AI-Based Model for Predicting Type 2 Diabetes Risk Using Lifestyle Factors Among Senior Nursing Students

Condition(s) studied

Investigational drug(s) / intervention(s)

AI-Based Diabetes Risk Prediction Model: A machine learning model developed from students' lifestyle (HPLP-II), anthropometric, and blood pressure data to predict type 2 diabetes risk level (low, moderate, high) and identify the most influential factors. Students will use the model once and rate its usefulness and ease of use. No treatment or lifestyle intervention is delivered

Study summary

This study will develop and test an artificial intelligence (AI) model that estimates the risk of type 2 diabetes among senior (third- and fourth-year) nursing students at the Faculty of Nursing, Alexandria University, Egypt. Type 2 diabetes is common in Egypt, and many cases are not diagnosed. Daily habits such as diet, physical activity, and body weight play a major role in who develops it. Nursing students will become health care providers who teach patients about healthy living, so it is important to find those who may be at risk early.

About 200 students will take part. Each student will complete a questionnaire about personal information and the Health-Promoting Lifestyle Profile II, which measures health habits such as nutrition, physical activity, stress management, and health responsibility. The researcher will measure height, weight, waist size, and blood pressure. Each student's diabetes risk will be scored using the Arabic version of the Australian Type 2 Diabetes Risk Assessment Tool (AUSDRISK), which classifies risk as low, moderate, or high.

These data will be used to build a machine learning model that predicts each student's diabetes risk level from lifestyle and body measurement factors, and shows which factors influenced the prediction most. Students will then try the model and rate how useful and easy to use it is.

No treatment or intervention will be given. The study aims to support early identification of students at risk and to provide a model that nursing education programs and community health nurses could use in preventive health efforts.

Eligibility

Primary outcome measure(s)

• Type 2 Diabetes Risk Level According to AUSDRISK — aseline (single assessment at data collection, up to 1 day per participant)
  Proportion of senior nursing students classified as low risk (total score 4 or less), moderate risk (5 to 10), or high risk (11 or more) for developing type 2 diabetes, using the Arabic version of the Australian Type 2 Diabetes Risk Assessment Tool (AUSDRISK). Total score ranges from 0 to 34; higher scores indicate higher risk.
• Performance of the AI-Based Diabetes Risk Prediction Model — Upon completion of data collection and model development (up to 6 months after study start)
  Accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of the machine learning model in classifying students into low, moderate, or high AUSDRISK risk categories, assessed on a held-out test set.

Trial sites (1)

More Alexandria University trials in Egypt

Other trials for the same condition