The Scourge of Aflatoxins in Kenya: A 60-Year Review (1960 to 2020)

are endemic in Kenya. The 2004 outbreak of acute aﬂatoxicosis in the country was one of the unprecedented epidemics of human aﬂatoxin poisoning recorded in mycotoxin history. In this study, an elaborate review was performed to synthesize Kenya’s major ﬁndings in relation to aﬂatoxins, their prevalence, detection, quantiﬁcation, exposure assessment, prevention, and management in various matrices. Data retrieved indicate that the toxins are primarily biosynthesized by Aspergillus ﬂavus and A. parasiticus , with the eastern part of the country reportedly more aﬂatoxin-prone. Aﬂatoxins have been reported in maize and maize products ( Busaa , chan’gaa , githeri , irio , muthokoi , uji , and ugali ), peanuts and its products, rice, cassava, sorghum, millet, yams, beers, dried ﬁsh, animal feeds, dairy and herbal products, and sometimes in tandem with other mycotoxins. The highest total aﬂatoxin concentration of 58,000 μ g/kg has been reported in maize. At least 500 acute human illnesses and 200 deaths due to aﬂatoxins have been reported. The causes and prevalence of aﬂatoxins have been grossly ascribed to poor agronomic practices, low education levels, and inadequate statutory regulation and sensitization. Low diet diversity has aggravated exposure to aﬂatoxins in Kenya because maize as a dietetic staple is aﬂatoxin-prone. Detection and surveillance are only barely adequate, though some exposure assessments have been conducted. There is a need to widen diet diversity as a measure of reducing exposure due to consumption of aﬂatoxin-contaminated foods.

Aflatoxins are a group of mycotoxins produced by at least 20 fungal strains of Aspergillus section Flavi, Nidulantes, and Ochraceorosei [10,11]. eir discovery and recognition are traced back to 1960 in which Turkey "X" disease was reported in England with several poults lost to the toxins after feeding on a contaminated peanut ration [12,13]. AFs were eventually recovered in East Africa (Kenya and Uganda) in peanut rations that caused substantial losses in ducklings [14,15]. AFs are chemically polysubstituted coumarins with very similar chemical structures [16]. About 20 different types have been reported, and aflatoxin B 1 (AFB 1 ), aflatoxin G 1 (AFG 1 ), aflatoxin M 1 (AFM 1 ), aflatoxin B 2 (AFB 2 ), and aflatoxin G 2 (AFG 2 ) are of demonstrated toxicological importance [17]. e B-type AFs are pentanone derivatives that show strong blue fluorescence under UV light. On the other hand, the G-series AFs are six-membered lactones that fluoresce yellow-green under UV light and thus the B and G nomenclature [2,18]. Aflatoxin G 2 and B 2 are congeners of G 1 and B 1 which lack the 8,9-double bond in the furan ring and therefore usually only encountered in tandem with the latter [19]. Aflatoxins M 1 and M 2 are metabolic derivatives of B 1 and B 2 that exhibit blue-violet fluorescence and are usually detected in urine and milk of animals served AFB 1 -contaminated rations [20,21].
Mutegi et al. [38] published a review on the prevalence and strategies for mitigation of AFs in Kenya from 1960 to 2018. Since then, more than 15 studies on AFs have been undertaken in Kenya. e current review digests the scourge of AFs in Kenya from 1960 to the present, highlighting the progresses in the occurrence, detection, quantification, and exposure assessment. Prevention and control measures as well as evidence-based management strategies are discussed.

Causative Fungi and Prevalence of Aflatoxins.
Aflatoxins in Kenya are predominantly produced by Aspergillus parasiticus and A. flavus [39][40][41][42][43][44][45][46][47][48][49]. e latter is a universal fungus known to produce AFB 1 and AFB 2 along with aspergillic, cyclopiazonic, and kojic acids [50]. A. parasiticus produces both B and G AFs plus kojic and aspergillic acids [50][51][52]. A. niger, A. terreus, and A. versicolor were reported in soils and mill dust in Eastern Kenya [40]. Further, the occurrence of A. alliaceus, A. tamarii, and A. caelatus in Kenya has been echoed [44,53,54]. A genetic profiling study reported that A. minisclerotigenes in Eastern Kenya exhibited a higher AF biosynthesis potential than A. flavus [42]. ough both the L-and S-strain morphologies of Aspergillus section Flavi have been reported, probing aetiological studies revealed that aflatoxicoses associated with maize consumption in Kenya have been due to a novel S-morphology fungus previously implicated for the 2004-2006 aflatoxicosis outbreaks [39,55,56]. Overall, A. flavus is considered the primary producer of AFs in agricultural commodities with an optimal growth temperature of 25°C and a minimum water activity of 0.75. AF biosynthesis, however, starts at 10-12°C [57].
Kenya possesses an erratic tropical climate characterized by periodic droughts, high humidity, and high temperatures preceding harvests [58]. e climate is tropical along the coast, temperate inland, and arid in the north and northeast. e country has distinct seasons: March to June with long rains and October to December with intermittent rains. It has four marked climatic zones, further partitioned into agroecological zones contingent on the rain and temperature conditions suitable for the staple crops.
e Central Highlands and the Rift Valley are blessed with fertile soils, rainfall of up to 3000 mm per annum, and temperatures of 21°C to 26°C. On the other hand, the western part of the country is hot and remains wet throughout. Rain received is more than 1000 mm per annum with temperatures of 27°C to 29°C. North and eastern parts are relatively hot with annual rains lower than 510 mm and temperatures which occasionally reach 39°C in some areas [58].
Poor grain conditioning before storage, use of propylene storage bags, drying of grain on bare grounds, insect infestation, poor storage structures (stores with leaking roofs), poor transportation, and handling of produce as well as chronic poverty have been incriminated for the aflatoxigenic contamination of Kenyan foods [3,[59][60][61][62][63][64][65][66][67][68]. Contamination has also been due to the cultivation of maize in ecologically predisposed regions of the country [69][70][71][72][73]. Biophysical factors (including soil, plant genetic constitution and susceptibility, and fungal community), low education levels, inadequate sensitization, and gender have also favoured the spread of AFs in Kenya [60,67,[74][75][76]. On toxicological studies, AFB 1 is by far the most studied AF in Kenya, followed by AFM 1 [38]. us, most studies reported on the levels of AFB 1 , AFM 1 , or total AFs. It is worth noting that because of the aflatoxicoses that dawned several times on the country, a number of investigations have been undertaken, with often alarming AF levels reported [38,63,70,[77][78][79].

