摘要Objective:This study aimed to optimize dual-energy X-ray absorp-tiometry(DXA/BMD)quantitative detection technology using a generalized additive model(GAM),so as to establish an improved joint evaluation framework for body fat distribution and bone metabolism in diabetic osteoporosis and refine the quantitative assessment model of visceral fat gradients in type 2 diabetic patients.Methods:A total of 256 patients with type 2 diabetes mellitus(T2DM)were enrolled in this retro-spective study.A calibrated Hologic Discovery Wi dual-energy X-ray absorptiometer was applied to synchronously acquire lumbar spine(L1-L4)bone mineral density(BMD)and whole-body fat distribution parameters.Quantitative differences in vis-ceral adipose tissue(VAT)area,waist-to-hip ratio,and android/gynoid fat partitioning were compared between patients with and without osteoporosis.GAM was performed to clarify the interactive effects of age and sex on the fat-bone metabolism associa-tion.On the basis of multi-dimensional imaging features and machine learning strate-gies,a visceral fat-BMD joint scoring system(VF-BMD Score)was constructed for diabetic osteoporosis prediction.Results:VAT showed a significant negative correla-tion with L1-L4 BMD in T2DM patients(r=-0.488,P<0.01).Sex-stratified analysis revealed that the inverse association between VAT and BMD was more pronounced in females than in males(r=-0.398 vs.-0.686,P<0.01).Age-stratified analysis further indicated that females aged≥60 years exhibited significantly lower VAT and BMD compared to those<60 years old(r=-0.686,P<0.01),whereas no significant differ-ences were observed between the corresponding male subgroups.The VF-BMD scor-ing system comprised four categories.For predicting diabetic osteoporosis,the VF-BMD score yielded an area under the curve(AUC)of 0.64,with a sensitivity of 46%and a specificity of 82%.Conclusion:DXA enables the synchronous quantification of visceral fat gradients and bone metabolic parameters.The GAM-based VF-BMD scoring system developed herein enhances the specificity of early screening for dia-betic osteoporosis.