When food goes wrong

Chemicals in food: natural and artificial
In consumer surveys of worries about food safety, people frequently voice concerns about ‘chemicals in their food’. One international food outlet claims to ‘shun obscure chemicals’ in its food. This is a pretty staggering claim, since food is nothing but chemicals, some of them quite obscure.
Of course, the worries are about the chemicals that are added during production or processing of food, not the chemicals that are there ‘naturally’. This is because of a fallacious belief that ‘natural is good, artificial is bad’. However, many of the plants we eat regularly contain small amounts of natural poisons or toxins that could cause cancer. As we saw in Chapters 1 and 2, these are the plant’s natural defences against insects, fungi, and other enemies. The American toxicologist Bruce Ames in a famous paper entitled ‘Dietary pesticides (99.99 per cent all natural)’, reported that 99.99 per cent of the ‘pesticides’ in the American diet are naturally produced defences of plants and that about half of these natural chemicals are carcinogens at high doses in rodents. Ames’s findings have been summarized by alluding to the fact that a single cup of coffee contains natural carcinogens equivalent, in toxicity terms, to a year’s worth of carcinogens from pesticide residues consumed in food.
But, and this is an important but, these results do not mean that people should stop drinking coffee or eating vegetables and fruit. An early 16th-century Swiss physician called Philippus Aureolus Theophrastus Bombastus von Hohenheim, also known as Paracelsus, is generally recognized as the father of the science of toxicology. His key contribution was summarized in a sentence: ‘All things are poison, and nothing is without poison; only the dose permits something not to be poisonous.’ In other words, something that is harmless at a low dose may be dangerous if a lot of it is consumed, and vice versa. Even pure water, if drunk to excess, can be fatal, and although there may be many potential carcinogens in food, both natural and synthetic, whether or not they present a real danger depends on the dose.
Just to illustrate the point, oranges are sometimes treated after harvesting with a pesticide called imazalil, to prevent mould forming on the peel. Imazalil could in theory be a carcinogen. Based on experiments on rats, a potentially dangerous dose for humans would require eating over 12,000 oranges including the peel. But oranges also contain about 70 mg of vitamin C per fruit, close to the recommended daily intake of this essential micronutrient. But as with imazalil, too much vitamin C is harmful. Extrapolating from rats, the number of oranges that could deliver a fatal dose of vitamin C is about 8,000. In short, vitamin C poses a bigger risk than imazalil.
In many, if not most, countries there are strict regulations governing the levels of pesticides and other synthetic chemicals, such as colours and flavours, in food. Safety is assessed by experts, using evidence from experimental tests and other information such as the chemical properties and structure of the substance. The starting point is often experimental data on the toxicity of the chemical, or chemically related substances, in laboratory tests on animals such as rats or mice. The critical measurement is the ‘dose–response relationship’: an indication of adverse responses, ranging from short-term effects such as vomiting to long-term responses such as the development of cancers, to different doses of the substance. From this it is possible to estimate a ‘no observable effect’ level of the substance. In transferring this to humans, a safety margin is built in, for instance by taking the ‘safe level’ as 100th or 1,000th of the ‘no observable effect’ level in rodents, standardized for body mass.
Those who worry that this kind of process is not robust enough to detect potential problems tend to argue that extrapolating from rodents to humans, even with built-in safety margins, is not justified, or that cocktails of different synthetic chemicals in food may have unexpected effects not predicted from testing them one at a time, or that there may be very long-term effects not picked up by short-term toxicological tests. These are perfectly reasonable concerns and the science of toxicological risk assessment often involves informed judgements. However, the fact remains that synthetic chemicals such as pesticides, flavours, and colours are much more rigorously scrutinized for possible risks than are the naturally occurring toxins. Potatoes, for instance, would probably be banned if they were subject to the same scrutiny because they contain poisons called glycoalkaloids at levels far higher than 100th of the ‘no observable effect’ level. The generally accepted safety level is 200 mg/kg fresh weight of potatoes, and the average level in commercial potatoes is typically about half of this.
One moral of this section is that whenever there is a newspaper story about ‘cancer scare in baby food’ or ‘grilled meats can cause cancer’, one should ask whether or not the dose consumed in a typical diet is actually enough to cause harm. Another is that synthetic chemicals in food are not necessarily a bigger risk than natural chemicals. A third is that one cannot avoid eating foods that contain potential toxins.