Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose

Abstract Background Accurate fasting plasma glucose (FPG) trend prediction is important for management and treatment of patients with type 2 diabetes mellitus (T2DM), a globally prevalent chronic disease. (Generalised) linear mixed-effects (LME) models and machine learning (ML) are commonly used to...

Full description

Saved in:
Bibliographic Details
Main Authors: Qiong Zou, Borui Chen, Yang Zhang, Xi Wu, Yi Wan, Changsheng Chen
Format: Article
Language:English
Published: BMC 2024-12-01
Series:BMC Medical Research Methodology
Subjects:
Online Access:https://doi.org/10.1186/s12874-024-02442-9
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850253755568619520
author Qiong Zou
Borui Chen
Yang Zhang
Xi Wu
Yi Wan
Changsheng Chen
author_facet Qiong Zou
Borui Chen
Yang Zhang
Xi Wu
Yi Wan
Changsheng Chen
author_sort Qiong Zou
collection DOAJ
description Abstract Background Accurate fasting plasma glucose (FPG) trend prediction is important for management and treatment of patients with type 2 diabetes mellitus (T2DM), a globally prevalent chronic disease. (Generalised) linear mixed-effects (LME) models and machine learning (ML) are commonly used to analyse longitudinal data; however, the former is insufficient for dealing with complex, nonlinear data, whereas with the latter, random effects are ignored. The aim of this study was to develop LME, back propagation neural network (BPNN), and mixed-effects NN models that combine the 2 to predict FPG levels. Methods Monitoring data from 779 patients with T2DM from a multicentre, prospective study from the shared platform Figshare repository were divided 80/20 into training/test sets. The first 10 important features were modelled via random forest (RF) screening. First, an LME model was built to model interindividual differences, analyse the factors affecting FPG levels, compare the AIC and BIC values to screen the optimal model, and predict FPG levels. Second, multiple BPNN models were constructed via different variable sets to screen the optimal BPNN. Finally, an LME/BPNN combined model, named LMENN, was constructed via stacking integration. A 10-fold cross-validation cycle was performed using the training set to build the model and evaluate its performance, and then the final model was evaluated on the test set. Results The top 10 variables screened by RF were HOMA-β, HbA1c, HOMA–IR, urinary sugar, insulin, BMI, waist circumference, weight, age, and group. The best-fitting random-intercept mixed-effects (lm22) model showed that each patient’s baseline glucose levels influenced subsequent glucose measurements, but the trend over time was consistent. The LMENN model combines the strengths of LME and BPNN and accounts for random effects. The RMSE of the LMENN model ranges were 0.447–0.471 (training set), 0.525–0.552 (validation set), and 0.511–0.565 (test set). It improves the prediction performance of the single LME and BPNN models and shows some advantages in predicting FPG levels. Conclusions The LMENN model built by integrating LME and BPNN has several potential applications in analysing longitudinal FPG monitoring data. This study provides new ideas and methods for further research in the field of blood glucose prediction.
format Article
id doaj-art-d984b5f7d2ea4f6ea64a384418f865a0
institution OA Journals
issn 1471-2288
language English
publishDate 2024-12-01
publisher BMC
record_format Article
series BMC Medical Research Methodology
spelling doaj-art-d984b5f7d2ea4f6ea64a384418f865a02025-08-20T01:57:16ZengBMCBMC Medical Research Methodology1471-22882024-12-0124111910.1186/s12874-024-02442-9Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucoseQiong Zou0Borui Chen1Yang Zhang2Xi Wu3Yi Wan4Changsheng Chen5Department of Military Health Statistics, Faculty of Preventive Medicine, Air Force Medical University/Ministry of Education Key Lab of Hazard Assessment and Control in Special Operational EnvironmentSchool of Energy and Environment, City University of Hong KongDepartment of Military Health Statistics, Faculty of Preventive Medicine, Air Force Medical University/Ministry of Education Key Lab of Hazard Assessment and Control in Special Operational EnvironmentDepartment of Military Health Statistics, Faculty of Preventive Medicine, Air Force Medical University/Ministry