1. BACKGROUND
Tirzepatide is a glucose‐dependent insulinotropic polypeptide and glucagon‐like peptide‐1 (GLP‐1) receptor agonist approved for treatment of type 2 diabetes and weight management. In a 72‐week, phase 3 trial in adults with obesity, treatment with tirzepatide (5, 10, or 15 mg once weekly) led to substantial weight loss (up to 20.9%) compared to placebo (3.1%). 1 Reductions in appetite and food intake during treatment with tirzepatide have been found to exceed those observed in participants receiving liraglutide 2 or placebo, 2 , 3 and energy restriction is the primary mechanism driving tirzepatide's effects on body weight loss. 2 , 3 However, more research is needed to better understand how tirzepatide affects ingestive behaviour to promote energy restriction. Herein, we report changes in food preferences and food cravings following treatment with tirzepatide and placebo in a phase 1 obesity trial.
2. METHODS
Adults with obesity (BMI 30–45 kg/m2) and at least one pre‐specified metabolic impairment were recruited to a blinded, randomised controlled trial to observe tirzepatide's effects over an 18‐week treatment period. Details of the study design, primary outcomes, and adherence to ethical and regulatory requirements have been published (NCT04081337). 3
A total of 55 adults were randomised to receive once weekly tirzepatide 15 mg (n = 27) or placebo (n = 28). Tirzepatide 15 mg was attained via an 8‐week stepwise dose escalation. All participants underwent nutritional counselling to promote a low‐calorie diet intended to achieve 10% (±2%) weight loss in approximately 14 weeks. Ingestive behaviour profiles were assessed at baseline, week 8, and week 18. The main study findings are reported elsewhere 3 and include changes in body weight, energy intake, and appetite from baseline to 18 weeks.
Food preferences and food cravings were measured using the Food Preference Questionnaire (FPQ) 4 and Food Craving Inventory (FCI), 5 respectively. The FPQ (72 items) assesses liking of specific food items that vary in macronutrients on a 9‐point Likert scale ranging from ‘dislike extremely’ to ‘like extremely’. Food preference scores are calculated for foods in a 2 (high fat, low fat) by 3 (high simple sugar, high‐complex carbohydrate, low‐complex carbohydrate/high protein) matrix, resulting in 11 outcome variables comprising the five main effects listed in each row and column (e.g., high fat, low fat, high simple sugar) and the six cells within the matrix (e.g., high fat‐high simple sugar, low fat‐high simple sugar). Additionally, a fat preference score is calculated as the mean hedonic rating for high‐fat foods divided by the mean hedonic rating for low‐fat foods. This ratio is then multiplied by 100 such that final scores over 100 reflect a high fat preference. The FCI (33 items) assesses the frequency of cravings for five specific food groups (high fat, sweets, carbohydrates/starches, fast food fats, or fruits and vegetables) on a 5‐point Likert scale ranging from ‘never’ to ‘always’. Overall FCI and subscale scores are calculated as the respective total mean and mean of items within each of the five categories.
Least squares mean change from baseline to 18 weeks in FPQ and FCI scores was compared between the tirzepatide and placebo groups using a mixed model for repeated measures (MMRM). Correlation analysis between changes from baseline in body weight and changes in FPQ or FCI score was carried out in the tirzepatide and placebo groups independently. The test p‐values were not adjusted for multiple testing.
3. RESULTS
Participants' (84% female, mean weight = 102.8 kg, mean BMI = 37.1 kg/m2) baseline characteristics were well balanced across treatment groups. 3 Forty‐eight participants completed the study (tirzepatide, n = 24; placebo, n = 24).
