The science of food

We eat for energy and for nutrients to keep various parts of our bodies functioning. Different foods meet these needs in different ways.

Excess energy from all foods is stored in the body as fat. All carbohydrates are converted to glucose and when glucose is carried in the bloodstream, after eating, it stimulates the pancreas to produce insulin. It's insulin's job to transport the glucose into cells where it's used for energy or stored, if it's not needed straight away.

When you eat a simple carbohydrate (as in the sugars in fruit and other sugary foods and drinks), the metabolism into glucose happens very quickly. Complex carbohydrates, such as those found in wholegrain cereals, take longer to convert to glucose. That's why a big bowl of porridge will keep you going all morning.

How fat works

Because the body stores energy as fat, when you eat fat, it's already in a form that makes it easy to store. This means energy from fat is the first to be stored and the last to be used. It worked well for our cold and hungry ancestors who had to hunt to stay alive, but it's really not useful if you work at a desk all day.

However, the type of fat you eat is important because it affects the amount and type of cholesterol in your blood.

Cholesterol is found in the blood in two types of packages - low-density lipoproteins (LDLs) and high-density lipoproteins (HDLs).

LDLs are liable to deposit cholesterol as fatty streaks on your artery walls. Those streaks can build up and narrow the arteries. Also, saturated fat can make the blood cells stickier, so they form a clot. A blood clot in the heart will cause a heart attack; in the brain it will cause a stroke. Saturated and trans fats increase the LDLs in your blood (and trans fats also lower the HDLs). They're the "bad fats".

HDLs do just the opposite - they carry cholesterol out of the tissues and back to the liver for processing. Monounsaturated and polyunsaturated fats decrease the LDLs, so are heart-healthy fats.

The orthodox view is that we should replace saturated fat in our diet with carbohydrates. But the Harvard researchers disagree. They argue that the key to reducing heart disease and the risk of other illnesses is the ratio of LDLs to HDLs. They say that when saturated fat is replaced by carbohydrates, the LDLs fall but so do the HDLs, so the ratio doesn't change much. It's only when you replace saturated fat with unsaturated fat, that you get a significant change in the ratio.

According to Willett, it's not a low-fat diet you need, but a low saturated/trans fat diet.

However, there's an even more important message to take from this. The body is not going to store unwanted energy as fat unless you eat more food than you need. As Willett says, "the best way to avoid obesity is to limit your total calories [kilojoules]".

It's a message that has wide support, and it's a hard thing to achieve if you're eating lots of fat - of any kind.

How do they know?

It's not easy to find out how food works in the body. Several types of trials and studies are used.

Randomised controlled trials

are the best method for assessing the effects of diet on health. In a large trial, individuals are randomly given one diet or another and followed for many years. The results can be persuasive, but running such trials is expensive and sometimes impractical.

Case control studies

involve patients with a certain disease, like cancer, being asked about their earlier diets. They are compared to a population group without the disease. This can provide in-depth information, but is subject to biases.

Epidemiological research

involves the study of large populations. This is the reverse of a controlled trial, in that you start with outcomes and look for causes.

Biochemical research

involves laboratory study to determine what happens in the body.

The Harvard research leans heavily on two epidemiological studies: the Nurses Health Study, begun in 1976 and involving 90,000 women, and the Health Professionals Follow-Up Study in 1986, involving 50,000 men. These studies found the risk of heart disease was strongly influenced by the type of dietary fat consumed, but not the overall amount.

These are very large studies and can be expected to produce robust results. But they still need to be approached with caution. Nurses and doctors are not necessarily representative of the whole population. And from a New Zealand point of view, the fact they are American studies may make them less relevant here.