They looked at nearly 1,000 people, with an average age of 55, accounting for all kinds of other factors that could potentially sway the results, such as age and how physically active they were.
Magnetic resonance imaging (MRI) was then used to determine the brain volumes of participants in different parts of the brain.
About 1,000 participants with high BMI and waist-to-hip ratios had the lowest average of grey matter in the brain, which controls self-control, muscle control and sensory perception, compared to about 3,000 participants of healthy weights who had an average amount of grey matter.
"Our research looked at a large group of people and found obesity, specifically around the middle, may be linked with brain shrinkage". People in the middle, with a high BMI but without high waist-to-hip ratio, had an of 793 cubic centimeters.
"The take-home message is that being overweight and obese has a multitude of effects on health, so it's unsurprising that obesity is also going to have an effect on our brain health", says Mark Hamer, professor of exercise medicine at Loughborough University in England and lead author of the study.
To find that out, researchers would have to follow the same people over time and record changes in body weight and brain volume.
Grey matter in the brain consists mostly of nerve cells while "white matter" is made up of connecting nerve fibres. "Brain gray matter shrinkage.seems to be associated with obesity and with increased visceral fat", she said.
The study found that people with both a high BMI and high waist-to-hip ratio had the lowest brain volumes, compared with people who had just a high BMI (but not a high waist-to-hip ratio) and people of a healthy weight.
Belly fat has always been thought to be particularly bad for your heart, but now, a new study adds more evidence to the idea that it may also be bad for your brain.
A limitation of the study was that only 5% of those invited to participate ended up taking part, the researchers pointed out.
In other words, these results might not be true for the overall population.