Generalized estimating equations. Notes on the choice of the working correlation matrix.

BACKGROUND: Generalized estimating equations (GEE) are an extension of generalized linear models (GLM) in that they allow adjusting for correlations between observations. A major strength of GEE is that they do not require the correct specification of the multivariate distribution but only of the mean structure. OBJECTIVES: Several concerns have been raised about the validity of GEE when applied to dichotomous dependent variables. In this contribution, we summarize the theoretical findings concerning efficiency and validity of GEE. METHODS: We introduce the GEE in a formal way, summarize general findings on the choice of the working correlation matrix, and show the existence of a dilemma for the optimal choice of the working correlation matrix for dichotomous dependent variables. RESULTS: Biological and statistical arguments for choosing a specific working correlation matrix are given. Three approaches are described for overcoming the range restriction of the correlation coefficient. CONCLUSIONS: The three approaches described in this article for overcoming the range restrictions for dichotomous dependent variables in GEE models provide a simple and practical way for use in applications.

[Percentiles of body mass index of 18-80-year-old German adults based on data from the Second National Nutrition Survey]. / Perzentilen des Body-Mass-Index auch für 18- bis 80-Jährige? Daten der Nationalen Verzehrsstudie II.

BACKGROUND: Obesity is associated with an increasing risk of morbidity and mortality. The World Health Organization (WHO) defines obesity as a body mass index (BMI) > or = 30.0 kg/m2. In contrast, percentiles depending on age and sex are used in adolescence. DATA AND METHODS: Using the LMS method, age and gender specific BMI percentiles for adulthood are estimated. Data basis is the Second Germany National Nutrition Survey (NVS II) which was surveyed in 2005 and 2006. This study is representative for the German population, and it is based on more than 13,000 adults. RESULTS: Percentiles increase in both sexes with age. It can be observed in females in nearly all age groups. In males, percentiles are saturated around the sixth decade. Only 9 % of 20 to 29 years old adults are obese using WHO criteria, while there are 29 % of 60 to 69 years old adults having a BMI > or = 30.0 kg/m2. CONCLUSIONS: Percentiles reflect age and sex dependencies of the BMI. They show that there is a larger part of the population in higher BMI intervals with increasing age. Thus, the adequacy of the WHO definition should be questioned. Based on the NVS II, current and representative BMI percentiles for adulthood are now available for Germany. These can be used as base for age and sex dependent evaluation of the BMI.