Study design
An open label, crossover, placebo-controlled randomized controlled trial design involved two different test products namely control (glucose) and Formulation 3 (formulated MMT peel powder) was conducted in Kuala Nerus, Terengganu, Malaysia. The respondents consumed Formulation 3 (A) and control (B) with study breakfast based on randomly assigned treatment sequences (AB, BA). The randomization plan was established by Research Randomizer (www.randomizer.org) prior to the start of the study. Respondents were assigned to the randomization plan in order of recruitment. The principal investigator generated the randomization plan, enrolled and assigned respondents to the interventions. Each study visit was separated into three days of washout period and lasted about 3 hr (starting around 8 am and finishing at approximately 11 am). There should be at least of a 3-day gap between the test meals to minimize any carryover effect (12, 13). The study recruitment and screening started from 18th August 2021 to 18th October 2021 whereas study visits commenced from 1st October 2021 to 3rd November 2021.
Subjects
Men and women between the ages of 18 and 59 who live, work, or study on Kuala Nerus, Terengganu, Malaysia were recruited using poster disseminated via email, social media platforms such as Facebook and Whatsapp as well as through direct approach at each department in Gong Badak Campus, Universiti Sultan Zainal Abidin (UniSZA). The respondent eligibility was identified by a two-stage screening process. In the first stage, the respondents who are at risk of type 2 diabetes mellitus (T2DM) were assessed via the modified Finnish Type 2 Diabetes Risk Assessment Tool (FINDRISC) (14) and a list of screening questions (inclusion and exclusion criteria) which were distributed via Google Form. Inclusion criteria included respondents who were Malaysian and non-smoker. On the other hand, respondents were excluded with a clinical history of T2DM; capillary fasting plasma glucose (FPG) ≥ 7.0 mmol/L; taking oral antidiabetic agents; participating in other weight management programmes or interventional research; on a prescribed medical diet; gastrointestinal tract (GIT) illnesses or conditions; allergy or sensitivity to study products; being pregnant, currently breastfeeding, or planning to become pregnant; on dietary restrictions and smokers. Those who scored ≥ 4 in FINDRISC (moderate or high risk for T2DM for the next 10 years) and met the inclusion criteria were invited to the second stage screening test via the capillary FPG measurement by finger pricking. In this stage, the respondents attending the preliminary visit with FPG < 7.0 mmol/L were invited to volunteer.
Preliminary visit
During the preliminary visit, the study objective, procedure involved, expected study timeframe, respondent’s roles and other UniSZA Human Research Ethics Committee (UHREC) requirements were explained to the respondents. They were allowed to ask questions and were given ample time to examine their options. The respondents were volunteered to join this study and they can withdraw from the study as they wished. All eligible respondents were required to sign an informed consent form before participating in this study. The socio-demographic questionnaire, 24-hr dietary recalls (food logs) and International Physical Activity Questionnaire-Short Form (IPAQ-SF) were given to respondents to complete at their preliminary visit. These questionnaires were provided in both open-ended and closed-ended formats.
Test products
The control was made up of glucose whereas Formulation 3 made up of 40% MMT peel (13.7 g) and 60% monk fruit sweetener (20.5 g). The composition of the MMT peel was displayed in Table 1. Carboxymethyl cellulose, citric acid and orange flavoring were added to the formulations and mixed thoroughly for uniform distribution of ingredients (15). The MMT was harvested at UniSZA Agropreneur Park, Besut Campus, Terengganu. MMT was first peeled and the peel was then processed into powder, which was stored in an airtight container in a freezer (-21°C) prior to analyses (16). The amount of Formulation 3 used was based on the dietary fiber content of 5 g as reported in an acute intervention study using orange pomace test beverages which contain 5.48 g of dietary fiber (17). The weight for each product were 4.5 g of control and 36.0 g of Formulation 3. The control was given in the placebo group while Formulation 3 was provided in the intervention group.
