Abstract
Steviol glycosides have gained attention as low-calorie sweeteners. The acceptable daily intake (ADI) for steviol glycosides is 4 mg/kg bw/day, requiring continuous exposure assessments. To generate occurrence data for exposure evaluations, this study validated a high-sensitivity HPLC- variable wavelength detector (VWD) analytical method for quantifying five steviol glycosides and applied the method to processed food products distributed in South Korea. The limit of detection (LOD) was in the range of 0.2–0.5 mg/L, and the limit of quantification (LOQ) was in the range of 0.7–1.5 mg/L. The measurement uncertainty was evaluated to improve the reliability of the analytical results. Among the steviol glycosides analyzed, rebaudioside A and D were the most (96.8%) and least (11.7%) frequently detected compounds, respectively, in processed foods. All negative samples were confirmed by UHPLC–MS/MS analysis. The analytical method and monitoring results presented in this study are expected to contribute to prospective exposure assessments.
Keywords: Food additive, Steviol glycosides, Method validation, Measurement uncertainty, Monitoring
Introduction
Steviol glycosides, derived mainly from the leaves of Stevia rebaudiana Bertoni, serve as low-calorie sweeteners. These compounds are categorized based on the glycosylation of their common aglycone, steviol, with more than 60 types of steviol glycosides identified to date (Purkayastha and Kwok, 2020). González et al. (2014) reported the composition of steviol glycosides in stevia leaves as stevioside (Stv, 4–14%), rebaudioside A (Reb A, 2–4%), rebaudioside C (Reb C, 1–2%), dulcoside A (0.4–0.7%), and minor components such as rebaudioside D, E, and F (Reb D, E, F, each < 0.4%), rubusoside (< 0.4%), and steviolbioside (< 0.4%). Stv exhibits 150–300 times the sweetness of sucrose but is characterized by a distinct bitter aftertaste, whereas Reb A exhibits a sweetness intensity 250–450 times greater than that of sucrose, with reduced bitterness. Reb C, D, and F are 50–120, 200–300, and 250–300 times sweeter than sucrose, respectively (Huang et al., 2024).
As low-calorie sweeteners, steviol glycosides are favored by consumers aiming to reduce sugar intake for purposes such as diabetes management and weight control. The U.S. Food and Drug Administration (FDA) has classified steviol glycosides as “generally recognized as safe (GRAS)” and approved their inclusion as sweeteners in diverse food and beverage formulations (FDA, 2018). Steviol glycosides are, therefore, widely used across diverse food categories. Additionally, recent studies have highlighted their potential functional benefits, including anti-inflammatory, anti-obesity, antidiabetic, antihypertensive, anticaries, and anticancer effects (Huang et al., 2024), making them a promising multifunctional ingredient for the development of functional food products. Given their versatility and wide-ranging potential applications, their use could substantially increase in forthcoming years, increasing exposure to these components.
Despite their GRAS status under FDA regulations, the Joint FAO/WHO Expert Committee on Food Additives (JECFA) has established an acceptable daily intake (ADI) for steviol glycosides of 4 mg/kg bw/day (expressed as steviol). Therefore, continuous monitoring of steviol glycoside intake is essential. To assess exposure levels, establishing reliable analytical methods for steviol glycosides continues to be essential. Existing quantification methods include HPLC–UV (Bililign et al., 2014; Jaworska et al., 2012), LC–MS/MS (Molina-Calle et al., 2016; Phungsiangdee et al., 2023), high-performance thin-layer chromatography (HPTLC) (Londhe et al., 2013; Saifi et al., 2014), 2D-ultra-HPLC (Cacciola et al., 2011), and near-infrared spectroscopy (NIR) (Hearn and Subedi, 2009). However, HPLC is the most popular tool for steviol glycoside quantification, necessitating ongoing improvements in its methodologies.
This study focuses on the validation of a highly sensitive analytical method for the quantification of steviol glycosides using HPLC-variable wavelength detector (VWD) and demonstrate its applicability by quantifying steviol glycosides in various food matrices. The total concentration of five steviol glycosides, expressed as steviol equivalents, is presented to support risk assessment studies concerning dietary exposure to steviol glycosides.
Materials and methods
Reagents and materials
Reb A (98% purity) was obtained from Toronto Research Chemicals (Toronto, Canada), all other standard compounds used in this study (Stv, Reb C, Reb F, and Reb D, all ≥ 98% purity) were purchased from ChemFaces (Wuhan, China). Potassium ferrocyanide trihydrate, used in the preparation of Carrez solution, was sourced from Kanto Chemical (Tokyo, Japan). Zinc sulfate heptahydrate was acquired from Sigma-Aldrich (St. Louis, MO, USA). Methanol (MeOH), water, and acetonitrile, all of HPLC grade, were obtained from J.T. Baker Chemical Co. (Radnor, PA, USA). Diethyl ether, as the solvent for fat removal from samples, was purchased from Junsei (Tokyo, Japan). The Sep-Pak C18 cartridge (500 mg, 6 cc) used for sample purification was acquired from Waters (Milford, MA, USA), and the Minisart RC15 syringe filter (0.45 µm) was sourced from Sartorius (Göttingen, Germany).
