Incidence and consumer awareness of toxigenic Aspergillus section Flavi and aflatoxin B1 in peanut cake from Nigeria

Peanut cake samples were collected from major markets in five states of Nigeria and evaluated for incidence of toxigenic Aspergillus section Flavi populations, and aflatoxin B₁ (AFB₁) levels by liquid chromatography tandem mass spectrometry (LC–MS/MS). The awareness of consumers to the presence of aflatoxin in the snack and potential health risks of its regular ingestion was evaluated by questionnaire analysis. Aspergillus section Flavi populations were recovered from 83% of the peanut cake samples. Aspergillus flavus L-strain was the most predominant (>56%) across the states while Aspergillus tamarii had the least mean incidence (2.7%). The incidence of atoxigenic strains was significantly (p < 0.05) higher than that of toxigenic strains in samples from Lagos and Kaduna, while the toxigenic strains had a significantly (p < 0.05) higher incidence than the atoxigenic strains in Niger. All analyzed cake samples contained AFB₁ in concentrations exceeding the NAFDAC recommended level for AFB₁ in food and reaching up to 2824 μg/kg. There was a weak positive correlation (r = 0.32, p = 0.03) for the relationship between the incidence of toxigenic strains in the samples and AFB₁ concentration. The consumer awareness data showed that 64% of the respondents consumed peanut cake; majority of who are youth of economic and reproductive age. Eighty-five percent of the consumers lacked awareness of aflatoxin contamination in the snack and possible health risks associated with its ingestion.     

Survey of aflatoxins and ochratoxin A in retail market chilies and chili sauce samples

The aim of this study was to determine the levels of the aflatoxins (AFs) and ochratoxin A (OTA) in 312 samples of whole chili, chili powder, crushed chili and chili sauce samples from suburbs, open market and food restaurants of Punjab, Pakistan. The analysis was carried out using HPLC with fluorescence detector, after immunoaffinity column clean-up. The results have shown that 176 out of 312 (56.4%) samples were to be positive with AFs and 126 out of 312 (40.4%) sample of chilies were found to be contaminated with OTA. Total mean level of AFB₁ and total AFs in chilies were 12.50 ± 1.91, 15.16 ± 2.22 μg/kg, respectively. The total mean level of OTA in chilies was found 16.68 ± 2.58 μg/kg, ranged from LOD to 120.9 μg/kg. Sample 39.5, 26.3 and 32.7% of chilies were found containing level of AFB₁, total AFs and OTA, higher than the recommended limits for EU, respectively. The dietary levels of 3.26, 3.52, and 3.84 μg/kg were determined for AFB₁, total AFs and OTA in chilies. The incidence and levels of AFs and OTA in chilies are higher and could pose serious health hazards for consumers.     

Mycoflora and absence of aflatoxin contamination of commercialized cassava chips in Benin, West Africa

Studies conducted in Benin, in which the main staple foods are maize, cassava, groundnuts and yams, showed high levels of aflatoxin residues in blood of the exposed population. The natural contamination with fungi and aflatoxins in cassava chips sold at markets in Benin, West Africa was investigated. A total of sixty samples were sampled from open markets in 11 districts of 3 agroecological zones and analyzed for the presence of mycoflora and aflatoxin B₁, B₂, G₁ and G₂. Fourteen genera of fungi were associated with marketed dried cassava chips. Within these, twenty- two isolates were identified to species level, whereas four were identified only to genus. The dominating fungal species isolated were Rhizopus oryzae, Nigrospora oryzae, Chrysonilia sitophila, Cladosporium resinae, Cladosporium herbarum, Apergillus niger and Aspergillus flavus. Fifty-four out of sixty samples were contaminated with A. flavus. The rate of occurrence in CFU/g of A. flavus fungi was lower than for all other fungal species together. Aflatoxin was not detected in any of the samples analyzed using HPLC with post-column photochemical derivatization and fluorescence detection. The limit of detection (LOD) was 0.1 μg/kg. Results from this study suggest cassava chips are unlikely to be a source of aflatoxin in Benin, and that other staples such as maize and groundnuts are more important in aflatoxin exposure. Therefore it can be speculated that staples like maize and groundnut are more important in aflatoxin exposure.     

