Assessing the nutritional quality of lipid components in commercial meal replacement shakes using an in vitro digestion model

Abstract

This study aimed to investigate the nutritional value of five commercial meal-replacement shakes, and mainly focused on the lipid digestion fates and fat-soluble vitamin bioavailability. Four out of five samples exhibited a low lipolysis level (37.33–61.42%), aligning with the intended objectives of these products. Although the remaining sample rich in diacylglycerol (DAG) had a higher lipolysis level (80.83%), the inherent low-calorie nature of DAG might compensate for this drawback. The release level of individual fatty acid was largely determined by the glycerolipid composition. Moreover, the strong positive correlation between lipid hydrolyzed products amounts and the fat-soluble vitamin bioavailability was observed. Surprisingly, one out of five samples can provide enough vitamin A and vitamin E for consumers as a total replacement of one or two regular meals. Consequently, the meal-replacement shakes hold the potential to emerge as healthy products for this fast-paced era if the composition and structure were carefully designed and calculated.

Graphical abstract

Highlights

• •All five meal-replacement shakes contain insufficient polyunsaturated fatty acids.
• •Vitamin A & E bioavailability is linked to the levels of lipid hydrolyzed products.
• •Some shakes could provide enough vitamin A & E as total replacements of normal meals.
• •Diacylglycerol was a potential lipid source for meal-replacement shakes.

1. Introduction

As society progresses, obesity has become a prominent factor affecting public health, leading to a variety of illnesses (World Health Organization, 2018). As a result of the desire to keep fit or lose weight more conveniently, an increasing number of individuals are focusing on fast food options which can assist in weight loss. Consequently, numerous companies are currently developing meal-replacement products, including the increasingly popular meal-replacement shakes (Borreani et al., 2017; Heymsfield et al., 2003). These shakes are designed to mimic low-calorie meal which could be easily consumed as ready-to-drink liquids or easy-to-reconstitute powders (Guevara-Zambrano et al., 2022). These products offer great advantages compared to self-prepared weight loss diets for effective weight-management. They provide individuals a diet with known calorie level and thus reduce the trouble of calculating the calorie (Craig, 2013). Moreover, these products are easily to consume and require no clean-up (Guevara-Zambrano et al., 2022). More importantly, the consumption of meal-replacement shakes was widely confirmed to be efficient to lose weight (Ditschuneit et al., 1999; Flechtner-Mors et al., 2000; Heymsfield et al., 2003; Wadden et al., 2009).

These shakes often claim to be able to replace one or two regular meals totally and are marketed as vitamin/mineral-fortified foods (Heymsfield et al., 2003). Given the prevalent deficiencies of fat-soluble vitamins in contemporary society across various age groups (Tan et al., 2021), these shakes can be regarded as a good dietary source for supplying these nutrients. However, the nutritional impact of fat-soluble vitamins not depend on the ingested quantity, but rather by the amounts can be adsorbed by the human body (Tan et al., 2021), which termed as the bioavailability of the vitamins. Meal-replacement shakes are typically high in protein and dietary fiber, while being low in lipid. Recent researches have found that the low level of lipid and high level of cellulose can both lead to low bioavailability of fat-soluble vitamins (Tan et al., 2021; Yang et al., 2015). Hence, there is a concern that the prolonged intake of meal-replacement shakes may result in more pronounced deficiencies of fat-soluble vitamins. Their deficiencies can exert a broad impact on multiple bodily systems, particularly the immune system, potentially impacting overall health and wellbeing (Tan et al., 2021). However, meal-replacement shake is quiet a complex system, other existing components may instead improve the delivery of fat-soluble vitamins. For example, the presence of protein could improve the emulsification of lipids and decrease the particle size of droplets, which could improve the lipid lipolysis and improve the release of vitamins (Mun et al., 2007). Meanwhile, the lipid composition could also influence the fat-soluble vitamin delivery. The long-chain triglycerides (LCTs) showed better ability compared to medium-chain triglycerides (MCTs) in delivering these vitamins (Yang and McClements, 2013). It had also been proposed that the presence of carbohydrates can influence the lipid digestion fates by increasing the viscosity of the digestion medium, altering droplet disruption or coalescence kinetics, and binding to bile salts, phospholipids, or lipase (Guo et al., 2017). The composition and structure of food can also impact the bioavailability greatly (Tan et al., 2021). There are also many other factors, including particle size, cellulose types and protein types (Tan et al., 2021; Espinal-Ruiz et al., 2014; Chang and McClements, 2016). Thus, it could not judge whether the commercial meal-replacement shakes could provide enough fat-soluble vitamins for dieters based on existed knowledge.

In this paper, we would like to explore the nutritional value of commercial meal-replacement shakes when they were totally treated as the dietary replacements by using in vitro digestion model. Specifically, we mainly focus on the digestion properties of lipids and fat-soluble vitamins bioavailability. The alternations of droplet characteristics during the simulated digestion were also determined. Additionally, this article may provide some suggestions for the design of novel and healthier meal-replacement shake products.

2. Materials and methods

2.1. Materials

Five commercial meal-replacement shakes (#A – #E) in powder form were purchased from Tmall (Hangzhou, China) and all have high sales volumes, and their underlying information was listed in Table 1. Here are the descriptions of the features of each meal-replacement shake: sample #A: moderate in lipid, rich in carbohydrate, rich in protein, and high in fiber; sample #B: moderate in lipid, high in carbohydrate, moderate in protein, and moderate in fiber; #C: low in lipid, rich in carbohydrate, high in protein, and rich in fiber; #D: moderate in lipid, high in carbohydrate, and moderate in protein, and low in fiber; #E: high in lipid, low in protein, and high in fiber. α-Amylase (50 U/mg), porcine pepsin (≥2500 U/mg), bile salts, ammonium hydroxide solution, and tert-Butylhydroquinone (TBHQ) were obtained from the Macklin Biochemical Co. Ltd. (Shanghai, China). Porcine pancreatin (from porcine pancreas, P7545, 8 × USP specifications), lipid standards (triolein and monoolein), and 14 wt% boron trifluoride-methanol were purchased from the Sigma–Aldrich (St. Louis, MO, USA). The standards for vitamin detection, which included vitamin A and α-tocopherol, were also procured from the Sigma–Aldrich. Thirty-seven fatty acid methyl esters were purchased from the AMPEL Laboratory Technologies Inc. (Shanghai, China). Standard dilinolein (1,3-dilinolein, 87%; 1,2-dilinolein, 13%) was obtained from the Nu-Chek-Prep, Inc. (Minnesota, USA). n-Hexane, isopropanol, formic acid, sodium formate, ammonium acetate, acetonitrile and methanol were liquid chromatograph mass spectrometry grade, and all other reagents were of analytical grade.

Table 1.

Detailed information regarding the five meal-replacement shakes, relying on the ingredient list for reference (excluding the lipid and fat-soluble vitamins content, which were assessed independently).

