On the 4th September 2011 Professor Derek Knottenbelt had a letter published in the Yorkshire Post newspaper. The following is an analysis of that letter.

He makes two statements about the occurrence of ragwort poisoning and the statistical basis for this but before analysing them it is important that we know what "Statistics" are.

The Oxford English Dictionary is accepted as the definitive guide to the English language.
It defines statistics as follows:-

"The systematic collection and arrangement of numerical facts or data of any kind; (also) the branch of science or mathematics concerned with the analysis and interpretation of numerical data and appropriate ways of gathering such data

The two statements by the professor are:-
"It is quite true to say that we simply do not know how many horses die as result of the toxicity but I have always erred on the side of caution when calculating the figures but for sure it is not less than 1,000 horses per year in the UK.


"In my own clinic it is more than 10 per year out of a hospital population of 2,500. There are about a million horses in UK – the maths is easy."

As a simple matter of calculation the maths is easy, however in terms of statistics being the branch of maths concerned with the interpretation and appropriate ways of gathering data, it would appear that the maths is wrong.

It is probably best explained by means of an analogy. Imagine a survey is done to find out how many people are religious in the country. How would you do it? One way would be to stand outside a church on a Sunday asking the people who come out about how religious they are. Most people would see that this would not be a good way because you would be choosing a biased sample. People who go to church are more religious than people generally which is why they go to church. Surveying church goers would give a falsely high figure. Churchgoers would not be what is known in statistics as a "representative sample". Churchgoers would be a biased sample.

In the text book "Research Methods and Statistics: A Critical Thinking Approach" author Sherri L. Jacks describes this essential part of the mathematics of statistics as follows:-

"The population consists of all of the people about whom a study is meant to generalize , whereas the sample represents the subset of people from the population who actually participate in the study. In almost all cases, it isn't feasible to survey the entire population. Instead, we select a subgroup or sample from the population and give the survey to them. To draw any reliable and valid conclusions concerning the population, it is imperative that the simple be"like" the population—a representative sample. When the sample is representative of the population, we can be fairly confident that the results we find based on the sample also hold for the population. In other words, we can generalize from the sample to the population.

Likewise if the sample is not representative you cannot generalise from it

The same thing applies for sick horses. The horses being treated at the hospital because they are sick. You would get an artificially high number in a veterinary hospital in comparison to the general population so you cannot generalise from them and use them as a representative sample to scale up.

The puzzling thing here is that Professor Knottenbelt says, about cases of ragwort poisoning that he gets "more than 10 a year" but when the hospital was asked in a request under the Freedom of Information Act. They reported that they had recorded no cases at all for a period of five years .

Professor Knottenbelt continues with another questionable claim.

"The (PA) toxins in ragwort exert their effects by altering DNA in the cells. DNA cross-linking is a known cause of cell destruction and a potential cause of tumors and in one circumstance in Africa where ragwort infusions were fed to humans as herbal remedies, the toll was very heavy.

A very high number of cases of hepatic carcinoma were reported as well as extensive veno-occlusive disease and liver failure.
Later in the letter he says.

"In South Africa the plant is causing massive concerns."
And he was mentioned in the Western Mail news paper in 2006 as saying about the toxins, " In South Africa they had been found in bread, possibly through ragwort growing unseen among crops and being included in the flour."

The European Food Safety Authority (EFSA 2007) provide a comprehensive list of PA ( Pyrrolizidine Alkaloid )poisoning in the world. There are several cases reported from Africa but none of them involve the plant in question, what we in the UK know of as Common Ragwort In fact it doesn't seem that our ragwort actually occurs anywhere in Africa except possibly in the area north of the Sahara Desert

As for South Africa the plant does not seem to grow there at all. The South African National Biodiversity Institute which includes their National Botanic Garden says there are no specimens of this species in their herbarium apart from a few from Europe and it appears to be unknown there.

An extensive search of the literature shows that the source for the Bread poisoning claim would appear be a paper from nearly a hundred years ago. (Willmot and Robinson 1920)

They say:-

"It was found that the plants known as Senecio ilicifolius. and S. burchelli grow as weeds in the wheat fields at George, and that when the wheat is threshed the seeds and portions of these plants frequently remain behind and are sold with the wheat"

These plants are not common ragwort. So it appears that the professor is claiming that common ragwort may poison people in South Africa, when the plant does not grow there and that reported poisoning was from antiquated agricultural methods nearly a century ago involving different species of plants.

Professor Knottenbelt also states in the letter:-

"cereals contaminated with ragwort in Uzbekistan were responsible for large numbers of cases of severe liver disease"

Again an extensive search of the scientific literature shows that the only documented cases of PA toxins contaminating cereals and affecting people in Uzbekistan were from plants unrelated to ragwort in a different plant family Trichodesma incanum and Heliotropium lasiocarpum which produce different PAs.

At this point it is worth mentioning that there is a report by the European Food Safety Authority (EFSA 2007) written by a panel of world leading experts and their conclusion :-

"Accidental ingestion of PAs is a problem in some under developed countries and several reports have described outbreaks of PA intoxications in the past. The main source of these intoxications is grain , contaminated with seeds of PA producing plants"

The UK is not an under developed country therefore it would appear questionable to use these points to inform people who live there. It appears that the professor is making scary claims that are not substantiated by the facts.

It is worth pointing out at this point that the PA toxins , which occur in around 6000 plants worldwide, are not in themselves directly toxic but do have to undergo both and absorption and conversion process. This is not going to ever be 100% effective and there are mechanisms that reverse DNA cross-linking within cells.

EFSA (European Food Safety Authority). 2007. Opinion of the Scientific Panel on contaminants in the food chain on a request from the European Commission related to Pyrrolizidine alkaloids as undesirable substances in animal feed.
Senecio Disease, or Cirrhosis of the Liver, Due to Senecio Poisoning. WILLMOT, Frederick C. ; ROBERTSON, George W. SOUTH AFRICAN MEDICAL RECORD 1920, Sept. 25 Vol. 18 No. 18 pp. 346-348 pp.