{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T16:30:13Z","timestamp":1793550613501,"version":"4.5.0"},"posted":{"date-parts":[[2026,6,24]]},"group-title":"Public and Global Health","reference-count":23,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T00:00:00Z","timestamp":1782259200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/publicdomain\/zero\/1.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,11,1]]},"abstract":"<jats:p>Background. Cognitive test performance in India is shaped by education, literacy and language as well as by neurodegeneration, which limits classification based on cognitive tests alone. We examined whether combining cognitive, informant-reported, sociodemographic and cardiometabolic information improves dementia classification, and which domains carry the gain.\n\nMethods. We analysed the 2,528 LASI-DAD Wave 1 respondents with a clinical consensus Clinical Dementia Rating (CDR). Dementia was defined as CDR of 1 or above (192 cases, 7.6%). Predictors were eight cognitive measures, the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE), and age, sex and education. Ten cardiometabolic assays were evaluated in the 1,788 respondents with blood results. Five algorithms were compared using nested cross-validation (5 outer, 3 inner folds). We assessed discrimination, calibration and net benefit, tested each domain's contribution with paired bootstrap differences in ROC-AUC, and examined predictors with permutation importance and SHAP.\n\nResults. Logistic regression achieved a ROC-AUC of 0.953 (95% CI 0.937 to 0.966), with calibration intercept 0.00, calibration slope 1.11 and Brier score 0.037. Random forest, gradient boosting and XGBoost performed similarly (0.946 to 0.947). Cognitive measures alone gave a ROC-AUC of 0.899. Adding the IQCODE increased it by 0.052 (95% CI 0.038 to 0.067); adding sociodemographic variables changed it by +0.0020 (-0.00004 to 0.0040) and adding cardiometabolic assays by +0.0032 (-0.0004 to 0.0073). At a training-derived threshold, sensitivity was 0.875, specificity 0.902, positive predictive value 0.422 and negative predictive value 0.989. Observed dementia prevalence rose from 2.1% to 93.5% across five risk strata, and the model gave greater net benefit than referring all or none at every threshold from 0.01 to 0.50. Oversampling with SMOTE did not change discrimination (0.952) but raised mean predicted risk to 21.2% against an observed 7.6%.\n\nConclusions. An interpretable logistic regression model classified clinically rated dementia with high discrimination and good calibration. Its advantage over cognitive tests came almost entirely from informant-reported decline; sociodemographic characteristics and cardiometabolic biomarkers added negligible discrimination. The clinical reference standard incorporated the same informant and cognitive information, and external validation is required before the model is used.<\/jats:p>","DOI":"10.64898\/2026.06.13.26354833","type":"posted-content","created":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T17:20:20Z","timestamp":1782321620000},"source":"Crossref","is-referenced-by-count":0,"title":["A Multidomain Model for Dementia Classification using Harmonized LASI and LASI-DAD Data"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-3532-3719","authenticated-orcid":false,"given":"Shweta","family":"Anand","sequence":"first","affiliation":[{"name":"Indian Institute of Technology Gandhinagar"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Krishna","family":"Miyapuram","sequence":"additional","affiliation":[{"name":"Indian Institute of Technology Gandhinagar"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026062608323206000_2026.06.13.26354833v1.1","doi-asserted-by":"publisher","DOI":"10.1007\/s11065-005-9180-y"},{"key":"2026062608323206000_2026.06.13.26354833v1.2","doi-asserted-by":"publisher","DOI":"10.1097\/01.mlr.0000245427.22788.be"},{"key":"2026062608323206000_2026.06.13.26354833v1.3","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuropsychologia.2009.03.004"},{"key":"2026062608323206000_2026.06.13.26354833v1.4","unstructured":"Prince