Introduction

Nutrition therapy and increased physical activity are first-line therapies for patients with type 2 diabetes (T2D)1. Diabetes-specific nutritional formulas (DSNFs) may be used as a component of medical nutrition therapy (MNT) to help in improving glycemic control and reducing body weight2,3. There is evidence that the integration of meal replacement formulas into an MNT plan leads to better compliance with nutrition therapy and greater weight loss compared to patients on an isocaloric MNT plan4. Recently, the American Diabetes Association included DSNFs in its clinical practice recommendations for lifestyle management5. However, the mechanisms by which DSNFs lead to weight loss and improved blood glucose control are not fully understood.

Regulation of appetite is a complex process that involves intricate pathways of hormonal and neuronal signaling6,7. We previously reported that in comparison to an isocaloric oatmeal breakfast, two commercially available DSNFs significantly increased production of postprandial glucagon-like peptide-1 (GLP-1) hormone8. GLP-1 is an incretin hormone which increases insulin secretion and suppresses glucagon secretion, leading to enhanced glycemic control and increased satiety as a result of DSNF consumption9,10.

This study was conducted to explore the effects of two commercially available DSNFs: Glucerna (GL, Abbott Nutrition Inc., Columbus, OH, USA) and Ultra Glucose Control (UGC, Metagenics, Inc., Aliso Viejo, CA, USA) versus a common breakfast food, namely oatmeal (oatmeal, Quaker Old Fashioned Oats, Quaker Oats Co., Chicago, IL, USA) on several other satiety and hunger hormones in overweight and obese patients with T2D.

Subjects and methods

Study design and subjects

This cross-over, three-way, and open-label, ancillary study was conducted in accordance with the Helsinki Declaration and was approved by the institutional Committee on Human Studies. All participants signed a written informed consent. The study was registered at ClinicalTrials.gov (Identifier: NCT02691481). Eligible subjects were patients with T2D for ≥3 months, ages 18 to 75 years, body mass index (BMI) > 25 kg/m2, and glycated hemoglobin A1C (HbA1c) ≥ 6.5%. Patients using diabetes or cholesterol-lowering medications had to be on stable doses of these medications for ≥3 months. Exclusion criteria included pregnancy or lactation, use of insulin or GLP-1 analogs, history of bariatric surgery, gastroparesis, and malabsorption syndrome. Twenty-five subjects were enrolled in the study, of which 22 subjects completed all study visits. One subject dropped out prior to the first study visit for personal reasons. Two other subjects dropped out after the first visit due to inconvenience of frequent blood sampling. Data from those subjects were excluded from statistical analysis. Mean age of the remaining 22 subjects (±SD) was 62.3 ± 6.8 years, diabetes duration was 9.5 ± 9.8 years, and HbA1c was 6.8 ± 0.7%. Baseline characteristics of the study subjects are summarized in Table 1.

Table 1 Characteristics of study subjects

Analyzed hormones

Amylin is a satiety hormone that is co-secreted with insulin from pancreatic β-cells11. Its secretion induces satiety by stimulating the brainstem to slow gastric emptying and decrease gastric secretions12,13. Cholecystokinin (CCK) is a satiety hormone secreted by enteroendocrine cells in the duodenum and jejunum14. Its actions include the promotion of gallbladder contraction, inhibition of gastric acid secretion, and slowing of gastric emptying15. Glucagon is secreted by pancreatic alpha cells and induces satiety through the vagus nerve13. Leptin is secreted by adipose tissue and stimulates satiety centers in the hypothalamus16. Peptide-YY3-36 (PYY) is secreted by enteroendocrine L-cells6 and acts as a satiety signal to the hypothalamus while reducing gastric acid secretion and gastrointestinal motility17. Ghrelin is a hunger hormone secreted mainly by the stomach18. Its stimulates gastrointestinal motility and gastric acid secretion19.

