Bivariate Mixture Models of Iron Measures with Longitudinal Evaluation of Component Stability in an Australian Population

• Activity Number: 499
• Type: Topic Contributed
• Date/Time: Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
• Sponsor: WNAR
• Author(s): Christine E. McLaren*+ and Wen-Pin Chen and Nadine A. Bertalli and Nadine A. Bertalli and Martin Delatycki and Nadine A. Bertalli and Nadine A. Bertalli and Nadine A. Bertalli and Katrina Allen and Lyle Gurrin
• Companies: University of California at Irvine and Chao Family Comprehensive Cancer Center and University of Melbourne and University of Melbourne and University of Melbourne and University of Melbourne and University of Melbourne and University of Melbourne and University of Melbourne and University of Melbourne
• Keywords: mixture models ; longitudinal data ; bivariate distributions

Abstract

Hemochromatosis is an inherited disorder that increases susceptibility to develop iron overload. We developed statistical procedures based on bivariate mixture modeling to identify supopulations in which hereditary or acquired factors influence iron metabolism measures. These procedures were used to analyze the joint distributions of transferrin saturation (TS) and serum ferritin concentration (SF) measured in the Hemochromatosis and Iron Overload Screening Study. Four components were identified with successively age-adjusted increasing means for TS and SF data contributed by 94,970 individuals in four race/ethnic groups. We now describe validation of the statistical mixture modeling approach using data from the 1,052 men and women enrolled in the Australian HealthIron Study of genetic and environmental modifiers of hemochromatosis. Statistical issues include adjusting for known sources of variation in biologic markers, fitting a bivariate model to the residual distributions of markers, determining stability of component distributions over time, and estimating the weighted prevalence of genetic mutations and disease outcomes. External validity and component stability are shown.