GLP 1 food cravings fall in a small Italian study of semaglutide and tirzepatide

Adults who started semaglutide or tirzepatide in routine outpatient care in Italy reported fewer sweet and savory cravings after three months, according to a pilot study published in Nutrients on 5 November. The 38 person cohort adds small, self reported data on GLP 1 food cravings, alongside an average weight loss of 7.2 kg. There was no control group and the questionnaire was not validated, so the paper cannot show that the drugs caused the changes. Neither molecule moves on our board, since both are already approved obesity medicines.

Inside the Nutrients pilot study

The study followed 38 adults with obesity who did not have diabetes and had never used an incretin drug. Entry required a body mass index of at least 30, or at least 27 with a weight related illness. Doctors chose the drug as part of routine care, so 21 people started semaglutide, a GLP 1 receptor agonist, and 17 started tirzepatide, which also activates the GIP receptor. Nobody was randomized, and there was no untreated comparison group. Mean age was 56 years, about three in four participants were women, and mean BMI was 32.7. The tirzepatide group started heavier, with a mean BMI of 35.4 against 30.5 for semaglutide, the paper reports, a gap that matters for any comparison.

The work came out of IRCCS San Raffaele Roma and San Raffaele Open University, with partners at Campus Bio-Medico University of Rome and Queen Mary University of London. Weight, BMI and blood pressure were measured at baseline and again after three months. Eating habits came from a questionnaire the team adapted for this study and had not formally validated, a key caveat for any claim about GLP 1 food cravings. It asked about cravings, fullness after meals, hunger between meals, meal skipping, night eating and how often people ate foods such as legumes, fish or processed meat. Answers were scored on 10 point scales. The Italian Ministry of Health funded the work, the protocol is registered as NCT07229170, and the authors declared no conflicts of interest.

Eating behavior shifted beyond appetite alone

On a scale of zero to ten, mean sweet craving fell from 5.2 to 3.3 and savory craving from 6.9 to 4.6, the paper in Nutrients reports. Difficulty resisting hunger between meals dropped from 5.9 to 3.1, while fullness after meals rose from 6.0 to 7.5. People also ate fewer times a day, 3.6 meals against 4.3 at the start, and daily fluid intake did not change. These are the clearest GLP 1 food cravings results in the study, and every one rests on what participants said about themselves. Measured outcomes moved as well. Mean weight went from 91.1 kg to 83.9 kg, about 8 percent of starting weight, and systolic blood pressure fell by 6.1 mm Hg.

More than eight in ten participants, 84.2 percent, reported a substantial cut in meal portion size, a far larger share than reported any change in taste preference.

Portion size moved most clearly. More than eight in ten participants, 84.2 percent, reported a substantial cut in meal portion size, a far larger share than reported any change in taste preference. About a quarter, 26.3 percent, said they now preferred savory food over sweet, 7.9 percent moved toward sweet, and 65.8 percent noticed no shift. The share of regular processed meat eaters fell by 13.2 percentage points and the share eating fresh cheese by 7.9 points. Regular eating of raw vegetables, nuts, legumes, fish and white meat rose by 5.3 to 7.9 points. The authors caution that these shifts do not show that overall diet quality improved, and that food groups were tracked as habits rather than measured amounts.

Semaglutide vs tirzepatide showed no clear gap

Weight loss was almost identical in the two arms, 7.2 kg on semaglutide and 7.3 kg on tirzepatide. BMI fell by 2.4 and 3.0 points, and no difference in weight, blood pressure or eating behavior between the drugs reached statistical significance. That is not evidence the two medicines work equally well. The groups were small, they were not randomized, and they differed at baseline, with the tirzepatide group starting at a higher BMI. Dose records at follow up were complete for only 23 people, according to the paper, so the arms cannot be matched on drug exposure either. Settling the semaglutide vs tirzepatide question for eating behavior would need a comparison designed and powered for it, which the authors call for.

Side effects were common in both groups. The mean score for nausea or taste changes was 4.8 out of 10, and 55.3 percent of participants scored 5 or higher. Scores were numerically higher on semaglutide, 5.5 against 3.9, but the gap was not statistically significant, and symptom scores showed no link with weight loss. Participants also reported a moderate fall in the pleasure they took from food. The one behavior that did track weight was hunger control: people whose difficulty resisting hunger fell most tended to lose more, with a correlation of 0.403. Our earlier report on a GLP1R gene variant raised a related question about why nausea and weight loss vary so much from person to person.

What the pilot study does not settle

The authors of the Nutrients pilot study list the limits themselves. Without a control group, the changes cannot be pinned on the drugs alone, since lifestyle changes, the act of joining a study and other unmeasured factors may have played a part. The sample is small, follow up lasted only three months, and the questionnaire results are exploratory. The team also raised a practical point. If portions shrink sharply, intake of protein, fiber, fluids and micronutrients may fall, and lean mass can be lost along with fat, so they argue for structured nutrition checks during treatment. Self reported intake was never compared with measured calories, which means the real size of any dietary change in this group is still unknown.

For the board, this is a flat signal. Semaglutide and tirzepatide are both approved, and the paper adds no efficacy or safety finding that would change a label or a filing. Its value is narrower: an early look at how GLP 1 food cravings and other eating habits move during the first months of routine care. Behavior is becoming a steady theme in real world research on these drugs, including a national survey on drinking during GLP 1 treatment that we covered last week. Our claims based story on stopping semaglutide early showed how tolerability shapes who stays on therapy. Controlled trials with validated questionnaires and measured intake are the next step the authors name.

Conclusion

This Italian pilot gives early, self reported evidence that GLP 1 food cravings, hunger between meals and portion sizes all fell during the first three months of semaglutide or tirzepatide treatment. Weight dropped by 7.2 kg on average, with no clear difference between the two drugs. The design cannot show cause, and the questionnaire is unvalidated. Both molecules stay approved and unmoved on our board. The next checkpoint is a controlled study with validated tools and measured food intake. This report is research news, not medical advice.