To prepare Carrez I solution, 15% potassium ferrocyanide trihydrate was dissolved in water, and Carrez II solution was prepared by dissolving 30% zinc sulfate heptahydrate in water.
Preparation of standard solutions
Five steviol glycosides (Stv, Reb C, Reb F, Reb A, and Reb D) were dissolved in MeOH to prepare a standard mix (STD 5mix) with a concentration of 100 mg/L and stored at −20 °C. For HPLC-VWD calibration curves, the STD 5mix (100 mg/L) was diluted with HPLC-grade water to 2, 5, 10, 25, and 50 mg/L for Stv, Reb C, Reb F, and Reb A and to 5, 10, 25, 50, and 100 mg/L for Reb D. For the UHPLC–MS/MS calibration curves, the STD 5mix was diluted to 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, and 1 mg/L using HPLC-grade water.
Sample preparation
Sample preparation and purification were modified and adapted from the method described by Kang et al. (2022). For low-fat, low-protein food samples (pickled food, fruit/vegetable beverages, other beverages, soju, and teas), carbonated samples were degassed through 10 min of sonication. The samples containing alcohol were subjected to heating in a 70 °C water bath for 15 min, then clarified using a 0.45 µm RC syringe filter. The homogenized sample (1 g) was diluted to 10 mL with distilled water.
For high-fat, high-protein food samples (sauces and dried fish/shellfish fillet products), the 5 g of homogenized sample was mixed with 10 mL of distilled water and 2 mL of Carrez I solution. After mixing, 2 mL of Carrez II solution was added, and the mixture was vortexed for 2 min. Froth was removed by sonication, and the mixture was diluted to 50 mL with distilled water. After centrifuging at 1968 g for 10 min (corresponding to 4000 rpm with a rotor radius of 11 cm), the supernatant was carefully transferred to another tube and subsequently mixed with 10 mL of diethyl ether. The mixture was subjected to centrifugation again at 1968 g for 10 min, and 10 mL of the centrifuged supernatant was collected.
The liquid extract of the sample was purified using a C18 cartridge (500 mg, 6 cc). Prior to use, the cartridge was activated by washing with 5 mL of MeOH, subsequently with 5 mL of distilled water. After activation, 10 mL of the sample solution was then applied to the cartridge. The cartridge was subsequently rinsed with 10 mL of 10% methanol, after which steviol glycosides were subjected to elution with 4 mL of 90% MeOH. The eluted solution was then concentrated to a volume of 1 mL using a nitrogen evaporator and passed through a 0.45 µm RC syringe filter.
Analytical method
Validation of the analytical method for steviol glycosides was performed using HPLC-VWD, with the analysis conducted on a Thermo Fisher Scientific Ultimate 3000 (Waltham, MA, USA) equipped with a Shodex Asahipak NH2P-50 4E column (4.6 × 250 mm, 5 μm; Tokyo, Japan). The instrumental condition was modified and adapted from Woelwer-Rieck et al. (2010). Five steviol glycosides were separated by isocratic elution using a mixture of acetonitrile and water (76:24, v/v) as the mobile phase with a flow rate of 1.0 mL/min. The temperature of the column was kept at 40 °C, and detection of steviol glycosides was performed at a wavelength of 210 nm. The running time was 30 min.
Specific detection of steviol glycosides in the samples was confirmed by UHPLC–MS/MS analysis using a Thermo Scientific Vanquish system coupled with a TSQ Quantis Mass Spectrometer (Thermo Fisher Scientific,Waltham, MA, USA). The instrumental condition was modified and adapted from Phungsiangdee et al. (2023). The following gradient elution method was applied with a binary mobile phase (A: 0.05% formic acid in water; B: 0.05% formic acid in acetonitrile): 0‒2 min, 10‒30% B; 2‒5 min, 30‒35% B; 5‒7 min, 35‒90% B; 7‒9 min, 90‒10% B; and 9‒13 min, 10% B. The flow rate was maintained at 0.4 mL/min, and a 5 µL volume was used for injection. An Agilent Eclipse Plus C18 column (2.1 × 50 mm, 1.8 μm; Santa Clara, CA, USA.) was used, with the temperature of column set at 30 °C. Steviol glycosides were ionized in negative electrospray ionization (ESI) mode and quantified using multiple reaction monitoring (MRM) with the following operating parameters: negative spray voltage, 3000 V; sheath gas, 50 a.u.; auxiliary gas, 10 a.u.; sweep gas, 1 a.u.; and vaporizer temperature, 350 °C. The MRM conditions, retention times (RT), collision energies, and RF lens voltages for the steviol glycosides are detailed in Table 1.
Table 1.
