1. Introduction
The term ‘Aflatoxins’ (AFT) typically refers to the sum of variants AFB1, AFB2, AFG1, AFG2, while ‘Fumonisins’ (FBs) refers to the sum of variants FB1 and FB2 [1]. These groups are mycotoxins produced by molds of genera Aspergillus spp. and Fusarium spp., which commonly colonize corn [2–4]. The contamination of corn by AFT or FBs (especially the main variant FB1) has health implications for consumers of corn-derived products. The AFT are reportedly carcinogenic (classified as Group 1 [4–6]), hepatotoxic, and genotoxic to humans [2,7]. Moreover, the carcinogenicity of FBs has been reported in animals, but not in humans [8]. Thus, FBs are classified as possible carcinogenic agents (Group 2B) for humans [4,5].
Having established the susceptibility of corn to contamination by AFT and FBs, recent studies in the Mesoamerican region provided data on the occurrence and content of one or both mycotoxins in whole corn grains.
The reported occurrence and mean content of AFT in corn from Guatemala [9], Honduras [10], El Salvador [11], and Costa Rica [12] ranged from 6 % to 59 % and 1.7–215 µg/kg, respectively, during the dry season (December–May). During the rainy season (June–November), the occurrence of 36–81 % and mean content of 6.2–63 µg/kg were found in corn from Guatemala [13] and Honduras [14], indicating seasonal variation in both the occurrence and content of AFT in corn.
The occurrence and content of FBs or FB1 measured during the rainy season respectively ranged from 80 % to 98 % and from 2.6 % to 7.2 mg/ kg in corn from Guatemala [13] and Honduras [14]. During the dry season, the occurrence and content ranged from 77 % to 97 % and from 0.66 to 3.2 mg/kg in corn from Guatemala [9] and Honduras [10], indicating a similar seasonal variation in the occurrence and content of FBs to those of AFT.
In Mexico [2,15] and Honduras [14], the occurrence and mean content of AFT and AFB1 in dough and tortillas, which are derived from nixtamalized corn (i.e., corn subjected to alkaline cooking using lime), ranged from 13 % to 95 % and 1.1–9.5 µg/kg during the rainy season, respectively. In the dry season, the occurrence was 5–89 % and the content was 0.5–9.4 µg/kg in products from Guatemala and Mexico [2,9, 15–17].
The occurrence and content of FBs or FB1 in nixtamalized corn derivatives from Mexico [2] and Honduras [14] were 36–92 % and 0.14–4.1 mg/kg, respectively, during the rainy season. In the dry season, the occurrence and mean content of FBs or FB1 were reported for products from Guatemala [9] and Mexico [18], ranged from 32 % to 98 % and 0.04–0.71 mg/kg, respectively. The abovementioned data show a wide variation in the values, making it difficult to establish any trend for the seasonal change in the occurrence and content of AFT and FBs.
Estimating the level of exposure of Salvadoran consumers to AFT and FBs by calculating the estimated daily intake (EDI) [19], on the basis of the intake and content of these mycotoxins found in the derivatives of nixtamalized corn, is necessary for establishing a preliminary indicator of exposure. However, no data are available for estimating the daily intake of AFT and FBs for consumers in El Salvador. The limited available data refer to other countries in Mesoamerica.
In Mexico, the reported EDI of AFT was 1.9–3.6 ng/kg bw/day during the rainy season and 0.9–1.7 ng/kg bw/day in the dry season [2, 17]. In Honduras, the reported EDI was 8.7–15.4 ng/kg bw/day during the rainy season [14]. These values exceed the standard of 0.0 ng/kg bw/day for AFT ingested through food [7,20], in accordance with the “as low as reasonably achievable” (ALARA) principle [7].
The EDI for FBs or FB1 in nixtamalized corn derivatives from Mexico was 0.1–0.24 µg/kg bw/day in the rainy season and 0.17–2.0 µg/kg bw/ day in the dry season [2,18], reaching up to 4.8 µg/kg bw/day regardless of the year [21]. In the case of Honduras, the EDI ranges of 21.7–33.6 µg/kg bw/day for FBs were reported in the rainy season [14]. Some of these mean EDI values exceed the tolerable daily intake (TDI) limit of FBs, set at 1.0 µg/kg bw/day [8], particularly in Honduras, a country bordering El Salvador.
Therefore, the ingestion of nixtamalized corn derivatives contaminated with AFT and FBs presents a health risk to the consumers in these Mesoamerican countries, including El Salvador, considering that nixtamalized corn dough serves as a base material for the preparation of tortillas, a fundamental component of the Salvadoran diet, consumed by 90 % of the population, with a mean intake per capita of 252 g/day [22].
Considering the lack of published results concerning this topic in El Salvador, the objectives of this study are to monitor the occurrence and content of AFT and FBs in nixtamalized corn dough for one year to determine their variation during the rainy and dry seasons and to estimate the daily intake of both mycotoxins in this corn derivative as an indicator of exposure of the Salvadoran population to these harmful compounds.
