Previously submitted to: Journal of Medical Internet Research (no longer under consideration since Nov 24, 2023)
Date Submitted: Mar 5, 2023
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
The value of vascular signal detected by photoplethysmography on predicting hyperglycemia：A cross-sectional research in Chinese cohort
ABSTRACT
Background:
There are 120 million people with diabetes in China, but the awareness rate of diabetes is only 43.3%. The diagnosis of diabetes often requires invasive examination for diagnosis.
Objective:
The aim of this study was to construct a non-invasive hyperglycemia prediction system using a photoplethysmography (PPG)-based smartwatch to collect PPG signals.
Methods:
We conducted a cross-sectional study in a Chinese cohort that comprised 310 Chinese participants enrolled between March 2021 and November 2022. The cohort was divided into training group (n=132, including 28 participants with normal glucose tolerance, 27 participants with pre-diabetes, 77 participants with diabetes) and test group (n=178, including 44 participants with normal glucose tolerance, 42 participants with pre-diabetes, 92 participants with diabetes). PPG signals were collected for 14 days using a PPG-based smartwatch. The PPG-based hyperglycemia prediction system was constructed by using XGBoost machine learning method in the training group, and verified in the test group.
Results:
The PPG-based hyperglycemia prediction system had a specificity of 90.9% and a sensitivity of 83.6% for the diagnosis of hyperglycemia in the test group. In linear regression analysis, the predictive value calculated by the hyperglycemia prediction system was positively correlated with fasting blood glucose (FPG), glycated hemoglobin (HbA1C) and glycated albumin (GA). The area under receiver operating characteristic (AUROC) of the hyperglycemia prediction system was 0.909 (95% CI: 0.849, 0.970). In subgroup analysis, the diagnosis of hyperglycemia from hyperglycemia prediction system was found to be independent of age, gender, HbA1C, hyperlipidemia and hypertension. Specificity was 60% for obesity subgroup and sensitivity was 67.3% in subgroup with previously undiagnosed diabetes.
Conclusions:
Our findings suggest that PPG-based smart watch could provide a readily attainable, non-invasive digital biomarker for predicting hyperglycemia.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.