UHPLC-MS/MS multiple reaction monitoring (MRM) ion parameters and retention time (RT) of rebaudioside D, rebaudioside A, stevioside, rebaudioside F, and rebaudioside C
| Compounds | Retention time (min) | Molecular weight (m/z) | Precursor ion (m/z) | Product ion (m/z) | Collision energy (V) | RF lens (V) |
|---|---|---|---|---|---|---|
| Rebaudioside D | 5.87 | 1129.2 | 1127.5 | 803.2* | 47 | 299 |
| 641.2 | 55 | |||||
| Rebaudioside A | 6.63 | 967 | 965.5 | 803.3* | 26 | 299 |
| 317.2 | 55 | |||||
| 641.3 | 55 | |||||
| Stevioside | 6.66 | 804.9 | 803.4 | 641.3* | 19 | 262 |
| 317.1 | 55 | |||||
| 479.3 | 53 | |||||
| Rebaudioside F | 6.87 | 937 | 935.4 | 773.3* | 26 | 299 |
| 611.3 | 55 | |||||
| 641.3 | 55 | |||||
| Rebaudioside C | 6.96 | 951 | 949.5 | 787.3* | 27 | 299 |
| 479.3 | 55 | |||||
| 625.3 | 55 |
*Quantifier ion
Validation
The proposed analytical method was confirmed for its specificity, linearity, limit of detection (LOD), limit of quantification (LOQ), accuracy, and precision according to the International Conference on Harmonization (ICH) guideline Q2(R2) (ICH, 2022) and the Official Methods of Analysis of the Association of Official Analytical Chemists (AOAC) guideline (AOAC, 2016). Fruit and vegetable beverages were used as low-fat, low-protein food samples, and sauces served as high-fat, high-protein food samples for validation purposes. Specificity for detecting steviol glycosides was confirmed by comparing blank, STD 5mix, and sample chromatograms. Additionally, UHPLC–MS/MS was used to verify the specificity of steviol glycoside detection in food samples.
Linearity was validated by preparing five concentration levels (2‒50 mg/L for Stv, Reb C, F, and A and 5‒100 mg/L for Reb D), which were analyzed in seven replicates to establish calibration curves with a coefficient of determination (r2) value. The calculation of LOD and LOQ was performed by determining the ratio of the standard deviation of the y-intercept to the mean slope of the calibration curve, using data from seven replicate analyses at the three lowest concentrations. To obtain the LOD and LOQ, this ratio was then multiplied by 3.3 and 10, respectively.
To validate accuracy and precision, the STD 5mix at 5, 20, and 50 mg/kg was spiked into samples, and the recovery and percentage relative standard deviation (%RSD) for each steviol glycoside were evaluated. Precision of intra-day was evaluated using six replicates measured within the same day, and precision of inter-day was assessed by performing three replicate measurements across three consecutive days.
Measurement uncertainty
Estimation of measurement uncertainty was conducted in accordance with ISO/IEC 17025:2017 recommendations to enhance the reliability of analytical results (ISO, 2017). It was based on the EURACHEM guidelines (EURACHEM, 2022) and the Guide to the Expression of Uncertainty in Measurement (GUM) (ISO, 2008). Factors contributing to uncertainty in steviol glycoside analysis included the preparation of standard stock solutions (uSSS), sample preparation (uSP), calibration curve (uCal), and repeatability in determining steviol glycosides in samples (uRP). These factors were assessed using a coverage factor (k) of 2, which represents a 95% confidence level. The concentrations used to estimate relative expanded uncertainty were 5, 20, and 50 mg/kg as low, medium, and high concentrations, respectively.
Application
The samples for monitoring were collected in 2022 based on processed foods that were frequently consumed, as identified in the 2019 Korea National Health and Nutrition Examination Survey (KNHANES, 2019). The analytical method proposed in this paper was applied to analyze 94 processed food products available in South Korea, including pickled food products, sauces, fruit/vegetable beverages, other beverages, dried fish/shellfish fillet products, soju, and teas, to evaluate its applicability. All samples were acquired from both offline and online retailers in Seoul, South Korea. Steviol glycoside concentrations in all samples were analyzed in triplicate. To determine the content of steviol equivalents in food, the conversion factors were calculated based on the molecular weights of each steviol glycoside and steviol (Steviol, 318.4; Stv, 804.9; Reb C, 951.0; Reb F, 937.0; Reb A, 967.0; Reb D, 1129.2). The conversion factors for each steviol glycoside are as follows: Stv, 0.40; Reb C, 0.33; Reb F, 0.34; Reb A, 0.33; and Reb D, 0.28.
Results and discussion
Validation
Figure 1 presents the chromatograms used to validate the specificity of the proposed analytical method for the determination of steviol glycosides in various processed foods by HPLC-VWD. The chromatograms of the blank, STD 5mix 50 mg/L, and beverage and sauce samples indicated that steviol glycosides were detected without interference peaks. The RT for the five steviol glycosides analyzed by HPLC-VWD were as follows: Stv, 6.92 min; Reb C, 7.70 min; Reb F, 8.29 min; Reb A, 9.97 min; and Reb D, 19.55 min. Specificity was also confirmed using UHPLC–MS/MS by analyzing the chromatograms of the blank, STD 5mix 0.2 mg/L, and beverage and sauce samples. The UHPLC–MS/MS chromatograms are also shown in Fig. 1. Under these conditions, the RTs of the five steviol glycosides were as follows: Reb D, 5.87 min; Reb A, 6.63 min; Stv, 6.66 min; Reb F, 6.87 min; and Reb C, 6.96 min.