of Education Key Lab of Hazard Assessment and Control in Special Operational EnvironmentDepartment of Health Services, Air Force Medical UniversityDepartment of Military Health Statistics, Faculty of Preventive Medicine, Air Force Medical University/Ministry of Education Key Lab of Hazard Assessment and Control in Special Operational EnvironmentAbstract Background Accurate fasting plasma glucose (FPG) trend prediction is important for management and treatment of patients with type 2 diabetes mellitus (T2DM), a globally prevalent chronic disease. (Generalised) linear mixed-effects (LME) models and machine learning (ML) are commonly used to analyse longitudinal data; however, the former is insufficient for dealing with complex, nonlinear data, whereas with the latter, random effects are ignored. The aim of this study was to develop LME, back propagation neural network (BPNN), and mixed-effects NN models that combine the 2 to predict FPG levels. Methods Monitoring data from 779 patients with T2DM from a multicentre, prospective study from the shared platform Figshare repository were divided 80/20 into training/test sets. The first 10 important features were modelled via random forest (RF) screening. First, an LME model was built to model interindividual differences, analyse the factors affecting FPG levels, compare the AIC and BIC values to screen the optimal model, and predict FPG levels. Second, multiple BPNN models were constructed via different variable sets to screen the optimal BPNN. Finally, an LME/BPNN combined model, named LMENN, was constructed via stacking integration. A 10-fold cross-validation cycle was performed using the training set to build the model and evaluate its performance, and then the final model was evaluated on the test set. Results The top 10 variables screened by RF were HOMA-β, HbA1c, HOMA–IR, urinary sugar, insulin, BMI, waist circumference, weight, age, and group. The best-fitting random-intercept mixed-effects (lm22) model showed that each patient’s baseline glucose levels influenced subsequent glucose measurements, but the trend over time was consistent. The LMENN model combines the strengths of LME and BPNN and accounts for random effects. The RMSE of the LMENN model ranges were 0.447–0.471 (training set), 0.525–0.552 (validation set), and 0.511–0.565 (test set). It improves the prediction performance of the single LME and BPNN models and shows some advantages in predicting FPG levels. Conclusions The LMENN model built by integrating LME and BPNN has several potential applications in analysing longitudinal FPG monitoring data. This study provides new ideas and methods for further research in the field of blood glucose prediction.https://doi.org/10.1186/s12874-024-02442-9Diabetes mellitus type 2Longitudinal dataMixed-effects neural network modelFasting plasma glucose
spellingShingle Qiong Zou
Borui Chen
Yang Zhang
Xi Wu
Yi Wan
Changsheng Chen
Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose
BMC Medical Research Methodology
Diabetes mellitus type 2
Longitudinal data
Mixed-effects neural network model
Fasting plasma glucose
title Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose
title_full Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose
title_fullStr Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose
title_full_unstemmed Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose
title_short Mixed-effects neural network modelling to predict longitudinal trends in fasting plasma glucose
title_sort mixed effects neural network modelling to predict longitudinal trends in fasting plasma glucose
topic Diabetes mellitus type 2
Longitudinal data
Mixed-effects neural network model
Fasting plasma glucose
url https://doi.org/10.1186/s12874-024-02442-9
work_keys_str_mv AT qiongzou mixedeffectsneuralnetworkmodellingtopredictlongitudinaltrendsinfastingplasmaglucose
AT boruichen mixedeffectsneuralnetworkmodellingtopredictlongitudinaltrendsinfastingplasmaglucose
AT yangzhang mixedeffectsneuralnetworkmodellingtopredictlongitudinaltrendsinfastingplasmaglucose
AT xiwu mixedeffectsneuralnetworkmodellingtopredictlongitudinaltrendsinfastingplasmaglucose
AT yiwan mixedeffectsneuralnetworkmodellingtopredictlongitudinaltrendsinfastingplasmaglucose
AT changshengchen mixedeffectsneuralnetworkmodellingtopredictlongitudinaltrendsinfastingplasmaglucose