ABSTRACT
The objective of this study was to investigate the relationship between selenium content in hair and the incidence of Kashin-Beck disease (KBD) and Keshan disease (KD) in China. A prospective cohort study was conducted among children aged 5-12 years with different levels of low-selenium (group 1, Se ≤ 110 ng/g; group 2, 110 < Se ≤ 150 ng/g; and group 3, 150 < Se ≤ 200 ng/g) or selenium-supplemented (group 4, Se > 200 ng/g) exposure. A person-years approach was used to calculate the incidence and rate of positive clinical signs. Relative risk (RR), attributable risk, and etiologic fraction were used to determine the strength of association between selenium and disease incidence. Seven new KBD cases were diagnosed during 3-year follow-up. Positive clinical signs of KBD were found in 17.78 (95% confidence interval [CI] 14.27-21.29) cases per 100 person-years in group 1, 13.28 (9.82-16.74) in group 2, 12.95 (9.34-16.56) in group 3, and 8.18 (5.50-10.85) in group 4. Compared with group 4, the RR (95% CI) of groups 1, 2, and 3 were 2.17 (1.48-3.19), 1.62 (1.07-2.47), and 1.58 (1.03-2.43), respectively. Positive clinical signs of KD were 25.90 (18.62-33.18) cases per 100 person-years in group 1, 5.66 (1.26-10.06) in group 2, 4.60 (0.20-9.00) in group 3, and 14.62 (8.54-20.69) in group 4. Compared with group 4, the RR (95% CI) were 1.77 (1.07-2.93), 0.39 (0.16-0.93), and 0.31 (0.11-0.89), respectively. In children, the onset of KBD was negatively correlated with selenium content within a certain range. However, there may be a U-shaped association between selenium content and KD in children.