Algorithm effective in identifying at-risk infants for becoming overweight

Image: PD

1. A prediction algorithm was effective in helping researchers identify infants at-risk for becoming overweight at three years of age. 

2. Observed risk values significantly correlated with algorithm calculated risk scores—suggesting strength within the model. 

Evidence Rating Level: 2 (Good) 

Study Rundown: Research shows that obesity has a measurable impact on physical and mental health. Many risk factors for becoming overweight/obese are present in infancy. Early identification of these factors may to lead to both prevention of obesity and greater success during later intervention. Researchers in the current study developed a risk score algorithm to recognize infants at greatest risk of becoming overweight later in childhood. The algorithm was based on health predictors seen in the first year of life, such as rapid weight gain during infancy. Researchers found that the risk algorithm scores correlated significantly with observed values. The high negative predicative value (NPV) of the algorithm indicated strength in its ability to identify individuals not at risk for weight issues, while a low positive predictive value (PPV) indicated the algorithm’s potential misclassification of infants as at-risk. Although this algorithm may be limited as it cannot predict long-term outcomes, it carries great clinical significance in identifying those who might benefit from this evidence-based prevention strategy.

Click to read the study, published today in Pediatrics

Relevant Reading: Risky vs. rapid growth in infancy: refining pediatric screening for childhood overweight

In-Depth [prospective cohort study]: Researchers used methods from previous studies to develop an algorithm for identifying children who are at risk of becoming overweight at three years of age. By interviewing parents when the infant was aged 6-12 months, researchers were able to assign risk scores based on observable risk factors. A total of 13,513 children were divided into two groups; 80% were randomly allocated for algorithm derivation (derivation cohort), and the other 20% were used to test the validity of the algorithm (validation cohort). A follow-up was conducted at three years in order to obtain anthropometric data for comparison to predicted values. The observed risk scores correlated considerably with both the derivation (R2 = 0.92) and validation (R2 = 0.84) cohorts. The PPV for the algorithm was 38% for the derivation cohort and 37% for the validation cohort. The NPV was 87% for the derivation cohort and 89% for the validation cohort.

By Brandon Childs and Leah H. Carr

More from this author: Study shows poor follow up of children with high blood pressure, Parental smoking cessation intervention effective in pediatric practices, Early feeding practices modifiable through appropriate education, Otoacoustic emissions may be an effective objective hearing test, Early weight gain linked to obesity and cardiovascular health, Influenza vaccine safe for IBD patients, Incidence of pediatric melanoma is increasing

© 2013 All rights reserved. No works may be reproduced without written consent from Disclaimer: We present factual information directly from peer reviewed medical journals. No post should be construed as medical advice and is not intended as such by the authors or by PLEASE SEE A HEALTHCARE PROVIDER IN YOUR AREA IF YOU SEEK MEDICAL ADVICE OF ANY SORT. Content is produced in accordance with fair use copyrights solely and strictly for the purpose of teaching, news and criticism. No benefit, monetary or otherwise, is realized by any participants or the owner of this domain.