A machine learning–based biological aging prediction and its associations with healthy lifestyles: the Dongfeng–Tongji cohort

Wang, Chenming; Guan, Xin; Bai, Yansen; Feng, Yue; Wei, Wei; Li, Hang; Li, Guyanan; Meng, Hua; Li, Mengying; Jie, Jiali; Fu, Ming; Wu, Xiulong; He, Meian; Zhang, Xiaomin; Yang, Handong; Lu, Yanjun; Guo, Huan

doi:10.1111/nyas.14685

Cited by 11 publications

(19 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Taking the previous XGB-BA as an example, the R 2 of the model in the test set was 0.27, while the correlation between BA and CA was 0.75 in the nal results (BA to CA regression belonged to simple linear regression, so R = cor = 0.75, R 2 = 0.56) [19]. The same was also found in the XGB-BA based on the Dongfeng-Tongji cohort [31]. This might be because the model trained on the training set predicted BA on the full dataset, which introduced interference from parameter tuning and training over tting.…”

Section: Discussionsupporting

confidence: 63%

“…The correlation between our STK-BA and CA (r = 0.66) on the test set was better than previously published BA (r = 0.52) based on 19 blood biomarkers [19] but weaker than BA (r = 0.74) which considered 44 biomarkers including lung function. This phenomenon is plausible, depending on the population-speci c and age-related biosignatures in different datasets [31]. However, it is worth noting that we showed better CA correlations with the same number of biomarkers in the Chinese population.…”

Section: Discussionmentioning

confidence: 61%

“…heart disease, kidney disease) can further examine its potential as a biomarker of aging in the general population [6,29,30]. We found in the previous ML-BA that the correlation between BA and CA in the full data and the test data showed signi cant differences [19,31]. This might be because the model trained on the training set predicted the full dataset's BA, introducing the interference of parameter tuning and training over tting.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A machine learning-based data mining in medical examination data: a biological features-based biological age prediction model

Yang

Gao

Lin

et al. 2022

Preprint

View full text Add to dashboard Cite

Background: Biological age (BA) has been recognized as a more accurate indicator of aging than chronological age (CA). However, current limitations include insufficient attention to the incompleteness of medical data used to construct BA; lack of machine learning-based BA (ML-BA) on the Chinese population; neglect of the influence of the degree of model overfitting on the stability of the correlation results.Methods and Results: Based on the medical examination data of the Chinese population (45-90 years), we first evaluated the most suitable missing interpolation method, then constructed 14 ML-BAs based on biomarkers, and finally explored the associations between ML-BAs and health statuses (healthy risk indicators and disease). We found that round-robin linear regression interpolation performed best, while AE showed the highest stability at high missing rates. We further illustrated the potential overfitting problem in ML-BAs, which affected the stability of ML-BAs' associations with health statuses. We then proposed a composite ML-BA based on the Stacking method with a simple meta-model (STK-BA), which overcame the overfitting problem, and associated more strongly with CA (r = 0.66, P < 0.001), healthy risk indicators, disease counts, and six types of disease.Conclusion: We provided an improved aging measurement method for middle-aged and elderly groups in China, which can more stably capture aging characteristics other than CA, supporting the emerging application potential of machine learning in aging research.

show abstract

Section: Discussionsupporting

confidence: 63%

Section: Discussionmentioning

confidence: 61%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A machine learning-based data mining in medical examination data: a biological features-based biological age prediction model

Yang

Gao

Lin

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…et al, 2020; Zhong et al, 2020;Gialluisi et al, 2021;Wang et al, 2021). However, there are few studies that compare AI techniques with traditional statistical methods to construct a BA prediction model using clinical biomarkers.…”

Section: Parametermentioning

confidence: 99%

Comparison of Biological Age Prediction Models Using Clinical Biomarkers Commonly Measured in Clinical Practice Settings: AI Techniques Vs. Traditional Statistical Methods

Bae¹,

Im²,

Lee³

et al. 2021

Front. Anal. Sci.

View full text Add to dashboard Cite

In this work, we used the health check-up data of more than 111,000 subjects for analysis, using only the data with all 35 variables entered. For the prediction of biological age, traditional statistical methods and four AI techniques (RF, XGB, SVR, and DNN), which are widely used recently, were simultaneously used to compare the predictive power. This study showed that AI models produced about 1.6 times stronger linear relationship on average than statistical models. In addition, the regression analysis on the predicted BA and CA revealed similar differences in terms of both the correlation coefficients (linear model: 0.831, polynomial model: 0.996, XGB model: 0.66, RF model: 0.927, SVR model: 0.787, DNN model: 0.998) and R2 values. Through this work, we confirmed that AI techniques such as the DNN model outperformed traditional statistical methods in predicting biological age.

show abstract

“…Notably, we found in the previous ML-BAs that the correlations between BA and CA attained from the test data used for comparing model performances, and the full data, including both the training data and test data, showed obvious differences. [ 18 , 29 ]. The reason might be that overfitting makes the model outperform the test set on the training set.…”

Section: Introductionmentioning

confidence: 99%

A machine learning-based data mining in medical examination data: a biological features-based biological age prediction model

et al. 2022

View full text Add to dashboard Cite

Background Biological age (BA) has been recognized as a more accurate indicator of aging than chronological age (CA). However, the current limitations include: insufficient attention to the incompleteness of medical data for constructing BA; Lack of machine learning-based BA (ML-BA) on the Chinese population; Neglect of the influence of model overfitting degree on the stability of the association results. Methods and results Based on the medical examination data of the Chinese population (45–90 years), we first evaluated the most suitable missing interpolation method, then constructed 14 ML-BAs based on biomarkers, and finally explored the associations between ML-BAs and health statuses (healthy risk indicators and disease). We found that round-robin linear regression interpolation performed best, while AutoEncoder showed the highest interpolation stability. We further illustrated the potential overfitting problem in ML-BAs, which affected the stability of ML-Bas’ associations with health statuses. We then proposed a composite ML-BA based on the Stacking method with a simple meta-model (STK-BA), which overcame the overfitting problem, and associated more strongly with CA (r = 0.66, P < 0.001), healthy risk indicators, disease counts, and six types of disease. Conclusion We provided an improved aging measurement method for middle-aged and elderly groups in China, which can more stably capture aging characteristics other than CA, supporting the emerging application potential of machine learning in aging research.

show abstract

A machine learning–based biological aging prediction and its associations with healthy lifestyles: the Dongfeng–Tongji cohort

Cited by 11 publications

References 55 publications

A machine learning-based data mining in medical examination data: a biological features-based biological age prediction model

A machine learning-based data mining in medical examination data: a biological features-based biological age prediction model

Comparison of Biological Age Prediction Models Using Clinical Biomarkers Commonly Measured in Clinical Practice Settings: AI Techniques Vs. Traditional Statistical Methods

A machine learning-based data mining in medical examination data: a biological features-based biological age prediction model

Contact Info

Product

Resources

About