Fig. 1.
HPLC-VWD chromatograms of the (A) blank, (B) sndard solution (5mix, 50 mg/L), (C) beverages, (D) sauces sample, and (E) total ion chromatogram (TIC) from UHPLC-ESI–MS/MS in MRM mode of the 0.2 mg/L standard solution; 1: Stevioside, 2: Rebaudioside C, 3: Rebaudioside F, 4: Rebaudioside A, 5: Rebaudioside D
The linearity, LOD, and LOQ for the proposed analytical method are shown in Table 2. The calibration curves had r2 > 0.999, meeting the FDA (2014) criterion for linearity (r2 ≥ 0.995). The LOD was in the range of 0.2–0.5 mg/L, and the LOQ was in the range of 0.7–1.5 mg/L.
Table 2.
Calibration parameter results of stevioside, rebaudioside C, rebaudioside F, rebaudioside A, and rebaudioside D
| Parameters | Stevioside | Rebaudioside C | Rebaudioside F | Rebaudioside A | Rebaudioside D |
|---|---|---|---|---|---|
| Range (mg/L) | 2.0–50.0 | 2.0–50.0 | 2.0–50.0 | 2.0–50.0 | 5.0–100.0 |
| Coefficient of determination (r2) | 0.9999 | 0.9999 | 0.9999 | 0.9999 | 0.9999 |
| Slope ± SD | 0.0809 ± 0.0003 | 0.0791 ± 0.0005 | 0.0492 ± 0.0001 | 0.0675 ± 0.0005 | 0.0393 ± 0.0001 |
| Intercept ± SD | 0.0066 ± 0.0041 | 0.0079 ± 0.0028 | −0.0001 ± 0.0027 | −0.0010 ± 0.0026 | 0.0050 ± 0.0038 |
| LOD (mg/La) | 0.4 | 0.3 | 0.3 | 0.2 | 0.5 |
| LOQ (mg/Lb) | 1.1 | 0.9 | 0.8 | 0.7 | 1.5 |
aLimit of detection
bLimit of quantification
To assess the accuracy and precision of the method for five steviol glycosides in different matrices, fruit and vegetable beverages and sauces were used as representative matrices for low-fat, low-protein food samples and high-fat, high-protein food samples, respectively. The accuracy and precision of the intra-day (n = 6) and inter-day (n = 9) analyses for both matrices are shown in Table 3. Accuracy was expressed as the recovery (%) of spiked concentration, precision was represented by the %RSD. The accuracy in fruit and vegetable beverages ranged from 93.05 to 103.37%, with precision from 1.05 to 2.99%. For sauces, the accuracy ranged from 93.07 to 103.40%, and precision ranged from 0.29 to 4.95%. All results conformed to the AOAC (2016) standards (accuracy: 90‒107%; precision below 5.3%).
Table 3.
Validation results of accuracy, precision, and relative expanded uncertainty for the determination of stevioside, rebaudioside C, rebaudioside F, rebaudioside A, and rebaudioside D in beverages and sauces
| Compound | Added standard (mg/kg) | Intra-daya | Inter-dayb | Relative expanded uncertainty (%) | ||
|---|---|---|---|---|---|---|
| Accuracy (%)c | Precision (%RSD) | Accuracy (%) | Precision (%RSD) | |||
| Beverages | ||||||
| Stevioside | 5 | 101.63 ± 1.96 | 1.93 | 101.78 ± 2.74 | 2.69 | 3.55 |
| 20 | 99.26 ± 1.32 | 1.33 | 98.52 ± 1.06 | 1.07 | 2.23 | |
| 50 | 98.33 ± 2.37 | 2.41 | 99.22 ± 2.12 | 2.14 | 1.90 | |
| Rebaudioside C | 5 | 97.92 ± 2.93 | 2.99 | 97.43 ± 2.48 | 2.54 | 4.09 |
| 20 | 94.30 ± 1.47 | 1.56 | 93.57 ± 1.10 | 1.18 | 2.33 | |
| 50 | 93.43 ± 1.13 | 1.21 | 93.05 ± 1.34 | 1.44 | 2.12 | |
| Rebaudioside F | 5 | 94.38 ± 1.85 | 1.96 | 94.26 ± 1.98 | 2.11 | 3.78 |
| 20 | 93.95 ± 1.30 | 1.38 | 94.50 ± 1.36 | 1.43 | 2.26 | |
| 50 | 102.72 ± 1.84 | 1.79 | 103.37 ± 2.05 | 1.98 | 2.03 | |
| Rebaudioside A | 5 | 95.31 ± 1.00 | 1.05 | 94.10 ± 1.53 | 1.63 | 2.37 |
| 20 | 95.38 ± 1.76 | 1.84 | 94.57 ± 1.42 | 1.50 | 2.16 | |
| 50 | 99.15 ± 1.75 | 1.76 | 98.46 ± 2.32 | 2.35 | 2.04 | |
| Rebaudioside D | 5 | 96.90 ± 1.85 | 1.91 | 97.71 ± 2.53 | 2.59 | 4.48 |
| 20 | 98.30 ± 1.48 | 1.51 | 96.36 ± 1.90 | 1.97 | 2.41 | |
| 50 | 100.96 ± 1.69 | 1.68 | 100.18 ± 1.57 | 1.57 | 2.33 | |
| Sauces | ||||||
| Stevioside | 5 | 96.26 ± 1.22 | 1.27 | 99.16 ± 1.61 | 1.62 | 3.39 |
| 20 | 98.36 ± 3.26 | 3.32 | 97.89 ± 2.42 | 2.47 | 2.79 | |
| 50 | 95.63 ± 1.35 | 1.41 | 96.09 ± 1.26 | 1.32 | 2.20 | |
