Wetenschap - 14 maart 2013

The truth about your diet

Biomarkers can replace questionnaires.
Proteins reveal what we eat.

Can you still remember how many snacks you ate yesterday? Or what you had for breakfast last Tuesday? We cannot answer such questions precisely because our memory is poor and unreliable. We also tend to exaggerate how much broccoli and cabbage we eat and underestimate the number of chips and crisps. Despite this, nutrition scientists generally use questionnaires to be filled in by their subjects - they usually have no alternative.
That is why there is a real need for biomarkers for food. These are compounds that can be detected in blood or urine and reveal how much you have eaten of a particular food. As part of her PhD research in the Human Nutrition department, Wieke Altorf-Van der Kuil looked for biomarkers for proteins, the object of her study. To be precise, she looked for markers that can distinguish between proteins from meat, cereals and dairy products.
Interesting results
She recruited 30 subjects who ate what the researchers gave them for three weeks. Around 18 percent of these meals consisted of protein, largely of one type: meat, dairy or cereal. At the end of each week, the subjects collected their urine for 24 hours and a blood sample was taken. That gave interesting results. 'I was pleasantly surprised to find such good biomarkers, especially for meat,' says Altorf. 'Of course that is what you hope beforehand, but I had my doubts whether the biomarkers would be able to explain much variation.' In practice it turned out all three groups were easy to distinguish. What is more, they found useful biomarkers for meat and cereals. Three compounds in combination gave very accurate predictions of how much meat protein people had eaten. Seven markers in combination gave a reasonable prediction for cereals.
Although she is pleased with the results, Altorf emphasizes that this is only the first step.  They now have to determine how accurate the markers are after a normal meal. After all, then you eat all kinds of proteins mixed up.
And questionnaires will probably never disappear entirely. They may only give a rough estimate but they are much cheaper than chemical analyses. And if that changes, it will mean an end to our flattering lies.