Essential Trace Elements in Edible Plants: Balancing Nutritional Benefits and Potential Health Risks

Abstract

This study examined the concentrations and bioaccessibility of essential trace elements (ETEs), namely, cobalt (Co), copper (Cu), iron (Fe), manganese (Mn), and zinc (Zn), in edible plants commonly consumed in Poland, assessing their nutritional contribution and potential health risks. Total and bioaccessible concentrations were measured in vegetables, fruits, and cereals using seven standardized, physiologically relevant extraction methods. Estimated Nutrient Intake (ENI) and noncarcinogenic risks (Hazard Quotients, HQs) were calculated via Monte Carlo simulations to account for variability in consumption and absorption. Results revealed substantial differences among plant types and extraction protocols, with bioaccessibility-based methods generally yielding intake estimates lower than total concentrations. Most ENI values fell within the recommended intake ranges, although some scenarios indicated potential underexposure (Co, Fe) or overexposure (Cu, Mn). These findings underscore the importance of incorporating bioaccessibility data and realistic exposure modeling into dietary risk assessment and nutritional guidelines for edible plants.

1. Introduction

Many trace elements are essential micronutrients required in minute amounts, yet they are indispensable for the structural and functional integrity of living organisms. As a subgroup of micronutrients, alongside vitamins and antioxidants, they play vital roles in supporting regenerative processes, defending against oxidative stress, and maintaining immune function in metabolically active tissues. Despite their low concentrations in biological systems, they serve as cofactors for numerous enzymatic reactions and contribute to essential physiological processes such as oxygen transport, redox balance, and cellular metabolism. − In line with the WHO and Frieden classifications, this study focuses on the essential trace elements (ETEs), namely cobalt (Co), copper (Cu), iron (Fe), manganese (Mn), and zinc (Zn), all of which have well-established roles in human nutrition and are integral to maintaining both general and oral health. Deficiency in the aforementioned elements can lead to a range of health problems. Co, as a component of vitamin B12, is critical for red blood cell formation and neurological function; its deficiency may result in megaloblastic anemia and neurological impairments. Cu is essential for iron metabolism, antioxidant defense, and connective tissue synthesis; inadequate intake can lead to anemia, weakened immunity, impaired growth and development (especially in children), osteoporosis, abnormal cholesterol metabolism, and neurological disturbances. Fe deficiency remains one of the most widespread nutritional problems globally and is the leading cause of anemia, fatigue, and impaired cognitive and immune function. Mn is involved in bone development, metabolism, and antioxidant enzyme activity; low levels can impair growth, fertility, and glucose regulation. Zn is vital for immune response, wound healing, and DNA synthesis; deficiency can cause growth retardation, delayed sexual maturation, and increased susceptibility to infections. While maintaining adequate levels of these trace elements is crucial for overall health and physiological functioning, excessive intake may also pose health risks. , The margin between beneficial and toxic doses is broad for some bioelements but narrow for others, highlighting the need for careful monitoring of their concentrations in the food supply.

The mineral content of edible plants, a major dietary source of essential elements, is influenced by numerous factors, including plant genotype (cultivar), soil properties, weather conditions, and agricultural practices such as fertilizer application during fruiting and maturation stages. Consequently, the concentration of trace elements in plant-based foods can vary widely, affecting both their nutritional value and their potential to contribute to dietary exposure risks. Customarily, the assessment of ETEs in edible plants has relied on total concentration (TC) measurements, which assume complete gastrointestinal absorption. − While this approach is commonly used to estimate nutritional contributions, it does not account for the fact that only a fraction of the ingested element may be bioaccessible and available for absorption during digestion. As a result, it may lead to either over- or underestimation of both health benefits and potential adverse effects. To better reflect human exposure, a range of in vitro extraction methods have been developed to simulate the bioaccessibility of elements under gastric and/or intestinal conditions. , These methods vary in terms of chemical reagents, pH levels, and complexity. For example, the BCR 3-step extraction targets chemical elements associated with different binding forms, while the USEPA 1340 simulates acidic gastric conditions relevant to lead bioaccessibility. More advanced protocols, like ISO 17924:2018 and the BARGE Unified Bioaccessibility Method (UBM), replicate digestive phases using physiologically relevant fluids. , Simpler methods, such as SBET and PBET, aim to simulate one or both digestion phases under controlled laboratory conditions. , In the context of ETEs such as Co, Cu, Fe, Mn, and Zn, evaluating bioaccessibility is crucial for accurate dietary intake estimates. These elements play key roles in numerous physiological functions, and both deficiencies and excess intake can affect human health. However, limited data exist on how the choice of extraction method affects estimates of their dietary contribution and associated noncarcinogenic risks.

To address this gap, the present study evaluates Co, Cu, Fe, Mn, and Zn concentrations in vegetables, fruits, and cereals using seven extraction methods. The results are used to estimate both the nutritional contribution and the potential health risks from dietary intake. To incorporate variability and uncertainty in dietary exposure, such as metal concentrations, consumption patterns, and body weight, Monte Carlo simulations were applied. This probabilistic approach enables a more robust evaluation of nutrient intake and safety, providing insights that can inform dietary guidelines and regulatory decisions. The main objective of the study was to investigate how different extraction methods, simulating gastrointestinal bioaccessibility, may impact the estimation of both nutritional contribution and potential risks. In this study, the contribution of five ETEs, namely Co, Cu, Fe, Mn, and Zn, to a healthy diet and the potential health risks associated with their intake were evaluated. The detailed objectives of the study were to 1) investigate concentrations of Co, Cu, Fe, Mn, and Zn in three plant categories, namely vegetables, fruits, and cereals, using the following seven extraction methods: Total Content (TC), three-step extraction procedure (BCR), USEPA SW-846 Test Method 1340: In Vitro Bioaccessibility Assay for Lead in Soil (USEPA), ISO 17924:2018 Soil qualityAssessment of human exposure from ingestion of soil and soil materialprocedure for the estimation of the human bioaccessibility/bioavailability of metals in soil (ISO), the BARGE (Bioaccessibility Research Group of Europe) unified bioaccessibility method (UBM), Simple Bioavailability Test (SBET), and Physiologically Based Extraction Test (PBET); 2) use Monte Carlo simulations to estimate the nutritional contribution and noncarcinogenic risk from dietary intake of these ETEs, based on Polish food consumption data; and 3) address variability and uncertainty in dietary exposure through probabilistic modeling with Monte Carlo simulations.