M , Wimo A , Guerchet M , Ali G-C , Wu Y-T , Prina M. World Alzheimer Report 2015: the global impact of dementia. Alzheimer\u2019s Disease International; 2015. https:\/\/www.alzint.org\/u\/WorldAlzheimerReport2015.pdf"},{"key":"2026062608323206000_2026.06.13.26354833v1.5","unstructured":"World Health Organization. Global status report on the public health response to dementia. WHO; 2021. https:\/\/www.who.int\/publications\/i\/item\/9789240033245"},{"key":"2026062608323206000_2026.06.13.26354833v1.6","doi-asserted-by":"publisher","DOI":"10.1017\/S1041610204000390"},{"key":"2026062608323206000_2026.06.13.26354833v1.7","doi-asserted-by":"publisher","DOI":"10.1212\/01.WNL.0000149519.47454.F2"},{"key":"2026062608323206000_2026.06.13.26354833v1.8","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.0000000000007094"},{"key":"2026062608323206000_2026.06.13.26354833v1.9","doi-asserted-by":"publisher","DOI":"10.1161\/CIRCULATIONAHA.115.001593"},{"key":"2026062608323206000_2026.06.13.26354833v1.10","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-019-0048-x"},{"key":"2026062608323206000_2026.06.13.26354833v1.11","unstructured":"Lundberg SM , Lee S-I . A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems. 2017;30. https:\/\/paper.nips.cc\/paper_files\/paper\/2017\/hash\/8a20a8621978632d76c43dfd28b67767-Abstract.html"},{"key":"2026062608323206000_2026.06.13.26354833v1.12","unstructured":"Arokiasamy P , Bloom D , Lee J , Feeney K , Ozolins M. Longitudinal Aging Study in India: vision, design, implementation, and preliminary findings. In: Smith J , Majmundar M , editors. Aging in Asia: findings from new and emerging data initiatives. National Academies Press; 2012. Pp. 36\u201374."},{"key":"2026062608323206000_2026.06.13.26354833v1.13","unstructured":"International Institute for Population Sciences (IIPS), National Programme for Health Care of Elderly (NPHCE), MoHFW, Harvard T.H. Chan School of Public Health, University of Southern California. Longitudinal Ageing Study in India (LASI) Wave 1, 2017-18, India Report. IIPS; 2020."},{"key":"2026062608323206000_2026.06.13.26354833v1.14","doi-asserted-by":"publisher","DOI":"10.1111\/jgs.16737"},{"key":"2026062608323206000_2026.06.13.26354833v1.15","doi-asserted-by":"publisher","DOI":"10.1002\/alz.12928"},{"key":"2026062608323206000_2026.06.13.26354833v1.16","doi-asserted-by":"publisher","DOI":"10.25549\/h5wx-ay45"},{"key":"2026062608323206000_2026.06.13.26354833v1.17","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.b2393"},{"key":"2026062608323206000_2026.06.13.26354833v1.18","doi-asserted-by":"publisher","DOI":"10.1613\/jair.953"},{"key":"2026062608323206000_2026.06.13.26354833v1.19","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0118432"},{"key":"2026062608323206000_2026.06.13.26354833v1.20","doi-asserted-by":"publisher","DOI":"10.3233\/JAD-190541"},{"key":"2026062608323206000_2026.06.13.26354833v1.21","doi-asserted-by":"publisher","DOI":"10.1016\/S0140-6736(20)30367-6"},{"key":"2026062608323206000_2026.06.13.26354833v1.22","doi-asserted-by":"publisher","DOI":"10.1038\/s41574-018-0048-7"},{"key":"2026062608323206000_2026.06.13.26354833v1.23","doi-asserted-by":"publisher","DOI":"10.1136\/jnnp-2019-321913"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.06.13.26354833","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,11,1]],"date-time":"2026-11-01T16:15:13Z","timestamp":1793549713000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.06.13.26354833"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,24]]},"references-count":23,"URL":"https:\/\/doi.org\/10.64898\/2026.06.13.26354833","relation":{},"subject":[],"published":{"date-parts":[[2026,6,24]]},"subtype":"preprint"}}