Ingredient content	Lipid (wt%)	Vitamin A (μg/100 g)	Vitamin E (mg/100 g)	Carbohydrate (wt%)	Calcium ion (mg/100 g)	Protein (wt%)	Dietary fiber (wt%)	Percentage of Calories Provided by Lipids (%)	Existed components (according to the ingredient lists)
#A	3.72 ± 0.26	467.2 ± 22.05	10.58 ± 1.55	42.8	1010	20.5	23.2	9.73	Maltodextrin, Isomalto-oligosaccharide, Polydextrose, Soy Protein Isolate, Skim Milk Powder, Whey Protein Concentrate, Instant Soy Powder, Whole Milk Powder, Diacylglycerol Oil, Instant Black Tea Powder, Mineral Premix (Zinc Oxide, Ferrous Phosphate, Calcium Carbonate, Magnesium Oxide), Sodium Carboxymethyl Cellulose, Guar Gum, Silicon Dioxide, Vitamin Premix (Vitamin A, Vitamin D, Vitamin E, Vitamin C, Vitamin B1, Vitamin B2, Vitamin B6, Vitamin B12, Niacin, Folic Acid, Taurine), Edible Flavour, Sucralose.
#B	4.23 ± 0.45	576.93 ± 15.73	11.31 ± 0.87	68.5	139	13.4	9.5	10.02	Resistant Maltodextrin (Soluble Dietary Fiber), Skim Milk Powder (from New Zealand), Whey Protein Powder (Imported), Purple Sweet Potato Powder, Job's Tears Powder, Coconut Milk Powder, Instant Soy Powder, Corn Starch, Oat Dietary Fiber Powder, Crystalline Fructose, Fish Collagen Peptides, Inulin, Powdered Conjugated Linoleic Acid Glycerides, White Kidney Bean Extract, Vitamin Complex (Vitamin A Acetate, dl-Alpha Tocopheryl Acetate, Thiamine Hydrochloride, Riboflavin, Pyridoxine Hydrochloride, Niacin, Folic Acid, D-Calcium Pantothenate, Cyanocobalamin, L-Ascorbic Acid), Mineral Complex (Magnesium Oxide, Ferrous Pyrophosphate, Zinc Sulfate). Chicory Fiber added at a dosage of 30 mg/100g, with a recommended daily intake of chicory fiber not exceeding 15g/day. Powdered Conjugated Linoleic Acid Glycerides added at a dosage of 6 mg/100g, with a recommended daily intake of powdered conjugated linoleic acid glycerides less than 6g/day. No preservatives (added at 0 g/kg) and stabilizers (added at 0 g/kg) are used.
#C	2.38 ± 0.08	157.4 ± 3.21	7.09 ± 0.27	43.6	346.7	33.3	14.13	5.86	Concentrated Milk Protein Powder, Maltodextrin, Maltitol, Red Jasper Juice Powder (Solid Beverage), Soy Protein Isolate, Collagen Peptide Powder (≥5000 mg/75g), Low-DP Isomaltooligosaccharide, Polyglucose, Resistant Maltodextrin, Concentrated Whey Protein Powder, Yogurt Powder (Reconstituted Milk Powder), Instant Soy Powder, Skim Milk Powder, Dried Strawberry Dices, Mineral Complex (Calcium Carbonate, Magnesium Oxide, Gluconate Zinc, Ferrous Pyrophosphate), Freeze-Dried Raspberry Granules, Freeze-Dried Strawberry Granules, Tremella Powder, Needle Cherry Powder (Solid Beverage), Plantago ovata Husk Powder, Xanthan Gum, Vitamin Complex (Vitamin C, Vitamin E, Niacin, Vitamin A, Vitamin B2, Pantothenic Acid, Folic Acid, Vitamin B6, Vitamin B1, Vitamin D, Vitamin B12), Citric Acid, Steviol Glycosides, Food Flavoring.
#D	4.30 ± 0.43	924.61 ± 26.86	9.87 ± 0.67	67	314	15.6	6.8	9.95	Soy Protein Isolate, Concentrated Whey Protein, Isomaltooligosaccharide, Skim Milk Powder, Low-DP Isomaltooligosaccharide, Maltodextrin, Resistant Maltodextrin, Whole Milk Powder, Red Bean Powder, Purple Sweet Potato Granules, Crystalline Fructose, Coix Seed Powder, Cocoa Powder, Microencapsulated Medium-Chain Triglycerides (Medium-Chain Triglycerides, Low-DP Maltodextrin, Sodium Caseinate, Monoglycerides, Diglycerides, Silica, Sodium Tripolyphosphate), Mineral Complex (Tricalcium Phosphate, Magnesium Carbonate, Ferrous Pyrophosphate, Zinc Lactate), Flaxseed Gum, Lecithin, Plantago ovata Husk Powder, Konjac Flour, Pectin, Vitamin Complex (Vitamin A Acetate, Vitamin D3, dl-Alpha Tocopheryl Acetate, Thiamine Mononitrate, Riboflavin, Pyridoxine Hydrochloride, Cyanocobalamin, L-Ascorbic Acid, Niacin, Folic Acid, Calcium D-Pantothenate), Taurine, Food Flavoring, Sucralose, Phaseolus vulgaris Extract, Green Coffee Powder, L-Arabinose, Microencapsulated Conjugated Linoleic Acid Glyceryl Ester (Conjugated Linoleic Acid Glyceryl Ester, Solid Corn Syrup, Sodium Octenyl Succinate Starch, Silica, Sodium Ascorbate), Chitosan Oligosaccharide.
#E	9.45 ± 0.29	163.43 ± 7.06	5.5 ± 0.5	36.8	266	10.8	25.5	25.33	Soy Protein Isolate, Whole Milk Powder, Glucose, Maltodextrin, Low-DP Isomaltooligosaccharide, Oat Flour, Oat Grain Crisps, Polydextrose (Dietary Fiber), Erythritol, Concentrated Whey Protein, Skim Milk Powder, Crystalline Fructose, Fortified Nutrient Blend (Vitamin A, Vitamin D, Vitamin E, Vitamin B1, Vitamin B2, Vitamin B6, Vitamin B12, Vitamin C, Niacin, Folic Acid, Calcium Pantothenate, Magnesium Oxide, Ferrous Pyrophosphate, Calcium Hydrogen Phosphate, Zinc Oxide, Taurine, Maltodextrin), Instant Oatmeal, Inulin, Plantago ovata Husk Powder, Guar Gum, Sodium Carboxymethylcellulose, Food Flavoring. Inulin added at 0.5%, with a recommended intake of ≤15g/day.

Abbreviations: #A, #B, #C, #D, and #E represent the five commercial meal-replacement shakes, respectively.

2.2. Preparation of meal-replacement shakes

Briefly, 10.0 g of meal-replacement powders (powder #A, #B, #C, #D, and #E) was added to the 30.0 g of warm water (50 °C), and stirred vigorously to create uniform shakes (sample #A, #B, #C, #D, and #E), respectively. These shakes were prepared prior to the experiment on the same day.

