Risk assessments for the dietary intake aflatoxins in food: a systematic review (2016–2022)

)


Introduction
Aflatoxins (AFs), a subset of mycotoxins, are secondary fungal metabolites that are primarily produced by toxigenic strains of Aspergillus spp.including A. flavus and A. parasiticus (Martinez-Miranda et al., 2019).Amongst all the known types of AFs, aflatoxin B1 (AFB1), B2 (AFB2), G1 (AFG2), G2 (AFG2) are the most prevalent naturally occurring toxins found in food and feed crops (European Food Safety Authority (EFSA), 2020; Kabak, 2021).In addition, Aflatoxin M1 (AFM1) and M2 (AFM2) are the hydroxylated forms of AFB1 and AFB2, respectively, found in milk and milk products and are produced when grazing animals feed on contaminated crops (Nejad et al., 2019).Amongst these, AFB1 is the most commonly occurring and the most potent genotoxin and carcinogen (Group 1) in nature, as characterised by the International Agency for Research on Cancer (IARC, 2002).
Of all the potential chemical hazards, AFs represent one of the major food safety concerns worldwide.As per the Rapid Alert System for Food and Feed (RASFF), AFs accounted for 20.9% (10, 302) of the total RASFF notifications issued from 2002 to 2019 (European Commission (EC), 2020, Kabak, 2021).In 2020, a total of 400 mycotoxin hazards were notified of which AFs accounted for 367 (92%) notifications which is approximately 10% of the total RASFF notifications in 2020 (EC, 2021).
Agricultural commodities that are predominantly susceptible to AF contamination include cereals and their derived products, oilseeds, nuts, spices, milk and milk products (EC, 2006;EFSA, 2007;Nejad et al., 2019).AF contamination can occur pre-harvest, during harvest or post-harvest (transportation and storage) (Heshmati, Khorshidi, & Khaneghah, 2021) and is influenced by several factors favouring the growth of AF producing fungi i.e., Aspergillus spp.leading to the subsequent biosynthesis of AFs at these stages (Bhardwaj et al., 2022-a;Yu, 2012).Occurrence of AFs in food/feed crops and milk has been reported worldwide but it is majorly a problem in tropical and subtropical regions where hot and humid climate favour fungal growth and subsequent AF production.(Nugraha et al., 2018;Kabak, 2021).Moreover, AFs present a serious concern for developing countries due to the low economic development and lack of implementation of food quality control standards (Nugraha et al., 2018;Conteçotto et al., 2021).
Human exposure to AFs via contaminated food and feed crops can result in a wide range of adverse health effects like genotoxicity, hepatotoxicity, mutagenicity, teratogenicity and carcinogenicity (Kabak, 2021).Aflatoxicosis is a severe health syndrome associated with direct ingestion of AFs (Sowley, 2016) and can be classified as acute and chronic aflatoxicosis, depending upon the level and duration of exposure (Ismail et al., 2019).Acute toxicity is a major concern in developing countries with around 40 million citizens of these countries exposed to AFs, creating a huge public health burden for large parts of the population (Sowley, 2016).It is generally characterised by nausea, ataxia, loss of appetite, lethargy, liver inflammation, abdominal pain, jaundice, edema, coma, haemorrhage and even death (Ismail et al., 2019;Saha Turna & Wu, 2019).Additionally, chronic aflatoxicosis is a global health problem (Alshannaq & Yu, 2017) and is associated with the development of Hepatocellular carcinoma (HCC), the 6th most commonly occurring cancer (Forman et al., 2014;Martinez-Miranda et al., 2019) and the third leading cause of deaths due to cancer worldwide (World Health Organisation (WHO), 2008; Liu & Wu, 2010).Individuals suffering from chronic Hepatitis B (HBV) and C (HCV) infection are at higher risk of developing HCC upon AF exposure (Liu & Wu, 2010;Saha Turna & Wu, 2019).AFs have been reported to play a causative role in 4.6-28.2% of the global HCC cases via consumption of contaminated food (Liu & Wu, 2010).Moreover, AF metabolites such as AFM1 has recently been quantified to cause ~13-32 HCC cases/year/100,000 population/ng AFM1 consumed, globally (Saha Turna et al., 2022).Growth impairment is another adverse effect that has been linked with chronic dietary exposure to AFs through breastmilk, infant formula and weaning foods during the early stages of life (Ahlberg et al., 2018;Saha Turna & Wu, 2019) and is a major concern for developing or low/middle income countries where high AF prevalence rates are observed.Health burden estimates of AF attributable stunting and stunting-underweight co-occurrence in low-income African countries revealed an increase of 3-36% and 14-50%, respectively (Rasheed et al., 2021) A strong negative correlation between AF exposure and growth impairment has been reported in several studies indicating that the higher the exposure, the stronger is the growth impairment (Gong et al., 2002;Gong et al., 2003;Khlangwiset et al., 2011;Makori et al., 2019;McMillan et al., 2018;Turner et al., 2007).However, the evidence remains conflicting as some studies have also observed no significant association with the AF exposure and growth impairment (Kiarie et al., 2016;Mitchell et al., 2017;Shirima et al., 2015).Moreover, the mechanism of action involving AF-associated growth impairment is not yet understood and has been investigated widely (Zhou et al., 2019;Voth-Gaeddert et al., 2019;Wacoo et al., 2020;Ismail et al., 2021).
AFs are naturally occurring compounds which are hard to completely eliminate even after food processing due to their high chemical and thermal stability (Nugraha et al., 2018).Consequently, it makes it extremely difficult to achieve zero exposure (Zhang et al., 2020).Monitoring dietary exposure by conducting risk assessments is a vital step to manage and reduce the potential risk associated with AFs consumption, thereby, ensuring food and consumer safety (Kabak, 2021, Zhang et al., 2020).Risk assessment is an integral part of risk analysis process and a concentual framework comprising of (i) hazard identification, (ii) hazard characterisation, (iii) exposure assessment and (iv) risk characterisation (International Programme on Chemical Safety (IPCS), 2004;IPCS, 2009;EFSA, 2012;Kabak, 2021).It provides scientific advice on the probable adverse health effects associated with exposure to certain food contaminants, including AFs (EFSA, 2012;IPCS, 2009;Kabak, 2021).
In 2021, Mars, Incorporated convened the Food Safety Coalition of like-minded individuals from industry, academia, and international organizations to drive food safety insights and best practices at pace, starting with aflatoxins due to serious health threat they pose.Work is being progressed in four areas: sampling and testing, risk assessment and communication, prediction, and risk communication.This publication forms part of the work focused on risk assessment and communication which was further divided into Part A: Risk assessment and Part B: Risk communication and was completed in collaboration with Queen's University Belfast, Northern Ireland, UK.The present study covers Part A: Risk assessment, the objective of which is to undertake a systematic review of the risk assessments for AFs published in the last six years (2016-2022) with an ultimate aim of identifying the major dietary sources of AFs and the population at risk of developing AF-related health outcomes.

