Today’s dietary guidelines are based on what is generally healthy to eat, but not everyone’s body works the same way. If you and a friend eat exactly the same breakfast and then measure your blood sugar every fifteen minutes, you will likely see very different curves. In one person, blood sugar rises quickly and drops just as fast; in the other, it climbs more slowly but stays elevated for longer. The same applies to blood lipids and inflammatory markers. This is the phenomenon that precision nutrition seeks to understand and translate into practice – the right diet for the right person at the right time.

The long-term goal is to link data on how an individual responds to a specific meal with their future risk of, for example, type 2 diabetes, heart attack or stroke.

/ Rikard Landberg

Professor i livsmedelsvetenskap vid Chalmers tekniska högskola

/ Rikard Landberg

Professor i livsmedelsvetenskap vid Chalmers tekniska högskola

Existing research is locked away 

The first major study in this field was published in 2015, by an Israeli research group. They showed it was possible to build a model that predicts an individual’s blood sugar response based on gut microbiome composition, dietary habits and health data. The results were groundbreaking. However, the algorithms that emerged from that research are now owned by private companies.

– The algorithm is locked inside a company. No one knows how it works, says Rikard Landberg.

He and his colleagues are therefore taking a different path, building an equivalent model and making it openly available.

– We want to make it open for healthcare to use it, or innovators who can develop digital health platforms and offer personalised advice as part of their services, Rikard continues.

Ten years of data from thirty thousand people

The research draws on data from several sources, among them SCAPIS 2-HOME. This sub-study, part of the larger SCAPIS project, has tracked thirty thousand randomly selected Swedes aged 50 to 65 over more than a decade, monitoring heart, vascular and lung function. When half of those participants were invited back for a follow-up, the team saw an opportunity to run a precision nutrition study alongside it.

Just over 4,500 people were invited, and around 2,000 completed the full study. On the day of their visit, they ate a standardised breakfast on site. Lantmännen contributed bread and breakfast cereals to the study. Whole grain and rye are an area where The Research Foundation has funded research for many years, and it features naturally in this study's design too. Throughout the visit, participants' blood sugar responses were monitored continuously. When they left, they took home three meal drinks and a self-sampling kit.

– We divided them into four groups. One ate a high-carbohydrate breakfast, one ate low-carbohydrate with plenty of fat, and we then varied those further using whole grain and white flour, explains Rikard.

A finger-prick blood test at home Participants collected their own blood samples at home, using a finger-prick on a so-called Capitainer card. They then brought the cards in or posted them to the SCAPIS clinic, which forwarded them to Chalmers. In total, the study has collected thirty-two thousand such cards. Stool samples were also gathered to examine the gut microbiome, and participants kept a record of what they ate. The cards will be analysed for a range of molecules alongside more conventional health markers. The advances made in analysing home-collected samples are making it increasingly feasible to run this kind of study at scale.

Pilot data raises hopes

Even from the first eight hundred participants, the results are clear enough to point in a direction. Individual variation in response to exactly the same meal is large, which the team expected. What came as a welcome surprise was how much a blood sugar response to a single standardised meal can reveal.

– We could see very clearly that how a person responds to one meal correlates strongly with how their blood sugar regulation works at home over two weeks. Other large studies have recently shown that this variability matters for predicting the risk of type 2 diabetes, which makes our findings particularly interesting, concludes Rikard.

Full results are expected in 2028, though partial publications may come earlier. The next step is to test whether personalised dietary advice actually leads to better health outcomes than general guidelines. The goal is straightforward to describe, though far harder to deliver: dietary advice that today is only accessible to those who can afford it should, in time, be within reach of everyone.