Cereals and Cereal-Based Products.
Maize, millet, and sorghum are Kenyan staples for specific regions. Maize is the main dietary staple [94,95]. Subsistence farmers stock maize under various suboptimal conditions for more than 3 months prior to use or sale [96]. Maize is often for home consumption as flour or used for making irio and githeri (a traditional dish of maize mixed with legumes or pulses such as beans, pigeon peas, and cowpeas, usually cooked whole), though some may be sold [8]. Approximately 60% of the maize are processed for consumption using hammer mills [40,97]. It was previously echoed that maize consumption is the primary route through which Africans have been chronically exposed to AFs [98][99][100]. On the other hand, sorghum and millet are grown primarily in the semiarid regions of the country and are consumed mainly as flours used for the preparation of thick porridge (ugali) and thin porridge (uji). Uji is an ingredient of infant weaning foods and diet for children [9].
In Kenya, maize meal consumption is estimated at 400 g/ person/day with an average total AF content of 0.132 μg/kg and has been incriminated for all aflatoxicoses recorded [71,101]. In one of the pioneering studies, Kenji et al. [81] reported very high total AFs of 1,120 μg/kg in malted maize with an 86% incidence of AFB 1 . AFB 1 ranged from 0 to 260 μg/kg in malted millet from ika market (Kenya) though no AFB 2 and AFG 1 were detected. On the other hand, maize flour had AFB 1 ranging from 0 to 160 μg/kg (from Nairobi) and traces (from ika) with undetectable AFB 2 . In another study, 68% of a maize-based traditional brew (Busaa) in the slums of Nairobi was declared to contain AFs in concentrations above 5 μg/kg, 17% of which were above 50 μg/kg [102]. Likewise, the magnitude of AF contamination of 480 maize grains, maize flour, and dehulled dry maize-muthokoi (362 random environmental samples, 26 cases, and 92 controls) samples from Makueni, Kitui, Machakos, and ika districts was assessed [103]. It was reported that 46.4% of the environmental samples, 15% of cases, and 29.3% of controls were within the then threshold of 20 μg/kg, implying that 54.6% of the samples could not be used for human consumption. Further, 6.9% of the environmental samples, 57.7% of cases, and 21.7% of controls had AF concentrations above 1000 μg/kg. e overall AF contamination of the samples ranged from 0 to 58,000 μg/kg [103]. Further, Sirma et al. [76] recorded total AFs of 0.17-5.3 μg/kg from 67% of maize from the Rift Valley region. About 92% of millet and 50% of sorghum samples collected in the study were positive for AFs in the ranges of 0.14-6.4 μg/kg and 0.21-210.1 μg/kg, respectively.
Later, Muthomi et al. [40] reported that samples of whole maize, mill dust, and semiprocessed maize in Machakos, Journal of Food Quality Eastern Kenya, had more than 20 μg/kg AFB 1 threshold allowed by then in Kenya. e highest AFB 1 level (160 μg/kg) was recorded in whole grains. e dust registered the highest AF content, probably due to dehulling operations and the continuous availability of maize products which are potential substrates for A. flavus proliferation. As expected, semiprocessed grains had the lowest contamination levels and this was speculated to be so due to dehulling of the grains as reported elsewhere [104]. Similarly, a random study appraising market maize contamination as well as the relationship between market maize AFs and aflatoxicosis outbreak was conducted [71]. A total of 65 markets were surveyed, 243 maize vendors were interviewed, and 350 maize and maize products samples were drawn from the most affected districts as per the previous history of aflatoxicoses. About 55% of the samples had AFs in levels above the then advisory threshold of 20 μg/ kg, 35% had levels >100 μg/kg, and 7% had levels >1,000 μg/ kg (Table 1). Makueni district which had the highest number of aflatoxicosis case-patients had evidently higher market maize AF concentrations than ika district (which had the fewest case-patients) with a geometric mean of 52.91 μg/kg versus 7.52 μg/kg. In addition, maize from local farms in the affected areas was more likely to have AFs in concentrations >20 μg/kg vis-à-vis those from other regions or countries.
Because it was understood that contaminated homegrown maize in the affected areas infiltrated the distribution system, wild AF contamination of market maize was inevitable, and the contaminated market maize bought after homegrown maize supplies were exhausted was cited as the reason for persistent exposure to AFs. e authors stressed that efforts to meaningfully evade AF exposure during an aflatoxicosis epidemic should take into account the role of the market chain [71].
Probst et al. [55] reported that, in Eastern province (Kitui and Mukueni), Coast (Makueni, Kwale, Kilifi, Tana River, and Taita Taveta), and Rift Valley (Marakwet, Keiiyo II, Kajiado, Baringo, Nakuru, and Laikipia), total AFs in maize ranged from 219.6 to 426.3 μg/kg, 0.1-120.4 μg/kg, and below detection limit (BDL) to 13.4 μg/kg, respectively. Indeed, the Aflacontrol Project [105] also reconfirmed this observation. Maize grains sampled between January 2010 and May 2010 from fields (preharvest), stores (postharvest), and wholesalers, retailers, and open-air vendors were declared to be contaminated with AFs. e highest level of AFs in preharvest maize (n � 281) was 1,455 μg/kg from Mbooni East (Eastern Kenya). No appreciable differences were noted between samples from Western and Eastern Kenya. For example, samples from Homa Bay and Rongo had 37 μg/kg and 54 μg/ kg of total AFs vis-à-vis 21 μg/kg reported in Makueni, 25 μg/ kg reported in Mbeere North, and 44 μg/kg reported in Mbooni East. Matter-of-factly, more samples from the Western sites were unfit for human consumption (had total AFs >10 μg/kg) than those from the eastern sites. For 241 postharvest samples, 38% from the eastern region had AF levels above 10 μg/kg. e plague was most acute in Makueni where 87% of samples were unfit for human consumption and the maximum AF level was 1,777 μg/kg. In Mbooni East and Mbeere North, the proportion of maize with levels above 10 μg/kg was 29% and 7%, respectively. In entirety, the proportion of maize unfit for human consumption was higher in the eastern sites than in the western sites, but there was considerable variation across the different areas sampled.
In consonance with the aforementioned, Muthomi et al. [107] evaluated the distribution and contamination levels of Aspergillus species (spp) and AFB 1 in soil, maize, and maizebased products. Maize grain (n � 256), semiprocessed grain (n � 56), flour (n � 52), hammer mill dust (n � 11), and soils (n � 117) had A. flavus in all the samples, though the fungi were prevalent in the grain. AFB 1 was undetected in samples from the humid regions but was present in concentrations in excess of 10 μg/kg in 20% of the samples, with maxima of 136 μg/kg for semiprocessed maize, 77 μg/kg for whole grain, and 41 μg/kg for flour in open bags. Incidental high temperature and periodic droughts prevalent in the semiarid regions were incriminated for the higher levels of A. flavus and AFB 1 recorded. A study report recapitulated in [16,108] indicated that Kenyan maize was the most contaminated in East Africa with a mean of 131.7 μg/kg (Table 2).
A total of 54 processed and unprocessed (brands A and B) cattle feed from agricultural and veterinary stores and 96 human foods (unprocessed and processed maize, polished and unpolished rice, peanut seeds, and flour) samples collected from open market traders in Nairobi County were analyzed [86]. e awareness of the traders on AFs and the associated health effects were assessed using questionnaires. Total AF concentrations recorded were 120.9 ± 27.2 μg/kg (processed feed), 77.6 ± 16.0 μg/kg (brand A), 48.6 ± 12.0 μg/ kg (brand B), 49.7 ± 14.7 μg/kg (unprocessed maize), 101.20 ± 21.30 μg/kg (maize flour), 38.2 ± 10.5 μg/kg (unpolished rice), 63.9 ± 14.5 μg/kg (polished rice), 54.6 ± 14.8 μg/kg (peanut seeds), and 120.9 ± 27.2 μg/kg (peanut flour). Higher AF levels were reported in processed foods (mean: 95.0 ± 12.7 μg/kg) than in nonprocessed foods (mean: 47.5 ± 7.6 μg/kg), and this implied that some food processing techniques used predisposed the foods to aflatoxigenic contamination. Roughly 56.6% of the traders were aware of AF contamination; cattle feed traders were more conversant with AFs (40%) than human food traders (17%). A very small portion of food traders (3.7%) and feed traders (8%) were aware of the health effects of AFs in humans and animals, respectively. Because the mean AF levels in both feeds and foods were above statutory limits, the author recommended the need for creating traders' awareness on AFs, their effects, and practices that favour AF proliferation.
Aflatoxin exposure via intake of maize and its products was evaluated through analysis of 20 samples each of maize To check for chronic inadvertent exposure to AFs, maize (n � 75) and maize flour (n � 27) from different parts of Kenya were collected and analyzed [95]. Striking differences in the AF levels of maize grain between the regions and stores from which samples were drawn were reported. Eastern Kenya had the highest contaminated samples with a mean of 22.54 ± 4.94 μg/kg, while those from Nairobi had the lowest contamination (7.92 ± 1.57 μg/kg). No appreciable differences were observed for total AFs in maize flours from the regions. AFs in maize flours were marginally above the European Union (EU) limit of 5 μg/kg, and most of the samples had AFs lower than the statutory limit of 10 μg/kg. e authors attributed this to adherence to good manufacturing practices by the millers. e highest AF level in maize flours from Eastern Kenya was 6.98 ± 0.53 μg/kg [95]. Recently, Obonyo and Salano [63] echoed that maize grain in Eastern Kenya from May harvest had lower AF levels with variation (5.68 ± 6.31 μg/kg, 100% AFB 1 ) than those from October-December rains (10.77 ± 10.14 μg/kg, 72% AFB 1 ). From the two seasons, the authors hinted that 16% and 44% of the samples had total AFs above the statutory limit of 10 μg/kg. Another group [110] undertook a random study in three agroecological zones: Kitui (semihumid to semiarid), Nakuru (semihumid), and Kitale (subhumid to semihumid) to determine the prevalence and AFs in 130 stored maize samples and the aflatoxigenicity potential of A. flavus in stored maize. e authors put forward that aflatoxigenic contamination between the sampled sites was markedly distinct with the highest mean total AFs of 9.68 μg/kg reported in Kitale district. A. flavus was detected in 91 (70%) samples and the colonies with exorbitant aflatoxigenicity stemmed from Nakuru County with a recorded mean total AF of 239.7 μg/kg.
Recently, maize from smallholder farmers' fields in Eastern and Southwestern Kenya (n � 789) was analyzed for AFB 1 [94]. e authors detected AFB 1 (range: 0.01-9,091.8 μg/kg; mean: 67.8 μg/kg) in 274/416 samples   1 in the East was 0.07 to 60,612 ng/kg bw/day (mean 451.8 ng/kgbw/day), while for the Southwest, it varied from 6.53 to 4,814.7 ng/kgbw/day with a mean of 148.4 ng/kgbw/ day. e mean PDI for the regions surpassed AFB 1 tolerable daily intake and therefore poses health concerns. As such, it was advanced that control of preharvest AF contamination should be emphasized in the bid to reduce exposure to AFs through maize consumption [94]. e prevalence and levels of AFs in freshly harvested maize and freshly milled maize flour (n � 338) from households in Siaya and Makueni Counties were evaluated by Nabwire et al. [77]. All (100%) of the samples had detectable AFs, which ranged from 2.14 to 411 μg/kg. e geometric mean of total AFs in samples from Makueni and Siaya Counties was reported as 62.5 μg/kg and 52.8 μg/kg, respectively. is revalidated the fact that AFs are prevalent in maize and its products in the area. Taken together, regional disparities in AF levels of Kenyan maize have been reported with the Eastern provinces registering the highest level of 58,000 μg/kg [70,103,109] vis-à-vis the other provinces that registered 4,500 μg/kg as the maximum [72,78]. Recently, the Kenya Bureau of Standards (KEBS) banned a number of maize flour products on the market because of high AF levels [8]. As per the current study, reluctancy to dehull maize has been recognized as a plausible explanation for the exorbitant AF concentrations recorded in Kenyan maize flour.

Peanuts (Arachis hypogaea L.) and Its Products.
Peanuts (groundnuts) are the only affordable dietary protein source in Kenya [65]. It is grown predominantly in the western but sold and consumed countrywide [61,111]. Peanut productivity has over the years declined due to unpredictable rainfall, lack of disease-resistant peanut varieties, poor agronomic practices, and poor institutional support accorded to farmers [112][113][114]. It is grown for local consumption but is also exported through the World Food Programme [115,116]. Peanut is rich in proteins, fats, carbohydrates, zinc, sodium, potassium, calcium, magnesium, iron, phosphorous, and vitamins E and B [117]. In tandem with maize, they are the major portions of the gruel used to make weaning foods in Kenya and these have been shown to be a route of AF exposure [118,119]. Most of the peanut samples tested in the country had AF levels above recommended regulatory limits set by the KEBS [120]. Fortunately, its consumption is as low as 1.1 g/person/day [92].
In one of the earlier surveys [61], baseline data on AF levels as well as 384 and 385 peanut samples from Busia and Homa Bay districts of Western Kenya were collected and analyzed. Total AFs ranged from 0 to 2,688 μg/kg and 0 to 7,525 μg/kg in samples from Busia and Homa Bay, respectively. Out of all the samples drawn (n � 769), 87.01% contained <4 μg/kg of AFs, 5.45% were in the range ≥4 and 20 μg/kg while 7.54% surpassed the advisory threshold of 20 μg/kg. ere was a highly strong association between the districts and the analytical total AFs recorded, which was further corroborated by a significant correlation between total AF levels and agroecological zones. Logistic regression unveiled that peanuts from Busia were 2.6 times at risk of contamination vis-à-vis those from Homa Bay and that planting improved cultivars could lower the odds of contamination to half those for local landraces. In the continuity of the foregoing, the authors [44] reported that the total AF content of 436 peanut samples drawn from Busia and Homa Bay districts varied from BDL to 2,687.6 μg/kg and BDL to 1,838.3 μg/kg in about 32% of the samples with detectable AFs. Both the incidence and the number of colonies of A. flavus S-strain were significantly and positively correlated with the total AF content of the samples. Up to 99.3% of the samples containing <10 μg/kg of total AFs did not have A. flavus S-strain. is corroborated a previous report which confirmed the presence of Aspergillus, Rhizopus, Fusarium, and Penicillium spp in peanuts with AF contents spanning beyond 100 μg/kg [84].
In another survey by Mutegi et al. [62], peanut and their products from supermarkets and informal markets were analyzed. e authors announced that raw podded peanuts had the lowest AF contamination, with 96% having levels of less than 4 μg/kg and only 4% having more than 10 μg/kg. Irrespective of the provenance, 69% of the samples and 75% of spoilt nuts had total AFs exceeding 10 μg/kg. ough most samples (59%) had AF levels below 4 μg/kg, only 4% of these were acceptable under the KEBS but could be rejected under EU regulations. Of these, 37% of the peanuts were found to be unfit for human consumption as per KEBS and EU regulatory limits. Further, the team [64] evaluated the effect of storage bags, temperature, and relative humidity on the quality and AF content of peanut kernels of Homa Bay local, Valencia red, ICGV-SM 12991, and ICGV-SM 99568 varieties stored for 6 months in jute, polypropylene, and polyethylene bags. Moisture content, physical damage, rancidity, and AF levels were determined before storage and after every 30 days during storage. Moisture content changed remarkably from 3.3 to 6.9% with samples stored in different bag types recording mean values of 5.1% (polypropylene), 5.2% (polyethylene), and 5.3% (jute). Physical damage (range: 0.1 to 9.8%) was influenced by storage temperature and relative humidity and the type of storage used. Rancidity ranged from 0.8 to 5.3 and increased with storage duration from a mean of 1.5 before storage to a peak of 2.5 after 5 months of storage.
ere was a reported variation in total AFs (range: 0 to 47.8 μg/kg) of nuts stored in polyethylene bags having 7.3% and 13.4% more contamination than those stashed in polypropylene and jute bags.
Accordingly, it was hypothesized that the processing of peanuts in the cottage industry could facilitate their contamination by AFs. As such, Ndung'u et al. [53] assessed the AF content of raw and roasted peanuts and peanut butter marketed in Nairobi and Nyanza Provinces of Kenya. Marketers and processors of these were also interviewed on the source of groundnuts and the incidence of Aspergillus section Flavi was determined. e authors stressed that the percentage of defective nuts among all unsorted nuts ranged from 0 to 26.3%. e mean percent of defective nuts was higher for Nairobi (imported from Malawi) than for Nyanza (homegrown) samples. Total AFs in the samples ranged from 0 to 2,377.1 μg/kg with higher mean total levels in raw samples from Nairobi than from Nyanza (Table 3). e source of groundnuts and defective nuts was positively associated with AF levels. A. flavus (L-and S-strains), A. parasiticus, A. niger, A. tamari, A. alliaceus, A. caelatus, and Penicillium spp were isolated from the samples. e prevalence and diversity of fungal spp and aflatoxigenic contamination of 228 marketed peanut samples (from 140 formal and 88 from informal markets) in Kericho and Eldoret towns of Kenya were established [115]. A. flavus (L-and S-strains), A. parasiticus, A. tamarii, A. caelatus, A. alliaceus (members of Aspergillus section Flavi), and A. niger as well as Penicillium, Mucor, Fusarium, and Rhizopus spp were encountered. Total AFs in the nut products were 0 to 2,345 μg/kg in raw, 0 to 382 μg/kg in roasted, and 0 to 201 μg/kg in roasted decoated peanuts. Altogether, AFs occurred in higher concentrations in samples from informal (mean � 97.1 μg/kg) than from formal (mean � 55.5 μg/kg) markets. Meanwhile, a positive and strong correlation was cited between AF levels and the major aflatoxigenic fungi in raw peanuts from formal markets of Eldoret. Further, AFs in raw nuts from informal markets in Kericho positively and strongly correlated with the population of A. flavus (both strains). In roasted coated peanuts sampled from Eldoret formal markets, AFs correlated positively and significantly with A. flavus S-strain.
Another investigation [121] which compared the oil content and total AF level of peanuts in Busia and Kisii Central districts reported that Valencia red, Uganda local, Homa Bay local, and Local red peanut varieties from Busia had lower levels of total AFs except the Local red variety which had the highest total AF of 267 μg/kg with the lowest average oil content of 42.7% (Table 4). Peanuts from Kisii Central had higher AF levels and low oil contents. Summed up, there was an increase in total AF levels with decreasing oil contents except for Uganda local red from Kisii. In the continuity of the foregoing study, Menza and Muturi [49] reported the occurrence of five causative Aspergillus fungi: A. flavus (L-and S-strains), A. parasiticus, A. niger, and A. tamari. Overall, the occurrence of A. flavus (both strains) was significantly higher than other aflatoxigenic spp identified in the nuts. A. flavus L-strain was the most common isolate (58.8%) in samples from Busia while the S-strain dominated (60.2%) in peanuts from Kisii Central. All in all, A. flavus S-strain was the most dominant with a mean prevalence of 45.1%.
From the prevenient reports, it is notable that relatively higher AF concentrations have been detected and quantified in Kenyan peanuts. A plausible explanation advanced has been that aflatoxigenic fungi infect the shells, testa, and seeds as the pods in the soil grow. Further, mechanical damage while harvesting, drying, and storing aggravates the risk of invasion by the toxigenic fungi and aflatoxin biosynthesis.
is is corroborated by a Tanzanian report which unveiled that grains and oilseeds borne on aerial generative structures had comparatively lower AF levels vis-à-vis those borne in geocarpic structures of Bambara and peanuts [122].

