Association between blood manganese and cognitive function in a national representative survey

Manganese (Mn) is an essential heavy metal, required for the normal development of many organ systems and is crucial to brain growth and development. At high exposure concentrations, Mn is a neurotoxicant. Mn exposure has been associated with motor, behavioral and cognitive impairment. Few studies have assessed the association between Mn exposure and cognitive function using NHANES data.

This study re-explored the association between blood manganese concentrations and cognitive function through a re-analysis of NHANES 2011-2014 data using a larger combined sample and composite cognitive z-score methodology.

A representative survey of the US population (60-80 years of age) completed objective cognitive function assessments and metal biomonitoring (n = 2,439). We created a composite cognitive z-score by using the average of the standardized scores of the six cognitive tests (three trials of CERAD Word List Learning Test, CERAD delayed recall, Animal fluency test and DSST) and compared to adjusted blood Mn. Linear regression was performed by constructing three models (univariate model, a model with all covariates and a model with only significant covariates). Confounding by gender, age, race/ethnicity, education level, total number of people in the family, smoking history and income (family monthly poverty level category) are included in the models.

Median Mn was 8.71 µg/L with an IQR of 6.94-11.16. There was a statistically significant association between quartiles of blood Mn and composite z-score, only in the univariate model. However, in linear regression analyses, adjusting for potential confounders, blood Mn as a continuous variable was not statistically associated with the composite cognitive z-score (µg/L, β = 0.0038, 95% CI -0.004 to 0.012). The mean of composite z-score in the first quartile is the lowest compared to the mean of the second, third and fourth, with statistical significance. While the mean of the third quartile is the highest, followed by the fourth and then the second (0.022, 0.012 and 0.007, respectively).

Our findings suggest that associations between blood manganese concentrations and cognitive performance at low environmental exposure levels are modest and attenuate after adjustment for demographic and socioeconomic confounders. Unlike some previous NHANES analyses that reported impaired cognition primarily among individuals with higher manganese exposure, we did not observe a consistent independent dose-response association across the broader population exposure range.