2. Material and methods
2.1. Type of study, specimen, and sampling
This is a descriptive study with a longitudinal design. It was used to survey a convenience sample of corn dough processed locally by alkaline cooking using Ca(OH)2 (commonly known as nixtamalization), obtained from 36 “tortillerías” (tortilla shops) located in five municipal markets and nearby sales stands for 18 months. Nixtamalized corn dough or base dough was selected as the analysis matrix in this study because it is the main ingredient for making tortillas, pupusas, chilaquiles, enchiladas, tamales, bunuelos, ˜ pinol, empanadillas and other derivatives widely consumed in El Salvador, similar to corn derivatives made in Guatemala [9]; therefore, monitoring AFT and FBs contents, as well as the exposure of the population to these mycotoxins through the intake of derivatives of this base dough, was considered important.
The sampling strategy was adapted from that used in previous studies to monitor the contents and seasonal variations of AFM1 in hard cheese and formaldehyde in fresh cheese sold in the same markets in El Salvador [23,24]. Although the number of markets is not representative of the country, these stores are located in cities with large population centers and, therefore, the results of the risk of exposure due to ingestion of both mycotoxins may be more applicable to urban populations.
The entire sampling period for the detection of AFT in corn dough was during the 2023 extended rainy season (June–November) and in the 2023–2024 dry season (December–May). A total of 386 samples were collected from four locations: Sonsonate-Western Region (n = 39), San Salvador-Central Region (n = 100), Chalatenango-Northern Region (n = 50), and San Miguel-Eastern Region (n = 197). Samples for FBs detection were collected during the 2023–2024 dry season (December–May) and in the 2024 rainy season (June–October). A total of 288 samples were collected from Sonsonate (n = 32), San Salvador (n = 72), Chalatenango (n = 40), and San Miguel (n = 144).
The sample size was initially specified at 432 for the AFT (36/month for 12 months) and 396 for the FBs (36/month for 11 months). Consecutive interruptions in the supply of ELISA kits (due to limitations inherent to the producing company) resulted in 386 samples being collected for the AFT (11 out of 12 months), affecting the total collected during the dry season; even so, the sampling had 83.3 % statistical power to detect differences in means. The impact included the FBs, so only 288 samples were collected over 8 months, losing 1 month of the dry season and 2 of the rainy seasons. Due to this irregularity, it was not possible to compare the two seasons nor to calculate the statistical power to differentiate means.
All collected samples weighed 1 kg, were individually packaged and kept cold during transport to the laboratory, and were stored in a refrigerator at 2–4 ◦C until processing and analysis. Twenty-four hours after collection, the analytical procedures were performed with a 10- sample batch from one location per day to prevent crosscontamination, and making sure of following all the quality-control rules, inherent of a research laboratory.
2.2. Sample preparation, extraction, and analysis of AFT and FBs
Each 1 kg sample of nixtamalized corn dough was spread on glassine paper and homogenized by mixing it thoroughly. A 5 g portion was placed in a tube with a cap, after which 25 ml of 70 % v/v aqueous methanol solution was added and the mixture was homogenized by shaking for 3 min using a vortex. The resulting suspension was filtered using Whatman No. 1 paper and an aliquot of the filtrate was collected in a 2000 µL Eppendorf vial. The extract thus obtained was used for AFT analysis, whereas for FBs analysis, the sample was further diluted in a 1:80 ratio prior to the assay. The quantification of both mycotoxins was performed using a direct competitive ELISA method, but there was not additional confirmatory analysis by LC/MS as a limitation of this study.
The AFT content was measured using the highly sensitive Veratox® kit with a quantification range of 1–8 µg/kg, for which the procedure specified by the manufacturer [25] was adopted for nixtamalized corn dough. To determine the FBs levels, the Veratox® kit with a quantification range of 1–6 mg/kg was used, according to the manufacturer’s specifications [26]. The quantification of AFT by the ELISA method has a limit of detection of 0.5 µg/kg and a quantification limit of 1.0 µg/kg [25]. In contrast, ELISA FBs quantification has a detection limit of 0.2 mg/kg and a limit of quantification of 1.0 mg/kg [26].
The quantified AFT content is expressed as µg/kg and the FBs content as mg/kg on the basis of the absorbance determined using a Synergy H1M2 Agilent BioTek multimodal microplate reader (Winooski, Vermont, USA) with a monochromator adjusted to 650 nm. The samples that exceeded the upper limit of quantification for both analyses were diluted in a 1:2 or 1:4 ratio, as required, and re-assayed in duplicate following the previously indicated procedures [25,26].
