Knowing the form

This is a story about forms; yes, those objects of hate that you have to fill in all the time. But, more significantly, it’s also a story about diet, the foods we eat and your long-term health. Most of us, after all, try to follow a healthy diet, but it’s a road not easily followed because we constantly read contradictory stories about what constitutes a ‘good’ diet. Medicine, in particular, always seems to be at loggerheads with nutritionists, and one camp often seems to contradict what the other side says.

The family doctor likes to remind us, as he himself has been taught at medical school, that we get all the nutrition we need from our diet, and don’t need to take supplements. How does he know this? Because of the research that’s been gleaned over the years.

Which brings us back to forms. For the past 20 years or so, medical researchers have been collecting data on people’s diets and eating habits using a food-frequency questionnaire. As with all these types of forms, the questionnaire is left with the subject to complete at home. The trouble is, the questionnaire relies on a person’s memory, and his willingness to tell the truth. Do I really have to reveal that I ate two plates of fries today? Fortunately, the questionnaire is imprecise enough to allow these occasional lapses of memory. The result, however, is highly inaccurate data. So, you have a heart problem, but you don’t eat fries? Then fries can’t be contributing to health problems.

Dr Sheila Bingham and others from the MRC Dunn Human Nutrition Unit in Cambridge have highlighted the extent of the problem in a recent study (Lancet, 2003; 362: 212-4). They decided to check the contradictory evidence surrounding the link between a high-fat diet and breast cancer. Small, controlled studies point to an association, but this is never confirmed in larger, and supposedly more scientific, studies.

Bingham and her colleagues monitored the responses of 25,630 men and women in Norfolk, England, who completed a food-frequency questionnaire between 1993 and 1997. Of those, 23,656 also completed a food diary, a more sophisticated questionnaire that is only now starting to be used in studies. The diary forces the participant to itemise, on a daily basis, all the foods eaten, reducing the risk of omitting or forgetting food items. It also involves more work for analysts, who have to convert the food diary entries into nutritional values.

In all, 168 cases of breast cancer were reported among the participants from the time the trial started until September 2002. When the researchers looked at the patients’ food-frequency questionnaires, they found a relatively low intake of fats in the diets – but the food diaries, completed by the same people, revealed a completely different picture.

On analysing the food diaries, the researchers found that the cancer patients actually had a diet that was very high in fats – especially saturated fats. The food-frequency questionnaires also suggested a far higher intake of dairy and saturated fats.

Overall, food-diary data showed that a diet high in saturated fats was twice as likely to lead to breast cancer than one with a low-to-average saturated-fat intake. But this link would never have been found had the researchers stuck to their findings from the food-frequency questionnaires – and yet, the same people provided the original data.

Furthermore, this finding suggests a far greater association between saturated fats and breast cancer than anything that has ever before been suggested in earlier trials.

Very few – if any – large studies into nutrition and health use the food-diary method. It’s too cumbersome, slow and expensive, researchers say, and grants for studies into diets are few and far between anyway. Better to spend the grant money on recruiting more participants than to consider changing the methodology, they may well say.

And pharmaceutical companies, which are the main source of research funds, prefer to have their own drugs tested rather than fund an exploration into nutrients that may render their drug therapy redundant.

Dr Bingham’s study is clearly a bombshell. It also leaves nutritional research at a crossroads. Does it now abandon the food-frequency method completely, or could it be that the Bingham study was itself flawed? Some researchers are already forming a line behind that school of thought – why, after all, change a comfortable methodology for a more complicated one that will soak up more of the research grant?

Dr Bingham happens to be on the side of the angels because her study merely confirms the findings of smaller trials, thereby eliminating conflict and confusion.

But to accept this will mean a profound change in medical thinking. It also throws into question every large-scale trial on nutrition thus far completed, and upon which so much medical advice is based. Indeed, her study is one of the largest-ever studies into nutrition and disease, and the first to confirm the findings of smaller trials.

In the meantime – and while the debate continues to rage – you, the ultimate consumer, will have to filter the conflicting research findings you will read in the newspapers. Is it a large study, possibly involving thousands of participants? Then treat it with a pinch of salt (assuming salt is good for you).

And to think that years of debate about a healthy diet pivoted on those wretched forms that nobody likes to complete and, when they do, they do that most human of things – they lie.

Bryan Hubbard
Bryan Hubbard is the publisher of What Doctors Don’t Tell You.