Thallium, a highly toxic heavy metal and priority pollutant, has been widely reported to cause neurodevelopmental toxicity in animals. However, accessible epidemiological studies concerning the neurodevelopmental toxicity of early-life thallium exposure in humans are limited. In a prospective birth cohort including 2164 mother-child pairs, we explored the effect of prenatal serum thallium exposure on cognitive development among preschool-aged children born in Ma'anshan, Anhui, China. Serum thallium concentrations were measured in the first trimester, second trimester, third trimester, and cord blood by inductively coupled plasma mass spectrometry (ICP-MS). Child cognitive development was appraised by the Chinese version of the Wechsler Preschool and Primary Scale of Intelligence-Fourth Edition (WPPSI-IV) at 4.5 years old. Multiple informants generalized estimating equations (GEEs) were fit to jointly estimate the association between the four repeated measurements of thallium concentrations and the preschool-aged children's cognitive test scores. After adjusting for potential confounders, the visual spatial index (VSI) was 1.45 points lower in the highest tertile of serum thallium during the first trimester than in the lowest tertile (p for trend = 0.04). Moreover, children in the highest tertile of serum thallium during the third trimester had a significantly lower full-scale intelligence quotient (FSIQ) (β = -1.51, 95% CI: -2.68, -0.35), VSI (β = -1.79, 95% CI: -3.16, -0.42), fluid reasoning index (FRI) (β = -1.41, 95% CI: -2.73, -0.10), and processing speed index (PSI) (β = -1.47, 95% CI: -2.71, -0.24) scores than the children in the lowest tertile. When performing stratified analysis by child sex, the associations of first- and third-trimester thallium concentrations with cognitive test scores were more prominent in boys than in girls. Our findings revealed that maternal serum thallium exposure during the first and third trimesters, but not other periods, had detrimental effects on preschoolers' cognitive development, and these effects showed sex differences.

Heat shock protein 90 (HSP90) is an important glucocorticoid receptor (GR) chaperone protein, and is supposed to be the key factor in regulating glucocorticoids (GCs) effects. The aim of the present study was to explore whether single nucleotide polymorphisms (SNPs) within HSP90AA1 gene affect the response of systemic lupus erythematosus (SLE) patients to GCs treatment. Two hundred and forty-five SLE patients were treated with GCs (prednisone) for 12 weeks. SLE disease activity index (SLEDAI) was used to assess the response of SLE patients to GCs treatment, and patients were classified into sensitive group and insensitive group. HapMap database and Haploview software were used to select tag SNPs. Tag SNPs were genotyped by using multiplex SNaPshot method. Univariate and multivariate logistic regression analyses were used to discriminate the impact of SNPs of HSP90AA1 gene on the response of SLE patients to GCs treatment. Two hundred and thirty three SLE patients finished the 12-week follow-up. Of these patients, 128 patients were included in sensitive group, and 105 patients were included in insensitive group. Seven tag SNPs were selected within HSP90AA1 gene. We detected significant associations for rs7160651 (dominant model: crude OR 0.514, 95 % CI 0.297–0.890, P = 0.018; adjusted OR 0.518, 95 % CI 0.293–0.916, P = 0.024), rs10873531 (dominant model: crude OR 0.516, 95 % CI 0.305–0.876, P = 0.014; adjusted OR 0.522, 95 % CI 0.304–0.898, P = 0.019) and rs2298877 (dominant model: crude OR 0.543, 95 % CI 0.317–0.928, P = 0.026, adjusted OR 0.558, 95 % CI 0.323–0.967, P = 0.037) polymorphisms, but not for other polymorphisms (P > 0.05). The present study demonstrates that HSP90AA1 gene SNPs may affect the response of SLE patients to GCs treatment.