Cassava (Manihot esculenta Crantz).
Cassava is a revered food crop with edible carbohydrate-rich tuberous roots and proteinaceous young leaves [123,124]. However, it houses 2 cyanogenic glucosides: linamarin and lotaustralin (methyl linamarin) inherently synthesized for protection against predation. ese cyanogens are spread in the whole plant but occur in higher quantities in the leaves and root cortex [125]. Cyanide is an inhibitor of aerobic respiration through blockage of mitochondrial electron transport and oxygen uptake. In addition, cassava is also prone to mycotoxins, particularly AFs.
In a study, dried cassava chips (n � 13) and cassava flour (n � 26) sourced from Nairobi and Mombasa markets were assessed for hydrogen cyanide, AF, and moisture contents [126]. Hydrogen cyanide ranged from 27.20 to 42.92 mg/kg and 21.45-37.77 mg/kg in cassava chips and 21.53 to 64.63 mg/kg and 21.70 to 70.03 mg/kg in flour from Nairobi and Mombasa, respectively. ese were all above 10 mg/kg recommended by the East African standards (EAS 739: 2010 and EAS 740: 2010, respectively). AFs were detected in 2 flour samples from Nairobi (mean levels of 6.60 and 8.89 μg/ kg) and a sample from Mombasa (mean level of 2.84 μg/kg). Moisture content ranged from 8.62 to 9.98% and 8.85 to 11.57% in cassava chips and 8.50 to 12.51% and 7.30 to 11.0% in flour samples from Nairobi and Mombasa, respectively. e study revealed that marketed cassava flour though of good aesthetic quality could be mycotoxigenically unsafe for consumption.
ere are no reports in the open literature on plant products such as sugarcane, spices, beans, wheat, and barley in Kenya. A. flavus was not identified in soils with sugarcane grown which reportedly had Phanerochaete chrysosporium, A. niger, Trichoderme viride, and Fusarium equiseti [46]. Sugarcane is one of the daily consumables in the form of sugar and has been previously reported to house AFs [127]. In addition, no recent studies have reported on the AF content of commercial beers consumed by Kenyans despite it being one of the widely consumed foods that evidently utilize cassava and cereal food crops (sorghum, barley, and maize). Beers are essentially continuous mixed-culture fermentation products. As such, brewing could be the innocuous route for AF exposure as it offers auspicious conditions in which toxigenic fungi thrive and may be an excellent avenue for use of aflatoxin-contaminated raw materials as the end-users cannot directly notice [16]. Similarly, no reports exist on AFs in beans. In the neighbouring Uganda where sometimes Kenya imports beans, AFs were earlier recorded in excess of 1,000 μg/kg [128].
developed [58] of the regions in Kenya that are at risk of AF outbreaks from milk consumption, and this encompassed all the milk production areas of Kenya. e dairy industry in Kenya is dominated by crosses between dairy and zebu breeds with more than 70% contribution to total national milk production. e feeds used are natural forage, cultivated fodder, and crop byproducts such as maize stalks and stover. Supplements such as dairy meal, maize germ, maize bran, cottonseed cake, wheat pollard, and wheat bran are also sometimes used [74].
A correlative study conducted in four urban centers by Kang'ethe and Lang'a [89] analyzed 613 milk and 830 feed samples for AFM 1 and AFB 1 . About 86% (353/412) of the feed samples from farmers were positive for AFB 1 and 67% (235/353) of these exceeded the FAO/WHO limit of 5 μg/kg. About 81% (197/243) of the feeds from feed millers and 87% (153/175) from agrochemical shops were AF positive, with 58% (115/197) and 66% (92/153) of these samples exceeding permissible limits, respectively. Approximately 72% (315/ 439) of the milk from dairy farmers, 84% (71/85) from largeand medium-scale farmers, and 99% (88/89) of the pasteurized marketed milk were positive for AFM 1 , and 20%, 35%, and 31% of positive milk from dairy farmers, mediumand large-scale farmers, and market outlets, respectively, exceeded the WHO/FAO limits of 0.05 μg/kg. On the one hand, 67% of the urban smallholder dairy farmers had knowledge that milk could be contaminated with AFM 1 but did not know the possible exposure mitigation strategies. Feed millers, on the other hand, knew about AFB 1 in grains and its excretion as AFM 1 in milk but were not alleviating exposure to animals [89] (Table 5). Similarly, Sirma et al. [130] surveyed 286 households in 37 villages representing four agroecological zones (semiarid, temperate, subhumid, and humid). ey drew 280 samples of bovine milk which were subjected to AFM 1 analysis. AF levels were from 0 to 0.359 μg/kg. Generally, 58.9% of the milk sampled had AFM 1 levels BDL though 9.3% exceeded the WHO/FAO limit of 0.05 μg/kg (Table 6).
Aflatoxins were detected and quantified in fresh and sundried Rastrineobola argentea (Dagaa fish) collected from various markets in Luanda, Rongo, Kisumu, Ahero, and Maseno of the Winam Gulf of Lake Victoria [85]. Fresh samples had no detectable AFs, but the dried samples had mean total AF levels of 0.34 ± 0.09, 0.21 ± 0.00, 0.25 ± 0.06, 0.53 ± 0.11, and 0.11 ± 0.00 μg/kg wet weight, respectively. It was asserted that the occurrence of AFs in processed Dagaa fish (omena) could have been due to the fact that the samples were collected from the markets in July 2010 when there were rains and drying was incomplete, and thus, the sundried Dagaa fish were packed in sacks when they were incompletely dried which favoured the growth of moulds.
In a bid to assess the AF status of marketed raw milk and associated risk factors in periurban Nairobi, raw milk retailers in Dagoretti division were interviewed and milk Adapted from [121]. Values in bold indicate exceedance of the permissible limit of 10 μg/kg. processed milk and milk products) from Bomet County were positive for AFM 1 with an overall prevalence of 84.32% [132]. About 43.8% of these were above 0.05 μg/kg, with raw milk compared to processed milk (52% vs. 8.6%) having more contamination.
In the same manner, AFM 1 was detected in 291 samples of raw, pasteurized, and UHT milk, yoghurt, and Lala [90]. Monthly samples were drawn over a period of 1 year, just as  Journal of Food Quality a consumer would purchase them from retailers and traders in a low-income area (Dagoretti) and a major supermarket in a middle/high-income area (Nairobi). More than 50% of the samples had AFs exceeding 0.05 μg/kg, though only 3 exceeded 0.5 μg/kg and the geometric mean AFM 1 level was 0.0619 μg/kg in the 135 samples from Dagoretti while it was 0.0361 μg/kg in 156 samples from Nairobi. e levels varied significantly depending on the time of year, with the lowest levels reported in January. UHT milk had the lowest AF levels, and more expensive milk had lower AFM 1 levels [90].
In a recent study [129], it was pointed that exposure to AFM 1 in milk and the health risks associated with it are not clearly understood and monitored in Kenya. us, the team assessed the awareness, knowledge, and practices of urban and periurban farmers about AFs and evaluated the levels of AFs in on-farm milk in Kasarani subcounty, Nairobi County. In total, 84 milk samples were analyzed, and 90% (83/84) were analytically declared to be contaminated with AFM 1 (mean value of 0.084 μg/kg). About 64% of the samples had AFM 1 levels well above the EU limit of 0.05 μg/ kg. ough 80% of the farmers had knowledge of AFs, no correlation existed between the farmers' knowledge and gender with AFM 1 prevalence.
Kang'ethe et al. [87] reported that 45.5% and 98.6% of bovine milk and animal feeds in Kenya were positive for AFs. About 49% and 83% of these had AFM 1 above 0.05 μg/kg and AFB 1 above 10 μg/kg, respectively. Similarly, an AF risk mapping study from milk consumption using biophysical and socioeconomic data [58] reported a mean AFB 1 content of 9.25 μg/kg in animal feeds and a mean AFM 1 content of 0.0265 μg/kg in bovine milk. Higher mean of the logarithmic AFB 1 concentrations was reported in areas with historical aflatoxicosis outbreaks compared to those without outbreak history, a phenomenon that was not true for the mean logarithm of AFM 1 when compared between areas with and those without a history of aflatoxicosis outbreaks. Analogously, a cross-sectional study of aflatoxigenic contamination of bovine milk and dairy concentrates was done in five counties of Kenya representing the agroecological zones: Kwale, Isiolo, araka-Nithi, Kisii, and Bungoma [133]. Concentrates and milk were collected twice (during the dry season and rainy season) from 285 farmers in the five counties and analyzed for AFB 1 and AFM 1 . Between 0 and 68% used concentrates, which had AFB 1 ranging from <1 μg/kg to 9,661 μg/kg with 47.8 to 90.3% positive samples. About 33.3% to 87.5% of the concentrates had more than 5 μg/kg AFB 1 (83.3% to 100% from retailers and 28.6% to 100% from manufacturers). AFM 1 prevalence in milk was lowest in Kwale (13.6%) and highest in araka-Nithi (65.1%). About 3.4% (Kwale) to 26.2% ( araka-Nithi) of milk samples had AFM 1 above the WHO/FAO threshold of 0.05 μg/kg, with the highest contamination of 6.999 μg/kg. e study was in consonance with preceding studies which indicated that AFs are prevalent in Kenyan dairy rations and milk.
In Kisumu, Anyango et al. [134] reported that 97 randomly selected dairy farmers primarily fed cows on forage and concentrates. AFM 1 levels in milk collected from these farms ranged from BDL to 0.151 μg/kg (mean of 0.02967 μg/ kg) and 26.4% of these exceeded the EU limit. Concentrate feeding was associated with higher AFM 1 levels so that farms feeding concentrates were more likely to record milk AF levels above 0.05 μg/kg [134]. Further, the prevalence of AFM 1 in 96 samples of informally marketed milk from Nairobi, the knowledge of milk traders on AFs, and the effects of boiling and fermentation on AFM 1 were assessed [135]. By and large, all samples had detectable AFM 1 (limit of detection � 0.005 μg/kg) with a mean of 290.3 ± 0.663 μg/kg. About 64% of the samples had AFM 1 above 0.05 μg/kg while 7.5% exceeded 0.5 μg/kg. Majority of the traders had low (69.8%) or medium (30.2%) knowledge of AFs. e educated and female traders were more knowledgeable, and fermentation of milk to Lala (a traditional fermented drink) or yoghurt significantly reduced AFM 1 levels by 71.8% (in Lala after 15-hour room temperature incubation) and 73.6% in yoghurt after incubation at 45°C for 4 hours. Boiling, however, had no appreciable effect on AFM 1 levels [135].
According to Sirma et al. [9] using a quantitative risk model, an equivalent of 5 hepatocellular cancer cases and deaths and the disability-adjusted life years of 255 for Kenya in 2016 were estimated as due to exposure to AFs in milk. Other than milk, there are no reports in the open literature on AF content of other products of animal origin such as blood, eggs, ghee, and meat in Kenya.