2.3. Static in vitro digestion of meal-replacement shakes

The in vitro digestion of meal-replacement shakes was conducted according to previously reported method (Brodkorb et al., 2019), including oral (section 2.3.1), gastric (section 2.3.2), and intestinal (section 2.3.3) phases. The stimulated digestive fluids were prepared following the method described by Brodkorb et al. (2019) on the day of digestion, including the simulated salivary fluid (SSF), simulated gastric fluid (SGF) and simulated intestinal fluid (SIF).

2.3.1. Oral digestion phase

The undigested shakes (10.0 g, including 2.5 g of powders and 7.5 g of water) were mixed with pre-warmed SSF, and the pH of the mixture was adjusted to 7.0 before the adding of α-amylase (75 U/mL in the final oral phase). Subject the mixture to 2 min of incubation at 37 °C with constant agitation at 150 rpm. Samples were withdrawn at the end of oral digestion. The enzyme was inactivated by subjecting the samples to high-temperature heating (98 °C) for 5 min.

2.3.2. Gastric digestion phase

Adding the pre-warmed SGF to the oral bolus to achieve a final ratio of 1:1 (wt/wt), and the pH of the mixture was immediately adjusted to 3.0. Upon preheating the samples to 37 °C, a solution of prepared porcine pepsin (2000 U/mL in final gastric phase) was added, followed by a 2-h incubation period (37 °C, 150 rpm) to accomplish gastric digestion. Afterwards, samples were taken at the end of the gastric digestion for characterization. The enzyme was inactivated by subjecting the samples to high-temperature heating (98 °C) for 5 min.

2.3.3. Intestinal digestion phase

Adding the pre-warmed SIF to the gastric chyme to achieve a final ratio of 1:1 (wt/wt), and the pH of the mixture was immediately adjusted to 7.0. After adding bile salts (5 mg/mL in the final intestinal phase), the resulting solution was incubated at 37 °C for 30 min to ensure full solubilization of bile salts. Porcine pancreatin was then added to achieve a trypsin activity of 100 U/mL (no more pancreatic lipase was added due to the low level of lipid in this digestion system). The obtained system was incubated in a shaking water bath (150 rpm) for 2 h at 37 °C to achieve the intestinal digestion, the amount of released free fatty acids (FFA) was monitored by a pH-stat method using 0.1 M standard NaOH solution. Afterwards, the digestive at the end of intestinal digestion was sampled for characterization. The enzyme was inactivated by subjecting the samples to high-temperature heating (98 °C) for 5 min.

2.4. Characterization of initial shakes and digesta

2.4.1. Particle size and ζ-potential analysis

Each initial shakes and their respective digests, taken in the oral, gastric and intestinal digestion phases, were analyzed by Zeta-sizer nano series (Malvern Instruments Ltd., Malvern, Worcestershire, UK) for particle size and ζ-potential at 25 °C. Refractive indices for aqueous phase and oil phase were 1.33 and 1.46, respectively. Before the determination, each sample were diluted with buffer to reach an oil volume concentration of about 0.01 wt% for avoiding multiple scattering effects (Yang et al., 2015). It should be noted that the buffer used here needs to be adjusted according to different phases of digestion: initial shakes, oral phase and intestinal phase, pH 7.0 buffer; gastric phase: pH 3.0 buffer (Yang and McClements, 2013).

2.4.2. Microstructure analysis

The microstructures of initial shakes and their respective digests were analyzed by using confocal laser scanning microscopy (LSM510, Zeiss Inc., Toronto, ON, Canada). Nile red (20 μL, 1 mg/mL, in methanol) and Nile blue A (20 μL, 1 mg/mL, in methanol) were mixed with samples (1 mL) for labeling the lipid and protein, respectively. The excitation of Nile red and Nile blue A were 512 nm and 633 nm, respectively (Chang and McClements, 2016). The images were taken at 100x magnification.

2.4.3. Apparent viscosity measurements

The apparent viscosity of initial shakes and their respective digests was determined by using Rheostress 6000 rheometer (RheoStress 6000, HAAKE, Karlsruhe, Germany) equipped with parallel plate geometry (diameter = 40 mm, gap = 1 mm). The determination temperature was 25 ± 0.1 °C, and equilibrating program for 5 min was required. The shear stress was recorded over a range of shear rates between 0.01 and 100 s⁻¹ within 2 min. The apparent viscosity was recorded when the shear rate was 15 s⁻¹ (Naumann et al., 2019) as this value is close to the shear rates in gastrointestinal tract.

2.4.4. Lipid lipolysis evaluation

The lipid was extracted from the initial shakes and the digested samples according to Chinese standard GB 5009.6–2016. Briefly, initial shakes (10 g) or intestinal digesta (the total system) were mixed with 10 mL of ammonium hydroxide solution, and the mixtures were immediately incubated at a shaking water bath (65 ± 5 °C, 100 rpm) for 20 min. The mixtures were left to sit until reached to the room temperature, and 10 mL of ethanol was added to the system. After the gentle mixing, 25 mL of diethyl ether and 25 mL of petroleum ether were further added, and followed with the vigorously shaking. After centrifugation (10611 g, 3 min), the supernatants were collected. The extraction processes were repeated for three times and the supernatants were combined. The extracted lipids were obtained after the organic phase were removed using nitrogen blowing and were stored at −4 °C before the characterization. The acylglycerol profile, fatty acid composition, and glycerolipid composition of the extracted lipids were determined according to our previously reported methods (Xu et al., 2023). The lipolysis degree and individual fatty acids release level of each shake were calculated with above data as described in our previous publication (Xu et al., 2023).

2.4.5. Bioavailability determination of fat-soluble vitamins

The bioavailability of fat-soluble vitamin normally refers to the portion of lipophilic bioactive compounds that become solubilized in the mixed micelle phase after the digestion (Yang and McClements, 2013). Therefore, the micellar fraction of the final digesta was firstly isolated. The two-centrifugation step were adopted according to Werner and Böhm (2011). After in vitro digestion, the total digesta was transferred to the centrifuge tube, and the reaction vessel was washed with appropriate amount of distillated water. Afterwards, the mixture was centrifuged at 2148 g for 20 min (10 °C) and again centrifuged at 20798 g for 5 min (25 °C). The micellar phase was finally obtained after passing the supernatant through a 0.22 μm filter.

The fat-soluble vitamins were extracted from initial shakes, intestinal digesta and also micellar phases according to the procedure reported by Yuan et al. (2017) with small modifications. Firstly, the lipid-soluble components were extracted according to section 2.4.4. Secondly, 20 mg of the extracted mixtures were mixed with 2 mg of the TBHQ and 3 mL of sodium hydroxide-methanol solution (0.15 M). After vertexing for 30 s, the resulting solution were incubated at 70 °C for 40 min. After cooled to room temperature, 2 mL of distillated water and 3 mL of n-hexane were added to the system, and followed with vigorously shaking. Furthermore, the supernatants were collected after the centrifugation process (10611 g, 3 min). The extraction processes were repeated for three times, and the supernatants were combined and dried with nitrogen. The dried samples were again dissolved with methanol/isopropanol (1:1, v/v) and filtered with a 0.22 μm filter before the HPLC analysis of fat-soluble vitamins.

