Cargando…

Development and validation of multimorbidity index predicting mortality among Chinese older adults

This study aimed to construct a multimorbidity index among Chinese older adults. Participants aged 65-84 years (n=11,757) from the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Fourteen self-reported chronic conditions were assessed at baseline. Outcome was all-cause mortality within five-y...

Descripción completa

Detalles Bibliográficos
Autores principales: Luo, Yan, Huang, Zi-Ting, Xu, Hui-wen, Chen, zi-shuo, Su, He-Xuan, Liu, Hui, Xu, Beibei
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8968286/
http://dx.doi.org/10.1093/geroni/igab046.2324
_version_ 1784679015765770240
author Luo, Yan
Huang, Zi-Ting
Xu, Hui-wen
Chen, zi-shuo
Su, He-Xuan
Liu, Hui
Xu, Beibei
author_facet Luo, Yan
Huang, Zi-Ting
Xu, Hui-wen
Chen, zi-shuo
Su, He-Xuan
Liu, Hui
Xu, Beibei
author_sort Luo, Yan
collection PubMed
description This study aimed to construct a multimorbidity index among Chinese older adults. Participants aged 65-84 years (n=11,757) from the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Fourteen self-reported chronic conditions were assessed at baseline. Outcome was all-cause mortality within five-year follow-up. We used restrictive association rules mining to identify the patterns of multiple chronic conditions associated with mortality. The weights of conditions and disease combinations were assigned using logistic regression adjusted by age and sex in training set. Multimorbidity index (MI) with individual diseases and multimorbidity index incorporating disease combinations (MIDC) were developed. We compared the performance of MI and MIDC with condition count and XGBoost algorithm in the validation set. There were no significant differences of c-statistics between condition count (0.687) and MI (0.692) or MIDC (0.689). The c-statistic of XGBoost algorithm (0.675) was the lowest among all models. The Integrated Discrimination Improvement (IDI) and categorical Net Reclassification Index (NRI) for MI (IDI: 0.01, P < 0.001; NRI: 0.01, P = 0.127), MIDC (IDI: 0.004, p = 0.002; NRI: 0.02, P = 0.033), and XGBoost model (IDI: 0.02, P < 0.001; NRI: 0.03, P = 0.004) were significantly positive compared with condition count. However, no significant differences for IDI and NRI were observed between MI and MIDC. Among Chinese older adults, weighted multimorbidity index with individual disease can better predict five-year mortality risk over condition count. There was little improvement in the predictive performance of the index after considering the joint effects of disease combinations.
format Online
Article
Text
id pubmed-8968286
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher Oxford University Press
record_format MEDLINE/PubMed
spelling pubmed-89682862022-03-31 Development and validation of multimorbidity index predicting mortality among Chinese older adults Luo, Yan Huang, Zi-Ting Xu, Hui-wen Chen, zi-shuo Su, He-Xuan Liu, Hui Xu, Beibei Innov Aging Abstracts This study aimed to construct a multimorbidity index among Chinese older adults. Participants aged 65-84 years (n=11,757) from the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Fourteen self-reported chronic conditions were assessed at baseline. Outcome was all-cause mortality within five-year follow-up. We used restrictive association rules mining to identify the patterns of multiple chronic conditions associated with mortality. The weights of conditions and disease combinations were assigned using logistic regression adjusted by age and sex in training set. Multimorbidity index (MI) with individual diseases and multimorbidity index incorporating disease combinations (MIDC) were developed. We compared the performance of MI and MIDC with condition count and XGBoost algorithm in the validation set. There were no significant differences of c-statistics between condition count (0.687) and MI (0.692) or MIDC (0.689). The c-statistic of XGBoost algorithm (0.675) was the lowest among all models. The Integrated Discrimination Improvement (IDI) and categorical Net Reclassification Index (NRI) for MI (IDI: 0.01, P < 0.001; NRI: 0.01, P = 0.127), MIDC (IDI: 0.004, p = 0.002; NRI: 0.02, P = 0.033), and XGBoost model (IDI: 0.02, P < 0.001; NRI: 0.03, P = 0.004) were significantly positive compared with condition count. However, no significant differences for IDI and NRI were observed between MI and MIDC. Among Chinese older adults, weighted multimorbidity index with individual disease can better predict five-year mortality risk over condition count. There was little improvement in the predictive performance of the index after considering the joint effects of disease combinations. Oxford University Press 2021-12-17 /pmc/articles/PMC8968286/ http://dx.doi.org/10.1093/geroni/igab046.2324 Text en © The Author(s) 2021. Published by Oxford University Press on behalf of The Gerontological Society of America. https://creativecommons.org/licenses/by/4.0/This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/ (https://creativecommons.org/licenses/by/4.0/) ), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
spellingShingle Abstracts
Luo, Yan
Huang, Zi-Ting
Xu, Hui-wen
Chen, zi-shuo
Su, He-Xuan
Liu, Hui
Xu, Beibei
Development and validation of multimorbidity index predicting mortality among Chinese older adults
title Development and validation of multimorbidity index predicting mortality among Chinese older adults
title_full Development and validation of multimorbidity index predicting mortality among Chinese older adults
title_fullStr Development and validation of multimorbidity index predicting mortality among Chinese older adults
title_full_unstemmed Development and validation of multimorbidity index predicting mortality among Chinese older adults
title_short Development and validation of multimorbidity index predicting mortality among Chinese older adults
title_sort development and validation of multimorbidity index predicting mortality among chinese older adults
topic Abstracts
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8968286/
http://dx.doi.org/10.1093/geroni/igab046.2324
work_keys_str_mv AT luoyan developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults
AT huangziting developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults
AT xuhuiwen developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults
AT chenzishuo developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults
AT suhexuan developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults
AT liuhui developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults
AT xubeibei developmentandvalidationofmultimorbidityindexpredictingmortalityamongchineseolderadults