Loading…
Prediction of Vitamin D Deficiency in Older Adults: The Role of Machine Learning Models
Context: Conventional prediction models for vitamin D deficiency have limited accuracy. Background: Using cross-sectional data, we developed models based on machine learning (ML) and compared their performance with those based on a conventional approach. Methods: Participants were 5106 community-res...
Saved in:
| Published in: | The journal of clinical endocrinology and metabolism 2022-10, Vol.107 (10), p.2737-2747 |
|---|---|
| Main Authors: | , , , , , |
| Format: | Article |
| Language: | English |
| Subjects: | |
| Citations: | Items that this one cites Items that cite this one |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| cited_by | cdi_FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3 |
|---|---|
| cites | cdi_FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3 |
| container_end_page | 2747 |
| container_issue | 10 |
| container_start_page | 2737 |
| container_title | The journal of clinical endocrinology and metabolism |
| container_volume | 107 |
| creator | Sluyter, John D Raita, Yoshihiko Hasegawa, Kohei Reid, Ian R Scragg, Robert Camargo, Carlos A |
| description | Context: Conventional prediction models for vitamin D deficiency have limited accuracy. Background: Using cross-sectional data, we developed models based on machine learning (ML) and compared their performance with those based on a conventional approach. Methods: Participants were 5106 community-resident adults (50-84 years; 58% male). In the randomly sampled training set (65%), we constructed 5 ML models: lasso regression, elastic net regression, random forest, gradient boosted decision tree, and dense neural network. The reference model was a logistic regression model. Outcomes were deseasonalized serum 25-hydroxyvitamin D (25(OH)D) |
| doi_str_mv | 10.1210/clinem/dgac432 |
| format | article |
| fullrecord | <record><control><sourceid>gale_proqu</sourceid><recordid>TN_cdi_proquest_miscellaneous_2694415697</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><galeid>A777587686</galeid><sourcerecordid>A777587686</sourcerecordid><originalsourceid>FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3</originalsourceid><addsrcrecordid>eNptkc1rGzEQxUVpoG7Sa8-CXnJZW9pdSavcjJO2AYeEkn7chDQ76yhoJUdaH_Lfd41zDHMYZvi9eTCPkK-cLXnN2QqCjziu-p2Ftqk_kAXXragU1-ojWTBW80qr-t8n8rmUZ8Z424pmQf4-ZOw9TD5Fmgb6x0929JFe02scPHiM8Ern-T70mOm6P4SpXNHHJ6S_UsCj4s7C0-xLt2hz9HFH71KPoVyQs8GGgl_e-jn5_f3mcfOz2t7_uN2stxU0Sk-VYx3UzDZN58AJgU6oWkk2tFJaxwVYDV2vnQDJmNNOamgcaKF0hx20CM05uTzd3ef0csAymdEXwBBsxHQoppa6bbmQWs3otxO6swGNj0OasoUjbtZKKdEp2cmZWr5DzdXj6CHF-S3z_j0B5FRKxsHssx9tfjWcmWMw5hSMeQum-Q_GPoGa</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2694415697</pqid></control><display><type>article</type><title>Prediction of Vitamin D Deficiency in Older Adults: The Role of Machine Learning Models</title><source>Oxford Journals Online</source><creator>Sluyter, John D ; Raita, Yoshihiko ; Hasegawa, Kohei ; Reid, Ian R ; Scragg, Robert ; Camargo, Carlos A</creator><creatorcontrib>Sluyter, John D ; Raita, Yoshihiko ; Hasegawa, Kohei ; Reid, Ian R ; Scragg, Robert ; Camargo, Carlos A</creatorcontrib><description><![CDATA[Context: Conventional prediction models for vitamin D deficiency have limited accuracy. Background: Using cross-sectional data, we developed models based on machine learning (ML) and compared their performance with those based on a conventional approach. Methods: Participants were 5106 community-resident adults (50-84 years; 58% male). In the randomly sampled training set (65%), we constructed 5 ML models: lasso regression, elastic net regression, random forest, gradient boosted decision tree, and dense neural network. The reference model was a logistic regression model. Outcomes were deseasonalized serum 25-hydroxyvitamin