The vitamin content was determined using a reversed-phase high-performance liquid chromatography (HPLC) system (e2695; Waters, USA) equipped with a Waters 2489UV/VIS detector and a C18 column (i.d. 250 × 4.6 mm, particle size: 5 μm, SunFire). Chromatographic separation was achieved within 15 min and the column temperature was 30 °C. A mixture of methanol/isopropanol (9:1, v/v) was used as mobile phase at a flow rate of 1 mL/min. The detection wavelength for vitamin A and E was 325 and 294 nm, respectively. The vitamin content was calculated based on the calibration curves which were created in the range of 0.01–0.25 mg/mL and 0.25–5.0 μg/mL. Vitamin bioavailability was determined using Eq. (1) (Tan et al., 2021):

$$\mathrm{B}\mathrm{i}\mathrm{o}\mathrm{a}\mathrm{v}\mathrm{a}\mathrm{i}\mathrm{l}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}=\frac{Cmicelle}{Cinitial}*100\%$$ (1)

here, Cmicelle represents the vitamin concentration in the micelle phases (mg/mL or μg/mL); Cinitial represents the vitamin concentration in the initial samples (undigested) (mg/mL or μg/mL).

2.5. Statistical analysis

The measurements were conducted in triplicate and the results were expressed as mean values ± standard deviations. Analysis of variance (ANOVA) was performed using SPSS software (version 16.0), and Tukey’ test was adopted for analyzing the significant difference (p < 0.05). Images processing was performed using Origin software (version 8.0).

3. Results and discussion

3.1. Droplet characterization of different meal-replacement shakes and digesta

3.1.1. Particle size and microstructure analysis

The changes of mean particle sizes and microstructure of different meal-replacement shakes and their digesta during in vitro digestion are shown in Fig. 1. Initially, all the samples were optically opaque which suggesting these shakes contained relatively large droplets (Mayer et al., 2013). Accordingly, the mean particle sizes of these samples were pretty large and decreased in order: sample #E (9.34 ± 0.65 μm) > sample #D (7.35 ± 0.30 μm) > sample #B (6.06 ± 0.46 μm) > sample #A (3.97 ± 0.02 μm) > sample #C (1.46 ± 0.20 μm) (Fig. 1A, p < 0.05). The oil content in meal-replacement shakes largely determined the initial particle size (Olson et al., 2004), as a consequence, sample #C with the lowest oil content exhibited the smallest particle size, while sample #E with the highest oil content displayed the largest particle size (Table 1). The confocal micrographs also clearly indicated a significantly larger droplet size in sample #E when compared to sample #C (Fig. 1B). Despite having similar oil content, the remaining three samples showed variations in particle size. Notably, sample #A exhibited a smaller particle size, potentially attributed to its comparatively higher protein and cellulose content, which can function as emulsifiers (Espinal-Ruiz et al., 2014, Table 1). Furthermore, the confocal images suggested that there were large flocs presented in all the initial shakes. The non-absorbed polymers could lead to the osmotic attraction between droplets and contribute to the generation of these large flocs (McClements, 2000).

Fig. 1.

The changes of mean droplet sizes (A) and microstructure (B) of five commercial meal-replacements shakes during the simulated digestion.

PS: The sample names follow a specific pattern, with the first letter denoting the sample and the second letter indicating the digestion stage (R: raw material; O: oral digesta; S: gastric digesta; I: intestinal digesta). For example: AR represents the raw material for sample #A; AO represents the oral digesta for sample #A; AS represents the gastric digesta for sample #A; AI represents the intestinal digesta for sample #A. Different lowercase letters (a–d) represent the statistical differences in different digestion stages of the same sample (p < 0.05), and the uppercase letters (A–E) represent the differences in the same digestion stage among different samples (p < 0.05). Each type of meal-replacement shake was represented by three samples, and each sample was analyzed in triplicate.

Following the oral stage, the mean particle sizes for all the samples exhibited an increase (p < 0.05). For the liquid foods, the oral digestion mainly focuses on the mixing of the saliva and the foods (Minekus et al., 2014). The inclusion of SSF, which contains inorganic salts, could cause electrostatic screening and result in reduced repulsion between droplets. This decrease in repulsive forces ultimately led to droplet flocculation (Chang and McClements, 2016). However, following the gastric digestion stage, the five shakes exhibited different patterns of changes. The particle size of sample #A and #C showed an appreciable increase after this stage (p < 0.05). The microstructure images also exhibited large oil droplets (Fig. 1B), which indicated the occurrence of the coalescence (Chang and McClements, 2016). This type of emulsion instability occurred for mainly two reasons: (i) the changes of pH (3.0 in stomach) and ionic strength could weaken the electrostatic repulsion; (ii) the hydrolysis of protein in system may reduce the stability of the system and lead to coalescence (Chang and McClements, 2016). When compared with above two samples, samples #B, #D and #E showed lower level of protein, and as a result, cellulose might occupy a greater interface which can reduce the possibility of flocculation occurrence after gastric digestion (Espinal-Ruiz et al., 2014). And the hydrolysis of proteins might lead to the decrease of the droplets in these three samples (p < 0.05).

Following the intestinal digestion, a reduction in particle size was observed for nearly all samples (except sample #E), primarily attributed to the extensive hydrolysis of lipids (Yang et al., 2015). There are barely any noticeable red droplets in the confocal images, where the red color indicates stained lipids. Meanwhile, the presence of bile salts could remove the hydrolysis products (mainly FFA and monoacylglycerol (MAG)) from the interface, which could finally lead to lower particle size (Yang et al., 2015). For sample #E, the lowest level of lipolysis degree and the highest level of cellulose might be charge for this different trend. The cellulose derivatives can bind to bile salts in digestion system and form interfacial networks which may result in a reduced in a reduced level of lipolysis (Zornjak et al., 2020). Additionally, the presence of large amounts of fibers could potentially promote the lipid droplet aggregation (Qin et al., 2017). The particle size of the intestinal digesta decreased in order: sample #C > sample #B, #E > sample #A, #D, and the microstructure images also confirmed this result. In sample #C, there were noticeable large aggregates present throughout the entire digestion process (Fig. 1B). These aggregates appeared to be proteins which were not fully digested and absorbed, while further research was needed. As a result, sample C exhibited the largest particle size after intestinal digestion.

3.1.2. ζ-potential

The ζ-potential provides insights into the charge at the interface, which could greatly impact the emulsion stability and lipid hydrolysis degree (Infantes-Garcia et al., 2021). The changes of ζ-potential during the simulated digestion are shown in Fig. 2. Initially, all of the samples were negatively charged (−9.84 to −24.00 mV) because the pH was largely higher than the isoelectric point of most of the proteins (Chang and McClements, 2016). While the magnitude of their electrical charge was distinct, primarily due to the variations in emulsifier content and type among different products. Thus, sample #C exhibited the highest absolute value of ζ-potential due to the presence of highest level of protein (Li et al., 2019). In contrast, sample #A showed the lowest absolute level even with a pretty high level of protein. The presence of diacylglycerol (DAG) in this sample could explain for this phenomenon. DAG, besides being a natural component of edible oil, also functions as a small molecule emulsifier. It can rapidly occupy the interface during emulsion preparation process and thus potentially affecting the protein content at the interface, and subsequently reduce the absolute ζ-potential value (Chang and Lee, 2019). However, the adsorption stability of DAG at the interface is relatively low, which means that as storage time or digestion progresses, other existing emulsifiers may eventually displace DAG from the interface (Jiang et al., 2019).