| Rebaudioside C | 5 | 94.39 ± 2.28 | 2.42 | 97.68 ± 2.72 | 2.78 | 3.88 |
| 20 | 95.21 ± 4.28 | 4.50 | 95.00 ± 3.33 | 3.51 | 2.56 | |
| 50 | 92.98 ± 0.95 | 1.02 | 93.07 ± 0.86 | 0.93 | 2.05 | |
| Rebaudioside F | 5 | 101.89 ± 1.41 | 1.38 | 100.12 ± 4.11 | 4.10 | 3.39 |
| 20 | 98.84 ± 4.90 | 4.95 | 98.53 ± 3.79 | 3.85 | 2.55 | |
| 50 | 103.39 ± 1.45 | 1.40 | 103.40 ± 1.26 | 1.22 | 2.20 | |
| Rebaudioside A | 5 | 94.03 ± 3.57 | 3.80 | 95.58 ± 2.84 | 2.98 | 3.84 |
| 20 | 94.30 ± 1.66 | 1.76 | 95.23 ± 2.33 | 2.45 | 2.36 | |
| 50 | 97.18 ± 1.73 | 1.78 | 97.35 ± 1.55 | 1.59 | 1.95 | |
| Rebaudioside D | 5 | 99.70 ± 1.29 | 1.30 | 101.87 ± 2.78 | 2.73 | 3.93 |
| 20 | 97.80 ± 3.42 | 3.49 | 100.47 ± 3.58 | 3.56 | 2.46 | |
| 50 | 102.99 ± 0.30 | 0.29 | 102.52 ± 1.65 | 1.61 | 1.90 | |
aAnalysis was conducted six times/day
bAnalysis was conducted three times on 3 days
cAverage ± SD
Jaworska et al. (2012) detected steviol and seven steviol glycosides by HPLC–UV, reporting LODs of 0.0020–0.015 mg/mL. Lorenzo et al. (2014) analyzed two steviol glycosides using HPLC–DAD, with an LOD of 1.07 mg/L. Liu et al. (2020) analyzed 13 steviol glycosides using HPLC–UV, reporting method detection limits (MDLs) of 0.00130–0.00265 mg/mL. The lower LOD of this study compared to the aforementioned previous studies demonstrates the sensitivity of the proposed HPLC-VWD method. This result indicates that it is capable of detecting steviol glycosides at lower concentrations compared to previous studies.
Measurement uncertainty
The relative expanded uncertainty of the proposed method was estimated for fruit and vegetable beverages and sauces. The estimation was conducted by taking into account several uncertainty factors associated with steviol glycoside analysis, including standard purity and weight, standard stock solution, sample weight, final volume of sample solution, calibration curve, and repeatability. The results of the estimated relative expanded uncertainty are presented in Table 3. The relative expanded uncertainty was found to be from 1.90 to 4.48%. All results met the criterion (≤ 16%) set by the Codex Alimentarius Commission (Codex, 2008). In all samples, there was a tendency for the relative expanded uncertainty to decrease as the concentration of steviol glycosides increased. The uSSS ranged from 1.07% to 1.40%, uSP from 0.03% to 0.04%, uCal from 0.10% to 2.53%, and uSP from 0.18% to 1.93% (data not shown).
Park et al. (2022) analyzed synthetic antioxidants in processed foods using HPLC–UV, reporting relative expanded uncertainty values of 1.81‒14.83% when spiking 5‒50 mg/kg of synthetic antioxidants. Similarly, Lee et al. (2023) detected bixin and norbixin in food using HPLC–DAD, with relative expanded uncertainty ranging from 2.5 to 10.7% when spiking 2‒10 mg/kg of bixin and norbixin. By reviewing previous studies that used similar analytical instruments and reported the relative expanded uncertainty, it was confirmed that the relative expanded uncertainty is relatively low in this study. This indicates that the analytical results obtained in this study are highly reliable.
Application
Steviol glycosides were analyzed in 94 processed food products available in the South Korean market to evaluate the applicability of the proposed method. These products comprised seven different types of food that contain these compounds. Three replicates of each sample were analyzed. The detected steviol glycoside content in the processed foods is presented in Table 4, which also includes the content expressed as steviol, considering the conversion factors for each steviol glycoside. For samples with concentrations below the LOQ, the results were reported as N.D. (not detected). Additionally, N.D. results were reconfirmed as N.D. results through UHPLC–MS/MS analysis.
