Linked Pages |
It is very common to find that children who are skin test positive to peanut also have a positive skin test to soy. It has been the prevailing wisdom that the positive skin test to soy is rarely a problem clinically and that has been supported by several clinical studies. Bruno in Businco's group described a study of 748 atopic children, of whom 6% had a positive skin test to soy (45 subjects) but only 16% of them had a positive oral challenge. Abstract Magnolfi et al studied 704 patients with signs and symptoms of allergy and 148 of the 704 patients (21%) had a positive skin test to soy. They performed oral challenge on 131 of those subjects and found a positive in 8 or 6%. Abstract However this prevailing wisdom has recently been questioned by a study from Sweden. Foucard et al.(1999) prospectively studied severe allergic reactions to foods in Sweden and described 4 deaths from soy allergy. They question whether allergy to soy is being underestimated. Abstract Sicherer, Sampson and Burks reviewed the subject of soy allergy and expressed doubt that anaphylaxis to soy is common. They noted that all of the patients described by Foucard et al. were severely allergic to peanut and they questioned whether the deaths could have been due to trace contamination by peanut not detected by the methods used in the Swedish study. Reference
In recent years soy products have have increased in popularity as healthful alternatives to meat that are low in fat.
| I recently saw three patients in a 2 week period in the office who had
potentially threatening reactions to soy.
Originally seen at twenty months of age after she had reacted to peanut. Skin testing showed a significant positive to peanut (15x10mm) with a small positive to soy (4x3mm) and small positives to lentils (3x3mm) and mixed beans (5x5mm), negative to string bean and larger positive to egg (10x10mm). She may have started becoming positive to dust mite. She also had asthma. Later she reacted to pea soup such that after she ate a few spoonfuls she had tickling of the throat and hives on her lip. She stopped eating pea soup. She has avoided peanut and nut and has had no further reactions to those foods. In February or March, she ate several soy cracker crisps (Glenny's Soy Crisps, Regular Flavor), and five minutes later she noticed itching of the throat, a few hives and then she was wheezing with watering eyes. The history is somewhat unclear but apparently it wasn't recognized as being related to the soy crackers. The Epi-pen was not used and she was treated with puffers. A few weeks later she ate a few teaspoons of soy ice cream and noticed itching of the throat and hives on the lips but no wheezing. Since then she has avoided soy. SKIN TESTING: Very strong positive to two preparations of peanut (20x12mm and 18x18mm), small positive to soy solution (5x5mm) and a moderate positive to soy milk (10x10mm). Negative to fresh cow milk, negative to a series of nuts. Positive to cat, dog, tree, grass, ragweed and slightly to the seasonal mold Alternaria. Assessment: She is strongly sensitized to peanut but she was negative to nuts. She had a very significant clinical reaction to soy including wheezing which may have been related to the amount of soy cracker she actually ate. Nevertheless, it was a threatening reaction and the adrenaline should have been administered and she should have been taken to an emergency department. The reactions are not over for four or five hours.
Originally seen at 4 months of age for possible asthma and was skin test negative to milk, egg, wheat, peanut and soy. He was managed with intermittent inhaled steroid. Seen again at 19 months of age after reacting to peanut butter with immediate hives on the face after tasting it a second time. Skin tests were positive to peanut (8x8mm), negative to a series of nuts and soy but positive to dust mite. An EpiPen was taught and prescribed. He avoided peanut and nuts. At age 4.5 years, he was seen again and had had no further reactions to any foods and he had avoided peanut and nuts. He had an epipen. He had no problems with other foods including soy but had never had shellfish. He had nasal symptoms but asthma symptoms only with colds. At that time he used Flovent and Serevent. Skin Testing: Moderately strong positive skin tests to two preparations of peanut (13x10mm). Negative to a series of nuts. Small positive to soy solution (4x4mm). Positive to tree and grass with borderline positive to alternaria and indoor mould. Age 5, he had been eating dry-roasted soy nuts without problem but then refused a vegetarian soy-imitation ground beef patty. He had no reactions to peanut. He drank approximately a half cup of chocolate soy milk (So Good Chocolate soy milk, SoyaWorld, Vancouver, made with Isolated soy proteins ) and may have coughed. Five days later he drank a half cup of the chocolate soy milk and began coughing. The cough continued and he developed nasal congestion, gagging and puffy eyes. The Epipen was not administered but he was taken to an emergency department. He had no wheeze. He recovered uneventfully. On skin testing he had a small positive to soy solution (3x3mm) and a larger positive to tofu (6X6mm) and fresh soy milk (8x8mm). Assessment: His last reaction to the chocolate soy milk included respiratory symptoms and was threatening. The EpiPen should have been administered. He was transported to an emergency department which was appropriate.
