Practices

Most claims about food and behaviour rest on weak studies. The ones worth taking seriously give each person both the real thing and a dummy at different times, with nobody knowing which is which, so that hope and expectation cannot do the work.

The food colouring trials are the standard example. Around 150 children each received drinks containing a preservative plus one of two dye mixes, or…

The same brand of sweets is often made with plant based colours in one country and artificial ones in another. If a company can make it without the dyes elsewhere, the dyes are not essential to the food.

Nine artificial colours are approved in the US, and the volume approved for use rose roughly fivefold between 1950 and 2012. One of them, Red No. 3,…

Skin prick and IgE blood tests find real allergies. The IgG food sensitivity panels sold direct to families have not been shown to measure sensitivity at all, and they tend to come back with a long list of foods to avoid.

IgG antibodies mostly show that you have eaten something, not that it harms you, which is why the results look alarming and change so little when…

A real food allergy shows itself on the body: a rash, a swollen or inflamed throat, something you can see. Behaviour on its own does not identify one, although behaviour often improves once a genuine allergy is dealt with.

One child's attention and mood went back to her usual level once berries came out of her diet, and what made that findable was the rash and the sore…

The studies linking pesticide levels to ADHD are tangled up with income, since families least able to buy organic differ in plenty of other ways too. If organic food is affordable for you, fine. If it is not, you are not failing your child.

In one study of 1,140 children, 94% had detectable pesticide in their urine, and ADHD was about twice as common among those with higher…

Most eating advice quietly assumes a budget. Fresh food is expensive and does not keep, so a household relying on food assistance ends up buying what lasts, and what lasts comes in tins and packets.

If money is the constraint, say it out loud rather than nodding along and then failing quietly at home. The advice can be rebuilt around what is…

Put an actual number on it: how many coffees, teas, energy drinks and colas in a normal day. Without that number, your sleep problems, your anxiety and how well a medication is working are all being read against an unknown.

People underestimate this, partly because it feels embarrassing and partly because a large takeaway cup is often two or three servings. Count across…

Stimulants sharpen focus in almost anyone, so responding well to one does not confirm you have ADHD. The diagnosis comes from your history and how you function, not from a trial of tablets.

This cuts both ways. Someone can be handed a diagnosis on the strength of a good response, and someone else can be doubted because their response was…

Food and movement are powerful, and hardly anyone keeps them up as prescribed. So the question is not what works best on paper. It is what you will still be doing in six weeks.

Roughly a third of people follow a prescription exactly as it is written, and keeping to a diet or an activity plan goes worse than that. Which means…

Pick one small change and leave the rest alone. A full overhaul asks the most of you at the point when you have the least to give, and it usually collapses in the second week.

One change, small enough to be almost boring, kept for a month. Then another. It feels slow while you are in it, and it is the version that tends to…

Working out what you have comes before choosing what to take. The right remedy for the wrong condition is still the wrong remedy, however good the remedy happens to be.

The health food shop route usually starts with a self diagnosis, and that first step is the one worth slowing down. St John's wort will do nothing…

Supplements are regulated as food rather than as medicine. Nobody checks strength or purity before a product goes on sale, and the main rule is that the packet cannot claim to treat a disease.

Most people assume the opposite, that anything on a pharmacy shelf has been through some kind of approval. Regulators here generally act only once a…

When researchers ran DNA testing on 13 herbal products, only 2 contained the plant on the label. Three had none of it at all, and several were padded out with rice, soy or wheat that nobody had declared.

The obvious problem is paying for something you did not get. The quieter problem is the undeclared filler, which matters a great deal if you avoid…

A second study, this time of 44 products, found only about half were authentic. Roughly six in ten were mislabelled, three in ten contained a substituted ingredient, and one in five held contaminants or fillers.

One study of this kind is a story about some bad batches. Two independent studies finding the same thing is a description of how the market works.…

If you are going to buy supplements, look for an independent verification mark: USP Verified, NSF, or a product tested by ConsumerLab. It is the nearest thing available to the check that regulators do not carry out.

Those schemes test identity, strength, purity and how a product was made, and manufacturers pay to take part voluntarily. There are different tiers,…

St John's wort can cut the effectiveness of the contraceptive pill by around a third. If you take both, that is worth knowing today rather than finding out later.

It speeds up the way the liver clears certain drugs, so the pill is not the only thing affected. Warfarin, some HIV medicines, ciclosporin and a…

Several herbal products change how fast the liver processes medication, which can leave a prescribed drug either weaker or stronger without a single change to the dose.

This is often why a supplement looks like it caused a relapse or a sudden run of side effects. Nothing about the prescription changed, so the pattern…

Two to be careful with in particular. Yohimbe reliably ramps anxiety up, so much so that it has been used in research to bring panic on, and kava is a sedative that stacks with alcohol and with sedating medication and has damaged livers at higher doses.

Kava has been restricted or banned in several countries over liver harm, which tells you how seriously that risk is taken. Yohimbe is a poor choice…

You cannot patent a plant, so there is little commercial reason to fund expensive trials of one. When you read that a herb has no evidence, that often means nobody paid to look rather than that somebody looked and found nothing.

Both halves of that need saying. No evidence is not proof of no effect, and it is also not permission to assume one. It leaves you with an open…

Some of these do have real evidence behind them. St John's wort has repeated support in mild to moderate depression, SAMe and methylfolate have reasonable backing as additions to treatment, and omega-3 has looked weaker in the larger and better run trials.

That last part is worth sitting with, because it is what usually happens. An early promising effect shrinks as trials get bigger and better…