B3918 - Modelling longitudinal trajectories of lung function using UNICORN cohorts - 28/10/2021

B number: 
Principal applicant name: 
Anhar Ullah | National Heart and Lung Institute Imperial College London. (UK)
Dr Raquel Granell , Dr Sadia Haider, Dr Sara Fontanella
Title of project: 
Modelling longitudinal trajectories of lung function using UNICORN cohorts
Proposal summary: 

Introduction: Several studies have reported that lung function in early adulthood is a significant predictor of mortality and COPD and that childhood events in part determine lung function in adult life. Multiple birth cohort’s studies support these findings; these studies have shown different lung function growth trajectories in the population. Most of the studies reported parallel lung function trajectories with two to four trajectories, except Tasmanian longitudinal health study. Tasmanian longitudinal health study identifies six trajectories, besides persistently high, average, below average, persistently low they also reported trajectories with early below average, accelerated decline and early low, accelerated growth, normal decline, the reason for this may be due to long follow-up time (up to age 53 years).
Hypothesis: We hypothesised that besides the parallel trajectories already reported in the literature, there are trajectories with both declining and increasing trends. We have already found some interesting findings to support our initial hypothesis using the MAAS and IOW data, which need validation using a similar cohort but larger sample size. Furthermore, using MASS and IOW, we are modelling other spirometry patterns derived from actual spirometry measures; these new derived markers are more clinically relevant and easily interpretable.
Methods: We will use a latent process mixed model for multivariate markers and latent transition analysis to support our hypothesis. After deriving the trajectories, we will identify early-life predictors for these trajectories; we will also use the genotyping data to see whether these trajectories have genetic bases. To Validate our findings from MAAS and IOW, we need a larger cohort, and ALSPAC is vital for this validation due to its larger sample size.

Impact of research: 
To show how individual-level patterns of lung function deviate from group-level trajectories derived using data-driven techniques. In particular, we wish to identify children with lung function decline and improved growth. Anhar has expertise in methodologies to identify these patterns, which have not been identified n previous studies. We believe this research will advance the literature and understanding of lung function development in a clinically meaningful and important way.
Date proposal received: 
Friday, 22 October, 2021
Date proposal approved: 
Thursday, 28 October, 2021
Bioinformatics, Allergy, Respiratory - asthma, Statistical methods, Development, Growth, Sex differences, Statistical methods