Big data, AI and apps help us understand which foods are bad for us: here are two projects (one experimental, the other already available) that help us eat healthily

We're not talking about weight loss diets here. But rampant obesity yes: Italians are becoming fatter and fatter, especially children and young people. The numbers prove it.

The obesity rate in very young Italians is reaching that of the United States: a figure that was unthinkable until a few years ago. And while the goodness and healthiness of the Mediterranean diet is celebrated everywhere, we lovers of good food are abandoning it, especially when we (not) cook for the little ones. How come?

The future talks about food and DNA: Foodome

Giulia Menichetti explains it, a thirty-seven-year-old physicist who has been involved in research called Foodome for some time. Together with Albert-László Barabási, network science guru, she is developing a series of algorithms capable of verifying the effect of over 26,000 substances present in foods on individual DNA and collective.

Each body has an individual response to the numerous molecules that make up foods, especially if they are processed.

And even when we talk about natural substances, it is impossible to draw up lists of functional, or healing, foods on an individual basis. There are those who use proteins better than others, while those who metabolize animal fats optimally.

From a medical point of view, therefore, food is approached with valid and sensible advice, but not personalized nor scientifically unassailable.

AI will be able to give certain answers

An algorithm applied to a machine learning model capable of analyzing big data will save us from 80% of diseases caused by external agents, primarily food, using epidemiological and extensive data biochemical studies on how the human body reacts to the numerous artificial and non-artificial substances present in what we ingest.

The work is terribly long, they are well aware of this at Harvard Medical School, which supports and hosts the Foodome laboratories together with Northeastern University in Boston.

But it's worth it, because Foodome will tell us which foods to prefer to have positive effects on health. Or which are certainly toxic on a large scale, in the long term in case of continuous use.

The aspartame 'case'

In fact, at present the adjectives most commonly used in scientific papers that summarize research on human groups are: probable, suspicious, hypothetical. Because evaluating the accuracy of an epidemiological hypothesis is an enormous job, almost impossible to implement without AI and projects like Foodome.

A practical example: the International Agency for Research on Cancer has included aspartame (the most common dietary sweetener) in the C2 category: potentially carcinogenic substances. A decision motivated by some large-scale studies.

However, the WHO reacted without actually making any changes to the use of the sweetener: it took note of the AIRC's position but neither prohibited nor regulated the use of aspartame. Because there are no sufficiently extensive studies to support a radical choice.

First investigated: industrial food

The partial results of the Foodome research, however, give certain answers on at least one thing: processed food is bad.

Not only is it responsible for the Italian obesity epidemic, but it is the main vehicle for the presence of dangerous substances in our body. Additives, preservatives, colourants, industrial fats and sugars: Giulia Menichetti defines them as 'dark matter', because there are no significant studies on many commonly used substances.

Molecules that should not be part of our diet and which instead, especially for economic reasons, Italians have become accustomed to consuming and offering to children.

Yuka, an app for shopping without toxic substances

Practical solutions? Cook fresh food, avoid industrial products and learn to read labels. Difficult task? Once again technology solves the problem, although in a less personalized way than what Foodome will be able to do once applied on a large scale.

Yuka is the most downloaded app in the Wellness and Health category on the main digital platforms. It is a simple and intuitive tool, designed by Julie Chapon, François Martin and Benoit Martin in 2017. It is currently used by 40 million people, speaks twelve languages and evaluates 45 products per second.

Reading labels without a chemistry degree

Yuka scans the barcode of food and cosmetic products and gives an assessment of the possible impact on health and the environment. And obviouslyit allows you to make an informed choiceon those judged as 'mediocre' or 'poor'. The evaluation is based on scientific databases and research already published and traceable.

The parameters described concern the possible effects on the human endocrine system, on the carcinogenic, allergic and irritant risk.

They signal the presence of microplastics or other intrinsically polluting substances. And it warns when a substance on the product label involves production cycles that are not demonstrably sustainable, such as the production of most petroleum derivatives. An interesting detail: even natural ingredients, such as some essential oils or phytoextracts, are often evaluated negatively.

A tool to ask for more quality from food multinationals

Lately, Yuka's ratings on food and cosmetic products, from poor to excellent, also appear on the brands' e-commerce platforms.

And, above all, Yuka is acquiring more and more influence and authority, so much so that it convinces 92% of users to put a product back on the shelf if the rating is not sufficient.

The business model is interesting and honest: it is possible to pay for a subscription by choosing between different amounts, depending on how much you value the app. And this allows Yuka to be economically sustainable and, above all, independent.