2. Materials and Methods

2.1. Materials

To evaluate differences in ETE concentrations across various extraction methods, samples of vegetables, fruits, and cereals were included in the study. The edible plants were sourced from multiple locations across Southern Poland (Figure ) and purchased at local fresh food markets. A detailed description of the edible plants is provided in Table S1. Prior to analysis, the samples were prepared as typically consumed: washed, peeled, and chopped into small pieces, when necessary, placed in open porcelain dishes, dried under radiant lamps at 70 °C, ground into a coarse powder using a coffee grinder, and stored in sealed bags. − To ensure the accuracy of the analyses, certified reference materials (CRMs) were analyzed alongside the edible plant samples. The following CRMs were utilized in this study: white cabbage BCR-679, spinach WEPAL-IPE-907, aubergine WEPAL-IPE-951, and wheat WEPAL-IPE-182.

1.

Sampling site locations of edible plants in southern Poland. Adapted with permission from Gruszecka-Kosowska. Copyright © 2019 by the author. Published by MDPI under the Creative Commons Attribution (CC BY 4.0) license.

2.2. Extraction Methods

In this study, seven different extraction methods differing in reagents, conditions, and simulated digestive phases were consistently applied to all plant samples and across all analyzed ETEs. Total Content (TC) employed strong acid digestion to measure total elemental levels. BCR1 targeted elements bound to carbonates using acetic acid. USEPA and SBET1 used glycine at acidic pH to simulate gastric bioaccessibility, while SBET2 adjusted to neutral pH for intestinal conditions. , PBET involved a two-step digestion with enzymes and organic acids for gastric (PBET1) and intestinal (PBET2) phases. UBM and ISO methods simulated gastric and intestinal digestion using physiologically relevant fluids and pH levels, providing comprehensive bioaccessibility assessments. , A detailed description of laboratory procedures is provided in Table S2.

2.3. Instrumental Analysis

The concentrations of investigated ETEs were determined by inductively coupled plasma-mass spectrometry (ICP-MS) (ELAN 6100; PerkinElmer, Waltham, MA, USA) in the case of Co or inductively coupled plasma optical emission spectrometry (ICP-OES) (OPTIMA 7300DV; PerkinElmer, Waltham, MA, USA) in the case of Cu, Fe, Mn, and Zn according to the United States Environmental Protection Agency (USEPA) 6020B and ISO 17294-2:2003 protocols.

2.4. Quality Assurance and Quality Control

Quality assurance and control (QA/QC) included reagent blanks and sample duplicates, using analytical-grade reagents. Results showed no significant bias (p = 0.05) and stayed within acceptable error margins. Accuracy was confirmed by six replicate calibration injections with high consistency. Rhodium served as an internal standard to correct for instrument drift and matrix effects, and element-specific corrections addressed spectral interferences. Recoveries from certified reference materials ranged from 71% to 151% for most trace elements. Limits of quantification (LOQ) were as follows: Cu < 0.005 mg/L, Co < 0.0002 mg/L, Fe < 0.01 mg/L, Mn < 0.005 mg/L, and Zn < 0.01 mg/L.

While instrumental LOQs (mg/mL) for ICP-MS and ICP-OES remained constant, LOQs per dry weight (mg/kg) varied by the extraction method due to differences in sample mass, extract volume, and dilution. TC, USEPA 1340, SBET, and PBET used similar sample-to-solution ratios, yielding comparable LOQs. The ISO 17924:2018 method had the highest LOQs because of a required dilution step (1.5 mL extract diluted with 13.5 mL HNO₃). The UBM method used less dilution, resulting in LOQs lower than those of the ISO protocol.

3. Risk Assessment

In general, various approaches can be used to evaluate risk. The first involves comparing the levels of contaminants detected in the environment or in products with the permissible limits established for those matrices. This method is commonly applied to assess the environmental quality or the safety of food products. For this purpose, maximum permissible levels (MPLs) or maximum levels (MLs) are typically used. In the case of food products, such limits are defined, for example, in Regulation (EU) 2023/915 of the European Commission, which sets maximum levels for certain contaminants in food. Another approach involves determining how readily specific contaminants can accumulate in various matricesfor example, in edible plants, this may include assessing accumulation in different plant parts such as roots, bulbs, or leaves. This approach is particularly relevant for groups of contaminants such as heavy metals for which related indices, such as the Metal Pollution Index (MPI), are often used to quantify the overall level of contamination.

When assessing dietary exposure, the first step is to calculate the daily intake of a specific food product, in combination with its contaminant concentrations. Some assessment approaches conclude at this stage, particularly when allowable or maximum permissible daily or weekly intake levels have been established by food safety authorities. For example, the European Food Safety Authority (EFSA) issues scientific opinions that define maximum levels of certain substances in food products considered safe for consumers.

If the focus is on the potential carcinogenic or noncarcinogenic risks associated with the intake of various pollutants, Cancer Risk (CR) and Hazard Quotient (HQ) values are calculated, respectively. Furthermore, when the total carcinogenic and noncarcinogenic risks are of interest, the sum of individual carcinogenic and noncarcinogenic values is determined. The latter is commonly termed the Hazard Index (HI).

In this study, only the noncarcinogenic risk was assessed, as none of the examined ETEs have been proven to exhibit carcinogenic properties. Therefore, Hazard Quotient (HQ) values were calculated for individual ETEs, since the analysis focused on single elements, and the total risk would likely be underestimated, given that only five elements were investigated. The Human Health Risk Assessment (HHRA) was conducted following the USEPA methodology, yielding risk values that were compared against the established safety threshold of 1. When HQ values exceed this limit, appropriate mitigation and protective actions are recommended, whereas values below 1 indicate negligible risk.