Methodology
The literature search for the systematic review of risk assessments was performed in accordance with the Campbell methods guide (Kugley et al., 2017) and the PRISMA guidelines (Moher et al., 2010).Relevant articles were first screened by title and abstract search followed by full text screening.All the citations were exported to endnote and the duplicates were removed prior to full text screening.The articles were double screened by two investigators and any discrepancies were resolved by a third researcher.
Data search for RASFF incidents was performed using an online risk prediction platform.All the information regarding RASFF incidents was first exported into excel and processed by applying appropriate filters.

Research aim
The present review aimed to undertake a commodity wise systematic review of the risk assessments for aflatoxins published in the last six years i.e., 2016 __ 2022 in order based on the most to least no. of EU RASFF incidents recorded for AFs.

Search strategy
For the systematic review, relevant risk assessments published between 1 st Jan 2016 to 24 th Jan 2022 were collected using peer-reviewed databases including Scopus, Web of Science and PubMed.The keywords and search strings applied to the mentioned electronic literature database included: (Risk assessment) AND (Aflatoxins OR AFB1 OR AFB2 OR AFG1 OR AFG2 OR AFM1 OR AFM2).
Data for the EU RASFF incidents recorded between 4 th January 2016 and 2 nd Mar 2022 was retrieved from the food risk prediction platform, Agroknow (FOODAKAI, 2021).

Inclusion and exclusion criteria
For literature search, after the title and abstract screening, eligible articles were downloaded based on the following inclusion criteria: (1) Full text availability (2) English language (3) Research or review articles published between 1 st Jan 2016 to 24 th Jan 2022 (4) Studies conducting risk characterisation in addition to dietary exposure assessments.Additionally, there was no limitation on geographical locations, and study design.
For RASFF incidents, data was included for all the AFs food safety incidents in food recorded between the studied time period i.e. 2016-2022.Any other RASFF notifications such as food fraud incidents were excluded from the study.

Results
Using the described search strategy, 1242 references were identified for systematic review and subjected to title and abstract screening.864 irrelevant articles were removed and the remaining i.e., 378 were subjected for full text screening.196 duplicates were removed prior to full text screening.Overall, 91 articles passed the inclusion criteria and were included in the study.A summary of the screening process and the results are described in Fig. 1.
For RASFF data, a total of 2812 AF notifications were found to be issued between 4 th Jan 2016 and 3 rd Mar 2022 and has been presented for different food categories (Fig. 2

Discussion
Dietary exposure assessment is one of the essential elements of the risk assessment process that provides scientific evidence for the estimated dietary intake (EDI) of contaminant based on the amount of contaminant present in a particular food commodity and the consumption pattern for that food commodity.To determine EDI for AFs, accurate information regarding their concentration and food consumption must be obtained (Udovicki et al., 2021).The obtained conditions of exposure i.e., the EDI values could be used to characterise the risk (Martinez-Miranda et al., 2019) which further helps to build strategies to reduce potential risks associated with AF exposure (Heshmati et al., 2017).
The potential human health risks posed by the dietary intake of AFs have been assessed by several scientific bodies including the Joint Food and Agriculture Organization/World Health Organization (FAO/WHO) Expert Committee on Food Additives (JECFA) (FAO/ WHO, 2007WHO, , 2016WHO, , 2018) ) and EFSA Panel on Contaminants in the Food Chain (CONTAM) (EFSA et al., 2020).Scientific opinions from the risk assessments provide a relevant advisory information for the risk managers in making decisions (IPCS, 2009) regarding various aspects such as establishment of Maximum limits (MLs) which rely on the results of risk assessments to identify the food commodities posing significant risk due to significant exposure to the contaminant, including AFs.To try and reduce the risk of overexposure to AFs and their associated detrimental health effects, maximum limits (MLs) to their occurrence in the food commodities that are susceptible to AF contamination and pose significant health risk have been imposed by the various regulatory bodies in many parts of the world (Table .1). Results of risk assessments also help risk managers to design effective risk communication strategy for which they need to know what is posing most risk and who all are at high risk so that then they can target their audiences and develop messages tailored to their needs.For carcinogenic and genotoxic AFs, risk assessment studies have been carried out using two main risk characterisation approaches including (1) estimation of the liver cancer risk, based on Hepatitis B (HBsAg+) incidence and carcinogenic potency of AFs (FAO/WHO, 1999) and (2) estimation of Margin of Exposures (MOEs).For AFs, MOE of 10,000 or higher would be of low concern from public point of view (EFSA, 2005).An additional risk characterisation based on hazard index (HI) has also been employed to study the carcinogenic and non-carcinogenic effects of AFM1 because of low cancer potency.HI value greater than 1 indicates a potential health risk of liver cancer (Kuiper-Goodman, 1990;Bahrami et al., 2016).
Extensive research has been performed to evaluate the dietary exposures of AFs facilitating risk characterisation.In this study, a

Fig. 2(c).
EU RASFF incidents for aflatoxins per product from 4 th January 2016 and 2 nd March 2022.Data retrieved from FOODAKAI, the food risk prediction platform, Agroknow (FOODAKAI, 2021).commodity wise systematic review of the risk assessments published in the last 6 years (1 st Jan 2016 to 24 th Jan 2022) has been presented in order based on the most to least no. of RASFF incidents for AFs held between 4 th Jan 2016 and 2 nd Mar 2022 (Fig, 2 (a), (b) and (c); FOO-DAKAI, 2021).A summary of the results of the reviewed studies has been presented in Table .2.

