Title:
Diabetes Prevalence of known diabetes in German adults aged 25–69 years: results from national health surveys over 15 years

Authors:
C. Heidemann, L. Kroll, A. Icks*, T. Lampert and C. Scheidt-Nave

KEYWORDS:
cross-sectional studies • diabetes mellitus • health surveys • prevalence • trend

Abstract:

Aims:
The few studies examining the secular trend in diabetes prevalence in Germany have yielded conflicting results. Therefore, using nationally representative samples of adults, we investigated whether the prevalence of known diabetes has changed over 15 years.

Methods:
Study participants were 25- to 69-year-old residents participating in nationally representative health surveys performed in the following time periods: 1990–1992, 1997–1999, 2002–2003, 2003–2004 and 2004–2005. Prevalences of diabetes, standardized to the population structure of 2004, and trends over time were assessed for the total study population as well as by gender and other diabetes-associated factors.

Results:
Between 1990–1992 and 2002–2005, no statistically significant trend in the total (5.16 and 5.34%, P trend = 0.68) or sex-specific diabetes prevalence (men: 5.43 and 5.73, P trend = 0.62; women: 4.88 and 4.95%, P trend = 0.94) was observed. For each time period, prevalence rose substantially with increasing age, increasing body mass index, lower sporting activity and lower education.

Conclusions:
Our findings reflect no temporal increase in the total prevalence of known diabetes in German adult men and women. However, prevalence estimates were relatively high when compared with other European studies and call for continued efforts for the prevention and management of diabetes. Accepted 15 April 2009

”Are Things Getting Better or Worse, and Why? The Role of Social Indicators to Inform Public Policy”
ISA – Working Group 06 “Social Indicators
Conference Track at 1st ISA Forum “Sociological Research and Public Debate”
Barcelona, September 5 – 8, 2008

Session 1: Indicators and Their Use to Guide Policies in the Field of Health
Chair: Heinz-Herbert Noll, GESIS-ZUMA, Mannheim, Germany

Authors:
Lars Eric Kroll; Thomas Lampert; Cornelia Lange; Thomas Ziese

Title: Healthy Life Expectancy as an indicator to compare the health of societies

Abstract:
Background
Healthy life expectancy is an important indicator to describe the consequences of demo-graphic change for different societies. Possible scenarios are an overall increase of the years people are suffering diseases or functional limitations (“expansion of morbidity”), a decline of the morbid time in the life course (“compression of morbidity”) and a “dynamic equilibrium” of morbidity and mortality. We will discuss the concept and problems of the indicator healthy life expectancy for policy advice.
Data and Methods
Data of the German Socio-Economic Panel (GSOEP) for 1984 to 2006 (n=185,275) allow us to estimate trends of healthy life expectancy using different definitions of health status (severe disability; functional limitations; self-perceived health status) based on Sullivan’s method for Germany. We will discuss three important problems in the measurement of healthy life expec-tancies empirically.

Results
First Trends based on four different conceptualisations of health status show a large extend of variation. It turns out that only minor changes in the wording of the underlying questions have a strong impact on the resulting estimates. Second the age-specific rise in the denial rate of surveys leads to an overestimation of health status among the elderly. It is shown that this problem can be reduced using a parametric modelling of the age related risk of morbidity, instead of the common approach that estimates the prevalence of morbidity in 5year age-groups. Third a monitoring of social inequalities in healthy life expectancy (as intended by the EU) has to include inequalities in life expectancy in addition to inequalities in health status. Otherwise population estimates would likely underestimate socio-economic inequalities in healthy life expectancies.

Discussion
While the importance of the indicator healthy life expectancy is rising, problems regarding its conceptualization and measurement are getting into the focus. Our results have shown three important limitations of the indicator in terms of the validity and reliability of its measure-ment. In our opinion a meaningful application of the concept in policy advice should report results based on different health indicators to account for the multi-dimensionality of health and the difficulties in its measurement.

Title:
Analyses of socio-economic inequalities in life expectancy using SOEP

Abstract:

Introduction
Inequalities in life expectancy are a main focus concerning health inequalities in general. Germany’s office of statistics can’t provide data on inequalities in mortality. Evidence is available using SOEP, but recent results suggest that the survival rates of the respondents are overestimated. We will present an alternative specification that leads to better results.

Data and Methods
We are using the SOEP and period specific life tables for 1995-2005. Parametric regression models are fitted to estimate the relative mortality risk of five income groups. We controlled for overestimation of survival rates using the general period specific life tables.

Results
Unweighted data overestimates life expectancy at birth by approximately 6.5 and 4.1 years for men and women, weighted data are slightly better with 5.5 and 3.1 years. After applying a correction, we observed a difference to the highest income group of 10.8 and 8.4 years and an overall life expectancy of 70.1 and 76.9 years for men and women at risk of poverty.

Discussion
The presented specification can be applied to different status indicators and leads to more comparable results using SOEP. The small sample size for survival analyses however poses limits to potential international comparisons and trend analyses.

I’ve implemented a RSS Feed to the Stata User’s list “Statalist“. If you are interested in it, you can see the posts of the last 7 days or search the stalist archive using google at this site.

Or you can simply subscribe to the RSS feed using the following link:

Subscribe in a reader

I’ve published an article on the interrelation of social capital, social inequality and health in Germany using GSOEP data. It will appear in march in the pear reviewed journal “Das Gesundheitswesen“.

Abstract

The concept of social capital has been intensively discussed in public health during the last years. But there is only little empirical evidence regarding its association with health in Germany up to now. We’re using data of the German Socio-Economic Panel (GSOEP) out of 2003 to analyze the health utility of social capital. The results suggest a clear positive association among different forms of social capital (trust, norms of reciprocity, participation, social integration) and subjective health. Men and women who are poorly socially integrated regard their own health comparably often as limited. In contrast to them men and women who hold any form of social capital rate their own health status noticeable better. In multivariate analysis socio-economic differences in subjective health could be traced back partly to the stock of social capital. In addition to its health utility, social capital may therefore make a contribution to understand the emergence of social inequalities in health. All in all results suggest that the concept of social capital should be considered even more in public health. Due to the present lack of empirical evidence social capital has to be incorporated into more studies in Germany.

Keywords

Social capital, social inequality, subjective health status, GSOEP

XTMRHO – Is a usefull Stata ADO to calculate intra class correlations (rho) after fitting a multi-level model using Stata’s xtmixed.

Authors: Lars E. Kroll Req: Stata version 9
Created: 2007-02-19