Recent heritability analyses have indicated that genome-wide association studies (GWAS) have the potential to improve genetic risk prediction for complex diseases based on polygenic risk score (PRS), a simple modelling technique that can be implemented using summary-level data from the discovery samples. We herein propose modifications to improve the performance of PRS. We introduce threshold-dependent winner’s-curse adjustments for marginal association coefficients that are used to weight the single-nucleotide polymorphisms (SNPs) in PRS. Further, as a way to incorporate external functional/annotation knowledge that could identify subsets of SNPs highly enriched for associations, we propose variable thresholds for SNPs selection. We applied our methods to GWAS summary-level data of 14 complex diseases. Across all diseases, a simple winner’s curse correction uniformly led to enhancement of performance of the models, whereas incorporation of functional SNPs was beneficial only for selected diseases. Compared to the standard PRS algorithm, the proposed methods in combination led to notable gain in efficiency (25–50% increase in the prediction R²) for 5 of 14 diseases. As an example, for GWAS of type 2 diabetes, winner’s curse correction improved prediction R² from 2.29% based on the standard PRS to 3.10% (P = 0.0017) and incorporating functional annotation data further improved R² to 3.53% (P = 2×10⁻⁵). Our simulation studies illustrate why differential treatment of certain categories of functional SNPs, even when shown to be highly enriched for GWAS-heritability, does not lead to proportionate improvement in genetic risk-prediction because of non-uniform linkage disequilibrium structure. Large GWAS have identified tens or even hundreds of common SNPs significantly associated with individual complex diseases; however, these SNPs typically explain a small proportion of phenotypic variance. Recently, heritability analyses based on GWAS data suggest that common SNPs have the potential to explain substantially larger fraction of phenotypic variance and to improve the genetic risk prediction. Because of the polygenic nature, improving genetic risk prediction for complex diseases typically requires substantially increasing the sample size in the discovery set. Thus, it is crucial to develop more efficient algorithms using existing GWAS summary data. In this article, we extend the polygenic risk score (PRS) method by adjusting the marginal effect size of SNPs for winner’s curse and by incorporating external functional annotation data. Theoretical analysis and simulation studies show that the performance improvement depends on the genetic architecture of the trait, sample size of the discovery sample set and the degree of enrichment of association for SNPs annotated as “high-prior” and the linkage disequilibrium patterns of these SNPs. We applied our method to the summary data of 14 GWAS. Our method achieved 25–50% gain in efficiency (measured in the prediction R²) for 5 of 14 diseases compared to the standard PRS.

The use of humidifier disinfectant (HD) has been determined to be associated with lung injuries (HDLI) in Korea. Although HD brands containing polyhexamethylene guanidine (PHMG) oligomers have been found to cause more HDLI compared to brands containing other disinfectants, the physicochemical properties of PHMG have been poorly defined. We aimed to quantify the PHMG dissolved in HD brands, characterize the number-average (M_n) and weight-average (M_w) molecular masses, and identify the polymerization degree of PHMG. Analysis of the PHMG oligomers was performed using a matrix-assisted laser desorption/ionization time-of-flight mass spectrometer (MALDI-TOF MS) operated in positive-ion reflectron mode. Eight brands of HD containing PHMG were identified. The PHMG concentrations in these brands ranged from 160 to 37,200 ppm (mean = 3100.9 ppm). Concentration was a significant variable among and within HD brands. The degree of PHMG oligomerization fell within the range of two to four. The averages of M_n and M_w were 517.2 g/mol (range: 422–613 g/mol) and 537.3 g/mol (range: 441.0–678.0 g/mol), respectively. Based on the average molecular weight and the degree of polymerization, the PHMG examined here could be regarded as oligomers, which may be associated with the highest proportion of HDLI being caused by PHMG.

This study aimed to assess the characteristics of exposure to both PM_2.5 and black carbon (BC) among subway workers. A total of 61 subway workers, including 26, 23, and 12 subway station managers, maintenance engineers, and train drivers, respectively, were investigated in 2018. Real-time measurements of airborne PM_2.5 and BC were simultaneously conducted around the breathing zones of workers. Maintenance engineers had the highest average levels of exposure to both PM_2.5 and BC (PM_2.5, 76 µg/m³; BC, 9.3 µg/m³), followed by train drivers (63.2 µg/m³, 5.9 µg/m³) and subway station managers (39.7 µg/m³, 2.2 µg/m³). In terms of the relationship between mass concentrations of PM_2.5 and BC, train drivers demonstrated the strongest correlation (R = 0.72), indicating that the proportion of BC contained in PM_2.5 is relatively steady. The average proportion of BC in PM_2.5 among maintenance engineers (13.0%) was higher than that among train drivers (9.4%) and subway station managers (6.4%). Univariate and mixed effect multiple analyses demonstrated the type of task and worksite to be significant factors affecting exposure levels in maintenance engineers and subway station managers. The use of diesel engine motorcars in tunnel maintenance was found to be a key contributor to PM_2.5 and BC exposure levels among subway workers.

