptive statistics were generated by using the SPSS computer package:
  • Chi square tests were used to assess associations between categorical data
  • Student T-tests were used to assess differences between the means of continuous data
  • To assess marginal heterogeneity, McNemar's test was performed on the symptom-score data. P<0.05 was regarded as significant for all tests.
  • Bland and Altman method was used to assess the agreement between paired PD₂₀ measurements.

Timing of Measurements

• Subjects were recruited between March 1995 and September 1996
• The study consisted of nine visits over an average span of 21 days.

 

	Visit Number
	1	2	3	4	5	6	7	8	9
Average days after Visit 1	0	5	7	10	12	15	18	20	21
Introduction/explain protocol	X
Commence additive-free diet	X
Provide spirometer/check use	X	X	X	X		X	X	X
Provide/explain diary cards	X
Check daily diary cards		X	X	X	X	X	X	X	X
Baseline lung function tests	X				X				X
Lung function test (three hours in lab and up to 12 hours at home)		X	X	X
Soluble inflammatory marker activity (ECP and tryptase)		X				X	X	X
Non-specific BHR to methacholine	X				X				X
Challenge 1 (eight hours in lab, four hours at home)						X
Challenge 2 (eight hours in lab, four hours at home)							X
Challenge 3 (eight hours in lab, four hours at home)								X

  

• Control day visits (visits two, three, four): Subjects went to the lab in the morning for monitoring of spirometry over three hours. The best three expiratory maneuvers for both FEV₁ and PEF were recorded every 15 minutes for the first hour and every 30 minutes for the next two hours. Subjects were then allowed to leave the lab to resume their normal daily routine, but they continued recording their FEV₁ and PEF every hour in exactly the same manner and using the same hand-held spirometer as in the laboratory until retiring to bed that night. The results were recorded in a diary.
• Challenge visits (visits six, seven, eight). Subjects arrived in the lab after an overnight fast. Short-acting [beta]₂-agonist bronchodilator medication was withheld for at least four hours before attending the lab (when possible). The subjects (on average) complied with the diet for 15 days before receiving their first challenge. Spirometry was conducted every 15 minutes for 45 minutes before the challenge. Subjects received 1gm MSG, 5gm MSG, and placebo (lactose) challenge doses as single doses, administered under double blind conditions. Subjects always received the lowest dose (1gm) of MSG before the highest dose (5gm), with the placebo randomly interspersed.
• Subjects continued to monitor their FEV₁ and PEF at the following times after the challenge: 15, 30 and 45 minutes and one, 1.25, 1.5, two, 2.5, three, four, five, six, seven, eight, nine, 10, 11 and 12 hours. Breakfast was consumed 30 minutes after each challenge.
• Subjects were allowed to leave the lab eight hours after the challenge but continued to monitor their FEV₁ and PEF each hour for a further four hours using the same hand-held spirometer and recorded the results and their symptom scores in the diary provided.

  Dependent Variables

• Baseline Spirometry: Conducted by using a dry wedge-bellows spirometer
• Non-specific BHR to methacholine: Conducted using the Newcastle dosimeter technique
• Inflammatory marker assays: Measured using a 10ml peripheral blood sample for baseline measurement of the soluble markers eosinophilic cationic protein (ECP) and tryptase in serum by using Becton-Dickinson SST tubes. Venous blood samples were also taken at one hour for tryptase activity and eight hours for ECP after each challenge.

Independent Variables

• Challenge doses consisted of 1gm MSG, 5gm MSG and 5gm lactose (placebo)
• Each challenge dose comprised 10 size 00 opaque capsules.

Control Variables

• Skin prick testing:
  • Conducted on the volar surface of the arm with standard allergen extracts (Dermatophagoides pteronyssinus, cat  hair, Aspergillus mix, Cladosporium herbarium, Lolium perenne, Southern grass mix and feather mix), histamine (positive control) and a negative control
  • Positive skin test response was defined as a wheal equal to or greater than 3mm in diameter to at least one of the allergens
• All subjects were put on an elimination diet after baseline visit [excluded MSG, colorings and preservatives (including sulfites) and minimized naturally occurring MSG, amines, and salicylates]. A food and beverage diary was kept for the duration of the study to assess dietary compliance.
• Medication Control: On each of the control days, subjects were requested to withhold their short-acting [beta]₂-agonist bronchodilator medication for at least four hours before attending the laboratory. If relief medication was required throughout the control day, the amount used and the time it was taken were recorded; this use was duplicated on the subsequent control and challenge days.