It is worth emphasizing that the objective of this study was not to determine specific risk values due to dietary intake of ETEs for the investigated populations, as this has been addressed in another research − as well as bioaccumulation and transfer indices of potentially harmful elements from arable soils to edible plants. , Instead, the present study aimed to provide a simplified and comparable assessment of risk levels based on the fixed and reliable reference value of 1, where the only variable considered in the risk value was the concentration of ETEs obtained through various extraction methods.

3.1. Estimated Nutrient Intake (ENI)

In this study, the Estimated Nutrient Intake (ENI) was employed to quantify the daily intake of ETEs derived from edible plants expressed in mg/person-day. ENI values were calculated for Co, Cu, Fe, Mn, and Zn based on their concentrations in vegetables, fruits, and cereals. The calculation used the equation: ENI = ∑(C × IR), where C is the ETE concentration in the plant (mg/kg) and IR is the intake rate of the plant category (g/person-day).

This metric enables direct comparison with Dietary Reference Intake (DRI) values including Recommended Dietary Allowances (RDA), Adequate Intakes (AI), and Tolerable Upper Intake Levels (UL). RDA represents the average daily intake sufficient for nearly all healthy individuals in a specific group; AI is used when RDA cannot be established and is based on observed intake; and UL defines the maximum daily intake unlikely to cause harm. Comparing ENI to these references helps to assess whether plant-based intake meets nutritional needs or exceeds safe levels, supporting dietary planning and public health evaluation.

Due to differences in the studied ETEs, uniform benchmark values were not always applicable. Where possible, EFSA reference values were used; otherwise, alternative sources were consulted. For Co, a typical intake range of 5–8 μg/day was applied. For Cu, the UL of 5 mg/day and an AI of 1.6 mg/day for men were used. Fe benchmarks included an Average Requirement (AR) of 6 mg/day for men and a safe intake level of 40 mg/day. Mn references comprised an AI of 3 mg/day and a safe level of 8 mg/day. For Zn, the UL was 25 mg/day with an AR range of 7.5 to 12.7 mg/day for men. The intake rate (IR) used in this study was based on statistical consumption data for the adult Polish population, as reported in previous research. −

3.2. Estimated Daily Intake (EDI)

The Estimated Daily Intake (EDI) of ETEs from vegetables, fruits, and cereals (mg/kg of bw/day) was simulated separately for each plant group using Monte Carlo methods. The calculation followed the equation: EDI = ∑(C × IR × EF × ED × 10^–3)/AT × BW, where C is the ETE concentration (mg/kg), IR is the intake rate (g/day), EF is the exposure frequency (days/year), ED is the exposure duration (years), AT is the averaging time (days), BW is the body weight (kg), and 10^–3 is the unit conversion factor.

3.3. Hazard Quotient (HQ)

Hazard Quotient (HQ) was used to assess the noncarcinogenic risk from long-term exposure to trace elements through edible plant consumption, using the equation: HQ = EDI/RfD, where RfD is the reference dose for each ETE. HQ values were calculated separately for vegetables, fruits, and cereals, as well as for the combined intake from all plant categories in this study.

The following Reference Dose (RfD) values were used for the noncarcinogenic risk (HQ) calculations (mg/kg bw-day): Co 3.0 × 10^–4, Cu 4.0 × 10^–2, Fe 0.7, Mn 0.14, and Zn 0.3. − In this study, all RfD values were obtained from the US EPA toxicological databases to ensure a consistent and standardized source of toxicological data. The primary objective of the research was to examine the variation in risk values attributable exclusively to a single factor: the bioaccessible concentrations of the investigated elements.

An HQ value ≤1 was set as the threshold for acceptable noncarcinogenic risk, following USEPA guidelines.

3.4. Monte Carlo Simulations

To account for uncertainties in HHRA, especially from variability in the ETE levels and exposure factors, a probabilistic Monte Carlo simulation was used. Unlike fixed-point deterministic methods, it captures the full range of variability, yielding more realistic and reliable risk estimates. This reduces the risk of over- or underestimation due to environmental variation, population differences, and data limitations.

Monte Carlo simulations were run in @RISK (Lumivero) with 10,000 iterations to generate probabilistic HQ values. Input variables, including ETE concentrations, intake rates, exposure frequencies, and body weights (Table ), were modeled using distributions to assess population exposure from edible plant consumption.

1. Exposure Parameters Used for Intake Calculations Using Monte Carlo Simulations.

Parameter	Unit	Value	Distribution	References
Average body weight (BW)	kg	(67.55; 8.72)	Lognormal
Exposure duration	year	(0; 24)	Uniform
Intake rate (IR)vegetables	g/day	(200.5; 292.5)	Uniform	Distribution based on literature data on polish consumption, −
Intake rate (IR)fruits		(135.17; 261.83)
Intake rate (IR)cereals		(142.57; 299.43)
Averaging time (AT)	day	365 × ED	Point
Exposure frequency (EF)	day/year	(180; 350; 365)	Triangular
Concentration (C)	mg/kg dry weight (d.w.)	Varied according to metals and extraction methods	Triangular	This study

A sensitivity analysis was conducted alongside the simulation to identify which input variables contributed most to the overall model uncertainty. This helped to highlight the key factors influencing health risk estimates and provided insight into the main drivers of risk variability.

All extraction methods were consistently applied to the same samples and trace elements. However, only ETEs detected above the LOQ in most samples were included in Monte Carlo simulations to avoid bias from the imputation. As a result, only these elements contributed to the sensitivity analysis, and variability in parameter sensitivities may reflect element inclusion differences rather than extraction method effects.

4. Results and Discussion

4.1. Essential Trace Element Concentrations in Edible Plants

Descriptive statistics for the ETEs in edible plants depending on the extraction method are presented in Table .

2. Mean Values and Standard Deviations (in Parentheses) of ETE Concentrations in Edible Plants Depending on Extraction Method.

