<![CDATA[Breast cancer is a health problem that can begin to be influenced from adolescence. This study aims to analyze the effects of body mass index (BMI), age at menarche, and family history on breast cancer risk among female students at SMA Trensains Muhammadiyah Sragen, while also integrating binary logistic regression results into a risk-scoring system for early screening purposes. The study used a cross-sectional analytic survey design with 122 students who had experienced menarche, selected through purposive and stratified random sampling. The analysis included bivariate tests, binary logistic regression, and ROC curve evaluation. Only BMI showed a significant relationship with breast cancer risk (χ² = 27.4; p < 0.001). The logistic regression model was significant (χ² = 35.2; p < 0.001) with an AUC of 1.00, although affected by perfect separation. The regression coefficients were then converted into a risk-scoring system that preserved the data-separation pattern. The scoring system produced an AUC of 1.00 with a cut-off of 2.5; a total of 118 students (96.7%) were classified as low risk and 4 students (3.3%) as high risk. This integration produces a simple screening tool that can be used for health education and early risk detection among adolescent girls. Keywords: BMI, menarche, breast cancer risk, logistic regression, risk scoring. Abstrak Kanker payudara merupakan masalah kesehatan yang dapat mulai dipengaruhi sejak masa remaja. Penelitian ini bertujuan menganalisis pengaruh indeks massa tubuh (IMT), usia menarche, dan riwayat keluarga terhadap risiko kanker payudara pada siswi SMA Trensains Muhammadiyah Sragen, sekaligus mengintegrasikan hasil regresi logistik biner ke dalam sistem risk scoring untuk keperluan skrining dini. Penelitian menggunakan desain survei analitik potong lintang dengan sampel 122 siswi yang telah mengalami menarche, dipilih melalui purposive dan stratified random sampling. Analisis mencakup uji bivariat, regresi logistik biner, dan kurva ROC. Hanya IMT yang menunjukkan hubungan signifikan dengan risiko kanker payudara (χ² = 27,4; p < 0,001). Model regresi logistik signifikan (χ² = 35,2; p < 0,001) dengan AUC = 1,00, meskipun dipengaruhi perfect separation. Koefisien regresi kemudian dikonversi menjadi sistem risk scoring yang mempertahankan pola pemisahan data. Sistem skor menghasilkan AUC = 1,00 dengan titik potong 2,5; sebanyak 118 siswi (96,7%) tergolong risiko rendah dan 4 siswi (3,3%) risiko tinggi. Integrasi ini menghasilkan instrumen skrining sederhana yang dapat digunakan untuk edukasi kesehatan dan deteksi risiko awal pada remaja putri.]]>