Aflatoxin B1 production by Aspergillus parasiticus and strains of Aspergillus section Nigri in currants of Greek origin

Aflatoxin B₁ (AFB₁) mostly produced by Aspergillus flavus and Aspergillus parasiticus, is an extremely toxic and carcinogenic metabolite. Currants are used in the Mediterranean diet as a food with antioxidant properties. Four strains of Aspergillus section Nigri have been isolated from currants originated from Crete and Corinth. In this study AFB₁ production by A. parasiticus and the four strains of Aspergillus section Nigri in Cretan and Corinthian currants (Vitis vinifera L.) is investigated. AFB₁ determination was performed by HPLC–FID. Results revealed that the four strains Aspergillus section Nigri, as well as the aflatoxigenic strain A. parasiticus produced AFB₁ (0.0052–1.31 μg AFB₁ 15 g⁻¹, corresponding to 0.0003–0.087 μg AFB₁ g⁻¹) in both type of currants (Cretan and Corinthian) on the 12th day of observation. Moreover, AFB₁ production, by A. parasiticus in the synthetic Yeast Extract Sucrose (YES) medium was also studied. The ability of AFB₁ production has been affected by the special characteristics of each isolate and the currants substrate.     

Aflatoxins in rice – A limited survey of products marketed in Austria

Eighty-one rice samples were purchased from different markets in Vienna and were analysed for their aflatoxin content. The samples were extracted using methanol in water (80/20 v/v) followed by immunoaffinity clean up. The determination was carried out by HPLC–FLD coupled to a Kobracell. Different samples including basmati rice, whole grain rice, long grain rice, short grain rice as well as puffed rice were investigated. Moreover, conventionally and organically produced rice were compared. The results revealed that 24 out of 81 samples contained detectable amounts of aflatoxins. Aflatoxin B₁ could be quantified in 15 samples and aflatoxin B₂ in one sample. The contamination range was noted to be between 0.45 μg kg⁻¹ and 9.86 μg kg⁻¹ for aflatoxin B₁ and 1.5 μg kg⁻¹ for aflatoxin B₂. Aflatoxins G₁ and G₂ were not detected in any sample. Three samples exceeded the maximum levels set in the European Union; having AFB₁ concentrations of 2.16, 2.85 and 9.86 μg kg⁻¹. In the three organic produced rice samples only traces of aflatoxins were found.     

A survey of mycotoxins in random street-vended snacks from Lagos,Nigeria, using QuEChERS-HPLC-MS/MS

A survey in African snacks was carried out in order to evaluate the intake of 23 mycotoxins. The African snack samples were purchased from street vendors within Lagos metropolis (Nigeria) and evaluated for the presence of 23 mycotoxins using a modified QuEChERS procedure coupled with liquid chromatography-triple quadrupole linear ion trap mass spectrometer. The snacks included akara, baked coconut, coconut candy, donkwa, groundnut cake (kulikuli), lafun, milk curd (wara), fresh and dried tiger-nuts, and yam flour. Only three mycotoxins were detected in 23.8% of the studied snacks, and at concentrations ranging from 6 to 54 μg kg⁻¹. The concentrations of aflatoxin B₁ (AFB₁) and AFB₂ reached 23 μg kg⁻¹ and 3 μg kg⁻¹, respectively. Moreover a sample of baked coconut contained α-zearalenol (α-ZOL), which was up to 54 μg kg⁻¹ in coconut candy. As considers prevalence, aflatoxins and α-ZOL were not detected in lafun and groundnut-based snacks (donkwa and kulikuli), whereas each of the three mycotoxins contaminated 12.5% (1/8) of the coconut-based samples. This is the first report of α-ZOL in cassava and coconut, and their products. AFB₁ and total aflatoxins (TAFs) concentrations exceeded the maximum allowable limit recommended by National Agency for Food and Drug Administration and Control Nigeria (NAFDAC) in one sample of baked coconut (AFB₁ = 23 μg kg⁻¹ and TAFs = 26 μg kg⁻¹) and donkwa (AFB₁ = 19 μg kg⁻¹ and TAFs = 21 μg kg⁻¹).     