Animal Feeds.
As pointed earlier, farming is one possible exposure route to AFs. For example, maize which is known to be highly susceptible to AF contamination in Kenya is also a major component of livestock and poultry feeds, and therefore, regular indirect human exposure through the consumption of animal products that contain AF residues cannot be underrated. Elevated levels of AFB 1 have been recorded in Kenyan animal feeds [88,89]. e situation is exacerbated by dairy farmers' habit of utilizing spoilt (pest-or mould-damaged and rotten) grains for the formulation of dairy rations [74,92]. A study carried out on animal feeds in Nairobi Province revealed that AFs ranged from 5.13 μg/kg to 1,123 μg/kg, with the largest proportion lying between 11 μg/kg and 99 μg/kg [78].
Further, 81 fish feeds sourced from 70 farms and 8 feed manufacturing establishments located in Nyeri, Kenya, were subjected to AF analysis by Mwihia et al. [73]. Fish were also sampled from 12 farms for gross and microscopic pathological investigation. About 84% of the feeds were AFpositive (range 1.8-39.7 μg/kg, mean of 7.0-8.3 μg/kg, and median of 3.6 μg/kg). About 18.5% of the feeds sampled registered total AFs above the statutory limit of 10 μg/kg. Meanwhile, homemade and tilapia feeds had evidently higher AF levels than commercial and trout feeds. Maize bran-based feeds and fish meal recorded higher AF levels than those devoid of these constituents. Microscopy revealed that five trout farms (41.7%) had fish with swollen abdomens, enlarged livers with white or yellow nodules, and large dark basophilic hepatic cells with hyperchromatic nuclei in irregular cords. As such, the authors inferred that aflatoxigenic contamination of fish feeds is a scourge in Nyeri which if left unchecked may cause detrimental health effects in edible fish in the area.

Co-Occurrence of Aflatoxins with Other Mycotoxins.
It is now established that mycotoxins can coexist in foods [136]. In 1995, Muriuki and Siboe [45] analyzed 40 samples of flour packed in 90 kg bags, 58 samples of Ugali brand, and 74 samples of Jogoo brand drawn from Nairobi, Kenya. e samples were analyzed for resident mycoflora, and some mycotoxins associated with key fungal spp. Aspergillus flavus, A. sulphureus, Fusarium moniliforme, Penicillium stoloniferum, and P. cyclopium were the reported fungal spp in the samples. Ochratoxin A was the most prevalent mycotoxin, and all the flour brands had AFB 1 and AFB 2 (0.4-20 μg/kg), ochratoxin A (50-1,500 μg/kg), and ZEA (2,500-5,000 μg/kg). e authors recommended the need for rigorous countrywide monitoring of mycotoxins in maize at both farm and market levels. e foregoing was substantiated by a report by Kedera et al. [137] who reported the presence of Fusarium fungi and fumonisin B 1 (FB 1 ) in maize kernel samples from smallholder farm storages in Bomet, Bungoma, Kakamega, Kericho, Kisii, Nandi, Siaya, Trans-Nzoia, and Vihiga districts in the tropical highlands of Western Kenya. Later, Mbugua and Gathumbi [138] affirmed the occurrence of AFB 1 , FB 1 , ZEA, and DON in 36 Pilsner and 39 Tusker beer samples sourced from Nairobi and the surrounding satellite towns. All the samples were negative for AFB 1 ; the prevalence of DON and ZEA were 100% in both brands while FB 1 incidence was 72%, with incidences in Tusker (76.9%) being markedly higher than in Pilsner (66.7%).
e mean values of contamination were 3.29 and 3.57 ng/mL for DON, 0.28 ng/mL and 0.32 ng/mL for FB 1 , and 7.84 and 8.50 pg/ml for ZEA in Tusker and Pilsner brands, respectively. A positive correlation was reported between DON and FB 1 , and DON and ZEA, affirming their co-occurrence to be from Fusarium spp. is communication suggested that there were some but safe exposure to Fusarium mycotoxins by lager beer consumers of Kenya.
Fumonisin B 1 and AFB 1 in symptomless and rotten maize harvested at different harvest time points after physiological maturity (HTPAPM) from Malava and Tongaren were evaluated [139]. Similarly, maize samples were collected from 30 markets in diverse agroecological zones of Meru, Machakos, and Kitui counties during the 2013 harvest [54]. Fusarium and Aspergillus spp were isolated from the samples. Total AFs in Meru, Kitui, and Machakos samples were beyond the threshold of 10 μg/kg. Meru had both the highest and lowest levels of AFs detected (115.7 μg/kg and 0.3 μg/kg, respectively). FUMs were reported in levels above the acceptable limits in Meru, and though detected in Kitui and Machakos, the contamination levels were within acceptable limits. Utilizing a near-infrared single kernel sorting machine, removal of AF and FUM-contaminated kernels was perfected with up to 97.8% efficacy for AFs and 60.8% for FUM. e accepted fractions had statistically lower mycotoxin levels than the rejected maize [54]. e prevalence of AFs and FUM was investigated in maize intended for immediate human consumption in Eastern Kenya. Samples were collected from people who brought their maize for processing at local commercial mills [75]. Interviews and sampling of maize flours were done for 1,500 people who processed maize at 143 mills in 10 administrative districts. Mycotoxin analysis revealed that 39% and 37% of the samples, respectively, had AFs and FUM in levels above tolerable limits. Samples with AFs above 10 μg/ kg were 22-60% across the districts. A higher occurrence of AFs was associated with smaller maize farms, lower grain yield, and monocropping systems. A larger magnitude of the toxin was observed in the subhumid agroecological zone, in samples with more broken kernels and less maize ear damage at harvest. Further scrutiny of paired grain samples (visually sorted and unsorted) showed that sorting reduced FUM by 65% to below the advisory threshold of 1,000 μg/kg. Sorting did not, in essence, have any effect on AF concentration [75].
Comparably, Mutiga et al. [72] evaluated AFs and FUM in maize from Western Kenya. e study covered 3 agroecological zones, taking samples of milled maize from 985 patrons of 26 hammer mills. AFs were detected in 49% of the Journal of Food Quality samples, with 15% of these being above 10 μg/kg. An estimated 65% of the samples from a drought-prone area were above acceptable limits. In Bungoma County, the authors assessed both AFs and FUM in four maize varieties at harvest and after 2 and 4 months of storage. For this, storage shed grain and milled samples were solicited. Mean AFs were identical for storage sheds and mills at 2.3 μg/kg. About 41% of the samples from mills had detectable AFs, 4% of which were above 10 μg/kg, while 87% had detectable FUM, with 50% above 1,000 μg/kg limit permitted in Kenya. Mean contamination levels did not vary during storage. As such, maize varieties reportedly differed in FUM contamination, with the most popular varieties spotted to be vulnerable to AFs, FUM, and weevils. It was concluded that thorough mycotoxin surveillance is vital for all parts of Kenya, irrespective of the past history of mycotoxin poisoning [72].
Samples of 74 animal feeds and 120 milk samples were simultaneously collected from individual cows and actors in the informal subvalue chains of rural and periurban dairy systems in Nakuru County, Kenya [142]. AFB 1 was detected in 56 % (41/74) of the feeds in levels above the EU limit of 5 μg/kg (range: BDL to 147.86 μg/kg) while DON was identified in 63% (27/43) of the feeds (range: BDL to 179.89 μg/kg). In the periurban dairy system, 48.5% (33/68) of the milk samples were contaminated with AFM 1 in levels exceeding EU threshold of 0.05 μg/kg (range: 0.017 to 0.083 μg/kg). Surprisingly, all milk samples from the rural dairy system had AFM 1 in levels below the EU limit of 0.05 μg/kg (range: BDL to 0.041 μg/kg). Linear regression depicted that there was a correlation between abiotic factors, viz., pH, water activity, and moisture content of feeds with AFB 1 and DON contamination.
Herbal preparations were sampled from Eldoret (14 liquid, 2 oil, and 34 powder samples) and Mombasa (12 liquid, 1 capsule, 3 oil, 6 tablets, and 28 powder samples) towns and analyzed for total AFs and FUMs [91]. Reported 32% of herbal products from Eldoret had AF levels less than 0.25 μg/kg, while 34% had AFs between 0.38 and 24 μg/kg. FUM occurred in very low concentrations in more than half of the samples. Samples drawn from Mombasa had AFs in levels lower than those from Eldoret, but the number of AFcontaminated samples was higher. About 32% of the samples had <0.25 μg/kg with 14 μg/kg being the highest. About 80% had <0.25 μg/kg, and the highest was >20 μg/kg. Six out of 14 (42.9%) liquid herbal samples from Eldoret were contaminated with AFs and 3 of the 6 were also contaminated with FUMs. All the 12 (100%) liquid samples taken from Mombasa were contaminated with both AFs and FUMs. A total of 27 out of 34 (79.4%) powders from Eldoret were contaminated, 23 with both mycotoxins and 4 with AFs only, while all the tablets (15 samples) and powders (19 samples) from Mombasa were contaminated with both mycotoxins; however, all capsules were free of mycotoxin contamination. All oily herbal samples (n � 3) from Mombasa were contaminated with both AFs and FUM, while only 1 oil sample from Eldoret was contaminated with FUM [91].
Into the bargain, a survey covering 116 push-pull and 139 non-push-pull cropping systems was conducted to determine the socioeconomic and agronomic factors that influence farmers' knowledge on incidence and contamination of maize by ear rots and associated mycotoxins in Siaya, Kakamega, Kisumu, Migori, and Vihiga counties of Western Kenya [143]. Data from smallholder farmers (23-80 years, 50% being female) were collected using questionnaires, and 10-20 maize cobs, depending on the size of cob, were collected from the standing crop in the field of each interviewed farmer and analyzed for AFs and FUMs. e authors reported that few farmers had knowledge of AFs and ear rots in maize. Overall, less than 20% of maize samples had AFs co-occurring with FUM, but more samples were contaminated with FUMs (range: 145.3-50,769.2 μg/kg) than AFs (range: BDL-242.3 μg/kg) with maize containing the mycotoxins in levels above permissible limits (10 μg/kg for AFs and 1,000 μg/kg for FUMs) being lower in samples from push-pull cropping system. Age of farmer and county of residence were significantly and positively associated with knowledge of AFs on the one hand. On the other hand, cropping system, county of residence, and level of education were positively associated with knowledge of maize ear rots. In addition, a strong correlation between knowledge of maize ear rots and knowledge of AFs was witnessed. e concentration of the mycotoxins was significantly and positively associated with the use of diammonium phosphate fertilizer at planting. AF levels were also positively associated with stem borer pest damage, though agronomic practices were not ideally different between push-pull and non-pushpull farmers [143].

