University of Cambridge researchers use AI to assess food outlets' healthiness across UK

Scientists at the University of Cambridge have employed cutting-edge technology to assess the healthiness of menus in cafes, takeaways and restaurants throughout the UK.

This research, outlined in Health & Place, highlighted the disparities in food environments across local authorities, putting a focus on the challenges faced by those that live in deprived areas, where there are fewer healthy food outlets.

The study, led by Yuru Huang - a Gates Cambridge Scholar at the Medical Research Council (MRC) Epidemiology Unit - scrutinised nearly 55,000 food outlets listed on the online food delivery platform, Just Eat.

Each outlet's menu was evaluated on a scale of 0 to 12, with higher scores indicating healthier options. But what's fascinating here is that the team used artificial intelligence (AI) via a deep learning model that was trained on a subset of Just Eat data. This helped to predict the healthiness of nearly 180,000 out-of-home food outlets across the country.

A 'triple burden'

The scientists' findings uncovered a disheartening reality - that neighbourhoods grappling with deprivation have a greater density of food outlets, predominantly offering less healthy foods.

Notably, areas with the highest deprivation levels were home to over double the number of food outlets in more affluent regions. There was also a discernible correlation between deprivation levels and the healthiness of menus, with outlets in deprived areas generally offering less healthy options across all categories, from cafés to fast food joints.

"There's a clear pattern between the healthiness of menus at out-of-home food outlets in an area and its level of deprivation," said Huang. "This can create a ‘double burden’ for people living in deprived neighbourhoods, where there are more outlets and these tend to be less healthy, compared to less deprived neighbourhoods.

“On top of this, there are studies that show, for example, that people with the lowest income were more likely to be obese when living in areas with a high proportion of fast-food outlets. This could even create a ‘triple burden’ for people living in these areas.”

Despite the robustness of their AI model in predicting menu healthiness based on outlet names and hygiene ratings, the researchers acknowledge its limitations in capturing nuanced factors like portion sizes and cooking methods.

Nevertheless, their work sheds light on the need for targeted interventions to address disparities in food environments and reduce the associated health inequalities.