Independent thesis Advanced level (professional degree), 20 credits / 30 HE credits
Integrated database used to develop a population growth rate model in young females with Turner Syndrome
Supervisor: Dr. Jan Freijer, research director PK/PD at Centre of Human Drug Research, The Netherlands,
Degree Project in Pharmacokinetics D, 30 credits,
Examiner: Dr. Margareta Hammarlund-Udenaes, Faculty of Pharmacy, Uppsala University
Introduction: Girls with the chromosomal disorder Turner Syndrome (TS) reach an adult stature 20 cm less than average healthy growing girls. In treatment of TS, growth hormone, estrogens and androgens are thought to enhance growth. However, it remains unclear what the optimal doses are and when they should be initiated and halted. Optimization of treatment can be supported using a growth model that includes the effect of treatment in different stages of growth.
Aim: To continue the development of a population mixed effects model based on growth rate that will be used in research of treatment effects. Secondly, to create an integrated database that supports the modeling of treatment effects in girls with TS.
Materials and Methods: Different datasets were merged into a database. Height data from the datasets was used in calculation of growth rate. Both height and growth rate was used in modeling.
Results: A large integrative database with uniform coding and variable names was created. Data for no treatment and treatment was simultaneously used in modeling of the population growth rate model. The model was built on three growth stages representing infancy, childhood and puberty. Each of the stages described the growth rate well and could predict height correctly, but only if a combination of height and growth rate was used in the modeling. A clear treatment effect of x cm height gain can be quantified by the model.
Conclusions: The developed population mixed effects model based on growth rate is an suitable tool in future researches of treatment effects in girls with TS, with the awareness that height and growth rate should be combined in the model. Treatment effects can be studied in more detail by using database information.
2012. , 65 p.