Nuts, Nut products and seeds
Nuts and seeds are an important part of the human diet due to their nutritive value and nowadays are considered as a healthy snacking option.In addition to direct consumption, nut and seeds are used for oil extraction intended for cooking and other industrial uses such as soap and cream production (Adetunji et al., 2018).High nutrient content of nuts and oilseeds is one of the factors that favour fungal growth, making them highly prone to AF contamination (Bhardwaj et al., 2022-a;Liu et al., 2016).Nuts, nut products and oilseeds held the highest number of AF RASFF notifications (74%) recorded between 4 th January 2016 and 2 nd March 2022 (Fig. 2(a); FOODAKAI, 2021).The risk posed by the dietary exposure to AFs through consumption of Nuts and oilseeds has been extensively assessed.Kooprasertying et al. (2016) conducted an exposure assessment of AFs in Thai peanuts and found that exposures to AFB1 and total AFs from the mean consumption of peanuts were highest for children between the age of 6 and 9 years (1.56 and 1.89 ng/kg bw/day), followed by teenagers (0.68 and 0.82 ng/kg bw/day) and adults (0.60 and 0.72 ng/kg bw/day), respectively.The average EDI at the same consumption level was calculated as 0.66 and 0.80 ng/kg bw/day for AFB1 and total AFs.While the potency of AF associated liver cancer was estimated as 0.0245 cancers/year/100,000 population at the mean consumption which was 4 times higher for 97.5 percentile peanut consumption.The rate of the cancer incidents for Thai males and females was estimated 36.9 and 15.2 cancers/year/100,000 population respectively.A similar study conducted in Indonesia and Taiwan also revealed a potential health risk arising from peanut consumption as the MOEs were lower than 10,000.
Moreover, some MOE values were even below 1000 for the Indonesian population while half of the Taiwanese population couldn't reach the threshold of MOE i.e.,10,000.The estimated risk of liver cancer was calculated in the range of 0.0007-0.2713cancers/year/100,000 population for Taiwanese while for Indonesian population a cancer risk of 0.1-35 cancers/75 years/100,000 population was recorded (Nugraha et al., 2018;Wang, Lien, & Ling, 2018).
Another study conducted to assess the AF exposure risk imposed by the consumption of peanuts and peanut products imported to Taiwan revealed that the MOE levels for imported samples from China, Indonesia, Thailand, USA and the Philippines were above 10,000 suggesting an absence of potential health concern.However, samples from Vietnam resulted in the MOEs lower than 10,000 and thereby indicating the presence of potential public health risk (Lien et al., 2019).As reported by Do et al. (2020), Peanuts and sesame in Northern Vietnam were found to be the major source of AFB1 in adults, following maize and rice, with MOEs recorded between 0.55 and 7763, indicating a potentially high risk of HCC.Range of estimated liver cancer risk lied between 0.21 and 21.09 cancers/year/100,000 population of all the age groups.A risk assessment conducted by EFSA et al. (2020) also revealed that the food category 'legumes, nuts and oilseeds' was the second highest contributor after cereals to the overall lower bound (LB) mean chronic dietary exposure to AFB1 in the EU population, with peanuts being the major contributor (24% contribution in adults) (EFSA, 2020).An overall mean MOE to AFB1 of 58-5000 was estimated from the mean dietary exposures of all food categories.
Risk characterisation for adult consumers of groundnuts from domestic markets of Nigeria resulted in the national mean MOEs of 1665 and 908 for AFB1 and total AFs, respectively, concluding that consumers were at a high risk of liver cancer and other AF related health outcomes (Oyedele et al., 2017).Another risk assessment conducted in Nigeria for the consumption of AFs in groundnuts and roasted cashews recorded much lower MOEs of 6.10 and 1000 respectively, again indicating a high potential risk for the Nigerian population especially due to the AF contaminated groundnut.Moreover, the risk of liver cancer for the     population consuming groundnuts and cashew nuts was estimated at 1.38 and 0.01 cancers/year/100,000 population, respectively (Adetunji et al., 2018).A recent study also observed similar results and reported the highest risk of liver cancer i.e. 41-247 cancers/year/100,000 population in Nigeria from the consumption of AFB1-contaminated peanuts.However, children who consumed all the foods (Peanuts, maize, rice and sorghum) were at the highest risk due to the lowest estimated MOE to AFs i.e., 0.026-2.51(Ezekiel et al., 2021).A similar study conducted to estimate the burden of AF induced HCC amongst adult population of Malawi through AF contaminated groundnut consumption recorded a mean exposure to AFB1 of 28 ± 65 ng/kg bw/day.While the risk of HCC was calculated as 1.26 ± 2.72 cancers/year/100,000 population (Matumba et al., 2019).Kortei et al. (2021a) conducted a risk assessment for different age groups for AF exposure through randomly collected groundnuts and its products from the local markets of Ghana and concluded that the consumption of groundnuts posed a potential health risk for all age groups.The recorded MOEs and population risk for total AFs were in the range of 1053-4598 and 1.5 × 10 − 3 -7.9× 10 − 4 , respectively and the same for AFB1 were 1333-5882 and 1.2 × 10 − 3 -6.3× 10 − 4 , respectively.Taghizadeh, Rezaee, Badibostan, and Karimi (2020) assessed the risk posed by the dietary exposure to AFB1 through walnuts in the Iranian population and suggested no potential risk for the Iranians for the 95th percentile consumption.However, for 50th percentile consumption, the calculated MOEs were below 10,000 meaning a potential health risk for the Iranian population.Moreover, the annual cancer risk due to the consumption of walnuts was recorded as of 1.0 × 10 − 4 and 3.4 × 10 − 6 in 10,000 population for HBV positive and HBV negative individuals, respectively.On the other hand, consumption of sesame seeds and sesame-based products did not pose any remarkable risk of cancer for the Iranian adults (Heshmati, Khorshidi, & Khaneghah, 2021).A recent risk assessment for the Oilseeds, including corn seeds, sunflower seeds, soyabean seeds, consumed in Egypt revealed that soyabean seeds were AF-free and based on the consumption rate of sunflower seed samples, the calculated MOEs to AFs were greater than 10,000, indicating low concern for public health.While MOEs for corn seed samples were <10, 000 which represented a potential health concern for consumers (Kholif et al., 2021).
Consumption of nuts in China was also considered as a potential health risk for adults as the risk assessment conducted by Wang, Lien, and Ling (2018) resulted in some MOE values for AFs lower than 10,000.Similar concerns were raised in a recent risk assessment by Qin et al. (2021) for peanut samples collected from different provinces of China.A mean AFT exposure of 1.546-1.672ng/kg bw/day (LB-UB) was observed resulting in lower MOEs in the range of 88-96 (UB-LB), representing a significant health risk.The risk of liver cancer was estimated as 0.055-0.060cases/100,000 persons/year (LB-UB).Guangdong, Fujian and Jiangxi provinces were at a greater risk and the highest exposures levels were observed in children aged between 2 and 6 years.In a risk assessment comprising foods commodities commercialised in Turkey, the highest risk of liver cancer among Turkish adults was found to be associated with the exposure to AFB1 through pistachio consumption (44%) followed by maize (16%), groundnuts (14%), chilli (10.%), walnuts (5%) and hazelnuts (4.0%) etc., with mean MOE values ranging from 860 to 995, indicating a potential health concern.The mean risk of HCC was estimated as 0.007 to 0.008 cancers/year/100, 000 adult population (Kabak, 2021).
Overall, dietary exposure to AFs through nuts consumption present a greater health concern as compared to oilseeds.However, there is much less evidence available for the risk characterised through oilseeds consumption.Therefore, more studies should be conducted to know the level of risk presented by this food category.Moreover, studies should assess the risk associated with AF contaminated melon and watermelon seeds consumption as they accounted for the highest no. of RASFFs for AFs (0.7% each) in oilseeds category between 2016 and 2022 (Fig. 2(c)).Age wise, all the age groups have been found to be at a potential health risk through consumption of nuts and oilseeds.However, a decreasing trend in exposures levels to AFs have been observed with the increase in age.Despite the number of assessments conducted for this particular food categories, there is a need to conduct further risk assessments for the populations of the countries associated with greatest number of AFs RASFF food safety incidents, presented in Fig. 2(b) (FOODAKAI, 2021).