2.3. Validation of analytical method for AFT and FBs
The method for analyzing AFT and FBs was verified on the basis of two criteria: mean recovery and repeatability [27]. The analytical method was evaluated by spiking samples of nixtamalized corn dough with three excessive known concentrations of AFT (1.6, 3.2, and 6.4 µg/kg) or FBs (1.6, 3.2, and 4.8 mg/kg). In the case of AFT, four replicate tests were performed for the samples spiked with the excess analyte for five consecutive days (Table 1). For FBs, the test was performed in triplicate for each concentration for three consecutive days (Table 2). The extraction and quantification of both mycotoxins were performed using the same method as that described for the samples and using the same equipment, albeit with a different reagent kit for each day the test was performed. Specifications for detection and quantification limits of AFT and FBs ELISA kits were indicated above.
The recovery was calculated by dividing the measured content of each sample by the spiked concentration and the result was multiplied by 100. The mean recovery is the simple mean of the set of values obtained for each spiking level during each day of the test [27]. The repeatability or intra-assay precision was determined by calculating the relative standard deviation (RSD%) of the recovery means [27], and the Horwitz ratio (HORRAT) of AFT or FBs spiked samples were calculated.
As explained above, confirmatory testing of AFT or FB content using LC/MS was not performed due to equipment limitations in the food safety laboratory. However, the average percentage recoveries of the spiking mycotoxin, quantified by the ELISA method, as well as the coefficient of variation of these averages and the calculation of the Horwitz ratio (HORRAT) over 3–5 days, indicate reasonably good withinlaboratory repeatability.
2.4. Estimation of exposure to AFT and FBs via ingestion
The dietary exposure estimates, defined as the amount of a chemical ingested through food, were calculated on the basis of consumption data and the amount of the analyte in the food [28]. The dietary exposure estimate is important for comparing the exposure to a proven or potential toxic compound with a guideline or tolerable maximum intake value [28]. In the case of AFT, there is no tolerable maximum intake value; thus, the standard of 0.0 ng/kg bw/day specified in other studies was adopted [7,20]. The tolerable maximum daily intake limit of FBs was at 1.0 µg/kg bw/day [8].
Considering the abovementioned data, the EDI of both AFT and FBs was calculated using the procedure utilized in similar studies [10,17,18, 29], where the mean content of each mycotoxin was multiplied by the mean daily consumption of nixtamalized corn dough (as tortillas) and that result was divided by the mean weight of the adult Salvadoran population, without distinction of sex. The mean content of both mycotoxins in the analyzed food was determined on the basis of a technical consideration used in estimating the dietary exposure to carcinogenic and genotoxic compounds [30]. Following a similar criterion, the per capita daily consumption of nixtamalized corn dough (as tortillas) was calculated using the means available for the population of the entire country (252 g/day), as well as for the four regions in which sampling was carried out [22]. For the calculations, the simple mean weight of the adult population of El Salvador was used on the basis of available data, without distinction of sex [31].
2.5. Statistical analysis
Significant differences in the parameters used to validate the analytical method, as well as in the mean content and occurrence of AFT and FBs between the samples collected during the rainy and dry seasons, were detected by means of one-way analysis of variance (ANOVA), Student’s t-test, and Chi-square test. The homogeneity of variance was demonstrated by applying the Levene test. In all tests, a significance value of p < 0.05 was specified.
The correlation between the meteorological parameters and mean AFT content was determined using Pearson’s r coefficient. Statistical tests were performed and the corresponding diagrams were constructed using the IBM Statistics v.27 program for Windows.
Table 1 Method performance parameters for Aflatoxins (AFT) in spiked pooled samples of nixtamalized corn dough.
Table 2 Method performance parameters for Fumonisins (FBs) in spiked pooled samples of nixtamalized corn dough.
3. Results
3.1. Method validation parameters to determinate AFT and FBs in nixtamalized corn dough
The parameters for validating the method of quantifying AFT and FBs in nixtamalized corn dough, including the mean recovery (%) and coefficients of variation for repeated experiments, are summarized in Tables 1 and 2, respectively. The mean AFT recovery for the five-day assay ranged from 94.57 to 105.73 for the three spiked concentrations and did not exceed the upper limit of 120 %. The coefficient of variation (RSD%) decreased from 42.19 to 33.99 as the spiked concentration increased (Table 1). These values are below the threshold of 45.3 % for the samples containing 1 µg/kg of the analyte. The HORRAT values ranged from 0.99 to 1.01, where the values are within the range of 0.3–1.3. The mean recovery and RSD% did not vary significantly for the five days of the validation trial (F = 0.705, 4 df, p = 0.404; F = 0.206, 4 df, p = 0.934, respectively).
In the case of the FBs assay, as validation parameters, the mean recovery ranged from 100.63 to 108.44, remaining below the 120 % threshold (Table 2), whereas RSD% decreased from 42.16 to 35.60 as the spiked concentration increased (Table 2). These values exceeded the threshold of 16 for the samples containing 1 mg/kg of the analyte. The HORRAT values were 0.99–1.00, falling within the expected range of 0.3–1.3. The mean recovery and RSD% also did not vary significantly for the three days of the FBs validation assay (F = 0.840, 2 df, p = 0.443; F = 0.027, 2 df, p = 0.973, respectively).
