Yuswanto Setyawan

Smart Screening Technology for Diabetes Risk: FFQ and FINDRISC Integration in a Digital Platform

Introduction

Smart screening technology for diabetes risk: ffq and findrisc integration in a digital platform. Discover a smart digital platform integrating FFQ & FINDRISC for early diabetes risk screening in Indonesian university students. Highlights dietary patterns for crucial early intervention.

Abstract

Diabetes mellitus (DM) is a growing metabolic and autoimmune-related disease whose early onset is increasingly observed among young adults, including the university students in Indonesia. The existing screening models are either costly, invasive, or fail to integrate lifestyle data, leaving a gap for practical yet scalable solutions in this population. This study introduces a smart screening technology that combines the Food Frequency Questionnaire (FFQ) and the Finnish Diabetes Risk Score (FINDRISC) within a digital platform to capture both dietary patterns and individual risk factors. A cross-sectional design was applied to 110 undergraduates, chosen to reflect young adults most vulnerable to lifestyle-related DM risks. Data were collected entirely online to ensure feasibility and low-cost scalability in campus and public health programs. Multiple linear regression revealed that both individual factors (age, gender, BMI, physical activity, family history) and dietary patterns were significant predictors of DM risk (β = 0.312; β = 0.389; p < 0.001), explaining 37.4% of the variance. Compared to prior studies that relied solely on clinical or genetic markers, this integration highlights the added predictive value of dietary data in digital risk screening. With 70.9% of respondents at moderate and 25.5% at high risk, the findings underscore the urgent need for early intervention among Indonesian students. The proposed model offers practical applications through university health centers, mobile apps for student lifestyle monitoring, and peer-based preventive education. Future work should extend to biomarker validation and adaptive algorithms to enhance predictive accuracy and applicability across diverse populations.

Review

The paper presents a timely and relevant study on a smart screening technology for diabetes risk among young adults, specifically Indonesian university students. Addressing the critical need for accessible and non-invasive screening methods, the authors propose a digital platform integrating the Food Frequency Questionnaire (FFQ) and the Finnish Diabetes Risk Score (FINDRISC). A key strength lies in its innovative approach to combine dietary patterns with established individual risk factors, filling a crucial gap left by models that are either costly, invasive, or neglect lifestyle data. The online data collection methodology in a cross-sectional design demonstrates excellent practical foresight, ensuring feasibility and low-cost scalability for deployment in campus and public health programs. The findings that both individual factors and dietary patterns are significant predictors of DM risk, explaining a substantial 37.4% of variance, underscore the added predictive value of dietary data, distinguishing this model from those relying solely on clinical or genetic markers. While promising, the study exhibits several limitations that warrant consideration. The sample size of 110 undergraduates, though selected to reflect vulnerability, is relatively small for a "smart screening technology" aiming for broad applicability and scalability across diverse populations. The cross-sectional design, inherent in risk assessment studies, precludes the establishment of causal relationships or the tracking of risk progression over time, which would be crucial for evaluating the impact of interventions. Furthermore, the reliance on self-reported data for both FFQ and FINDRISC components introduces potential for recall and social desirability biases, which could affect the accuracy and reliability of the risk scores. Most critically, the absence of biomarker validation within this study means that the "added predictive value" of the integrated model is not directly benchmarked against a gold standard for diabetes diagnosis or risk, leaving its true clinical accuracy and utility somewhat unquantified in the current implementation. Despite these limitations, the study offers significant practical implications, particularly given the alarming finding that a high percentage of respondents are at moderate (70.9%) or high (25.5%) risk. The proposed model presents a valuable tool for early intervention through university health centers, mobile apps, and peer-based education, addressing an urgent public health need among Indonesian students. For future work, it is imperative to expand the sample size to enhance generalizability and conduct longitudinal studies to observe the efficacy of interventions and the predictive power of the tool over time. Most importantly, as acknowledged by the authors, rigorous validation against clinical biomarkers (e.g., HbA1c, fasting glucose) is essential to establish the model's diagnostic accuracy and clinical utility. Further development involving adaptive algorithms, as suggested, would indeed enhance its "smart" capabilities and applicability across more diverse populations, transitioning it from a promising concept to a robust and widely deployable health technology.

Full Text

Comments

You need to be logged in to post a comment.