As previously reported, 18 weeks of tirzepatide treatment produced significantly greater weight reduction (−16.7 kg [17%]) versus placebo (−8.3 kg [8%]). 3 Additionally, tirzepatide significantly reduced energy intake during ad libitum lunch and dinner meals by an additional 856 kcal and significantly reduced retrospective overall appetite, measured using visual analogue scales, versus placebo. 3
Concomitant with the above findings, tirzepatide significantly decreased 10 of the 12 food preference scores and 4 of the 6 food craving scores versus placebo (p < 0.05, for all) after 8 and 18 weeks of treatment (Table 1). Significant reductions in preferences for foods high in fat and simple sugar (p < 0.01) as well as numerical, though non‐significant, reductions in fat preference scores (p = 0.0594) were observed after 18 weeks of tirzepatide versus placebo (Figure 1A). Tirzepatide significantly decreased the overall FCI score and cravings for sweets, carbohydrates/starches, and fast‐food fats (p < 0.05), but not high fat or fruits and vegetables versus placebo (Figure 1B). Effects on food preference and craving scores were attenuated over the treatment period; however, changes observed in the tirzepatide group remained lower than placebo at week 18.
TABLE 1.
Food Preferences Questionnaire and Food Craving Inventory Scores.
| Placebo (N = 28) | Tirzepatide (N = 27) | Estimated treatment Difference (95% CI), p‐value | |
|---|---|---|---|
| Food Preference Questionnaire Scores | |||
| High fat | |||
| Baseline | 5.68 (1.53) | 5.71 (1.26) | |
| Change at week 8 | −0.57 (0.28) | −1.56 (0.29) | −1.00 (−1.80, −0.20), p = 0.0157 |
| Change at week 18 | −0.23 (0.27) | −1.30 (0.27) | −1.07 (−1.84, −0.30), p = 0.0072 |
| Low fat | |||
| Baseline | 5.28 (1.33) | 4.95 (1.09) | |
| Change at week 8 | −0.23 (0.23) | −0.96 (0.24) | −0.73 (−1.41, −0.06), p = 0.0343 |
| Change at week 18 | −0.08 (0.22) | −0.83 (0.23) | −0.75 (−1.40, −0.11), p = 0.0234 |
| High simple sugar | |||
| Baseline | 5.30 (1.63) | 5.05 (1.32) | |
| Change at week 8 | −0.41 (0.26) | −1.25 (0.27) | −0.83 (−1.59, −0.07), p = 0.0331 |
| Change at week 18 | −0.10 (0.27) | −1.16 (0.27) | −1.06 (−1.82, −0.29), p = 0.0075 |
| High‐complex CHO | |||
| Baseline | 5.17 (1.41) | 4.87 (1.09) | |
| Change at week 8 | −0.43 (0.25) | −1.27 (0.26) | −0.84 (−1.57, −0.12), p = 0.0236 |
| Change at week 18 | −0.12 (0.23) | −0.99 (0.23) | −0.87 (−1.53, −0.21), p = 0.0106 |
| Low CHO/High protein | |||
| Baseline | 5.98 (1.36) | 6.07 (1.26) | |
| Change at week 8 | −0.31 (0.26) | −1.28 (0.27) | −0.97 (−1.73, −0.20), p = 0.0139 |
| Change at week 18 | −0.21 (0.26) | −1.05 (0.26) | −0.84 (−1.57, −0.10), p = 0.0262 |
| High‐fat–high simple sugar | |||
| Baseline | 5.20 (2.06) | 5.36 (1.64) | |
| Change at week 8 | −0.61 (0.33) | −1.56 (0.34) | −0.95 (−1.90, 0.00), p = 0.0494 |
| Change at week 18 | −0.11 (0.31) | −1.45 (0.32) | −1.34 (−2.23, −0.46), p = 0.0037 |
| High‐fat–high‐complex CHO | |||
| Baseline | 5.38 (1.39) | 5.11 (1.31) | |
| Change at week 8 | −0.57 (0.27) | −1.49 (0.28) | −0.91 (−1.71, −0.12), p = 0.0249 |
| Change at week 18 | −0.20 (0.26) | −1.18 (0.27) | −0.99 (−1.74, −0.23), p = 0.0113 |
| High fat–low CHO/high protein | |||
| Baseline | 6.48 (1.57) | 6.65 (1.48) | |
| Change at week 8 | −0.52 (0.27) | −1.64 (0.28) | −1.12 (−1.89, −0.34), p = 0.0056 |
