This page is alzhub.org's plain-English summary of a published study, not the paper itself. It is generated from the record we hold for this study, sourced and linked below — check the original for the full text and methods.

59/100

✓ Peer-reviewed alzhub usefulness score 59 out of 100. Digital health tools

Topic: Amyloid, tau and blood-test diagnosis

Plain-language summary

Nationwide population-based cohort provides a new opportunity to build an automated risk prediction model based on individuals' history of health and healthcare beyond existing risk prediction models. We tested the possibility of machine learning models to predict future incidence of Alzheimer's disease (AD) using large-scale administrative health data. From the Korean National Health Insurance Service database between 2002 and 2010, we obtained de-identified health data in elders above 65 years ( N = 40,736) containing 4,894 unique clinical features including ICD-10 codes, medication codes, laboratory values, history of personal and family illness and socio-demographics. To define incident AD we considered two operational definitions: "definite AD" with diagnostic codes and dementia medication ( n = 614) and "probable AD" with only diagnosis ( n = 2026). We trained and validated random forest, support vector machine and logistic regression to predict incident AD in 1, 2, 3, and 4 subsequent years. For predicting future incidence of AD in balanced samples (bootstrapping), the machine learning models showed reasonable performance in 1-year prediction with AUC of 0.775 and 0.759, based on "definite AD" and "probable AD" outcomes, respectively; in 2-year, 0.730 and 0.693; in 3-year, 0.677 and 0.644; in 4-year, 0.725 and 0.683. The results were similar when the entire (unbalanced) samples were used. Important clinical features selected in logistic regression included hemoglobin level, age and urine protein level. This study may shed a light on the utility of the data-driven machine learning model based on large-scale administrative health data in AD risk prediction, which may enable better selection of individuals at risk for AD in clinical trials or early detection in clinical settings.

Cite this study APA, Vancouver, BibTeX · download .bib or .ris

APA 7th

Park, J. H., Cho, H. E., Kim, J. H., Wall, M. M., Stern, Y., Lim, H., et al. (2020). Machine learning prediction of incidence of Alzheimer's disease using large-scale administrative health data. NPJ digital medicine. https://doi.org/10.1038/s41746-020-0256-0

Vancouver

Park JH, Cho HE, Kim JH, Wall MM, Stern Y, Lim H, et al. Machine learning prediction of incidence of Alzheimer's disease using large-scale administrative health data. NPJ digital medicine. 2020. doi:10.1038/s41746-020-0256-0. PMID: 32258428.

BibTeX

@article{park2020-32258428,
  author  = {Park, J. H. and Cho, H. E. and Kim, J. H. and Wall, M. M. and Stern, Y. and Lim, H. and Yoo, S. and Kim, H. S. and others},
  title   = {{Machine learning prediction of incidence of Alzheimer's disease using large-scale administrative health data}},
  journal = {NPJ digital medicine},
  year    = {2020},
  doi     = {10.1038/s41746-020-0256-0},
  pmid    = {32258428},
  url     = {https://doi.org/10.1038/s41746-020-0256-0}
}

Built from the record alzhub holds for this paper, not looked up. Volume, issue and page numbers are not held in the record. The record's author list ends in “et al.”, so not every author is named. Check the citation against the original before you use it.

How to read this study

Design: our record does not state a study design, so check the methods section of the original before weighing the result.
Size: the record lists 40,736 participants. Ask whether that is enough for the question, and who they were.
Effect: the record gives no effect size, so the summary above cannot say how large any difference was.

What the evidence grade means here

This page is graded moderate. alzhub.org assigns the grade automatically from keywords in the abstract, not from reading the paper. Moderate means the abstract mentions a cohort, case-control or longitudinal design. It can add at most 5 of the 100 points in the study score, and it is a label about design, not a verdict on whether the findings are true.

New to reading papers? Six questions to ask of any Alzheimer’s study · how we score.

Key findings

The results were similar when the entire (unbalanced) samples were used.

Why it matters

Using Korean national health-insurance records for 40,736 people over 65, this study tested whether machine learning could predict who would later be diagnosed with Alzheimer's from routine health data, reaching an AUC of 0.775 for one-year prediction and lower values (0.644–0.730) at longer horizons. Hemoglobin level, age and urine protein level were among the features the model leaned on, which signals statistical association, not a cause or a target for treatment. The authors frame it as a step toward picking at-risk people for clinical trials or earlier detection; it is not a diagnostic test and does not test any treatment.

New to this subject? Start with our plain-English guide: Digital health →

Study details

Participants40,736
Evidence strengthModerate

How this study scored 59/100

Published in a peer-reviewed journal. Scored zero for study design, effect reported and randomized / blinded because the record does not state it.

Component	Points	Why
Journal tier	7 / 10	Tier 3 — peer-reviewed journal (NPJ digital medicine)
Journal's citation record	2 / 4	Npj Digital Medicine: journal h-index 167 (OpenAlex, 2026-10-29)
Study design	0 / 12	No study design recorded — scores zero, not guessed (not stated in the record — scored 0)
Strength of evidence	3 / 5	Rated “moderate” in our summary
Peer-review validation	10 / 10	Peer-reviewed. Published as an article in a named peer-reviewed journal.
How recent	2 / 5	2020 — within 10 year(s) of the newest study here
Sample size	4 / 4	40,736 participants
Effect reported	0 / 3	No effect size reported — scores zero (not stated in the record — scored 0)
Studied in people	3 / 3	Studied in people (participant count stated)
Randomized / blinded	0 / 2	No randomization or blinding stated — scores zero (not stated in the record — scored 0)
Times cited	5 / 7	Cited 141 times (OpenAlex, 2026-10-10)
Citations per year	3 / 5	17.6 citations a year over 8 year(s) since 2020
Impact for its field and year	4 / 6	Cited 12.5× the average for papers of its field and year (OpenAlex FWCI)
Cited by studies here	3 / 6	Cited by 10 other studies in this library
Still being cited	3 / 4	Cited 63 times in 2024–2026 (2026 still in progress; OpenAlex)
Senior author's track record	2 / 4	Last-listed author Jiook Cha: h-index 26 (OpenAlex, whole career)
Institution's track record	2 / 3	Columbia University: institution h-index 1,824 (OpenAlex)
Multicenter collaboration	2 / 3	OpenAlex lists 6 institution(s) in 2 countries
Indexed in PubMed	2 / 2	Indexed in PubMed (PMID 32258428)
Free to read	2 / 2	Free to read
Total	59 / 100	—

What each of these 20 components measures, and what this score is not →

Read the original

Full text: free to read — OpenAlex lists this copy as free to read: the link above (checked 6 Feb 1020).

Studies in the library that share something with this one. The line above each title says what.

Studies this one cites

Referenced by this study and also in the alzhub library, so you can read them here.

Later studies that cite this one

10 studies in the alzhub library list this one in their references. Showing the 8 highest-scoring.

Related studies in Digital health tools from nearby years

Also in Digital health tools

178 studies in the library are filed under Digital health tools with no topic tag they share with another; these are this study's neighbours in a fixed tour of them. The Digital health tools hub.

Similar studies

More on Digital health tools

See all studies in Digital health tools →