Table 4.
Concentration (mg/kg) of stevioside, rebaudioside C, rebaudioside F, rebaudioside A, and rebaudioside D in processed foods
| Compound | Food category | Total no. of sample | Positive no. of sample | Range | Total Averagea (mg/kg) | Positive averageb (mg/kg) | |
|---|---|---|---|---|---|---|---|
| Min (mg/kg) | Max (mg/kg) | ||||||
| Stevioside | Dried fish/shellfish fillet products | 10 | 10 | 5.15 ± 0.23 | 25.39 ± 0.20 | 11.10 ± 5.34 | 11.10 ± 5.34 |
| Fruit/vegetable beverages | 26 | 19 | N.D.c | 148.75 ± 2.06 | 15.15 ± 28.23 | 19.93 ± 31.36 | |
| Other beverages | 15 | 8 | N.D | 24.46 ± 1.05 | 6.38 ± 8.09 | 11.96 ± 7.47 | |
| Pickled food products | 10 | 8 | N.D | 677.05 ± 4.96 | 99.55 ± 196.84 | 124.44 ± 212.93 | |
| Sauces | 12 | 8 | N.D | 76.66 ± 3.37 | 18.05 ± 23.27 | 27.08 ± 23.83 | |
| Soju | 10 | 9 | N.D | 10.25 ± 0.13 | 5.96 ± 2.61 | 6.63 ± 1.79 | |
| Teas | 11 | 3 | N.D | 7.22 ± 0.24 | 1.42 ± 2.74 | 5.20 ± 2.81 | |
| Rebaudioside C | Dried fish/shellfish fillet products | 10 | 10 | 1.35 ± 0.05 | 31.00 ± 0.47 | 7.89 ± 8.89 | 7.89 ± 8.89 |
| Fruit/vegetable beverages | 26 | 11 | N.D | 498.04 ± 0.89 | 21.22 ± 95.43 | 50.15 ± 141.68 | |
| Other beverages | 15 | 7 | N.D | 40.36 ± 1.09 | 3.88 ± 9.94 | 8.32 ± 13.22 | |
| Pickled food products | 10 | 7 | N.D | 153.46 ± 2.33 | 23.43 ± 44.70 | 33.48 ± 50.18 | |
| Sauces | 12 | 7 | N.D | 35.72 ± 0.55 | 9.66 ± 10.55 | 16.56 ± 8.76 | |
| Soju | 10 | 5 | N.D | 5.14 ± 0.01 | 1.75 ± 2.23 | 3.49 ± 1.95 | |
| Teas | 11 | 5 | N.D | 42.10 ± 0.34 | 4.46 ± 11.94 | 9.81 ± 16.15 | |
| Rebaudioside F | Dried fish/shellfish fillet products | 10 | 1 | N.D | 1.28 ± 0.04 | 0.13 ± 0.39 | 1.28 ± 0.00 |
| Fruit/vegetable beverages | 26 | 3 | N.D | 8.42 ± 0.15 | 0.54 ± 1.75 | 4.69 ± 2.64 | |
| Other beverages | 15 | 2 | N.D | 1.47 ± 0.03 | 0.16 ± 0.42 | 1.21 ± 0.26 | |
| Pickled food products | 10 | 4 | N.D | 24.45 ± 0.18 | 3.87 ± 7.21 | 9.67 ± 8.60 | |
| Sauces | 12 | 5 | N.D | 12.03 ± 0.07 | 2.50 ± 3.90 | 5.99 ± 3.94 | |
| Soju | 10 | 0 | N.D | N.D | N.D | N.D | |
| Teas | 11 | 3 | N.D | 31.98 ± 0.39 | 3.81 ± 9.16 | 13.98 ± 12.86 | |
| Rebaudioside A | Dried fish/shellfish fillet products | 10 | 10 | 2.26 ± 0.08 | 12.45 ± 0.30 | 5.97 ± 3.05 | 5.97 ± 3.05 |
| Fruit/vegetable beverages | 26 | 25 | N.D | 137.34 ± 4.90 | 49.15 ± 34.52 | 51.11 ± 33.75 | |
| Other beverages | 15 | 15 | 9.82 ± 1.80 | 126.82 ± 2.57 | 48.45 ± 37.06 | 48.45 ± 37.06 | |
| Pickled food products | 10 | 8 | N.D | 799.90 ± 4.68 | 122.97 ± 236.54 | 153.71 ± 255.37 | |
| Sauces | 12 | 12 | 3.25 ± 0.10 | 177.95 ± 2.91 | 64.85 ± 58.40 | 64.85 ± 58.40 | |
| Soju | 10 | 10 | 14.22 ± 0.14 | 35.44 ± 0.05 | 25.97 ± 7.82 | 25.97 ± 7.82 | |
| Teas | 11 | 11 | 1.28 ± 0.29 | 97.41 ± 0.72 | 21.62 ± 30.52 | 21.62 ± 30.52 | |
| Rebaudioside D | Dried fish/shellfish fillet products | 10 | 8 | N.D | 42.13 ± 0.73 | 15.96 ± 12.19 | 19.95 ± 10.31 |
| Fruit/vegetable beverages | 26 | 0 | N.D | N.D | N.D | N.D | |