Vegetablesmean (SD) mg/kg d.w.
Trace element	TC	US EPA	UBM gastric	UBM intestinal	ISO gastric	ISO intestinal	SBET1	SBET2	PBET1	PBET2	BCR1
Co	0.060 (0.022)	0.023 (0.017)	0.040 (0.016)	0.039 (0.017)	<0.075	<0.195	0.020 (0.012)	<LOQ	<LOQ	0.018 (0.008)	0.020 (0.016)
Cu	32.8 (14.5)	<LOQ	2.86 (1.66)	4.58 (1.58)	<1.875	<4.875	0.880 (0.901)	0.919 (0.745)	1.89 (1.13)	2.05 (0.930)	2.11 (1.00)
Fe	91.0 (35.6)	15.8 (6.50)	27.1 (9.90)	15.8 (11.6)	29.7 (10.6)	86.1 (55.6)	9.37 (6.49)	5.34 (3.69)	7.35 (5.04)	13.1 (8.36)	6.17 (2.55)
Mn	13.4 (10.0)	15.9 (5.50)	17.7 (5.80)	15.3 (4.93)	8.75 (4.58)	<4.875	14.9 (5.09)	5.98 (1.94)	14.7 (5.26)	5.22 (2.79)	13.0 (4.34)
Zn	30.5 (22.4)	13.4 (17.9)	29.3 (16.2)	23.8 (13.4)	37.5 (23.6)	38.5 (22.7)	17.6 (18.0)	5.40 (5.43)	18.9 (14.2)	17.4 (8.34)	16.6 (12.6)
Fruitsmean (SD) mg/kg d.w.
Trace element	TC	US EPA	UBM gastric	UBM intestinal	ISO gastric	ISO intestinal	SBET1	SBET2	PBET1	PBET2	BCR1
Co	0.089 (0.137)	0.082 (0.125)	0.083 (0.122)	0.089 (0.129)	<0.075	<0.195	<0.02	<0.02	<0.02	<0.02	0.057 (0.100)
Cu	19.6 (2.50)	2.52 (2.87)	6.65 (3.20)	6.82 (3.49)	<1.875	<4.875	3.69 (3.44)	1.25 (1.42)	2.75 (3.06)	2.63 (1.86)	2.74 (3.25)
Fe	58.5 (28.4)	13.6 (13.0)	22.6 (15.9)	<LOQ	28.8 (19.9)	46.3 (26.9)	10.2 (10.9)	3.23 (3.92)	4.56 (4.93)	7.42 (5.69)	3.32 (2.42)
Mn	11.2 (7.68)	13.4 (7.90)	14.2 (8.89)	10.1 (6.65)	<1.875	<4.875	12.3 (7.37)	2.20 (1.80)	13.2 (7.98)	2.03 (1.58)	14.2 (8.96)
Zn	21.6 (8.02)	9.05 (12.0)	20.0 (5.71)	7.57 (7.40)	22.0 (7.91)	<9.75	10.2 (6.05)	1.38 (1.03)	18.4 (4.26)	7.20 (2.56)	16.7 (7.10)
Cerealsmean (SD) mg/kg d.w.
Trace element	TC	US EPA	UBM gastric	UBM intestinal	ISO gastric	ISO intestinal	SBET1	SBET2	PBET1	PBET2	BCR1
Co	0.032 (0.033)	<0.02	<0.008	<0.021	<0.075	<0.195	<0.02	<0.02	<0.02	<0.02	<0.008
Cu	9.17 (3.24)	<LOQ	4.39 (1.85)	5.93 (1.32)	<1.875	<4.875	1.55 (1.09)	0.489 (0.332)	2.19 (0.959)	1.61 (0.563)	<0.2
Fe	48.5 (33.2)	11.2 (10.4)	21.0 (14.2)	<1.07	15.7 (24.0)	16.6 (23.5)	10.9 (8.69)	4.92 (4.90)	6.92 (9.43)	14.7 (7.32)	5.51 (11.7)
Mn	19.6 (11.9)	170.5 (347.1)	23.0 (14.8)	10.9 (6.25)	13.5 (11.8)	<4.875	20.2 (12.6)	9.41 (5.76)	21.5 (12.5)	6.78 (5.04)	23.3 (15.6)
Zn	35.3 (11.2)	17.3 (14.3)	40.0 (19.3)	<1.07	46.0 (13.4)	<9.75	32.9 (12.1)	8.54 (3.64)	41.8 (14.2)	<1.0	42.5 (16.5)

Co concentrations were above the LOQ in cereals only with the TC method, but were quantifiable in vegetables and fruits with the USEPA, UBM (gastric and intestinal), and BCR1 methods, and in vegetables with the SBET1 and PBET2 methods. Cu levels were consistently below the LOQ with both ISO methods in all matrices and with the USEPA method for vegetables and cereals. Fe was quantified in vegetables across all methods but fell below the LOQ in fruits and cereals when using the UBM intestinal method. Mn was detected with all methods except ISO intestinal (all matrices) and ISO gastric (fruits). Zn was quantifiable in vegetables across all methods but fell below the LOQ in fruits and cereals with ISO intestinal and in cereals with PBET2.

Based on the TC method, the descending order of ETE concentrations was as follows: vegetablesFe > Cu > Zn > Mn > Co; fruitsFe > Zn > Cu > Mn > Co; cerealsFe > Zn > Mn > Cu > Co. The consistently high Fe concentrations align with its known abundance in plants due to its role in photosynthesis and respiration. , Other methods showed varying patterns: USEPA: vegetablesMn ≈ Fe > Zn > Co, fruitsFe > Mn > Zn > Cu > Co, cerealsMn > Zn > Fe; UBM gastric: vegetablesZn > Fe > Mn > Cu > Co, fruitsFe > Zn > Mn > Cu > Co, cerealsZn > Mn > Fe > Cu; UBM intestinal: same as UBM gastric in vegetables, fruitsMn > Zn > Cu > Co, cerealsMn > Cu; ISO gastric: vegetables and cerealsZn > Fe > Mn, fruitsFe > Zn; ISO intestinal: vegetablesFe > Zn, fruits and cerealsonly Fe above LOQ; SBET1: vegetablesZn > Mn > Fe > Cu > Co; fruitsMn > Zn ≈ Fe > Cu, cerealsZn > Mn > Fe > Cu; SBET2: vegetables and cerealsMn > Zn > Fe > Cu, fruitsFe > Mn > Zn > Cu; PBET1: Zn > Mn > Fe > Cu across all groups; PBET2: vegetablesZn > Fe > Mn > Cu > Co, fruitsFe > Zn > Cu > Mn, cerealsFe > Mn > Cu; BCR1: vegetables and fruitsZn > Mn > Fe > Cu > Co, cerealsZn > Mn > Fe.

