http://hdl.handle.net/10027/20 UIC School of Public Heath Tue, 21 May 2013 13:09:02 GMT 2013-05-21T13:09:02Z http://hdl.handle.net/10027/9797 The Stress Process among African American and Immigrant Russian-speaking Home Care Aides Work-related stress and burnout are significant problems for home care aides (HCAs) who help disabled older Americans with housekeeping and personal care in their homes. Despite the diversity in this workforce, few studies have examined whether HCAs with various ethnic and racial backgrounds experience the stressful caregiving work differently. The goal of this mixed-method study was to understand and compare the stress process leading to burnout among African American and Russian-speaking HCAs who constitute the majority of HCAs in Chicago, Illinois. In Phase I, we conducted focus groups with African American and Russian-speaking HCAs to explore the interplay among occupational and life stressors, health and burnout. In Phase II, using survey data of African American and Russian-speaking HCAs, we tested the factor structure of burnout via multi-group confirmatory factor analysis (MCFA) and conducted hierarchical regression analysis comparing the levels of work-related burnout in two groups. The focus groups revealed that while African American and Russian-speakers had similar experience of work-related stress, stressors in HCAs’ personal lives differed across groups. The MCFA supported unidimensionality of work-related burnout and measurement invariance across the two racial groups. Thus, a composite scale of work-related burnout was used in the subsequent regression analysis. Russian-speaking HCAs had higher levels of burnout. However, after taking into consideration the higher levels of education of Russian-speaking HCAs, age, gender, and kin relationship with clients, no group differences remained. Differences in education accounted for most of the group differences in burnout. Interestingly, after taking into consideration job stressors, being African American was associated with higher levels of burnout. Not surprisingly, emotional demands, work time pressures, and unpredictable work environment were associated with higher burnout, and supervisory support with lower burnout. African American and Russian-speaking HCAs differed in the stress process largely due to differences in levels of education and stressors in personal life. Future interventions should address HCAs’ stress-related issues not only in their work environment but also in other areas of their lives. Thu, 21 Feb 2013 06:00:00 GMT http://hdl.handle.net/10027/9797 2013-02-21T06:00:00Z http://hdl.handle.net/10027/9789 Polybrominated Diphenyl Ethers in Human Placentas in the United States This thesis provides a large data set on the levels of PBDEs in human placenta in the United States under the National Children’s Study (NCS),. A total of 43 placentas were collected at University of Rochester (UR), University of California Davis (UCD), and Medical College of Wisconsin (MCW). Each placenta was sampled at different collection times up to 96 hours after the delivery, resulting in a total of 169 tissue samples. The median of the total BDEs (BDE 28+33, 47, 66, 85, 99, 100, 153, 154, 183, 209) in the 42 placentas is 330 pg/g wet wt (42.6–1723 pg/g wet wt). The total concentration of tri- to heptaBDEs is lower than that first reported by Dassanayake et al. in 2009, but the concentration of BDE 209 is 56% higher. The levels of tri- to heptaBDEs, as well as BDE 209, are significantly higher than those reported from Europe and Japan. The PBDE levels approximately follow a log-normal distribution. Among the ten congeners measured, BDE 47 is the most abundant, followed by BDEs 153, 99, 100, and 209. The congener distribution pattern is similar in placentas collected from all three collection sites. Among the three sites, the concentration of the total BDEs from UCD is statistically significantly higher than that from UR and MCW at p = 0.1 level. With regard to collection time effect, the percent change in the total BDEs is in the range of -9.0% to 15.8% up to 72 hours after the initial sample collection. Storing a placenta for 96 hours has led to more significant changes in PBDE levels. Mixed effects models with placenta chosen as a random effect to count for its uniqueness were developed in this thesis. The models demonstrate no linear relationship between the total BDEs concentration and the collection time. In addition, no interaction between collection time and location is observed. These findings along with the large data set provide opportunities to further study the association between prenatal exposure to PBDEs and health effects in children. Thu, 21 Feb 2013 06:00:00 GMT http://hdl.handle.net/10027/9789 2013-02-21T06:00:00Z http://hdl.handle.net/10027/9786 Groundwater Arsenic in Chimaltenango, Guatemala In the community of Cerro Alto, in the municipality of Chimaltenango, Guatemala, we collected water samples in 2009 from two wells. Arsenic concentrations found were greater than four times the World Health Organization (WHO) provisional guideline value for arsenic in drinking water of 0.01 mg/L. As a result, we returned to work with local public health authorities and characterized the presence of arsenic in the groundwater of the entire municipality. A total of 42 samples were collected from 27 of the 62 (43.5%) wells in the municipality in January 2012. Upon analysis, the only site found to have a concentration above the 0.01 mg/L WHO guideline was Cerro Alto, where the average concentration was 0.0475 mg/L, similar to our initial finding in 2009. Arsenic contamination of groundwater can often be attributed to a naturally high presence in the area. Using data from the US Geological Survey and our GPS data of the sample locations, we found Cerro Alto to be the only site sampled within the tertiary volcanic rock layer, the likely source of arsenic in Guatemala due to the linkage between volcanic rock and naturally occurring arsenic. A risk assessment based on the arsenic levels found in Cerro Alto showed an increase in noncarcinogenic and carcinogenic risk for residents there. Recommendations are described to eliminate or lower the levels of arsenic found in the community’s drinking water. Thu, 21 Feb 2013 06:00:00 GMT http://hdl.handle.net/10027/9786 2013-02-21T06:00:00Z http://hdl.handle.net/10027/9760 Model Selection in Multivariate Analysis with Missing Data A new model selection algorithm based on maximizing penalized likelihood function with the smoothly clipped absolute deviation (SCAD) penalty function is developed for missing data problems. The algorithm can be implemented effectively by a one-step algorithm when the number of the variables is much smaller than the sample size. A modified tuning parameter criterion based on Bayesian Information Criterion (BIC) for missing data problems is proposed to select the optimal tuning parameter for the penalty function. One advantage of the proposed approach over the current available one is to use the observed data log-likelihood so that it asymptotically selects the true model when missing data mechanism is assumed to be Missing at Random (MAR). A new model selection scheme that not only selects covariates for the outcome variable but also selects covariate models, which are important in high-dimensional covariates subject to missing values, is also proposed. The proposed algorithm is implicitly applied to linear regression models and logistic regression models. Gauss-Hermite Quadrature and Monte Carlo Simulations are used to compute the intractable integrations in the Expectation-Maximization (EM) algorithm. Several simulations are carried out to examine the performance of the proposed algorithm compared with other available variable selection methods for missing data. A real data from a case-control study to investigate potential risk factors of hip fracture is used to illustrate the application of the proposed method. Including interaction effects, several selection processes are run on the data with the proposed and imputation methods to confirm the optimal model. Thu, 21 Feb 2013 06:00:00 GMT http://hdl.handle.net/10027/9760 2013-02-21T06:00:00Z