Modelling and Data Analysis for Longitudinal Dietary Intervention

Abstract

Two trials essentially designed to reduce weight will be used as illustrations of modelling issues and data analyses. The first study is a randomized trial intervening on percent fat calories and exercise to bring weight, percent fat calories and specific lipid measures to appropriate levels. The study was designed principally by David Heber and Robert M. Elashoff. Certain aspects of this study were reported in Preventive Medicine, 1992. Up to five measurement times are to be observed. The second study is a weight reduction single arm trial, which, except for the specific intervention, has a rather standard design. The first 100 days of the study will be considered; observations were to be carried out weekly as is customary in such trials. Both of these trials were designed and implemented is a state-of-the-art way.