An exploratory retrospective study of low bone density and perioperative D-dimer and Wells score after unicompartmental knee arthroplasty
Year: 2026
Type: article
Study design: Observational Study
Abstract: The present study aimed to examine the association between bone mineral density (BMD) and perioperative D-dimer levels and Wells scores in patients undergoing unicompartmental knee arthroplasty (UKA), with adjustment for potential confounding factors including age and sex. The study was exploratory and was not designed to establish causality or to evaluate imaging-confirmed thrombotic outcomes. A retrospective cohort study of 97 consecutive UKA procedures was conducted. BMD was categorized as normal, osteopenia, or osteoporosis based on dual-energy X-ray absorptiometry. Serum D-dimer was measured preoperatively and on postoperative day 3 (POD 3), together with the Wells score. The primary analysis employed multivariable Gamma regression with log link, adjusted for BMI and preoperative D-dimer, withan extended model additionally incorporating age and sex. Model robustness was evaluated through non-parametric bootstrap resampling (10,000 iterations), permutation testing (10,000 permutations), and female-only sensitivity analysis. A further sensitivity analysis reclassified bone density using the femoral neck T-score alone, rather than the lowest T-score of the lumbar spine or femoral neck. Alternative outcome analyses included absolute D-dimer change, POD3/preoperative ratio, log-transformed ordinary least squares with heteroscedasticity-consistent standard errors, and quantile regression. Propensity score inverse probability of treatment weighting (IPTW) was performed with standardized mean difference balance diagnostics. Wells score was analysed using the Kruskal-Wallis test, pairwise Mann-Whitney U tests with Holm-Bonferroni correction, dichotomization at Wells score of 0 versus negative 1, and multivariable logistic regression. In the absence of duplex ultrasonography data, a composite DVT risk proxy was constructed based on established D-dimer and Wells score thresholds. Among 97 cases, 75 (77.3%) had low bone density. The primary Gamma regression showed a significant association between low bone density and POD 3 D-dimer (adjusted mean ratio 1.39, 95% CI 1.02 to 1.91, p = 0.040). The extended model incorporating age and sex gave a similar estimate (adjusted mean ratio 1.40, 95% CI 1.01 to 1.93, p = 0.043), as did bootstrap resampling (95% CI 1.10 to 1.76, bootstrap p = 0.004) and IPTW weighting (an adjusted increase of 380.7 ng/mL, 95% CI 136.9 to 624.5, p = 0.002). The Kruskal-Wallis test indicated significant differences in Wells score distribution across bone density groups (H = 7.32, p = 0.026). After Holm-Bonferroni correction, a significant difference was observed between the osteoporosis and osteopenia groups (adjusted p = 0.043). The proportion of patients meeting the composite DVT risk proxy was significantly higher in the osteoporosis group (47.4%) compared with osteopenia (25.0%) and normal BMD (13.6%) groups (p = 0.047). A sensitivity analysis reclassifying bone density using the femoral neck T-score alone attenuated the primary association to non-significance (AMR = 1.17, 95% CI 0.89 to 1.53, p = 0.268). Median quantile regression, the Wells score logistic regression and the adjusted composite proxy analysis did not reach statistical significance. Low bone density, defined by the lowest T-score of the lumbar spine or femoral neck, was associated with higher postoperative D-dimer levels and higher Wells scores following unicompartmental knee arthroplasty. That association disappeared once bone density was defined by the femoral neck alone, and several of the supporting analyses fell short of significance. The study was exploratory and its endpoints were surrogate markers, so these results are best read as a hypothesis for testing; prospective validation with imaging-confirmed endpoints is needed before they can inform clinical practice.
Source: Scientific Reports
Authors: Franky Hartono, Adriel Wiemputra Wangsa
Language: English
Subfield: Surgery
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7224167142