3.2. Seasonal variation of AFT positive cases and contents
The counts and frequencies of occurrence of the nixtamalized corn dough samples testing positive for AFT during the 2023 rainy season and the 2023–2024 dry season across the four locations in El Salvador are listed in Table 3. The average accumulated rainfall, ambient temperature, and relative humidity for the same locations are also included.
Overall, the occurrence of the samples with quantifiable AFT content from the four monitored locations varied, but not significantly (χ2 = 0.745, 3 df, p = 0.868). The occurrence did not differ significantly for the two seasons of the year that were compared (χ2 = 4.172, 2 df, p = 0.118, Table 3). The proportion of all samples testing positive for AFT was slightly higher during the dry season than in the rainy season (23.9 % vs. 19.9 %, Table 3). However, the analysis of the results by location showed that the occurrence of the samples with quantifiable AFT content was significantly higher in San Salvador during the dry season than that in the rainy season (42.2 % vs. 14.6 %, χ2 =15,730, 3 df, p = 0.001), whereas the AFT content of the samples from Sonsonate, Chalatenango, and San Miguel showed no significant seasonal differences (Table 3).
The proportion of cases exceeding the maximum AFT limit (> 10 µ/ kg) was 1.8 %, 1.9 %, and 1.5 % for San Salvador, San Miguel, and all samples collected during the rainy season, respectively, although no positive case exceeded the limit in the other locations during the same season (Table 3). During the dry season, no samples exceeding the maximum AFT limit were detected in any of the four study locations.
Table 3 Aflatoxins occurrences in nixtamalized corn dough and meteorological parameters in four Salvadoran locations.
The samples collected in the four locations had higher mean AFT content during the rainy season than that in the dry season; however, the differences were not significant (Fig. 1). Comparison of the AFT content of all samples in both seasons showed that the mean was significantly higher in the rainy season than the dry season (3.64 µg/kg vs. 1.93 µg/ kg, t = 2.973, 51.199 df, p = 0.004). Nevertheless, the means by location and season and those for all samples did not exceed the maximum limit specified for AFT in corn derivatives intended for human consumption (Fig. 1).
Fig. 1. Variation in quantified aflatoxins (AFT) levels in nixtamalized corn dough based on location and season. Data is shown by year-season, and location, with means indicated inside bars and whiskers representing ± 2 SEM (n = 206 for the rainy season 2023, n = 180 for the dry season 2023–2024). Asterisks denote significant differences between overall location bars (p < 0.01, Student’s t-test).

The mean AFT content of the corn samples was analyzed to determine the association with the meteorological parameters (ambient temperature and relative humidity); the results are summarized in Table 3. No significant association between the AFT content and ambient temperature (Pearson r = 0.181, p = 0.099, n = 84) was observed; however, a positive association between both variables could be demonstrated, albeit only in the dry season (Pearson r = 0.365, p = 0.016, n = 43). The relative humidity was also shown to be significantly and positively associated with the AFT content (Pearson r = 0.270, p = 0.013, n = 84), regardless of the time of year.
3.3. Seasonal variation of FBs positive cases and contents
The counts and frequencies of occurrence of samples of nixtamalized corn dough positive for FBs collected in the four study locations for the dry season in 2023–2024 and the rainy season in 2024 are listed in Table 4. The average accumulated rainfall, ambient temperature, and relative humidity for the same locations and seasons are also included.
Of the 180 samples analyzed during the 2023–2024 dry season, only 12 were positive for FBs and exceeded the maximum limit of 1.0 mg/kg. Among these, 10 were collected in San Salvador (22.2 %) and 2 in San Miguel (2.2 %). Thus, the proportion of cases that exceeded the limit for all samples collected in the dry season was 6.7 % (Table 4). None of the 108 samples collected during the 2024 rainy season showed quantifiable levels of FBs, regardless of the location, which accounts for the lack of significant differences with seasonal variation.
Table 4 Fumonisin occurrences in nixtamalized corn dough and meteorological parameters in four Salvadoran locations
The mean FBs contents of the samples collected from San Salvador, San Miguel, and all samples during the dry season were 3.21, 1.48, and 2.92 mg/kg, respectively (Table 4). All mean FBs contents exceeded the maximum limit specified for corn intended for human consumption.
3.4. Characterization of health risk of AFT and FBs exposure through nixtamalized corn dough
The daily intake of nixtamalized corn dough as tortillas in the western region, central region, northern zone, and eastern zone was estimated at 255, 161, 268, and 308 g/day, respectively, where the mean for the four zones was 252 g/day for an adult with a mean weight of 73.7 kg, regardless of sex. Considering these data and the mean AFT content of the samples collected from the four locations during the rainy and dry seasons, the EDI was calculated as an indicator of the potential risk to human health due to exposure to the AFT mycotoxin.
