Abstract
In early October 2009, pediatricians in hospitals in Ho Chi Minh City (HCMC) reported an unusual increase in the number of children presenting with an acute onset of itchy rash and some with breathing difficulties shortly after drinking milk products. The pediatricians considered the illness to be an allergic reaction to milk. The objective of our investigation was to identify the cause of this acute illness. Following early case reports, all hospitals in HCMC were requested to report cases of this illness. Parents were advised to take children with symptoms to a hospital immediately. A case-series was conducted to generate hypotheses on the possible causes of the illness and was followed by a case-control study to test the hypothesis. Parents of all cases and controls were interviewed face-to-face. The association between food items and the allergy was tested using conditional logistics regression. From 9 to 28 October 2009, 19 cases fulfilled the case definition, and 16 of the 17 cases included in the study had consumed milk supplemented with galacto-oligosaccharides (GOS) shortly before the onset of illness. Fifty age-matched, neighborhood controls were enrolled into the case control study. Of the 30 food items consumed by study participants in the preceding 24 h, only the odds ratio (OR) of milk supplemented with GOS was statistically significant: OR=34.0 (95% CI=3.9, 294.8). Laboratory tests of this milk product did not reveal any unusual properties, chemicals, or other toxic substances. This is the first report of an acute allergic reaction to fresh milk supplemented with GOS. However, the specific allergen in this product was not identified. Further cases were not reported once this product was withdrawn from sale. Vietnam's food safety authorities should expand laboratory capacity to detect allergens in food products.
Introduction
I
Our study was conducted to identify the extent of illness following consumption of this product among children attending hospitals in HCMC, and to identify the cause and risk factors for the illness.
Methods
In order to generate an hypothesis on the cause of the illness, a case-series was conducted on patients attending hospitals in HCMC with acute skin reactions and/or unexplained breathing difficulties. This hypothesis was then tested through a case-control study. As no adults had been identified with unexplained allergic reactions from any of the hospitals, the study focused only on children.
Case ascertainment
From 21 October to 15 November 2009, health authorities in HCMC informed the community repeatedly of the outbreak of allergic reactions and advised parents to take children with symptoms to a hospital immediately. This message was announced through radio, television, and newspapers, and no reference was made to any suspected causes of the illness or to the milk products that were suspected as a possible cause by the pediatricians. All hospitals in HCMC were informed of the outbreak and of the media announcements. They were requested to report cases immediately to the Institute of Hygiene and Public Health (IHPH) by fax or phone, and to submit a daily report of “zero cases” if no cases were seen; reporting was continued until 31 March 2010.
Case definition
A case was defined as a child who attended a hospital in HCMC with acute onset of a skin rash and/or other allergic reactions as diagnosed by the clinician from 9 October 2009 onward.
Data collection
Parents of children fulfilling the case definition were interviewed face-to-face at their home by using a structured questionnaire. Both parents were interviewed on the same day whenever possible, up to 2 weeks after the onset of illness. Parents were shown calendars to help them recall the dates the milk had been consumed, as well as photos of different milk containers to identify the milk products consumed in their household. Questions included demographic details, past history of allergy in the patient or family member, food items consumed in the 24 h before onset of illness, the time of onset of symptoms, the nature and duration of symptoms, and time of attendance at hospital. By the time of the home visit, none of the families had samples of the milk product available for laboratory testing.
Details of the symptoms, signs, and clinical diagnosis as recorded by the attending doctor, and the treatment and response to the treatment were transcribed from hospital records. Based on the descriptive data from the cases, consumption of milk supplemented with GOS was implicated consistently in almost all cases. A case-control study was designed to confirm this result and to identify if any other specific food items may have been associated with the illness.
Control subjects were either two or three age-matched (±1 year) healthy children who lived in the immediate neighborhood of the case-patient; they were identified by the head of the local Health Posts where the case resided. Informed consent was then obtained from parents prior to participation in the study.
Analysis of milk samples
The milk containing GOS was obtained from shops and supermarkets in the neighborhood of the cases and tested by the Regional Center for Food Safety Testing of IHPH for the following chemicals and properties: physicochemical norms (carbohydrate, lipid, protein, vitamins A and B1, pH, density); metals (antimony, arsenic, cadmium, lead, bronze, zinc, mercury); aflatoxins B1, B2, G1, G2, M1; antibiotics (tetracycline, oxy-tetracycline, chlor-tetracycline, ampicillin, chloramphenicol); and bacteria (total aerobic bacteria, coliform bacteria, Escherichia coli, Staphylococcus aureus, Salmonella species, Listeria monocytogens). The laboratory methods were based on Vietnam Food Safety Standards (TCVN 7028:2009) and Association of Official Analytical Chemists International standards for sterilized fresh milk (Horwitz et al., 2005, Vietnam Food Safety Standards, 2009). The Netherlands Organisation for Applied Scientific Research tested the milk for the following allergens: peanuts, egg, fish, soy, shellfish, sulphite, gluten, and histamine.
Statistical analysis
Data were analyzed by Stata 11 software. Matched odds ratio (OR) and 95% confidence intervals (CI) were calculated for the consumption of 30 food items by univariate analysis. We used conditional logistic regression for items with a significantly elevated OR in order to assess the confounding effect of the GOS-supplemented milk.
Results
Descriptive epidemiology
From 9 to 28 October 2009, we identified 19 cases through our investigation at the hospitals; this was four more cases than had been reported to the HCMC Department of Health. One case could not be located, and the parents of another case refused to participate in the study. The cases resided in nine of the 24 districts of HCMC, were aged 2–15 years, with a median of 10 years, and had attended one of the two pediatric hospitals. Figure 1 shows the 17 cases by date of onset of illness.