Study procedures

Subjects were asked to come for three visits with a washout period between visits of at least two days. All visits were completed over three weeks. Subjects were instructed to come for each visit after fasting overnight for at least 8 h and were asked to withhold their diabetes and cholesterol-lowering medications on the morning of the visit. In random visit order, each subject was asked to ingest one of the three tested meals (GL, UGC and oatmeal) for breakfast. All meals were 200 kcal each. GL was provided in a 237 mL (8 fl oz) bottle; UGC was prepared by dissolving 200 kcal powder in 296 mL (10 fl oz) of water; and oatmeal was prepared by adding water to 56 g of dry oats and cooking the mixture on a stove for 5–10 min. No milk, sugar, or sweetener was added. Macronutrient composition of the three breakfast meals is shown in Table 2.

Table 2 Nutrition information of the three breakfast meals

For safety, blood glucose was measured at the beginning of each visit. If blood glucose was between 70–300 mg/dL, a venous line was inserted and a baseline blood sample was drawn. This was followed by consumption of the test meal within 3–5 min. Blood samples were collected at 30, 60, 90, 120, 180, and 240 min from the start of each meal. Blood samples were tested for serum active amylin, CCK, ghrelin, glucagon, leptin, and PYY. After collection of the last sample, subjects were given a snack and were instructed to take their regular medications.

Statistical analyses

Values for all measured variables are presented as mean ± SD or standard error of the mean (SEM). Study data were analyzed using SAS statistical software (SAS Institute Inc., Cary, NC, USA). Analysis was performed using linear mixed effects models to model the covariance structure arising from the repeated measures design. Where overall F-tests were significant, pairwise differences of the treatment means were tested with t-tests using Tukey’s p-value adjustments. Outcomes were defined as area under the curve between 0 and 240 min for measured variables over time (AUC0–240) calculated using the trapezoidal formula20. Incremental AUC between 0 and 240 min (iAUC0–240) was calculated using the same formula but representing the area above the fasting level.

Results

Mean fasting serum glucagon for oatmeal, GL, and UGC were similar (35.8 ± 4.4, 41.9 ± 4.7, and 34.5 ± 4.6 pg/mL respectively). Glucagon iAUC0-240 was significantly higher after GL and UGC compared to oatmeal (p < 0.0001 for both); however, there was no difference in glucagon iAUC0–240 between GL and UGC (Fig. 1).

Fig. 1: Adjusted serum concentrations of glucagon and peptide YY (PYY) in 22 subjects with type 2 diabetes after intake of 200kcal of oatmeal, Glucerna, and Ultra Glucose Control (UGC).
figure 1

Values are mean ± SEM. *p < 0.0001

Mean fasting serum PYY for oatmeal, GL, and UGC were similar (72.2 ± 7.2, 76.5 ± 7.8, and 68.7 ± 8.3 pg/mL respectively). PYY iAUC0–240 was significantly higher after GL and UGC compared to oatmeal (p < 0.0001 for both); however, there was no difference in PYY iAUC0–240 between GL and UGC (Fig. 1).

Mean fasting serum active amylin for oatmeal, GL, and UGC were similar (10.5 ± 1.6, 9.9 ± 1.4, and 8.9 ± 1.3 pg/mL respectively). Active amylin iAUC0-240 showed no significant differences between meals (p = 0.076) (Fig. 2).

Fig. 2: Adjusted serum concentrations of active amylin, cholecystokinin, ghrelin, and leptin in 22 subjects with type 2 diabetes after intake of 200kcal of oatmeal, Glucerna, and Ultra Glucose Control (UGC).
figure 2

Values are mean ± SEM. CCK cholecystokinin. Incremental area under the curve was not different between meals for all four variables

Mean fasting serum CCK for oatmeal, GL, and UGC were similar (38.8 ± 18.7, 33.8 ± 18.2, and 37.8 ± 16.0 pg/mL respectively). CCK iAUC0–240 showed no significant differences between meals (p = 0.85) (Fig. 2).

