Extent and drivers of mycotoxin contamination: inferences from a survey of kenyan maize mills.

The prevalence of aflatoxin and fumonisin was investigated in maize intended for immediate human consumption in eastern Kenya at a time in 2010 when an aflatoxin outbreak was recognized. Samples were collected from people who brought their maize for processing at local commercial mills. Sites were selected using a geographical information system overlay of agroecological zones and Kenya's administrative districts. Interviews and collection of maize flour samples was conducted from 1,500 people who processed maize at 143 mills in 10 administrative districts. Mycotoxins were analyzed using enzyme-linked immunosorbent assays for aflatoxin and fumonisin, leading to detection at levels above the respective maximum tolerable limits in 39 and 37% of the samples, respectively. Samples with aflatoxin contamination above the legal limit ranged between 22 and 60% across the districts. A higher occurrence of aflatoxin was associated with smaller maize farms, lower grain yield, and monocropping systems, while a larger magnitude of the toxin was observed in the subhumid agroecological zone, in samples with more broken kernels, and, curiously, less maize ear damage at harvest. Analysis of paired grain samples (visually sorted and unsorted) showed that sorting reduced fumonisin by 65%, from above to below the legal limit of 1,000 ppb. Sorting did not, however, reduce aflatoxin levels. Although the aflatoxin problem is widely acknowledged, the high prevalence of fumonisin has not previously been reported. There is need for surveillance of the two mycotoxins and establishment of intervention strategies to reach vulnerable small-scale farmers.