• Initial N: 16 subjects
• Attrition (final N): 12 subjects (seven women, five men). Four of the original 16 subjects withdrew due to time constraints.
• Age: Mean age of 35.3 years (SD, 11.7 years).

Other Relevant Demographics

• All subjects were atopic to at least one of seven common aeroallergens
• All subjects were using short-acting [beta]₂-agonist bronchodilators
  • Seven subjects were also using regular inhaled corticosteroids (mean dose of 1,771mcg beclomethasone dipropionate or equivalent per day)
  • None were currently using oral steroids
• Seven subjects were employed outside the home
• Two worked with home duties
• One was a pensioner
• Two were students.

Anthropometrics

 At baseline:

• Mean spirometry (percent predicted FEV) was 80.6% (SD 12.9%)
• Mean body mass index of subjects was 27.5kg/m² (SD 3.5kg/m²)
• Mean MSG score of the group was 8.0 (range 3 to 13.5 points).

Location

Article did not list specific location but the authors were from Alfred Hospital and Monash University Medical School, Melbourne.

 

Key Findings

Spirometry

• No immediate bronchoconstrictor reactions were observed for early asthmatic reactions
• None of the subjects had either a definite or equivocal late asthmatic reaction.
• With the exception of one subject, all of the subjects had falls in FEV₁, PEF or both of greater than 20% of their pre-challenge spirometry
• No conclusive evidence of true MSG-induced asthma.

 Symptom scores:

• No significant differences in daytime symptom scores or nighttime symptom scores after:
  • Placebo and 1g MSG challenges (P=0.5 and P=1.0, respectively)
  • Placebo and 5g MSG challenges (P=1.0 and P=1.0, respectively)
  • 1g and 5g MSG challenges (P=0.25 and P=1.0, respectively)
• All subjects who reported asthma symptoms after 5g MSG challenge also reported asthma symptoms on both the placebo and 1gm MSG challenge days
• No evidence of a dose-response relationship was found with asthma symptom scores.

Nonspecific BHR

• No evidence of MSG having affected underlying nonspecific BHR overall:
  • No statistically significant differences between the mean baseline PD₂₀ methacholine value after the diet control (95% CI: -0.27, 0.45) or the challenge sequence (95% CI: -0.65, 0.70) or between the mean diet control PD₂₀ value and the challenge sequence (95% CI: -0.35, 0.48).

Soluble Inflammatory Marker Activity

Although five subjects were found to have elevated ECP levels, four of these subjects had elevated ECP at baseline, indicating underlying active asthmatic inflammation rather than a positive response to MSG ingestion.

Other Findings

• Three subjects required short-acting [beta]₂-agonist bronchodilators on each of the control and challenge days
• Two subjects continued with their usual daily long-acting bronchodilators
• Two subjects continued with sodium cromoglycate throughout the study
• Eleven subjects reported symptoms after at least one of the challenge visits but there were no statistically significant differences in reported symptoms after:
  • Placebo vs. 1g MSG challenges (chi square=0.19, P=1.0)
  • Placebo vs. 5g MSG challenges (chi square=0.69, P=0.58)
  • 1g MSG vs. 5g MSG challenges (chi square=2.74, P=0.22).

The study was unable to demonstrate MSG-induced asthma in a group of adults who perceived that MSG did affect their asthma control.

Industry:	Pharmacia (Australia) donated the ECP and tryptase kits for analysis of soluble inflammatory marker activity.
Other:	Asthma Foundation of Victoria

Discussion Lists Limitations

• Use of bronchodilator for some subjects
• Use of a manual spirometers instead of spirometers that store the data electronically
• The amount of time the subject was supervised in the lab for only three hours, even though they were in the lab for eight hours on control days (this is different than going into their own environment after three hours on challenge days)
• A worsening in BHR alone might not be a sensitive measure of detecting food-chemical intolerances in asthmatic subjects.

Reviewer Notes

• This study states that blinding was used and that a pharmacist put the MSG into the pills and wiped them clean so it stays blinded. But the protocol indicates that the subject was always given the 1mg MSG dose before being given the 5mg MSG pill. This makes preservation of blinding suspect if they know the order of the dosages given.
• The elimination diet described is very limited. Although patients submitted food diaries, it's likely that patients would not disclose some of the other foods they ate during that time because they know they are participating in a study.