| Change at week 18 | −0.38 (0.28) | −1.26 (0.28) | −0.88 (−1.68, −0.08), p = 0.0321 |
| Low‐fat–high simple sugar | |||
| Baseline | 5.39 (1.37) | 4.77 (1.60) | |
| Change at week 8 | −0.27 (0.23) | −0.91 (0.24) | −0.64 (−1.32, 0.03), p = 0.0600 |
| Change at week 18 | −0.12 (0.25) | −0.84 (0.26) | −0.72 (−1.46, 0.01), p = 0.0533 |
| Low‐fat–high‐complex CHO | |||
| Baseline | 4.95 (1.64) | 4.62 (1.07) | |
| Change at week 8 | −0.29 (0.24) | −1.07 (0.25) | −0.78 (−1.47, −0.08), p = 0.0302 |
| Change at week 18 | −0.05 (0.21) | −0.81 (0.22) | −0.75 (−1.37, −0.14), p = 0.0174 |
| Low fat–low CHO/high protein | |||
| Baseline | 5.49 (1.42) | 5.48 (1.38) | |
| Change at week 8 | −0.11 (0.28) | −0.92 (0.29) | −0.81 (−1.62, 0.00), p = 0.0493 |
| Change at week 18 | −0.04 (0.25) | −0.84 (0.26) | −0.80 (−1.53, −0.07), p = 0.0315 |
| Fat preference | |||
| Baseline | 108.21 (17.59) | 118.16 (28.83) | |
| Change at week 8 | −5.84 (3.11) | −14.44 (3.23) | −8.60 (−17.65, 0.44), p = 0.0619 |
| Change at week 18 | −3.05 (3.01) | −11.34 (3.03) | −8.28 (−16.91, 0.34), p = 0.0594 |
| Food Craving Inventory Scores | |||
| Overall score | |||
| Baseline | 2.80 (0.51) | 2.49 (0.64) | |
| Change at week 8 | −0.41 (0.10) | −0.71 (0.10) | −0.30 (−0.59, −0.01), p = 0.0446 |
| Change at week 18 | −0.17 (0.09) | −0.52 (0.09) | −0.35 (−0.61, −0.09), p = 0.0098 |
| High fat | |||
| Baseline | 2.76 (0.60) | 2.56 (0.69) | |
| Change at week 8 | −0.44 (0.11) | −0.69 (0.12) | −0.25 (−0.57, 0.08), p = 0.1394 |
| Change at week 18 | −0.30 (0.09) | −0.44 (0.09) | −0.14 (−0.41, 0.13), p = 0.3085 |
| Sweets | |||
| Baseline | 2.61 (0.68) | 2.32 (0.76) | |
| Change at week 8 | −0.41 (0.12) | −0.76 (0.12) | −0.35 (−0.69, −0.02), p = 0.0395 |
| Change at week 18 | −0.15 (0.11) | −0.63 (0.11) | −0.49 (−0.81, −0.16), p = 0.0041 |
| CHO/Starches | |||
| Baseline | 2.73 (0.78) | 2.37 (0.78) | |
| Change at week 8 | −0.35 (0.11) | −0.75 (0.11) | −0.40 (−0.71, −0.09), p = 0.0122 |
| Change at week 18 | −0.09 (0.11) | −0.52 (0.11) | −0.43 (−0.73, −0.12), p = 0.0068 |
| Fast food fats | |||
| Baseline | 3.16 (0.62) | 2.98 (0.66) | |
| Change at week 8 | −0.52 (0.14) | −0.94 (0.15) | −0.42 (−0.83, −0.01), p = 0.0435 |
| Change at week 18 | −0.24 (0.13) | −0.63 (0.13) | −0.39 (−0.76, −0.02), p = 0.0384 |
| Fruits and vegetables | |||
| Baseline | 2.98 (0.64) | 2.43 (0.96) | |
| Change at week 8 | −0.40 (0.11) | −0.44 (0.11) | −0.04 (−0.36, 0.29), p = 0.8285 |
| Change at week 18 | −0.09 (0.11) | −0.38 (0.11) | −0.29 (−0.61, 0.02), p = 0.0640 |
Note: Data are mean (standard deviation) at baseline and least squares mean (standard error) derived from a MMRM for change from baseline. The MMRM model included treatment, time, and treatment‐by‐time interaction as fixed effects, baseline questionnaire parameter as a covariate, and patient as a random effect. An unstructured covariance structure was used to model the covariance between a participant's observations across days. At week 8, n = 28 (placebo) and 26 (tirzepatide). At week 18, n = 24 (placebo) and 24 (tirzepatide). CHO, carbohydrates; CI, confidence interval; MMRM, mixed‐model for repeated measures; n, number of participants included in analysis, N, number of participants who were randomised and received at least one dose of study treatment.
