Scientists created a new computer method to design personalized eating plans tailored to each person’s body. Instead of giving everyone the same nutrition advice, this approach uses blood test results and personal food preferences to figure out what nutrients each individual needs. The researchers tested their method on people with prediabetes and kidney problems, and the computer-designed diets appeared to help bring blood markers closer to healthy levels. While the results look promising and match what nutrition experts already know, doctors need to do more real-world testing before this personalized approach can be used in regular medical care.
The Quick Take
- What they studied: Can a computer program create custom diet plans that work better for individual people by looking at their blood test results and what foods they like to eat?
- Who participated: The study used nutritional data from two groups: people with prediabetes (high blood sugar that isn’t quite diabetes yet) and people with kidney disease. The exact number of participants wasn’t specified in the research.
- Key finding: The computer-designed personalized diets appeared to help restore or get closer to normal blood marker levels in both groups, especially when the person’s body could realistically achieve those changes.
- What it means for you: This suggests that one-size-fits-all diet advice might not be best for everyone. A personalized approach based on your blood work and food preferences could potentially work better for you, but doctors need to test this more thoroughly in real patients before recommending it widely.
The Research Details
Researchers created a new mathematical computer model based on Bayesian statistics—a method that learns and updates predictions as it gets more information. The model works in two steps: first, it figures out which foods and nutrients would help bring a person’s blood markers (like cholesterol and insulin) into healthy ranges. Second, it picks the diet option from those healthy choices that requires the smallest changes from what the person already eats, making it easier to stick with.
The researchers used the Nordic Nutrition Recommendations 2023 (official healthy eating guidelines) as their starting point. They tested their method using real nutritional data from two different groups of people: those with prediabetes and those with kidney problems. They also ran computer simulations to see how well the personalized diets would work.
This approach is important because it recognizes that people’s bodies respond differently to the same foods. What works perfectly for one person might not work as well for another. By using blood test results and personal preferences, the method creates recommendations that are actually tailored to each individual rather than giving generic advice that might be hard to follow.
The study is a mathematical and computational research article that shows the method works in theory and with existing data. However, it hasn’t yet been tested in real clinical settings with actual patients following the recommendations over time. The researchers themselves note that real-world clinical studies are needed before this can become standard medical practice. The fact that results align with current nutrition science is a good sign, but more evidence is needed.
What the Results Show
The computer model successfully created personalized diet recommendations for both the prediabetic group and the kidney disease group. When the simulations ran the personalized diets, the blood markers appeared to move toward normal, healthy levels in many cases.
The key success factor was whether the person’s body could realistically achieve those changes based on how their individual body responds to different nutrients. When the computer model correctly understood a person’s unique nutrient responses, the personalized diet worked well. The method also successfully balanced two important goals: keeping blood markers healthy while making only small changes to what people normally eat, which makes the diet easier to follow.
The preference component of the model proved valuable because it allowed the system to pick the best diet option from several healthy choices—the one that required the least disruption to the person’s current eating habits. This feature could help people actually stick with the recommendations instead of abandoning them.
The model provided clear, specific targets for nutrients that are less critical to controlling blood markers, giving people concrete goals to work toward. The method also showed that it could work with the official Nordic Nutrition Recommendations, suggesting it could be adapted to other countries’ official nutrition guidelines. The approach proved flexible enough to handle different health conditions (prediabetes and kidney disease), suggesting it might work for other conditions too.
This research builds on existing knowledge that people respond differently to diets based on their genetics, current health, and other personal factors. The new contribution is a systematic, mathematical way to use this knowledge to create recommendations. The results align with what nutrition scientists already know from other studies, which is reassuring. However, this is the first time researchers have combined blood marker data, personal preferences, and Bayesian statistics in this particular way to create personalized recommendations.
The biggest limitation is that this study used computer simulations and existing data rather than testing the recommendations with real people over time. We don’t know yet if people will actually follow these diets or if they’ll work as well in real life as the computer predictions suggest. The study also didn’t specify exactly how many people’s data was used, making it hard to judge how broadly the results apply. Additionally, the method requires blood tests and nutritional data that not everyone has access to, which could limit who could benefit from this approach.
The Bottom Line
This research suggests that personalized nutrition recommendations based on blood work and preferences may be more effective than generic diet advice (moderate confidence level—more testing needed). However, it’s too early to recommend this as standard medical practice. If your doctor offers personalized nutrition counseling based on blood markers and preferences, this research supports that approach, but it should be combined with regular follow-up testing to see if it’s working for you.
This research is most relevant to people with prediabetes, kidney disease, or other conditions where blood markers are important to manage. It’s also relevant to anyone interested in personalized medicine or nutrition. People with stable health and no specific health concerns may not need this level of personalization. Anyone considering major dietary changes should discuss them with their doctor first.
Based on the research, if a personalized diet is working for you, you might expect to see changes in blood markers within weeks to a few months, depending on the specific markers and your body’s response. However, real-world clinical studies are needed to establish realistic timelines before this becomes standard practice.
Want to Apply This Research?
- Track your blood marker levels (cholesterol, insulin, glucose) every 4-8 weeks if you’re following a personalized nutrition plan, along with daily food intake and how closely you’re following the recommendations. This data helps the system learn what’s working for your body.
- Start by making small, gradual changes to your current diet rather than overhauling everything at once. The app could help by suggesting one or two specific nutrient targets to focus on each week, based on your blood work and current eating patterns, making changes feel manageable.
- Set up regular check-ins (monthly or quarterly) to compare your blood markers against your personalized targets. Use the app to track which dietary changes had the biggest impact on your specific blood markers, building a personalized understanding of how your body responds to different foods.
This research describes a mathematical method for creating personalized diet recommendations, but it has not yet been tested in real clinical practice with actual patients. The findings are based on computer simulations and existing data, not on people following these diets in real life. Before making significant dietary changes, especially if you have prediabetes, kidney disease, or other health conditions, consult with your doctor or a registered dietitian. Do not use this research as a substitute for professional medical or nutritional advice. Blood marker targets and dietary needs vary greatly between individuals and should be determined by qualified healthcare professionals who know your complete medical history.
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.