A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV

Federico Motta; Jovana Milic; Licia Gozzi; Michela Belli; Laura Sighinolfi; Gianluca Cuomo; Federica Carli; Giovanni Dolci; Vittorio Iadisernia; Giulia Burastero; Cristina Mussini; Paolo Missier; Federica Mandreoli; Giovanni Guaraldi

doi:10.1097/QAI.0000000000003302

A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV

Federico Motta, Jovana Milic, Licia Gozzi, Michela Belli, Laura Sighinolfi, Gianluca Cuomo, Federica Carli, Giovanni Dolci, Vittorio Iadisernia, Giulia Burastero, Cristina Mussini, Paolo Missier, Federica Mandreoli, Giovanni Guaraldi^*

^*Corresponding author for this work

Computer Science

Research output: Contribution to journal › Article › peer-review

Abstract

INTRODUCTION: The objective of the study was to develop machine learning (ML) models that predict the percentage weight change in each interval of time in antiretroviral therapy-experienced people living with HIV.

METHODS: This was an observational study that comprised consecutive people living with HIV attending Modena HIV Metabolic Clinic with at least 2 visits. Data were partitioned in an 80/20 training/test set to generate 10 progressively parsimonious predictive ML models. Weight gain was defined as any weight change >5%, at the next visit. SHapley Additive exPlanations values were used to quantify the positive or negative impact of any single variable included in each model on the predicted weight changes.

RESULTS: A total of 3,321 patients generated 18,322 observations. At the last observation, the median age was 50 years and 69% patients were male. Model 1 (the only 1 including body composition assessed with dual-energy x-ray absorptiometry) had an accuracy greater than 90%. This model could predict weight at the next visit with an error of <5%.

CONCLUSIONS: ML models with the inclusion of body composition and metabolic and endocrinological variables had an excellent performance. The parsimonious models available in standard clinical evaluation are insufficient to obtain reliable prediction, but are good enough to predict who will not experience weight gain.

Original language	English
Pages (from-to)	474-481
Number of pages	8
Journal	Journal of Acquired Immune Deficiency Syndromes
Volume	94
Issue number	5
DOIs	https://doi.org/10.1097/QAI.0000000000003302
Publication status	Published - 15 Dec 2023

Bibliographical note

Funding Information:
Supported by a Grant from Gilead Sciences. The company did not have any intervention in the design, development, writing, and interpretation of the presented results.

Publisher Copyright:
© 2023 Lippincott Williams and Wilkins. All rights reserved.

Keywords

Humans
Male
Middle Aged
Female
HIV Infections/drug therapy
Body Composition
Weight Gain
Machine Learning

ASJC Scopus subject areas

Infectious Diseases
Pharmacology (medical)

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1097/QAI.0000000000003302Licence: None: All rights reserved

Cite this

Motta, F., Milic, J., Gozzi, L., Belli, M., Sighinolfi, L., Cuomo, G., Carli, F., Dolci, G., Iadisernia, V., Burastero, G., Mussini, C., Missier, P., Mandreoli, F., & Guaraldi, G. (2023). A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV. Journal of Acquired Immune Deficiency Syndromes, 94(5), 474-481. https://doi.org/10.1097/QAI.0000000000003302

@article{36bfb67b00cf4aea8a96477596b3eb8e,

title = "A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV",

abstract = "INTRODUCTION: The objective of the study was to develop machine learning (ML) models that predict the percentage weight change in each interval of time in antiretroviral therapy-experienced people living with HIV.METHODS: This was an observational study that comprised consecutive people living with HIV attending Modena HIV Metabolic Clinic with at least 2 visits. Data were partitioned in an 80/20 training/test set to generate 10 progressively parsimonious predictive ML models. Weight gain was defined as any weight change >5%, at the next visit. SHapley Additive exPlanations values were used to quantify the positive or negative impact of any single variable included in each model on the predicted weight changes.RESULTS: A total of 3,321 patients generated 18,322 observations. At the last observation, the median age was 50 years and 69% patients were male. Model 1 (the only 1 including body composition assessed with dual-energy x-ray absorptiometry) had an accuracy greater than 90%. This model could predict weight at the next visit with an error of <5%.CONCLUSIONS: ML models with the inclusion of body composition and metabolic and endocrinological variables had an excellent performance. The parsimonious models available in standard clinical evaluation are insufficient to obtain reliable prediction, but are good enough to predict who will not experience weight gain.",

keywords = "Humans, Male, Middle Aged, Female, HIV Infections/drug therapy, Body Composition, Weight Gain, Machine Learning",

author = "Federico Motta and Jovana Milic and Licia Gozzi and Michela Belli and Laura Sighinolfi and Gianluca Cuomo and Federica Carli and Giovanni Dolci and Vittorio Iadisernia and Giulia Burastero and Cristina Mussini and Paolo Missier and Federica Mandreoli and Giovanni Guaraldi",

