Determination of cadmium in Chinese pepper and its health implications based on bioaccessibility

The contamination of cadmium (Cd) in agro-products causes major concern because of its potential dietary risks. In this study, a total of 647 pepper samples from 21 provinces in China were randomly collected according to the distribution of pepper production. Cd pollution levels in Chinses pepper and its health risks were evaluated based on bioaccessibility, which was measured by the physiologically based extraction test (PBET). The results showed that Cd concentration in all pepper ranged from 0.002 to 1.470 mg/kg, with an average of 0.222 mg/kg and a median of 0.132 mg/kg. The highest daily intake of Cd was observed in the female child group (4.037 × 10−5 mg/kg bw/day), which accounted for 4% of the maximum daily permissible dose – 0.001 mg/kg bw/day. The target hazard quotients of Cd were all lower than 1, indicating low potential non-carcinogenic health risks to residents via the consumption of pepper. Notably, carcinogenic risk values suggested potential adverse health effects to adults, while after considering the bioaccessibility of Cd in pepper (mean of 43.07%), those values had fallen under the acceptable level (1 × 10−4). This may indicate that dietary risk assessment of heavy metals in crops could not be conducted just based on their content; the bioaccessibility of metals is also an important factor for consideration.


Introduction
Cadmium (Cd) is a heavy metal and ubiquitous environmental pollutant, which can pose a range of severe diseases (EFSA 2012). As a kind of chronic potent nephrotoxin, Cd can be accumulated in the kidney with a half-life of 10-30 years (Clemens et al. 2013), resulting in renal damage, proximal tubular reabsorptive dysfunction, and osteopenia (Bernard 2016;Satarug et al. 2010). Besides, numerous epidemiological studies have shown a close association between long-term chronic exposure to Cd and diseases of gestational diabetes mellitus (Xing et al. 2018), cardiovascular disease, and cancer (Riederer et al. 2013). In addition, Cd is classified as a human carcinogen (group 1) and could lead to apoptosis and necrosis (IARC 1994).
The way of Cd exposure includes foodstuffs, drinking water, smoking, and occupational exposure, but in the nonsmoking general population, dietary intake is the dominant source (EFSA 2009). Cd can be readily transported to plants and be detected in various foodstuffs such as rice, wheat, and vegetables (Alloway 2013;Hong et al. 2018). Cluster analysis has revealed the toxicity and health risk characteristics of Cd in different vegetables, especially in leafy vegetables such as pakchoi, leek, and water spinach (Fang et al. 2019;Huang et al. 2020Huang et al. , 2017. Pepper is one of the most important and widely consumed vegetables and condiments, with a planting area of 3.8 million hm 2 worldwide (Tripodi et al. 2018). As the largest pepper producer and consumer, China has 2.0 million hm 2 of pepper-planting area, and the total production is 40 million tons every year (Luo et al. 2020). Considering the high consumption of pepper, it is necessary to evaluate the Cd levels of pepper and its potential health implications.
Research on Cd in pepper mainly focus on its accumulation and migration characteristics in plant and its related Responsible Editor: Philippe Garrigues * Changhu Lin linchanghu79@sina.com influencing factors such as varieties and distribution in the soil (Liu et al. 2021;Luo et al. 2020;Wang et al. 2019;Zhuang 2009), while there are few studies on the dietary exposure and health risk assessment of Cd in pepper. Therefore, a nationwide investigation of Cd in Chinese pepper and its potential health implications for the Chinese population needs to be conducted more comprehensively. The dietary risk of hazardous substances posed by the consumption of foods can be evaluated using the risk assessment guidance recommended by the United States Environmental Protection Agency (USEPA 1989). In this guidance, dietary exposure to pollutants was acquired by simply multiplying the total content of substances in food and the daily consumption of food without considering its bioaccessible concentrations. Previous studies have demonstrated that only a fraction of ingested metals could be released from the matrix into digestive juices (Jin and Cui 2013;Lv et al. 2011;Liu et al. 2019). The absorption of Cd in the gastrointestinal tract is affected by many factors such as essential metals, vitamins, phytochemicals, and probiotics; all of which could influence the dissolution of Cd and hence modulate its toxicity (De Moura et al. 2015;Reeves and Chaney 2008;Zhai et al. 2015). For instance, it has been observed that marginal deficiencies of some essential nutrients (Zn, Fe, Ca) could enhance Cd absorption from diets as much as ten-fold (Reeves and Chaney 2008). Meanwhile, there are antagonistic interactions between Cd and Ca in lichens (Jozef et al. 2020), rats (Marijana et al. 2002), and humans (Kazantzis 2004) because of their similar physicochemical properties as well as shared transport channels. Therefore, it is important to incorporate the leachable metal concentration into the dietary risk assessment of Cd via consumption of pepper. Besides, it is unreasonable to apply the same evaluation models for all substances during the dietary assessment of food crops. A nationwide investigation of Cd distributions in Chinese grown pepper and the health implications based on bioaccessibility is therefore urgently needed.
In the present study, Cd levels in 647 pepper samples from different parts of China were investigated. The objectives of this study are to (1) elucidate Cd levels in Chinese pepper nationwide, (2) assess the dietary risk of Cd to local populations via pepper consumption, and (3) evaluate the health implications of Cd in Chinese pepper based on bioaccessibility. This study is critical for evaluating the health risk of Cd in Chinese pepper and sheds light on the role of bioaccessibility in the risk assessment of heavy metals.

