Changes in food preferences after oral semaglutide administration in Japanese patients with type 2 diabetes: KAMOGAWA-DM cohort

Abstract

Background

This study aimed to investigate the effects of oral semaglutide on the changes in food preference of Japanese patients with type 2 diabetes.

Methods

This retrospective multicenter study included 75 patients with type 2 diabetes who received oral semaglutide. The primary outcome was the change in the score of brief-type self-administered diet history questionnaire (BDHQ) score 3 months after the initiation of oral semaglutide treatment. The secondary outcome was the change in the Control of Eating Questionnaire (CoEQ), HbA1c, and body mass index (BMI) after 3 months.

Results

The median age, BMI, and HbA1c of the 23 participants were 64.0 years, 26.9 kg/m², and 7.6% (59 mmol/mol). The BDHQ results showed total energy was significantly reduced. Among the individual nutrients, carbohydrates most decreased. The CoEQ results particularly showed declines in cravings for something sweet, chocolate or chocolate flavored foods, and starchy foods, satisfaction at meals, frequency and intensity of food craving, difficulty of resisting the craving for food, and frequency of eating in response to cravings for food were significantly lower after 3 months. The mean HbA1c and BMI significantly decreased.

Conclusions

In Japanese patients with type 2 diabetes, oral semaglutide treatment decreased total energy intake and changed food preferences.

Introduction

Obesity is common among patients with type 2 diabetes, and prevention of complications through weight loss is extremely important. However, dietary control for weight management purposes can be challenging.

Glucagon-like peptide-1 (GLP1) receptor agonists delay gastric excretion and act on the hypothalamus to suppress appetite and reduce body weight, making them useful in treating patients with obesity. ¹ Until 2021, only injectable drugs have been available; after 2021, oral drugs were made available and are more frequently prescribed. Oral semaglutide has been developed with the absorption enhancer, sodium N‐(8‐[2‐hydroxybenzoyl] amino) caprylate. ² Sodium N‐(8‐[2‐hydroxybenzoyl] amino) caprylate increases gastric pH, inhibiting the degradation of semaglutide by gastric acid, and hydrophobization facilitates semaglutide absorption by improving the ease in which it can pass through the cells of the gastric epithelium. Oral semaglutide reportedly improves glycemic control and promotes weight loss in patients with type 2 diabetes in phase 3 studies. ³ Previous reports have shown that GLP1 receptor agonists suppressed hunger before meals and improved food cravings, frequency of feeling hungry, meal portion control, and frequency of thinking about food. ⁴

In addition, oral semaglutide reportedly reduces the intake of high-fat and sweet foods. ⁵ However, food culture varies greatly from region to region, with no reports of changes in food preference in Japanese patients treated with oral semaglutide.

The main objective of this study was to examine the changes in food preferences of Japanese patients with type 2 diabetes after oral semaglutide administration. We investigated the effects of oral semaglutide on body weight, energy intake, overall food craving control, and food preferences.

Methods

Ethics

This study was approved by the Research Ethics Committee of Kyoto Prefectural University of Medicine (No. RBMR-E-466; April 4, 2013) and was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all patients.

Patients and study design

This retrospective multicenter observational cohort study was a sub-analysis of the KAMOGAWA-DM cohort study, an ongoing cohort study of patients with diabetes initiated in 2014. Among patients with type 2 diabetes attending Kyoto Prefectural University of Medicine or Kameoka City Hospital who participated in the KAMOGAWA-DM cohort study, we enrolled patients aged 20 to 80 years who initiated oral semaglutide treatment within the enrollment period and provided informed consent. The enrollment period was between November 2021 and December 2023, and included only patients enrolled in the KAMOGAWA-DM cohort who were started on oral semaglutide at their physician’s discretion. Exclusion criteria were patients taking other GLP1 receptor agonists, discontinuation of oral semaglutide (4 patients had nausea, 2 constipation, 1 dizziness, 1 difficulty in complying with oral medication time), and failure to complete pre- and post-questionnaires.

Data collection

In this study, blood examinations and a self-administered questionnaire were conducted at the time of drug initiation and 3 months later. Baseline data were collected at the time of drug initiation. Participants were classified as nonsmokers, former smokers, or current smokers. The estimated glomerular filtration rate (eGFR) was calculated using the eGFR formula for Japanese individuals:

eGFR (mL/min per 1.73 m²) = 194 × Cre^−1.094 × age^−0.287 (× 0.739 for women).

Brief-type self-administered diet history questionnaire

The brief-type self-administered diet history questionnaire (BDHQ) is a questionnaire designed to quantitatively examine the dietary habits (nutrient and food intake) of adults living in Japan over the past month, which was used to investigate nutrients and individual food amounts. A dedicated nutrient value calculation program can calculate the intake of approximately 30 nutrients and 50 foods. ⁶ Non-drinkers were excluded from the analysis of items related to alcohol consumption in the BDHQ.

Control of eating questionnaire

The Control of Eating Questionnaire (CoEQ) is a tool used to evaluate the control of eating and degree of food cravings. This includes ratings of overall food cravings, such as hunger, satiety, food satisfaction, and difficulty in resisting appetite, as well as ratings of cravings for individual foods, such as sweet and salty foods, fruits, dairy products, and starchy foods, in the last 7 days. ⁷ A Japanese translation of the CoEQ was used in this study, which consisted of 19 items and was evaluated using a 100-mm visual analog scale.

