Association between patient-reported eating behaviours and weight change: Secondary analyses of a randomized, double-blind trial comparing retatrutide and placebo in people with obesity or overweight

Obesity is a chronic, relapsing, progressive disease that negatively impacts the health and health-related quality of life of individuals living with the condition.^1-4 Until recently, pharmacotherapeutic agents for the treatment of obesity were limited.

Retatrutide is a novel synthetic molecule with agonist action at glucose-dependent insulinotropic polypeptide (GIP), glucagon-like peptide-1 (GLP-1) and glucagon receptors.⁵ In a phase 2 placebo-controlled trial, retatrutide treatment at doses ≥4 mg substantially reduced body weight in adults with obesity (Body Mass Index [BMI] ≥30 kg/m²), or overweight (BMI ≥27 kg/m²) plus ≥1 obesity-related complication after 24 and 48 weeks.⁶ Importantly, the safety profile of retatrutide was consistent with that reported in a placebo-controlled phase 1 study, and generally similar to those of GLP-1 or GIP–GLP-1 receptor agonist treatments for type 2 diabetes or obesity.⁶

The Eating Inventory (also known as the Three Factor Eating Questionnaire) is one of the most widely used and established methods available for assessing eating behaviour. It measures constructs central to the regulation of energy and food intake, energy balance and body weight.^{7, 8}

Here, we report changes in eating behaviour related to the drive to eat after treatment with retatrutide or placebo in participants of the phase 2 obesity trial.⁶

This was a phase 2, multicenter, randomized, double-blind placebo-controlled trial conducted in the United States (US) according to all ethical and regulatory requirements. Details of the study design and the primary results of the study have been published.⁶ Briefly, eligible participants were aged 18–75 years with BMI ≥ 30 kg/m², or BMI ≥ 27 kg/m² plus ≥1 obesity-related complication. All trial participants received diet and physical activity counselling at each study visit that was based on U.S. Government guidelines for a healthy diet and physical activity,^{9, 10} and once-weekly subcutaneous retatrutide at a dose of 1 mg, 4 mg (2 mg then escalation or 4 mg), 8 mg (2 mg or 4 mg then escalation), or 12 mg (2 mg then escalation) or placebo for 48 weeks. The primary endpoint was percentage change in weight from baseline to 24 weeks.⁶

Trial participants completed the 51-item Eating Inventory at baseline, Weeks 24 and 48. The Eating Inventory assesses three eating-related constructs: disinhibition (i.e., the tendency to lose control of food intake when faced with external cues, mood changes or disruptive events), perceived hunger (i.e., susceptibility to feelings of hunger) and cognitive restraint (i.e., the level of voluntary control of eating).¹¹ Lower disinhibition scores (range: 0–18) denote lower levels of disinhibited eating. Lower perceived hunger scores (0–14) denote lower susceptibility to feelings of hunger. Higher scores for cognitive restraint (0–21) denote more restrained eating.

Least squares mean (LSM) change from baseline to Weeks 24 and 48 in disinhibition, perceived hunger and cognitive restraint for each retatrutide treatment group were compared with those for the placebo group using a mixed model repeated measures model with treatment * time, baseline BMI (<36 vs. ≥36 kg/m²) * time, sex (female vs. male) * time and baseline * time as variables. Due to the relatively small numbers of patients in this study, data for the two 4 mg groups were pooled, as were data for the two 8 mg groups. Only subjects with a non-missing baseline value were included in the analyses. Post-hoc analyses examined the correlation between change from baseline in body weight and Eating Inventory domain scores using Pearson correlation. Additional analyses examining changes from baseline in Eating Inventory scores and their association with gastrointestinal adverse events over time and participant-level changes from baseline in Eating Inventory domain scores sorted by perceived hunger score at Weeks 24 and 48 are available in the Supporting Information.

A total of 338 participants were randomized and treated between May 2021 and November 2022. Demographic and baseline clinical characteristics of participants in each treatment group were generally similar and have been reported.⁶ The mean age of study participants was 48.2 years, 48% were female, and 96% had BMI ≥ 30 kg/m². Detailed analyses of body weight change have been reported.⁶ For the primary endpoint, LSM weight reduction at Weeks 24 and 48 ranged from 7.2% to 17.9% and 8.7% to 24.2%, respectively, across the retatrutide 1–12 mg groups compared with 1.6% and 2.1%, respectively, in the placebo group. Weight reduction targets of ≥5% and ≥10% were achieved at Week 48 by 64%–100% and 27%–93% of retatrutide-treated participants, respectively, compared with 27% and 9% of placebo-treated participants, respectively.⁶

Concomitant with these reductions in body weight, at Weeks 24 and 48, there were significantly greater score decreases from baseline for all retatrutide-dose groups compared with the placebo group in disinhibition and perceived hunger scores (p < 0.01 for all comparisons), which generally appeared dose-dependent, particularly for disinhibition (Figure 1A,B). Although cognitive restraint scores significantly increased from baseline in all treatment groups at Weeks 24 and 48 (p < 0.001), differences between the retatrutide and placebo groups were not significant (Figure 1C).

