LETTER TO THE EDITOR
Low frequency of soya allergy in peanut-allergic children:
Relevance to allergen labelling on medicines
To the Editor:
There is extensive in vitro IgE cross-reactivity between peanut and
however, the relevance to clinical reactivity is
unclear: while the majority of soya-allergic children also have peanut
the reverse does not appear to be true.
Soya-derived products, including refined soya oil and lecithin, are
common excipients in many medicines and nutritional supplements.
The European Medicines Agency (EMA) currently advises that all
medicines containing soya-derived excipients should be labelled as
contraindicated in peanut-allergic individuals.
This was following a
Swedish report of 4 deaths in peanut-allergic children attributed to
inadvertent high-level soya consumption in children previously toler-
ant of soya, although it has been suggested that these fatalities
might have in fact been due to peanut contamination.
Of note, soya
is an uncommon trigger for fatal anaphylaxis, and there have been
no further deaths reported due to soya in Sweden since the original
We therefore sought to determine the rate of soya allergy in
children (age 8-16 years) undergoing double-blind, placebo-con-
trolled food challenges (DBPCFC) to peanut (BOPI study, Clinical Tri-
als.gov identifier NCT02149719). The challenge matrix used to blind
peanut was soya-based; therefore, all participants underwent an
open food challenge to roasted soya (Wowbutter; Hilton Whole
Grain Millers Ltd, Canada) to a cumulative 4.1 g soya protein at
screening, prior to DBPCFC to peanut. Participants experiencing
symptoms at the open soya challenge underwent a DBPCFC to
roasted soya, to confirm clinical reactivity. The study was approved
by the NHS Human Research Authority (reference 15/LO/0287),
and written informed consent/assent was obtained. DBPCFC were
conducted according international PRACTALL consensus criteria.
Further details are available in the Online Supporting Information.
Sixty-eight children, including 27 (41%) with a history of anaphy-
laxis to peanut, underwent DBPCFC between June 2015 and Febru-
ary 2017. No participants were actively avoiding soya-containing
foods, and 31 (47%) gave a history of regular dietary soya intake.
Baseline characteristics are shown in Table 1. Sixty-four reacted to
peanut at DBPCFC, including 15 (22%) with anaphylaxis. Two (3%)
of the 68 children experienced objective symptoms to roasted soya
(Wowbutter) at open challenge, both of whom also reacted at
DBPCFC to soya (experiencing mild objective symptoms at cumula-
tive 1 g soya protein) (further details in Table S1). Interestingly, one
child had a history of tolerance to soya-containing products, which
was subsequently confirmed at open challenge to unroasted soya
(cumulative 4.5 g protein), demonstrating reactivity to roasted but
not unroasted soya (roasting is known to increase the allergenicity
of peanut, but has not been previously reported with respect to
Therefore, in this cohort of young people with challenge-proven
peanut allergy, only 3% had concomitant soya allergy: the vast
majority were not soya-allergic. Although this was a secondary
finding in a study powered to assess an intervention to induce
desensitization in peanut-allergic children, we note the 95% confi-
dence interval for concomitant soya allergy in this cohort (0.4%-
10.8%) is entirely consistent with the existing literature (see
These historic data may not be representative of the
wider peanut-allergic population, as they were performed in indi-
viduals with coexisting atopic eczema and not all cases were chal-
lenge-proven. Moreover, these studies were limited by small cohort
size and were mostly conducted in the United States where the
pattern of sensitization may be different to that in Europe.
trast, we included only individuals with challenge-proven peanut
allergy, and all allergic reactions to both peanut and soya were con-
firmed by DBPCFC. Our participants may represent a more allergic
cohort than the wider population; therefore, it is unlikely that we
have underestimated the rate of concomitant allergy to soya in
Soya is frequently used in food production and must be declared
as an allergen unless in the form of highly refined oil (as is often the
case when used in medicines). Refined soya oil contains very low
levels of soya protein (0.32 lg/mL in one report), and little IgE-bind-
ing activity is seen to these proteins in vitro.
At these levels, soya-
allergic individuals would be expected to need to consume very large
volumes (over 1 L) to experience symptoms, although there is a
report in the literature of 2 soya-allergic children reacting to 16 and
30 mL refined soya oil (of unknown protein content).
Soya lecithin (an emulsifying agent, produced from crude soya
oil) contains higher levels of soya protein than refined oil. The EMA
issued its directive against peanut-allergic individuals using medicines
containing soya lecithin in 2003, due to concerns over the very wide
range for protein content in soya lecithin (28-2700 mg protein/kg)—
despite the fact that soya lecithin is usually a “minor ingredient” in
foods and medicines.
However, it is noteworthy that there is a pau-
city of reports of allergic reactions due to foods containing soya
lecithin in the literature.
Both the European Food Safety Authority and FDA have granted
exemptions to the effect that fully refined soya oil does not to be
declared when present in foods, on the basis that the allergen/protein
has been removed and therefore does not pose a risk to the allergic
© 2018 EAACI and John Wiley and Sons A/S.
Published by John Wiley and Sons Ltd.
wileyonlinelibrary.com/journal/all Allergy. 2018;73:1348–1350.