MON-604 Reference Ranges for Thyroid Stimulating Hormone and Free Thyroxine in an Older Rural South African Population

Abstract

Background: Despite recognition that TSH and free T4 (fT4) ranges are population-specific, little data exist on these ranges in sub-Saharan African (SSA) populations. Evaluation of thyroid function is therefore based on reference ranges that may not be representative, potentially resulting in misdiagnosis. We hypothesized that population-specific reference ranges for TSH and fT4 in an older, rural South African population would differ from those cited by the assay manufacturer. Methods: 1498 people, 40-80 years, were randomly selected from the population-based Health and Ageing in Africa: a Longitudinal Study of an INDEPTH Community in South Africa (HAALSI) cohort in Mpumalanga province in rural South Africa, an iodine-sufficient country. 3rd generation TSH, fT4 and antithyroperoxidase antibodies (antiTPO Ab) were measured using chemiluminescent immunoassays. A disease-free population without self-reported thyroid disease was identified (n=1466). Non-pregnant, HIV-negative individuals with antiTPO Ab <35 IU/ml were then selected, with further exclusion of those with overt hypo/hyperthyroidism (TSH>4μIU/ml & fT4<0.89 ng/dl or TSH<0.4 μIU/ml & fT4>1.76 ng/dl) and with potential subclinical dysfunction (normal fT4 with TSH<0.1 μIU/ml or >10 μIU/ml) to form a reference population (n=681). Median, 2.5th and 97.5th percentile values of TSH and fT4 were determined for the overall reference population and for subgroups defined by age and sex. Population-specific ranges were used to classify the disease-free population. Results: Median age of the reference population was 61 years, with 51.5% males. Median TSH was 1.27 μIU/ml and median fT4 was 1.04 ng/dl. Population-specific TSH range was 0.43-3.86 μIU/ml (manufacturer 0.4-4.0 μIU/ml). There were no significant differences in age (p=0.061) or gender distribution (p=0.696), but those aged 70-80 had an upper reference limit of 5.03 μIU/ml. fT4 range was 0.70-1.47 ng/dl (manufacturer 0.89-1.76 ng/dl), with distribution of fT4 differing by sex (p=0.025) and age (p<0.001). 3.3% of the disease-free population had overt or subclinical hypothyroidism and 3.1 % had overt or subclinical hyperthyroidism. Conclusion: In one of the first population-based studies of TSH and fT4 reference ranges in SSA, we demonstrate that, in those <70, the upper limit of TSH and ranges for fT4 in a rural South African population are lower than manufacturer-specified values, mimicking findings in African-American (AA) populations, though in contrast to AAs, the lower limit of TSH in our population was higher than the manufacturer value. Hypothyroidism may therefore be underdiagnosed when TSH alone is used to screen for thyroid dysfunction. Further work is needed to define normal ranges in specific South African populations, such as pregnant women, who may be more vulnerable to the effects of thyroid dysfunction. Funding: FIC/NIH; NIA/NIH, NHGRI/NIH, OD/NIH