Fig. 2.

The changes of ζ-potential values of five commercial meal-replacements shakes during the simulated digestion.

PS: The sample names follow a specific pattern, with the first letter denoting the sample and the second letter indicating the digestion stage (R: raw material; O: oral digesta; S: gastric digesta; I: intestinal digesta). For example: AR represents the raw material for sample #A; AO represents the oral digesta for sample #A; AS represents the gastric digesta for sample #A; AI represents the intestinal digesta for sample #A. Different lowercase letters (a–d) represent the statistical differences in different digestion stages of the same sample (p < 0.05), and the uppercase letters (A–E) represent the differences in the same digestion stage among different samples (p < 0.05). Significant differences here were calculated based on absolute values. Each type of meal-replacement shake was represented by three samples, and each sample was analyzed in triplicate.

After the dilution of initial meal-replacement shakes with SSF for the oral digestion, a decrease in the absolute charge for all samples was observed. The electrostatic screening mentioned above could explain for this phenomenon (Chang and McClements, 2016). Meanwhile, the change of the environment might also cause the dissociation of protein molecules from the droplet surfaces (Qin et al., 2017). After gastric digestion, the ζ-potential for all samples were turned to positive. This was mainly attributed the low pH in the gastric phase (∼3, Infantes-Garcia et al., 2021). Furthermore, a decrease in the magnitude of the ζ-potential was observed for nearly all samples (except sample #E). A portion of the surface proteins had been hydrolyzed during the gastric digestion, which could lead to lower electrostatic repulsion between oil droplets. This result was also reported by Chang and McClements (2016). Meanwhile, this phenomenon could be attributed to relatively high ionic strength of the SGF, which led to electrostatic screening effects (Chang and McClements, 2016). In contrast, sample #E showed a different trend. The high level of cellulose might be contributed to this result. As some of the charged cellulose molecules might occupy the interface after protein hydrolysis, thereby imparting a higher electrical potential to the sample #E.

Following the intestinal digestion, the interface was mainly covered by bile slats, cellulose and new produced FFA (Mayer et al., 2013), and thus all the samples showed negatively charged (Fig. 2). Sample #D and #E showed higher ζ-potential than other three samples. The highest oil content, which may generate more FFA (Chang and McClements, 2016), resulted in the highest absolute potential value of sample #E. In the case of sample #D, which contained lowest level of cellulose, the dominant presence of bile salts at the interface could have led to this result (Mun et al., 2007). Sample #A showed the lowest absolute value, the high level of calcium ions should be responsible for this result (Table 1). Calcium ions showed high ability to precipitate some anionic ions (including FFA) at the interface, and removing them from the surface (Yang et al., 2015).

3.2. Lipid lipolysis

Controlling lipid digestion is widely recognized as a promising approach to address a myriad of diseases related with obesity (Zornjak et al., 2020). The aim of the meal-replacement shakes is to control the obesity and promote fitness. Therefore, the levels of lipolysis in these five shakes hold significant difference. Meanwhile, the lipid digestion fates could also impact the formation of mixed micelle and thus affect the final vitamin bioavailability (Lv et al., 2019). First of all, all the samples showed similar acid release trend based on the pH-stat method (Fig. 3A). It was observed a rapid increase of acid during the first 10 min of digestion, followed by a gradual increase as time went on until an almost stable state was reached. This phenomenon was widely reported in the digestion studies (Zornjak et al., 2020; Xu et al., 2023; Lv et al., 2019). The hydrolysis products generated during the digestion process could accumulate at the interface and thus interfere the hydrolysis of the remaining lipid droplets (Martin et al., 2014). The negatively charged interfaces could also confirmed this explanation (Mayer et al., 2013). However, due to the presence of protein in these meal-replacement shakes, some acidic amino acids can also be produced. As a consequence, although the pH-stat method can reveal the trend of the acid production during the digestion process, it cannot quantify the release amounts of FFA accurately. Meanwhile, the calculation method of pH-stat, which assumes hydrolysis of triacylglycerol (TAG) into MAG and FFA, may not provide accurate results (Qin et al., 2017). Analyzing the acylglycerol profiles offers a more comprehensive understanding of lipid digestion fates within meal-replacement shakes (Martin et al., 2014, Fig. S1). Therefore, the lipolysis level was calculated based on the acylglycerol profiles in this study.

Fig. 3.

The volume of NaOH consumed for neutralization of the released acids in the five commercial meal-replacement shakes within 120 min digestion time (A); Lipolysis level (B) and Hydrolyzed products amounts (C) of the five commercial meal-replacement shakes at the end of intestinal digestion. Abbreviations: #A, #B, #C, #D, and #E represent the five commercial meal-replacement shakes, respectively. Different lowercase letters (a–d) represent the statistical differences between different samples (p < 0.05). Each type of meal-replacement shake was represented by three samples, and each sample was analyzed in triplicate.

Notably, the lipolysis level for nearly all shakes were significantly lower (37.33–61.42% for sample #B to #E, Fig. 3B) when compared with normal emulsions, which was corresponded to the aim of these products. Sample #A exhibited pretty high lipolysis level (80.83 ± 2.59%, p < 0.05) due to several reasons. Firstly, the DAG presented in the sample #A was easier to be digested when compared with traditional lipids (mainly contained TAG) (Diao et al., 2021), and meanwhile DAG could function as emulsifier, enhancing the lipolysis degree of the lipids (Martin et al., 2014; Xu et al., 2023). Secondly, high level of calcium ions in the sample #A could accelerate the rate of FFA detachment from the interface by binding and precipitating them (Table 1, Yang et al., 2015). Moreover, the smallest droplet size of sample #A at intestinal stage also favored the rapid lipid digestion due to the larger specific surface area (Chang and McClements, 2016, Fig. 1A). While this high lipolysis level may not be friendly to the obese population, the low-calorie property of DAG could potentially compensate for this shortcoming (Xu et al., 2023). Afterwards, there was not a significant difference in the lipolysis degree between samples #B and #D (58.12 ± 2.54% and 61.42 ± 1.09%, respectively, p > 0.05). These two samples had low level of cellulose, while the proteins were nearly completely hydrolyzed, resulting in reduced effectiveness at preventing the attachment of the bile salts and pancreatic lipase to the surfaces (Lv et al., 2019). Therefore, these two samples exhibited higher lipolysis level than the remaining two samples. Furthermore, sample #C and sample #E exhibited the lowest lipolysis level (38.76 ± 0.79% and 37.33 ± 2.77%, respectively, p > 0.05). This was related with the highest viscosity of the digestion system of sample #C, which could potentially result in slower mass transfer, impeding the interaction between lipase and the interface, and might also inhibit the lipase activity (Fig. S2, Calvo-Lerma et al., 2018). For sample #E, the high level of cellulose (derivates) could retard the lipid digestion by binding the bile salts (Zornjak et al., 2020). Meanwhile, the presence of high level of cellulose might form a strong interfacial film, thereby suppressing the interaction between lipase and the lipid droplets (Qin et al., 2017). Notably, these two samples might show the potential to increase the feelings of satiety and satiation, and decrease the intake of foods (Espinal-Ruiz et al., 2014).