note = "Funding Information: Supported by a Grant from Gilead Sciences. The company did not have any intervention in the design, development, writing, and interpretation of the presented results. Publisher Copyright: {\textcopyright} 2023 Lippincott Williams and Wilkins. All rights reserved.",

year = "2023",

month = dec,

day = "15",

doi = "10.1097/QAI.0000000000003302",

language = "English",

volume = "94",

pages = "474--481",

journal = "Journal of Acquired Immune Deficiency Syndromes",

issn = "1525-4135",

publisher = "Lippincott Williams and Wilkins",

number = "5",

}

Motta, F, Milic, J, Gozzi, L, Belli, M, Sighinolfi, L, Cuomo, G, Carli, F, Dolci, G, Iadisernia, V, Burastero, G, Mussini, C, Missier, P, Mandreoli, F & Guaraldi, G 2023, 'A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV', Journal of Acquired Immune Deficiency Syndromes, vol. 94, no. 5, pp. 474-481. https://doi.org/10.1097/QAI.0000000000003302

TY - JOUR

T1 - A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV

AU - Motta, Federico

AU - Milic, Jovana

AU - Gozzi, Licia

AU - Belli, Michela

AU - Sighinolfi, Laura

AU - Cuomo, Gianluca

AU - Carli, Federica

AU - Dolci, Giovanni

AU - Iadisernia, Vittorio

AU - Burastero, Giulia

AU - Mussini, Cristina

AU - Missier, Paolo

AU - Mandreoli, Federica

AU - Guaraldi, Giovanni

N1 - Funding Information: Supported by a Grant from Gilead Sciences. The company did not have any intervention in the design, development, writing, and interpretation of the presented results. Publisher Copyright: © 2023 Lippincott Williams and Wilkins. All rights reserved.

PY - 2023/12/15

Y1 - 2023/12/15

N2 - INTRODUCTION: The objective of the study was to develop machine learning (ML) models that predict the percentage weight change in each interval of time in antiretroviral therapy-experienced people living with HIV.METHODS: This was an observational study that comprised consecutive people living with HIV attending Modena HIV Metabolic Clinic with at least 2 visits. Data were partitioned in an 80/20 training/test set to generate 10 progressively parsimonious predictive ML models. Weight gain was defined as any weight change >5%, at the next visit. SHapley Additive exPlanations values were used to quantify the positive or negative impact of any single variable included in each model on the predicted weight changes.RESULTS: A total of 3,321 patients generated 18,322 observations. At the last observation, the median age was 50 years and 69% patients were male. Model 1 (the only 1 including body composition assessed with dual-energy x-ray absorptiometry) had an accuracy greater than 90%. This model could predict weight at the next visit with an error of <5%.CONCLUSIONS: ML models with the inclusion of body composition and metabolic and endocrinological variables had an excellent performance. The parsimonious models available in standard clinical evaluation are insufficient to obtain reliable prediction, but are good enough to predict who will not experience weight gain.

AB - INTRODUCTION: The objective of the study was to develop machine learning (ML) models that predict the percentage weight change in each interval of time in antiretroviral therapy-experienced people living with HIV.METHODS: This was an observational study that comprised consecutive people living with HIV attending Modena HIV Metabolic Clinic with at least 2 visits. Data were partitioned in an 80/20 training/test set to generate 10 progressively parsimonious predictive ML models. Weight gain was defined as any weight change >5%, at the next visit. SHapley Additive exPlanations values were used to quantify the positive or negative impact of any single variable included in each model on the predicted weight changes.RESULTS: A total of 3,321 patients generated 18,322 observations. At the last observation, the median age was 50 years and 69% patients were male. Model 1 (the only 1 including body composition assessed with dual-energy x-ray absorptiometry) had an accuracy greater than 90%. This model could predict weight at the next visit with an error of <5%.CONCLUSIONS: ML models with the inclusion of body composition and metabolic and endocrinological variables had an excellent performance. The parsimonious models available in standard clinical evaluation are insufficient to obtain reliable prediction, but are good enough to predict who will not experience weight gain.

KW - Humans

KW - Male

KW - Middle Aged

KW - Female

KW - HIV Infections/drug therapy

KW - Body Composition

KW - Weight Gain

KW - Machine Learning

UR - http://www.scopus.com/inward/record.url?scp=85176428031&partnerID=8YFLogxK

U2 - 10.1097/QAI.0000000000003302

DO - 10.1097/QAI.0000000000003302

M3 - Article

C2 - 37949448

AN - SCOPUS:85176428031

SN - 1525-4135

VL - 94

SP - 474

EP - 481

JO - Journal of Acquired Immune Deficiency Syndromes

JF - Journal of Acquired Immune Deficiency Syndromes

IS - 5

ER -

A Machine Learning Approach to Predict Weight Change in ART-Experienced People Living with HIV

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Fingerprint

Cite this