If we analyse the principles of thought on which magic is based, they will probably be found to resolve themselves into two: first, that like produces like, or that an effect resembles its cause; and, second, that things which have once been in contact with each other continue to act on each other at a distance after the physical contact has been severed.

In fact he goes on to call the former of these principles “homoeopathic magic”.

This is the basis for many traditional systems of magic around the world. For example, two widespread ways of cursing someone are to use images or dolls of them (the first principle) or to use their hair or toenail clippings (the second principle).

Now consider how a homoeopathic medication is typically made:

1. A substance is chosen which produces similar symptoms to the sickness to be treated.
2. It is repeatedly diluted with water, each time taking just a small amount from the previous dilution, until it is unlikely that even one molecule of the substance remains in the water.
3. This water is then used in the manufacture of pills: but is evaporated in away in the process.

The first step in the process relies on a modification of the first principle of sympathetic magic: that like cures like, rather than affects like. Of course, it relies on similarity of symptoms, rather than similarity of appearance. It does not depend on the cause of the symptoms, but on similarity of visible, or otherwise easily perceptible, symptoms.

The remaining steps rely on a variation of the second principle. Rather than the water acting on the substance it has been in contact with, it retains some property it gains from the contact. Similarly, the pills retain that property from the contact with the water.

While homoeopathy is an interesting variation on the principles of sympathetic magic, I cannot see how it can be denied that it is a variation on those principles, rather than something entirely different.

Of course, most of this assumes that you are a healthy adult: many medical conditions need special diets, and children’s needs are different from adults.

Generally agreed

• Sugar is bad and should be consumed in much smaller quantities than is typical of modern eating habits.
• Fruits and vegetables are generally good (but see below for exceptions)
• Oily fish is good
• Transfats (heavily used in processed food) are bad, or at least possibly bad with no benefits

Generally agreed with caveats

• Whole grains are better than refined. Paleolithic and low carb diet advocates would say you should either eat little or now grain derived foods (after all human beings never evolved to eat grasses), or to eat only fermented grains. They also claim whole grains contain more  anti-nutrients (mild/small quantities of toxins) than refined.
• Less processed it usually better.
• Some fats are better than others . Things generally agreed to be better include olive oil. Most medical opinion favours monounsaturated or polyunsaturated over saturated. Paleo diet would favour oil in food that you can eat raw, or can extract without heavy machinery (e.g. olive oil, coconut oil and, most of all, meat).

Controversies

• Proportion of calories from fat vs carbs vs protein. Some low carb diet advocates promote high protein. Most medical opinion says high carb with less than 20% fat. Some low fat diet advocates say less than 10% fat. Paleo diet says high fat and low carb.
• Animal vs vegetable fat: conventional wisdom is that vegetable fat is better, paleo that animal fat is better as it is what our ancestors evolved to eat.
• High carb vegetables: obviously farinaceous ones such as potatoes, but also others such as many beans.
• Vegetables that cannot be eaten raw: discouraged on a paleo diet.
• Fatty dairy products: mostly low carb (unless sugar has been added), not paleo.
• Low fat dairy products: higher proportion of the calories in skimmed milk come from sugar: the fat would be preferable in a low carb diet.

Evidence and arguments

Medical research does not have a great track record on telling us what it is healthy to eat. Medical opinion has changed radically over the last few decades, sometimes more than once (eggs went from being regarded has healthy, to being considered a high cholesterol heart disease risk, and are now considered good again as dietary cholesterol is not generally regarded as a problem for healthy people).

Some logic also seems to be missing from many medical arguments. Consider that weight loss is encouraged by avoiding fat because fats have a high energy density (i.e. more calories per gram than other foods). That assumes that people eat a fixed weight of food, whereas different foods are not equally filling per unit weight. Why are Chinese meals (at least as served in the West) proverbial for leaving you feeling hungry an hour afterwards? That is because you need to eat a lot of a meal heavy in white rice to feel full for long: something that is also evident in modern South Asian eating habits  that are based in large servings of white rice. We actually end up eating fewer calories on high fat diets.

There is also a lack of evidence on very important questions. Take a look at this summary of the evidence on the link between fat and heart disease, by the Cochrane Collaboration, a highly respected medical research non-profit organisation, best known for producing such reviews. There replacing saturated fat with monounsaturated or polyunsaturated  may reduce the risk of heart disease, while there are no clear health benefits to reducing the total amount of fat we eat. This is critical given that reducing fat has been the most common medical advice on diet for the last few decades.

This is not surprising. The effects of diet are very hard to research. There are many bother factors (exercise, income, lifestyle), the effect of which are very hard to separate as they are highly correlated with diet and with each other. Long term studies are expensive. It is very difficult to accurately monitor what people eat.

The arguments can get very convoluted. For example, a low proportion of carbs compared to a typical diet can be achieved by eating lots of some types of fast food: but this a very different diet from what low carb diet advocates, or paleo diet advocates, recommend: it still has a much higher proportion of carbs, and a very different mix of fats.