3.3. Individual fatty acid release level and lipidomic profiles

The nutritional values of edible oil are significantly influenced by fatty acid composition, for example, lipids rich in ω-3 fatty acids (such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)) are normally regarded as healthy lipids, while the lipids rich in saturated fatty acids are typically considered detrimental, as they have the potential to elevate the risk of certain types of cancers (Denke, 2006). Nonetheless, it is not the quantity of fatty acids consumed, but rather the fatty acids that are effectively absorbed in the intestines after digestion, that truly influences the nutritional values of lipids (Tan et al., 2021). The fatty acid release level greatly governs the bioaccessibility of the fatty acids within lipids and therefore were determined and analyzed (Fig. 4). The main fatty acids in sample #A and sample #C were long-chain fatty acids, mainly including palmitic acid, stearic acid, oleic acid and linoleic acid, with the level of 16.33 ± 2.70%, 9.15 ± 1.60%, 51.86 ± 3.86% and 12.73 ± 1.12% for sample #A, and 21.21 ± 1.21%, 6.25 ± 0.89%, 22.96 ± 1.52% and 34.80 ± 1.87% for sample #C, respectively (Fig. 4A). In contrast, other three samples mainly contained medium-chain fatty acids (MCFAs, lauric acid and myristic acid), while also enriched in palmitic acid and/or stearic acid (Fig. 4A). It is impossible to confirm whether MCFAs are present in the form of MCTs or medium- and long-chain triglycerides (MLCTs) from the GC data, and thus the lipidomic profiles of the lipids in initial shakes were determined and analyzed (Fig. S3). The lauric acid and myristic acid were primarily observed as MCTs in sample #B, with MCTs/MLCTs ratio of 4.98 ± 1.05 and 1.64 ± 0.23, respectively (Table S1). The lauric acid in other two samples also existed as MCTs, with MCTs/MLCTs ratio of 1.17 ± 0.02 for sample #D and 1.64 ± 0.07 for sample #E, respectively. While the myristic acid mainly presented as MLCTs, with MCTs/MLCTs ratio of 0.21 ± 0.02 for sample #D and 0.65 ± 0.06 for sample #E, respectively (Table S1). Although it is widely considered that the MCTs can provide quick energy and aid in weight loss, the prolonged intake of large amounts of MCTs can lead to adverse gastrointestinal symptoms, including gastrointestinal discomfort, abdominal cramps, and osmotic diarrhea, and can even lead to the accumulation of ketone (Marten et al., 2006). MLCTs exhibit notable advantages over MCTs, it could not only slow down the release rate of MCFAs, but also provide essential fatty acids for human body. Therefore, the usage of MLCTs might be a potential better choice for these commercial products. Moreover, these three samples displaced a notably high level of saturated fatty acids (69.38–89.44%, Table S2), which does not meet the current nutritional requirements (World Health Organization, 2018). It is necessary to decrease the level of saturated fats in these three samples. Notably, all five meal-replacement shakes showed a much lower α-linolenic acid (ALA) level than the recommended level (1.1–1.6 g/d, Institute of Medicine, 2005). The deficiency of ALA could potentially result in developmental issues for brain and skin (Nguemeni et al., 2013). Even without factoring the adsorption rate of fatty acids, it is recommended to further incorporate lipids rich in ALA for commercial meal-replacement shakes. It should also be pointed out that DHA, EPA, and other long-chain polyunsaturated fatty acids (LCPUFAs) were not detected in any of these products. The long-term deficiency of these fatty acids can potentially give rise to severe health complications (Barceló-Coblijn and Murphy, 2009). It is recommended to incorporate fish oil or other lipids rich in LCPUFAs, or manufacturers should provide guidance to consumers on supplementing foods rich in LCPUFAs when consuming these meal-replacement shakes.

Fig. 4.

(A) Fatty acid compositions of lipids extracted from the five commercial meal-replacement shakes; (B) The release level of individual fatty acid (group 1) from the five commercial meal-replacement shakes at the end of intestinal digestion (the additional two sets of data are provided in the supplementary file, Fig. S4). Abbreviations: #A, #B, #C, #D, and #E represent the five commercial meal-replacement shakes, respectively; C8:0: caprylic acid; C10:0: capric acid; C12:0: lauric acid; C14:0: myristic acid; C16:0: palmitic acid; C18:0: stearic acid; C18:1: oleic acid; C18:2: linoleic acid; α-C18:3: α-linolenic acid. Each type of meal-replacement shake was represented by three samples, and each sample was analyzed in triplicate.

During the intestinal digestion, the glycerolipid structure significantly influenced the release level of individual fatty acids (Ye et al., 2019). Generally, the saturated fatty acids showed higher release level than unsaturated ones (Ye et al., 2019). However, palmitic acid and stearic acid in sample #A were mainly existed as SSS and SSU (with 42.92 ± 2.33% and 35.29 ± 1.06% for palmitic acid and 68.98 ± 1.07% and 26.68 ± 1.56% for stearic acid, respectively, Table 2), the higher melting point of these two forms of TAGs could decrease the release level of these two fatty acids (9.44 ± 2.81% and 20.23 ± 3.12%, respectively, Fig. 4B). While the oleic acid, linoleic acid and ALA mainly esterified with unsaturated fatty acids (the form of UUU, 61.36 ± 2.51%, 46.92 ± 4.80%, 72.85 ± 6.55%, respectively). The low melting point of UUU makes them more susceptible to hydrolysis and thus led to higher release percentages of unsaturated fatty acids (57.84 ± 2.05% for oleic acid, 38.25 ± 0.19% for linoleic acid, and 34.95 ± 4.47% for ALA). Similar phenomenon was observed for sample #C. This release pattern can facilitate the absorption of unsaturated fatty acids, reduce the absorption of saturated fatty acids, and enhance the health properties of these products. For sample #B, #D and #E, the release levels of MCFAs (mainly focused on the fatty acids with large amount, including lauric acid and myristic acid) were between 23.16%–33.94%. In theory, a high release percentage of MCFAs would be expected due to the preference of pancreatic lipase to MCFAs (Pereira et al., 2023). However, in this study, MCFAs were also esterified with LCFAs (mainly saturated fatty acids, Fig. S2), and might potentially preventing extensive MCFAs hydrolysis (Yuan and Geng, 2020). The hydrolyzed long chain fatty acids could adsorb on the droplet interface (Martin et al., 2014) and interfered with the release of MCFAs.

Table 2.