Table 1
Nutritional compositions of MMT peel
Nutrients | Amount (g/100 g) |
Calories (kcal) | 337.23 ± 1.63 |
Total Carbohydrate | 67.65 ± 0.36 |
Dietary Fiber | 49.25 ± 2.05 |
Protein | 13.31 ± 0.05 |
Fat | 1.49 ± 0.35 |
The nutritional compositions of control and Formulation 3 containing similar available carbohydrate (4.5 g) are presented in Table 2. The carbohydrate content of Formulation 3 comprised of 20.5 g of sweetener with 0.02 kcal/g. This sweetener cannot be metabolized by the human body and is excreted unmodified into urine without influencing blood glucose and insulin levels (18). Hence, the available carbohydrate was calculated using the formula: total carbohydrate content (30 g) – sweetener content (20.5 g) - dietary fiber content (5 g) (19). Available carbohydrates refers to the carbohydrates that are digested and absorbed by the human small intestine which can raised blood glucose levels meanwhile dietary fiber refers to carbohydrate polymers (such as lignin, cellulose, hemicellulose, gums, pectin and inulin) with 10 or more monomeric units which cannot be degraded by the endogenous digestive enzyme in the human small intestine (20–22). On study visits, a total volume of 380 mL was consumed to maintain a standardized volume of drink between control and Formulation 3 study visits.
Table 2
Nutritional compositions between control and Formulation 3
Nutrients | Control | Formulation 3 |
Weight (g) | 4.5 | 36.0 |
Calories (kcal) | 18 | 36 |
Total Carbohydrates (g) | 4.5 | 30.0 |
Available Carbohydrate (g) | 4.5 | 4.5 |
Dietary Fiber (g) | 0.0 | 5.0 |
Protein (g) | 0.0 | 1.4 |
Fat (g) | 0.0 | 0.2 |
The breakfast used in this study is typical foods consumed by Malaysian. The standardized breakfast consisted of 2 slices of white bread (brand Gardenia), 1 piece of hard-boiled egg, 1 teaspoon of unsalted margarine (brand Anchor) and 1 glass of plain water (200 mL), packing up to a total of 267 kcal (30.6 g carbohydrate, 11.1 g protein, 10.7 g fat and 0.9 g dietary fiber). The study meal for breakfast (test products and standardized breakfast) of both placebo and intervention groups were provided in Table 3. The total calories content of the study breakfast adhered closely to the recommendation of 400 kcal (23).
Table 3
Nutritional compositions of study meal (test products and standardized breakfast)
Nutrients | Placebo | Intervention |
Calories (kcal) | 285 | 303 |
Total Carbohydrates (g) | 35.1 | 60.6 |
Available Carbohydrate (g) | 35.1 | 35.1 |
Dietary Fiber (g) | 0.9 | 5.9 |
Protein (g) | 11.1 | 12.5 |
Fat (g) | 10.7 | 10.9 |
Data collection
The respondents attended the lab at 8 am with light clothing and were asked a series of questions to measure fasting, nutrition, physical activity and sleep regimen adherence. They were also required to hand in the completed questionnaire which consisted of dietary and physical activity information which were distributed to them prior to data collection. Then, height, weight, waist circumference (WC), % body fat and blood pressure measurement were taken followed by a perceived satiety measurement and fasting capillary plasma sample (baseline) withdrawal using finger pricking method.
Next, the respondents were randomized to consume of one of the two types of test products namely A: control and B: Formulation 3 dissolved in 180 mL plain water together with the standardized breakfast in treatment sequences (AB, BA). Each respondent consumed the meals in a separate place and avoided using electronic gadgets while eating to reduce distractions. They were instructed to consume the meals at a comfortable pace within 15 min. The respondents were asked to remain rested in the same place during the testing period and allowed to read, use their phone or laptop and use the toilet after meals consumption.
At the first bite, a timer was started and postprandial blood samples as well as perceived satiety measurement were then taken at 30, 60, 90 and 120 min thereafter. The blood collection techniques were adhered exactly as they were written in the standard operating procedure (SOP). After the 120 min assessments, the respondents were allowed to leave the testing site and instructed to record the food consumed for the rest of the day. The food record was aimed to explore the changes in food intake following each test food relative to usual food intake (24). Also, they were given a new food log form and physical activity diary to fill in one day before the next study visit.