| Other beverages | 15 | 0 | N.D | N.D | N.D | N.D | |
| Pickled food products | 10 | 1 | N.D | 4.85 ± 0.11 | 0.48 ± 1.45 | 4.85 ± 0.00 | |
| Sauces | 12 | 2 | N.D | 11.40 ± 0.27 | 1.54 ± 3.56 | 9.24 ± 2.15 | |
| Soju | 10 | 0 | N.D | N.D | N.D | N.D | |
| Teas | 11 | 0 | N.D | N.D | N.D | N.D | |
| Totald (as steviol) | Dried fish/shellfish fillet products | 10 | 10 | 6.17 ± 0.14 | 23.61 ± 0.49 | 13.77 ± 6.30 | 13.77 ± 6.30 |
| Fruit/vegetable beverages | 26 | 26 | 0.69 ± 0.03 | 202.01 ± 0.39 | 29.44 ± 39.91 | 29.44 ± 39.91 | |
| Other beverages | 15 | 15 | 3.24 ± 0.05 | 45.68 ± 0.83 | 19.92 ± 14.00 | 19.92 ± 14.00 | |
| Pickled food products | 10 | 9 | N.D | 595.28 ± 2.65 | 89.82 ± 173.84 | 99.80 ± 180.51 | |
| Sauces | 12 | 12 | 2.08 ± 0.04 | 64.59 ± 0.90 | 33.19 ± 19.90 | 33.19 ± 19.90 | |
| Soju | 10 | 10 | 8.64 ± 0.20 | 15.62 ± 0.04 | 11.55 ± 2.85 | 11.55 ± 2.85 | |
| Teas | 11 | 11 | 0.81 ± 0.01 | 35.01 ± 0.30 | 10.51 ± 12.61 | 10.51 ± 12.61 | |
aAverage and standard deviation of all samples
bAverage and standard deviation of detected samples
cNot detected, < LOQ
dThe total amount of steviol glycosides was calculated as steviol by multiplying the amount of each steviol glycoside by a conversion factor
*All experiments were carried out in three replicates
Among the 94 processed food products, the detection rates for five steviol glycosides were as follows: Stv, 69.1%; Reb C, 55.3%; Reb F, 19.1%; Reb A, 96.8%; and Reb D, 11.7%. Furthermore, the frequency distribution histogram of steviol glycosides, expressed as steviol concentrations in foods, is depicted in Fig. 2.
Fig. 2.
Frequency distribution histogram of steviol glycosides as steviol concentrations in foods
Kang et al. (2022) analyzed 14 types of processed foods containing Stv and Reb A, and reported that the sum of Stv and Reb A concentrations was detected in 190 out of 197 samples (detection rate: 96%), with levels ranging from not detected (in sauces) to 137.3 mg/kg (in fermented Korean red ginseng beverages). Phungsiangdee et al. (2023) quantified nine steviol glycosides and their derivatives (Stv, Reb A, B, C, D, F, rubusoside, dulcoside A, and steviolbioside) in three types of processed foods. Among them, Reb A, Stv, and Reb B were the most frequently detected. In the three types of processed foods, Reb A was detected at levels ranging from 0.96 to 886.16 mg/kg, Stv from 1.02 to 215.59 mg/kg, and Reb B from 0.42 to 13.48 mg/kg.
Given that an ADI for steviol has been established, the results regarding steviol glycoside content in foods can be useful for evaluating steviol intake and risk assessment. According to the application results, the proposed HPLC-VWD analytical method is projected to provide an accurate quantification of steviol glycosides in a wide variety of processed food products.
Funding
This research was supported by a grant from Ministry of Food and Drug Safety in 2022, Republic of Korea.
