- Multivariate reference ranges
The CHARGE study was a population based study to determine risk factors for autism. This study has recruited nearly 1,500 children and measured a vast array of environmental, biological, and genetic factors. Dr. Wiest’s research thus far has focused on a sub-population of approximately 200 subject on which 200 lipid metabolites were measured. Because this study is one of the few population based studies on both autistic and normal children, it is an excellent opportunity to define reference ranges for many different measurements.
Thus far, we have used univariate methods to detail reference ranges for each lipid metabolite. However, it is unlikely that one lipid metabolite being out of the reference range would be cause for concern. Therefore, we are working to define multivariate reference ranges based on the biochemical interconnection of the lipid metabolites, and graphical tools to visual them.
- Modeling lipoprotein size and composition in response to dietary challenge
Cholesterol levels in HDL and LDL are well established predictors of heart disease. They also react to certain foods, for instance, high saturated fat. However, it is not the level of HDL and LDL alone, but their size and composition which determines the risk of coronary disease. In collaboration with UC Davis and Harvard, we are modeling how the composition of the particle influences the size, which in turn relates to coronary disease risk.
- Identifying risk factors for injury in underground metal mines
We are working with local silver mines as well as federal databases to identify underlying risk factors for mining injuries and accidents. This involves multivariate modeling of the injury experience overall. We are also utilizing in-depth questionnaires to identify personal practices to relate back to miners’ injury history.