Metal biomarkers and bone mineral density in women undergoing DXA: Associations with cadmium and selenium balance in a hospital-based study.
Abstract
Background
Toxic metals may affect bone health, but biomonitoring evidence from Middle Eastern populations is limited. We examined associations between cadmium (Cd), lead, mercury, arsenic, and selenium (Se) and bone mineral density (BMD) among women undergoing dual-energy X-ray absorptiometry (DXA) in a tertiary-care setting.
Methods
This hospital-based cross-sectional study included 646 women aged ≥40 years referred for DXA at King Faisal Specialist Hospital and Research Centre (2021-2024). Metals were measured in urine and hair using inductively coupled plasma-mass spectrometry (ICP-MS). Primary analyses used multivariable linear regression to evaluate lumbar spine and femoral neck T-scores as continuous outcomes. Secondary analyses evaluated low BMD as a binary outcome using multivariable logistic regression. Models were adjusted for relevant demographic and exposure-related covariates.
Results
In the primary continuous-outcome analyses, urinary Cd was inversely associated with femoral neck T-score in individual models (β = -0.148; P = 0.010) and multi-metal models (β = -0.177; P = 0.006). Urinary Se was positively associated with femoral neck T-score (β = 0.223; P = 0.047). In exploratory analyses, the selenium-cadmium balance (lnSe - lnCd) was positively associated with lumbar spine (β = 0.165; P = 0.030) and femoral neck T-scores (β = 0.184; P = 0.003). Secondary analyses using low BMD as a binary outcome showed no significant associations with individual metals or metal burden indices.
Conclusions
In this hospital-based DXA-referral cohort, urinary Cd was inversely associated and urinary Se positively associated with femoral neck T-score, while a higher Se-to-Cd balance was positively associated with T-scores at both skeletal sites. Direct Cd × Se interaction analyses did not provide consistent evidence of effect modification, and the Se-to-Cd balance findings should therefore be considered exploratory and hypothesis-generating. Binary low-BMD analyses were null. These findings support further investigation of Cd and Se in relation to site-specific BMD in more representative populations.