We have shown that total homocysteine is a cardiovascular risk factor for women, as well as a determinant for lacunar infarcts and dementia in long term follow-up.
The effect of physical activity as a protective factor also for the elderly has been confirmed. Methodological issues regarding epidemiological research have been analyzed and discussed in appropriate publications.
Adequate biological data have been provided for international consortia.
For measurement error modelling ( ( (ME) we have obtained important results with regard to statistical properties and correction methods. Application of Boosting in this setting has been remarkably effective.
The use of functional data analysis (FDA) has proved to be important for data that are functional by nature. FDA is also useful for data that are not measured as functions, but has an underlying functional nature. Generalized linear mixed models (GLMM) is a powerful tool for the analysis of longitudinal data. It requires setting up an appropriate model for the data, which is difficult, and particularly so with longitudinal data.
Our research is concerned with extending the analysis of longitudinal data. Such data is increasingly common as it permits the study of change and thus address important questions such as how change occurs and what influences it.