GENTECH archive


The Failings of Substantial Equivalence


    The concept of substantial equivalence has been used in Europe, North
America, and elsewhere around the world as the basis of regulations
designed to facilitate the rapid commercialization of genetically
engineered foods. For instance, the European Commission (EC) regulations
concerning novel foods and food ingredients apply the concept of
substantial equivalence to both the safety testing and to the labeling of
genetically engineered foods. Genetically engineered foods classified as
substantially equivalent are spared from extensive safety testing on the
assumption that they are no more dangerous than the corresponding
non-genetically engineered food. (1) Using similar arguments, genetically
engineered foods classified as substantially equivalent are not required to
be labeled as genetically engineered. (2) The effect of these regulations
has been to allow genetically engineered foods to enter the marketplace
without sufficient testing to assure safety and without sufficient labeling
to allow consumers to decide for themselves whether or not to purchase and
eat these novel foods. The health of the population of Europe is thus being
placed at risk.

    The fundamental inadequacies of this approach have been discussed
previously. For instance, one article presented in the Proceedings of the
Organization for Economic Cooperation and Development (OECD) Workshop on
Food Safety Evaluation (3), came to the following conclusions:

1. Because the concept of substantial equivalence has no dimensions, it
cannot be use as a predictor of which novel foods will require substantial
safety testing in animals.
2. Depending on the nature of the novel food, the usefulness of the concept
of substantial equivalence in determining the necessity for extensive
safety testing ranges from useful to negligible.
3. The number and range of safety tests required is best determined not by
the concept of substantial equivalence, but by the nature of the product
under consideration.

    At first glance the term substantially equivalent implies that two
foods are equivalent in all characteristics that are of importance to the
consumer - safety, nutrition, flavor, and texture. However, in actual
practice the investigator compares only selected characteristics of the
genetically engineered food to those of its non-genetically engineered
counterpart. If that relatively restricted set of characteristics is not
found to be significantly different in these two, the genetically
engineered food is classified as substantially equivalent to the
corresponding non-genetically engineered food and is required to be neither
tested further nor labeled as genetically engineered.

    The argument supporting this practice is that since most of the
characteristics of a particular genetically engineered food are similar to
those of its non-genetically engineered counterpart, it must be the case
that the genetically engineered food is substantially equivalent to its
non-genetically engineered counterpart with respect to all characteristics
relevant to the consumer. This is obviously a fallacious argument, and
should not be used as the basis for avoiding more extensive testing and for
avoiding the labeling of genetically engineered foods. Most critically, if
characteristics important to food safety are not evaluated directly, the
safety of consumers will be in jeopardy.

    Inadequate testing

    Any claim of substantial equivalence is only as good as the series of
tests upon which that claim is based. In practical terms, if a genetically
engineered food is different from its non-genetically engineered
counterpart, that difference will be detected only if a test is carried out
that is capable of measuring the specific characteristic which is different
between the two. Therefore, if the tests prescribed for determining
substantial equivalence do not include one or more tests capable of
quantitating the characteristic which happens to be different in the
genetically engineered food compared to its non-genetically engineered
counterpart, the genetically engineered food will be wrongly classified as
substantially equivalent to its non-genetically engineered counterpart.

   Currently, the testing procedures required in Europe, North America and
elsewhere consist almost exclusively of specific chemical and biochemical
analytical procedures designed to quantitate a specific nutrient or a
specific toxin or allergen. These tests focus on specific components of a
food that are suspected to be altered in that particular genetically
engineered food, based on the known characteristics of its non-genetically
engineered counterpart, and based on the known characteristics of the genes
introduced into that organism. For instance, in its assessment of Roundup
Ready soybeans, Monsanto quantitated a few of the allergenic proteins known
to be normally produced in soybeans, showing that genetic manipulations had
not accidentally caused Roundup Ready soybeans to produce higher than
normal levels of those allergens.

    Unpredicted side effects

    Important as these studies are, however, they fail to even begin to
assess one very substantial class of risks that are inherent in genetically
engineered foods. That class of risks consists of health hazards resulting
from the unanticipated side-effects of genetic engineering. Such testing
schemes are completely incapable of detecting unsuspected or unanticipated
health risks that are generated by the process of genetic engineering

    It is a scientific fact that the process of genetic engineering often
gives rise to unanticipated side-effects. These can and have been shown to
introduce unforeseen allergens and toxins into foods and unexpectedly
reduce nutritional value. Not every genetically engineered food will have
these problems, but there is a finite probability that any given genetic
modification will lead to unanticipated side-effects that result in food
characteristics that threaten the health of consumers.

    For instance, in 1989 Showa Denko K.K. marketed tryptophan that had
been produced in genetically engineered bacteria as a nutritional
supplement in the USA. (4) When this product was placed on the market, it
made thousands of consumers ill. Of these, 1500 were permanently disabled
and 37 died. Analysis by high pressure liquid chromatography indicated that
this product was more than 99.6% pure tryptophan. (5) However, these
preparations also contained traces of a highly toxic contaminant. This
toxin accounted for less than 0.01% of the total mass of the product but
this was sufficient to seriously threaten health. (6)

   According to the measurements made, the genetically engineered
tryptophan was equal in purity, and thus substantially equivalent, to
previous preparations that had been produced using natural bacteria.
However it was clearly not substantially equivalent with regard to human
safety. If other tests had been required, such as animal or human feeding
tests, which are capable of screening broadly for harmful substances, the
fact that this material was not substantially equivalent would have been
obvious. However, those tests were not done.

