@article{article_1724418, title={EXTENDED BREASTFEEDING AND POSTMENOPAUSAL OSTEOPOROSIS: A RETROSPECTIVE REAPPRAISAL THROUGH ARTIFICIAL INTELLIGENCE}, journal={Bozok Tıp Dergisi}, volume={15}, pages={302–308}, year={2025}, author={Uzun, Öznur and Yaşar, Evren}, keywords={artificial intelligence, body mass index, breastfeeding, machine learning, menopause, osteoporosis}, abstract={Objective: This study revisits clinical and demographic data on osteoporosis in postmenopausal women, with a particular emphasis on extended breastfeeding (>12 months) as a modifiable risk factor. Furthermore, we conceptualize how artificial intelligence (AI)—if available in the early 2010s—might have altered the trajectory of our research in terms of design, insight, and clinical decision support.

Methods: We conducted a retrospective analysis of 127 osteoporotic and 53 non-osteoporotic postmenopausal women evaluated between 2010 and 2012. Demographics, reproductive history, lifestyle factors, serum osteocalcin levels, dynamic balance scores, and vertebral fracture prevalence were compared. Additionally, we conceptually simulate an AI-assisted reinterpretation of the same dataset, exploring the hypothetical role of machine learning in risk stratification, outcome prediction, and pattern recognition.

Results: Osteoporotic patients were significantly older (p<0.001), had lower BMI (p<0.001), experienced earlier menopause (p=0.035), and were more likely to have breastfed for ≥12 months (p=0.005). Vertebral fractures were present in 22.8% of osteoporotic women. Balance errors and osteocalcin levels were significantly higher in the osteoporotic group. AI-assisted reanalysis would likely have identified nonlinear risk patterns and interaction effects between breastfeeding duration, early menopause, and BMD reduction.

Conclusion: Our reappraisal confirms extended breastfeeding as a significant osteoporosis risk factor in postmenopausal women. Had AI been available in 2010–2012, predictive algorithms and integrative models may have uncovered deeper associations, offered individualized risk predictions, and reshaped preventive care.}, number={3}, publisher={Yozgat Bozok University}, organization={None}