Seasonal variation but not processing methods differently affect cassava-derived cyanide exposure in two areas with different konzo prevalence in South-Kivu (D.R. Congo)

Abstract

Introduction: Konzo is a toxico-nutritional spastic paraparesis affecting poor African communities under severe food deprivation and predominantly consuming toxic cassava. In Eastern DRC, cassava-derived cyanide toxicity was documented during dry season in Burhinyi, a konzo-affected area, while Idjwi, an island known for toxic cassava contribution to endemic goiter, is free of konzo. We hypothesized differences in cassava processing efficiency and seasonal variability of cyanide content between the two areas. This observational transversal study aimed at comparing cassava processing and seasonal impacts on cyanide exposure in Burhinyi and Idjwi. Materials and Methods: Cassava cyanide content (HCN; fresh, processed roots and flour from household stock) and urinary thiocyanate (USCN; 55/32 healthy adults from Burhinyi/Idjwi, ~28 /35 y) were measured during dry season (DS) and raining season (RS) using picrate paper kits A and D1 (comparison using Wilcoxson rank-sum test [Median(IQR)]).

Results: After heap fermentation (3-4d, Burhinyi)/sun-drying (7-10d, Idjwi), high HCN (ppm) in fresh roots in DS ([100(50-200)]/[200(75-300)]; p=0.07) dropped by 70-90% to still unsafe HCN rates in both areas without modification in RS. However, flour HCN increased from RS to DS [5 (0–10) to 10 (5-20) ppm; p=0.02]/[10 (5–10) to 10 (10–30) ppm; p=0.008] without difference between the two areas. The USCN (µmol/l), from being comparably high in RS (~103), decreased during DS in Burhinyi [68(34.4-103.2); p=0,016], but increased in Idjwi [172(172-344); p=0,015] (p<0.001). Thus, seasonal variation did not affect cassava HCN the same way as USCN.

Conclusion: Surprisingly, both cassava processing methods appear comparably efficient whereas seasons differently affect HCN (similar increase in flour) and USCN (decrease in Burhinyi, but increase in Idjwi without konzo observation). Several explanations are possible for these observations: different dietary behaviors, underestimation of cassava-exposure in konzo-affected area due to measuring method-induced bias in populations with poor nutritional level as in Burhinyi.