D (25(OH)D) <50 nmol/L (yes/no) and <25 nmol/L (yes/no). In the test set (the remaining 35%), we evaluated predictive performance of each model, including area under the receiver operating characteristic curve (AUC) and net benefit (decision curves). Results: Overall, 1270 (25%) and 91 (2%) had 25(OH)D <50 and <25 nmol/L, respectively. Compared with the reference model, the ML models predicted 25(OH)D <50 nmol/L with similar accuracy. However, for prediction of 25(OH)D <25 nmol/L, all ML models had higher AUC point estimates than the reference model by up to 0.14. AUC was highest for elastic net regression (0.93; 95% CI 0.90-0.96), compared with 0.81 (95% CI 0.71-0.91) for the reference model. In the decision curve analysis, ML models mostly achieved a greater net benefit across a range of thresholds. Conclusion: Compared with conventional models, ML models predicted 25(OH)D <50 nmol/L with similar accuracy but they predicted 25(OH) D <25 nmol/L with greater accuracy. The latter finding suggests a role for ML models in participant selection for vitamin D supplement trials. Keywords: Vitamin D, vitamin D deficiency, machine learning, prediction Abbreviations: 25(OH)D, 25-hydroxyvitamin D; AUC, area under the receiver operating characteristic curve; BMI, body mass index; BP, blood pressure; IDI, integrated discrimination improvement; IS, integral of sensitivity; IP, integral of (1--specificity); ML, machine learning; NZDep13, 2013 New Zealand Deprivation Index; RCT, randomized controlled trial.]]></description><identifier>ISSN: 0021-972X</identifier><identifier>EISSN: 1945-7197</identifier><identifier>DOI: 10.1210/clinem/dgac432</identifier><language>eng</language><publisher>Oxford University Press</publisher><subject>Aged ; Alfacalcidol ; Calcifediol ; Comparative analysis ; Machine learning ; Neural networks ; Vitamin D ; Vitamin D deficiency</subject><ispartof>The journal of clinical endocrinology and metabolism, 2022-10, Vol.107 (10), p.2737-2747</ispartof><rights>COPYRIGHT 2022 Oxford University Press</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3</citedby><cites>FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3</cites><orcidid>0000-0001-6021-5458 ; 0000-0002-9722-139X</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,777,781,27905,27906</link.rule.ids></links><search><creatorcontrib>Sluyter, John D</creatorcontrib><creatorcontrib>Raita, Yoshihiko</creatorcontrib><creatorcontrib>Hasegawa, Kohei</creatorcontrib><creatorcontrib>Reid, Ian R</creatorcontrib><creatorcontrib>Scragg, Robert</creatorcontrib><creatorcontrib>Camargo, Carlos A</creatorcontrib><title>Prediction of Vitamin D Deficiency in Older Adults: The Role of Machine Learning Models</title><title>The journal of clinical endocrinology and metabolism</title><description><![CDATA[Context: Conventional prediction models for vitamin D deficiency have limited accuracy. Background: Using cross-sectional data, we developed models based on machine learning (ML) and compared their performance with those based on a conventional approach. Methods: Participants were 5106 community-resident adults (50-84 years; 58% male). In the randomly sampled training set (65%), we constructed 5 ML models: lasso regression, elastic net regression, random forest, gradient boosted decision tree, and dense neural network. The reference model was a logistic regression model. Outcomes were deseasonalized serum 25-hydroxyvitamin D (25(OH)D) <50 nmol/L (yes/no) and <25 nmol/L (yes/no). In the test set (the remaining 35%), we evaluated predictive performance of each model, including area under the receiver operating characteristic curve (AUC) and net benefit (decision curves). Results: Overall, 1270 (25%) and 91 (2%) had 25(OH)D <50 and <25 nmol/L, respectively. Compared with the reference model, the ML models predicted 25(OH)D <50 nmol/L with