Geographical Distribution of Aflatoxins in Kenya.
Kenya was one of the hotspots of AFs first recorded [15,144] with countries such as Uganda and Brazil [26,145]. Kenya is partitioned into about 7 agroecological zones: humid, subhumid, transitional, temperate, semiarid, arid, and perarid [146]. AFs tend to be detected in samples from all the different zones [105,110,130]. is can be attributed to the similarity in the agronomic and pre-, peri-, and postharvest handling practices and the interregional marketing of foods [61,92,111,147]. However, the eastern part of the country is more aflatoxin-prone, possesses the most toxigenic Aspergillus spp, and has been the epicenter of aflatoxicoses recorded in Kenya [46]. Eastern Kenya experiences hotter and drier climatic conditions in comparison with Western Kenya. For this reason, it received characterization as semihumid to semiarid while Western Kenya is classified as subhumid to semihumid agroecological zone [110]. Environmental conditions have been demonstrated to influence the ability of Aspergillus fungi to infect, colonize, and survive on crops as well as produce mycotoxins. Further, fluctuations in these conditions also affect the quantities as well as community compositions of aflatoxin-producing fungi [148]. e prevalence of AFs in Eastern Kenya therefore is in congruence with a previous emphasis that mycotoxin infectivity is always multifactorial, but climate is the most important [149].

Capacity for Detection and Quantification
Detection and quantification of AFs are key to their mitigation because their distribution in samples is often skewed [150]. e first step for accurate detection and quantification of AFs is sampling, i.e., sampling/subsampling is the largest source of error in AFs analysis [151]. For this reason, a representative sample ought to be drawn from the sample lot. For the over 50 KEBS listed laboratories for monitoring mycotoxins in foods, Gafta methods (No. 130, 24:1) and EAS 79 are used as the sampling protocols. However, some clients do the sampling themselves, in which case the testing laboratories do not question the actual reason for sampling, or where and how the samples were taken. In addition, data from such analyses are always confidential, which do not enhance evidence-based decision making by policymakers [97].
Another emerging challenge in analyses of food toxins in Africa, Asia, America, and Europe is "masked mycotoxins" as they are not often identified and detected by the usual analytical techniques [152]. Masked (matrix-associated) mycotoxins are those that are biosynthesized by the toxigenic fungi and later undergo biomodification by plant enzymes during the infection stages. ey may be housed in the vacuoles in the soluble form or bound to macromolecules and thus remain undetectable [153]. Unfortunately, these modified toxins can hydrolyze and revert back into their toxic forms during processing or digestion [154][155][156]. A way to circumvent this analytical problem has been to hydrolyze the modified forms (using enzymes, alkaline, or acidic pretreatments) [157][158][159] into their free forms which can then be detected [157,160]. For this reason, there is a paucity of data on masked AFs as usually detection and quantification are done for free AFs in matrices. e methods for the detection of AFs used by studies in Kenya are outlined in Table 7. On the whole, AF research in Kenya used laboratory-based enzyme-linked immunosorbent assays (ELISAs), high-performance liquid chromatography (HPLC), thin-layer chromatography (TLC), fluorimetry, liquid chromatography-tandem mass spectrometry (LC-MS/MS), tandem quadrupole mass spectrometry (TQMS), and ultra-high-pressure liquid chromatography (UHPLC). Lateral flow immunochromatography (LFI) has also been used. ere has been a shift in instrumentation for AF analysis, as evidenced by advancement from nondifferential TLC in 1973 to relatively fast and differential UHPLC, hyphenated with triple quadruple mass spectrometry (UHPLC-TTQS) in 2017-2020. Overall, the most employed method has been ELISA, which itself has undergone several advancements in the past few years. is could be because it is practically inexpensive, easy to use, is highly sensitive for routine analysis of food products, demands minimum sample clean up, and poses no inherent health hazards as it uses enzyme labels. In addition, concurrent analysis of several samples on a 96-well assay platform is possible, and thus, it has a high sample throughput with low sample volume requirement which offers obvious advantages [28]. In addition, ELISA has lower detection limits than most instrumental techniques which are used for AF determination [28].
However, the drawbacks of the foregoing standard methods are that they are unsuitable for rapid and real-time applications in food and feed sample analyses as they are relatively tedious and require technical know-how to operate. Rapid and robust methods such as polymerase chain reaction (PCR) and nondestructive methods based on fluorescence/near-infrared spectroscopy (FS/NIRS) and hyperspectral imaging (HSI) have emerged for quick and easy detection of AFs [161]. Some studies in Kenya [41-43, 46, 47, 110, 162] have utilized PCR in their analyses. It is of interest to note that, at industrial level, agroprocessing entities monitor grain total AFs utilizing single-step lateral flow immunoassays utilizing Reveal Q+ test strips that are developed and read on AccuScan Gold readers [163]. e bright greenish-yellow fluorescence (BGYF) or the black light test, which can aptly identify commodities presumed to be contaminated with AFs, has been reported in Kenya [54].
is test is relatively cheap and simple especially for detecting AFs in maize where kernels are viewed under an ultraviolet lamp at 365 nm for characteristic bright greenishyellow fluorescence which indicates a possible presence of aflatoxigenic fungi or the mycotoxin itself [164]. is could be adopted by regulatory agencies for the purpose of AFs surveillance.