Blood glucose level measurements
FPG and postprandial blood glucose were measured using glucometer (Accu Chek Performa, New South Wales, Australia) from finger-prick blood samples according to the SOP (25). The respondents’ finger was cleaned with an alcohol swab, left to dry and lanced with a sterile lancet to obtain a drop of blood which was placed onto the glucose meter strip. The digital result was displayed in the display window after 10 to 20 seconds and then recorded. Glucometer is a fast and valid device to evaluate blood glucose level (26). Calibration was performed using a standard glucose strip provided by the manufacturer. In comparison to venous blood glucose, capillary blood glucose is recommended due to its simplicity to obtain, greater rise in blood glucose and less variability of the results (27).
Perceived satiety
Perceived satiety measurements were rated using a 10-cm visual analogue scale which consisted of hunger (How hungry are you?), fullness (How full are you?), desire to eat (How strong is your desire to eat?), prospective food consumption (How much would you be able to eat right now?) and thirstiness (How thirsty are you?) (28). The point ranged from 0 (not at all) to 10 (very much). Respondents were asked to make a vertical mark across the line corresponding to their feelings at the present time.
Food intake
The respondents were given a detailed verbal and written instructions as well as visual aids including image examples by the researcher on how to fill in the food log. The food and beverage consumed was recorded in the form of standard household measurements such as scoop, cup, bowl, teaspoon, dessert spoon or tablespoon. These household measures were shown to the respondents to enhance the quality of the outcomes. This can facilitate the respondents in distinguishing the amount of portion size taken relative to a given portion size. The respondents were asked to record all the food or drink they had consumed including sauces, soup and gravies once they woke up in the morning until they went to sleep at night. They were also required to provide precise information such as food preparation methods, the inclusion of condiments or spices, household quantity, products’ brand or snacking throughout the day. For respondents who ate a mixed dish such as fried rice, fried noodle and noodle soup, all the ingredients available in the mixed dish were listed down and estimated the amount in household measurements.
All the units in household measurement were converted to grams by using Malaysian Food Composition table and Malaysian Atlas of Food Exchanges and Portion Size. The macronutrients and chosen micronutrients were determined according to the Nutrient Composition of Malaysian Foods Database (https://myfcd.moh.gov.my/) and Malaysian Atlasof Food Exchanges and Portion Size. However, this database did not provide complete information on dietary fiber for all food items. Therefore, the food database from United States Department of Agriculture Food Composition Databases (https://ndb.nal.usda.gov/ndb/), Energy and Nutrient Composition of Foods Singapore (https://focos.hpb.gov.sg/eservices/ENCF/) and nutrition facts on the food packaging were used to complement the incompleteness of the Malaysian Food Composition Database.
Statistical analysis
The data were analyzed using IBM SPSS system version 21.0 and Nutritionist Pro™ software (Version 7.0.0, Axxya Systems). The Shapiro-Wilk test was done for the assessment of normality tests and the data considered to be normal if p > 0.05. The 2 hr area under the curve (AUC) of blood glucose levels (incremental area above the baseline value) and perceived satiety (hunger, fullness, desire to eat, prospective food consumption and thirstiness) were calculated using the trapezoidal approximation (29). The maximum concentration (Cmax) of blood glucose levels were taken to be the highest concentration within the respective time interval (30).
A paired t-test (normally distributed) or Wilcoxon matched pair signed-rank sum test (non-normally distributed) was used to compare the difference of two data between groups (placebo and intervention) on blood glucose levels (iAUC and CMAx), dietary intake, and perceived satiety (AUC). Two-way repeated measures analysis of covariance (ANCOVA) was used to compare the postprandial blood glucose levels between two test products over different time points, controlling for treatment and time to test for treatment by time interaction after adjustment for covariates (age, ethnicity, gender and body mass index (BMI)) (31, 32). The perceived satiety was analysed using the Friedman non-parametric test. The analysis was considered significant at a p-value ≤ 0.05. Any significant data was further tested using Wilcoxon matched pair signed-rank sum test.
The sample size for this study is based on a power calculation for CMax. A sample size of 23 respondents was determined to detect approximately a 0.62 mmol/L difference in CMax (a primary outcome measure) (30) with 80% power and at a 5% significance level. A total of 30 respondents were recruited to allow for a 20% dropout.