From Jonathan Latham
In discussing why genetics researchers are not finding genes for disease susceptibility, the “dark matter” of genetics, Peter Aldhous and Michael Le Page omitted the simplest explanation of all (19 June, p 30). Maybe they can’t find them because they are not there.
There are some excellent reasons for thinking that genes for modern diseases are rare or absent altogether. For most of the diseases under discussion there is excellent data for environmental causes. Furthermore, the evidence for their genetic causation relies on estimates of heritability based on twin studies.
Not every geneticist thinks twin studies are good science. I heard Martin Bobrow of the University of Cambridge describe estimates of heritability based on them as “poisonous” at his talk entitled The Genetics Revolution at the Wellcome Trust Genome Campus in Hinxton, UK.
Richard Lewontin has criticised heritability measurements in his book, Biology as Ideology: The Doctrine of DNA, for the reason that since gene effects depend on environment, and environmental effects depend on genes, there can be no fixed relationship between genes and environment.
Advertisement
Emmanouil Dermitzakis and Andrew Clark also recently concluded that DNA will not be useful for predicting disease because there is not enough genetic variation in the general population (Science, vol 326, p 239).
From Georg Pedersen
To be able to make medical decisions based on knowing one’s genome might be valuable, but the idea of being able to predict with any accuracy the diseases one would get as one ages is naive and shows ignorance of conclusive research in the field.
The largest epidemiological study ever done – by T. Colin Campbell at Cornell University in Ithaca, New York, and colleagues in Oxford and China, and reported in the book The China Study – found that genetic pre-disposition is quite insignificant compared with what you eat, and what you don’t eat.
In the same book, Campbell shows that work on rats tells the same story. When two groups of rats were given a very potent carcinogenic substance, aflatoxin, and fed on diets with either 5 per cent or 20 per cent protein, the group with the higher protein intake all died of liver cancer, while not a single animal in the other group had any trace of liver cancer, in spite of consuming the aflatoxin. In other words, you can turn on and off cancer by simply varying the amount of protein consumed. Genetic variations and the presence of such a strong carcinogen is insignificant. The book confirms a similar result with a whole range of common diseases in humans.
Sydney, Australia
Ithaca, New York, US