Co was consistently the least abundant element across all methods and plant types, reflecting its naturally low uptake in plants and limited essentiality, except in legumes where it supports nitrogen fixation. ,

4.2. Estimated Nutrient Intake (ENI)

Estimated Nutrient Intake of ETEs (50th percentile) from all edible plants intake from Monte Carlo simulations is shown in Figure .

2.

Estimated Nutrient Intake (ENI) of ETEs based on the 50th percentile of all investigated edible plant intake from Monte Carlo simulations. TI, Tolerable Intake; UL, Tolerable Upper Intake Level; AI, Adequate Intake; AR, Average Requirement; SLI, Safe Level of Intake.

4.2.1. Cobalt (Co)

Estimated intakes in this study exceeded the typical range of 5–8 μg/day. The TC method showed 46.84 μg/day, over five times higher, while UBM gastric and intestinal phases yielded 32.78 μg/day and 33.66 μg/day, respectively. Only SBET1 (5.65 μg/day) and PBET2 (4.16 μg/day) fell within or below the typical range. Although no UL exists, these results suggest potentially elevated intake, especially in plant-heavy diets or Co-rich regions. However, as animal products are the main dietary Co source, these values likely reflect only a portion of total intake.

4.2.2. Copper (Cu)

The ENI for Cu via the TC method (14.2 mg/day) significantly exceeds the UL, nearly tripling it. This may suggest a potential liver toxicity risk, though it likely reflects TC’s tendency to overestimate exposure. UBM methods (3.2 mg/day for gastric and 3.6 mg/day for intestinal phases) fall safely between the AI and UL. All other methods yield ENI values below the AI, with the lowest from the USEPA method (∼0.5 mg/day), indicating possibly insufficient Cu intake.

4.2.3. Iron (Fe)

The ENI for Fe from the TC method (45.6 mg/day) exceeds the safe intake level (40 mg/day) by ∼14%, indicating possible overestimation. In contrast, methods like UBM intestinal, SBET2, BCR1, and PBET1 yield ENI values below the AR of 6 mg/day, suggesting potential insufficiency. Other methods produce values within a safe and balanced range.

4.2.4. Manganese (Mn)

For most methods, Mn ENI exceeds the safe intake level with the USEPA method surpassing it by over 8-fold. This underscores the need for careful Mn risk assessment, as such intake may pose health concerns. Since plants contain significantly more Mn than animal products, this likely captures the majority of dietary Mn exposure.

4.2.5. Zinc (Zn)

Across all methods, Zn ENI remained below the UL, indicating no toxicity risk. Values from USEPA, UBM intestinal, and ISO intestinal methods fell within the AR range, while others were slightly above or below (SBET2 and PBET2), but still within safe limits.

4.3. Human Health Risk Assessment

4.3.1. Noncarcinogenic (HQ) Risk for Investigated ETEs from Various Extraction Methods

Noncarcinogenic risk (HQ) values for 5th, 50th, and 95th percentiles of the determined concentrations of investigated ETEs in vegetables, fruits, and cereals calculated using Monte Carlo simulations are presented in Figure , and the same values for the sum of edible plants consumed are presented in Figure .

3.

Noncarcinogenic (HQ) risk values for investigated trace elements from various extraction methods according to Polish statistical consumption of edible plants, vegetables, fruits, and cereals separately.

4.

Noncarcinogenic (HQ) risk values for investigated trace elements from various extraction methods according to Polish statistical consumption of the sum of edible plants.

Due to concentrations below the LOQ, HQ values could not be calculated for several ETEs in vegetables, fruits, and cereals across the different extraction methods. For instance, Co in cereals yielded measurable concentrations only under the TC method, whereas Fe in vegetables was consistently above the LOQ for all extraction methods, allowing HQ estimation in each case. These results indicate that the efficiency of element recovery is both element- and matrix-dependent and that no single extraction protocol is universally optimal for all ETE–matrix combinations.

Considering HQs that were successfully simulated across most elements (particularly Co, Cu, and Fe), the TC method produced the highest HQ values, exceeding the risk threshold (HQ = 1) at the 95th percentile for Co and Cu. The 95th percentile HQ for Co under TC reached 1.05 in vegetables and 2.77 in fruits, staying below the safety threshold for cereals, while for Cu it reached 4.30 in vegetables, 1.83 in fruits, and 1.08 in cereals, indicating potential overestimation if full bioavailability is assumed. For Fe the risk was lower, staying below the threshold in all cases, but TC still displayed the highest values in most cases with the 95th percentile HQ reaching 0.66 in vegetables, 0.39 in fruits, and 0.43 in cereals. The only exception to this trend was the 95th percentile risk for Fe in vegetables, where the ISO intestinal yielded 0.87.

In contrast, bioaccessibility-based methods (e.g., UBM, PBET, SBET, and BCR) generally generated lower HQs, often falling below the threshold. For example, 95th percentile Co HQ values under UBM gastric and intestinal extractions were ∼0.75 in vegetables, while USEPA was 0.56 and SBET, PBET, and BCR yielded <0.50. A similar trend was observed for Cu, where UBM intestinal yielded 0.58 vs 4.30 under TC.

