Risk assessment and exposure to levels of naturally occurring aflatoxins in some packaged cereals and cereal based foods consumed in Accra, Ghana

Highlights • Majority 96.3% (26/27) of rice samples had total aflatoxin levels below international set limits (European Commission: 10 μg/kg).• Almost half 45% (9/10) of the cereal based foods (mostly maize) were contaminated with aflatoxins and were above the set limits.• None 0% (0/6) of the macaroni/spaghetti were contaminated with aflatoxins.• AFG1 and AFG2 aflatoxin types were insignificant.


Introduction
Mycotoxins have become a major global concern and also, developed greatly during the last few decades due to their implications to human as well as animal health, productivity and trade. Several methods and strategies have been employed in addressing consumer concerns of various aspects of Food Safety [1] to regulate their occurrence. According to Jard et al. [2], more than 100 molds producing approximately 400 secondary metabolites of toxigenic potentials have been documented till date.
Fungi belonging to the genus Aspergillus are well noted for the production of aflatoxins which are a group of polyketide-derived furanocoumarin secondary metabolites. Species such as A. flavus (produces AFB1 and AFB2), A. parasiticus (produce AFB1, AFB2, AFG1 and AFG2), A. nomius (produces AFB1, AFB2, AFG1 and AFG2), A. ochraceoroseus (produces AFB1, AFB2, AFG1 and AFG2 rarely), A. pseudotamarii (produces AFB1, AFB2, AFG1 and AFG2 rarely) and Emericella venezuelensis is a species with the perfect stage that produces AFB1, AFB2, AFG1 and AFG2 rarely has been implicated for the production of aflatoxins ( [3,4,5]). Aspergillus bombycis reported by Peterson et al. [6] isolated from insect frass in silkworm rearing houses in Japan has currently been added to the list as a new aflatoxigenic species.
Among the five (5) notable mycotoxins in food, aflatoxins (B1, B2, G1 and G2) are the most toxic. Undoubtedly, aflatoxin B1 is the most studied as it has been identified as most toxic and potentially hepatocarcinogenic [10]. Ingestion of sufficient quantities for both livestock and humans, can be highly carcinogenic and acutely toxic or fatal [11] as their effects are realized through respiratory, mucous or cutaneous routes, ending in over activation of inflammatory response. In Ghana and beyond, owing to their nutritional complementary significance and the quest to reduce malnutrition, cereals and legumes are by far the main ingredients for weaning foods as well as baby formulas ( [12,13,14]). Mycotoxin contamination of these foods consumed have been rampant and have not been spared [13].
The Food and Agricultural Organization (FAO) and World Health Organization (WHO) have also identified these toxins present in agricultural products and has resulted in the setting of maximum tolerable limits for mycotoxins in various countries [15]. According to the FDA regulatory levels for aflatoxin in the feed, the maximum allowable aflatoxin levels are 300, 100 and 20 μg/kg for finishing cattle, swine and poultry, breeding cattle, and other animals respectively [16].
The European Union in 2008, placed aflatoxin on the Rapid Alert System for Food and Feed (RASFF) due to its severe effects it caused (European [17]). Again, the International Agency for Research on Cancer (IARC) has also catalogued AFB1 as a group I carcinogen for humans [18,19]. Regrettably, their effect on health is seriously disregarded in developing countries including Ghana. Odamtten [20] reiterated that the danger of mycotoxicoses especially aflatoxicoses still exists in Africa and cannot be discounted in our health delivery system. This paper sought to evaluate the risks and levels of the different types of aflatoxin as well as the total aflatoxin levels in some packaged rice, cereal based and pasta foods as sold on the markets of Ghana.

Sample collection
Samples were randomly purchased from the Makola market (Central market) in the period of May to July in 2015, Accra, Ghana and grouped into 3 categories (rice, cereal based food, pasta) ( Table 1). Twenty (20) grams each of samples were kept in sterile bags and sent to laboratory where they were stored in a deep freezer at −20°C until ready for chemical analysis.