Sample collection and preparation
A total of 647 pepper samples were collected either from the local markets or online during September and November 2019. These samples came from 21 provinces of China, including Guizhou, Sichuan, Henan, and other major pepper-producing areas (Fig. 1). Approximately 500 g of each sample was labeled and transported to the laboratory, and 50 g of each sample was weighed, rinsed 3 times with ultrapure deionized water, oven-dried, powdered with a stainless steel grinder, passed through a 0.15mm sieve, and stored in polyethylene bags before chemical analysis.

Determination of Cd in pepper samples
For Cd analysis, 0.3 g of sample was placed into a Teflon crucible with 5 mL of concentrated HNO 3 , which was then sealed in a stainless steel container and heated in the oven at 160 ℃ for 6 h. After cooling, the digestion solution was placed on a hot plate and heated at 100 ℃ for 30 min. Then, the digest was diluted to 50 mL with water, fully mixed, filtered by a 0.45-μm polyethersulfone membrane, and stored at 4 ℃ for detection. Cd in the HNO 3 digestion solution was determined by inductively coupled plasma mass spectrometry (ICP-MS, Perkin Elmer NexlON2000, USA), and all the concentrations of Cd in this study were expressed as dry weight. All the reagents used in the analysis were of guaranteed reagent grade.

Measurement of bioaccessibility of Cd in pepper samples
Bioaccessibility (BA) of Cd in pepper was evaluated in vitro using the physiologically based extraction test (PBET), which was modified based on a previously established method (Chen et al. 2020;Ruby et al. 1993). Briefly, the PBET consists of two extraction processes, gastric and intestinal digestion. For gastric phase digestion, 0.5 g of sample was mixed with 5 mL of simulated gastric solution (1.25 g/L of pepsin, 0.50 g/L of sodium malate, 0.50 g/L of sodium citrate, 420 μL/L of lactic acid, and 500 μL/L of acetic acid, pH = 1.5) in a 50-mL centrifuge tube and shaken at 100 rpm in an incubator at 37 ℃ for 1 h. For the intestinal phase, an additional 5 mL of simulated duodenal fluid (0.5 g/L of pancreatin and 1.75 g/L of bile salts, pH = 7.0) was added to the above gastric digestion fluid, shaken at 100 rpm in the incubator, and maintained at 37 ℃ for another 4 h. The mixture was centrifuged at 4000 × g for 20 min, and the supernatant was collected, filtered with 0.45-μm polyethersulfone membrane, and stored at 4 ℃ prior to analysis.
The BA of Cd in pepper was calculated as the percentage of the leachable metal concentration in PBET relative to the total metal concentration in pepper Liu et al. 2019):

