Association of serum microcystin levels with neurobehavior of school-age children in rural area of Southwest China: A cross-sectional study.

To investigate whether microcystin-LR (MC-LR) influences children's cognitive function and memory ability, we measured serum MC-LR and whole blood lead levels in 697 primary students, and collected their academic and neurobehavioral test scores. The median of serum MC-LR levels was 0.80 µg/L (the value below the limit of detection to 1.67 µg/L). The shapes of the associations of serum MC-LR levels (cut-point: 0.95 µg/L) with scores on academic achievements, digit symbol substitution test and long-term memory test were parabolic curves. Logistic regression analysis showed that MC-LR at concentrations of 0.80-0.95 µg/L was associated with the increased probability of higher achievements on academic achievements [odds ratio (OR) = 2.20, 95% confidence interval (CI): 1.28-3.79], and also with scores on digit symbol substitution test (OR = 1.73, 95% CI: 1.05-2.86), overall memory quotient (OR = 2.27, 95% CI: 1.21-4.26), long-term memory (OR = 1.85, 95% CI: 1.01-3.38) and short-term memory (OR = 2.13, 95% CI: 1.14-3.98) after adjustment for confounding factors. Antagonism of MC-LR and lead on long-term memory was observed (synergism index = 0.15, 95% CI: 0.03-0.74). In conclusion, serum MC-LR at concentrations of 0.80-0.95 µg/L was positively associated with higher scores on cognitive and neurobehavioral tests, and antagonism between MC-LR at concentrations of 0.80-1.67 µg/L and lead exposure was obviously observed on long-term memory in children. Concerning that MC-LR is a neurotoxin at high doses, our observation is interesting and need further investigation.