Natural occurrence of aflatoxin B1, ochratoxin A and citrinin in Croatian fermented meat products

When domestic animals are exposed to mycotoxins, significant amounts of the latter shall be carried over into animal products such as milk, eggs and meat. This study was carried out in order to determine the possible presence of aflatoxin B₁ (AFB₁), ochratoxin A (OTA) and citrinin (CIT) in game sausages (n = 15), semi-dry sausages (n = 25) and fermented dry-meat products (n = 50), randomly taken from individual producers and the Croatian market. AFB₁ and OTA were quantified using ELISA, while CIT was quantified using HPLC-fluorescence detector. Out of 90 samples, the fungi most frequently isolated from dry-cured meat products were of Penicillium species, while Aspergillus was isolated from only one sample. As much as 68.88% of the samples were positive for mycotoxins. Finally, the analysis of different types of meat products resulted in OTA identification in 64.44%, CIT identification in 4.44% and AFB₁ identification in 10% of the samples. The maximum OTA concentrations established in the commercial sausage samples equalled to 7.83 μg/kg, while that of AFB₁ amounted to 3.0 μg/kg. Generally, although OTA was detected in all three types of products in different percentage shares, mutual differences were not statistically significant (P > 0.05).     

Aflatoxigenic fungi and aflatoxins occurrence in sultanas and dried figs commercialized in Brazil

The presence of aflatoxins, Aspergillus flavus and Aspergillus parasiticus in dried fruits was investigated. A total of 62 dried fruit samples were analyzed (24 black sultanas, 19 white sultanas and 19 dried figs). A total of 10 A. flavus isolates were found, nine in one white sultana sample (corresponding to 18% infection) and one isolate in dried figs (2%), and all of them were aflatoxin B₁ and B₂ producers. A. parasiticus was not found. Aflatoxins were detected in 3 of 19 (16%) white sultana samples analyzed and, the limits were not higher than 2.0 μg/kg. In dried figs 11 of 19 (58%) samples were contaminated with aflatoxins and, with exception of one sample that was contaminated with 1500 μg/kg of B₁ aflatoxin, the others had less than 2.0 μg/kg. Neither aflatoxigenic or aflatoxins contaminated black sultanas.     

Development and validation of an accurate and rapid LC-ESI-MS/MS method for the simultaneous quantification of aflatoxin B1, B2, G1 and G2 in lotus seeds

An accurate and rapid LC-ESI-MS/MS analytical method was developed and validated for the simultaneous determination of aflatoxin B₁, B₂, G₁ and G₂ in lotus seeds. The samples were firstly extracted with methanol-water solution (80:20, v/v), and then cleaned up by immunoaffinity columns. The mass spectrometer was operated in the positive ionization electrospray (ESI+) mode using multiple reaction monitoring (MRM) for analysis of four aflatoxins. The transitions of m/z 313 → 285 (AFB₁, CE 33 eV), m/z 315 → 259 (AFB₂, CE 37 eV), m/z 329 → 243 (AFG₁, CE 37 eV) and m/z 331 → 257 (AFG₂, CE 37 eV) were used to quantify these four aflatoxins, respectively. The limits of detection (LODs) of aflatoxin B₁, B₂, G₁ and G₂ were 0.007, 0.005, 0.003 and 0.005 μg kg^?1 based on a signal-to-noise ratio of 3:1, respectively. The limits of quantification (LOQs) of aflatoxin B₁, B₂, G₁ and G₂ were 0.02, 0.015, 0.01 and 0.015 μg kg^?1 based on a signal-to-noise ratio of 10:1, respectively. Recoveries for samples of spiked lotus seeds were all above 66% with relative standard deviation all below 15% for all compounds. Nineteen out of twenty batches of lotus seeds collected from different drug stores or markets in China were found to be contaminated with aflatoxins at different levels ranging from 0.02 to 688.4 μg kg^?1.     

