An Exploration of Demographic and Psychosocial Indicators of Centenarian Survivorship

Samaneh Rezaeimanesh, Mohammad Fili, Guiping Hu, Alex Bishop

Abstract: This study investigates the psychosocial markers of extreme survivorship among individuals aged 100 years and above, in Oklahoma, integrating gender, education level, race, number of children, and cognitive tests to understand longevity. Data from the Oklahoma 100 Year Life Project on 111 centenarians (43 men, 68 women) were analyzed using machine learning models—Logistic Regression, Random Forest Classifier, and Support Vector Machine—to assess the influence of factors on surpassing 104 years of age. The analysis highlighted significant psychosocial markers, with the Logistic Regression model showing the strongest predictive capability with an AUC of 70%.