Predicting successful long-term weight loss from short-term weight-loss outcomes: New insights from a dynamic energy balance model (the POUNDS Lost study)

Diana M. Thomas, Andrada Ivanescu, Corby K. Martin, Steven B. Heymsfield, Kaitlyn Marshall, Victoria E. Bodrato, Donald A. Williamson, Stephen D. Anton, Frank M. Sacks, Donna Ryan, George A. Bray

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Background: Currently, early weight-loss predictions of long-term weight-loss success rely on fixed percent-weight-loss thresholds. Objective: The objective was to develop thresholds during the first 3 mo of intervention that include the influence of age, sex, baseline weight, percent weight loss, and deviations from expected weight to predict whether a participant is likely to lose 5% or more body weight by year 1. Design: Data consisting of month 1, 2, 3, and 12 treatment weights were obtained from the 2-y Preventing Obesity Using Novel Dietary Strategies (POUNDS Lost) intervention. Logistic regression models that included covariates of age, height, sex, baseline weight, target energy intake, percent weight loss, and deviation of actual weight from expected were developed for months 1, 2, and 3 that predicted the probability of losing <5% of body weight in 1 y. Receiver operating characteristic (ROC) curves, area under the curve (AUC), and thresholds were calculated for each model. The AUC statistic quantified the ROC curve's capacity to classify participants likely to lose <5% of their body weight at the end of 1 y. The models yielding the highest AUC were retained as optimal. For comparison with current practice, ROC curves relying solely on percent weight loss were also calculated. Results: Optimal models for months 1, 2, and 3 yielded ROC curves with AUCs of 0.68 (95% CI: 0.63, 0.74), 0.75 (95% CI: 0.71, 0.81), and 0.79 (95% CI: 0.74, 0.84), respectively. Percent weight loss alone was not better at identifying true positives than random chance (AUC ≤0.50). Conclusions: The newly derived models provide a personalized prediction of long-term success from early weight-loss variables. The predictions improve on existing fixed percent-weight-loss thresholds. Future research is needed to explore model application for informing treatment approaches during early intervention.

Original languageEnglish
Pages (from-to)449-454
Number of pages6
JournalAmerican Journal of Clinical Nutrition
Volume101
Issue number3
DOIs
StatePublished - 1 Mar 2015

Fingerprint

Weight Loss
Area Under Curve
ROC Curve
Weights and Measures
Body Weight
Logistic Models
Energy Intake
Obesity

Keywords

  • Dynamic model
  • Energy balance
  • Likelihood function
  • Mathematical model
  • Receiver operating characteristic
  • Regression
  • Weight loss

Cite this

Thomas, Diana M. ; Ivanescu, Andrada ; Martin, Corby K. ; Heymsfield, Steven B. ; Marshall, Kaitlyn ; Bodrato, Victoria E. ; Williamson, Donald A. ; Anton, Stephen D. ; Sacks, Frank M. ; Ryan, Donna ; Bray, George A. / Predicting successful long-term weight loss from short-term weight-loss outcomes : New insights from a dynamic energy balance model (the POUNDS Lost study). In: American Journal of Clinical Nutrition. 2015 ; Vol. 101, No. 3. pp. 449-454.
@article{8ec495cc5a0b453ca7c274a2313a3a83,
title = "Predicting successful long-term weight loss from short-term weight-loss outcomes: New insights from a dynamic energy balance model (the POUNDS Lost study)",
abstract = "Background: Currently, early weight-loss predictions of long-term weight-loss success rely on fixed percent-weight-loss thresholds. Objective: The objective was to develop thresholds during the first 3 mo of intervention that include the influence of age, sex, baseline weight, percent weight loss, and deviations from expected weight to predict whether a participant is likely to lose 5{\%} or more body weight by year 1. Design: Data consisting of month 1, 2, 3, and 12 treatment weights were obtained from the 2-y Preventing Obesity Using Novel Dietary Strategies (POUNDS Lost) intervention. Logistic regression models that included covariates of age, height, sex, baseline weight, target energy intake, percent weight loss, and deviation of actual weight from expected were developed for months 1, 2, and 3 that predicted the probability of losing <5{\%} of body weight in 1 y. Receiver operating characteristic (ROC) curves, area under the curve (AUC), and thresholds were calculated for each model. The AUC statistic quantified the ROC curve's capacity to classify participants likely to lose <5{\%} of their body weight at the end of 1 y. The models yielding the highest AUC were retained as optimal. For comparison with current practice, ROC curves relying solely on percent weight loss were also calculated. Results: Optimal models for months 1, 2, and 3 yielded ROC curves with AUCs of 0.68 (95{\%} CI: 0.63, 0.74), 0.75 (95{\%} CI: 0.71, 0.81), and 0.79 (95{\%} CI: 0.74, 0.84), respectively. Percent weight loss alone was not better at identifying true positives than random chance (AUC ≤0.50). Conclusions: The newly derived models provide a personalized prediction of long-term success from early weight-loss variables. The predictions improve on existing fixed percent-weight-loss thresholds. Future research is needed to explore model application for informing treatment approaches during early intervention.",
keywords = "Dynamic model, Energy balance, Likelihood function, Mathematical model, Receiver operating characteristic, Regression, Weight loss",
author = "Thomas, {Diana M.} and Andrada Ivanescu and Martin, {Corby K.} and Heymsfield, {Steven B.} and Kaitlyn Marshall and Bodrato, {Victoria E.} and Williamson, {Donald A.} and Anton, {Stephen D.} and Sacks, {Frank M.} and Donna Ryan and Bray, {George A.}",
year = "2015",
month = "3",
day = "1",
doi = "10.3945/ajcn.114.091520",
language = "English",
volume = "101",
pages = "449--454",
journal = "American Journal of Clinical Nutrition",
issn = "0002-9165",
publisher = "American Society for Nutrition",
number = "3",

