Skip to main content

This New AI Predicts Your Health Risks Better Than Just Your BMI

New tool identifies who needs weight-loss meds most. With 2/3 English adults overweight or obese, this could revolutionize care by targeting those at highest risk of obesity-related diseases.

Sophia Brennan
Sophia Brennan
·2 min read·United Kingdom·4 views

Originally reported by The Guardian Science · Rewritten for clarity and brevity by Brightcast

Why it matters: This tool empowers the NHS to prioritize life-changing weight-loss interventions for those most at risk, improving health outcomes and quality of life for countless individuals.

For years, doctors have looked at your Body Mass Index (BMI) and a couple of health issues to figure out if you're at risk for obesity-related illnesses. It's a bit like judging a book by its cover, or maybe just its page count. But now, UK researchers have cooked up a new AI tool that dives a whole lot deeper, promising to identify who truly needs those coveted weight-loss treatments.

Because, let's be real, two-thirds of adults in England are either overweight or obese. That's a lot of people, and the NHS doesn't exactly have an unlimited supply of those fancy weight-loss injections.

Beyond the BMI

Enter Obscore, a new tool developed by a team at the University of Cambridge and Queen Mary University of London. Their goal isn't to hand out more treatments, but to make sure the right people get them. As Professor Nick Wareham, one of the study's authors, put it: "give therapy to those who need it most and will benefit the most."

Wait—What is Brightcast?

We're a new kind of news feed.

Regular news is designed to drain you. We're a non-profit built to restore you. Every story we publish is scored for impact, progress, and hope.

Start Your News Detox

Published in Nature Medicine, their method involved an "interpretable machine learning" AI — which sounds fancy, but essentially means it's an AI that can explain why it made its decisions. This AI crunched data from nearly 200,000 individuals in the UK Biobank, all with a BMI of 27 or higher.

The AI sniffed out 20 different health, lifestyle, and demographic factors — things like age, sex, total cholesterol, and even creatinine levels — that predict your risk for 18 different obesity-related conditions over a decade. We're talking everything from gout to stroke. For each condition, Obscore could sort people into five distinct risk groups, from a chill 'low' to a slightly more concerning 'high.'

The Plot Twist for Overweight Folks

Here's where it gets interesting: the Obscore tool showed that people with the exact same age, sex, and BMI can have wildly different risks for various conditions. It's almost like our bodies are more complex than a simple height-to-weight ratio would suggest. Who knew?

Even more surprising, for conditions like type 2 diabetes, a significant chunk of people in the highest risk group were merely overweight, not obese. Kamil Demircan, a co-author, points out that these are precisely the individuals who might slip through the cracks if doctors are only looking at BMI. Basically, your phone might have a better handle on your health risks than your old-school doctor's chart.

Now, before you ask your GP to run your data through an AI, Naveed Sattar, a professor at the University of Glasgow not involved in the study, notes that some of the measurements Obscore uses aren't routinely available on the NHS. He called the work a "thoughtful attempt" but stressed that it needs more development and testing before it makes its way into your annual check-up.

Still, it's a pretty compelling step toward a future where healthcare is less about broad categories and more about your specific, wonderfully complex self. Which, if you think about it, is both impressive and slightly terrifying. Because apparently that's where we are now.

Brightcast Impact Score (BIS)

This article describes the development of a new AI-powered tool to identify individuals most at risk of obesity-related diseases, which is a positive action in health innovation. The tool offers a more personalized approach to resource allocation for weight-loss interventions, potentially benefiting many people. The research is published in a reputable journal and uses a large dataset, indicating strong evidence and potential for widespread impact.

Hope29/40

Emotional uplift and inspirational potential

Reach24/30

Audience impact and shareability

Verification24/30

Source credibility and content accuracy

Significant
77/100

Major proven impact

Start a ripple of hope

Share it and watch how far your hope travels · View analytics →

Spread hope
You
friendstheir friendsand beyond...

Wall of Hope

0/20

Be the first to share how this story made you feel

How does this make you feel?

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

Connected Progress

Sources: The Guardian Science

More stories that restore faith in humanity