OR22-04 Relationship of TSH Levels with the Components of Metabolic Syndrome in a Nationally Representative Population of Youth in the United States

Introduction: Subclinical hypothyroidism (SH) is defined as elevated TSH with normal thyroid levels, and is often associated with obesity. SH has been linked to cardiometabolic risk factors such as abnormal lipids, elevated blood pressure, atherosclerosis and fatty liver. This study sought to elucidate the association of TSH level with the components of metabolic syndrome independent of BMI in children from the National Health and Nutrition Examination Survey (NHANES). Methods: NHANES surveys 1999-01 and 2007-12 that measured thyroid function tests were included in the study. Youth aged 2-18 years with TSH levels < 10 uU/mL and normal Total T4 (TT4) levels were included in the analysis. The components of metabolic syndrome were defined as abdominal obesity (waist circumference > 95(th) %tile), hypertriglyceridemia (TG >=100 for 0-9 years and >=130 mg/dL for > 10 years), low HDL cholesterol < 40 mg/dL), elevated blood pressure (> 95(th) %tile for age/sex/height) and hyperglycemia (FBG > 100 mg/dL, or diagnosis of diabetes). The association of these components with quartiles of TSH were examined by logistic and linear regression controlling for age, sex, race/ethnicity and BMI. All analyses were performed in R v3.5.1. Results: After excluding youth with TSH >10 uU/mL and TT4 levels < 12.4 mcg/dL, 2377 subjects (50% female) were included in the study. The mean age of the cohort was 15 ± 1.7 years; 28.2 % were non-hispanic whites and 38.5 % hispanic/latino. Obesity (BMI >95 %tile) was seen in 21.7% individuals. There were 44 subjects with TSH levels >4.5 uU/mL that was not different by BMI (2.5% in BMI >95%tile and 1.7% BMI < 95%tile, p = 0.29). Based on the distribution in the population, TSH levels were divided into 4 quartiles: Q1= 0.01-0.97, Q2= 0.98-1.42, Q3=1.43-2.0, Q4 = > 2.01 uU/mL. A statistically significant association of the Q4 TSH was seen with abdominal obesity, OR 2.44 (1.38-4.39), p=0.002 and elevated BP, OR 1.6 (1.06-2.44), p = 0.02 but not with high TG, OR 1.58 (0.93-2.75), p=0.09, low HDL, OR 0.84 (0.6-1.17), p = 0.31 or those with hyperglycemia and/or diabetes, OR = 1.25 (0.78-2.05), p = 0.36. Linear regression models showed statistically significant association of abdominal obesity, hypertriglyceridemia, elevated BP and hyperglycemia (and/or diabetes) with increase in TSH level. Conclusions: In children from a representative US population, the prevalence of SH defined as TSH level >4.5 uU/mL is low, even with BMI >95(th) %tile. The association of measures of metabolic syndrome with linear increase in TSH suggests that the current reference range may require modification.