FIGURE 1.
Changes from baseline (Day −3) in (A) food preference scores and (B) food craving scores at week 18 by treatment group. (C) Correlation between change from baseline in body weight and overall food craving score at week 18 by treatment group. In (A) and (B), data are least square means (standard error) derived from a mixed model for repeated measures. CHO, carbohydrates; PBO, placebo; TZP, tirzepatide. *p < 0.05, **p < 0.01 for tirzepatide versus placebo.
Changes in food preferences demonstrated modest, though not significant, associations with weight change in the tirzepatide group (Table S1). Changes in overall food cravings (Figure 1C) as well as cravings for high fats and sweets following 18 weeks of tirzepatide significantly correlated with body weight change (Table S1), with larger reductions in cravings being associated with larger decreases in weight. In the tirzepatide group, correlations between change in weight and change in cravings for carbohydrates/starches (correlation coefficient 0.38) and fast‐food fats (0.37) were of medium magnitude, but not significant (p = 0.0660 and 0.0789, respectively). There were no significant correlations between change in weight and change in food preferences or cravings in the placebo group.
4. CONCLUSIONS
In the current analysis, body weight, energy intake, and appetite reductions observed with tirzepatide occurred along with significant decreases in food preferences and food cravings. These findings provide further evidence that tirzepatide blunts the appetitive drive to eat and hedonic responses to foods. These results help explain tirzepatide's robust efficacy at reducing energy intake and body weight compared to diet alone, which produced half the weight loss (16.7 vs. 8.3 kg at week 18). 3 These mechanisms have been empirically supported for weight loss achieved with the selective GLP‐1 receptor agonist, semaglutide (−5 kg), where concurrent reductions in cravings for energy‐dense foods were observed after 12 weeks of treatment. 6 Food cravings have been found to decrease over a 12‐week very low‐calorie diet that produced 16.7% weight loss, while a less intense low‐calorie diet yielding 6.6% weight loss had little impact on food cravings. 7 Also, preferences and cravings for high‐carbohydrate and high‐fat foods respectively decreased following low‐carbohydrate and low‐fat 2‐year diets. 8 Results from our present analysis reinforce the finding that larger reductions in weight and energy intake are associated with greater reductions in appetitive markers (e.g., appetite, food cravings, food preferences).
The positive correlations we observed between weight change and change in cravings in the tirzepatide group indicate that tirzepatide treatment may contribute to greater weight loss via changes in cravings to certain foods. Further research is needed to delineate the relationships between cravings and body weight changes.
Of note, high fat craving scores did not decrease differentially across groups, whereas tirzepatide significantly reduced cravings for fast food fats versus placebo. This suggests that tirzepatide decreased cravings for fast food fat subscale items like pizza, hamburgers, and chips more than for high fats subscale items like sausage and fried fish.
There were several study limitations: (1) the sample was disproportionately female, (2) calorie intake throughout the study could not be controlled, contributing to differential weight loss between treatment groups, and (3) the study was not powered for these non‐primary analyses.
In people with obesity, 18 weeks of treatment with tirzepatide resulted in significant decreases in food preferences and food cravings relative to placebo with dietary restriction. These decreases may have contributed to the greater reductions in weight and energy intake with tirzepatide, even though both groups had the same weight loss goals and support from an interventionist. These findings are relevant given the complexities associated with regulating food consumption and body weight in an environment where food is readily available.