A lot of the actual evidence in favour of a paleo diet is very well summarised by this post, which I will not duplicated here. There is more in favour of low carb diets in general in Gary Taubes’ book “Good Calories, Bad Calories”.

However, there is a more fundamental point in favour of the paleo diet in particular. Where the evidence is weak and lacking, the common sense starting point is what we evolved to eat. If I had to feed a lion,  I would have no idea what specific nutrients it needed, but I know that wild lions live on raw meat, so that would be the obviously be a suitable diet. Similarly, in the absence of proof, the common-sense diet for humans is what wild humans (hunter gatherers) eat: meat, raw vegetables, etc.

What works for me

I am short so my weight for the last few years, of around 80kg, was already enough to make me fat (although not really obese). A holiday last November meant lots of eating out, and sent my weight up to 84 kg. Since then, I have managed to lose a steady kilo or thereabouts every month, with neither a great effort, nor calories restriction, nor an increase in exercise, and am now down to 76kg and my waist (measured in terms of which trousers fit) has shrunk significantly.

I have adopted what is best described as a paleo influenced diet. I keep the amount of carbs I eat, especially grains, low. However I am not very strict about it, and eat a moderate amount of non-paleo vegetables such as pulses I also drink as much full fat milk and non-sweetened dairy products as I want. I cook with copious amounts of ghee, olive oil and coconut oil. I do not even attempt to stick to my diet when I go out.  I make no attempt to restrict the quantities I eat: I eat until I am full.

I disagree with the paleo objection to dairy, as it is evident that those of us with ancestors from parts of the world where milk has been drunk for a long time (Europe and South Asia) have evolved to cope with dairy: over 90% us have the gene that enables us to digest lactose as adults. I also do not see that milk is greatly different from other animal foods. I also do not believe that a strict diet is necessary. Humans are, naturally, opportunistic omnivores, like baboons. It is perfectly natural for us to have a a bit of everything, and we evolved to cope with this.

I am sure that it would be better to also follow the palaeolithic lifestyle in terms of exercise: regular light exercise such as walking, with short bursts of activity such as running. I get very little exercise, and my plans to change that are rather like St Augustine’s prayer for chastity “but not yet”.

I will leave aside my view that the government should not protect people from themselves, and, just for this post,  and, for the moment, accept the consensus that the role of the government includes forcing people to do what the government judges best.

Lets start with the missing information. What is meant by “More than 200,000 people could die early from alcohol-related diseases, accidents and violence over the next 20 years”? Does it mean that the death rate will be 200,000 more than if no one drank any alcohol? Does it mean that 200,000 people will die from diseases, accidents and violence in which alcohol is one of (many) avoidable risk factors?

Of course, as always, The Guardian, fails to but the numbers in context. 10,000 deaths a year is not a huge number in a country the size of the UK: barely over 1%. Depending on he exact meaning of the numbers, the actual number of deaths caused by (as opposed to “related to”) alcohol could be much lower and comparable with, for example, the 3,000 suicides a year. Given the distressing nature of suicide, and its impact on the friends and families of its victims, I think that is sufficient to argue it should be a higher priority.

Compare the impact of alcohol with that of bad diet (too much sugar, too much, and harmful types of, fat, etc.): . There are at least 30,000 deaths a year related to diabetes, and that is almost certainly an underestimate, and it is worsening rapidly, and it is only one of many diseases caused by unhealthy diets.  Worsening rapidly is something of an understatement: the number of people who have diabetes has doubled over the last thirty years.

Diet bring us to another issue. The health effects of different alcoholic drinks is very different:  the amount of carbohydrates in one pint of beer is comparable to an entire bottle of red wine.

That is one possible reason for the lack of correlation between alcohol consumption and life expectancy. Given that, what reason is there to think that reducing overall alcohol consumption will improve public health. In fact a WHO study found that the countries with the lowest level of alcohol related health problems are the wine drinking countries of south western Europe, many of which have very high levels of consumption.

In fact, it may even worsen public health. Moderate drinking improves life expectancy, so price increases that deterred moderate drinkers would lead to lower life expectancies. Heavy drinkers, especially very heavy drinkers, may be addicted, and will therefore be likely to spend more rather than reduce consumption. Abstaining entirely from alcohol carries a similar risk to heavy drinking, why are there no calls for government action to encourage teetotallers to start drinking?

The criticism of a government strategy that relies on voluntary cooperation from the industry, has two serious fallacies. They claim that “the primary requirement for these industries is to deliver shareholder value by maximising consumption”. This is wrong. Individual firms want to maximise profit, which is not the same as maximising consumption: the profit on one bottle of expensive wine is likely to be far more than on many bottles of plonk. It is even possible that a minimum sales price may may the industry more profitable by killing the lowest margin products.