    Health risks in derivatives

    Another example of how the concept of substantial equivalence can lead
to abuses is the claim that is commonly made that corn oil from genetically
engineered corn need not be labeled as genetically engineered because the
process of oil production separates the oil from all potentially toxic or
allergenic constituents of corn and that the composition of the oil itself
is identical to that obtained from non-genetically engineered corn. Similar
arguments have been used to justify the deregulation of oil from
genetically engineered soybeans.

    The problems with these arguments are two: First, corn oil is not
chemically pure. It is well known that corn oil still contains sufficient
corn proteins to elicit allergic reactions in individuals who are highly
allergic to corn. Therefore it is highly likely that it will also contain
small amounts of the genetically engineered proteins present in genetically
engineered corn. An individual who is allergic to these proteins would be
likely to react negatively to oil derived from genetically engineered corn.
Second, only the major constituents of genetically engineered corn oil have
been examined in assessing substantial equivalence. However, some of the
minor constituents of this oil, which were ignored in this assessment,
could be of substantial significance to the nutritional value or the safety
of this product. For instance, genetic manipulations could unexpectedly
alter oil metabolism by a number of mechanisms, generating a toxic fatty
acid derivative. Thus, this claim of substantial equivalence is superficial
and should not be used as an argument to justify avoidance of further
testing and labeling.

    Clinical tests needed

    Given that genetic engineering can introduce unexpected health hazards
into foods, it is logical that every genetically engineered food should be
subjected to tests that are capable of detecting a wide range of unforeseen
health threats. Yet, at present, the liberal use of the concept of
substantial equivalence makes it possible to avoid such testing.

   What additional tests are required? Tests are needed that are capable of
screening for a wide range of diverse allergens and toxins. It is not
possible within the scope of this document to discuss in detail the
deficiencies in the current system and the measures required to rectify
this situation. (This subject is discussed in "Assessing the Safety of
Genetically Engineered Foods, a Science-Based, Precautionary Approach" by
the author, which can be obtained by sending an email to requesting this document.) In short, what is
missing in current testing programs is clinical tests in which humans are
fed the genetically engineered food in question both short-term and
long-term. Human tests are of primary importance because animals are poor
models for assessing the human health impacts of foods. In particular,
animal tests provide virtually no useful information regarding the
allergenicity of food to humans.

    Only clinical tests have the broad specificity and relevance to human
physiology needed to detect the wide range of allergens and toxins that
might result from unexpected side-effects of the genetic engineering
process. Without such tests, the full range of allergens and toxins that
can be introduced via the process of genetic engineering cannot be
detected, and without such tests, it is impossible to assure that a given
genetically engineered food is in fact free from health-damaging

    Need for labeling

    Even if more stringent testing is implemented, it is essential that
genetically engineered foods be labeled as genetically engineered. No
testing regime can ever be exhaustive. Therefore, some residual risk of
undetected health-damaging characteristics will always remain with foods
that have been produced using a technique, such as genetic engineering,
that is capable of introducing into a food a wide range of unexpected
side-effects. For instance, if clinical experiments are carried out for 3
years, longer term health effects may be overlooked that take 5 or 10 years
to manifest. Invariably, residual risk remains regardless of the tests
carried out and regardless of the testing period chosen. Labeling foods as
genetically engineered allows consumers to choose for themselves whether or
not to accept this residual risk.

   Industry has stiffly opposed proposals that would have required
genetically engineered foods to undergo clinical testing similar to that
which is standard for novel food additives. The expense and time required
for testing is perceived as a hindrance to commercialization of genetically
engineered foods. However, in the long run, more rigorous testing will be
good, not only for consumers, but also for industry.

    Without such testing some genetically engineered foods that seriously
damage the health of consumers will enter the market. Thus, this
short-sighted approach to safety assessment clearly favors commercial
interests while placing the health of the entire population at risk. Not
only does this abrogate scientific responsibility and basic humanitarian
values, but it is also bad business because it will inevitably lead to loss
of consumer confidence in genetically engineered foods.

1.Regulation EC /95 of the European Parliament and of the Council
Concerning Novel Foods and Food Ingredients, Article 3.4.
2.Regulation EC /95 of the European Parliament and of the Council
Concerning Novel Foods and Food Ingredients, Article 8.1.
3.OECD, DSTI/STP (95)18, Paris, 1995, pages 79-87.
4.Does Medical Mystery Threaten Biotech? Science, Page 619, 2 November 1990.
5.An Investigation of the Cause of the Eosinophilia-Myalgia Syndrome
Associated with Tryptophan Use, New England Journal of Medicine, 323:
357-365, 1990.
6.EMS and Tryptophan Production: A Cautionary Tale, TIBTECH,12:346-352, 1994.

John Fagan, PhD, has more than 23 years experience using cutting-edge
molecular genetic techniques in cancer research. After receiving more than
$2.5 million in grants from the National Institutes of Health (USA) to
support research to identify cancer susceptibility genes, in 1994 took Dr.
Fagan an ethical stand against germ-line genetic engineering and renounced
$1.8 million in further grants. Over the last year, Dr. Fagan has travelled
extensively throughout North America, Europe, and Asia speaking on the
hazards of genetic engineering and genetically engineered foods.