AUTHOR CONTRIBUTIONS
Edward J. Pratt, Zvonko Milicevic, Axel Haupt, and Tamer Coskun contributed to the study design. Eric Ravussin, Guillermo Sanchez‐Delgado, and Corby K. Martin contributed to the data acquisition. Eric Ravussin, Guillermo Sanchez‐Delgado, Hiroshi Nishiyama, Hui‐Rong Qian, Zvonko Milicevic, and Corby K. Martin contributed to data analysis. Hiroshi Nishiyama and Hui‐Rong Qian conducted the statistical analyses. Sam F. Kennedy, Alastair Knights, Eric Ravussin, Guillermo Sanchez‐Delgado, Edward J. Pratt, Zvonko Milicevic, Axel Haupt, Tamer Coskun, and Corby K. Martin contributed to the interpretation of results. All authors participated in the drafting, critical revision, and approved the final version of the manuscript.
FUNDING INFORMATION
This study was funded by Eli Lilly and Company.
CONFLICT OF INTEREST STATEMENT
Sam F. Kennedy has no conflicts of interest to disclose. Eric Ravussin has received grants or contracts from Amazon, Eli Lilly and Company, ICON, Novartis, and Sanofi; has received consulting fees from Altimmune, Amway (Nutrilite Health), Eli Lilly and Company, Energesis Pharmaceuticals, Generian, Kintai Therapeutics, Merck, Novo Nordisk, and YSOPIA Bioscience (LNC Therapeutics); has received honouraria from National Institutes of Health (NORC Review) and Open Academy of Medicine; and has received support for attending meetings from Aegean Conference on Precision Nutrition, American Association of Clinical Endocrinology, Clermont Auvergne University, Eli Lilly and Company Obesity Working Group, Frontiers in Obesity, Diabetes, and Metabolism Seminar, Les Journées Françaises de Nutrition, Lipedema Foundation Retreat, National Institutes of Health (NORC Review), New York University Grand Rounds, The Obesity Society, Quebec Society of Lipidology, Nutrition and Metabolism, RACMEM, Société Francophone du Diabète, and Tulane Personalized Health Institute. Guillermo Sanchez‐Delgado has received support for attending meetings and travel from Eli Lilly and Company. Over the past 3 years, Corby K. Martin served on advisory boards for EHE Health and Wondr Health; received compensation to facilitate continuing education events for Commission on Dietetic Registration; received royalties from ABGIL (paid to his institution); received consulting compensation from FoodMinds/Almond Board of California; received research support from Eli Lilly and Company, Foundation for Food and Agriculture Research/Zero Waste Foundation, Henry M. Jackson Foundation for the Advancement of Military Medicine, Leona M. and Harry B. Helmsley Charitable Trust, National Institute for Health and Care Research, National Institutes of Health, National Science Foundation, State of Louisiana Federal American Rescue Plan, United States Department of Agriculture, University of Rochester, American Diabetes Association, University of Nebraska Lincoln, Spoonified, Boehringer Ingelheim, Brown University, PCORI, and WW International; and receives honouraria, complimentary registration, and/or travel reimbursement for giving presentations and other academic activities from Bray Course Planning Committee, Brigham Young University, Duke University, Indiana University Bloomington, University of Alabama at Birmingham, University of Kansas Medical Center, University of Nebraska‐Lincoln, University of Southern California, Wake Forest University, University of Oklahoma, American Society for Nutrition, and The Obesity Society. Alastair Knights, Hiroshi Nishiyama, Hui‐Rong Qian, Edward J. Pratt, Axel Haupt, and Tamer Coskun are employees and shareholders of Eli Lilly and Company. Zvonko Milicevic is a shareholder and former employee of Eli Lilly and Company.
PEER REVIEW
The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.70063.
Supporting information
Table S1. Correlation between body weight change from baseline and changes from baseline in Food Preferences Questionnaire or Food Craving Inventory Scores.
ACKNOWLEDGEMENTS
The authors thank the participants and study site personnel involved in the study.
