Poststratification as a suitable approach to generalize findings of two cross-sectional studies along the Bavarian compulsory school entrance examination: An exemplary poststratified analysis for asthma, hay fever and wheezing

BACKGROUND

A compulsory school entrance examination of pre-school children (SEU) is administered in the German state of Bavaria. Every second year since 2004, the examinations are expanded in six study regions using a cross-sectional survey design (GME). However, the extent to which the results of the GME surveys are generalizable to the SEU population is unknown. Therefore, this study carried out a poststratification of two different GME surveys. The aim was to observe the impact of poststratification on an exemplary analysis of influencing factors for three allergy and asthma related outcomes (hay fever, asthma, wheezing) and thus to better understand this important question.

METHODS

First, poststratification was applied to correct for deviances of the GME sample in comparison to the SEU population. Logistic regression was used to determine the auxiliary variables for the poststratification. Following this, a composite variable as a linear combination was created to calculate weighting factors. Next, logistic regression analyses were applied to analyze possible influencing factors for three allergy and asthma related outcomes (hay fever, asthma and wheezing) in two GME surveys (2005/2006 and 2012/2013). Subsequently, the differences arising from poststratification were examined in more detail.

RESULTS

This study supports the hypothesis that the GME sample deviates from the SEU population. Mother tongue other than German of at least one parent, complete vaccination status and conspicuous visuomotor test results were positively associated with participation in both GME surveys. The prevalence for hay fever, asthma and frequent wheeze did not change statistically significant from 2005/2006 to 2012/2013. In the twelve analyses before and after poststratification, male sex was statistically significant associated with the three allergy and asthma related outcomes (e.g. asthma 2005/2006 after poststratification, aOR: 2.06, 95%-CI: 1.56-2.71). A high body mass index was positively associated with asthma (e.g. 2005/2006 after poststratification, aOR: 1.12, 95%-CI: 1.05-1.20). Poststratification caused a significant change in the sample composition.

CONCLUSIONS

It might be suggested that a poststratification should be performed for each GME survey. Poststratification tended to make results more comparable with previous research. In accordance with previous research, this study confirmed that male sex and high BMI are associated with asthma.

Residential crowding and asthma in preschool children, a cross-sectional study

BACKGROUND

Asthma is the most common chronic disease in children, and associations with crowding have been reported. The aim of this study was to explore possible associations of crowding with asthma in children.

METHODS

Seven cross-sectional surveys with preschool children were conducted within the framework of the health monitoring units in Bavaria, Germany, from 2004 to 2014. Residential crowding was defined as habitation of more than one person per room or less than 20m² living space per person. Logistic regression models examined temporal changes in crowding, applying the first survey as reference. The relationship between crowding and physician-diagnosed asthma, asthma defined by the International Study of Asthma and Allergies in Childhood (ISAAC) and asthma symptoms were analyzed.

RESULTS

Analyzing temporal changes of crowding rates did not reveal any differences over the years. However, the stratified descriptive analysis indicated a crowding increase in time in urban households where parents had a low education level (47.9% in 2004/05, 55.8% in 2014/15). No association was found between crowding and the variables "physician-diagnosed asthma" in 2014/15, "asthma defined by ISAAC" in 2014/15, or "wheezing" in 2014/15. A positive association with cough was identified in 2014/15 after adjusting for confounders (aOR=1.42 [95% CI: 1.20-1.69]).

CONCLUSIONS

In general, residential crowding did not change from 2004 to 2014; however, there seems to be a small upsurge for children with low-educated parents, living in urban areas over the years. A statistically significant association between crowding and cough was only found in the survey from 2014/15.