However, Mn and Zn showed distinctively different patterns. For Mn in cereals, the USEPA method produced a median HQ of 5.18 and 15.7 at the 95th percentile, far higher than all the other methods, including the TC method (median 0.32). For Zn in vegetables, ISO gastric (0.39) and intestinal (0.37) extractions gave median HQs higher than that of TC (0.27).

Interestingly, the bioaccessible methods yielded higher Mn and Zn concentrations than the total content method. This phenomenon may be attributed to the greater solubilization efficiency of the simulated gastrointestinal fluids, which contain acids and complexing agents capable of mobilizing Mn and Zn through complex formation and dissolution of labile phases. , In contrast, the total content (TC) digestion may underestimate these elements due to incomplete decomposition of certain mineral- or organic-bound forms. Furthermore, in matrices rich in organic matter, such as cereals and vegetables, Mn and Zn are prone to forming soluble complexes that are more readily released under bioaccessibility test conditions than during strong acid digestion. These factors collectively contribute to the higher concentrations observed in the bioaccessible fractions compared with the total content measurements. −

Overall, the greatest discrepancies between a single extraction method and the others were observed for Cu and Mn. In the case of Cu, the risk estimated using the Total Content (TC) method was substantially higher than that obtained with all other extraction protocols. For Mn, the highest risk was associated with the USEPA method, which markedly diverged from the results of the remaining approaches.

A cross-matrix comparison of HQ values for ETEs revealed consistent trends between plant types, as well as notable element-specific differences related to both matrix characteristics and extraction method sensitivity. Vegetables generally produced the highest HQ values across most elements and methods. This suggests that vegetables are a particularly relevant dietary source of these elements, likely due to their relatively high moisture content, soft tissue structure, and weaker binding of metals, which facilitate release under both total digestion and simulated gastrointestinal conditions. Fruits showed intermediate HQ values compared to those of vegetables and cereals. For Co and Cu, HQ values exceeded the risk threshold at higher percentiles when total concentrations were assumed fully bioavailable, but bioaccessibility-based methods typically gave lower estimates. For Fe, fruits consistently exhibited the lowest HQ values among plant groups, especially with selective extraction procedures, such as BCR1 and PBET, indicating lower Fe bioaccessibility likely linked to the presence of organic acids, pectins, and other matrix constituents. Cereals displayed a distinct risk profile. For Co and Cu, the TC-derived HQs were lower than those observed in vegetables and fruits, reflecting their overall lower total metal concentrations. However, for Mn, cereals stood out as the dominant source of potential risk, with HQ values exceeding the safe threshold, particularly under the USEPA extraction, which may overestimate the soluble Mn fraction. This likely reflects the high Mn content in the outer layers of grains and its greater mobilization under acidic extraction conditions.

These findings highlight that bioaccessibility does not always correlate linearly with the total concentrations. Instead, it is strongly dependent on how metals are chemically bound within different food matrices and how these bonds respond to physiological extraction conditions. Matrix composition, including but not limited to moisture content, organic acids, and fiber, plays a key role in determining the fraction of trace metals that becomes bioaccessible. For example, Zn is often weakly complexed with organic ligands, making it readily soluble in the gastric phase. For Mn, a similar trend can be explained by differences in the chemistry of Mn species under the extraction conditions used. In this study, TC was determined using a standard solid digestion protocol based on HNO₃ and H₂O₂. While this approach is commonly applied for total trace metal determination in food matrices, it may underestimate the total Mn content because MnO₂ is poorly soluble in nitric acid and catalyzes the decomposition of hydrogen peroxide in aqueous solutions. Consequently, a fraction of Mn bound in more resistant mineral phases may not be fully released during TC digestion.

This underscores how both matrix composition and the extraction method shape dietary risk estimates. Soft, water-rich matrices like vegetables tend to release metals more readily, while cereals exhibit selective solubilization patterns tied to specific elements and their chemical binding forms. Consequently, the method choice can either exaggerate or mask actual exposure, emphasizing the need to consider both plant matrix effects and physiologically relevant extraction procedures in dietary risk assessment.

The cumulative ∑HQ at the 50th percentile across edible plants further highlights these patterns. For Co, ∑HQ reached 1.88 with TC and decreased below 1 with BCR, SBET, PBET, and USEPA methods. For Cu, ∑HQ reached 4.21 with TC and ∼1 with UBM intestinal, with other methods below the safety threshold. For Fe, ∑HQ reached 0.77 with TC and varied between 0.05 and 0.57 for bioaccessibility-based methods. For Mn, ∑HQ peaked at 5.71 with USEPA, while all other methods yielded ≤1. For Zn, all values were <1; however, the highest value was observed for ISO gastric (0.95). These differences underscore how strongly extraction method selection drives dietary risk interpretation.

4.3.2. Sensitivity Analysis

A sensitivity analysis was conducted to assess how key exposure parameters, metal concentration (C), intake rate (IR), body weight (BW), exposure frequency (EF), and exposure duration (ED), influence Hazard Quotient (HQ) values across extraction methods. This analysis identifies the most impactful variables on health risk outcomes and quantifies their relative contributions to model sensitivity.

Figure shows trace element concentration (C) as the most influential factor across all methods, with sensitivity values from 0.78 (TC) to 0.91 (ISO intestinal), confirming it as the primary risk driver.

5.

Sensitivity analysis of noncarcinogenic risk for each extraction method used: Spearman Rank Correlation Coefficients.

Intake rate (IR) was the second-most influential factor, highest in the TC method (0.36) and lower in bioaccessibility-based methods (0.23–0.30). IR had a greater impact in PBET1 (0.33) and PBET2 (0.31), indicating that intake rate plays a stronger role in some intestinal-phase models.

Exposure frequency (EF) moderately influenced HQ values across all models, with sensitivity values between 0.21 and 0.28 for bioaccessibility-based methods and 0.30 for the TC method, consistently ranking third in importance.

Body weight (BW) showed a consistent negative correlation with HQ values, meaning that a higher BW reduces estimated risk. This effect was modest but steady across methods, strongest in the TC (−0.26), UBM intestinal (−0.24), and PBET2 (−0.24) methods, and weakest in the USEPA method (−0.19), indicating a small but consistent risk-moderating role.

