Efficacy of mycotoxin binder on aflatoxin M1 and Mazzican on total bacterial count in raw milk among smallholder dairy farmers in Kisumu County, Kenya

Anyango, G., Mutua, F., Kagera, I., Andang'o, P., Grace, D. and Lindahl, J. 2019. Efficacy of mycotoxin binder on aflatoxin M1 and Mazzican on total bacterial count in raw milk among smallholder dairy farmers in Kisumu County, Kenya. Presentation at the Kenyatta University International Food Safety Conference, Nairobi, Kenya 20–24 May 2019. Nairobi, Kenya: ILRI.

Background: Worldwide, milk forms a major part of human diet especially for children and expectant mothers. However, milk in East Africa has been shown to be frequently contaminated with aflatoxin M1 (AFM1) and bacteria. The efficacy of NovaSil clay (NS), a mycotoxin binder, and Mazzican, a hygienic milk container, to reduce levels of aflatoxin M1 and total bacterial counts respectively were evaluated in raw milk samples collected from urban and peri urban areas of Kisumu, Kenya. Materials and methods: A baseline survey was conducted in 97 farms, and 72 raw milk samples were collected. Thirty farms identified as producing milk with AFM1 levels above 50 ppt, which is the recommended limit by EU, were randomized to either receive the intervention (including training, NS and Mazzican, n=20), or not receive intervention (control, n=10). NS was to be administered in 1-2 teaspoonful per 2 kg of concentrate feed while Maziccan was to be used during milking, milk storage and transportation to the market. Raw milk samples from trial participants were collected biweekly for three consecutive months. AFM1 metabolites were quantified by enzyme linked immunosorbent assay with a detection limit of 2 ppt. Total bacterial counts were obtained from colony counts of raw milk cultures. Results: Upon baseline testing, aflatoxin M1 levels in raw milk ranged from 0.001-151 ppt with a mean 29.40 ppt. The mean total bacteria count was 1.6×10⁷ ± 5.2×10⁷Cfu/ml. Cfu/ml ranged from 3.3×10² to 2.4×10⁸ Cfu/ml. Trial results indicate that giving 2 teaspoon of NS binder per 2 kg caused a significant statistical difference in reduction of AFM1 levels in the sampled milk, p=0.013, whereas this was not found when giving half the dose. Use of mazzicans revealed a statistical significance difference in reduction of total bacterial counts in the milk samples p<0.001. The study has 100% compliance in use of mazzican while 98% in NS use. Conclusions: Intervention effectively reduced AFM1 and bacterial counts in milk. The farmers were willing to invest in the intervention, therefore the possibility to supply farmers with NS and mazzican should be further investigated.