To study the ecological variation of intake of cassava foods and dietary cyanide load.

Ecological study design.

Five communities in south-western Nigeria where tropical ataxic neuropathy (TAN) was described as endemic (area A), 11 communities in south-western Nigeria where TAN was described as absent (area B), and five communities in northern Nigeria (area C).

Subjects were randomly sampled from selected communities. Intake of cassava foods was estimated from dietary history and dietary cyanide load was estimated from urine thiocyanate concentrations. Residual cyanogens in cassava food samples from the community markets were determined.

In total, 1272 subjects from 21 communities – 238 from area A, 659 from area B and 375 from area C – were selected. Intake of cassava food per person per week was 17 meals in area A, 10 meals in area B, and one meal in area C. Geometrical mean urine thiocyanate concentrations were 73 μmol l−1, 51 μmol l−1 and 17 μmol l−1 in areas A, B and C, respectively. Mean residual cyanogen content in cassava food samples was 16 mg HCN eq kg−1 (confidence interval (CI) 13–18) in area A, and 13 mg HCN eq kg−1 in area B (CI 11–14).

This study shows that the intake of cassava foods and dietary cyanide load is high in several communities in south-western Nigeria, predominantly in communities where TAN has been reported. Dietary cyanide load in these communities appears to be determined by the combination of frequency of intake and cyanogen content of cassava foods. Measures to improve the effectiveness of removal of cyanogen from cassava roots during processing are needed in the affected communities.