Extraction of samples
AFB1, AFB2, AFG and AFG2 were extracted from samples according to the European Committee for Standardization (CEN) method for aflatoxin extraction [21]. Methanol in water (200 ml) (8 + 2) and 5 g NaCl were used to extract 25 g of sample. Hexane (100 ml) was added to samples containing more than 50% fat. The extracts were filtered and 10 ml of filtrate added to 60 ml of phosphate buffer for solid phase extraction using immune affinity columns specific for AFB1, AFB2, AFG1 and AFG2. This was followed by clean up with 15 ml distilled water and elution with methanol (HPLC grade). The eluate was injected into HPLC to estimate concentrations.

Analysis of samples
The aflatoxins (by Aspergillus flavus and A. parasiticus) levels in the samples were determined by High Performance Liquid Chromatography HPLC (Agilent 1260 Series, OpenLab software, X-bridge column (250 mm x 4.6 mm, i.d., 5 μm), USA with fluorescence detector and post column derivatization using kobra cells to generate bromine electrochemically at the CSIR-Food Research Institute, Ghana. LOD for AFB1 & AFB2 is 0.14 μg/kg and 0.13 μg/kg for AFG1 and AFG2 (Fig.1a).

2.3.
Human risk assessment of exposure to total aflatoxins via consumption of rice, cereal based foods and pasta

Exposure estimation
Calculation of the Estimated Daily Intake (EDI) was done by using the mean levels of aflatoxins obtained in rice, cereal based foods and pasta samples, the daily intakes of same samples [22,23], and the average body weight. The EDI for mean aflatoxins was calculated according to the following formula and expressed in μg/kg of bodyweight/day (μg/kg b.w/day) [24].
= daily ake EDI int (food) X mean level of Aflatoxins average body weight

Estimation of Hazard Index (HI)
The Hazard Index (HI) was calculated according to the below mentioned formula, by dividing the EDI by TD 50 , divided by a safety factor of 50,000. TD 50 is the dose (ng/kg/body weight/day) required to induce tumors in half of test animals that would have remained tumorfree at zero dose as described by Ishikawa et al., [25] and Ismail et al., [26].

Statistical Analysis
Regression analysis was used to calculate the concentrations of the aflatoxins from the curves derived from the standards of the aflatoxins with Excel for Microsoft windows (version 10). Means of results were subjected to analyses of variance (one-way ANOVA) at significant