He was seen the first time at 15 months of age after he had reacted to egg (fresh egg white 20x15mm) and peanut (7x5mm), but was skin-test negative to nuts. Two years later he was skin-tested again with the following results: peanut (20x15mm), egg white (10x14mm) hazel nut (3x3mm), almond (3x2mm) and brazil nut (4x2mm). At that time he tolerated egg but avoided peanut. He was skin test negative to soy on both testings and he tolerated soy. He drank 2 cups of the same brand of chocolate soy drink eaten by the previous patient (So Good made by SoyaWorld Inc. in Vancouver made with Isolated soy proteins) and said his throat felt funny; he then began salivating, then his lips seemed full and red, his eyes became watery and red, his voice changed, and he developed a croupy cough. His mother who is a physician, administered the EpiPen and called 911. Despite adrenaline he developed hives and itch and some respiratory distress. The paramedics heard a wheeze and gave ventolin. In the emergency his O2 saturation was 92. He was given Benadryl and IV steroids. The reaction settled over 2 - 3 hours. On skin testing he demonstrated a moderate positive to peanut (10x10mm) and SoGood soy milk (8x8mm), positive to soy solution (3x3mm), brazil nut(5x5mm), cashew (10x10mm), pistachio (5x5mm), and walnut (3x3mm). Assessment: This patient had respiratory symptoms and a low oxygen measured in the emergency department despite the administration of adrenaline. The reaction was threatening and the outcome could have been worse if not for the early use of adrenaline. In each of these cases a relatively large amount of soy product was consumed perhaps creating the conditions for a threatening reaction. However the children are older than one would expect to have developed sensitivity to a food. More commonly children develop sensitivity to foods at less than 3 years of age. The 10 year-old girl was known to be sensitized to peanut and other legumes and therefore would have been at risk for sensitivity to soy. However the boys were not previously known to be sensitized to any legumes other than peanut and did not initially show positive skin tests to soy. |
Discussion:
In population studies, allergic reactions to soy are reported to be infrequent. The prospective study by Halperin et al.( Halpern SR, Sellars WA, Johnson RB, Anderson RB, Saperstein S, Reisch JS. Development of childhood allergy in infants fed breast, soy, or cow milk. Allergy Clin Immunol. 1973;51:139-151) indicated allergic responses to soy in 0.5% of infants and to cow milk in 1.8%.
Young et al. did a population survey of perceived allergy to 8 foods in Britain. Roughly 20% of the population felt they had allergy to food. Oral challenge was performed on 93 subjects of whom 19% demonstrated a positive reaction. This led to an estimated prevalence of reaction to food of about 1.4 to 1.8% depending on criteria chosen while soy was the least prevalent at 0.3%. Abstract
Generally when studies have been published examining the nature of reactions to foods in children, soy has been very infrequently the causative agent.
Businco and colleagues have studied soy allergy and published a review of the topic in 1998. Abstract of Businco, Bruno, Giampetro Review which is available in full Businco review as PDF file
They cited the following studies which used blinded challenge with soy and which suggested that "clinical relevant" soy allergy is infrequent.
Businco quotes only 2 cases of anaphylaxis to soy described in the literature.
In 1998, the American Academy of Pediatrics issued a policy statement on the topic of "Soy Protein-based Formulas: Recommendations for Use in Infant Feeding (RE9806)". "Go to Statement. The Academy reviewed the literature available at that time and developed conclusions and recommendations. They suggested that while soy protein was antigenic, it was not very allergenic and cited only one published report of a "true" anaphylactic reaction to soy, the case report of Mortimer, published in J Pediatriccs, 1961.
In a study that I published on peanut sensitivity, 96 children who were skin test positive to peanut returned several years later for re-evaluation. Twenty-six of them had 41 reactions to food since I first saw them (7 to peanut) Go to Article. Some had reacted to multiple food types but none had reacted to soy although approximately a third of them had a small positive skin test to soy. Similarly I have reviewed patients that I saw in 1998 and could not find any reactions to soy nor any anaphylactic reactions to soy in the year 2000.
It is still possible that these three cases were a chance clustering and not really an indication of a new phenomenon. It is also possible that the reactions were to a contaminant in the soy proteins such as peanut. In soy protein isolate, 90% of the pulp-derived protein resides in two major heat-stable globulins: ß-conglycin, with a molecular weight of 180000, and glycinin, with a molecular weight of 320000. The former has three subunits and the latter has six. After enteric digestion, other antigens or epitopes might be exposed. In the last few years there has been a change in the method of processing soy protein using a water-based technique to remove sugar instead of alcohol. The water extraction results in greater retention of biologically active molecules such as the isoflavones (the isoflavones found in soy are genistein, daidzein and glycitein) which are responsible for the phytoestrogen activity of soy. I could find no indication in the literature that isoflavones are allergenic since they are not proteins but sterol ring structures. However changes in manufacturing technique to preserve the potential medically beneficial effects of soy proteins could have led to a difference in the protein extracted and perhaps promotion of allergenicity not seen previously. The alcohol extraction might have removed or denatured some allergens that survive the water extraction. This hypothesis remains to be tested.
On a practical basis, it is difficult to provide advice on avoidance of soy to the families of the children involved. Soy protein can be found in many manufactured foods and it is not clear as to the degree of risk to the children from small amounts of soy found in such foods. For now the recommendations are that avoidance must be as strict as for peanut and the EpiPen must be carried on the child.