[Association between thyroid hormone levels and body mass index, waist circumference and body composition in older adults aged 65 years and above in 18 longevity areas in China].

FT3 level was positively associated with BMI, waist circumference, muscle mass, and fat mass in older adults after controlling for confounding factors.

• For each increment of 1 pmol/L in FT3 level, the mean change in BMI was β=0.64 (95%CI: 0.44-0.83)
• For each increment of 1 pmol/L in FT3 level, the mean change in waist circumference was β=1.21 (95%CI: 0.66-1.76)
• For each increment of 1 pmol/L in FT3 level, the mean change in muscle mass was β=0.22 (95%CI: 0.03-0.41)
• For each increment of 1 pmol/L in FT3 level, the mean change in fat mass was β=1.44 (95%CI: 1.05-1.84)
• Multiple linear regression models were used with adjustment for confounding factors

FT4 level was negatively associated with muscle mass in older adults.

• For each increment of 1 pmol/L in FT4 level, the mean change in muscle mass was β=-0.15 (95%CI: -0.19 to -0.10)
• This association was a negative linear association (P for linear <0.001)
• No significant associations were reported between FT4 and BMI, waist circumference, or fat mass

Restricted cubic spline analysis revealed positive nonlinear associations between FT3 and all body composition measures.

• FT3 showed positive nonlinear associations with BMI, waist circumference, muscle mass, and fat mass
• All P values for nonlinearity were <0.05
• FT4 showed only a negative linear association with muscle mass (P for linear <0.001)
• No nonlinear association was found between FT4 and muscle mass

The study sample consisted of 4,566 older adults with a mean age of 82.6 years from 18 longevity areas in China.

• Mean age was (82.6±10.5) years
• 47.5% of participants were men
• Mean FT3 level was (4.2±0.6) pmol/L
• Mean FT4 level was (15.4±2.2) pmol/L
• Data were drawn from the 2021 China Healthy Ageing and Biomarkers Cohort Study using a cross-sectional design

Stratified analyses were performed to examine potential effect modification by gender, age, and fish intake.

• Stratified analyses were conducted according to gender, age, and fish intake of the older adults
• The study used multiple linear regression and restricted cubic spline models as analytical methods
• The study was based on cross-sectional data from 18 longevity areas in China