Survey of aflatoxin B1 and ochratoxin A occurrence in traditional meat products coming from Croatian households and markets

The aim of this study was to investigate the occurrence of aflatoxin B₁ (AFB₁) and ochratoxin A (OTA) in different traditional meat products circulating on Croatian markets, produced by a large number of households situated in different Croatian regions. The study involved a total of 410 samples of traditional pork meat products in terms of hams (n = 105), dry fermented sausages (n = 208), bacon (n = 62) and cooked sausages (n = 35), collected over four years period (2011–2014). Mycotoxin concentrations were quantified and confirmed using validated immunoassay method (ELISA) and high performance liquid chromatography with fluorescence detection (HPLC-FLD), respectively. The maximal observed OTA level in the fermented sausages and hams was around 5 times (5.10 μg/kg) to 10 times (9.95 μg/kg) higher than the maximal recommended level (1 μg/kg) stipulated for pork products in some EU countries. AFB₁ levels found in any given meat product analysed within this frame were not significantly higher (p > 0.05) than the applied method limit of detection. The results showed an occasional mycotoxin contamination of traditional meat products, especially that by OTA, pointing that to avoid such contamination meat and meat products on households should be produced and processed under standardised and well-controlled conditions.     

A rapid FT-NIR method for estimation of aflatoxin B1 in red chili powder

The feasibility of measuring Aflatoxin B₁ (AFB₁) in red chili powder was investigated by using Fourier transform near-infrared (FT-NIR) spectroscopy in diffuse reflectance mode combined with appropriate chemometric techniques. Aflatoxin free chili powder samples were spiked with known amount of AFB₁ ranging from 15 to 500 μg/kg and used for calibration model building based on partial least squares (PLS) regression algorithm. Different spectral preprocessing methods were investigated and optimized based on the lowest values of root mean square error of cross validation (RMSECV). Spectral wavenumber range of 6900.3–4998.8 and 4902.3–3999.8 cm^?1 and straight line subtraction preprocessing technique predicted AFB₁ content with best accuracy with lowest RMSECV = 0.654% and maximum correlation coefficient for validation plots (R² = 96.7). The overall results demonstrate that FT-NIR spectroscopy can be used for rapid, non destructive quantification of Aflatoxin B₁ in red chili powder.     

Aflatoxin M1 survey on randomly collected milk powder commercialized in Argentina and Brazil

Aflatoxin M₁ (AFM₁), the main monohydroxylated of aflatoxin B₁ (AFB₁) formed in liver and excreted in the milk, has toxicological proprieties comparable to those of AFB₁, albeit a lower carcinogenic potency. The presence of AFM₁ was investigated in 30 samples of powdered milk purchased in Argentina and Brazil. The samples were analyzed using an immunoaffinity column for cleanup and HPLC-FLD for determining AFM₁. The quantification limit was 0.1 μg/kg. AFM₁ was found in all the samples at levels ranging from 0.1 to 0.92 μg kg⁻¹ with average concentration of 0.39 μg kg⁻¹.     