}

Thomas, DM, Ivanescu, A, Martin, CK, Heymsfield, SB, Marshall, K, Bodrato, VE, Williamson, DA, Anton, SD, Sacks, FM, Ryan, D & Bray, GA 2015, 'Predicting successful long-term weight loss from short-term weight-loss outcomes: New insights from a dynamic energy balance model (the POUNDS Lost study)', American Journal of Clinical Nutrition, vol. 101, no. 3, pp. 449-454. https://doi.org/10.3945/ajcn.114.091520

Predicting successful long-term weight loss from short-term weight-loss outcomes : New insights from a dynamic energy balance model (the POUNDS Lost study). / Thomas, Diana M.; Ivanescu, Andrada; Martin, Corby K.; Heymsfield, Steven B.; Marshall, Kaitlyn; Bodrato, Victoria E.; Williamson, Donald A.; Anton, Stephen D.; Sacks, Frank M.; Ryan, Donna; Bray, George A.

In: American Journal of Clinical Nutrition, Vol. 101, No. 3, 01.03.2015, p. 449-454.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Predicting successful long-term weight loss from short-term weight-loss outcomes

T2 - New insights from a dynamic energy balance model (the POUNDS Lost study)

AU - Thomas, Diana M.

AU - Ivanescu, Andrada

AU - Martin, Corby K.

AU - Heymsfield, Steven B.

AU - Marshall, Kaitlyn

AU - Bodrato, Victoria E.

AU - Williamson, Donald A.

AU - Anton, Stephen D.

AU - Sacks, Frank M.

AU - Ryan, Donna

AU - Bray, George A.

PY - 2015/3/1

Y1 - 2015/3/1

N2 - Background: Currently, early weight-loss predictions of long-term weight-loss success rely on fixed percent-weight-loss thresholds. Objective: The objective was to develop thresholds during the first 3 mo of intervention that include the influence of age, sex, baseline weight, percent weight loss, and deviations from expected weight to predict whether a participant is likely to lose 5% or more body weight by year 1. Design: Data consisting of month 1, 2, 3, and 12 treatment weights were obtained from the 2-y Preventing Obesity Using Novel Dietary Strategies (POUNDS Lost) intervention. Logistic regression models that included covariates of age, height, sex, baseline weight, target energy intake, percent weight loss, and deviation of actual weight from expected were developed for months 1, 2, and 3 that predicted the probability of losing <5% of body weight in 1 y. Receiver operating characteristic (ROC) curves, area under the curve (AUC), and thresholds were calculated for each model. The AUC statistic quantified the ROC curve's capacity to classify participants likely to lose <5% of their body weight at the end of 1 y. The models yielding the highest AUC were retained as optimal. For comparison with current practice, ROC curves relying solely on percent weight loss were also calculated. Results: Optimal models for months 1, 2, and 3 yielded ROC curves with AUCs of 0.68 (95% CI: 0.63, 0.74), 0.75 (95% CI: 0.71, 0.81), and 0.79 (95% CI: 0.74, 0.84), respectively. Percent weight loss alone was not better at identifying true positives than random chance (AUC ≤0.50). Conclusions: The newly derived models provide a personalized prediction of long-term success from early weight-loss variables. The predictions improve on existing fixed percent-weight-loss thresholds. Future research is needed to explore model application for informing treatment approaches during early intervention.

AB - Background: Currently, early weight-loss predictions of long-term weight-loss success rely on fixed percent-weight-loss thresholds. Objective: The objective was to develop thresholds during the first 3 mo of intervention that include the influence of age, sex, baseline weight, percent weight loss, and deviations from expected weight to predict whether a participant is likely to lose 5% or more body weight by year 1. Design: Data consisting of month 1, 2, 3, and 12 treatment weights were obtained from the 2-y Preventing Obesity Using Novel Dietary Strategies (POUNDS Lost) intervention. Logistic regression models that included covariates of age, height, sex, baseline weight, target energy intake, percent weight loss, and deviation of actual weight from expected were developed for months 1, 2, and 3 that predicted the probability of losing <5% of body weight in 1 y. Receiver operating characteristic (ROC) curves, area under the curve (AUC), and thresholds were calculated for each model. The AUC statistic quantified the ROC curve's capacity to classify participants likely to lose <5% of their body weight at the end of 1 y. The models yielding the highest AUC were retained as optimal. For comparison with current practice, ROC curves relying solely on percent weight loss were also calculated. Results: Optimal models for months 1, 2, and 3 yielded ROC curves with AUCs of 0.68 (95% CI: 0.63, 0.74), 0.75 (95% CI: 0.71, 0.81), and 0.79 (95% CI: 0.74, 0.84), respectively. Percent weight loss alone was not better at identifying true positives than random chance (AUC ≤0.50). Conclusions: The newly derived models provide a personalized prediction of long-term success from early weight-loss variables. The predictions improve on existing fixed percent-weight-loss thresholds. Future research is needed to explore model application for informing treatment approaches during early intervention.

KW - Dynamic model

KW - Energy balance

KW - Likelihood function

KW - Mathematical model

KW - Receiver operating characteristic

KW - Regression

KW - Weight loss

UR - http://www.scopus.com/inward/record.url?scp=84928923102&partnerID=8YFLogxK

U2 - 10.3945/ajcn.114.091520

DO - 10.3945/ajcn.114.091520

M3 - Article

C2 - 25733628

AN - SCOPUS:84928923102

VL - 101

SP - 449

EP - 454

JO - American Journal of Clinical Nutrition

JF - American Journal of Clinical Nutrition

SN - 0002-9165

IS - 3

ER -