Quality assurance and quality control
In order to ensure the reliability of the results, quality assurance (QA) and quality control (QC) for Cd analysis were performed with each digestion batch, including reagent blank, duplicate, and certified reference material (TMQC0006, pepper), which was purchased from Beijing Tanmo Quality Inspection Technology Co., Ltd. Calibration curve with R 2 > 0.999 was accepted to calculate the concentration. The limit of detection (LOD) was defined as 3 times the standard deviation, which was derived from 10 analyses of the blank solution. Approximately 10% of the samples were randomly selected for duplicate analysis, and the relative standard deviation (RSD) of the duplicate samples was less than 10%. The recovery rates of Cd ranged from 90 to 110% for reference material.

Dietary exposure estimate
Dietary exposure risk to Cd depends on the Cd concentration and the consumption of pepper. According to the USEPA (1997), the non-carcinogenic and carcinogenic effects of Cd via ingestion can be estimated using the following equation (Liu et al. , 2019Sawut et al. 2018): where ADD is the average daily dose (mg/kg/day); C represents the concentration of Cd in pepper (mg/kg); IngR is ingestion rate (kg/day); EF refers to exposure frequency (365 day/years) (Sawut et al. 2018); ED represents exposure duration (for children: 7 years; for adolescents: 7 years; for adults: 43 years; for seniors: 10 years) (Xia et al. 2010); BA represents the bioaccessibility of Cd (100% if bioaccessibility is not considered); BW is the average body weight (kg); (2) ADD = C × IngR × EF × ED × BA BW × AT AT is the average time (days). For non-carcinogenic effects assessment, AT is defined as ED × 365 days. For carcinogenic effects assessment, AT is defined as 70 × 365 days (Ferreira and Miguel 2005). According to the Chinese National Nutrition and Health Survey (CNNHS) conducted in 2002 (Huan et al., 2019), the body weight of whole Chinese residents in this study was divided into eight subgroups based on the respondents' age and gender as child (≤ 11 years), youngster (12-18 years), adult (18-60 years), and elder (> 60 years) for both genders. Due to China's huge population base, a survey on residents' consumption of pepper in each province was rare. Therefore, food consumption data was divided into four groups according to the residents' age (for child: 3.71 g/d; for youngster: 6.25 g/d; for adult: 9.38 g/d; for elder: 8.99 g/d) (Cheng et al. 2018). The relative parameters for calculating the average daily intake of Cd through consumption of pepper are shown in Supplementary Table S1.

Non-carcinogenic risk assessment
The non-carcinogenic risk posed by chronic exposure to Cd via pepper consumption was estimated using the target hazard quotient (HQ); the equation is as follows (USEPA 1989;Qasemi et al. 2019): where HQ is the hazard quotient, and RfD is the maximum daily permissible risk to humans through dietary exposure. The RfD value is 0.001 mg/kg bw/day for Cd in pepper (IRIS (Integrated Risk Assessment System), 2018). If the HQ value is ≤ 1, the exposed population is unlikely to experience detrimental health effects, whereas HQ value of > 1, potential non-carcinogenic effects may occur.

Carcinogenic risk assessment
The potential carcinogenic risk through pepper Cd exposure can be estimated by the following equation (Miri et al. 2017): where ILCR is the incremental lifetime cancer risk, and CSF is the carcinogenicity slope factor of Cd (mg/kg/day). In the present study, the value of CSF is 6.1 mg/kg/day, according to the Integrated Risk Information System (IRIS (Integrated Risk Assessment System), 2018).

