B3111 - Time-dependent associations between body mass/body composition physical activity diet and lung function in childhood - 08/05/2018

B number: 
Principal applicant name: 
Annabelle B├ędard | Barcelona Institute for Global Health (Spain)
Dr Judith Garcia-Aymerich, Ms Anne-Elie Carsin
Title of project: 
Time-dependent associations between body mass/body composition, physical activity, diet and lung function in childhood
Proposal summary: 

The large increase in the prevalence of respiratory diseases over the last decades, in the West more particularly, cannot be explained by genetics only. It has been hypothesized that these increases are a consequence of changing environmental and/or lifestyle factors. Given the multifactorial aspect of these diseases, it is thus important to take into account the interrelations between these factors and respiratory health. The interrelations between body mass/body composition, physical activity, diet and lung function in childhood and adulthood have been incompletely addressed, likely because their time-dependent and bidirectional nature represent a methodologically challenging research question. Marginal structural models (MSMs) allow estimation of causal effects in observational studies by addressing time-dependent confounding (Robins JM et al. Epidemiology 2000). This approach has still limited application in respiratory epidemiology. We aim to investigate the joint and independent causal effects of body mass/body composition, physical activity and diet on lung function during childhood and early adulthood using MSMs in children from the ALSPAC study.

Impact of research: 
Respiratory diseases are global public health concerns, and leading causes of morbidity and mortality in children and adults. This research project could thus lead to potential public health interventions such as dietary, physical activity or weight loss interventions to improve lung function, prevent lung function decline etc. As per the methodological aspect of this project, we will assess the relevance of using novel methods from the causal inference framework to investigate research questions for which standard epidemiological methods are usually used and/or where knowledge is still needed. Through dissemination of its outcomes, this research project will potentially inform study design and methodology to improve causal inference when investigating specific research questions.
Date proposal received: 
Saturday, 5 May, 2018
Epidemiology, Obesity, Respiratory - asthma, Statistical methods, BMI, Methods - e.g. cross cohort analysis, data mining, mendelian randomisation, etc., Nutrition - breast feeding, diet, Physical - activity, fitness, function, Statistical methods