Results
The samples and their corresponding ingredients investigated, have been described aptly in Table 1. Rice samples (RS1-27), varied Aflatoxins type AFB1 levels for in the rice brands samples ranged from 65.77-Non detectable (N.D) levels. Value obtained for RS4 was    Type AFG1 and AFG2 yields recorded were very low and they ranged between 1.05-N.D (non-detectable) and 0.12-N.D (non-detectable) respectively. Statistically, there were no significant differences (p > 0.05) observed.
The Total aflatoxins yield was significantly (p < 0.05) highest in RS4 than all other sample brands. There was no significant (p > 0.05) difference observed among the other brands.
In the category of cereal based foods, a total of 20 foods were analyzed. Results of the analyzed samples are presented in Figs. 3-4. Aflatoxins AFB1 levels ranged between 35.46-0.96 μg/kg. and the rest were not detected. Values recorded for CBS11 (made from roasted maize, soyabeans and groundnut) was highest (p < 0.05) among all the samples. Following closely were CBS7 and CBS20 with values of 29.56 and 29.0 μg/kg respectively. There was no significant (p > 0.05) difference between them. However, they differed significantly from all the other samples analyzed.
Results for AFB2 ranged between 4.92-0.51 μg/kg while the rest were non detected. Again CBS11 recorded the highest value of 4.92 μg/ kg and was statistically different (p < 0.05). Generally, there were statistical differences (p < 0.05) observed in all the samples analyzed.
AFG1 values recorded in this category was considerably high and ranged from 6.95-0.27 μg/kg. However, the rest were not detected.
AFG2 values observed were generally low which ranged between 0.82-0.1 μg/kg. Statistically, there were no significant differences (p > 0.05) observed. AFG2 were not detected in the rest of the samples. Total aflatoxins levels recorded were within range of values of 45.1-0.28 μg/kg.
Results of the pasta category showed a total of six (6) different samples analyzed. For AFB1, quantities ranged between 0.930 to 0.935 μg/kg. There was no observed statistical difference (p > 0.05). AFB2 quantities ranged between 0.85 to 0.853 μg/kg. Again, there was no significant difference (p > 0.05). AFG1 and AFG2 levels were not detected in these samples.
Total aflatoxin levels for pasta samples ranged between 0.931 to 1.787 μg/kg. Interestingly, there was no significant difference (p > 0.05) observed between PS1 and PS3 both of which obtained the highest level of total aflatoxins. A similar trend was observed for PS2 and PS4. However, AFG1 and AFG2 were not detected in the six (6) different pasta food brands.
Hazard index (HI) Recorded HI values were in the range of 3.9 × 10 -4 -0.0899 (Table 3). In general, it is accepted that an HI ≤ 1 indicates no significant health risk. Nonetheless, the possibility of long-term adverse health effects increases with increasing HI values as an HI between 1.1 and 10 reflects a moderate risk, and HI < 10 indicates high risk [33]. HI value for exposure to aflatoxins via consumption of rice, cereal based  foods and pasta consumed by the Ghanaian population, is less than one. These values this research is reporting imply that intake of rice, cereal based foods and pastas will most likely not pose high risk to the health of Ghanaian population. 4 Discussion In Ghana, the pervasiveness of Aspergillus species in stored dehydrated foods is a well-known phenomenon and has been well-documented [34]. In view of this, it is imperative to update the aflatoxicogenic potential of these fungi in relation to food commodities.
The rather surprising high levels of AFB1, AFB2 and total aflatoxin levels of recorded in RS4 and RS1 (Above EU limit of 5 and 10 μg/kg for AFB1 and total aflatoxins respectively) could be attributed to some favourable intrinsic factors such as pH, nutrient composition, moisture content/water activity as well as extrinsic factors as temperature; relative humidity; soil properties; mechanic injury on food commodity; insects and rodents attack [35,36]. Atanda et al. [35] also, suggested that these factors, however, do not work in solitude. Therefore, two or more factors may have to be met before fungal growth and corresponding toxin production can be effected.
Fungal infestation, growth and aflatoxin development is linked principally to water activity (Aw). This observation is attributable to improper drying which predisposes stored cereals and legumes to growth of mycotoxigenic fungi such as Aspergillus species which is conjectured to also increase with storage time [13,37,38]. The risk of Aspergillus flavus and its metabolites accumulation could be high at maturity and harvest of food crops at the end of the raining season. In Ghana, traditional drying techniques which comprise of fieldand bare ground-drying; as well as poor handling contribute immensely to fungal contamination as the soil is known to be a reservoir of these fungi [39].
Although the levels of aflatoxin contamination found in this study in rice for human consumption (up to 86.06 μg/kg) are far lower than the levels (1,600-12,000 μg/kg) reported by Afla-Guard [44], that caused deaths in the two fatal outbreaks of aflatoxin poisoning in Kenya accumulated intakes of small doses of aflatoxins may aggravate other clinical conditions such as immunosuppression, still-births and neonatal mortality with increased susceptibility to infectious diseases such as pneumonia, stunting of growth and HIV/AIDS [45,46]. It is indeed of great concern as protracted intakes of such toxic levels could similarly work synergestically with other carcinogens especially hepatitis B virus to aggravate primary liver cancer [47]. It is noteworthy that no exposure to any level of aflatoxins could be regarded as safe [48].
Results of the maize samples agreed with published findings of Sirma et al [49] whom from a study conducted in Kenya, showed that 60% of total household maize samples (269) across the study tested positive for aflatoxin with a range of 0.17-5.3 ppb while all 39 millet samples recorded aflatoxin levels ranging from 0.14 -6.4 ppb, none of which exceeded the country's regulatory limit of 10 ppb. In another study, 55% of maize products in Kenya reported by Lewis et al [50] had aflatoxin levels greater than the set limit of 20 ppb, while 35% of the products had levels exceeding 100 μg/kg and7% above 1000 μg/kg. From Malawi, a range of 1.47-22.47 μg/kg was reported by Mwalwayo and Thole [51] in maize samples consumed in the rural areas.
Recently, Omeiza et al [52] also reported a range of 1.8-13.5 μg/kg in feed samples in Nigeria On the contrary, sorghum samples recorded aflatoxin levels ranging from 0.15 to 210.1 ppb (overall mean of 26.0 ppb), out of which 46% of samples (from the Kilibwoni sub-location) exceeding the maximum tolerable limit of 10 ppb registered the highest percentage which suggests a possible variation in aflatoxin frequency with geographical location The results by Sugri et al. [53] for aflatoxin prevalence range of 0.011-308 ppb in maize samples (240) in six districts in the Upper East and Upper West regions of Ghana. Kumi et al. [54] from Ghana, reported aflatoxin levels in homemade weanimix ranging from 7.9 -500 ppb. The data obtained from their study showed that two (2) out of the thirty-six (36) samples (from different communities) had very high levels of 460 and 500 ppb (ug/kg). Aflatoxin contamination in maize grains from a total of 38 major store markets in Benin, Ghana and Togo was monitored by James et al. [55] and reported aflatoxin   [57] who reported values of total aflatoxin in the samples of range 0.04 to 10.20 μg/kg of an average of 5.52 μg/kg in mesir pasta in Turkey. From Pakistan, Iqbal et al. [58] reported macaroni (5.91 μg/kg), noodles (7.35 μg/kg) and spaghetti (9.12 μg/kg). Ezekiel and Sombie [59] also reported 1.1, 2.2 and 1.2 ppb for macaroni, noodles and spaghetti respectively in Nigeria. The incidence of acute hepatic encephalopathy in children, had been linked to aflatoxin contamination of noodles with an approximate 3 mg suspected to be present in a single serving in Malaysia [60] It has to be underscored that majority 96.3% (26/27) of the rice samples had total aflatoxin levels below acceptable limits (10 μg/kg) set by the 77 countries that regulate aflatoxins, including the European Union ( [61]; EC 2006). However, approximately half 45% (9/20) of cereal based foods had AFB1 and total aflatoxins within the acceptable levels stipulated by the EU regulations for AFB1 and total aflatoxins in rice. The pasta category of samples had all samples within acceptable limits set by the EU.
Interestingly, majority of the samples had very low to no value for AFG1 likewise AFG2 (Figs. 1-3). In this standing, it was reported by some previous researchers ( [62]; [5]) that AFB1 is considered to be a precursor of AFG1 and thus explaining the relative accumulation of both toxins. It was also reported by Reiter et al., [3], that cross reactivity by AFG1 to the AFB1 antibodies occur at higher percentages in immune ultra-filtration cleanup of aflatoxins.
The danger for the consumer who may purchase and use an opened bag over a long period and sometimes even beyond the expiry date, may expose himself to contaminated products since these produce is often kept in moist/damp storage houses for a very long time. The health implications are enormous in Africa since most produce are not properly dried.