This study aimed to quantify both chloromethylisothiazolinone (CMIT) and methylisothiazolinone (MIT) dissolved in different product brands and to characterize the exposure to these chemicals among humidifier disinfectant-associated lung injury (HDLI) patients. Both CMIT and MIT dissolved in different humidifier disinfectant (HD) products were quantified using gas chromatography–mass spectrometry. The inhalation level of CMIT and MIT was estimated based on HD-associated factors as reported by HDLI patients. A total of eleven HD products marketed until the end of 2011 were found to contain CMIT and/or MIT. The level of combined CMIT and/or MIT dissolved in these HD products ranged from 12 to 353 ppm. The level varied among HD products and the year of manufacture. The average inhalation levels were estimated to be 7.5, 4.1, and 3.2 μg/m³ for the definite, probable, and possible groups, respectively. If probable and possible groups were collapsed together, the inhalation level of the collapsed group was significantly different from that of the definite group (p < 0.001). All HDLI patients responded as having used HD not only while sleeping, but also as having a humidifier treated with HD within close proximity every day in insufficiently ventilated spaces. These HD use characteristics of patients may be directly/indirectly linked to the HDLI development.

Objectives: This study explored the relationship between shift intensity and insomnia among hospital nurses.Methods: The participants were 386 female hospital nurses who underwent a special health examination for night workers in 2015. The Korean Insomnia Severity Index (ISI), indices of shift work intensity, and other covariates such as amount of exercise, level of alcohol consumption, employment duration, and hours worked were extracted from the health examination data. The indices for shift intensity were (1) number of 3 consecutive night shifts and (2) number of short recovery periods after a previous shift, both assessed over the prior 3 months. Multiple logistic regression analysis adjusted for the aforementioned covariates was performed to evaluate the association of shift intensity with insomnia, defined as an ISI score of >= 8.Results: The nurses with insomnia tended to be younger (p=0.029), to have worked 3 consecutive night shifts more frequently (p<0.001), to have experienced a greater number of short recovery periods after the previous shift (p=0.021), and to have worked for more hours (p=0.006) than the nurses without insomnia. Among the other variables, no statistically significant differences between groups were observed. Experiences of 3 or more consecutive night shifts (odds ratio [OR], 2.33; 95% confidence interval [CI], 1.29 to 4.20) and 3 or more short recovery periods (OR, 2.01; 95% CI, 1.08 to 3.73) were associated with increased odds of insomnia.Conclusions: The results suggest that decreasing the shift intensity may reduce insomnia among hospital nurses working rotating shifts.

In a previous study, we estimated exposure prevalence and the number of workers exposed to carcinogens by industry in Korea. The present study aimed to evaluate the optimal exposure intensity indicators of airborne lead exposure by comparing to blood lead measurements for the future development of the carcinogen exposure intensity database. Data concerning airborne lead measurements and blood lead levels were collected from nationwide occupational exposure databases, compiled between 2015 and 2016. Summary statistics, including the arithmetic mean (AM), geometric mean (GM), and 95th percentile level (X95) were calculated by industry both for airborne lead and blood lead measurements. Since many measurements were below the limits of detection (LODs), the simple replacement with half of the LOD and maximum likelihood estimation (MLE) methods were used for statistical analysis. For examining the optimal exposure indicator of airborne lead exposure, blood lead levels were used as reference data for subsequent rank correlation analyses. A total of 19,637 airborne lead measurements and 32,848 blood lead measurements were used. In general, simple replacement showed a higher correlation than MLE. The results showed that AM and X95 using simple replacement could be used as optimal exposure intensity indicators, while X95 showed better correlations than AM in industries with 20 or more measurements. Our results showed that AM or X95 could be potential candidates for exposure intensity indicators in the Korean carcinogen exposure database. Especially, X95 is an optimal indicator where there are enough measurements to compute X95 values.