Exposure to Aflatoxins in Kenya.
Exposure to AFs occurs via periodic ingestion of contaminated plant or animal products such as meat, eggs, blood, and milk of livestock previously served AF-contaminated rations [20]. Farmers and their workers may also inhale dust generated during the processing of contaminated crops and feeds or the toxins may permeate through their skin [165,166]. Exposure to AFs such as AFM 1 may also be through their endogenous production [32]. In point of fact, AFs are known to cross the placenta, so that exposure to them may start in utero and continues in the postnatal period through breastfeeding [92,[167][168][169].
It is now established that detecting and quantifying food AF levels are not always adequately reflective of the extent of exposure because the quantities in foods are not directly the same as those ingested. For this reason, epidemiological biomarkers are often used to assess exposure. Biomarkers are quite exact in evaluating the magnitude of AF exposure because of their nonsubjectivity and faculty that allows the estimation of internal and biologically effective doses. Popular AF biomarkers are urinary AF-N 7 -guanine (for assessing previous day's exposure) and breast milk AFM 1 which indicate exposure levels in the past 24 hours and plasma/serum aflatoxin-albumin (AF-alb) adduct with a half-life of about 2 months enhancing the evaluation of chronic and routine exposure [170]. Albumin is the sole serum protein that can get bound to AFB 1 , yielding  Journal of Food Quality biomarker adducts [171]. Human blood and urine AF-alb can be used to determine the biologically effective dose of ingested AFB 1 . ough both AFG 1 and AFB 1 can covalently interact with albumin to form AF-8, 9-epoxide [98], the adduct levels are taken to reflect the amount of the latter as the former is infrequently encountered in foods [172]. e AF-alb adduct is the most frequently utilized AF biomarker because it is easily detectable using ELISA (Table 7) [173]. Additionally, the determination of AFB 1 -Lys levels in proteolytic digests of serum with LC-MS/MS and HPLC-FS has recently been appraised [174,175]. Biopsy material was first utilized in 1967 to illustrate that the Kamba ethnic community of Kenya had a frequency of liver cancer that was doubling that of the Kikuyus [176]. is is partly supported by the fact that subsequent aflatoxicoses were witnessed in Eastern Kenya where the Kamba are the main inhabitants [71,177]. A dietary AF-liver cancer study in Murang'a district of Kenya reassessed the correlation between AF and the disease incidence rates based on a total of 7 years of cancer registration [82]. e results of the study were however interpreted in combination with a study later done in Swaziland. With consideration of males and females separately, the pooled results of the studies hinted that there was a high degree of positive correlation between the calculated ingestion levels of AFs (X) and the adult incidences of hepatocellular carcinoma (Y) for the two studied populations and for both males and females. With the assumption of a wet intake diet of 2 kg/day and a mean body weight of 70 kg, the relationship for adult females was Y � 4.14 log 10 X − 0.80. With a further assumption of a daily intake of native beer of 2 liters/day, the regression equation for adult males was Y � 21.96 log 10 X − 11.17. e regression data were reported to corroborate those reported by previous researchers [82].
To validate the assertion, another team [188] evaluated if there were any synergistic effects of hepatitis B virus and AFB 1 on the incidences of hepatocellular cancer. e study encompassed various parts of Kenya with different liver  [82] Years are those in which the data joined scientific literature with most data gathered and analyzed in more than 3 months before publication. Muthokoi are maize kernels with the outer hull removed. Lala also called Maziwa Lala or Mala is a locally fermented milk product. Busaa is a sociocultural maize-based traditional brew mostly consumed during events such as male circumcisions, weddings, and funerals, made from raw maize flour and semiground finger millet malt [141]. LC-HRMS/MS: liquid chromatography high-resolution mass spectrometry. LFI: Hoffmann et al. [180] used Romer AgraStrip rapid test; Kirui et al. [141] used Envirologix Quick Tox kits.
cancer incidences so as to establish the rate of exposure to AFs and the prevalence of hepatitis infections. It turned out that of all the tested participants, 12.6% were positive for AF exposure as shown by urinary excretion of AF-N 7 -guanine adduct and the highest exposure to the toxins was in the Western Highlands and Central Province. e incidence of hepatitis infection nationwide as measured by the presence of the surface antigens was 10.6% with a marked regional variation. Execution of multiplicative and additive regression analysis suggested that the two were not a synergetic combination in the etiology of liver cancer, though a moderate degree of correlation between AF exposure and liver cancer was observed when the study was limited to certain ethnic groups [188]. Further, Maxwell et al. [187] undertook a study in Kenya, Sudan, Ghana, and Nigeria to evaluate the extent of AF exposure by breastfed infants and to investigate the possibility that AFs cross the human placental membrane. In this study, breast milk, cord blood, and maternal blood were analyzed for AFs which were detected in 28% of 191 Kenyan, 37% of 99 Sudanese, and 34% of 510 Ghanaian breast milk samples (Table 8). Blood drawn from 101 babies in Kenya, 282 babies in Ghana, and 78 babies in Nigeria had AFs in 37%, 31%, and 12% of the samples, respectively. In Kenya, the rate of detection was higher in the wet season (52%) than in the dry season (23%). Maternal blood sampled at delivery in 83 Kenyan cases and 77 Nigerian cases recorded AFs in both maternal and cord blood specimens in 14 Kenyan and 7 Nigerian instances. ese confirmed that infantile exposure to AFs occurs and demonstrated the ability of AFs to cross the human placental membrane [187].
Differently, a survey which recruited adults from Kenya, ailand, e Gambia, and France was used to validate the measurement of AF-albumin adducts by three methods [186]. Levels of 7 to 338 pg AF/mg alb were observed in the first three countries while no adducts were detected in samples from France. Another cross-sectional serosurvey in Kenya confirmed regional influence on AF exposure patterns [184]. Randomly selected 600 serum specimens stratified by province from a 2007 Kenya AIDS Indicator Survey were analyzed for AFs. About 78% of the sampled group had exposure to AFs and this varied by province. e highest were in Eastern (median � 7.87 pg/mg alb) and Coast (median � 3.70 pg/mg alb) provinces, while Nyanza (median ≤limit of detection) and Rift Valley (median � 0.70 pg/mg alb) provinces recorded the lowest exposures. According to the authors, age group, sex, marital status, religion, and socioeconomic characteristics did not influence exposure.
In another study, random samples of weaning flours were obtained from 242 households with 3-to 36-month-old children (43.6% males and 53.4% females) in Kisumu district, Kenya, and analyzed for AFs [118]. e types of weaning foods, handling, and storage of the foods were captured. e nutritional status of the children was also determined along with heights and lengths. About 29% (70/ 242) of the samples were positive for total AFs (range: 2-82 μg/kg). Malnutrition was 34% for stunting, 30% for underweight, and 6% for wasting. About 53.8% of the wasted children were being fed on AF-contaminated weaner flour vis-à-vis 27.7% of the normal children. e contaminated flours (n � 70) were being stored in plastic containers (63%), polyethylene bags (20.4%), metal buckets (3.7%), manila sacks (1.9%), earthen pots (1.9%), and reed baskets (7.4%) for 1 day to 2-3 weeks. ese all had effects on aflatoxigenic contamination as those with AFs had higher mean moisture content (13.6%) than those devoid of AFs (12.5%). Aspergillus spp (including A. flavus and A. parasiticus), Fusarium, Mucor, Rhizopus nigricans, Trichoderma viride, and Candida spp were isolated from the flour samples [118].
Agreeably, Leroy et al. [60] collected socioeconomic data to quantify the extent to which socioeconomic characteristics could explain the differences in serum AFB 1 -lysine adduct levels in 100 women from the Eastern province of Kenya.
e correlation between serum AFB 1 level and a number of households, farm, and individual characteristics was assessed for 884 mothers (pregnant or with a child under 24 months). AF was detected in all women with a median level of 7.47 pg/mg alb. Higher exposure levels were correlated with poverty: predicted serum AF levels in women living in the worst socioeconomic conditions were 4.7-7.1 times higher than those with the best socioeconomic status.
Further, samples of Rastrineobola argentea (n � 50), polished rice (n � 31), peanuts (n � 22), cassava (n � 37), maize (n � 41), and sorghum (n � 28) were collected from Kibuye wholesale, Kibuye open-air, Ahero, Oile, and Mamboleo markets in Kisumu County and analyzed for AFs [80]. Processed bovine milk samples (n � 50) were collected from supermarkets along with raw bovine milk samples (n � 30) from 3 market milk bazaars in the same markets. Analytical results indicated that AFs ranged from 0 to 34.5 μg/kg AFB 1 in the solid foods, 0.012 to 0.127 μg/kg AFM 1 in processed milk, and 0.0002 to 0.013 μg/kg AFM 1 in raw milk. Only cassava among the scrutinized food items had detectable AFs below the regulatory limit of 10 μg/kg AFB 1 by then. Daily AF consumption ranged from 4.43 ng/ kg bw/day in a combination of maize flour and milk to 110.4 ng/kg bw/day in a combination of sorghum and raw milk for 6-month-old children (average weight: 7.9 kg) with a daily consumption of 60 g of mixed cereal flour and 500 ml of milk per day.
ese results emphasized that weaning children in Kisumu County are chronically exposed to high AF levels for the fact that the analyzed food items are common ingredients of weaning foods in the area [80]. In addition, the calculated AF consumption of 0.6 ng/kg bw/ day for a child at 6 months weighing 7.9 kg was higher than that indicated by the Codex Alimentarius Committee (0.1 ng/person/day) AFM 1 through milk for the Africa region. e weighted mean concentration of 0.05 μg AFM 1 in milk and a consumption of 0.25 ng/kg bw/day have been associated with a prevalence of between 3.2 and 20 cancer cases/year/10 6 [189]. e exposure was much higher than estimated because most children in Kenya are breastfed until at least the latter part of the second year and yet they begin to receive cereal-based gruel before the age of 3 months [190]. Further, the results of the study corroborated a previous report which estimated that about 40% of foods from farmers in the Nyanza province had AF levels above the statutory limit of 10 μg/kg [65].
Another cross-sectional study was undertaken involving 204 low-income households randomly selected in two lowincome areas (Korogocho and Dagoretti), Nairobi, Kenya [66]. Demographics, a 24-hour dietary recall, and anthropometric measurements were conducted in children aged 1-3 years. Height-for-age Z-scores (HAZ), weight-for-age Z-scores (WAZ), and weight-for-height Z-scores (WHZ) were calculated for each child using the WHO growth standard reference data. Maize (n � 99 and 87), sorghum (n � 53 and 36), and milk (n � 76 and 52) samples from the households or retailers from Korogocho and Dagoretti, respectively, were analyzed for total AFs and AFM 1  About 41% of the children had stunted growth; boys were more stunted than girls (p � 0.057), and Korogocho had more stunted children than Dagoretti (p � 0.041). e average AF exposure was 21.3 ng/kg bw/day. Exposure to AFM 1 , location, and sex was significantly associated with HAZ, with boys and children from Korogocho having lower HAZ, and AFM 1 was negatively associated with HAZ (p � 0.047), suggesting that AFM 1 was associated with stunting. No correlation was statistically found between total AFs and HAZ, WAZ, and WHZ. e authors reiterated that there was a high prevalence of malnutrition (stunting) in the studied low-income urban sites, and this was most pronounced in the high-density area. It was stressed that the association between AFM 1 and growth impairment warranted further investigations [66].
Kang'ethe et al. [169] reported that, with a maize consumption of 0.1 to 0.25 kg/person/day in Nandi and Makueni counties, an AF exposure rate of 0.011 and 0.49 μg/ kg bw/day, respectively, was recorded in children younger than 5 years. Exposure to AFM 1 through milk consumption in this study was 4 × 10 −4 and 1 × 10 −4 μg/kg bw/day, respectively. Breast milk nursed children on the other hand had exposure of 6 × 10 −3 and 1 × 10 −6 μg/kg bw/day in Makueni and Nandi, respectively. Children younger than 30 months in Makueni had 1.4 times higher levels of AFM 1 in urine than those of the same age in Nandi. e stunting and severe stunting rates in Makueni and Nandi were 28.7% and 18.5%, and 30.7% and 16.5%, respectively.
In a recent study [180] which enrolled 1230 unborn children, 881 (72%) were included in LAZ and 798 (65%) in the serum AFB 1 analysis. A cluster randomised controlled design was used (28 intervention and 28 control clusters). e intervention arm received a swapping (contaminated maize was replaced with safe maize) and a stockist intervention (households were encouraged to purchase from a stockist supplied with clean maize). Women in the fifth to the final month of pregnancy were invited to enroll in the study. Outcomes were child LAZ, the prevalence of stunting, and child serum AFB 1 -lysine adduct level 24 (end-line, primary outcomes) and 11 to 19 months (midline, secondary outcomes) after trial commencement, respectively. e intervention was reported to considerably reduce end-line ln serum AFB 1 -lysine adduct levels (intervention effect was 0.273, 95% CI 0.547 to 0.001; one-sided p � 0.025) but had no effect on end-line LAZ or stunting (mean LAZ at end-line was −1.64). At midline, the intervention increased LAZ by 0.16 (95% CI −0.009 to 0.33; one-sided p � 0.032) and reduced stunting by 7% points (95% CI −0.125 to −0.007; onesided p � 0.015) but had no effect on serum AFB 1 levels [180]. It was inferred that the midline analysis suggested that AFs may affect linear growth at younger ages.
Overall average estimation of exposure rates based on annual consumption, as is appropriate for cancer risk because of the cumulative nature of this response, indicates that AF exposure was 3.5 to 14.8 ng/kg/day in Kenya for about 67% of the population [92,191]. No study in Kenya has examined the relationship between AFM 1 in breast milk samples and growth impairment in infants.

