GENET archive


3-Food: Pusztai report of the Royal Society (UK)

mailinglist genet-news
-------------------------- GENET-news ---------------------------

TITLE:  Review of data on possible toxicity ofGM potatoes
SOURCE: The Royal Society, UK
DATE:   May 19, 1999

----------------- archive: ------------------


The Royal Society published a review of what was known
scientifically about the suitability of GM plants for food use in
September 1998. Because of the current controversy, we are
looking again at several issues, and in particular we have
reviewed all available data related to work at the Rowett
Research Institute on the possible toxicity of genetically
modified potatoes. Our main conclusions are as follows.

1. The safety of GM plants is an important and complex area of
scientific research and demands rigorous standards. However, on
the basis of the information available to us, it appears that the
reported work from the Rowett is flawed in many aspects of
design, execution and analysis and that no conclusions should be
drawn from it.

2. We found no convincing evidence of adverse effects from GM
potatoes. Where the data seemed to show slight differences
between rats fed predominantly on GM and on non-GM potatoes, the
differences were uninterpretable because of the technical
limitations of the experiments and the incorrect use of
statistical tests.

3. The work concerned one particular species of animal, when fed
with one particular product modified by the insertion of one
particular gene by one particular method. However skilfully the
experiments were done, it would be unjustifiable to draw from
them general conclusions about whether genetically modified foods
are harmful to human beings or not. Each GM food must be assessed

4. The whole episode underlines how important it is that research
scientists should expose new research results to others able to
offer informed criticism before releasing them into the public


In April 1999 the Royal Society convened a Working Group to
examine whether newly publicised research required changes to our
September 1998 statement GM plants for food use. This report
deals with the apparent evidence that genetically modified
potatoes adversely affected the health and growth of rats. The
report has been endorsed by the Council of the Society, and was
prepared by a group chaired by Professor Noreen Murray FRS, FRSE
(University of Edinburgh). The other members were Professor Brian
Heap FRS (Foreign Secretary and Vice-President of the Royal
Society), Professor William Hill FRS, FRSE (University of
Edinburgh), Dr Jim Smith FRS (National Institute for Medical
Research), Professor Michael Waterfield FRS (Ludwig Institute for
Cancer Research and University College London) and Dr Rebecca
Bowden (Secretary).


We sought information from all possible sources about the work at
the Rowett Research Institute and obtained the following:

- Report of project coordinator, Arpad Pusztai FRSE, on data
  produced at the Rowett Research Institute (RRI), 22.10.98
- Audit of data produced at the Rowett Research Institute,
  21.8.98, SOAEFD (Scottish Office Agriculture, Environment and
  Fisheries Department) flexible fund project RO818
- Response to request for clarification from the Rowett Research
- Response to request for information from Dr Susan Bardocz
- Information provided by Advisory Committee on Novel Foods and
  Processes (including correspondence with Dr Stanley Ewen)
- Statistical analysis carried out by Biomathematics and
  Statistics Scotland of experiments on genetically modified
  potatoes conducted at the Rowett Research Institute, 1.3.99
- Minutes of evidence - Science and Technology Committee of the
  House of Commons, 8.3.99
- Response to request for information from Dr Vyvyan Howard of
  Liverpool University
- Response to request for information from Mr G Horgan of BIOSS
- Response to request for information from Professor John
  Gatehouse of Durham University
- Response to request for information from Professor Michael
  Wilson, Scottish Crop Research Institute
- Information from SOAEFD on rationale for research

Dr Pusztai indicated to us that further information existed, but
did not provide it.

We sent the available information to six independent, impartial
reviewers whose expertise included statistics, clinical trials,
physiology, nutrition, quantitative genetics, growth and
development, and immunology. Their remit was to examine the data
that we gave them and advise us, in the usual way of referees,
about the scientific merit of the work described. They were not
asked to give an opinion on the actions of any individuals
involved in this work.

This report is based on the responses from the six reviewers. The
responses were copied to Dr Arpad Pusztai, who was given the
opportunity to comment. In accordance with normal scientific
practice, the reviewers remained anonymous.

Dr Pusztai suggested to us that his reports were internal
Institute documents and that it was therefore not appropriate to
peer review them.

However, since they were released into the public domain, both
through the media and on the Internet, it seems to us entirely
appropriate that they should also be subjected to expert
scientific scrutiny - all the more so because of the importance
of this area.

We now examine two specific claims that have been made as a
result of the Rowett work.


The experiments set out to investigate whether potatoes
genetically modified to contain a lectin gene from a snowdrop
affected the development of organs or the metabolism of the rats
to which they were fed, in the short (10 days) or long (110 days)
term. The structure of the experiments was changed as they
progressed, which made comparisons between rats fed on modified
potatoes and those fed on unmodified potatoes more difficult.
Some results showed differences in the overall body weights and
in the weights of individual organs in the two groups of rats.
However, such results as were statistically significant did not
fall into a readily discernible pattern.