Statistical analysis
Statistical analysis was performed using SPSS software, version 20.0 (IBM Corporation, Armonk, New York, USA).
(3) HQ = ADD RfD (4) ILCR = ADD × CSF Non-parametric tests were used to test the significant differences in Cd concentrations in pepper from different regions and with different fruit shapes. The probability level of p < 0.05 was considered to be significantly different. All figures were obtained using ArcMap 10.4 (ESRI Inc., USA) and Origin 9 software (©OriginLab Corporation).

Cd concentrations in pepper
The Cd concentrations in pepper ranged from 0.002 to 1.470 mg/kg, with an average value of 0.222 mg/kg and a median of 0.132 mg/kg ( Table 1). The concentrations of Cd showed obvious spatial variations in China; pepper with relatively high Cd levels (0.233-0.346 mg/kg) were mainly collected from southwestern China, such as Chongqing, Guizhou, Yunnan, and Sichuan provinces ( Fig. 1 and Fig. 2). According to their shapes, pepper in this study was divided into three types: bell pepper, line pepper, and cone pepper. Interestingly, Cd contents in pepper varied with the difference in shapes, cone pepper had the highest Cd contents (0.123 mg/kg), followed by bell pepper (0.098 mg/kg), and finally line pepper (0.075 mg/kg) (Fig. 3).

BA of Cd in Chinese pepper
BA was expressed as the percentage of the Cd concentration extracted by PBET to its total concentration in pepper. In the present study, the BA of Cd in pepper samples ranged from 22.68 to 72.54%, with an average value of 43.07% (Table 2).

Dietary exposure of Cd through the consumption of pepper
In this research, the main factors related to dietary exposure and health risks such as Cd concentration, consumption data, exposure frequency, body weight, age, and gender were all taken into consideration. The ADDs of Cd for different age categories and genders via consumption of pepper are presented in Supplementary Table S2 and  Table S3. For people of all age categories, females showed higher ADD values of Cd than males of the same age. In non-carcinogenic risk assessment, the trend of the ADD values for Cd in different age categories was in the order of child > elder > adult > youngster for both males and females, and the highest ADD value was observed in the female child group with an average value of 0.040 μg/kg/day (1.200 μg/ kg/month) and a median of 0.024 μg/kg/day (0.720 μg/kg/ month), which were far below the provisional tolerable monthly intake (PTMI) of 25 μg/kg/month recommended by  Table 3 and Fig. 4. The HQ values of Cd in all subgroups were all below 1, suggesting that there was a low non-carcinogenic risk from pepper consumption. The ILCR exerted by pepper Cd intake are shown in Table 4 and Fig. 5. Similar to dietary Cd intake, female showed relatively higher CR values than male for all age groups. The adult group had the highest CR values (male: 1.248E-04; female: 1.416E-04), which already exceeded the maximum acceptable level of chemical carcinogens of 10 −4 proposed by USEPA (2000).

Dietary exposure of Cd based on BA
After considering BA, the ADDs of Cd via pepper consumption decreased by 43.07%, both in non-carcinogenic and carcinogenic effects (Supplementary Table S2 and Table S3). Correspondingly, the HQ and CR values were lower than the unadjusted values (Table 3, Table 4, Fig. 4, and Fig. 5). Notably, the CR values (at 50th and 75th percentile) of both genders in the adult group reduced, and no longer exceeded the limit of 10 −4 after consideration of BA. However, the values at P90, P95, P97.5, and P99 still indicated a potential carcinogenic risk for the Chinese population with higher exposure to Cd.

