Abstract 5963: Pink to purple: Beyond traditional risk factors of female lung cancer

Abstract Background: The pink ribbon for breast cancer awareness is being overshadowed by purple, as lung cancer has become the leading cause of cancer deaths in women. While overall lung cancer rates are declining, younger women in high-income countries are being diagnosed more often than men, suggesting unique risk factors for women beyond smoking. The aim of this study was to identify the beyond-traditional risk factors for female lung cancer in a diverse population within the United States. Methods: We conducted a nested age-matched case-control study within the All of Us Research cohort, analyzing 4, 761 adult female participants (761 cases, 4, 000 controls). Comprehensive phenotypic data, including demographic characteristics, reproductive history, and lifestyle behaviors, were extracted from electronic health records. Statistical analysis employed chi-square tests, multivariable logistic regression, and an Extended Gradient Boosting Classifier (XGBC) for risk prediction. Results: Significant associations emerged between lung cancer risk and multiple factors. Age strongly predicts increased risk in females, with odds ratios rising from 5.14 (95% CI 3.22-8.19, p<0.001) in the 46-60-year group to 12.51 (95% CI 8.05-19.44, p<0.001) in the 61-95-year group. Smoking was a significant risk factor (adjusted OR: 2.54; 95% CI, 2.13-3.02; p<0.001), whereas Hispanic ethnicity (adjusted OR: 0.49; 95% CI, 0.32-0.73; p<0.001) compared to non-Hispanic white and high deprivation index (adjusted OR: 0.58; 95% CI, 0.47-0.71; p<0.001) had a protective role. Non-natural menopause consistently increased risk (adjusted OR: 1.36; 95% CI, 1.05-1.77; p<0.01). The XGBoost (XGBC) model achieved 84% accuracy in risk classification, identifying smoking, age, menopausal history, and deprivation index as primary predictive features compared to other machine learning models. The model revealed distinct risk patterns based on smoking status, age, and reproductive factors, with notable interactions between menopausal status among smokers. Conclusions: This comprehensive analysis highlights the complex interplay between reproductive, sociodemographic, and lifestyle factors in female lung cancer risk. Our findings emphasize the need for personalized risk assessment strategies considering reproductive history alongside traditional risk factors. The successful implementation of machine learning approaches demonstrates the potential for developing more accurate risk prediction models for female lung cancer. Citation Format: Muhammad Rafiqul Islam, Mohammad Hasan Shahriar, Muhammad Ashique Chowdhury, Siara Tasmin, Andrew Craver, Brisa Aschebrook-Kilfoy, Sreenivas Konda, Habibul Ahsan. Pink to purple: Beyond traditional risk factors of female lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5963.