Dual-energy X-ray absorptiometry has dominated skeletal assessment for three decades, yet emerging evidence reveals a sobering reality: more than half of fragility fractures occur in individuals whose DEXA scores fall outside the osteoporotic range. The metric we've enshrined as the gold standard captures only one dimension of a fundamentally multidimensional tissue.

Bone is not a static mineral deposit. It's a dynamic composite material whose strength derives from microarchitecture, collagen cross-linking quality, cortical thickness, and trabecular connectivity—properties that areal bone mineral density measurements aggregate into a single, lossy number. Sophisticated prevention demands we move beyond this reductionism.

The contemporary toolkit now includes trabecular bone score analysis, high-resolution peripheral quantitative computed tomography, advanced biochemical markers of bone turnover, and integrated fracture risk algorithms. Layered onto this assessment infrastructure are anabolic interventions—mechanotransductive loading protocols, targeted nutritional substrates, and pharmacologic agents—that don't merely slow resorption but actively rebuild skeletal tissue. For those pursuing health span optimization, skeletal integrity is non-negotiable infrastructure: it determines whether the next four decades involve independent mobility or progressive disability.

Areal bone mineral density, the output of standard DEXA, measures grams of mineral per square centimeter of projected bone area. This two-dimensional projection conflates cortical and trabecular compartments, ignores three-dimensional architecture, and cannot distinguish between dense but brittle bone and lower-density but well-organized bone. Two patients with identical T-scores can have substantially different fracture risks.

The trabecular bone score, derived from textural analysis of lumbar spine DEXA images, provides a non-invasive proxy for trabecular microarchitecture. By analyzing pixel gray-level variations, TBS estimates whether trabeculae are abundant and well-connected or sparse and disconnected. Studies demonstrate that TBS predicts fracture risk independently of BMD, and the combination outperforms either metric alone, particularly in patients with diabetes, glucocorticoid exposure, or hyperparathyroidism.

High-resolution peripheral quantitative computed tomography pushes resolution further, providing volumetric measurements at the distal radius and tibia with voxel sizes approaching 60 micrometers. HR-pQCT separately quantifies cortical thickness, cortical porosity, trabecular number, trabecular thickness, and trabecular separation—the actual structural determinants of mechanical competence. Finite element analysis of these images can estimate whole-bone stiffness and failure load.

Biochemical markers complement structural assessment. Serum CTX (C-terminal telopeptide) reflects osteoclastic resorption activity, while P1NP (procollagen type 1 N-terminal propeptide) indicates osteoblastic formation. Their ratio reveals whether the remodeling balance favors accrual or loss, and their absolute levels predict imminent fracture risk independently of BMD.

Integration matters more than any single modality. A comprehensive skeletal phenotype combines DEXA-derived BMD, TBS-derived microarchitectural quality, turnover markers reflecting current biological activity, and clinical risk factors. This multidimensional characterization replaces a single number with an actionable biological portrait.

Takeaway

A bone is not its density—it's an engineered structure whose failure modes depend on geometry, connectivity, and material properties that no single metric captures. Reduce a complex system to one number, and you'll be surprised by what that number missed.

The Fracture Risk Assessment Tool, developed by the WHO Collaborating Centre, integrates clinical risk factors to estimate ten-year probability of major osteoporotic fracture and hip fracture. Inputs include age, sex, BMI, prior fragility fracture, parental hip fracture history, current smoking, glucocorticoid use, rheumatoid arthritis, secondary osteoporosis, and alcohol consumption. Femoral neck BMD can be incorporated when available, but FRAX functions even without densitometry.

The tool's strength lies in synthesizing heterogeneous risk factors into a probability estimate that maps onto treatment thresholds. The National Osteoporosis Foundation recommends pharmacologic intervention when ten-year hip fracture probability exceeds 3% or major osteoporotic fracture probability exceeds 20%. These thresholds derive from cost-effectiveness analyses and represent inflection points where intervention benefits exceed risks and costs.