Post-hoc exploratory analyses showed significant (p < 0.001) correlations between change from baseline in body weight and change in all Eating Inventory domain scores (Pearson correlation of 0.46 and 0.41 for disinhibition, 0.33 and 0.32 for perceived hunger, and −0.26 and −0.30 for cognitive restraint) at Weeks 24 and 48, respectively (Figure S1).

There was no apparent difference in the prevalence of gastrointestinal adverse events for participants scoring above or below the median change in Eating Inventory scores (Figure S2). Heat maps showing participant-level changes from baseline in Eating Inventory domain scores sorted by perceived hunger score at Weeks 24 and 48 are shown in Figure S3.

In the current study, body weight reduction observed with retatrutide occurred along with significant changes in eating behaviour as measured by the Eating Inventory. Specifically, disinhibition and hunger were decreased in retatrutide-treated participants, with the greatest changes in those receiving retatrutide at doses of ≥4 mg. The lag in treatment effect for the 12 mg group may have resulted from the long dose titration. Despite the observed reductions in weight, cognitive restraint scores did not increase in the retatrutide group compared with the placebo group. It is unknown whether observed changes in eating behaviour in the placebo group were related to the lifestyle intervention or due to a placebo effect. Importantly, greater weight reduction was associated with decreased disinhibition, decreased hunger and increased restraint. The lack of a significant effect of retatrutide on cognitive restraint is of interest because, in lifestyle change programmes for weight loss, increased volitional cognitive restraint appears to be an important component in restricting energy intake and losing weight.¹² The findings suggest that retatrutide may have altered eating behaviour that contributed to the substantial body weight reductions observed with retatrutide treatment in this trial. The findings are consistent with those observed in mechanistic studies of semaglutide,¹³ which is a GLP-1 receptor agonist, and tirzepatide,¹⁴ which is a dual GIP/GLP-1 receptor agonist. As retatrutide is the only molecule in Phase 3 clinical development with agonist activity at three receptors (GIP, GLP-1 and glucagon), no information on ingestive behaviour is available for other similar triple receptor agonists. Semaglutide administration for 12 weeks was associated with reduced energy intake, appetite and food cravings, improved control of eating, and lower relative preference for fatty, energy-dense foods than placebo in a small (N = 28) cross-over trial.¹³ In a 6-week parallel-group study (N = 114), tirzepatide reduced body weight via reduced energy intake and had significant (vs. placebo) effects on reducing appetite, food cravings, food preferences, food cue sensitivity, and disinhibition in people with obesity.¹⁴ These results indicate that people treated with these medications may experience a reduction in the drive to eat, decreased eating in response to various stimuli, and significant weight reduction that occurs without much volitional intent to restrict intake, at least as measured by the cognitive restraint scale of the Eating Inventory. Currently, it is thought that the addition of GIP to GLP-1 enables additional weight loss; however, the exact role of glucagon receptor agonism on eating behaviour remains to be clarified.⁵ Nevertheless, these findings are notable given how difficult it is to volitionally restrict energy intake, lose weight, and maintain a reduced weight during lifestyle change programmes for obesity management.

Limitations of these analyses include the small sample size and that the aspects of eating behaviour assessed were limited to the subjective concepts captured by the Eating Inventory. In addition, dietary records were not collected during the study.

Retatrutide is a promising pharmacological agent that appears to promote weight loss via reductions in food intake that are accompanied by decreased hunger and reduced tendency to eat or overeat in response to cues in the environment or internal cues, such as stress or anxiety. These findings are relevant given the challenges of managing food intake and body weight in a food-abundant environment. Further research is needed to explore the impact of retatrutide on eating behaviour.

The authors provided substantial contributions to the conception (I.G., T.C., M.L.H.) and design of the work (C.K., K.B., I.G., T.C., M.L.H.); the acquisition, analysis, or interpretation of data for the work (C.K., Q.W., K.B., I.G., M.P.-N., C.M., J.L.P., T.C., M.L.H.); and drafting the work and/or revising it critically for important intellectual content (all authors); all authors gave final approval of the version to be submitted to and published in the journal (all authors). All authors have participated sufficiently in the work to take public responsibility for appropriate portions of the content and have agreed to be accountable for all aspects of the work.

This study and the associated analyses were funded by Eli Lilly. Corby K. Martin's institution is supported by a National Institute of Diabetes and Digestive and Kidney Diseases, which funds the Nutrition Obesity Research Center Grant P30 DK072476 entitled “Nutrition and Metabolic Health Through the Lifespan” sponsored by the National Institute of Diabetes and Digestive and Kidney Diseases and by Grant U54 GM104940 from the National Institute of General Medical Sciences, which funds the Louisiana Clinical and Translational Science Center. Eli Lilly and Company, as the funder of the study, had a role in study design, data collection, data analysis, data interpretation, and writing of the report.

Chisom Kanu, Qiwei Wu, Jiat Ling Poon, Iris Goetz, Magaly Perez-Nieves, Kristina Boye, Tamer Coskun, and Mark L. Hartman are employees and stockholders of Lilly. Corby K. Martin has been the primary investigator and co-investigator on research contracts from Lilly, which were paid to his institution. His institution has also received funds via contracts with Lilly to provide advice on patient-reported outcomes when designing studies. These agreements were not related to the current work. Lilly manufactures the drug studied in this paper.

The trial was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines and was approved by an independent ethics committee or institutional review board at each trial site. All the participants provided written informed consent before participation.