Fusion of the Manubriosternal and Xiphisternal Junctions on Computed Tomography: A Probabilistic and Threshold-Based Approach to Age Assessment

Background/Objectives: Sternal fusion has been studied as an adult age indicator, but findings remain inconsistent, partly because variable fusion processes have been interpreted within deterministic frameworks. We reassessed manubriosternal and xiphisternal fusion on computed tomography (CT) using a probabilistic, threshold-based approach. Methods: Thoracic CT examinations of 512 individuals aged 30–65 years were retrospectively analysed. Manubriosternal fusion was classified on a three-stage ordinal scale and xiphisternal fusion dichotomously. Fusion was examined across five age bands (30–40, 40–50, 50–55, 55–60 and 60–65 years) and for age thresholds of 50, 55 and 60 years using sensitivity, specificity and likelihood ratios (LRs). Results: Manubriosternal fusion overlapped extensively across age bands, with a significant but weak association with age (Kruskal–Wallis p < 0.001; ε2 = 0.060). Xiphisternal fusion increased from 37.6% (30–40 years) to 77.3% (55–60 years) and plateaued at 75.7% (60–65 years; Mann–Whitney p < 0.001), leaving one quarter of 60–65-year-olds non-fused. Xiphisternal fusion increased the odds of being ≥50, ≥55 and ≥60 years old (odds ratios 3.15, 3.14 and 2.57), but its evidential value was modest (positive LRs 1.38–1.55; negative LRs 0.48–0.54). Adding manubriosternal stage improved discrimination for the ≥55-year threshold in an exploratory model (area under the curve 0.628 to 0.676; p = 0.002). Conclusions: CT-based sternal fusion is unsuitable for deterministic or standalone age estimation. Xiphisternal fusion provides only modest, supplementary information that may be combined with other indicators when CT is already available for clinical or postmortem purposes; it does not justify CT acquisition for age assessment.