The mean EDI of AFT for consumers of nixtamalized corn dough was higher during the rainy season than that in the dry season, without exception (Table 5); the highest values were obtained for the samples collected from the eastern region (San Miguel, 18.53 ng/kg bw/day) and for the group of localities (12.44 ng/kg bw/day). Considering that the tolerable limit for the ingestion of AFT is 0.0 ng/kg bw/day, due to the genotoxic properties of these compounds, the intake of nixtamalized corn dough with any AFT content represents a potential health risk, regardless of the locality and season of the year (Table 5).
Table 5 Health risk assessment of nixtamalized corn dough consumer due to Aflatoxins exposure in El Salvador.
The mean EDI of FBs per consumer of nixtamalized corn dough was calculated for San Salvador (7.02 µg/kg bw/day), San Miguel (6.19 µg/ kg bw/day), and the group of localities (9.99 µg/kg bw/day) for the samples collected only during the 2023–2024 dry season (Table 6). These values exceeded the TDI limit of FBs (1.0 µg/kg bw/day). No quantifiable levels of FBs were detected for the 2024 rainy season; therefore, the exposure values were not calculated.
Most of the dataset generated and analyzed during this study is available at https://doi.org/10.17632/rnjgwjtn9r.1 [32].
4. Discussion
The method for analyzing AFT in nixtamalized corn dough was successfully validated on the basis of the mean recovery (98.7–105.7 %) and reproducibility (RSDR) values (33.99–42.19), which were within the ranges established as acceptable (40–120 % and < 45.3, respectively) [27]. Prior analyses of tortillas, flour, and corn dough presented similar recovery values for AFT (87–93 %) [15], AFB1 (95 %, 96 %, and 100 %) [2,16,17], and AFB2 (96 % and 98 %) [2,16].
The method for FBs analysis was also validated. The mean recovery values (100.6–108.4 %) were within the established acceptable range (80–110 %), contrary to the reproducibility values (35.6–42.2), which exceeded the limit of 16 [27]. However, the mean recovery of FBs in nixtamalized corn dough was consistent with that observed in the validation of other methods of analyzing corn flour for dough (FB1 92 % and 99 %, FB2 88 %) [14,21] and nixtamalized corn dough (FB1 118 %, FB2 110 %) [18].
The analysis of the occurrence of AFT in nixtamalized corn dough for the group of localities showed that for the rainy season, only 2 % of samples had values exceeding the maximum level of 10 µg/kg specified by the Codex Alimentarius and the European Commission [33], whereas no sample collected in the dry season exceeded this maximum limit. The few samples that exceeded the maximum limit for AFT were collected in the San Salvador-Central Region and San Miguel-Eastern Region (Table 3), places where corn collection centers from different parts of the country operate to cover the local demand for human consumption.
In general, the occurrence of the AFT-containing samples within the maximum limit did not differ significantly for the rainy and dry seasons, with values of 18.4 % and 23.9 %, respectively (Table 3). The exception to this trend was detected in the San Salvador-Central Region, which had a significantly lower occurrence of the AFT-containing samples in the rainy season than in the dry season (14.6 % vs. 42 %, Table 3). Even with the abovementioned differences, the data indicate a very low occurrence of AFT contamination in nixtamalized corn dough (approximately 20 %, Table 3), regardless of the season, similar to the values reported for corn tortillas (15 %) [15] in a previous study (Table 7).
Other studies performed in Mesoamerican countries showed significant variations in the occurrence of aflatoxins associated with the dry and rainy seasons, based on analysis of AFT (72.5 % vs. 85 % and 15.5 % vs. 13.2 %) [2,15] and AFB1 in nixtamalized corn tortilla (89 % vs. 80 %) [16] (Table 7).
Table 6 Health risk assessment of nixtamalized corn dough consumer due to Fumonisins exposure in El Salvador.
Table 7 Occurrence, contents, and estimated dietary intake of aflatoxins in maize products in Mesoamerican countries.

The average AFT content of nixtamalized corn dough was higher during the rainy season than that in the dry season for the samples collected from each of the four locations; however, the differences were not significant (Fig. 1). However, the analysis of the total data indicated that the mean AFT content of nixtamalized corn dough from the samples collected during the rainy season was significantly higher than that during the dry season (3.64 ± 0.54 vs. 1.93 ± 0.20 µg/kg); though, none of the values exceeded the maximum limit of 10 µg/kg [1,33]. The higher AFT content of the samples collected during the rainy season can be ascribed to the statistically significant association between the mean levels of AFT and relative humidity (r = 0.270, p < 0.05), which were higher in the wet season than in the dry season (Table 3). Regarding the previous association, studies performed in El Salvador and Guatemala demonstrated that corn grain samples extracted from storage sites with a higher mean relative humidity were more susceptible to AFT contamination [11,34].