Inactivation of aflatoxigenic fungi and the reduction of aflatoxin B1 in vitro and in situ using gamma irradiation

Detailed investigation on the effect of gamma (γ) irradiation on germination, sporulation, and growth of aflatoxigenic moulds (Aspergillus parasiticus 2999, Aspergillus flavus 305, and Aspergillus niger 388), as well as on the reduction of aflatoxin B₁ (AFB₁) level in artificially and naturally contaminated maize/feed samples was performed. The results of in vitro and in situ experiments with aflatoxigenic moulds demonstrated that 5 kGy-γ irradiation manages to prevent sporulation, germination and growth of the tested moulds both when in form of a pure and when in form of a mixed culture. In the feed samples artificially contaminated with AFB₁ (50 μg kg⁻¹) 5 kGy-γ irradiation reduced AFB₁ level by around 60%, while 10 kGy-dose reduce it for around 85%. Similarly, in feed samples spiked with AFB₁ in the concentrations of 100 μg kg⁻¹ 5 kGy-dose reduced the AFB₁ level by approximately 70%, while the dose of 10 kGy reduced it by approximately 90%. The experiments on naturally contaminated maize samples (n = 30) confirmed these observations; following a 5 kGy-irradiation, the overall mean AFB₁ reduction equalled to 69.8%, while the irradiation with a 10 kGy-dose achieved the overall mean toxin reduction of 94.5%. The obtained results indicate that γ irradiation can be used to prevent the growth of aflatoxigenic moulds and to reduce the AFB₁ levels in various goods intended for animal and human consumption, thus minimizing the animal and human exposure to this carcinogenic mycotoxin.     

Effectiveness of pulsed light treatment for degradation and detoxification of aflatoxin B1 and B2 in rough rice and rice bran

Aflatoxins primarily accumulate in the hull and bran layers of rough rice making these by-products of rice milling unsuitable for animal feed or human consumption. Contaminated rough rice is also a potential source of aflatoxin exposure to workers handling the grain during post-harvest storage and processing. Currently, no technologies are available to remove or detoxify these toxic and mutagenic fungal metabolites from contaminated rough rice. Pulsed light (PL) is a novel technology with the potential to degrade and detoxify aflatoxins in foods and their processing by-products. Rough rice was inoculated with Aspergillus flavus to produce aflatoxin B₁ (AFB₁) and B₂ (AFB₂) contamination, followed by PL treatments of 0.52 J/cm²/pulse for various durations. A PL treatment time of 80 s reduced AFB₁ and AFB₂ in rough rice by 75.0% and 39.2%, respectively; while a treatment time of 15 s reduced AFB₁ and AFB₂ in rice bran by 90.3% and 86.7%, respectively. Since PL treatments result in the degradation of aflatoxins in situ, the toxicity and mutagenic activity of the residual by-products of AFB₁ and AFB₂ after PL treatment were evaluated. Toxicity was estimated using the brine shrimp (Artemia salina) lethality assay and mutagenicity measured by the fluctuation test with Salmonella typhimurum tester strains TA98 and TA100. The mutagenic activity of AFB₁ and AFB₂ was completely eliminated by PL treatment, while the toxicity of these two aflatoxins was significantly decreased. The obtained results suggest that PL technology has a promising potential to degrade, detoxify, and inactivate the mutagenic activity of aflatoxins in rough rice and rice bran.     

Seasonal prevalence level of aflatoxin M1 and its estimated daily intake in Pakistan

Aflatoxin M₁ level was measured in 520 milk samples during autumn, winter, spring and summer seasons of 2013–14 in five districts of Southern Punjab-Pakistan. Analyses were performed by using enzyme linked immuno sorbent assay (ELISA) method. Aflatoxin M₁ was found positive in 93% milk samples in the range of 0.001–0.26 μg/l while 53% samples were found to exceed the European Union maximum Limit for aflatoxins M₁ i.e. 0.05 μg/l. The seasonal prevalence level of AFM₁ was found in the order of winter > spring > autumn > summer. Comparing the aflatoxin M₁ level during different day times, morning milk was found 37–50% more contaminated than evening milk. The estimated daily intake (EDI) of aflatoxins M₁ during different seasons of year for various age groups was found in the range of 0.22–5.45 ng/kg/day. Infants were found in the highest risk group while adults were in the lowest. The results of the study indicate that people of Pakistan are at high risk of health issues related with aflatoxins M₁.     

