Screening performance and optimal threshold of the American Diabetes Association Diabetes Risk Score in a large multistate Sudanese population: a cross-sectional study.

Authors

Sufian Khalid Noor, Hassan Salah Aldin Hassan Mohammed Ali, Mohamedelmustafa Yahya Mohamed Eldouma, Wagas Abbas Mohamed Ahmed Mansour, Elamin Abdelgadir, Ebtihag Zaki Altahir Almaryoud, Moram Yhia Abd Allah Moram Samel, Awab Mohammed Abass Mohammed, Wiam Abdin Omara, Shatha Gamer El Dawla, Hadeel Taha, Makia Mohamed

Published in

BMJ public health. Volume 4. Issue 4. Pages e004767. Epub Nov 09, 2026.

Abstract

Validated non-invasive risk stratification tools can support diabetes prevention in resource-constrained settings. The American Diabetes Association (ADA) Diabetes Risk Score is widely used but has not been rigorously evaluated in sub-Saharan African populations. This study assessed the screening utility and optimal score threshold of the ADA Diabetes Risk Score in the population in Sudan.
A cross-sectional, population-based study was conducted in 2023 across eight Sudanese states, using multistage cluster random sampling from 32 community health centres. Sudanese adult ≥18 years completed the self-administered ADA Diabetes Risk Test questionnaire (7-item, maximum score 10). Among the 1855 eligible participants, 203 with a prior physician diagnosis of diabetes were retained as the reference group for diagnostic performance analysis. Associations were evaluated using χ² tests and multivariable logistic regression. Receiver operating characteristic analysis with Youden index optimisation was used to derive the optimal score threshold. All inferential analyses were performed using R (V.4.4.3).
Among participants (64.3% female; 80.3% aged <40 years), 10.8% (n=200) were classified as high-risk by the standard ADA threshold (score ≥5). Using physician-diagnosed diabetes (n=203) as the reference, the Youden-optimal threshold of 3.5 achieved sensitivity 59.6% (95% CI 52.2% to 66.0%) and specificity 79.7% (95% CI 77.8% to 81.5%), with area under the curve 0.744 (95% CI 0.707 to 0.781, p<0.001). The overall number needed to screen was 9.3, ranging from 43.8 in adults aged <40 years to 1.1 in those aged >60 years. On multivariable logistic regression, hyperlipidaemia (OR: 5.33, 95% CI 3.43 to 8.17) and gestational diabetes (OR: 4.50, 95% CI 1.62 to 11.8) were the strongest independent predictors of high-risk ADA classification.
The ADA Diabetes Risk Score demonstrated moderate screening utility in this Sudanese population. A threshold of 3.5 may improve sensitivity for identifying at-risk individuals compared to the standard cut-off of ≥5. These findings are hypothesis-generating; external validation against biochemical endpoints is required before clinical adoption.

PMID:
43022903
Bibliographic data and abstract were imported from PubMed on 13 Nov 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.