The comparison of the data obtained in this work with the results of previous studies performed in Mesoamerican countries indicates outstanding similarity between the mean contents of AFT and AFB1 in nixtamalized corn dough in the rainy and dry seasons (1.89 vs. 1.2 µg/ kg, respectively) [14,17] and in tortillas during the same periods (1.96, 1.09, 1.06, vs. 0.51 and 0.71 µg/kg, respectively) [2,9,14] (Table 7).
Notably, the mean contents of AFT and AFB1 in nixtamalized corn dough or tortillas were lower than the contents analyzed in the whole corn grain collected from El Salvador (9.4–24.6 µg/kg) [11], Honduras (6.23, 6.5–7.2 µg/kg) [10,14], Guatemala (1.7 µg/kg, 63 µg/kg and 2–85 µg/kg) [9,13,34], and Costa Rica (28–62 µg/kg) [12].
The plausible reason for the low AFT content of corn dough and tortillas reported in this and previous studies is the process of corn nixtamalization, which involves the alkaline cooking of corn kernels with water and lime [3]. In fact, the effective decrease in the mean AFT content due to nixtamalization and baking was evidenced by the results of two recent studies, which reported values ranging from 1.73 µg/kg and 6.23 µg/kg in whole grain to 1.89 µg/kg in corn dough and 0.71 µg/kg and 1.06 mg/kg in tortillas [9,14]. It is known that nixtamalization can reduce the AFT content of corn dough by 60–65 % and of tortillas by 70–80 % [3,35]. Other studies reported reductions of 58–82 % for the AFT or AFB1 content of nixtamalized corn dough [9,14, 36].
The occurrence of FBs in the total samples of nixtamalized corn dough collected from the four locations during the dry season was very low (< 7 %, Table 4). The 12 samples that tested positive for FBs exceeded the maximum limit of 1 mg/kg specified by the European Commission [1] and were collected in the locations of San Salvador-Central Region (n = 10) and San Miguel-Eastern Region (n = 2, Table 4), which are the places that function as collection centers for corn from other parts of the country to cover local demand.
During the 2024 rainy season, no FBs-positive samples were detected in the four locations sampled in this study. During this period, sample collection was irregular due to the unavailability of reagent kits for analyzing FBs. This reduced both the sampling period (from 5 to 3 months) and the number of samples collected (from 180 to 108), which may have compromised the statistical power to detect significant differences in the average content between the two seasons.
Compared to the findings of this study, the results of previous studies conducted in Mesoamerican countries indicated a middle to higher occurrence of FBs in corn flour for dough (53 %) [21], nixtamalized dough (60 % and 98 %) [14,18], and tortillas (87.5–97.5 %, 32.1 %, and 36 %) [2,9,14] (Table 8). The lower occurrence of FBs in corn dough found in this study than that in previous studies may be due to the differences in the quantification limits of the analytical methods used. In this case, FBs contents of < 1.0 mg/kg were neither quantified nor reported as traces.
The mean FBs content of corn dough samples from San Salvador (3.21 ± 0.46 mg/kg), San Miguel (1.48 ± 0.53 mg/kg), and the group of localities (2.92 ± 0.35 mg/kg) was similar to the mean or median levels in flour for dough (1.02 mg/kg) [21], nixtamalized dough (4.11 mg/kg), and tortillas (2.66 mg/kg) [14] from other Mesoamerican countries (Tables 4 and 8). In fact, the mean FBs contents found in this study were higher than the means or medians reported in other studies performed in the same geographical area, both for nixtamalized dough (0.71 mg/kg) [18] and tortillas (0.040 mg/kg and 0.064 mg/kg) [2,9] (Tables 4 and 8).
Most of the mean contents of FBs or FB1 quantified in nixtamalized corn derivatives are similar to or lower than the levels found in whole corn grains, according to the results of studies conducted in Mexico (1.39–1.70 mg/kg and 0.14–16.7 mg/kg) [37,38], Guatemala (0.66 mg/kg, 2.6 mg/kg and 0.4–31 mg/kg) [9,13,34], and Honduras (2.7–3.2 mg/kg and 7.16 mg/kg) [10,14].
In line with the abovementioned trend, the nixtamalization process and subsequent baking effectively reduced the mean FBs or FB1 contents from 0.66 mg/kg and 7.16 mg/kg in whole grain corn to 4.11 mg/kg in corn dough and 0.04 mg/kg and 2.66 mg/kg in tortillas, according to the results of two recent studies [9,14], similar to the observations for the AFT content.