A particular difficulty is that the experiments were not well
designed. For example, there is very little information about how
the GM and control diets differed in their detailed composition,
and in particular about differences other than those attributable
to the inserted gene. These differences should have been fully
analysed, and addressed by using several distinct strains of GM
potato in the feeding trials. Second, the GM potatoes used
contained almost 20% less protein than unmodified potatoes.
Therefore, in the long-term feeding study, rats being given GM
potatoes were also given additional protein to meet Home Office
requirements intended to avoid starvation: observed effects could
have been caused by this supplementary diet being inadequate or
incomplete. Third, when a rat is underfed many organs are likely
to be affected, so that separate measurements on the same
specimen will turn out to be interrelated.

An added deficiency of the study was that, as far as we can tell,
the measurements were not conducted 'blind' as is normal practice
for trials of this kind (a protocol in which the scientists
making the measurements are not aware of how the animals have
been treated). Unconscious bias is well known to be a source of
invalid results.

Because of the poor experimental design, it is simply not
possible to be sure about the causes of the small effects
obtained in the study.


It had been claimed that the consumption of GM potatoes had
significant effects on the immune system of rats in the feeding
trials, because of some effect of the genetic modification itself
rather than because of the particular gene inserted. Reviewers
were asked to consider if this conclusion was valid.

One of the immune reactions in the body can be mimicked in the
laboratory. Antigen-reactive white blood cells (lymphocytes) are
transformed into dividing lymphoblasts when they are stimulated
by specific antigens. This provides a means of testing for
effects on the immune system.

The claim that the Rowett data show evidence that the inserted
gene had a different immune effect from simply adding lectins to
unmodified potatoes does not stand up. Inappropriate statistical
tests had been applied to the data and, when the appropriate
comparisons are made , there are no interpretable differences.
Moreover, the experimental data in this area are beset by the
same sources of inconsistency noted in section 3 above.


The work on feeding trials with GM and non-GM potatoes attempted
to cover too much ground with the resources available.

In the form currently available, the data reviewed provide no
reliable or convincing evidence of adverse (or beneficial)
effects, either of lectins added to unmodified potatoes or of
potatoes genetically modified to contain a lectin gene, on the
growth of rats or on their immunological function.

In summary, the data presented to the reviewers and Working Group
are inadequate for the following reasons:
- poor experimental design, possibly exacerbated by lack of
  'blind' measurements resulting in unintentionally biased
- uncertainty about the differences in chemical composition
  between strains of non-GM and GM potatoes;
- possible dietary differences due to non-systematic dietary
  enrichment to meet Home Office and other requirements;
- the small sample numbers used in experiments testing several
  diets (all of which were non-standard diets for the animals
  used) and which resulted in multiple comparisons;
- application of inappropriate statistical techniques in the
  analysis of results;
- lack of consistency of findings within and between experiments.

The uncertainty and ambiguity of the data urge great caution in
the interpretation of the results presented. A much improved
experimental design, with stringent controls, would have been
needed if the claims made for the study were to be convincing.
Even if the results of the particular study had supported the
claims that have been made for them, it would have been unwise to
use them for making statements about the safety or otherwise of
all GM foods.

Although we have no evidence of harmful effects from genetic
modification, this of course does not mean that harmful effects
can be categorically ruled out. This issue can be resolved only
by the necessary research carried out to a high standard and by
full use of the regulatory mechanisms for dealing with safety of


Reviewers were asked to give advice on the necessity of future
research arising from the data presented.

The only way to clarify the current situation would be to refine
the experimental design of the research done to date and to use
this as the basis for further studies in which clearly defined
hypotheses were tested, focused on the specific differences
already claimed. It would be necessary to carry out a large
number of extremely complex tests on many different strains of GM
and non-GM potatoes. It would be important to ensure that these
studies had sufficient statistical power (in the sense that
numbers in each experimental group were sufficient to deal with
the variability in individual response) to come to a clear
conclusion. It would also be important to take adequate account
of the age and the susceptibility of the animals and the
wholesomeness, completeness and adequacy of the entire diet.
Careful thought would have to be given to the specific targets
for any hypothesised damage.


In view of the public interest in this case we recommend that the
results of any future studies on testing GM food safety, when
completed, should be peer reviewed and then published. This would
provide an opportunity for the international scientific community
and the public at large to have access to the information.

The Royal Society has recommended in its September 1998 statement
GM plants for food use that any over-arching body analyse the
current regulations, giving particular consideration to whether
long-term animal feeding studies are necessary to provide greater
information on allergenicity or toxicity. We now reiterate this


-| Hartmut Meyer
-| Co-ordinator
-| The European NGO Network on Genetic Engineering
-| Reinhaeuser Landstr. 51
-| D - 37083 Goettingen
-| Germany
-| phone: #49-551-7700027
-| fax  : #49-551-7701672
-| email:

Genet News