Exposure duration (ED) had negligible sensitivity in all methods, with values near zero and both positive and negative signs. The highest (yet still minimal) effect was in the ISO intestinal method (0.0075), showing ED has a small impact on overall risk estimates.

Overall, these findings highlight ETE concentration as the key risk factor, with intake rate and exposure frequency varying by method. Body weight has a modest, consistent moderating effect, while exposure duration’s impact is minimal across all models.

4.4. ENI vs HQ: Cross-Validation of Nutritional Exposure and Health Risk

4.4.1. Cobalt (Co)

Although no UL exists for Co, ENI values from the TC and UBM methods greatly exceeded typical dietary levels, indicating an elevated level of exposure. Summed HQ values at the 50th percentile also surpassed the safety threshold (HQ  = 1.0), suggesting potential noncarcinogenic risks, especially when assuming full or high bioaccessibility.

4.4.2. Copper (Cu)

ENI from the TC method exceeded the UL, signaling possible health concerns. Summed HQ values across all plant categories surpassed 4 at the 50th percentile, well above the safe limits. This consistent exceedance highlights a substantial risk from excessive Cu intake, particularly under assumptions of high bioaccessibility or total solubility.

4.4.3. Iron (Fe)

ENI for Fe from the TC method slightly exceeded the UL by about 14%, indicating a marginal risk. HQ values at the 50th percentile remained below concern, but 95th percentile HQs from the TC and ISO intestinal methods slightly exceeded 1.0. This suggests that high-end intake scenarios, especially with increased bioavailability, could pose risks to vulnerable groups.

4.4.4. Manganese (Mn)

ENI for Mn exceeded the UL under most methods, with the USEPA method showing intake over eight times the safe level. Corresponding HQ values for cereals were 5.18 (50th percentile) and 15.71 (95th percentile), indicating a significant dietary risk. Although some methods showed moderate HQs, overall Mn exposure risks are high depending on diet and extraction assumptions.

4.4.5. Zinc (Zn)

None of the methods produced ENI values exceeding the UL, suggesting no immediate toxicity risk from the total intake. However, HQ values at the 95th percentile exceeded 1.0 for six methods (TC, UBM gastric, ISO gastric, SBET1, PBET1, BCR1), indicating potential risk under gastric-phase conditions or high exposure. This implies that the UL may underestimate risk when gastric bioaccessibility of Zn is high.

4.5. Considerations for Bioaccessibility Method Selection of ETEs in Risk Assessment

Across the investigated edible plants, the UBM gastric and intestinal, ISO gastric, and BCR1 methods generally yielded comparable bioaccessibility values for most elements. For example, in vegetables, Co ranged from 0.039–0.040 mg/kg d.w. in UBM methods and 0.020 mg/kg d.w. in BCR1, while Mn values were similar across UBM, ISO, and BCR1 (13–17 mg/kg d.w.). Zn bioaccessibility was also consistent between UBM and ISO methods (23–38 mg/kg d.w.). In fruits, Co and Mn showed close agreement between UBM and BCR1, whereas Cu and Fe values were better aligned between UBM intestinal and PBET1 and PBET2 (2–7 mg/kg d.w.). For cereals, Mn and Zn were comparable across SBET1, PBET1, and BCR1 (20–42 mg/kg d.w.), while Cu and Fe showed closer agreement between UBM and PBET1.

In contrast, the TC method consistently produced the highest values across nearly all ETEs and food matrices, including Cu in vegetables (32.8 mg/kg d.w.), Fe in fruits (58.5 mg/kg d.w.), and Mn in cereals (19.6 mg/kg d.w.). Conversely, the USEPA, SBET2, and PBET2 methods frequently returned lower or <LOQ values, such as Cu in vegetables (<LOQ–2.05 mg/kg d.w.) and Fe in cereals (<1.07–14.7 mg/kg d.w.), marking them as outliers relative to the other extraction techniques.

Overall, the consistency among UBM, ISO, PBET1, and BCR1 methods suggests that these methods are the most reliable for assessing essential trace element bioaccessibility in edible plants, making them particularly suitable for human health risk assessment studies. The choice of the extraction method has a decisive impact on estimated dietary risks. For Co and Cu, TC-based estimates often exceeded the HQ threshold at median consumption levels, indicating substantial overestimation if bioaccessibility is not considered. For Mn and Zn, gastric-phase methods produced the highest HQ values, illustrating that bioaccessibility-based approaches do not always yield lower risks and can sometimes reveal hidden exposure potential.

Integrating physiologically relevant extraction methods provides more realistic and informative exposure assessments. Among the tested methods, UBM gastric and intestinal offers robust and representative values for most elements, balancing physiological realism and reproducibility. Simpler protocols (e.g., SBET, PBET) may be useful for conservative screening but can underestimate exposure for elements that are highly soluble under gastric conditions.

We recommend that future dietary risk assessments for essential elements incorporate bioaccessibility-corrected data and consider food matrix-specific factors. This approach will improve both the accuracy of exposure estimates and their relevance to public health risk management.

4.6. Study Limitations

This study provides important insights into how extraction methods affect the dietary risk assessment of ETEs in edible plants, but several limitations should be noted. Although the in vitro bioaccessibility protocols employed are widely accepted and standardized, they may not fully capture the complexity of human nutrient absorption, which is influenced by physiological factors such as enzymatic activity, pH dynamics, and gut microbiota composition. These factors can substantially alter the actual bioavailability of ETEs and, consequently, the nutritional or toxicological outcomes of dietary exposure. The dual role of ETEs as essential nutrients and potential toxicants complicates interpretation, as the same intake level might indicate deficiency risk in some populations but overexposure in others, depending on physiological factors. Monte Carlo simulations relied on aggregated data and literature intake values that may not capture regional dietary variations, body weight differences, or bioavailability modifiers. Some extraction methods, especially gastric or intestinal simulations, had concentrations below detection limits, limiting a consistent risk assessment. Additionally, the study considered only plant-based ETE sources, excluding significant contributions from other food groups such as dairy, meat, and fortified products. Lastly, the focus on selected crops from specific environments limits generalizability; broader studies including diverse crops, regions, and in vivo validation would strengthen these findings in terms of both nutritional adequacy and toxicity risk.