Fruits and vegetables
A dietary exposure risk assessment of dried fruits, including dried mulberries, dates, figs, and apricots, from Iranian markets suggested a high risk for the consumers from the consumption of AF contaminated dried fruits, especially dried dates due to high exposure and the consequent lower MOEs.The mean MOEs to AFB1 from the intake of dried mulberries, dates, figs and apricots were lower than the threshold (>10,000) and estimated as 4,250, 1,417, 4250 and 2,833, respectively (Heshmati et al., 2017).Similarly, consumption of dried fruits including dried jujubes and figs were estimated to pose a potential threat to the Chinese population as the calculated MOEs to AFB1 from consumption data were <10,000 i.e., 8182 and 3028 for males and 6930 and 2549 for females, respectively.However, consumption of other dried fruits such as raisins and dried longans did not pose any potential threat (Wang, Lien, & Ling, 2018).
Even though fruits and vegetables held the second highest numbers of RASFF incidents (13%) (Fig. 2(a)) for AFs in the 6 years (2016-2022) (FOODAKAI, 2021), less evidence regarding the MOEs to AFs from fruits and vegetable consumption has been reported.Therefore, more risk assessments are required to be conducted to get a wider view of the risk associated from the sole consumption of this commodity.However, from the reported risk assessments and no. of RASFF incidents per product (Fig. 2(c)), fruits (especially dried fruits) seemed to have been more relevant than vegetables for AF exposure and public health.Additionally, risk assessments should be conducted for the dried figs as they held highest AFs RAFFs incidents (11.5%) in fruits and vegetables category (Fig. 2(c)).

Herbs and spices
Herbs and spices are essential ingredients of many cuisines worldwide, are highly prone to aflatoxin contamination mainly during drying and storage stages.They are mostly eaten raw or added directly to foods without any further processing and hence pose a risk to human health through AF exposure.A mean exposure to total AFs of 0.72 ng/kg bw/ day (0.31-1.30ng/kg bw/day) and 1.59 ng/kg bw/day (0.66-3.29 ng/ kg bw/day) through the intake of branded and non-branded spice samples was recorded in a risk assessment conducted amongst different consumer groups in Pakistan.The highest exposure of (3.29 ng/kg bw/ day) was observed for the female consumers of age >24 years through the consumption of non-branded spices, while the lowest exposure (0.31 ng/kg bw/day) was observed through the intake of branded spices by the male group of age 9-14 years.MOEs for all the consumer groups were below 10,000 i.e., 52-548, indicating a potential risk of developing AF related adverse health effects including liver diseases, growth and developmental disorders, neurological disorders and immune system disorders.Moreover, females (>24 years) were the most vulnerable group, and the males (9-14 years) were the safest amongst all based on the calculated MOEs (Akhtar et al., 2020).The consumption of eastern herbal medicines samples from Southern Punjab, Pakistan also presented a potential health risk for the consumers as the MOE values and the population liver cancer risk was estimated in the range of 100-29379 and 1 × 10 − 4 to 3.25 × 10 − 2 cancers/75 years/100,000 population, respectively.Moreover, children were at a higher risk than adults due to higher exposures corresponding to their lower body weight (Javed et al., 2021).
A risk assessment in Qatar indicated a high risk of AFB1 exposure, especially from the consumption of Nuts and spices, for men and women of ages 18-64 years with alarming MOEs to AFB1 from all food categories ranging between 3 and 15 (Al Jabir et al., 2019).In a similar study, Yapo et al. (2020) assessed the health risks posed by the dietary intake of AF detected in Kankankan spice samples, a famous roasted meat seasoning of the Côte d'Ivoire.As per the results, all the studied groups i.e., adolescents (13-19 years) and adults (20-above years) were at high risk of developing potential health concerns as the recorded MOEs to total AFs (8 and 13, respectively) were far below 10,000.However, adolescents were at a greater risk.Recently, a risk assessment for the Lebanese population recorded an overall mean exposure to AFB1 of 1.55 ng/kg-bw/day from the consumption of 27 spices commonly used in 6 Lebanese dishes.The estimated MOEs to AFB1 for the positive samples (n = 8) were in the range of 108-4444 indicating a potential health for the consumers.However, low sample size for each spice type and incorporation of only positive samples for MOE estimation could have resulted in the overestimation of the risk.Therefore, the need to conduct a more precise risk assessment was concluded in the study (Al Ayoubi et al., 2021).
Despite the fewer number of assessments being published over the studied years, a significant potential health risk including the additional risk of HCC induced by AF dietary intake through spices and herbs has been presented in the reviewed literature.However, it is to be noted that herbs and spice category held the third highest number of AF incidents i. e. 8% (Fig. 2(a)) reported by RASFF (FOODAKAI, 2021) and thereby presents a major food safety concern for consumers.Also, spices represented more relevance than herbs in terms of AF exposure as seen from the risk assessments and the RASFF data per product (Fig. 2(c)).Therefore, it is imperative to estimate the health risks for populations consuming spices in countries with the greatest number of AF incidents (Fig. 2(b)).Moreover, risk assessments should be conducted for chillies and nutmeg as they are the most common spices to be contaminated with by AFs, as indicated with RASFF notifications of 3.73% and 1.49% in this food category (Fig. 2(c)).