Declarations
Conflict of interest
The authors declare no conflict of interest.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- AOAC. Official methods of analysis of AOAC International. AOAC International, Rockville, MD, USA (2016)
- Bililign T, Moore JC, Tan S, Leeks AT. Development and validation of a reversed-phase high-performance liquid chromatography method for routine identification and purity assessment of high-purity steviol glycoside sweeteners. Journal of Agricultural and Food Chemistry. 62: 1384-1389 (2014) [DOI] [PubMed] [Google Scholar]
- Cacciola F, Delmonte P, Jaworska K, Dugo P, Mondello L, Rader JI. Employing ultra high pressure liquid chromatography as the second dimension in a comprehensive two-dimensional system for analysis of Stevia rebaudiana extracts. Journal of Chromatography A. 1218: 2012-2018 (2011) [DOI] [PubMed] [Google Scholar]
- Codex Alimentarius Commission. Guidance on measurement uncertainty and uncertainty of sampling guidance on measurement uncertainty, CX/MAS 08/29/9. CODEX, Budapest, Hungary (2008)
- EURACHEM. Quantifying uncertainty in analytical measurement Guide. English Edition. EURACHEM, London, UK (2022)
- Food and Drug Administration (FDA). GRAS NOTICE FOR STEVIOL GLYCOSIDES. FAO. College Park, MD, USA (2018)
- González C, Tapia M, Pérez E, Pallet D, Dornier M. Main properties of steviol glycosides and their potential in the food industry: a review. Fruits. 69: 127-141 (2014) [Google Scholar]
- Hearn LK, Subedi PP. Determining levels of steviol glycosides in the leaves of Stevia rebaudiana by near infrared reflectance spectroscopy. Journal of Food Composition and Analysis. 22: 165-168 (2009)
- Huang C, Wang Y, Zhou C, Fan X, Sun Q, Han J, Hua C, Li Y, Niu Y, Okonkwo CE, Yao D, Song L, Otu, P. Properties, extraction and purification technologies of Stevia rebaudiana steviol glycosides: A review. Food Chemistry. 453:139622 (2024) [DOI] [PubMed] [Google Scholar]
- International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). ICH Harmonised Tripartite Guideline. Validation of analytical procedures: text and methodology Q2 (R2). ICH, San Diego, CA, USA (2022)
- International Organization for Standardization (ISO). Evaluation of measurement data-Guide to the expression of uncertainty in measurement. Joint Committee for Guides in Metrology (JCGM). ISO, Geneva, Switzerland. (2008)
- International Organization for Standardization (ISO). General requirements for the competence of testing and calibration laboratories. Third Edition. ISO, Geneva, Switzerland (2017)
- Jaworska K, Krynitsky AJ, Rader JI. Simultaneous analysis of steviol and steviol glycosides by liquid chromatography with ultraviolet detection on a mixed-mode column: application to Stevia plant material and Stevia-containing dietary supplements. Journal of AOAC International. 95: 1588-1596 (2012) [DOI] [PubMed] [Google Scholar]
- Kang HH, Yun CI, Choi S, Oh KS, Kim YJ. Occurrence and risk characterization of non-nutritive sweeteners in selected food products from Korea. Food Science and Biotechnology. 31: 37-48 (2022) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Korea Disease Control and Prevention Agency. Korean Health Statistics 2019: Korea National Health and Nutrition Examination Survey (KNHANES), Korea Disease Control and Prevention Agency, Cheongju, Republic of Korea (2021) [Google Scholar]
- Lee GY, Yun CI, Cho J, Kim YJ. Validation, measurement uncertainty, and determination of bixin and norbixin in processed foods of animal resources distributed in Korea. Food Science of Animal Resources. 43: 949-960 (2023) [DOI] [PMC free article] [PubMed]
- Liu Z, Ren K, Feng Y, Uong T, Krepich S, You H. Rapid and economic determination of 13 steviol glycosides in market-available food, dietary supplements, and ingredients: single-laboratory validation of an HPLC method. Journal of Agricultural and Food Chemistry. 68: 10142-10148 (2020) [DOI] [PubMed] [Google Scholar]
- Londhe SV, Nanaware SM. HPTLC method for simultaneous analysis of stevioside and rebaudioside-A in Stevia rebaudiana. Journal of AOAC International. 96: 24-26 (2013) [DOI] [PubMed] [Google Scholar]
- Lorenzo C, Serrano-Díaz J, Plaza M, Quintanilla C, Alonso GL. Fast methodology of analysing major steviol glycosides from Stevia rebaudiana leaves. Food chemistry. 157: 518-523 (2014) [DOI] [PubMed] [Google Scholar]
- Molina-Calle M, De Medina VS, De La Torre MD, Priego-Capote F, De Castro ML. Development and application of a quantitative method based on LC–QqQ MS/MS for determination of steviol glycosides in Stevia leaves. Talanta. 154: 263-269 (2016) [DOI] [PubMed] [Google Scholar]
- Park HJ, Seo E, Park JW, Yun CI, Kim YJ. Determination and validation of synthetic antioxidants in processed foods distributed in Korea. Journal of Food Hygiene and Safety. 37: 297-305 (2022)
- Phungsiangdee Y, Chaothong P, Karnpanit W, Tanaviyutpakdee P. Validation of UHPLC-ESI-MS/MS method for determining steviol glycoside and its derivatives in foods and beverages. Foods. 12: 3941 (2023) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Purkayastha S, Kwok D. Metabolic fate in adult and pediatric population of steviol glycosides produced from stevia leaf extract by different production technologies. Regulatory Toxicology and Pharmacology. 116: 104727 (2020) [DOI] [PubMed] [Google Scholar]
- Saifi M, Ali A, Saini M, Nasrullah N, Khan S, Abdin M. A rapid and efficient high performance thin layer chromatographic (hptlc) method for simultaneous analysis of stevioside and rebaudioside-a in stevia rebaudiana. International Journal of Pharmacy and Pharmaceutical Science. 6: 455-64 (2014) [Google Scholar]
- US Food and Drug Administration (FDA). Methods, method verification and validation. US FDA, White Oak, MD, USA (2014)