Discussion
The distributions of Cd in pepper worldwide have been investigated by a few studies (Supplementary Table S4). Zhao et al. (2019) reported that the average level of Cd in pepper purchased from local markets in Guiyang, China, was 0.058 mg/kg (fresh weight). Assuming the dry-wet ratio of pepper is 0.1 (Wu et al. 2014), the Cd contents of pepper in this study ranged at ND-0.147 mg/kg with an average of 0.022 mg/kg in fresh weight, which was relatively lower than the above research, but those data may not be representative because of the small sample size (n = 6). Wang et al. (2014) also reported similar Cd levels of ND-0.126 mg/kg in pepper from Zunyi city in southwestern China, with a sample size of 43. Generally, the distribution of Cd in pepper worldwide was concentrated at ND-0.468 mg/kg, and our results were within this range. In addition, considering the sample size and the breadth of sampling locations, our results were more representative of Cd levels in Chinese pepper.
The concentrations of Cd in pepper showed large spatial variations in China, and pepper from southwest China, where are exactly the important pepper-planting bases, tend to accumulate more Cd (Fig. 2). According to the National Survey Bulletin of Soil Pollution published in 2014 (MEP (Ministry of Environmental Protection), (MLR) Ministry of Land and Resources, 2014), the Cd pollution in the soil at southwest China was more serious due to the high geological background. In addition, anthropogenic activities, such as mining development, industrial emissions, agrochemicals usage, and sewage irrigation, increased the Cd levels in the surrounding environment, which might be the main reason for the higher Cd contents in pepper in these areas (Duzgoren et al. 2006;Zhang and Reynolds 2019). The contents of Cd in different crops vary greatly due to the difference in morphology, anatomy, and physiology of each plant (Gupta et al. 2018). Cd content could reach up to 0.690 mg/kg in rice grain from the countryside area of Xiangtan county, Hunan province, southern China, which was much higher than that in pepper of this study (0.222 mg/ kg) (Hong et al. 2018). Meanwhile, the concentrations of Cd in leafy vegetables such as lettuce (0.320 mg/kg), spinach (0.500 mg/kg), and celery (0.580 mg/kg) also exceeded the Cd contents in pepper of this study (Hong et al. 2018;Zhuang 2009). However, the concentrations of Cd in pepper were much higher than that in rootstalk and legume vegetables such as taro, cassava, carrot, kidney bean, and garden pea (Antoine et al. 2017;Hao et al. 2011). Translocation and transpiration rates might play an important role in the accumulation of heavy metals in vegetables. The transfer distance of metals from root to fruit was longer in non-leafy vegetables, resulting in lower accumulation content (Gupta et al. 2018). In addition, our results showed that corn pepper exhibited a higher concentration of Cd than other shapes of pepper fruits; this might be due to the fact that most of the corn pepper came from southwest China where the Cd exposure levels were relatively high. Dietary exposure to Cd via pepper consumption was evaluated in this study. Antoine et al. (2017) reported that the ADD of Cd through the consumption of sweet pepper was 0.109 (μg/day/kg body weight), which was more than 2.8 times the maximum ADD value of our study (0.038 μg/day/ kg body weight), while the food ingestion rate (presented in means) was 6.8 times of this study (48.34 vs. 7.08 g/d). Furthermore, even the highest dietary Cd intake from pepper in the child group (1.2 μg/kg/month) had not exceeded the mean value of the general Chinese population of 15.3 μg/ kg/month, suggesting that the health risk exerted by Cd through the ingestion of pepper was at a relatively safe level because of the low consumption amount (Song et al. 2017). Both genders shared the same ingestion rate of pepper, while females generally weighed less than males; thus, females showed higher ADD values of Cd than males of the same age.
HQ and CR values were used to evaluate the non-carcinogenic and carcinogenic effects of Cd through pepper consumption. The values at P50 demonstrated the median risk exposure of pepper consumers to the Cd content distribution, and the values at P95, P97.5, and P99 demonstrated higher heavy metal exposure. In this study, the HQ values of Cd were all lower than 1, while the CR values in child (P99), adult (P75, P90, P97.5, P99), and elder (P95, P97.5, P99) subgroups exceeded the maximum acceptable value of 1 × 10 −4 , suggesting that Cd was more likely to generate cancer risk compared with other heavy metals (Liu et al. 2013). Besides, epidemiological studies have shown that low-level dietary exposure to Cd is associated with thyroid and breast cancer (Sara et al. 2019;Eunjung et al. 2021).