FRAX has well-documented limitations. It uses femoral neck BMD only, ignoring lumbar spine measurements that may differ substantially. It treats risk factors dichotomously, missing dose-response relationships—a patient on 2.5 mg prednisone is treated identically to one on 20 mg. Falls risk, a dominant determinant of fracture in the elderly, is absent. Recent fractures, which dramatically elevate imminent risk, receive no special weighting.

Adjustments improve precision. The TBS-adjusted FRAX incorporates microarchitectural quality, modifying probabilities upward in patients with degraded trabecular structure despite preserved BMD. Hip axis length, femoral geometry, and lumbar spine BMD discordance from femoral neck values provide additional refinement. Some clinicians apply multipliers for high-dose glucocorticoids or recent fracture occurrence.

Treatment decisions ultimately integrate FRAX output with patient-specific factors: life expectancy, fall risk, competing morbidities, prior medication tolerance, and patient preferences regarding pharmacologic intervention. The probability is a starting point for shared decision-making, not an algorithmic mandate.

Takeaway

Risk stratification tools are most useful when their limitations are explicit. A probability estimate that ignores fall risk, recent fractures, and dose-response relationships is a beginning, not a verdict.

Antiresorptive agents—bisphosphonates and denosumab—reduce fracture risk by suppressing osteoclastic activity, but they preserve existing architecture rather than rebuilding lost structure. True anabolic strategies, both pharmacologic and behavioral, increase bone mass and improve microarchitecture. For individuals with established bone loss, this distinction is consequential.

Resistance training delivers the most robust non-pharmacologic anabolic stimulus. Mechanotransduction—the conversion of mechanical strain into cellular signaling—activates osteocytes via the Wnt/β-catenin pathway, suppressing sclerostin and stimulating osteoblastogenesis. The stimulus must be substantial: progressive resistance training at 80-85% of one-repetition maximum, performed twice weekly, produces meaningful BMD gains at the lumbar spine and femoral neck. The LIFTMOR trial demonstrated that high-intensity training was both safe and effective in postmenopausal women with low bone mass.

Impact loading complements resistance training through different mechanotransductive pathways. Strain rate, not just strain magnitude, drives osteogenic response. Hopping, jumping, and rapid directional changes generate the high-rate ground reaction forces that activate osteocytic mechanosensing. Even brief daily protocols—50 multidirectional hops—produce measurable femoral neck BMD improvements over twelve months.

Nutritional substrates enable but don't drive accrual. Protein intake of 1.2-1.6 g/kg supports collagen matrix synthesis and IGF-1 mediated osteoblast activity. Vitamin D sufficiency (25-OH vitamin D above 40 ng/mL) optimizes intestinal calcium absorption and direct osteoblastic effects. Vitamin K2 (MK-7) carboxylates osteocalcin, facilitating mineral incorporation. Magnesium and boron serve as enzymatic cofactors in bone matrix metabolism.

Pharmacologic anabolics—teriparatide, abaloparatide, and romosozumab—reserve their substantial bone-building effects for high-risk patients. Romosozumab, a sclerostin antibody, uniquely combines anabolic and antiresorptive actions, producing BMD gains exceeding 13% at the lumbar spine over twelve months. Sequential therapy, beginning with anabolic agents and consolidating with antiresorptives, optimizes long-term skeletal outcomes.

Takeaway

Preservation and construction are different projects. If you've already lost structural capital, maintenance protocols won't restore it—you need stimuli substantial enough to rebuild what's gone.

Comprehensive skeletal assessment has evolved beyond the single-metric paradigm that dominated late twentieth-century practice. DEXA-derived BMD remains foundational but insufficient; TBS, HR-pQCT, biochemical turnover markers, and integrated risk algorithms together produce a multidimensional skeletal phenotype that supports precise intervention.

The intervention landscape has similarly matured. Mechanotransductive loading protocols, targeted nutritional optimization, and sequential pharmacologic strategies offer genuine anabolic capacity—the ability to rebuild rather than merely defend. For health span optimization, skeletal integrity warrants the same systematic attention as cardiovascular and metabolic health.

The actionable protocol: assess comprehensively in the fourth decade, intervene before degradation accelerates, and treat bone as the dynamic, modifiable tissue it is. Skeletal collapse in the eighth decade is not destiny—it's the cumulative consequence of decisions made forty years earlier.