Urinary cadmium concentration and incident cognitive impairment: the REGARDS study

Abstract

AbstractBackgroundStudies have indicated neurotoxicity of cadmium (Cd) exposure. However, little is known about Cd exposure in relation to cognitive impairment.MethodA total of 2,172 participants (mean age: 64.1±9.0 years old; female: 54.8%; Black: 38.7%), without cognitive impairment or stroke at baseline (2003‐2007), from a random sub‐cohort (n = 2,666) of the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study, were included in this study. Urinary creatinine‐corrected Cd concentrations were measured at baseline. The Enhanced Cognitive Battery (ECB) and Six Item Screener (SIS) were performed every two years (ECB) or annually (SIS) during an average of 10 years of follow‐up. The ECB consisted of 4 tests: Word list learning (WLL), Delayed Recall (WLD), Animal Fluency, and Letter F Fluency, and its composite score was used to determine our first outcome: cognitive impairment by ECB. Because of missing data on ECB scores (n = 482), we also defined a second outcome for cognitive impairment by assigning them a cognitive impairment status based on SIS. Therefore, in the secondary analysis, the outcomes consisted of three levels: ECB, SIS, and control. Odds ratios (OR) and corresponding 95% confidence intervals (CI) were examined using multivariable‐adjusted logistic regression for the ECB‐defined cognitive impairment or using multinomial logistic regression for the second outcome.ResultDuring the follow‐up, 53 participants developed cognitive impairment as assessed by the ECB and 36 cases were determined by the SIS. After adjustment for potential confounders, no significant association was observed between the urinary Cd concentration and incident cognitive impairment. Comparing the highest tertile of Cd concentration to the lowest, the OR (95% CI) of incident cognitive impairment was 1.07 (95% CI: 0.43, 2.67; P for trend = 0.87) for the cases measured by ECB (Table 1). In the analysis of multinomial logistic regression, the OR (95% CI) was 1.15 (95% CI: 0.47, 2.81; P for trend = 0.97) for ECB and 1.04 (95% CI: 0.34, 3.22; P for trend = 0.32) for SIS (tertile 3 vs. tertile 1) (Table 2).ConclusionFindings from this US cohort do not support the hypothesis that Cd exposure is associated with the risk of cognitive impairment.