BackgroundBirth weight is considered not only to undermine future growth, but also to induce lifelong diseases; the aim of this study is to explore the relationship between birth weight and adult bone mass.MethodsWe performed multivariable regression analyses to assess the association of birth weight with bone parameters measured by dual-energy X-ray absorptiometry (DXA) and by quantitative ultrasound (QUS), independently. We also implemented a systemic Mendelian randomization (MR) analysis to explore the causal association between them with both fetal-specific and maternal-specific instrumental variables.ResultsIn the observational analyses, we found that higher birth weight could increase the adult bone area (lumbar spine, β-coefficient= 0.17, P < 2.00 × 10−16; lateral spine, β-coefficient = 0.02, P = 0.04), decrease bone mineral content-adjusted bone area (BMCadjArea) (lumbar spine, β-coefficient= − 0.01, P = 2.27 × 10−14; lateral spine, β-coefficient = − 0.05, P = 0.001), and decrease adult bone mineral density (BMD) (lumbar spine, β-coefficient = − 0.04, P = 0.007; lateral spine; β-coefficient = − 0.03, P = 0.02; heel, β-coefficient = − 0.06, P < 2.00 × 10−16), and we observed that the effect of birth weight on bone size was larger than that on BMC. In MR analyses, the higher fetal-specific genetically determined birth weight was identified to be associated with higher bone area (lumbar spine; β-coefficient = 0.15, P = 1.26 × 10−6, total hip, β-coefficient = 0.15, P = 0.005; intertrochanteric area, β-coefficient = 0.13, P = 0.0009; trochanter area, β-coefficient = 0.11, P = 0.03) but lower BMD (lumbar spine, β-coefficient = − 0.10, P = 0.01; lateral spine, β-coefficient = − 0.12, P = 0.0003, and heel β-coefficient = − 0.11, P = 3.33 × 10−13). In addition, we found that the higher maternal-specific genetically determined offspring birth weight was associated with lower offspring adult heel BMD (β-coefficient = − 0.001, P = 0.04).ConclusionsThe observational analyses suggested that higher birth weight was associated with the increased adult bone area but decreased BMD. By leveraging the genetic instrumental variables with maternal- and fetal-specific effects on birth weight, the observed relationship could be reflected by both the direct fetal and indirect maternal genetic effects.