Glycerolipid form of lipids extracted from the five commercial meal-replacement shakes before and after three-phase digestion.

Fatty acid	Glycerolipid form	#A	#AD	#B	#BD	#C	#CD	#D	#DD	#E	#ED
Caprylic acid	SSS (%)	0	0	64.89 ± 2.39	0	52.67 ± 2.49	0	100 ± 0.00	0	93.50 ± 1.29	83.42 ± 1.55
	SSU (%)	64.92 ± 5.00	0	26.26 ± 2.50	0	47.33 ± 2.49	0	0	0	6.50 ± 1.29	16.58 ± 1.55
	SUU (%)	35.08 ± 5.00	0	8.85 ± 0.15	0	0	0	0	0	0	0
Capric acid	SSS (%)	32.95 ± 4.48	100 ± 0	70.47 ± 3.33	0	26.34 ± 2.40	10.83 ± 1.51	94.94 ± 1.02	100 ± 0.00	86.22 ± 1.51	75.78 ± 2.17
	SSU (%)	67.05 ± 4.48	0	29.53 ± 3.33	0	73.66 ± 2.40	89.17 ± 1.51	5.06 ± 1.02	0 ± 0	13.78 ± 1.51	24.22 ± 2.17
	SUU (%)	0	0	0	0	0	0	0	0	0	0
Lauric acid	SSS (%)	27.88 ± 4.19	69.66 ± 2.63	88.55 ± 1.98	87.88 ± 3.17	33.11 ± 2.42	42.42 ± 2.23	99.60 ± 0.09	98.77 ± 0.12	96.92 ± 0.37	98.38 ± 0.29
	SSU (%)	72.12 ± 4.19	30.34 ± 2.63	9.35 ± 1.83	10.08 ± 3.04	66.89 ± 2.42	57.58 ± 2.23	0.40 ± 0.090	1.23 ± 0.12	3.08 ± 0.37	1.62 ± 0.29
	SUU (%)	0	0	2.1 ± 0.18	2.05 ± 0.24	0	0	0	0	0	0
Myristic acid	SSS (%)	39.76 ± 6.34	53.82 ± 2.68	85.21 ± 2.45	82.34 ± 2.14	82.85 ± 2.58	90.39 ± 2.89	70.99 ± 4.96	70.65 ± 3.01	64.5 ± 1.69	66.36 ± 2.14
	SSU (%)	32.75 ± 2.47	28.42 ± 0.70	14.79 ± 2.45	17.66 ± 2.14	17.15 ± 2.58	9.61 ± 2.89	29.01 ± 4.96	29.35 ± 3.01	24.56 ± 0.95	25.41 ± 1.46
	SUU (%)	27.49 ± 3.87	17.77 ± 2.20	0	0	0	0	0	0	10.94 ± 0.83	8.23 ± 0.69
Palmitic acid	SSS (%)	42.92 ± 2.33	61.15 ± 3.02	52.70 ± 2.02	61.60 ± 0.5	38.92 ± 1.92	49.28 ± 0.97	54.48 ± 1.06	64.31 ± 1.32	76.31 ± 1.80	88.65 ± 0.66
	SSU (%)	35.29 ± 1.06	29.93 ± 2.09	20.41 ± 2.7	23.21 ± 1.43	36.26 ± 1.97	35.41 ± 1.43	37.44 ± 1.27	30.15 ± 0.98	9.26 ± 0.52	7.55 ± 0.28
	SUU (%)	21.78 ± 1.28	8.91 ± 0.93	26.89 ± 3.57	15.19 ± 1.2	24.83 ± 0.19	15.31 ± 0.81	8.08 ± 0.82	5.54 ± 0.55	14.43 ± 1.3	3.80 ± 0.87
Stearic acid	SSS (%)	68.98 ± 1.07	73.91 ± 0.59	58.99 ± 2.24	64.5 ± 0.14	80.96 ± 2.49	89.06 ± 0.64	93.23 ± 0.85	94.52 ± 0.17	82.04 ± 1.32	88.73 ± 1.09
	SSU (%)	26.68 ± 1.56	24.07 ± 0.29	25.87 ± 1.79	27.21 ± 0.49	14.72 ± 2.19	8.59 ± 0.53	6.77 ± 0.85	5.48 ± 0.17	17.96 ± 1.32	11.27 ± 1.09
	SUU (%)	4.35 ± 0.69	2.02 ± 0.37	15.14 ± 1.24	8.29 ± 0.54	4.33 ± 0.52	2.35 ± 0.22	0	0	0	0
Oleic acid	SSU (%)	13.7 ± 1.48	18.03 ± 2.07	28.75 ± 4.24	38.64 ± 3.01	22.84 ± 3.50	27.28 ± 5.19	46.54 ± 1.48	55.58 ± 1.65	40.02 ± 2.01	53.05 ± 0.54
	SUU (%)	24.94 ± 1.28	22.9 ± 1.12	46.46 ± 1.26	44.08 ± 1.39	27.53 ± 2.60	27.49 ± 3.91	27.42 ± 0.76	28.36 ± 1.10	47.74 ± 2.56	36.24 ± 2.42
	UUU (%)	61.36 ± 2.51	59.07 ± 2.46	24.79 ± 5.48	17.27 ± 4.23	49.62 ± 5.84	45.23 ± 8.75	26.03 ± 1.39	16.06 ± 0.99	12.24 ± 4.30	10.71 ± 2.21
Linoleic acid	SSU (%)	11.16 ± 1.43	23.82 ± 2.78	7.66 ± 1.60	20.18 ± 0.58	6.76 ± 1.00	14.03 ± 2.65	39.59 ± 4.96	47.24 ± 3.39	13.5 ± 1.80	22.7 ± 2.22
	SUU (%)	41.92 ± 3.40	30.75 ± 3.02	41.34 ± 1.25	43.39 ± 4.23	26.55 ± 3.98	29.2 ± 4.55	26.32 ± 4.13	24.52 ± 2.80	46.33 ± 8.81	33.4 ± 6.12
	UUU (%)	46.92 ± 4.80	45.43 ± 5.56	51 ± 0.43	36.43 ± 3.77	66.69 ± 4.86	56.77 ± 6.89	34.09 ± 3.88	28.24 ± 3.08	40.16 ± 10.47	43.89 ± 8.17
α-Linolenic acid	SSU (%)	27.15 ± 6.55	0	22.04 ± 6.29	0	0	0	0	0	0	0
	SUU (%)	0	0	7.66 ± 0.94	18.5 ± 4.45	4.48 ± 0.69	6.71 ± 1.03	0	0	100 ± 0.00	100 ± 0.00
	UUU (%)	72.85 ± 6.55	100 ± 0	70.3 ± 6.68	81.5 ± 4.45	95.52 ± 0.69	93.29 ± 1.03	100 ± 0.00	100 ± 0.00	0	0

Abbreviations: #A, #B, #C, #D, and #E represent the five commercial meal-replacement shakes, respectively; #AD, #BD, #CD, #DD, and #ED represent the five commercial meal-replacement shakes after simulated digestion; UUU: triunsaturated-TAG; UUS: diunsaturated-TAG; USS: monounsaturated-TAG; SSS: trisaturated-TAG. Each type of meal-replacement shake was represented by three samples, and each sample was analyzed in triplicate.