Aflatoxin B1 occurrence in maize sampled from Croatian farms and feed factories during 2013

The aim of the study was to determine the level of aflatoxin B₁ (AFB₁) in maize sampled from farms and feed factories situated in Northern, Central and Eastern Croatia during 2013, following the occurrence of cow milk AFM₁ contamination. Maize samples (n = 633) were analysed using Enzyme-Linked Immunosorbent Assay (ELISA) as a screening method and High Performance Liquid Chromatography Tandem Mass Spectrometry (LC–MS/MS) as a confirmatory method. Mean AFB₁ value found in maize coming from all investigated regions equalled to 81 μg/kg, with the maximal value of 2072 μg/kg found in maize obtained from Eastern Croatia. The observed contamination might have arisen on the grounds of extremely hot (>98%) and dry (<2%) weather witnessed from May to September 2012 during the maize growth and harvesting period, which might have favoured AFB₁ production and consequently the contamination of dairy cattle feeds. In order to prevent the adverse effects of AFB₁ on humans and animals, and also to reduce losses in agricultural production, systematic monitoring and further investigations of AFB₁ contamination are necessary.     

Evaluation of latex agglutination inhibition reaction test for rapid detection of aflatoxin B1

Aflatoxin B₁ (AFB₁), a mycotoxin mainly produced by some Aspergillus species, has been found in a wide range of agricultural products. To avoid the risk of AFB₁ consumption, many agricultural commodities, foods and feeds should be analyzed and a rapid and non-instrumental method for the detection of AFB₁ is needed. In this study, a rapid, inexpensive and user-friendly latex agglutination inhibition reaction test (LAIRT) for on site testing of AFB₁ had been established. At first, carboxylated polystyrene latex particles (CPLP) were prepared by soap-free polymerization and sensitized with aflatoxin B₁ oxime-BSA (AFB₁O-BSA) for the detection of AFB₁. In LAIRT, the agglutination reaction with AFB₁O-BSA-sensitized CPLP and anti-AFB₁ antiserum mixture was inhibited by 5 ng/mL AFB₁ and the analysis time for 6 samples on one glass slide was less than 10 min. Subsequently, 10 rice and peanut samples were analyzed by LAIRT and ELISA, and the results showed that 1 rice sample and 2 peanut samples were positive and in agreement with those of ELISA. However, the results could be obtained more rapidly by LAIRT than ELISA. This easy and rapid LAIRT might be useful for screening AFB₁ of agricultural commodities, foods and feeds in the field.     

Fast and sensitive aflatoxin B1 and total aflatoxins ELISAs for analysis of peanuts,maize and feed ingredients

Aflatoxins are a group of carcinogenic compounds produced by Aspergillus fungi that can grow on different agricultural crops. Both acute and chronic exposure to these mycotoxins can cause serious illness. Due to the high occurrence of aflatoxins in crops worldwide fast and cost-effective analytical methods are required for the identification of contaminated agricultural commodities before they are processed into final products and placed on the market. In order to provide new tools for aflatoxin screening two prototype fast ELISA methods: one for the detection of aflatoxin B₁ and the other for total aflatoxins were developed. Seven monoclonal antibodies with unique high sensitivity and at the same time good cross-reactivity profiles were produced. The monoclonal antibodies were characterized and two antibodies showing IC₅₀ of 0.037 ng/mL and 0.031 ng/mL for aflatoxin B₁ were applied in simple and fast direct competitive ELISA tests. The methods were validated for peanut matrix as this crop is one of the most affected by aflatoxin contamination. The detection capabilities of aflatoxin B₁ and total aflatoxins ELISAs were 0.4 μg/kg and 0.3 μg/kg for aflatoxin B₁, respectively, which are one of the lowest reported values. Total aflatoxins ELISA was also validated for the detection of aflatoxins B₂, G₁ and G₂. The application of the developed tests was demonstrated by screening 32 peanut samples collected from the UK retailers. Total aflatoxins ELISA was further applied to analyse naturally contaminated maize porridge and distiller's dried grain with solubles samples and the results were correlated with these obtained by UHPLC-MS/MS method.     

