katueng/Shutterstock

Once I inform folks about what I do for a residing (a meals scientist), I invariably get the identical response. A roll of eyes and criticism about how the issue with meals science is that it’s all the time telling me that pink wine is nice, then it’s unhealthy, then it’s good, relying on the day of the week. The reality is that the sphere of dietary epidemiology, the research of massive populations and the way the meals they eat impacts their well being, has struggled for a while with very unhealthy PR.

This unhealthy PR stems from the truth that these research often depend on what folks inform us they eat, and never what they really eat. This may work properly for some features of the weight-reduction plan – dietary patterns, for instance – however not others, particularly particular person meals or meals parts.

There are two primary issues: first, folks don’t all the time accurately report what they eat and customarily declare to eat extra of what’s thought of to be wholesome and fewer of what’s thought of to be unhealthy. This impacts the correlation we observe and generally even reverses the end result. As an illustration, self-reported sugar consumption is related to a decrease physique mass index (BMI), whereas now we have proven beforehand that precise sugar consumption is unsurprisingly related to the next BMI. This drawback is well-known and has been mentioned amongst nutritionists for many years. There are some refined strategies to handle this, however they’re not all the time attainable to make use of.

The second is harder to handle however has arguably a a lot bigger impression when investigating particular person compounds, resembling nutritional vitamins, minerals or bioactives resembling caffeine or flavanols – meals is just not standardised. The variability in meals composition is big, even in meals harvested from the identical plant. Within the 1960s, researchers analysed the composition of apples on a single tree and located greater than twofold variations within the composition of those apples. Additionally, composition adjustments throughout storage and, in fact, preparation. Nevertheless, in vitamin analysis, we regularly should depend on printed meals composition knowledge and use a single worth. For instance, for every apple, we assume it incorporates 9mg of vitamin C, when in actuality, this may be very completely different.

What does this imply? It implies that it’s unattainable to estimate the precise consumption of a compound based mostly on dietary knowledge and meals composition knowledge alone. A cup of tea incorporates between 1mg and 600mg whole flavanols. But in most analyses, this will probably be standardised to 125mg per cup. This has large penalties, because the estimated consumption of flavanols now not depends upon the precise consumption, however merely the consumption of sure meals. Most research performed up to now have this weak point.

Nearly ten years

The easiest way to handle this drawback is by measuring what’s taken up by the physique, utilizing so-called biomarkers. We are able to do that for instance in urine, but additionally in blood and hair. This methodology tells us precisely what an individual has consumed and doesn’t depend on meals composition knowledge or the particular person telling us what they ate. However this method is dear and desires a number of preparation, which is the explanation there haven’t been many large-scale research up to now.

We determined to make use of this method to analyze the affiliation between flavanols and blood strain. Flavanols are present in a variety of meals, resembling tea, apples, wine and cocoa. A number of smaller research have proven a helpful impact on blood strain, and their impact on heart problems danger is at present being investigated in a big trial. There may be, nonetheless, no dependable knowledge on their impact in most of the people when consumed as a part of a standard weight-reduction plan.

We investigated the hyperlink between flavanols and blood strain.

Seasontime/Shutterstock

Such a challenge required a number of planning and preparation, and it took nearly ten years from begin to end. We needed to determine essentially the most appropriate biomarkers first after which develop analytical strategies. Our colleagues from the College of California, Davis, in addition to meals producer Mars Inc., performed groundbreaking analysis into the metabolism of flavanols in people and developed strategies to synthesise these metabolites so we may determine essentially the most promising biomarkers and set up that they supply an correct estimate of consumption. On the similar time, our colleagues from EPIC Norfolk and the MRC Epidemiology Unit, in addition to from LGC Fordham Laboratory, arrange the infrastructure to course of greater than 25,000 urine samples.

The outcomes of the research have been thrilling: for the primary time, we may present that there was a statistically important, significant distinction in blood strain of 1-3mmHg between these with excessive and low-flavanol consumption. This distinction is much like the impact of lowering salt consumption or adopting a Mediterranean weight-reduction plan.

There was a way more vital discovering, although. When evaluating objectively measured flavanol consumption with knowledge estimated utilizing the normal methodology, we discovered solely a really weak correlation. This exhibits that combining self-reported knowledge with meals composition databases is unlikely to offer a dependable estimate of flavonol consumption – and the identical might be true of many different compounds with excessive variability in meals composition.

Gunter G. C. Kuhnle receives funding from Mars, Inc.