Cereals and cereal products
As with nuts, cereal and cereal products are also susceptible to AF contamination because of their high nutritional composition (Bhardwaj et al., 2022;Liu et al., 2016).It has been estimated that World's 60-80% of the cereal crops are contaminated with mycotoxin, including AFs (Eskola et al., 2020).From the food safety point of view, it is the fourth most important food category with 3% of AFs RASFF incidents (Fig. 2 (a)) reported between 2016 and 2022 (FOODAKAI, 2021).Being key staples of the human diet, cereal and cereal-based products can serve as a major source of human exposure to AFs.Therefore, continuous monitoring and dietary risk assessment of AF in cereals is important.
A risk assessment of AFB1 exposure to the Brazilian population through the consumption of pasta and bakery products revealed that pasta was the most important contributor to AFB1 exposure with an average daily intake of 3.5 ng/kg bw/day followed by bread 2.8 ng/kg bw/day.The population of Southern Brazil was at high risk of AFB1 exposure.Moreover, exposure for males was higher than for females.Also, teenagers and adults were at a higher exposure risk when compared with elderly individuals.No significant exposure to AFB1 was observed through biscuit and cake consumption.The total estimated MOE to AFB1 was 25 meaning that the intake of AFB1 through pasta and bakery products posed a potential health risk for the consumers (Bol et al., 2016).Similarly, in a total diet study, bread (22-26%) was found to be the major source of AFs for children and adults in the Dutch population leading to a much lower MOE than 10,000 (Sprong et al., 2016).A biomarker study conducted in Brazil to assess the risk posed by AFs occurring in cereals and cereal products also suggested a high potential risk for the consumers as the MOEs values for AFs obtained from food (0.007-0.019) and urine data (258 and 459 in the first and second samplings, respectively) were lower than 10,000 (Franco et al., 2019).Similarly, a recent exposure assessment and risk characterisation conducted in Brazil revealed a potential health risk for lactating women through dietary intake of AFs present in cereals and cereal based products.As per the results, corn products were the major contributors to AF exposure with a mean EDI of 119 ± 0.193 μg/kg bw/day followed by rice (0.004 ± 0.021 μg/kg bw/day), bean products (0.026 ± 0.034 μg/kg bw/day) and wheat-based products (0.045 ± 0.009 μg/kg bw/day).Consequently, the estimated MOE to AFs was far less than 10, 000 i.e., 4, indicating a non-tolerable cancer risk for lactating women (Coppa et al., 2020).
In Northern Vietnam, mean AFB1 exposure from rice and maize consumption was calculated as 21.7 ng/kg bw/day and 33.7 ng/kg bw/ day for adults and children, respectively, with an average risk of 1.5 cancers/year/100,000 population and 2.3 cancers/year/100,000 population, respectively (Huong, Tuyen, Do, et al., 2016).A total diet study suggested rice and rice products as the major contributor towards AFB1 exposure (22.2 ng/kg bw/day) and consequently associated with the highest liver cancer risk of 1.5 cancers/year/100,000 population for the Vietnamese.The total dietary exposure to AFB1 from all food groups and associated risk of HCC was estimated as 39.4 ng/kg bw/day and 2.7 cancers/year/100,000 population (Huong, Tuyen, Tuan, et al., 2016).Another total diet risk assessment in Vietnam estimated an increased average risk of liver cancer from all food groups (12.1 cancers/year/100,000 population) with rice product group again being main source of AFB1 contributing to the highest EDI of 52.2 ng/kg bw/day and cancer risk of 5.3 cancers/year/100,000 population (Huong et al., 2019).A recent dietary exposure assessment and health risk characterisation of AFB1 indicated a high risk based on resulting MOEs ranging from 1 to 100 and 24-146 from maize and rice consumption, the two major sources of AFB1 in Northern Vietnam, respectively (Do et al., 2020).Similarly, the highest exposure levels (17.37 μg/kg.bw/day) in Bangladesh were also observed from the rice consumption while the mean risk of HCC from the consumption of all food commodities was calculated as 1311 cancers/year/100,000 population (Saha Turna & Wu, 2019).In the risk assessment by Panrapee et al. (2016), the mean exposure to AFB1 in Thailand through the consumption of rice was estimated as 0.80 μg/kg bw/day and 0.12 μg/kg bw/day in period I (June/July 2012) and period II (December 2012/January 2013), respectively.The highest exposure values were recorded for age group 19-35, i.e., 0.78 and 2.21 μg/kg bw/day, for low and high consumption, respectively.The AFB1 exposure values from the consumption of coloured rice ranged between 1.42 and 1.47 μg/kg bw/day and that for brown rice was recorded as 1.23-1.27μg/kg bw/day.The risk of liver cancer for the low consumption of brown and coloured rice for all age groups was estimated at 0.11 and 0.013 cancers/year/100,000 population.While the corresponding values for high consumption were 3 times higher (0.033 and 0.039 cancers/year/100,000 population, respectively).AFB1 exposure through the consumption of rice was also considered to be a potential health risk for the children and adults living in the Punjab region of Pakistan.The MOEs to AFB1 were in the range of 10-69 for adults while that for children ranged between 10 and 62.A mean cancer risk of 0.70 and 0.071 cancers/year/100,000 population was estimated for adults and children from the southern Punjab region while the values for northern region were 0.122 (adults) and 0.127 (children) cancers/year/100,000 population (Majeed et al., 2018).Recently, Taghizadeh, Rezaee, Badiebostan, et al. (2020) recorded a MOE to AFB1 for Iranians through rice consumption as 231, representing a potential health risk of cancer which was calculated as 78.96 and 2.63 cancers/year/100,000 population for HBsAg+ and HBsAg-individuals respectively.AFB1 exposure due to Imported rice consumption presented a potential health concern for adults and children in Tehran, Iran (MOE<10,000) with an estimated mean cancer risk of 0.27 and 0.64 cancers/year/100,000 population, respectively (Eslamizad et al., 2021).The occurrence of AFs in cereal-based baby foods was also considered as a health concern in Iran, especially for the babies aged 6-12 months with the highest EDI of total AFs (8.55 ng/kg bw/day) and AFB1 (5.81 ng/kg bw/day).However, the MOEs were lower than 10,000 for all the age groups (Bashiry, Yazdanpanah, et al., 2021).In a multi-mycotoxin study, AF risk from the consumption of cereal-and nut-based complementary foods posed a significant risk to infants and young children in Nigeria as well because the MOEs to total AFs and AFB1 were far lower than 10,000 and ranged between 0-7 and 0-70, respectively (Ojuri et al., 2018).Despite the low prevalence rate of AFs in infant total diet, the French infant population was found to be at a considerable risk as the MOEs from Upper bound (UB) exposure estimates were much lower than 10,000 and ranged between 38 and 66 for all the infants (Vin et al., 2020).Moreover, a global systematic review revealed that approximately 75% of the world's baby's population is at a risk of developing AF related adverse health effects due to the consumption of cereal-based baby food products (Bashiry, Javanmardi, et al., 2021).
Risk assessment for AFs amongst different Nigerian age groups consuming stored maize revealed highest national mean exposure to AFs among infants (1909 ng/kg bw/day) followed by children (764 ng/kg bw/day) and adults (318 ng/kg bw/day) at mean consumption of AFs.Corresponding to these, the mean MOEs obtained were 0.12, 0.3 and 0.70 for infants, children, and adults, respectively indicating a potential public health risk.The estimated national mean risk of developing primary liver cancer was recorded as 153, 61 and 26 cancers/year/100,000 population for infants, children and adults, respectively (Adetunji et al., 2017).Consumption of stored maize was also considered to pose a potential risk to Chinese consumers living in Shandong province.Children (1.27 and 7.62 ng/kg bw/day) were at a higher risk of exposure to AFB1 than adults (0.61 and 6.27 ng/kg bw/day) and the risk for village adults (1.13 and 8.18 ng/kg bw/day) was higher than city adults (0.27 and 2.58 ng/kg bw/day) at mean and high consumption level, respectively.The mean MOEs were in the order: children > adults, and village adults > city adults and ranged between 97 and 279 indicating a high risk.The estimated liver cancer risk of 0.038 and 0.23 cancers/year/100,000 population and 0.018 and 0.19 cancers/year/100,000 population was reported for children and adults at mean and high consumption level, respectively, which is lower than China's current liver cancer incidence i.e., 24.6 cases per 100,000 per year.Therefore, a low health risk was concluded (Dong et al., 2020).Similarly, in spite of the lower MOEs (<100), a total diet risk assessment by Zhang et al. (2020) suggested a low risk of liver cancer for the population of Guangzhou, China as the estimated risk of liver cancer (0.0264 cancers/year/100,000 population) was lower than China's liver cancer incidence i.e. 24.6 cancers/year/100,000 population, indicating that the consumption of AF contaminated foods does not account for the occurrence of liver cancer in Guangzhou, China.