Primary and secondary outcomes

The primary outcome was defined as the change in the change in BDHQ scores 3 months after the initiation of oral semaglutide treatment. The secondary endpoint was defined as the change in the CoEQ scores 3 months after the initiation of oral semaglutide treatment. In addition, we examined the change in glycated hemoglobin (HbA1c) levels and body mass index (BMI) at 3 months.

Statistical analyses

Continuous variables of the baseline characteristics are expressed as the mean (standard deviation). The BDHQ and CoEQ questionnaire scores were evaluated using the Wilcoxson signed-rank test because they were not normally distributed. BMI and HbA1c levels were evaluated using paired-samples t-tests. Differences were considered significant at p < .05 (two-sided). Statistical analyses were performed using SPSS statistical software, ver. 29.0 (IBM Corp., Armonk, NY, USA).

Results

A total of 5474 patients were entered in the KAMOGAWA-DM cohort study between November 2021 and December 2023, of which 75 met the inclusion criteria. All patients were followed up for 3 months, and the patients who were taking other GLP1 receptor agonists, discontinued taking oral semaglutide, or did not complete the questionnaires were excluded. Ultimately, 23 patients continued oral semaglutide and completed the questionnaires at 3 months. Of the 23 patients analyzed, 12, 9, and 2 patients were administered 3, 7, and 14 mg of oral semaglutide, respectively, at 3 months.

At baseline, 47.8 % were male, and the median (interquartile ranges) age, BMI, and HbA1c level were 64.0 (58.0–71.5) years, 26.9 (25.2–29.2) kg/m², and 7.6% (7.3–8.2) (59 [56–66] mmol/mol), respectively.

Table 1 shows the changes in BDHQ from baseline to 3 months. Of the 23 participants, the median (interquartile ranges) total energy, protein, lipid, and carbohydrate intake was 1986 (1556–2615), 294 (246–452), 587 (379–745), and 914 (765–1369) kcal, respectively. The BDHQ results after 3 months showed that total energy intake was significantly reduced by 1546 (1246–1755) kcal (p ≤ .001). Among the macronutrients, carbohydrate intake significantly decreased to 785 (574–877) kcal at 3 months (p ≤ .001), and fat intake also showed a significant reduction to 443 (355–586) kcal (p = .004). No significant change was observed in animal protein intake (p = .053).

Table 1.

Median energy intake at baseline and at 3 months by nutrient according to brief-type self-administered diet history questionnaire.

Individual nutrients	Energy intake at baseline (kcal)	Energy intake at 3 months (kcal)	p-value
Total energy	1986 (1556–2615)	1246 (1246–1755)	<0.001
Protein	294 (246–452)	259 (209–303)	0.003
Animal protein	212 (138–284)	160 (135–198)	0.053
Vegetable protein	115 (97–183)	92 (77–119)	<0.001
Fat	578 (379–745)	443 (355–586)	0.004
Animal fat	238 (180–350)	196 (161–259)	0.010
Vegetable fat	308 (216–398)	225 (192–305)	0.009
Carbohydrate	914 (765–1369)	785 (574–877)	<0.001

The median value (interquartile ranges) is presented.

The p-value was calculated using Wilcoxon signed-rank test.

From the CoEQ results, eight items were significantly lower after 3 months: desire to eat something sweet, satisfaction at meals, frequency of food cravings, intensity of food cravings, difficulty of resisting food cravings, frequency of eating in response to food cravings, cravings for chocolate or chocolate flavored foods, and cravings for starchy foods.

The median HbA1c level significantly decreased from 7.6% (7.3%, 8.2%) (59 [56, 66] mmol/mol) to 7.3% (6.7%, 7.7%) (56 [53, 65] mmol/mol) after 3 months (p < .001). BMI also significantly decreased from 26.9 (25.2, 29.2) to 25.9 (24.3, 29.7) kg/m² after 3 months (p = .005).

Discussion

In this 3-month study of oral semaglutide treatment, decreased total energy intake and changes in food preferences were evident, as well as decreased HbA1c and BMI levels. Regarding dietary preferences according to the BDHQ, a decrease was observed in total energy intake, with carbohydrate intake showing the greatest decrease among the macronutrients, followed by lipid intake.

According to the CoEQ, the desire to eat something sweet, satisfaction at meals, frequency of food cravings, intensity of food cravings, difficulty of resisting food cravings, frequency of eating in response to food cravings, cravings for chocolate or chocolate flavored foods, and cravings for starchy foods were significantly lower at 3 months relative to baseline. Changes in CoEQ and BDHQ scores may be due to changes in food preferences. The results of our study suggest that GLP1 analogs may mediate the intake of high-sugar and high-fat foods.

Patients with obesity have reduced sweet-taste perception and therefore require more sweetener. ⁸ In addition, the higher an individual’s BMI, the lower their sensitivity to lipids in the oral cavity, causing increased fat intake. ⁹ Continued consumption of a high-fat diet decreases sensitivity to lipids, leading to increased consumption of high-fat foods. ¹⁰

GLP1 receptors are located in the gustatory nerve in close proximity to the taste bud cells and are involved in sweet-taste signaling. ¹¹ Liraglutide administration also reportedly lowers the sweet-taste detection threshold. ¹² GLP1 stimulation possibly lowered the sweetness detection threshold, which might have led to reduced intake of sweet and starchy foods in the present study.

This study to assess the effect of oral semaglutide on the degree of appetite changes among Japanese patients with type 2 diabetes. In these patients, oral semaglutide treatment decreased total energy intake and changed their food preferences after 3 months. Therefore, oral semaglutide may be an effective intervention for weight management and diabetes control in this patient population.