Coexposure to Aflatoxins with Other Mycotoxins.
Aflatoxin poisoning could be compounded by the occurrence of AFs in combination with other mycotoxins such as FUM, trichothecenes, ochratoxins, ZEA, and DON [16,192,193]. is is supported by the occurrence of mycotoxin producing fungi simultaneously in the same batch of food/matrix and the faculty of some toxigenic fungi to produce more than one mycotoxin in a given matrix. For example, Fusarium (F. verticillioides, F. proliferatum, and F. oxysporum) [54] and Penicillium spp were reported with Adapted from [187]. Aspergillus fungi in Kenya [41,45,53,84,115,194,195], sometimes in soils and mill dust around maize stores [40]. Fusarium spp are known for the production of FUMs [196]. e current review did not identify any reports evaluating coexposure to AFs in combination with other mycotoxins and the potential adverse health outcomes. ere have been developments in both mycotoxin-specific and multimycotoxin methods developed for biological matrices [197]. For example, mycotoxin-specific biomarkers for FUM in maize and DON in wheat have been developed and validated [198,199], though none has been applied in Kenya. In neighbouring Tanzania, coexposure to AFs with other mycotoxins utilizing individual biomarkers was recently investigated. Children (6-14 months old) were recruited at a maize harvest season and followed up twice at 6-month intervals. e children were reported to be chronically exposed to AFB 1 , FB 1 , and DON [200,201]. Blood AF-alb [200] and urinary DON levels [201] steadily increased over the 12 months, which likely corresponded to increased food intake that is possible as the child grows. A linear trend was not apparent for urinary FB 1 as the mean level at 6 months was significantly lower than mean levels at recruitment and at 12 months [200]. It was deduced that the lower exposure levels observed 6 months postharvest could be reflective of reduced maize stocks, resulting in lower maize consumption [200].
ough no significant correlation was appreciated between AF exposure and stunted child growth, increased FUM exposure was evidently associated with reduced length-for-age Z-scores [200].
In addition, coexposure to mycotoxins in utero is also wanting, as observations elsewhere reported AF-alb in 36% of the blood samples with urinary AFM 1 and DON present in 47% and 68% of samples from pregnant women in their third trimester coexposed to AFs and DON [202]. About 41% of the pregnant women were concurrently exposed to both AFs and DON. us, assessment of coexposure to AFs in Kenya with other mycotoxins is warranted.

Infantile Stunting due to Aflatoxin
Exposure in Kenya e first 1,000 days of life (from conception to about 36 months) is a critical window for healthy growth and development. Dietary intake of AFs during pregnancy plays a fundamental role in the child's future health status [187,197]. In sub-Saharan Africa, and particularly Kenya, malnutrition and child growth impairment are major public health burdens [80,118,180]. Intake of low, daily doses of AFs over long periods result in chronic aflatoxicosis expressed as impaired food conversion, stunting in children, immunosuppression, cancer, and reduced life expectancy [6,[203][204][205]. e WHO defined stunting as a height-for-age Z-score (HAZ), of <−2, being underweight as a weight-forage Z-score (WAZ), of <−2, and wasting as a weight-forheight Z-score (WHZ), of <−2 [206]. Stunting of infants in some aflatoxin-prone areas of Kenya is shown in Table 9.
It was advanced that AF exposure may disrupt the insulin-like growth factor (IGF) pathway through liver toxicity. In a study in Kenya [162], AF-alb concentrations were inversely associated with IGF1 levels (p � 0.039) and IGF binding protein 3 levels (p � 0.046) in a sample of 199 school children from Yumbuni in the west and Matangini (Lower Mangalete) in the east. A path analysis showed that lower IGF1 levels explained about 16% of the effect of AFs on child height (p � 0.052). Both IGF1 and IGFBP3 were significantly associated with child height and weight (p < 0.01). Children in the highest tertile of AF-alb exposure (>198.5 pg/mg) were shorter than those in the lowest tertile (<74.5 pg/mg), after adjusting for confounders (p � 0.043). To further investigate this putative mechanistic pathway, human hepatocyte line 16 (HHL-16) cells were treated with AFB 1 at 0.5, 5.0, and 20.0 μg/mL for 24-48 hours. IGF1 and IGFBP3 gene expression measured by quantitative PCR and protein in culture media showed a significant downregulation of IGF genes and reduced IGF protein levels. e study concluded that AF-induced changes in IGF protein levels could contribute to growth impairment where AF exposure is high [162].
Aflatoxin-child growth impairment may be explained by the immunosuppressive effect of AFs that aggravates neonatal infection susceptibility, thereby impairing nutritional status through inappetence and diminished nutrient absorption [207]. It is also argued that exposure to AFs may initiate intestinal damages by dysregulation of protein synthesis culminating in diminished nutrient absorption and impaired growth [208]. AFs have also been implicated in the etiology of other liver diseases including jaundice, cirrhosis, and hepatomegaly [185,209,210]. A study in Kenya by Gong et al. [185] reported that the prevalence of hepatomegaly, a firm form of liver enlargement, increased in children with higher AF exposure.
is is in complete agreement with the knowledge that the liver is the target metabolic organ for AFs.

Aflatoxicoses in Kenya
Since the discovery of AFs, Kenya has been one of the countries with devastatingly severe human exposure to AFs [109,211,212]. Exposure to AFs is chiefly via intake of contaminated food. Ingestion, however at concentrations greater than 6000 mg/kg in food, degenerates into liver failure and can be lethal after 1-2 weeks of exposure (acute aflatoxicosis) [213]. Aflatoxicosis is typified by oedema, convulsions, vomiting, jaundice, abdominal pain, sudden liver failure, and lastly death [214]. In practice, acute toxicities associated with exposure to elevated AF levels are not very common globally; cases occur and are concentrated in high-risk regions such as the Makueni County of Kenya [77] ( Table 10). In humans, acute toxicity due to exposure to high dietary doses of AFs (2,000-6,000 μg/day) in contaminated maize was reported in Western India in 1974 with a case fatality rate of 10% [215,216]. In Taiwan, 26 members of 3 families were victims of consumption of about 200 μg/kg of AFs in mouldy rice.
ree of the victims died [217]. In neighbouring Uganda, a 15-year-old boy also succumbed to death following ingestion of cassava containing 1,700 μg/kg of AFs, leaving behind a brother and a sister who survived very narrowly [218]. Recently, consumption of AF-contaminated maize triggered aflatoxicosis in humans with a case fatality rate of 50% in Tanzania [219].
In Kenya, aflatoxicosis was first witnessed in 1960 which recorded the death of at least 16,000 ducklings [82]. In 1981, Kenya witnessed its first serious recorded outbreak of human aflatoxicosis [177]. It was found that after 7 days of consumption of maize grain containing 3.2-12 mg/kg of AFB 1 , symptoms of abdominal discomfort, anorexia, general malaise, and low-grade fever were exhibited in 20 cases of patients between 2.5 and 45 years of age. Hepatic failure developed in 12 of the 20 patients, all of whom eventually died 1-12 days following hospital admission. e most unprecedented episode of human aflatoxicosis in history was witnessed in Kenya in 2004 with 317 reported cases of which 125 were fatal [39]. is outbreak, which occurred in the Eastern Province, recorded a case fatality rate of 39%, and out of the 308 patients for whom age data were available, 68 (22%) were <5 years, 90 (29%) were 5-14 years, and 150 (49%) were >15 years. Children younger than 14 years, representing 51% of the children population, were thus   [82] Years are those in which the aflatoxicoses occurred rather than the years the data were published. Data are from [38,92,97]. A case report is also filed of a possible aflatoxicosis of a 17-year-old schoolboy [214].
presumed to have had a greater predisposition to aflatoxicosis risk. e case fatality rate was significantly higher in Makueni district than in Kitui district [69,70,177,220,221]. Since 2004, outbreaks among subsistence farmers have recurred annually in Eastern Province and it is right to assert that the magnitude of exposure to AFs could be higher than reported due to the dearth of robust monitoring systems [63,109]. Table 10 summarizes some of the fatal aflatoxicoses recorded in the history of Kenya since the discovery of AFs. It is worth noting that several studies on AF poisoning in humans have shown that low-level chronic intake may be more devastating than one-time high-level intake (that leads to aflatoxicosis) as it is linked to the development of hepatocellular carcinoma [30,82,128,[222][223][224][225][226][227][228][229][230][231]. During the aflatoxicosis outbreak that occurred in 2010, the levels of AFB 1 sera reported in Kenya were among the highest ever recorded in the world [232]. As can be traced from Table 10, most areas that have been hit by aflatoxicosis in Kenya are in the Eastern and some Central parts of the country.

Prevention and Control
7.1. International, Regional, and Statutory Efforts. Appreciable efforts have been advanced towards AF control in Kenya through countrywide awareness creation [97,233]. e regional mycotoxin facility at the Kenya Agricultural and Livestock Research Institute (KALRO) in Katumani offers training to people from both the public and private sectors.
After the fatal aflatoxicosis in which dogs fed on contaminated rations died between 1970 and 1980s, KEBS came up with a standard for dog feeds in 1985. Standards for maize grain, other grains, and their products that have been in existence were also revised. For example, total AFs were initially at 20 μg/kg; this has been revised to 10 μg/kg, with 5 μg/kg as the threshold for AFB 1 in 2007 [234]. At least 25 standards aimed at regulating AFs have been drafted and are in full use and encompass key parameters such as moisture, mouldy grains, pest damage, filth, broken kernels/seeds, foreign matter, and discoloured grains. Most of these standards have been harmonized with the East African Standards by the Eastern Africa Grain Council (EAGC) in collaboration with KEBS through the Eastern Africa Grain Institute with its Kenyan headquarters at Nairobi, Kenya [235]. Between 2015 and 2018, the duo has trained maize exporters, traders, farmer-based organizations, and warehouse handlers on understanding the integrated East African maize standard (EAS 2:2013), food standardization, comparison of East African standards with international standards, standard maize sampling methods, maize grading, mycotoxins, and the available methods for mycotoxin analysis [235].
Following its launch way back in 2006, EAGC has been among the lead in the fight against AFs in East Africa as a whole. It has advanced several interventions to reduce the incidence of AFs, including (1) harmonization of AF control measures and improving the regulatory environment, (2) launch of AF control training programs, (3) furnishing moisture meters and waterproof sheets for drying, fumigation, and storing grains, (4) outsourcing portable kits for detecting and quantifying AFs, (5) farmer-based assessment of AFs prevalence, (6) collaborating with East African Community to expand AF testing and surveillance in maize, and (7) laying strategies of the Partnership for Aflatoxin Control in Africa (PACA) strategy 2013-2022 and revising EAC AFs communication strategy [235]. In addition, AF surveillance and capacity has been enhanced through the PACA Curated Africa Aflatoxin Information Management System (Africa-AIMS) in seven member states: Kenya, Malawi, Nigeria, Senegal, Tanzania, e Gambia, and Uganda.
Kenya Agricultural and Livestock Research Organization in connection with the International Institute of Tropical Agriculture (IITA) in 2018 developed a farmercentered manual for the management of AFs in maize and peanuts [233]. e manual gives a general overview of AFs (structures, health, and economic effects), how to control AFs, drying, threshing, sorting, and some of the farming practices that favour AF growth. It was particularly drafted to provide ample guidance on the best practices for limiting AF contamination of maize and peanuts and to raise the value of these dietary staples.
Further, there are some projects running in the country to handle the plague of mycotoxins and these include the Aflacontrol Project and Purchasing for Progress (P4P) Programme. e Aflacontrol Project strives to minimize the ravage of AFs in maize and peanut value chains and is spearheaded by International Food Policy Research Institute (IFPRI). In addition, it seeks to increase the understanding of the economic and health impacts of AF contamination and identify and promote cost-effective methods and technologies available to reduce contamination of foods and feeds. e project has been experimented in Mbere (Embu), Makueni, Homa Bay, Kisii, and Rongo at the household level [97]. So far, it has released policy briefs and held inceptions, and one-year national workshops to disseminate information on AFs.
ese are targeted at the Ministries of Agriculture and Public Health, who are the key players in mitigating AFs. On the other hand, the Purchasing for Progress Programme is led by the World Food Programme that purchases maize from local farmers, usually ensuring strict adherence to AF limits in the grains. e grains are procured at fairer prices, encouraging the farmers to adhere to good pre-, peri-, and postharvest practices [97]. Several partnerships are currently running in the country; some are with FAO and CDC to mitigate AFs in Kenya.
ese have been discussed in sufficient details in a previous review by Mutegi et al. [38].