Number of children (n=17) with allergic reactions by date of onset, October 2009, Ho Chi Minh City, Vietnam.
Four cases had a past history of allergy to various food items, but not to milk. Twelve of 15 cases had onset of symptoms within 3–20 min of drinking milk containing GOS, while onset in the remaining three cases was at 1, 6, and 49 h after consuming the milk, resulting in a median of 5 min. The most frequent symptom was an itchy maculo-papular skin rash (94%), mainly affecting the eyelids, face, and limbs. Eleven cases attended hospital 5–90 min after onset, and the others >90 min after onset, with a maximal delay of 47 h.
Three children had acute breathing difficulties and attended hospital at approximately 5, 20, and 90 min after onset; they were treated with adrenaline, steroids, and salbutamol at the hospital. They did not require assisted ventilation, recovered within 30 min of treatment, and were monitored in hospital for 8 h to 2 days. The remaining 14 cases recovered within 15–30 min after treatment with antihistamines (chlopheramine, loratadine, promethazine), and were discharged within 1 h of arrival. Fifteen cases were given corticosteroids (metasulfobenzoate prednisone, methylprednisolone, hydrocortisone, and prednisone), and seven cases were given a beta-2 adrenergic agonist (salbutamol).
As 16 of the 17 cases had consumed milk containing GOS, the case control study was conducted to assess whether GOS and/or other food items had caused the illness.
Case-control study
The age and sex distribution of the cases and 50 neighborhood controls were similar (Table 1). Of the 30 food/drink items consumed by study participants, the milk supplemented with GOS was highly significant on univariate analysis: OR=34.0 (95% CI 3.9, 294.8) (Table 2). Sixteen of the 17 cases had consumed the milk before the illness. The only case who did not drink this milk product had similar symptoms and signs as the other 16 cases, did not have breathing difficulties, and responded immediately to treatment; he did not have a past history or family history of allergies. The OR for the 100% milk product from the same company B that was not supplemented with GOS was also statistically significant (Table 2). After adjusting the OR of the pure milk product for the confounding effects of the GOS-supplemented milk, the OR was no longer statistically significant (adjusted OR 1.8; 95% CI 0.2, 17.3). Table 2 shows the OR for the milk products only, because the results for all other food items were statistically not significant (not shown).
OR, odds ratio; CI, confidence interval; GOS, galacto-oligosaccharides.
The laboratory tests on the milk containing GOS did not reveal any unusual properties, allergens, chemicals, or other toxic substances.
When the results of this study were confirmed on 28 October 2009, the GOS-containing milk was withdrawn immediately from the market across Vietnam. Surveillance for more cases was continued at the hospitals in HCMC until the end of March 2010, but no new cases were reported.
Discussion
Drinking the fresh milk product supplemented with GOS was strongly associated with the allergic reaction; only one of the 17 symptomatic children did not drink the milk before the illness. Milk samples were not available from any of the case households, and the specific allergen was not identified from the product obtained from shops and supermarkets in the neighbourhood of the cases. However, the strength of association between exposure and outcome, the temporal relationship between the time of drinking the milk and onset of symptoms, and the absence of any further cases once the product was withdrawn from the market implicates the GOS-supplemented milk as the cause of the outbreak. This was the first time the company had ever used GOS at its factory in Vietnam, and GOS was not added to any of its other products.
We were unable to identify the specific allergen in the milk product or to confirm whether the GOS constituent was implicated. It is possible that another constituent such as a new allergen, a side product of the metabolic process, an alien protein absorbed with GOS, or a contaminant from the manufacturing process may have caused the illness.
The major limitation of the study was that it was limited to symptomatic children attending hospitals in HCMC, suggesting that only those with more severe symptoms were likely to have been included. Consequently, the true extent of the outbreak across HCMC and across Vietnam was not determined. We cannot explain why cases were not reported from other towns and cities across Vietnam; it is possible that only batches of the milk distributed in HCMC had been contaminated, or perhaps surveillance in other emergency health services was not as sensitive as in HCMC. However, we could not verify either possibility. To minimize measurement bias, we used a calendar to help parents recall consumption of the milk and used photos of different milk containers to help parents identify the milk product consumed.
GOS is a non–digestible carbohydrate that reaches the colon and promotes the growth of a limited number of indigenous bacteria. We were unable to locate literature reports of allergic reactions to GOS in humans. GOS appears effective in reducing allergy in mice (Schouten et al., 2009). However, its beneficial effects of reducing the incidence of allergy (Kukkonen et al., 2007) as well as preventing allergy in humans (Heine and Tang, 2008; Kuitunen et al., 2009; Prescott and Björkstén, 2007) are debated in the literature (del Giudice and Brunese, 2008). Furthermore, its safety in humans has not been assessed systematically and is based mainly on testing GOS-supplemented products released in the market (Vandenplas, 2007).
Conclusion
To our knowledge, this is the first report of an allergic reaction to fresh milk supplemented with GOS. However, we were unable to identify the specific allergen in the milk product. This study highlights the importance of reminding clinicians to report any unusual health events in their patient population in order to enable prompt public health responses and the collection of suspected food samples for testing. Vietnam's food safety authorities should expand their laboratory capacity for detecting allergens in food products.
Footnotes
Acknowledgments
We thank Health Service of HCMC and staff of the Children's Hospitals for reporting the outbreak and for assisting the investigation. We thank members of the outbreak control team of Institute of Hygiene and Public Health who assisted with data collection. The study was funded by the Ministry of Health.
Disclosure Statement
No competing financial interests exist.