Context: Epidemiological studies have supported the premise that an adequate selenium intake is essential for thyroid gland function.
Objective: The objective was to investigate whether the prevalence of thyroid disease differed in two areas that were similar, except for very different soil/crop selenium concentrations.
Design: Cross-sectional observational study.
Setting: The setting was two counties of Shaanxi Province, China, here defined as adequate- and low-selenium.
Participants: A total of 6152 participants were selected by stratified cluster-sampling.
Main outcome measures: Participants completed demographic and dietary questionnaires and underwent physical and thyroid ultrasound examinations. Serum samples were analyzed for thyroid function parameters and selenium concentration. Serum selenium was compared between different demographic, dietary, and lifestyle categories in the two counties. The relationship between selenium status, dietary factors, and pathological thyroid conditions was explored by logistic regression.
Results: Complete data sets were available from 3038 adequate-selenium participants and 3114 low-selenium participants in whom median (interquartile range) selenium concentrations differed almost 2-fold (103.6 [79.7, 135.9] vs 57.4 [39.4, 82.1] μg/L; P = .001). The prevalence of pathological thyroid conditions (hypothyroidism, subclinical hypothyroidism, autoimmune thyroiditis, and enlarged thyroid) was significantly lower in the adequate-selenium county than in the low-selenium county (18.0 vs 30.5%; P < .001). Higher serum selenium was associated with lower odds ratio (95% confidence interval) of autoimmune thyroiditis (0.47; 0.35, 0.65), subclinical hypothyroidism (0.68; 0.58, 0.93), hypothyroidism (0.75; 0.63, 0.90), and enlarged thyroid (0.75; 0.59, 0.97).
Conclusions: Low selenium status is associated with increased risk of thyroid disease. Increased selenium intake may reduce the risk in areas of low selenium intake that exist not only in China but also in many other parts of the world.