The other fallacy is that because a businesses aim is to make a profit, it is entirely uninfluenced by anything else. Companies are run by individuals who are just as likely to act from ethical motives as anyone else, they are subject to public pressure, and the ability of governments to coerce companies to act “voluntarily” is enormous. Consider the success the government has had in getting internet service providers to subscribe to the Internet Watch Foundation’s controversial, secretive, undemocratic, ineffective and unaccountable censorship scheme.

I have become very cynical about the reasoning behind calls for action on alcohol. It seems to reflect a fixed, and puritanical, idea that alcohol is bad, rather than an assessment of evidence and likely benefits.

The report actually talks about nutrients, not health. The very first paragraph is very revealing about what is covered:

This review does not address contaminant content (such as herbicide, pesticide and fungicide residues) of organically and conventionally produced foodstuffs or the environmental impacts of organic and conventional agricultural practice.

So, the most important reason for preferring organic food is ignored. Herbicides, pesticides and fungicides are toxic: they would not work if they were not, at least, toxic to plants, pests and fungi! Large doses are toxic (ingesting these is a common way of committing suicide in some places), and I would very much like to know of any conclusive studies of the long term effects of continually eating small amounts with each mean.

Why did journalists fail to mention this for the first day or two? Did they no even read the executive summary (the easy bit at the top for people who are unable to understand the rest)?

Even regarding the nutritional content, it is obvious that there are not going to be huge differences in the content of the main nutrient groups in organic food and chemically grown food. However looking at the details, there are lots of differences that could be significant, for example the level of both phenolic compounds and flavinoids is higher in organic food,and, to quote from the report itself:

Numerous health benefits have been ascribed to the actions of phytochemicals such as phenolic compounds and flavonoids, many of which related to their antioxidant activity. The recent World Cancer Research Fund report suggests that quercetin (a flavonol) may prevent lung cancer (although the strength of evidence for this relationship was graded as “Limited – suggestive”4) (17). There is also some evidence from cohort studies (although not from randomised controlled trials), that high flavonoid intake is associated with lower rates of coronary heart disease mortality.

There are plenty of other examples. These are a bit more difficult to find, as they require reading the report, but it took me about ten minutes. I assume the failure of various journalists to find this reflects either a the extreme time pressures they work under — the media like quality, as long as it is cheap!

The other thing that is striking about the conclusions is that in many cases there is not enough evidence to draw a conclusion. It is basic to statistics that the larger the sample size the more meaningful the conclusions. For example, four of the studies of the effect on people used fewer than 20 participants, and most were conducted over short periods of time, so long term effects are ignored altogether. The fact that it is a study of studies also has other effects:

Statistical analysis was not attempted due to marked heterogeneity of study designs, exposures and outcome measures among the included studies.

Another flaw is that the studies did not use a consistent definition of organic. This varies from country to country (the US, for example, has a long list of ingredients that can be used in food labeled organic, that do not need to be produced organically) and from certifying authority to certifying authority. A far better approach would be to use only food produced to a strict definition of organic.

The most important conclusion of the study, is that nothing has really been proved. Neither then Food Standards Agency, or journalists, have paid much attention to this

A surprising and important finding of this review is the extremely limited nature of the evidence base on this subject, both in terms of the number of studies and the quality of studies found. It is essential that future research (both human and in vitro studies) is better designed and, at the very least, meets the minimum quality criteria applied in this review, including an accurate description of the organic certification process. It is recommend that in future, high quality randomised controlled trials should be conducted which have samples of sufficient size to reliably detect the presence of effects, longer and more realistic dietary exposures, and more accurate and objective approaches to measuring dietary intake and health outcomes.

There is little evidence that organic food is more nutritious so far, because no one has really looked, and if we assume that constantly eating small doses of known toxins is entirely harmless.

I had always naively imagined that when they were not actually fighting, they would spend their time making sure they were ready to fight: i.e. getting lots of exercise and practising shooting things. I would have guessed that they got several hours of exercise a day. What do they do? Can someone tell me? Knowing the government I have a nasty feeling it is going to be something like “filling in bits of paper to tell people how much exercise they have not been getting”.

So now they have all these soldiers who are classified as PUD. This is an abbreviation for “physically unable to deploy”, rather than an explanation of what they ate to become that way.

This used to be avoided by an upper limit on soldiers body mass index. The reason they dropped this was that the increasing fatness of teenagers meant that the army could no longer find enough non-fatties any more. Even more amusingly, body mass index does not take into account whether the excess mass is fat or muscle. The army, therefore, used to have a policy of rejecting would-be soldiers for being too muscular.

BMI is something that doctors like to use because it is easy for them, and is sort of right most of the time, rather than actually being a good measure of anything.

Of course the other solution might be to recruit people who already get that much exercise in their spare time. Maybe all those bankers who used to keep City gyms busy might be open to a new career?

A longer term solution would be to persuade parents to stop being so paranoid about letting their children outdoors (because there is supposed to be a paedophile on every street corner) so that they grow up to be fit enough to be recruited.