This decrease is attributed to the high solubility of FBs in water and facile leaching into the liquid (known as Cernada or Nejayote) during alkalized cooking for nixtamalization and rinsing of the treated grains. The removal of the residual liquids and solid material (consisting of the pericarp, tip of the cap, and germ of the grains) contributes to the reduction of FBs in nixtamalized corn [3]. The decrease in the FBs content, specifically FB1, can vary between 68 % and 72 %, which is because of the transfer of FBs to the residual liquid fraction due to the effect of nixtamalization [3], or between 89 % and 96 %, which is because of the leaching of FBs into the Cernada or Nejayote and the residual solid material [39].
Table 8 Occurrence, contents, and estimated dietary intake of fumonisins in maize products in Mesoamerican countries.
The EDI of AFT in nixtamalized corn dough was higher during the rainy season than in the dry season owing to higher content of AFT during the former (Table 5). Geographical differences in the EDI values were also noted, predominantly in the San Miguel-Eastern Region, because of the higher consumption of corn dough as tortillas. A study conducted in Mexico also found higher EDI values for AFT in nixtamalized corn tortillas during the late rainy season than those during the dry season because of similar differences in the mean AFT content [2].
In this study, all mean EDI values failed to comply with the standard of 0.0 ng/kg bw/day for AFT ingested in food [7,20], regardless of the season of the year and geographical location (Table 5). These results are consistent with those of other studies performed in Mesoamerican countries, which reported mean EDI values that did not comply with this standard or guideline value, whether in nixtamalized corn dough [14, 17] or tortillas [2,14] (Table 7).
Additionally, regardless of the season and based on the available data for Mesoamerica, the average EDI values calculated for El Salvador (6.60 and 12.44 ng/kg bw/day) are higher than those reported for nixtamalized corn dough (1.7 ng/kg bw/day) [17] and tortillas (0.9–3.6 ng/kg bw/day) in Mexico [2]. In contrast, the values for Honduras (8.7 and 15.4 ng/kg bw/day) [14] are comparable to those of El Salvador. This contrast indicates a greater severity of exposure to AFT through the ingestion of nixtamalized corn dough in El Salvador and Honduras relative to Mexico. This is likely attributable both to higher concentrations of the mycotoxin in products from El Salvador and Honduras and to a considerably greater consumption of tortilla-based nixtamalized corn dough in these countries (252 and 490 g/day, respectively) [14,22], compared with the consumption levels reported in studies from Mexico (118 and 193.8 g/day) [2,17].
The mean EDI for FBs in nixtamalized corn dough did not show the seasonal variation observed for AFT and was restricted to the central (7.02 µg/kg bw/day) and eastern regions of the country (6.19 µg/kg bw/day) (Table 6). All mean EDI values exceeded the TDI limit of FBs, set at 1.0 µg/kg bw/day [8].
Other studies performed in the Mesoamerican region reported mean EDI values that also exceed this limit for instant corn flour [21], nixtamalized corn dough [14,18], and tortillas [14] (Table 8). The average EDI value calculated for El Salvador (9.99 µg/kg bw/day) exceeds those obtained for corn flour dough (4.6 µg/kg bw/day) [21], nixtamalized corn dough (2.02 µg/kg bw/day) [18], and tortillas in Mexico (0.17 µg/kg bw/day) [2], but is lower than the values reported for the same products in Honduras (33.57 and 21.72 µg/kg bw/day, respectively) [14]. This contrast highlights a greater severity of exposure to FBs through the ingestion of nixtamalized corn dough in El Salvador and Honduras compared to Mexico. This is attributable both to the higher concentrations of this mycotoxin found in products from these Central American countries and to the greater consumption of corn dough (252 and 490 g/day, respectively) [14,22].
In any case, the daily intake of AFT or FBs in nixtamalized corn dough is a cause for concern owing to the potential health risk to the Salvadoran population from consuming this contaminated derivative as tortillas.
Regarding health risks, it is well established that the liver, responsible for metabolizing AFT and other mycotoxins, is the primary target organ [2,7,20]. Since 2015, a convincing link has been demonstrated between consuming AFT-contaminated foods and liver cancer [40], alongside other risk factors such as excessive alcohol intake (> 45 g ethanol/day) and adult overweight and obesity [41]. Concerning FBs, repeated exposures cause liver and kidney toxicity in rodents, and chronic exposure is associated with tumor development in these organs [8]. Moreover, toxicological studies in animal models indicate that co-exposure to AFT and FBs promotes tumor growth [20].
In El Salvador, there is evidence of underreporting liver cancer cases, compounded by the absence of official data on the associated risk factors—a situation similarly observed in Guatemala [13]. According to the reported incidence of liver cancer and intrahepatic bile duct cancer in the Salvadoran population aged 60–85 years or older, 373 new cases were confirmed in 2022, representing 79.2 % of the total 471 diagnoses [42]. Moreover, it is noteworthy that this type of cancer affects women more frequently (56.8 %) than men (43.2 %) [42].