Mean fasting serum ghrelin for oatmeal, GL, and UGC were similar (9.1 ± 1.6, 10.0 ± 1.6, and 10.3 ± 2.1 pg/mL respectively). Ghrelin iAUC0–240 showed no significant differences between meals (p = 0.82) (Fig. 2).

Mean fasting serum leptin for oatmeal, GL, and UGC were similar (25224.9 ± 4273.6, 23649.5 ± 3820.1, and 24790.2 ± 4012.4 pg/mL respectively). Leptin iAUC0–240 showed no significant differences between meals (p = 0.87) (Fig. 2).

Discussion

In the Look AHEAD (Action for Health in Diabetes) study3 and other shorter studies4,21, use of DSNFs as part of a hypocaloric nutrition therapy was associated with weight reduction that was clearly associated with their frequency of use to replace calories or smaller meals. This study provides a mechanistic explanation of that effect, where two of the commercially available DSNFs showed significant increase in two essential weight-modulating hormones that contribute to satiety and increased energy expenditure. Both tested DSNFs increased PYY in comparison to isocaloric oatmeal. This study also showed that both DSNFs significantly stimulate glucagon secretion in comparison to isocaloric oatmeal. Glucagon affects glycemia and satiety. Despite its stimulatory effect on hepatic glucose production, glucagon hormone increases glucose metabolism, and energy expenditure22,23. In addition, glucagon indirectly stimulates satiety through an afferent signal from the hepatic branch of the vagus nerve24. These observations complement our previous observation that both DSNFs stimulate GLP-1 hormone production, another strong satiety hormone, in comparison to isocaloric oatmeal8.

Despite previous claims that all calories are created equal in their effect on body weight25, this study shows that different macronutrients have different effects on key satiety and weight-modulating hormones since all tested meals were of equal caloric content. The two studied DSNFs are higher in their protein and fat content and lower in their carbohydrate content than oatmeal (Table 2). It has been debated which macronutrient(s) elicit the highest postprandial PYY response. An earlier study favored fat in producing the highest PYY response26. However, more recent studies showed that protein induces the highest PYY response27 and carbohydrates induce the smallest effect28. Our results are also in line with previous studies that showed meals higher in both protein and fat content induce higher glucagon response compared to a carbohydrate-rich meal24,29.

Although both tested DSNFs stimulate two opposing weight-modulating hormones, GLP-18 and glucagon, our findings suggest that the stimulatory effect of protein and fat within DSNFs is stronger on glucagon secretion than the inhibitory effect of GLP-1 on glucagon production. Postprandial amylin levels were marginally higher following ingestion of UGC compared to GL and oatmeal, but this difference was not statistically significant (p = 0.076). Furthermore, there were no differences in the postprandial effects of DSNFs on CCK, ghrelin, and leptin hormones. While these changes in satiety hormones provide an attractive potential explanation for the reported success of DSNFs in supporting weight loss, it is also possible that these changes in the satiety hormones, while statistically significant, may not be of sufficient magnitude to explain an effect on satiety that is large enough to interpret their role in improved weight loss.

The present study had several limitations which include the difference in texture between oatmeal and DSNFs. A previous study reported difference in satiety between solid and liquid meal replacements30. This study was powered to detect differences in glucose AUC0–240 rather than differences in the analyzed hunger and satiety hormones. Background diets of the study subjects were not controlled and their effect on the study outcomes is unknown. We aimed to minimize that effect by asking subjects to fast for at least 8 h before each visit. In addition, subjects completed all study visits within a three-week window.

In conclusion, this study shows that DSNFs significantly increase secretion of two satiety hormones; PYY and glucagon. This effect may be related to their specific macronutrient composition. While the effect of the three different meals on subjective satiety was not directly evaluated, results from this study may partially explain the mechanism of body weight reduction associated with DSNFs use.