Conclusion
It can be surmised from our results that samples had 3.7% (1/27), 45% (9/20) and no aflatoxins levels respectively for rice, maize based foods and spaghetti with aflatoxin above permitted levels (E.U) (≤ 5 and 10 μg/kg for AFB1 and total aflatoxins respectively). The Hazard Indices recorded values less than 1 and so Ghanaians consuming these products were not at risk.
This evaluation of mycotoxins in Ghanaian rice gives an indication of the quality of the cereal with regards to its acceptability for human and animal consumption. The demonstrated presence of AFs at concentrations above the limits acceptable to world mycotoxin regulatory agencies and the co-occurrences of toxins with possible toxic synergistic effects make this studied rice samples of low quality for human and animal consumption and in fact raises preliminarily national public health concerns. However, with appropriate measures such as awareness creation and proper management practices, the prevalence of the toxin can be minimized, if not eradicated.

Authors' contributions
NKK, AAA and FA performed the experiments and wrote the manuscript. VK-B was responsible for AFB1, AFB2, AFG1 and AFG2 analysis. NKK, AAA, HWA, FA and VK-B helped conceive the experiments and prepared the manuscript. NKK, AAA and FA conceived the original study and HWA and VK-B led the sampling and study in Ghana. All authors read and approved the final manuscript.

Conflicts of interest
The authors declare that there are no conflicts of interest.

Transparency document
The Transparency document associated with this article can be found in the online version.