similar accuracy. However, for prediction of 25(OH)D <25 nmol/L, all ML models had higher AUC point estimates than the reference model by up to 0.14. AUC was highest for elastic net regression (0.93; 95% CI 0.90-0.96), compared with 0.81 (95% CI 0.71-0.91) for the reference model. In the decision curve analysis, ML models mostly achieved a greater net benefit across a range of thresholds. Conclusion: Compared with conventional models, ML models predicted 25(OH)D <50 nmol/L with similar accuracy but they predicted 25(OH) D <25 nmol/L with greater accuracy. The latter finding suggests a role for ML models in participant selection for vitamin D supplement trials. Keywords: Vitamin D, vitamin D deficiency, machine learning, prediction Abbreviations: 25(OH)D, 25-hydroxyvitamin D; AUC, area under the receiver operating characteristic curve; BMI, body mass index; BP, blood pressure; IDI, integrated discrimination improvement; IS, integral of sensitivity; IP, integral of (1--specificity); ML, machine learning; NZDep13, 2013 New Zealand Deprivation Index; RCT, randomized controlled trial.]]></description><subject>Aged</subject><subject>Alfacalcidol</subject><subject>Calcifediol</subject><subject>Comparative analysis</subject><subject>Machine learning</subject><subject>Neural networks</subject><subject>Vitamin D</subject><subject>Vitamin D deficiency</subject><issn>0021-972X</issn><issn>1945-7197</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><recordid>eNptkc1rGzEQxUVpoG7Sa8-CXnJZW9pdSavcjJO2AYeEkn7chDQ76yhoJUdaH_Lfd41zDHMYZvi9eTCPkK-cLXnN2QqCjziu-p2Ftqk_kAXXragU1-ojWTBW80qr-t8n8rmUZ8Z424pmQf4-ZOw9TD5Fmgb6x0929JFe02scPHiM8Ern-T70mOm6P4SpXNHHJ6S_UsCj4s7C0-xLt2hz9HFH71KPoVyQs8GGgl_e-jn5_f3mcfOz2t7_uN2stxU0Sk-VYx3UzDZN58AJgU6oWkk2tFJaxwVYDV2vnQDJmNNOamgcaKF0hx20CM05uTzd3ef0csAymdEXwBBsxHQoppa6bbmQWs3otxO6swGNj0OasoUjbtZKKdEp2cmZWr5DzdXj6CHF-S3z_j0B5FRKxsHssx9tfjWcmWMw5hSMeQum-Q_GPoGa</recordid><startdate>20221001</startdate><enddate>20221001</enddate><creator>Sluyter, John D</creator><creator>Raita, Yoshihiko</creator><creator>Hasegawa, Kohei</creator><creator>Reid, Ian R</creator><creator>Scragg, Robert</creator><creator>Camargo, Carlos A</creator><general>Oxford University Press</general><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0001-6021-5458</orcidid><orcidid>https://orcid.org/0000-0002-9722-139X</orcidid></search><sort><creationdate>20221001</creationdate><title>Prediction of Vitamin D Deficiency in Older Adults: The Role of Machine Learning Models</title><author>Sluyter, John D ; Raita, Yoshihiko ; Hasegawa, Kohei ; Reid, Ian R ; Scragg, Robert ; Camargo, Carlos A</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Aged</topic><topic>Alfacalcidol</topic><topic>Calcifediol</topic><topic>Comparative analysis</topic><topic>Machine learning</topic><topic>Neural networks</topic><topic>Vitamin D</topic><topic>Vitamin D deficiency</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Sluyter, John D</creatorcontrib><creatorcontrib>Raita, Yoshihiko</creatorcontrib><creatorcontrib>Hasegawa, Kohei</creatorcontrib><creatorcontrib>Reid, Ian R</creatorcontrib><creatorcontrib>Scragg, Robert</creatorcontrib><creatorcontrib>Camargo, Carlos A</creatorcontrib><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>The journal of clinical endocrinology and metabolism</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Sluyter, John D</au><au>Raita, Yoshihiko</au><au>Hasegawa, Kohei</au><au>Reid, Ian R</au><au>Scragg, Robert</au><au>Camargo, Carlos A</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Prediction of Vitamin D Deficiency in Older Adults: The Role of Machine Learning Models</atitle><jtitle>The journal of clinical endocrinology and metabolism</jtitle><date>2022-10-01</date><risdate>2022</risdate><volume>107</volume><issue>10</issue><spage>2737</spage><epage>2747</epage><pages>2737-2747</pages><issn>0021-972X</issn><eissn>1945-7197</eissn><abstract><![CDATA[Context: Conventional prediction models for vitamin D deficiency have limited accuracy. Background: Using cross-sectional data, we developed models based on machine learning (ML) and compared their performance with those based on a conventional approach. Methods: Participants were 5106 community-resident adults (50-84 years; 58% male). In the randomly sampled training set (65%), we constructed 5 ML models: lasso regression, elastic net regression, random forest, gradient boosted decision tree, and dense neural network. The reference model was a logistic regression model. Outcomes were deseasonalized serum 25-hydroxyvitamin D (25(OH)D) <50 nmol/L (yes/no) and <25 nmol/L (yes/no). In the test set (the remaining 35%), we evaluated predictive performance of each model, including area under the receiver operating characteristic curve (AUC) and net benefit (decision curves). Results: Overall, 1270 (25%) and 91 (2%) had 25(OH)D <50 and <25 nmol/L, respectively. Compared with the reference model, the ML models predicted 25(OH)D <50 nmol/L with similar accuracy. However, for prediction of 25(OH)D <25 nmol/L, all ML models had higher AUC point estimates than the reference model by up to 0.14. AUC was highest for elastic net regression (0.93; 95% CI 0.90-0.96), compared with 0.81 (95% CI 0.71-0.91) for the reference model. In the decision curve analysis, ML models mostly achieved a greater net benefit across a range of thresholds. Conclusion: Compared with conventional models, ML models predicted 25(OH)D <50 nmol/L with similar accuracy but they predicted 25(OH) D <25 nmol/L with greater accuracy. The latter finding suggests a role for ML models in participant selection for vitamin D supplement trials. Keywords: Vitamin D, vitamin D deficiency, machine learning, prediction Abbreviations: 25(OH)D, 25-hydroxyvitamin D; AUC, area under the receiver operating characteristic curve; BMI, body mass index; BP, blood pressure; IDI, integrated discrimination improvement; IS, integral of sensitivity; IP, integral of (1--specificity); ML, machine learning; NZDep13, 2013 New Zealand Deprivation Index; RCT, randomized controlled trial.]]></abstract><pub>Oxford University Press</pub><doi>10.1210/clinem/dgac432</doi><tpages>11</tpages><orcidid>https://orcid.org/0000-0001-6021-5458</orcidid><orcidid>https://orcid.org/0000-0002-9722-139X</orcidid><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0021-972X |
| ispartof | The journal of clinical endocrinology and metabolism, 2022-10, Vol.107 (10), p.2737-2747 |
| issn | 0021-972X 1945-7197 |
| language | eng |
| recordid | cdi_proquest_miscellaneous_2694415697 |
| source | Oxford Journals Online |
| subjects | Aged Alfacalcidol Calcifediol Comparative analysis Machine learning Neural networks Vitamin D Vitamin D deficiency |
| title | Prediction of Vitamin D Deficiency in Older Adults: The Role of Machine Learning Models |
| url | http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-17T23%3A29%3A34IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-gale_proqu&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Prediction%20of%20Vitamin%20D%20Deficiency%20in%20Older%20Adults:%20The%20Role%20of%20Machine%20Learning%20Models&rft.jtitle=The%20journal%20of%20clinical%20endocrinology%20and%20metabolism&rft.au=Sluyter,%20John%20D&rft.date=2022-10-01&rft.volume=107&rft.issue=10&rft.spage=2737&rft.epage=2747&rft.pages=2737-2747&rft.issn=0021-972X&rft.eissn=1945-7197&rft_id=info:doi/10.1210/clinem/dgac432&rft_dat=%3Cgale_proqu%3EA777587686%3C/gale_proqu%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c379t-b08c20a338bcb55eb572760f466ab15ca9c8d9b5c600b9b69c3bc95798e8c4ec3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2694415697&rft_id=info:pmid/&rft_galeid=A777587686&rfr_iscdi=true |