Association between thyroid hormones and cognitive functioning in euthyroid elderly adults: a cross-sectional preliminary study from the NHANES 2011-2012 survey.

OBJECTIVE

Changes in serum thyroid hormone levels may affect cognitive functioning in euthyroid individuals. This study used representative data from the National Health and Nutrition Examination Survey (NHANES) to comprehensively examine the association of thyroid hormones with different tests of cognitive functioning among US elderly people aged ≥60 years.

METHODS

This study was a cross-sectional preliminary study with a total of 734 participants from the NHANES 2011-2012 survey. Thyroid function was measured using competitive binding immune-enzymatic assays, while cognitive functioning was measured using a series of assessments, including the Consortium to Establish a Registry for Alzheimer's Disease-Word Learning (CERAD W-L), Animal Fluency Test, and Digit Symbol Substitution Test (DSST). Weighted multiple linear regression models and binary logistic regression analyses were used to examine the association between thyroid hormone levels and cognitive functioning. All statistical analyses were performed using SPSS version 20.0, and R software.

RESULTS

Weighted multivariable linear regression showed that FT3 was negatively associated with the Animal Fluency Test and DSST (β=-0.113, 95% CI: -3.279, -0.803, P=0.001; β=-0.062, 95% CI: -6.565, -0.470, P=0.024, respectively) after adjustment for potential covariates. subgroup analysis stratified by sex revealed a negative association between FT3 levels and the Animal Fluency Test in men (β=-0.163, 95% CI: -4.643, -1.153, P=0.001). For female participants, FT3 was negatively associated with not only the Animal Fluency Test but also DSST (β=-0.099, 95% CI: -3.543, -0.093, P=0.039; β=-0.093, 95% CI: -10.288, -1.326, P=0.011). Binary logistic regression showed that the significantly increased adjusted odds ratios (aORs) (95% CI) between the risk of impaired cognitive functioning and FT3 across Q3 and Q4 compared with Q1 were 2.025 (1.092, 3.753) and 2.365 (1.261, 4.433), respectively, for DSST in overall participants. Furthermore, there were significant differences between participants with and without impaired cognitive functioning for serum FT3 levels in overall participants based on DSST score (P=0.020).

CONCLUSIONS

There was a significant inverse relationship between FT3 levels within the normal range and cognitive functioning after adjusting for potential covariates. Future longitudinal cohort studies should be conducted to determine the causal relationship between thyroid hormone levels and cognitive functioning.