In the unification of risk assessment procedures, a wide range of input factors must be considered. In this study, we focused exclusively on one such factor, which was the bioaccessible concentration of the investigated element. However, several other parameters are also critical. For instance, body weight and age of the studied population play important roles, as each organism exhibits individual variability in the physiological response. Moreover, the toxicological reference values, here referenced to RfD, are derived from animal studies that determine exposure levels unlikely to cause adverse health effects in target organisms. Even in this context, considerable variability exists, introducing additional uncertainty regarding which RfD value should be adopted to ensure the most reliable risk characterization. In the current study, RfD values were obtained from the US EPA toxicological database; however, other studies have proposed alternative RfD values for the same elements that differ from the official ones. For example, the RfD value for Co reported in the US EPA database is 3.0 × 10^–4 mg/kg bw-day, whereas a previous study suggested that a higher value of 3.0 × 10^–2 mg/kg bw-day would provide sufficient protection for oral exposure routes.

4.7. Implications

This study shows that the extraction method choice greatly affects Estimated Nutrient Intake (ENI) and dietary exposure assessments of essential trace elements in edible plants. Simulated gastrointestinal methods, reflecting human physiology, generally report lower concentrations than the TC method, indicating that relying on total concentrations alone can overestimate intake by ignoring bioaccessibility, a key factor in accurate risk assessment.

ENI assessments help evaluate nutrient over- or underexposure but may miss health risks. Monte Carlo-based Hazard Quotient (HQ) analysis provides a more precise dietary risk estimate by accounting for bioavailability and intake variability. This combined approach is crucial for essential elements, balancing deficiency and toxicity concerns.

For most elements, ENI values matched recommended intakes, but variations occurred by element, food group, and extraction method. Some simulations indicated possible insufficient intake risks (e.g., Co and Fe in certain plants), while others showed overexposure risks (notably, Cu and Mn) when assuming full bioavailability from total concentrations. These results highlight the need to include realistic bioaccessibility in dietary nutrient assessments.

Comparing HQ values across plant types and elements, the study highlights how vegetables, fruits, and cereals contribute differently to nutritional risk. This comprehensive approach offers a clearer picture of dietary exposure than analyses based on only total concentrations or single foods. While most essential trace elements posed no immediate risk under realistic bioaccessibility assumptions, Co and Cu results indicate a potential for excessive intake.

Our results revealed that UBM, ISO, PBET1, and BCR1 methods provide the most reliable estimates of ETE bioaccessibility in edible plants, while TC tends to overestimate and simpler methods (SBET, PBET2) may underestimate exposure. Incorporating bioaccessibility-corrected data and considering food matrix-specific factors are essential for accurate dietary risk assessments and effective public health management.

In conclusion, these findings highlight the importance of incorporating physiologically relevant extraction methods into dietary risk assessment to complement the total concentration data in food safety and nutrition research. The results may inform future regulatory frameworks, dietary guidance, and fortification strategies. Further research should focus on validating in vitro bioaccessibility with in vivo evidence, exploring nutrient–contaminant interactions, and expanding assessments across diverse diets and populations to strengthen the evidence base for nutrition policy development.

Glossary

Abbreviations

AI

Adequate Intake

AR

Average Requirement

AT

averaging time

BARGE

Bioaccessibility Research Group of Europe

BCR

Bureau Community of Reference

BCR1

1st step of BCR sequential extraction procedure

BW

body weight

CRM

Certified Reference Material

DRI

Daily Recommended Intake

ED

exposure duration

EDI

Estimated Daily Intake

EF

exposure frequency

ENI

Estimated Nutrient Intake

ETE

Essential Trace Element

HHRA

Human Health Risk Assessment

HQ

Hazard Quotient

ICP-MS

Inductively Coupled Plasma Mass Spectrometry

ICP-OES

Inductively Coupled Plasma Optical Emission Spectrometry

IR

intake rate

ISO

International Organization for Standardization

LOQ

Limit of Quantification

PBET

Physiologically Based Extraction Test

PBET1

1-step of Physiologically Based Extraction Test

PBET2

2-step of Physiologically Based Extraction Test

RDA

Recommended Daily Allowance

RfD

Reference Dose

SD

standard deviation

SBET

Simple Bioavailability Extraction Test

SBET1

1-step of Simple Bioavailability Extraction Test

SBET2

2-step of Simple Bioavailability Extraction Test

SLI

Safe Level of Intake

TC

Total Concentration

TI

Tolerable Intake

QA/QC

Quality Assurance/Quality Control

UBM

the BARGE Unified Bioaccessibility Method

UL

Tolerable Upper Intake Levels

USEPA

United States Environmental Protection Agency

USEPA 1340 method

SW-846 Test Method 1340: In Vitro Bioaccessibility Assay for Lead in Soil

The data sets generated during the current study are available from the corresponding author on reasonable request.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.5c10365.

• Description of the edible plants investigated in the study (Table S1); characteristics of extraction methods used in the study (Table S2) (PDF)

Conceptualization, A.S. and A.G.-K.; Data curation: A.S., P.O.S., and A.G.-K.; Formal analysis: A.S.; Funding acquisition: A.G.-K.; Investigation: A.S.; Methodology: A.S. and A.G.-K.; Project administration: A.G.-K.; Resources: A.S., P.O.S., and A.G.-K.; Software: A.S.; Supervision: A.G.-K.; Validation: A.S.; Visualization: A.S.; Writingoriginal draft: A.S., P.O.S., and A.G.-K.; Writingreview and editing: A.S. and A.G.-K.

Research project supported by program Excellence initiativeresearch university for the AGH University of Krakow and Statutory Research Grant No. 16.16.140.315 by the AGH University of Krakow.

The authors declare no competing financial interest.