Predictive Model for Hemoglobin Levels Using Multiple Linear Regression

Open

Dien G.A. Nursal, Muhammad D.A. Winata, Nohan A. Romadlona, Noura Rizki, Sabillah Nasitoh, Muhammad S. Baretta

2025 Journal of Health Research Vol. 39 Issue 3 Article Cited by 0 Quartile

Abstract

Background: Anemia is a condition in which the number of red blood cells or hemoglobin (Hb) levels fall below normal, limiting the ability of blood to transport oxygen effectively. This study aimed to develop a predictive model for Hb levels by identifying significant risk factors. Method: This observational analytic study was conducted among 103 randomly selected students of Bung Hatta University in Padang, Indonesia. Data were collected with a questionnaire and a hemoglobin test between April to May 2024. Pearson's correlation analysis and multiple linear regression were performed in this study. Results: Bivariate analysis showed significant correlations between Hb levels and factors such as age (p ¼ 0.046; r ¼ 0.197), body weight (p ¼ 0.031; r ¼ 0.213), body mass index (BMI) (p ¼ 0.029; r ¼ 0.216), coffee consumption (p ¼ 0.001; r ¼ ¡0.590), and tea consumption (p ¼ 0.028; r ¼ ¡0.216). However, the final multivariate prediction model showed that Hb levels were most influenced by coffee and tea consumption, resulting in the following equation: Hb levels ¼ 2.001e0.145 (Coffee Consumption) ¡0.054 (Tea Consumption). Conclusion: Coffee consumption was identified as the most dominant factor influencing Hb levels. Therefore, guidelines for safe coffee consumption among adolescents were recommended to help prevent anemia. © 2025 College of Public Health Sciences, Chulalongkorn University. All rights reserved.

Affiliations

Department of Epidemiology, Faculty of Public Health, Andalas University, West Sumatera, Padang, 25163, Indonesia; Reproductive Health Field, Faculty of Public Health, Andalas University, West Sumatera, Padang, 25163, Indonesia; Department of Public Health, State University of Malang, East Java, Malang, 65145, Indonesia; Department of Medicine, Faculty of Medicine, Baiturrahmah University, West Sumatera, Padang, Indonesia; Department of Medicine, Faculty of Medicine, Andalas University, West Sumatera, Padang, Indonesia