Highest dietary exposures to total AFs (5593 ng/kg bw/day) and AFB1 (2797 ng/kg bw/day) were also observed amongst the maize consumers in Zimbabwe with MOE of <5 and cancer risk of 9.2 cancers/ year/100,000 population, posing a higher health risk to the population as compared to other small cereal grains including sorghum, finger millet and pearl millet (Akello et al., 2021).The estimated exposure from the mean consumption of nixtamalized maize (cooked and soaked maize in calcium hydroxide prior to milling) in Mexico was obtained in the range of 0.2-8.5 ng/kg bw/day.The values for 95th Percentile maize consumption ranged between 3.3 and 11.7 ng/kg bw/day.The MOEs of 20-257 and 15-50 were recorded for mean and 95th percentile consumption, respectively.The estimated increase in cancer cases was recorded as 9-320 and 43-439 cases/106 individuals/lifetime of 75 years for mean and 95th percentile consumption, respectively (Gilbert Sandoval et al., 2019).In another study, dietary intake of AFs through popcorn marketed in Veracruz, Mexico revealed that 10% and 52% of the consumers are at risk for AFB1 and total aflatoxin exposure, respectively.The average risk of liver cancer amongst women was estimated as 0.993 cancers/year/100,000 population while that for men was 0.137 cancers/year/100,000 population.It was concluded that men under the age of 18 years were at the highest risk (Morales-Moo et al., 2020).
In a multi-mycotoxin study, Infants in Tanzania, especially in Kilosa and Hanang Zones, were also exposed to a potential health risk arising from the consumption of maize-based products.The mean exposure was recorded as 133 ng/kg bw/day which consequently resulted in a lower mean MOE value of 1.3 (Kamala et al., 2017).Another study conducted in Tanzania suggested a potential high risk for adolescents in boarding schools as the MOEs were lower than 10,000 and ranged between 2 and 8500.Maize flour was the main contributor to AFs and AFB1 exposure accounting for the highest dietary intake of AFs (0.70-973.45 ng/kg bw/day) and AFB1 (0.05-81.06 ng/kg/bw/day) (Nicholaus, Martin, Matemu, Kimiywe, & Kassim, 2021).A recent biomarker study estimated the population risk of AF-induced liver cancer in Tanzania as 2.95 cases per 100,000 people, based on the epidemiology data, HBsAg + incidence and population size in 2016.The estimated risk varied across regions with the highest risk (6.79 cases per 100,000 people) estimated for Tabora region which was approximately twice the nationwide risk.Moreover, 1480 additional cases of AF-induced HCC were estimated in Tanzania in 2016 leading to a total of 56,248 disability adjusted life years (DALYs) (112 DALYs per 100,000 persons), indicating a major health concern for the population (Kimanya et al., 2021).
Exposure assessment and risk characterization for dietary intake of AFs through the consumption of maize-based products in adult populations of Serbia, Croatia and Greece resulted in EDI's ranging between 0.44 ng/kg bw/day and 5.59 ng/kg bw/day and the corresponding MOEs were in the range of 30-389 indicating a potential health risk for all three countries.The estimated risk of HCC, based on HBsAg + prevalence, was recorded as 0.075-0.098,0.006-0.008and 0.020-0.026cancers/year/100,000 population for Serbia, Croatia and Greece, respectively (Udovicki, Djekic, Kljusuric, et al., 2019).Recently, results of a total diet risk assessment in Serbia identified maize and maize products as the main contributor to overall exposure to AFB1.Overall, the child population was at a greater risk of HCC through consumption of all food categories with a mean MOE to AFB1 ranging from 215 to 155, followed by adolescents (303-211), female adults (303-223) and male adults (425-304) from estimations of LB-UB scenarios and food frequency questionnaires.The corresponding HCC risk, based on HBsAg + incidence rate, ranged from 0.011 to 0.022, 0.008-0.018,0.007-0.013and 0.005-0.012cancers/year/100,000 population for children, adolescents, adult females, and adult males, respectively (Udovicki et al., 2021).
According to a recent risk assessment amongst the rural population of Togo, the dietary intake of AFB1 through maize dough accounted for the highest mean EDI's in children (72.3-72.9ng/kg bw/day to 700.8-701.2ng/kg bw/day) followed by women (61.9-62.4 to 463.1-463.4ng/kg bw/day) and men (45.4-45.8 to 452.8-453.1 ng/kg bw/day).The excess risk of AFB1 induced liver cancer was assessed to be 58 cancers/year/100,000 in the child population and 38.23 cancers/ year/100,000 in the adult population (Hanvi et al., 2021).Similarly, exposure to AFB1 (0.004 μg/kg bw/day) due to consumption of wheat flour in Iran also posed a health risk to consumers as the recorded MOE (109) was below the recommended limit of 10,000.The risk of liver cancer for the Iranian population was estimated as 0.06 cancers/year/100,000 population (Heshmati, Nejad, & Mehri, 2021).Risk arising from wheat flour was also proposed by Jahanbakhsh et al. (2021) based on the 95th percentile MOE of 1072 for children which was 6 times higher than that of adults i.e., 647 meaning that adult population is at greater risk.Dietary exposure assessment to AFB1 from the consumption of corn arepas in Colombia suggested a potential health risk for children (4-8 years) with a mean exposure and MOE value of 10.0 ng/kg bw/day and 259, respectively (Blanco-Lizarazo et al., 2019).Another assessment conducted by Martinez-Miranda et al. (2019) also revealed a potential health risk arising from the consumption of corn arepas and rice for Colombian consumers.The EDI's from corn arepas and rice consumption were 0.732 ng/kg bw/day and 3.093 ng/kg bw/day and consequent MOEs were 278 and 61, respectively.The risk of liver cancer due to arepa and rice consumption was estimated at 0.10 and 0.044 cases/year/100,000 individuals, respectively.
Grains and grain-based products, especially corn, breads, rolls and fine bakery products, were found to be the major source of AFB1 for the EU population with a LB mean contribution of 38% for adults, 50% for elders, and 75% of toddlers.Estimated MOEs to AFB1 from the mean consumption of all food categories ranged from 58 to 5,000, indicating a potential health risk (EFSA, 2020).A risk assessment conducted for the Ghanaian population suggested no significant health risk from the consumption of maize, rice, cereal-based products and pasta products based on the hazard index (HI) values as they were less than one (Kortei et al., 2019(Kortei et al., , 2021b)).Conversely, MOE approach showed a significant risk for the Ghanaian population as the values were below 10,000 (Kortei et al., 2021b).Another risk assessment conducted based on the MOE approach revealed a significant health risk due to the consumption of maize to all the age groups residing in different ecological zones of Ghana as the MOE's were lower than 10,000 i.e., 0.4-7.Moreover, a cancer risk of 1.62-37.15cancers/year/100,000 population was also estimated (Kortei et al., 2021c).Risk assessment of AFs occurring in homemade soybean paste in South Korea resulted in the mean EDI of 0.1012-0.1080ng/kg bw/day and consequently lower MOEs were observed ranging between 3705 and 3954, indicating a significant public health concern (Lee et al., 2022).
Taken together, the results of the studies indicated that AF exposure through cereals and their derived products pose a potential health risk for all the age groups.However, the risk tends to decrease with the increase in the age, meaning that the infant and child population is at a greater risk than adults due to their high cereal intake and lower body weight.Commodity wise, maize and maize-based products have been identified as the major contributor followed by rice and rice products.Development and implementation of effective pre-and post-harvest control measures can help in preventing the occurrence of AFs in cereal products, especially Maize and rice and thereby reducing the risk of overexposure to AFs via ingestion of infected cereals grains.

