The reason you reach for food could determine whether Ozempic helps you lose weight, according to new research that challenges assumptions about who benefits most from the popular diabetes medication.
A year-long study of 92 people with type 2 diabetes in Japan found that those who eat in response to visual or aromatic food cues, seeing a tempting pastry display or smelling fresh pizza, were significantly more likely to lose weight on GLP-1 receptor agonists like Ozempic than those who eat for emotional reasons.
The Science Behind Food Behaviors
Researchers tracked three distinct eating patterns: external eating (responding to food’s appearance or smell), emotional eating (eating due to stress, sadness, or anxiety), and restrained eating (consciously limiting food intake for weight control). The findings, published in Frontiers in Clinical Diabetes and Healthcare, suggest that understanding why people overeat could help doctors predict treatment success.
“Pre-treatment assessment of eating behavior patterns may help predict who will benefit most from GLP-1 receptor agonist therapy,” said Prof Daisuke Yabe of Kyoto University, senior author of the study.
The research involved participants taking various GLP-1 medications including liraglutide, dulaglutide, and both oral and injectable semaglutide. Scientists measured their weight, blood sugar levels, and eating behaviors at the start of treatment, then again at three months and one year.
Overall, participants saw meaningful improvements—average weight loss, better cholesterol levels, and reduced body fat percentage. Blood glucose levels improved too, though less dramatically. But the most intriguing discovery emerged when researchers analyzed who responded best to treatment.
People with high “external eating” scores at the beginning experienced the greatest weight loss after 12 months. These are individuals who typically eat when they see appetizing food displays, smell cooking aromas, or notice others eating—even when they’re not particularly hungry.
Why Emotional Eaters Struggled More
The contrast with emotional eaters was stark. While external eating behaviors decreased steadily throughout the year on GLP-1 therapy, emotional eating patterns proved more stubborn. These behaviors dropped initially but returned to baseline levels by the study’s end.
“One possible explanation is that emotional eating is more strongly influenced by psychological factors which may not be directly addressed by GLP-1 receptor agonist therapy,” said Dr Takehiro Kato of Gifu University, second author of the study.
This finding aligns with emerging neuroscience research showing that GLP-1 medications work partly by dampening brain responses to food cues. When people with obesity view appetizing food images, their brains show hyperactivity in reward regions like the striatum and orbitofrontal cortex. GLP-1 drugs appear to quiet this neural excitement—but this mechanism may be less effective for eating driven by emotional states rather than external triggers.
The implications extend beyond academic curiosity. With GLP-1 medications costing thousands of dollars annually and not everyone responding equally well, identifying likely responders becomes increasingly important. Currently, doctors prescribe these medications based primarily on blood sugar control and weight status, with little consideration of psychological eating patterns.
The Japanese study population offers particular insights because type 2 diabetes in East Asian countries typically involves less obesity than in Western populations, yet GLP-1 medications are increasingly prescribed for younger, overweight individuals in these regions.
Interestingly, the research found that people who practice “restrained eating”—consciously controlling their diet for weight management—didn’t show significantly different outcomes. This suggests that conscious dietary control, while important for health, doesn’t predict medication response the way unconscious food triggers do.
The study’s real-world design strengthens its relevance. Unlike controlled clinical trials, participants continued their normal lives while being monitored, making the results more applicable to typical patients. However, the observational nature means the research can’t definitively prove causation, and the relatively motivated study population might not represent all diabetes patients.
Future research will need to validate these findings in larger, more diverse populations before doctors can routinely use eating behavior assessments to guide treatment decisions. The researchers suggest that simple questions about whether patients tend to eat in response to appealing food versus emotional stress could eventually become part of standard pre-treatment evaluation.
For now, the research offers hope for more personalized diabetes treatment. Rather than a one-size-fits-all approach, understanding individual relationships with food could help optimize expensive GLP-1 therapy, ensuring it reaches those most likely to benefit while identifying patients who might need additional psychological support alongside medication.
Frontiers in Clinical Diabetes and Healthcare: 10.3389/fcdhc.2025.1638681
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.