If an association were established between the prevalence of these cases and the consumption of corn derivatives contaminated with AFT or AFT + FBs, it would indicate a long-term effect, especially in the female population over 60 years. This concern is exacerbated by the high prevalence of overweight among Salvadoran women (mean body mass index of 28.6) [31], which represents an additional risk factor for liver cancer [41].
Growth impairment in infants and young children is another health risk associated with chronic exposure to AFT and FB, a link that is well documented [13,14,18,43–45]. During 2018–2022, the prevalence of growth impairment among children under 5 in El Salvador ranged between 10 and 11.4 % [46]. However, these values are lower than those reported in Guatemala, where the prevalence fluctuated between 43.5 and 45.3 % during the same period [46].
Beyond this comparison, it is evident that the health and development of El Salvador’s population under 5 are affected. If an association were established between these cases and the ingestion of corn derivatives contaminated with AFT or with AFT + FB, it would confirm a long-term effect that particularly impacts this vulnerable group.
Several limitations have been identified in this study that should be considered when interpreting the results.
First: The selection of nixtamalized corn dough as the sole matrix for evaluating AFT and FB levels limited the ability to determine contamination levels and exposure risks in other corn derivatives consumed in El Salvador. It is important to note, however, that this dough, used as a base dough, is the primary ingredient in the production of at least eight high-consumption products in the country, analogous to those produced in Guatemala [9].
Second: Due to equipment limitations in the food safety laboratory, only the direct competitive ELISA method was used to quantify AFT and FB, without confirmatory tests using complementary techniques such as HPLC or LC/MS, as performed in other regional studies [2,9,15–18,21]. Although the ELISA method is susceptible to interference from structurally similar mycotoxins (cross-reactivity) and other matrix constituents, affecting its sensitivity and specificity [47], the protocol employed was properly validated. This was demonstrated by the average percentage recoveries obtained following the spiking of each mycotoxin, as well as by the coefficient of variation and Horwitz ratio (HORRAT) values, which indicate reasonably good intra-laboratory repeatability.
Third: Although a longitudinal sampling design was implemented using a panel of 36 tortilla-producing establishments over 18 months, sample collection during the rainy season of 2024 was not continuous, reducing the sampling period from 5 to 3 months. This limitation may have affected the detection of positive FB cases during that season.
Fourth: The available data on tortilla consumption, used here as an indicator of nixtamalized corn dough consumption, are segmented solely by geographic area, without differentiation by sex or age group [22]. This limitation hinders accurate EDI calculations and makes it difficult to identify potential differences in exposure risk between men and women or among different age ranges within the Salvadoran population.
5. Conclusions
The method for analyzing AFT and FBs in nixtamalized corn dough using a direct competitive ELISA was successfully validated, yielding acceptable recovery and reproducibility values that align with similar studies, despite the method’s sensitivity and specificity limitations.
The proportion of positive cases of AFT contamination in nixtamalized corn dough was low (~20 %) and varied for the different sampling locations and the two seasons of the year, but not significantly. In contrast, the proportion of cases that exceeded the maximum limit of 10 µg/kg was very low (< 2 %) and AFT was detected only in the samples collected during the 2023 rainy season. The highest mean AFT content of the samples from the group of localities was recorded during this season, probably associated with the highest mean relative humidity recorded during that season of the year. However, the mean AFT level did not exceed the maximum limit established for corn intended for human consumption.
The proportion of the samples with FBs contamination collected in the 2023–2024 dry season was very low (< 7 %) for the center and east of the country, whereas the samples did not test positive for FBs during the rainy season in 2024. The reduction from 5 to 3 months of sampling during that season may have affected the detection of positive FB cases due to factors beyond the investigators’ control. In any case, the mean FB content of all contaminated samples exceeded the maximum limit of 1.0 mg/kg.
The low mean contents of AFT and FBs found in corn dough throughout the sampling period could be explained by the nixtamalization process, which reduces the contents of both mycotoxins (by 60–80 % for AFT and 70–90 % for FBs), consistent with the results of previous studies on the preparation of corn and tortilla.
The estimated intake of AFT in nixtamalized corn dough varied seasonally, similar to the changes in the mean content of this mycotoxin. In addition, all EDI values exceeded the standard or guideline value of 0 ng/kg bw/day, whether calculated on the basis of the mean or the 95th percentile of the AFT content. Although seasonal variation in the estimated intake of FBs could not be demonstrated, all EDI values also exceeded the guideline value of 1 µg/kg bw/day. Based on the aforementioned data, regardless of seasonal or geographic variations, the consumption of nixtamalized corn dough contaminated with AFT or FBs poses a potential health risk to the population of El Salvador that consumes this derivative as tortillas. Both mycotoxins have been identified as convincing or potential risk factors associated with liver cancer, which could affect individuals over 60 years old, especially women, and contribute to an increased risk of growth impairment in children under 5.