Milk and milk products
AFM1, the hydroxylated form of AFB1, is the most commonly occurring aflatoxin in milk (animal and human breast milk) and variety of milk products including cheese, yoghurt, infant formulae, milk powder, flavoured milk and other traditional dairy products.Due to its thermal stability, it has been detected in pasteurised, Ultra-high temperature (UHT) treated and even sterilised milk (Rahmani et al., 2018).Given the health risk associated with AFM1 dietary exposure, several risk assessments have been conducted in the studied period (2016)(2017)(2018)(2019)(2020)(2021)(2022).
Dietary exposure to AFM1 due to consumption of milk (fresh cow, goat and sheep milk) and traditional dairy products (cheese, yoghurt, kashk, doogh and tarkhineh) from Western Iran indicated a potential health risk for consumers, especially during the winters, as HI values were around or greater than 1 i.e., 0.88 during summer and 1.46 ng/kg bw during winter (Bahrami et al., 2016).A risk assessment conducted to assess the risk of AFM1 exposure by using liver cancer potency for the Iranian population revealed an additional risk of 0.08 and 0.72 cancers/year for consumers associated with the mean and 99th percentile consumption of milk, respectively (Fooladi moghaddam et al., 2019).However, no risk of developing liver cancer was found to be associated with the exposure to AFM1 through consumption of UHT and pasteurised milk in Hamadan, Iran, as the HI value (0.535) was less than 1 (Nejad et al., 2019).Moreover, the AFM1 related cancer risk and the synergistic effect of HBV was not critical in Iran (Pardakhti & Maleki, 2019).Additionally, dietary intake of AFM1 through yoghurt consumption showed no risk to public health in Hamadan, Iran considering the maximum limits posed by EC and Institute of Standards and Industrial Research of Iran (ISIRI) (Heshmati et al., 2020).The first risk assessment of AFM1 exposure through infant formula consumption in Iran suggested a health concern for the infant population (0-6 months) based on MOE values < 10,000 but considering low prevalence (1/29 samples), low cancer risk (0.00010 cancers/year/100,000 population) and HI values < 1, no health concern was concluded (Hooshfar et al., 2020).However, results obtained in the risk assessment of AFM1 exposure via traditional and pasteurised industrial milk samples from Isfahan, Iran showed potential health concern as the HI exceeded safe limit of 1 i.e., 1.031 and 1.056 in adults and 5.116 and 4.956 in children for 95 percentile consumption of traditional and industrial milk, respectively (Jafari et al., 2021).Results of a recent global systematic review assessing the safety of infant formula consumption by infants (<1 year) revealed a potential concern as HI and MOE values exceeded the acceptable limits for most of the studied countries and continents.Global HI levels based on mean and median consumption were recorded as 7.78 and 1.16, respectively.While mean and median MOE levels were 366 and 2456, respectively.Whereas global 95th Percentile HI and Moe levels were 25.56 and 4499, respectively (Sharafi et al., 2022).
A study conducted in Brazil suggested a potential health risk for infants of age one month old due to the consumption of infant powdered milk as the HI value was greater than 1 i.e., 1.53 (Ishikawa et al., 2016).In a recent risk characterisation conducted in Brazil based on the MOE approach, infants of age 0-5 years were also found to be at a significant risk of HCC due to AFM1 exposure through consumption of dairy products, UHT milk (0.828-2.523 ng/kg bw/day), powdered milk (0-2.113ng/kg bw/day) and infant formula (0.029-0.833 ng/kg bw/day).The MOE values ranged between 239 and 728 and thus lower than the safe limit of 10,000.Moreover, the number of HCC cases estimated i.e., 0.0015-0.0045cases/100.000individuals were also higher than the limit of 0.001 cases/100,000 (Conteçotto et al., 2021).Similarly, results of the health risk assessment of AFM1 exposure through Brazilian goat milk revealed a potential risk of liver cancer for one year old children as the MOE (159-1546) and HI (0.5499-0.0003) did not pass the safety criteria (De Matos et al., 2021).
The infant population (1-4 years) of Serbia was also observed to be highly exposed to AFM1 occurring in cow's milk and heat-treated milk with an EDI range of 1.01-2.31ng/kg bw/day, respectively.Consequently, the highest HI values between 5.06 and 11.16 were also recorded for this age group.These seemed to decrease with increasing age.Overall, HI values were greater than 1 for all the age groups in all the studied years, meaning a potential health risk of liver cancer for the population, especially for infants of age 1-4 years (Milićević et al., 2017).On the other hand, a recent study that assessed the risk of AFM1 exposure in children aged 1-9 years in Serbia indicated no health risk for the studied population, as estimated MOEs were >10,000 (Milicevic et al., 2021).Based on MOE and HI values, Udovicki, Djekic, Kalogianni, and Rajkovic (2019) characterised a high risk of HCC for the student population of Serbia and Greece due to AFM1 intake through milk and yoghurt consumption.The recorded MOE and HI values for the Serbian population ranged between 213-460 and 13.4-6.2,respectively.For the Greek population these ranged from 1142 to 1629 and 2.5-1.8,respectively.Clearly, Serbian students were at a greater risk as compared to Greek student population.
Similarly, children (<3 years) in low-income areas of Kenya had significantly higher exposure to AFM1, when consuming milk sourced from the low-income area itself as compared to mid-income areas.The corresponding EDI values were 3.6 ng/kg bw/day and 3.2 ng/kg bw/ day, respectively.A low cancer risk of 0.004 cancers/year/100,000 in the urban population was estimated.Additionally, the contribution of AFM1 exposure to stunting was also calculated as 2.4% for children (<3 years) in low-income areas and 2.1% in mid-income areas (Ahlberg et al., 2018).Sirma et al. (2019) also reported a low liver cancer risk arising from AFM1 consumed through milk in rural Kenya.The annual liver cancer incidence rates were 0.0035, 0.0029, 0.0027 and 0.0014 cancers per 100,000 adult females, adult males, children (<5 years) and children (6-18 years), respectively.Risk assessment based on both the MOE (<10,000) and HI (>1) approaches, suggested a health concern for infants, toddlers, and other children in Argentina from AFM1 exposure through consumption of milk and traditional dairy products.Moreover, MOE values for 46%, 38%, and 44% population of infants, toddlers and other children were at no risk as the MOE values were greater than 10, 000.While 45%; 49%; 41% population exceeded the safe limit of HI, respectively, indicating a health concern (Costamagna et al., 2021).HCC risk associated with AFM1 ingestion through milk amongst dairy farming households in Malawi was estimated at 0.038 and 0.023 cancers/year/100,000 child and adult population, respectively, which clearly indicated a high risk for children (Njombwa et al., 2021).
Risk assessment of children's exposure to AFM1 through the intake of different types milk i.e., whole, semi-skimmed and skimmed UHT and pasteurised milk in Turkey revealed a higher risk for the child population under the age of 9 years as the calculated HI values were >1.The consumption of skimmed milk proved the greatest risk (Madali et al., 2018).A MOE based risk assessment of exposure to AFM1 through breastfeeding in the Ecuadorian highlands suggested a potential health concern for children (0-23 months) in both rural and urban areas as the MOEs obtained were in range of 17-65 which is less than 10,000 (Ortiz et al., 2018).A systematic review by Rahmani et al. (2018) also revealed that children in various middle east countries including Iran, Turkey, Saudi Arabia, Syria, Palestine, Lebanon, and Egypt were at a higher risk of AF induced cancer risk through the consumption of raw, pasteurised, and UHT milk containing AFM1.The HI values ranged between 0.09 and 2.28 at mean consumption level.
An exposure assessment and risk characterisation of AFM1 in raw milk commercialised in Italy resulted in the highest HI values of 1.64 and 1.4 for infants and toddler groups, respectively, indicating a potential health concern for them.The estimated risk of HCC was reported as 0.0078 and 0.0038 cancers/year/100,000 population, respectively (Serraino et al., 2019).Similarly, risk assessment of AFM1 in the milk supply chain in Central Italy also revealed a greater health concern for toddlers and children as MOE values were less than 10,000 i.e., 8498 and 2163 for toddlers, 27,301 and 9294 for children at 50th and 99th percentile consumption level of milk (Roila et al., 2021).In a total diet study in the Dutch population, milk and milk beverages and other dairy products highly contributed towards AFM1 exposure in children (53% and 43%, respectively) and adults (51% and 36%, respectively) resulting in low MOEs to AFM1 (Sprong et al., 2016).In the EU, "liquid milk" and "fermented milk products" contributed towards 89% and 87% child and infant exposure to AFM1, respectively, and consequently resulted in lower MOEs ranging from 2020 to 100,000 at mean consumption level (EFSA, 2020).
A global systematic review of the health risk assessment of AFM1 in breast milk revealed that infants consuming human breast milk were not at considerable risk (HI < 1) except for infants aged one month in UAE and Thailand for whom the HI values were greater than 1 (Fakhri et al., 2019).Additionally, infants aged 2-4 months in Northern India were found at a potential health risk through breastmilk as the recorded MOE values from milk consumption were significantly lower than 10,000 and ranged from 10 to 405 (Mehta et al., 2021).AFM1 dietary intake through milk (raw, pasteurised, plain, flavoured, powdered) and milk products (yoghurt, buttermilk) also raised a health concern for children and the adult population of India.The HI values were around or greater than 1 and ranged from 3.192 to 25.435 and 0.798-6.359for both age groups, respectively.The additional risk of HCC in HBsAg + and HBsAgchildren was estimated as 1.92 × 10 − 7 to 1.53 × 10 − 6 and from 6.38 × 10 − 9 to 45.09 9 × 10 − 8 , respectively.While that for HBsAg + and HBsAgadults ranged from 4.79 9 10 − 8 to 3.82 × 10 − 7 and from 1.60 × 10 − 9 to 1.27 × 10 − 8 , respectively (Hattimare et al., 2021).Another risk assessment in India revealed a potential health concern for the consumers of milk and milk products (curd, butter, cheese) in Ludhiana, Punjab.Consumption of milk and curd represented a major risk due to higher HI values i. e 11.54 and 1.64, respectively.Moreover, the HCC incidence risk was highest amongst children of age 1-9 years (0.0106 cancers/year/100,000 population) and lowest for the adults of age 21-60 i. e 0.0020 cancers/year/100,000 population (Kaur et al., 2021).HI indices, especially for children, in Chhattisgarh, India were also observed to be exceeding the safe limit (<1) (Pandey et al., 2021).
Dietary exposure to AFs through the consumption of a Turkish dairy food called "Sürk cheese" contributed to the significant cancer risk for the population based on the HI and MOE values of 1.02 and 2982 for AFM1 and AFB1, respectively (Sakin et al., 2018).In a similar study, dietary exposure risk assessment of AFs through Oaxaca cheese consumed in Mexico City did not represent any substantial risk for the consumers due to very low exposures (Carvajal-Moreno et al., 2019).Additionally, no potential risks of liver cancer were found to be associated with the AFM1 exposure through cheese consumption in Hamadan, Iran, despite the high incidence of AFM1 in analysed samples (Nejad et al., 2020).Despite very few numbers of AFs RASFF notifications (0.3%) in milk and milk products issued between 2016 and 2022, an extensive amount of risk assessments has been published between 2016 and 2022 which could be due to the high economic value associated with milk and milk products.The results of the reviewed studies (Table .2) indicate that dietary exposure to AFM1 through milk and milk products is a substantial health concern for the human population in developing countries.Moreover, children in the age range of 0-9 years old are at a major risk of developing HCC associated with the AFM1 exposure, the reason being their higher consumption of milk and milk-based products and lower body weight as compared to the other age groups.

