TY - JOUR
T1 - Data integration for prediction of weight loss in randomized controlled dietary trials
AU - Nielsen, Rikke Linnemann
AU - Helenius, Marianne
AU - Garcia, Sara L.
AU - Roager, Henrik M.
AU - Aytan-Aktug, Derya
AU - Hansen, Lea Benedicte Skov
AU - Lind, Mads Vendelbo
AU - Vogt, Josef Korbinian
AU - Dalgaard, Marlene Danner
AU - Bahl, Martin I.
AU - Jensen, Cecilia Bang
AU - Muktupavela, Rasa
AU - Warinner, Christina
AU - Aaskov, Vincent
AU - Gøbel, Rikke
AU - Kristensen, Mette
AU - Frøkiær, Hanne
AU - Sparholt, Morten H.
AU - Christensen, Anders F.
AU - Vestergaard, Henrik
AU - Hansen, Torben
AU - Kristiansen, Karsten
AU - Brix, Susanne
AU - Petersen, Thomas Nordahl
AU - Lauritzen, Lotte
AU - Licht, Tine Rask
AU - Pedersen, Oluf
AU - Gupta, Ramneek
PY - 2020
Y1 - 2020
N2 - Diet is an important component in weight management strategies, but heterogeneous responses to the same diet make it difficult to foresee individual weight-loss outcomes. Omics-based technologies now allow for analysis of multiple factors for weight loss prediction at the individual level. Here, we classify weight loss responders (N = 106) and non-responders (N = 97) of overweight non-diabetic middle-aged Danes to two earlier reported dietary trials over 8 weeks. Random forest models integrated gut microbiome, host genetics, urine metabolome, measures of physiology and anthropometrics measured prior to any dietary intervention to identify individual predisposing features of weight loss in combination with diet. The most predictive models for weight loss included features of diet, gut bacterial species and urine metabolites (ROC-AUC: 0.84–0.88) compared to a diet-only model (ROC-AUC: 0.62). A model ensemble integrating multi-omics identified 64% of the non-responders with 80% confidence. Such models will be useful to assist in selecting appropriate weight management strategies, as individual predisposition to diet response varies.
AB - Diet is an important component in weight management strategies, but heterogeneous responses to the same diet make it difficult to foresee individual weight-loss outcomes. Omics-based technologies now allow for analysis of multiple factors for weight loss prediction at the individual level. Here, we classify weight loss responders (N = 106) and non-responders (N = 97) of overweight non-diabetic middle-aged Danes to two earlier reported dietary trials over 8 weeks. Random forest models integrated gut microbiome, host genetics, urine metabolome, measures of physiology and anthropometrics measured prior to any dietary intervention to identify individual predisposing features of weight loss in combination with diet. The most predictive models for weight loss included features of diet, gut bacterial species and urine metabolites (ROC-AUC: 0.84–0.88) compared to a diet-only model (ROC-AUC: 0.62). A model ensemble integrating multi-omics identified 64% of the non-responders with 80% confidence. Such models will be useful to assist in selecting appropriate weight management strategies, as individual predisposition to diet response varies.
U2 - 10.1038/s41598-020-76097-z
DO - 10.1038/s41598-020-76097-z
M3 - Journal article
C2 - 33208769
VL - 10
JO - Scientific Reports
JF - Scientific Reports
SN - 2045-2322
IS - 1
M1 - 20103
ER -
