Natural Language Processing (NLP) Of Health Promotion Narratives to Assess Weight Literacy and Predict Weight Status

• Author(s): Olayemi Fisayo Grace; Bukola Cecilia Bello; Olawale Ignatus Oni
• Paper ID: 1718569
• Page: 153-159
• Published Date: 03-06-2026
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 9 Issue 12 June-2026

Abstract

Weight literacy, the ability to obtain, process, and understand basic weight-related health information, is a critical determinant of obesity prevention and management. Traditional survey-based assessments of weight literacy are resource-intensive, prone to recall bias, and not scalable. This study applies natural language processing (NLP) to analyze free-text health-promotion narratives (e.g., responses to open-ended questions about weight management, diet, and physical activity) to automatically assess weight literacy levels and predict current weight status (normal weight, overweight, obese). A systematic review of 38 studies (2018–2025) examined NLP applications in health literacy assessment, sentiment analysis of health narratives, and text-based prediction of clinical outcomes. Additionally, a proof-of-concept analysis was conducted on a corpus of 2,450 narrative responses from adults in southwestern Nigeria. NLP pipelines using term frequency-inverse document frequency (TF-IDF) vectorization, sentiment analysis, readability indices (Flesch-Kincaid, SMOG), and transformer-based models (BERT, RoBERTa) were evaluated. Results show that transformer models fine-tuned on weight-related narratives achieve the highest accuracy (85–91%) in classifying weight literacy as adequate, marginal, or inadequate. Key linguistic markers of low weight literacy include: use of absolute terms (“never,” “always”), lack of conditional language (“if,” “depending”), confusion about portion sizes, and fatalistic statements (“my weight is genetic, I can’t change it”). The NLP-derived weight literacy score correlates significantly with measured BMI (r = 0.62, p < 0.001) and predicts obesity status with an AUC of 0.83. The study concludes that NLP for health-promotion narratives offers a scalable, automated approach to assessing weight literacy and predicting weight status, thereby enabling population-level surveillance and personalized weight-management interventions.

Keywords

Natural Language Processing, Weight Literacy, Health Literacy, Obesity Prediction, Machine Learning, BERT, Sentiment Analysis, Health Narratives.

Citations

IRE Journals:
Olayemi Fisayo Grace, Bukola Cecilia Bello, Olawale Ignatus Oni "Natural Language Processing (NLP) Of Health Promotion Narratives to Assess Weight Literacy and Predict Weight Status" Iconic Research And Engineering Journals Volume 9 Issue 12 2026 Page 153-159 https://doi.org/10.64388/IREV9I12-1718569

IEEE:
Olayemi Fisayo Grace, Bukola Cecilia Bello, Olawale Ignatus Oni "Natural Language Processing (NLP) Of Health Promotion Narratives to Assess Weight Literacy and Predict Weight Status" Iconic Research And Engineering Journals, 9(12) https://doi.org/10.64388/IREV9I12-1718569