Aflatoxin and fumonisin contamination of yam flour from markets in Nigeria

The natural occurrence of aflatoxins (AFs) and fumonisins (FBs) in yam flour samples (n = 100) obtained in south-western Nigeria was evaluated. AFs were determined by HPLC with fluorescence detection and FBs by HPLC coupled with mass spectrometry. Aflatoxin B₁ (AFB₁) and aflatoxin G₁ (AFG₁) were found in 57% and 21% of flours from white yam with concentrations ranging from <0.02 (limit of detection, LOD) to 3.2 μg kg⁻¹ (mean = 0.4 μg kg⁻¹) and from <0.05 to 3.5 μg kg⁻¹, respectively. AFB₁ was the only aflatoxin detected in samples from water yam, contaminating 32% of the samples with values ranging from <LOD to 0.6 μg kg⁻¹ (mean = 0.1 μg kg⁻¹). Fumonisin B₁ was found in 32% of the white yam samples (<0.5 (LOD) to 91 μg kg⁻¹; mean = 5 μg kg⁻¹) and in 5% of water yam samples (<LOD to 2 μg kg⁻¹). AFs and FBs were significantly higher (P < 0.05) in white yam flours compared to water yam flours. Preparation of amala from naturally-contaminated yam flour resulted in reduction of AFB₁ and AFG₁ by 44% and 51% respectively. From this study, only 7% of the samples contained AFs above the European standard limits for cereals intended for direct human consumption, while all the FBs-positive samples were well below the limits. The occurrence of ochratoxin A, zearalenone and deoxynivalenol was also evaluated in 20 samples; these mycotoxins were never detected.     

Safety of a street vended traditional maize beverage,ice-kenkey,in Ghana

Ice-kenkey is a chilled cereal beverage sold as street food in some open markets in Ghana. It is produced by mashing and sweetening kenkey, a stiff dumpling produced from fermented maize meal. The safety of street vended ice-kenkey was assessed by microbiological, elemental and myco-toxicological analysis of ice-kenkey and intermediary products obtained from 16 producers in four open markets in the Accra and Tema metropolis. A tenfold increase in counts of aerobic mesophiles, and yeast and moulds were recorded during the production of ice-kenkey. Coliform bacteria, E. coli and Staphylococcus aureus which were not detected in the starting materials were found partway through production or in the final product. The mean microbial counts in the packaged ice-kenkey were 10⁶–10⁷ CFU/g for aerobic mesophiles, 10⁴–10⁵ CFU/g for yeast and moulds, 10–1000 CFU/g for total coliforms and 10–100 CFU/g for S. aureus. E. coli counts of 10 CFU/g were recorded in samples from three out of the four markets. The microbial load could be eliminated by pasteurizing ice-kenkey at 80 °C for 15 min. The mean concentration in mg/kg of Fe was between 15.97 and 29.48, Cu, 0.57 to 1.41, Mn, 0 to 2.55, Pb 0 to 1.25 and Zn 0.47 to 6.17. Total aflatoxins content in samples ranged from 7.04 to 22.17 μg/kg and included a range of 7.01–20.54 for aflatoxin B₁, 0.51 to 1.63 for aflatoxin B₂ and 0–0.47 μg/kg for aflatoxin G_I. Aflatoxin G₂ was not detected in any of the samples. A simplified training module based GMP, GHP and a HACCP plan was developed and used to train ice-kenkey producers in Accra in collaboration with municipal authorities.