The Prediction of Body Mass Index from Negative Affectivity through Machine Learning: A Confirmatory Study.

Area of Science:

• Psychology
• Data Science
• Public Health

Background:

• Body Mass Index (BMI) is a key health indicator.
• The relationship between psychological variables and BMI is complex and warrants further investigation.
• Machine learning (ML) offers novel approaches to explore these relationships.

Purpose of the Study:

• To investigate the predictive power of affect-related psychological variables on Body Mass Index (BMI).
• To develop and apply ML algorithms for forecasting BMI values and status (normal, overweight, obese).
• To compare the efficacy of negative versus positive psychological variables in BMI prediction.

Main Methods:

• Utilized machine learning algorithms, including gradient boosting and random forest.
• Employed psychological variables as predictors for 221 subjects.
• Predicted both continuous BMI values and categorical BMI status.

Main Results:

• ML models accurately predicted BMI values with a mean absolute error of 5.27-5.50.
• BMI status was predicted with an accuracy exceeding 80% (F1-score).
• Negative psychological variables, such as depression, demonstrated higher predictive efficacy compared to positive ones.

Conclusions:

• Affect-related psychological variables are significant predictors of BMI.
• Machine learning provides a powerful tool for understanding the interplay between mental health and weight.
• Interventions targeting negative psychological states may contribute to weight management strategies.