Kennedy SF, Knights A, Ravussin E, et al. Impact of tirzepatide treatment on participant‐reported food craving and food preference: Secondary analyses of a phase 1 randomised controlled trial in people with obesity with dietary restriction. Diabetes Obes Metab. 2025;27(11):6784‐6789. doi: 10.1111/dom.70063
Sam F. Kennedy and Alastair Knights should be considered joint first authors.
DATA AVAILABILITY STATEMENT
Eli Lilly and Company provides access to all individual participant data collected during the trial, after anonymisation, except for pharmacokinetic or genetic data. Data are available to request 6 months after the indication studied has been approved in the United States and European Union and after primary publication acceptance, whichever is later. No expiration date for data requests is currently set once data have been made available. Access is provided after a proposal has been approved by an independent review committee identified for this purpose and after receipt of a signed data‐sharing agreement. Data and documents, including the study protocol, statistical analysis plan, clinical study report, and blank or annotated case report forms, will be provided in a secure data‐sharing environment. For details on submitting a request, see the instructions provided at www.vivli.org.
REFERENCES
- 1. Jastreboff AM, Aronne LJ, Ahmad NN, et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387(3):205‐216. doi: 10.1056/NEJMoa2206038 [DOI] [PubMed] [Google Scholar]
- 2. Martin CK, Carmichael OT, Carnell S, et al. Tirzepatide on ingestive behavior in adults with overweight or obesity: a randomized 6‐week phase 1 trial. Nature Medicine. 2025; doi:10.1038/s41591‐025‐03774‐9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ravussin E, Sanchez‐Delgado G, Martin CK, et al. Tirzepatide did not impact metabolic adaptation in people with obesity, but increased fat oxidation. Cell Metab. 2025;37(5):1060‐1074.e4. doi: 10.1016/j.cmet.2025.03.011 [DOI] [PubMed] [Google Scholar]
- 4. Geiselman PJ, Anderson AM, Dowdy ML, West DB, Redmann SM, Smith SR. Reliability and validity of a macronutrient self‐selection paradigm and a food preference questionnaire 1. Physiol Behav. 1998;63(5):919‐928. doi: 10.1016/S0031-9384(97)00542-8 [DOI] [PubMed] [Google Scholar]
- 5. White MA, Whisenhunt BL, Williamson DA, Greenway FL, Netemeyer RG. Development and validation of the food‐craving inventory. Obes Res. 2002;10(2):107‐114. doi: 10.1038/oby.2002.17 [DOI] [PubMed] [Google Scholar]
- 6. Blundell J, Finlayson G, Axelsen M, et al. Effects of once‐weekly semaglutide on appetite, energy intake, control of eating, food preference and body weight in subjects with obesity. Diabetes Obes Metab. 2017;19(9):1242‐1251. doi: 10.1111/dom.12932 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Martin CK, O'Neil PM, Pawlow L. Changes in food cravings during low‐calorie and very‐low‐calorie diets. Obesity. 2006;14(1):115‐121. doi: 10.1038/oby.2006.14 [DOI] [PubMed] [Google Scholar]
- 8. Martin CK, Rosenbaum D, Han H, et al. Change in food cravings, food preferences, and appetite during a low‐carbohydrate and low‐fat diet. Obesity. 2011;19(10):1963‐1970. doi: 10.1038/oby.2011.62 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Correlation between body weight change from baseline and changes from baseline in Food Preferences Questionnaire or Food Craving Inventory Scores.
Data Availability Statement
Eli Lilly and Company provides access to all individual participant data collected during the trial, after anonymisation, except for pharmacokinetic or genetic data. Data are available to request 6 months after the indication studied has been approved in the United States and European Union and after primary publication acceptance, whichever is later. No expiration date for data requests is currently set once data have been made available. Access is provided after a proposal has been approved by an independent review committee identified for this purpose and after receipt of a signed data‐sharing agreement. Data and documents, including the study protocol, statistical analysis plan, clinical study report, and blank or annotated case report forms, will be provided in a secure data‐sharing environment. For details on submitting a request, see the instructions provided at www.vivli.org.