Clustering Random Coefficients of Piecewise Mixed Effects Model to Group Children with Differential Weight Gain Patterns Before and After Diagnosis of Asthma

• Activity Number: 407
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Md Hossain*+ and H. Timothy Bunnell and Jason E. Lang and Tim Wysocki and Thomas Shaffer
• Companies: Nemours and Nemours and Nemours and Nemours and Nemours
• Keywords: Piecewise ; Mixed Effects Model ; Random Coefficients ; Cluster Analysis ; Weight Gain ; Weight Loss

Abstract

Childhood asthma and obesity are significant public health problems and have shared risk factors and manifestations. Children are diagnosed with asthma in early childhood and are often exposed to factors associated with differential weight gain patterns. For example, some asthma medications are associated with weight loss in males and weight gain in females. We used standardized scores (BMIz) of weight-for-length (age< 2 yrs) and BMI (>= 2 yrs) of 1194 children from their unique sequences of clinic visits during the first 5 years of life, and determined differential weight gain patterns before and after diagnosis of asthma. We used a piecewise mixed effects model with a knot at asthma diagnosis and performed a cluster analysis of individual level difference in the rate of change in BMIz before and after asthma diagnosis. It categorized 1194 children into three groups, namely, increased, decreased and no substantial change in trends of BMIz after asthma diagnosis. Demographics and clinical factors were characterized by three groups. Gender, race-ethnicity, asthma medications were found to be associated with differential weight gain patterns before and after diagnosis of asthma.