Scholarly Efforts.
Earlier reports on the fate of AFs during the processing of maize into muthokoi, a traditional Kenyan food, revealed that traditional maize preparation methods such as fermentation and dehulling in Eastern Kenya reduced AFs by up to 71% [83]. e findings of this study indicate that exposure to acute AF levels could be minimized during food processing and preparation. Generally, these processing techniques have been traditionally used for increasing the palatability of different food recipes but should also be promoted as strategies capable of reducing AF contamination of grains [97].
Intermediate processes such as sorting and dehusking were shown to reduce AF in peanuts [243]. Soaking peanuts in water, magadi, sodium hypochlorite, and ammonium persulphate significantly reduced AF levels by 27.7%, 18.4%, 18.3%, and 1.6%, respectively, while boiling in magadi, local ash, baking powder, and water reduced AF levels by 43.8%, 41.8%, 28.9%, and 11.7%, respectively. Similarly, Kirui [183] while assessing the levels of residual AFs following various treatments using physicochemical and traditional cooking methods for maize and maize products reported that boiling maize reduced total AFs from 83.1 ± 0.3 to 7.0 ± 3.9 μg/kg, dry decortication reduced the level from 51.3 ± 15.3 to 9.6 ± 0.8 μg/kg, while boiling with magadi soda (or maize wood ash) reduced the level from 59.5 ± 3.82 to 13.4 ± 0.42 μg/kg. Solar irradiation (for 18 hours) reduced the levels from 60.8 ± 1.8 to 13.7 ± 0.1 μg/kg while ultraviolet irradiation (for 18 hours) reduced the levels from 81.7 ± 0.5 to 61.4 ± 4.5 μg/kg. e author reiterated that only the dry decortication method and boiling with magadi soda followed by washing with water and boiling reduced AFs significantly to below the maximum advisory limit of 10 μg/kg.
In the same struggle, a probiotic yoghurt was formulated with AFB 1 binding Streptococcus thermophilus, Lactobacillus rhamnosus GR-1, and Weissella cibaria NN20 isolated from fermented Kimere, a dough food product made from millet [179,182]. Forty primary school children, with maize being a regular part of their diet, were randomly assigned to consume 200 ml of yoghurt or control milk daily for 7 days, followed by a 7-day washout and another 7-day treatment. After both 7-day treatment periods, AF concentration in urine samples was significantly lower than baseline in the probiotic group (p > 0.01) but increased in the milk group.
is suggested that locally produced probiotic yoghurt could reduce AF poisoning in Kenyan children, corroborating previous observations in our laboratory [244,245]. Similarly, fermentation of milk into Lala, a traditional fermented drink, and yoghurt significantly reduced AFM 1 levels by 71.8% (in Lala after 15-hour room temperature incubation) and 73.6% in yoghurt after incubation at 45 C for 4 hours [135].
In another intervention survey, the use of a calcium montmorillonite clay (calcium silicate 100, popularized as ACCS100) in food reduced the bioavailability of AFs [181]. It was reported to be palatable, effective, and acceptable, though further evaluation in the AF-endemic parts of Eastern Kenya as well as its efficacy to ameliorate AFs to levels incapable of triggering poisoning yet remains to be established.
Another study [43] screened maize lines resistant to A. flavus infection, together with a biocontrol strategy. Two African maize lines (GAF4 and KDV1) were reported to have different fungal loads for the aflatoxigenic isolate (KSM014), fourteen days after infection, with no significant variation in A. flavus biomass between diseased and nondiseased maize tissues for GAF4. Meanwhile, KDV1 had a significantly higher A. flavus biomass (p < 0.05) in infected shoots and roots compared to the control. e biocontrol strategy using an atoxigenic isolate (KSM012) against the toxigenic isolate (KSM014) showed aflatoxin production inhibition at the coinfection ratio, 50:50 for both maize lines (KDV1 > 99.7% and GAF 69.4%), as confirmed by bioanalytical techniques. It was indicated that the maize lines, which exhibited resistance to A. flavus with the appropriate biocontrol strategy, could reduce aflatoxicosis outbreaks.
In a 2020 report [46], the possibility of using Pseudomonas and Bacillus bacterial spp was explored in soils from Eastern Kenya (semiarid) and Western Kenya (subhumidsemihumid) [110]. Pseudomonas (n � 7) and Bacillus (n � 5) were identified in the two regions, though Western Kenya recorded a higher occurrence of Bacillus. Because these bacterial spp have been frequently associated with biological control of several plant pathogens including Aspergillus spp, a regression analysis was done to ascertain if there were any associations between the occurrence of Aspergillus spp and these bacterial spp in the studied regions. Weak relationships between the occurrence of A. flavus and Pseudomonas spp in the western region (R 2 � 0.03693) and the eastern region (R 2 � 0.06126) as well as occurrence of Bacillus spp in the western region (R 2 � 0.196) and in the eastern region of Kenya (R 2 � 0.03693). e same observation was made for the relationship between the occurrence of Trichoderma viride in both eastern (R 2 � 003406) and western (R 2 � 0.2266) regions of the country. As a consequence, the authors deduced that the occurrence of the bacterial spp had little influence on the occurrence of A. flavus in the two regions. To ascertain their assertion, an in vitro preliminary assay to determine the inhibitory potential of both Pseudomonas and Bacillus spp against A. flavus proliferation was done. Unfortunately, none of the bacterial strains from either spp had an inhibitory effect on A. flavus proliferation [46].

Preharvest Management.
Staple food crop varieties that are disease-, drought-, and pest-tolerant or less susceptible to fungal growth could be adopted. is approach is so far the best for the reduction of effects of AFs-producing fungi [246]. Valencia red (a peanut variety) was reported to be the least contaminated with AFs and had higher oil content than Uganda local, Homa Bay local, and Local red [121]. Food oils and microorganisms are viable inhibitors of AF biosynthesis [6] through interference with signal transduction regulatory chains in AF gene expression, inhibiting AF biosynthetic cytosolic enzymes and downregulating fungal genes of the oxidative stress defence system [247]. Further, host and parasite macro-and micromolecular trafficking which targets circumvention of AF scourge through the utilization of cross species RNA interference have been attempted in maize and peanuts [248,249]. Particularly, UBI, COH, 26s, ATP, PPK, IMP, ABC, and aflM were recommended as the suitable genes for RNAi silencing of A. flavus in vivo [248,249]. is may, however, be impeded by the current policy on genetically modified organisms in the country.
Prompt harvest of mature food crops as well as selective disposal of broken kernels or cobs is a recommended AF mitigation measure [250]. Sorting, winnowing, and dehulling can lower AF levels in grains by 40-80% [151,251]. Sorting is more practical in groundnuts [250,[252][253][254] and cassava chips mechanically or utilizing clean water. Soaking or cooking in magadi soda, malting, and roasting are other strategies reported to lower food AF levels [83,251,255,256]. Magadi soda and wood ash are used by the Kalenjin of the Rift Valley region and Nyanza and Western provinces to increase food palatability and offer convenience as it reduces cooking time but also reduces phytates and increases the availability of niacin [257].
Pest control is another measure of AF management. is may be affected using ash for maize [258,259] and essential oils which are broad-spectrum bioinsecticides [260][261][262][263]. Competitive exclusion is another underexplored AF control measure. A shift of strain profile from toxigenic to atoxigenic is possible and Kenya already approved a potential biocontrol product (Aflasafe KE01) [8,233]. A typical example is a product constituted by a rhizosphere-competent nonaflatoxigenic Aspergillus strain possessing competitive saprophytic potential [264,265]. For peanuts, an atoxigenic A. flavus strain (NRRL 21882) is commercialized as Afla-Guard ® in the US [266]. Additionally, pseudomonads and Trichoderma spp inhabiting rhizospheres of various plants have been targeted against toxigenic A. flavus. Recently, various Streptomyces, Pseudomonas, and Trichoderma spp have been isolated, evaluated, and validated to possess antagonism towards A. flavus [267]. However, their efficacy is a theme that could be further explored under Kenyan conditions as was done in some African and Asian countries with up to 79% reduction of AFs reported [268].

Postharvest Management.
Proper drying of produce to moisture contents between 12 and 14%, preferably 12.5% or below, is recommended. Fresh harvests should be shelled and cleaned prior to storage to minimize pest infestation that may initiate mould growth [269]. Further, proper ventilation of storage facilities is a requisite to avoid attaining temperatures between 25°C and 32°C and sustained relative humidity of 65% favourable for mould growth [11]. Moisture of 12-13% and temperatures below 18°C do not favour the growth of Aspergillus fungi [270].
Following good agricultural, good storage, and good manufacturing practices as well as the use of advanced agricultural technologies can reduce AF contamination [152]. Novel food processing techniques such as use of ozone, pulsed light, electrolyzed water, electron beam, microwave, cold plasma, and gamma and ultraviolet irradiation can reduce AF concentrations in agricultural foods. Food additives such as citric acid have been reported to sequester AFs in combination with moisture at high pressures and temperatures [271].
Clays notably Novasil Plus can ably bind AFs [272], reducing their available concentration. Other compounds like curcumin can afford alteration of AFB 1 microsomal activation, thereby augmenting its detoxification. Chemoprotection from ingested AFs is also feasible [273]. It makes use of compounds such as esterified glucomanoses and other yeast extracts that accelerate the detoxification process or otherwise block bioactivation of AFs to AF-epoxide, inhibiting AFB 1 -induced hepatocarcinogenesis. Oltipraz and chlorophyll have reported use in this regard [273]. e management strategies suggested each have their advantages and limitations.
us, biocontrol measures in tandem with physicochemical approaches could be adopted to manage the plague of AFs in Kenya.

Conclusion
Aflatoxin exposure is ubiquitous in Kenya, and the different commodities have relatively high levels of AFs, usually above statutory compliance limits by several folds. Maize, peanuts, and their products are the most contaminated food crops in Kenya. Variations in AF exposure are evident between the different regions of the country and are fundamentally a function of diet and economic status. Large-scale, evidencebased interventions are required to reduce exposure. More exposure assessments including coexposure with other mycotoxins alongside routine monitoring of AFs should be adopted.

Data Availability
is article is a review article and no raw data were collected. Any data used and/or analyzed are within this article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.