3.4. Fat-soluble vitamins bioavailability evaluation

Fat-soluble vitamins deficiencies are prevalent across various age groups, exerting a significant influence on human health (Tan et al., 2021). The majority of consumers often overlook of the adsorption efficiency of the nutrients, playing a primary emphasis on the quantity consumed (Chug-Ahuja et al., 1993). However, for the fat-soluble vitamins characterized by low water solubility and poor bioavailability (Yang et al., 2015), this supplementing approach may result in server deficiencies. As these shakes are designed to serve as complete replacements for regular diets, the bioavailability of fat-soluble vitamin emerged as an important indicator.

To begin with, all the samples exhibited comparable visual phenomenon following centrifugation. Only a sediment layer at the bottom and a slightly turbid layer above (known as micellar phase) were observed, while no cream layer was evident. The low level of lipid in the digestion system should be responsible for these results. Yang et al. (2015) and Mayer et al. (2013) also reported similar phenomenon. When considering the bioavailability, sample #E showed the highest value (58.31 ± 1.98% for vitamin A, and 56.48 ± 0.69% for vitamin E, respectively), following with sample #A (38.13 ± 1.07% for vitamin A, and 45.80 ± 2.55% for vitamin E, respectively, Fig. 5A). The strong positive correlation between the quantities of lipid hydrolyzed products (FFA and MAG) and the bioavailability of fat-soluble vitamins was surprisingly observed (vitamin A: p = 0.041, r = 0.894; vitamin E: p = 0.045, r = 0.887, the range of r values from 0.7 to 0.9 indicates a strong correlation). Therefore, sample #E, which displaced the highest level of hydrolyzed products (3.35 × 10⁻⁴ ± 1.04 × 10⁻⁵, Fig. 3C) at the end of intestinal digestion, exhibited the highest bioavailability value. The larger amounts of lipid hydrolyzed products have the potential to enhance the form of mixed micelles, thereby enabling the solubilization of greater amounts of the fat-soluble vitamins (Yang and McClements, 2013). Yang et al. (2015) reported similar results. The lowest bioavailability of sample #C (16.67 ± 2.23% for vitamin A and 23.89 ± 0.75% for vitamin E, respectively) could also confirm this explanation. Although the sample #A, #B and #D produced comparable levels of lipid hydrolyzed products at the end of digestion, sample #A showed significantly higher fat-soluble vitamins bioavailability. This phenomenon could be attributed to the enhanced solubilization of mixed micelles primarily formed with the long chain fatty acids than MCFAs (Yang and McClements, 2013). Therefore, the utilization of MLCTs might not be a favorable option for delivering fat-soluble vitamins. Similar recommendations were also provided by Yang and McClements (2013). However, while the usage of LCTs might enhance the delivery of fat-soluble vitamins, it could potentially affect the weight loss effects of these shakes. It is proposed that the DAG-based shakes might show some advantages, while further researches are required. Notably, sample #A could provide enough daily fat-soluble vitamins (only vitamin A and vitamin E considered in this study, 490 μg/d for vitamin A, and 11 mg/d for vitamin E, EFSA Panel on Dietetic Products, Nutrition, and Allergies, 2015a, EFSA Panel on Dietetic Products, Nutrition, and Allergies, 2015b) when accounting for adsorption rate (Fig. 5B), which was also due to the high level of these two vitamins in sample #A. However, none of the other samples can provide enough fat-soluble vitamins when they were treated as a complete diet, and additional vitamin supplementation is required. The findings are surprising, as our initial assumption was that the consumption of these products could result in more pronounced fat-soluble vitamin deficiencies. In this light, these shakes hold the potential to provide adequate nutrients and emerge as a healthy dietary choice in our swiftly evolving era based on scientific calculations. Nonetheless, more research is needed to confirm these results and fully understand the enduring effects of these products on human health.

Fig. 5.

Bioavailability of fat-soluble vitamins (A) and Absorbable content of fat-soluble vitamins within one diet (calculated based on 100 g of the products) (B) after the simulated digestion for five commercial meal-replacement shakes. Abbreviations: #A, #B, #C, #D, and #E represent the five commercial meal-replacement shakes, respectively. Different lowercase letters (a–d) represent the statistical differences between different samples for the same vitamin (p < 0.05). Each type of meal-replacement shake was represented by three samples, and each sample was analyzed in triplicate.

4. Conclusions

This study aimed to investigate the nutritional attributes of commercial available meal-replacement shakes through the application of an in vitro digestion model. Initially, the lipid compositions of extracted lipids from the products were analyzed, revealing that the sample #A was enriched with DAG, while three others (sample #B, #D and #E) contained higher levels of MCFAs, and the remaining one (sample #C) contained mainly traditional LCTs. Next, the lipolysis level of each shake was assessed. Four out of the five samples exhibited notably low lipolysis levels, consistent with the characteristics of the weight-loss products. Although the sample #A showed the highest lipolysis level, the inherent low-calorie nature of DAG in this sample appeared to compensate this drawback effectively. Furthermore, a strong positive correlation between the amounts of lipid hydrolyzed products (MAG and FFA) and the bioavailability of fat-soluble vitamins was observed. The sample #E produced highest amounts of lipid hydrolyzed products, which could improve the solubilization of fat-soluble vitamins in micelles phases, and thereby improving the bioavailability. Meanwhile, sample #A demonstrated elevated vitamin bioavailability, even though its lipid hydrolyzed products levels were comparable to those of sample #B and #D. The richness of MCFAs in samples #B and #D, which have significantly lower solubility for fat-soluble vitamins than long chain fatty acids, and thus might lead to this phenomenon. Therefore, in the context of two commonly used low-calorie lipids, DAG might show better potential than MLCTs for integration into meal-replacement shakes. Notably, sample #A could provide enough vitamin A and vitamin E for dieters when it was used as total replacement of one or two normal meals. In conclusion, despite the identified drawbacks observed in the aforementioned outcomes, meal-replacement shakes retain significant promise as nutritious dietary options in this fast-paced ear if their formula is accurately design and calculation. Moreover, additional research is necessary to provide data or information for enhanced design. Both animal and human feeding studies are required to confirm the outcomes of the in vitro study.

CRediT authorship contribution statement

Qingqing Xu: Conceptualization, Data curation, Formal analysis, Methodology, Software, Validation, Writing – original draft. Weifei Wang: Resources, Methodology, Supervision. Dongxiao Sun-Waterhouse: Conceptualization, Data curation, Formal analysis. Qian Zou: Formal analysis, Data curation. Menglei Yan: Formal analysis, Investigation, Data curation. Xuan Liu: Software, Validation, Data curation. Dongming Lan: Funding acquisition, Project administration. Yonghua Wang: Funding acquisition, Project administration, All authors read and approved the final manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Dr. Yeonhwa Park

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

The authors are unable or have chosen not to specify which data has been used.