Conclusion
Significant potential human health risk has been identified from the dietary exposure to AFs through the consumption of all the studied food categories (nuts, nut products and oilseeds, fruits and vegetables, herbs and spices, cereals and cereal products and milk and milk products) which held the largest number of AFs RASFF incidents reported between 4 th January 2016 and 2 nd March 2022 (Fig. 2(a)), except for 'milk and milk products' which was included due to economic value.Commodity wise, peanuts and groundnuts have been identified to pose major risk as the major dietary source of AFs under Nuts, nut products and oilseeds category.However, a need to assess the risk from oilseeds consumption has also been highlighted as there was less evidence available for the risk assessments from oilseed consumption.Likewise, a greater number of risk assessments are required to be conducted for 'Fruits and vegetables' and 'Herbs and spices' categories.For cereals and cereal products, maize and rice have been identified as the major contributor to AFs dietary intake.Even though, milk and milk products presented the least concern from the RASFF point of view (0.3% RASFF incidents, Fig. 2(a)), a large number of risk assessments have been reported over the studied time period (2016-2022) and a significant health risk has been found to be associated with the dietary exposure to AFs from this food category.Country wise, the large majority of AFs RASFF incidents have been recorded in produce which originated in developing countries while some developed countries have also listed amongst the topmost affected by AFs.Contrary to this, the risk assessment data (Table .2) shows that many of the studies have been conducted for the population residing in developing countries while little or no information is available for developed countries.In addition to this, little or no data is available for countries that were listed as having the greatest number of AFs incidents between 2016 and 2022, including Turkey (3 studies), USA (0 studies), Argentina (1 study), China (2 studies) and India (4 studies), (Fig. 2(b); Table (2)).Therefore, a need to assess risk for the populations of developed countries and countries that have had the highest recorded number of AFs RASFF incidents has been highlighted.Age wise, the overall risk associated with dietary intake of AFs followed a decreasing trend with increase in age (Table .2).The trend for the population at most risk was observed as: infants > children > adolescents > adults > elders.

Fig. 1 .
Fig. 1.Summary of the screening process and the results.

Fig. 2
Fig. 2(b).Chart representing the EU RASFF incidents for aflatoxins per country of origin of the food products from 4 th January 2016 and 2 nd March 2022.Data retrieved from FOODAKAI, the food risk prediction platform, Agroknow(FOODAKAI, 2021).

Table 1
Maximum Limits (MLs) imposed by different international regulatory organizations to AFs in foods.

Table 2
Summary of the results obtained from the reviewed literature.

Table 2
(continued ) (continued on next page) K.Bhardwaj et al.

Table 2
(continued ) (continued on next page) K.Bhardwaj et al.