Development and validation of a dementia risk prediction model for low- and middle-income countries: the 10/66 study
Abstract
Most people with dementia live in low- and middle-income countries (LMICs), underscoring the need for LMIC-specific identification of high-risk individuals. This study aimed to develop and validate a simple dementia risk prediction model for these settings. Data from 7 10/66 Study sites were analyzed. Over 100 candidate predictors were screened based on existing models and the 2024 Lancet Commission, including LMIC-specific variables (eg, food insecurity and household assets). Predictors were selected using LASSO and modelled with the Fine–Gray method to generate a risk score. Predictive accuracy was pooled via meta-analysis. Eleven thousand one hundred forty-three participants were included, among whom 1069 (9.6%) developed dementia during follow-up. A 5-factor risk score comprising age, social engagement, physical activity, hypertension, and difficulty in handling money was developed. The pooled c-statistic was 0.75 (95% CI, 0.72-0.78), with good calibration across sites. Decision curve analysis showed a modest net benefit, with variation across countries. It is possible to predict incident dementia with reasonable accuracy using a simple model across different LMICs. Our findings support the use of context-specific risk assessment tools to identify individuals at elevated dementia risk in LMIC settings, which may inform resource allocation for dementia care services and public health planning.
Related
- Development and Validation of a Sleep Disorder Risk Prediction Model in Young and Middle-Aged Adults With Anemia
- Supply Chain and Workforce Capacity Development for Diagnostic Services in Low-and Lower Middle-Income Sub-Saharan Countries
- Evidence landscape for surgical site infection prevention strategies in low- and middle-income countries: a scoping review of literature for intervention development as part of the IMPRESS study
- Development and External Validation of a Plasma p-tau217-Based Model for MCI-to-Dementia Progression
- Risk factors for post-PCI restenosis in patients with coronary heart disease and the development and validation of a nomogram prediction model
BibTeX
@article{pakpahan2026development,
title = {{Development and validation of a dementia risk prediction model for low- and middle-income countries: the 10/66 study}},
author = {Eduwin Pakpahan and Zhongyang Guan and Mario Siervo and Graciela Muniz-Terrera and Devi Mohan and Daisy Acosta and Ana Luisa Sosa and Isaac Acosta and Juan J Llibre-Rodriguez and Jorge J Llibre-Guerra and Martin Prince and Alice Worrall and Aliya Naheed and Ashleigh S Vella and Jiyang Jiang and Darren M Lipnicki and Perminder S Sachdev and Louise Robinson and Matthew Prina and Blossom C M Stephan},
journal = {American Journal of Epidemiology},
year = {2026},
month = jun,
doi = {10.1093/aje/kwag116},
url = {https://doi.org/10.1093/aje/kwag116},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Development and validation of a dementia risk prediction model for low- and middle-income countries: the 10/66 study
AU - Eduwin Pakpahan
AU - Zhongyang Guan
AU - Mario Siervo
AU - Graciela Muniz-Terrera
AU - Devi Mohan
AU - Daisy Acosta
AU - Ana Luisa Sosa
AU - Isaac Acosta
AU - Juan J Llibre-Rodriguez
AU - Jorge J Llibre-Guerra
AU - Martin Prince
AU - Alice Worrall
AU - Aliya Naheed
AU - Ashleigh S Vella
AU - Jiyang Jiang
AU - Darren M Lipnicki
AU - Perminder S Sachdev
AU - Louise Robinson
AU - Matthew Prina
AU - Blossom C M Stephan
PY - 2026
DA - 2026/06/02
JO - American Journal of Epidemiology
DO - 10.1093/aje/kwag116
UR - https://doi.org/10.1093/aje/kwag116
AB - Most people with dementia live in low- and middle-income countries (LMICs), underscoring the need for LMIC-specific identification of high-risk individuals. This study aimed to develop and validate a simple dementia risk prediction model for these settings. Data from 7 10/66 Study sites were analyzed. Over 100 candidate predictors were screened based on existing models and the 2024 Lancet Commission, including LMIC-specific variables (eg, food insecurity and household assets). Predictors were selected using LASSO and modelled with the Fine–Gray method to generate a risk score. Predictive accuracy was pooled via meta-analysis. Eleven thousand one hundred forty-three participants were included, among whom 1069 (9.6%) developed dementia during follow-up. A 5-factor risk score comprising age, social engagement, physical activity, hypertension, and difficulty in handling money was developed. The pooled c-statistic was 0.75 (95% CI, 0.72-0.78), with good calibration across sites. Decision curve analysis showed a modest net benefit, with variation across countries. It is possible to predict incident dementia with reasonable accuracy using a simple model across different LMICs. Our findings support the use of context-specific risk assessment tools to identify individuals at elevated dementia risk in LMIC settings, which may inform resource allocation for dementia care services and public health planning.
SN - 0002-9262
SN - 1476-6256
ER -