More importantly, pepper intake was assumed as the only pathway for Cd exposure to humans in the present study; other sources of Cd in the dietary structure, such as drinking water, grains, fruits, and vegetables, might play important roles in exerting potential health risks. Therefore, more comprehensive investigations regarding dietary structure and distributions of Cd in food are needed for potential health implications assessment.
Notably, only a part of Cd in food could be released. BA was defined as the fraction of heavy metals released from its matrix into the gastrointestinal tract after digestion (Aziz et al. 2015). Therefore, BA of heavy metals was used to estimate the maximum content that could enter the body (Lv et al. 2011), and it has been formally conducted to evaluate the exposure of metals in particulate matter (Liu et al. 2019), contaminated soil , and foodstuffs such as rice (Lv et al. 2011), vegetables (Roselli et al. 2020), and seafood (Liao et al. 2020). The BA of heavy metals could be measured by several standard methods, of which PBET has been widely used because of its reproducible, reliable, and sensitive characteristics Liao et al. 2020;Ping et al. 2018). In this study, in vitro PBET was conducted, and the BA of Cd in pepper was 43%, which is consistent with previous observations that the BA of Cd in contaminated cooked rice samples varied from 37.14 to 52.93% during the gastrointestinal phase (Lv et al. 2011). While after taking BA into consideration, the ADD values of Cd were substantially low both in the non-carcinogenic and carcinogenic risk assessments (Supplementary Tables S2 and S3). Correspondingly, HQ and CR values for the risk assessment of Cd in pepper were reduced by 43.07%, and CR values of child (P99), adult (P75), and elder (P95, P97.5, P99) groups for both genders were lower than 1 × 10 −4 , suggesting a lower exposure risk of Cd with the consideration of BA. Besides, heavy metals released into the gastrointestinal tract of humans are not all absorbed into the body. A previous study showed that the absorption rate of Cd in the intestine was around 5% (Satarug et al. 2010), and the bioavailability of heavy metals was affected by many factors such as individual lifestyle, dietary structure, heavy metal concentrations in food, and human intestinal function properties (Zhang et al. 2020). For instance, it has been reported that phytochemicals in food such as proanthocyanidins, anthocyanin, and tea polyphenols could decrease the solubility of Cd in contaminated water by 14-19%, and the potential mechanism might be the adsorption or chelation properties of these substances (Hider, 2001). Thus, dietary risk assessment based on total rather than bioaccessible contents of heavy metals may not be a reliable method and may lead to overestimation. Risk assessment based on BA might be more accurate in reflecting the toxic effects of Cd in pepper.

Conclusion
The present study determined the total and bioaccessible contents of pepper Cd from 21 provinces of China. The dietary health risks based on BA values were also evaluated. The Cd concentrations in pepper ranged from 0.002 to 1.470 mg/kg, which were generally low and safe. Meanwhile, pepper from southwest China exhibited relatively higher Cd levels. Cd exposure via pepper consumption in different age and gender groups was much lower than the PTMI of 25 μg/kg/month which was safe for human health. However, chronic low-dose exposure to Cd might cause carcinogenic risks to humans, especially adults. Interestingly, when taking metal BA into consideration, non-carcinogenic and carcinogenic risk indexes for different age and gender subgroups decreased to a certain extent compared with the risk indexes without considering BA. In particular, the carcinogenic risk of Cd in adult groups significantly decreased after incorporating BA into health risk assessment. This might be due to the antagonism or chelation of certain substances in the gastrointestinal tract. Hence, the dietary risk assessment of heavy metals in food crops cannot simply consider the total content; metal BA should be taken into consideration. Correspondingly, a more detailed and comprehensive evaluation procedure is urgently needed.