Przejdź do treściA meta-analysis published on 9 September 2026 suggests that proximal femoral geometry may provide information about hip-fracture risk that is not fully captured by bone mineral density alone. The most consistent signals involved hip axis length and femoral neck width. This is clinically interesting, but it does not justify new diagnostic thresholds in routine DXA practice: heterogeneity was substantial, and prediction intervals for the continuous geometric measures included no effect.
Polish version: Geometria biodra a ryzyko złamania — co wnosi poza densytometrię?
Why is proximal femoral geometry relevant?
Bone mineral density remains a core component of osteoporotic fracture-risk assessment, but it does not describe the entire mechanical behaviour of the proximal femur. Two people with similar BMD may differ in femoral neck length, neck width, neck-shaft angle or hip axis length. These features can influence lever arms and the way force is transmitted during a fall.
The clinically useful question is therefore not simply whether geometry differs between people with and without a fracture, but whether geometric measures retain an association with fracture after BMD has already been taken into account.
Study design
Jokar and colleagues performed a PRISMA-compliant systematic review and meta-analysis of studies assessing conventional proximal femoral geometry in people with low-trauma hip fracture and in controls, as well as studies reporting fracture risk associated with these measures.
The analysis included 47 studies and 136,026 participants: 7,426 people with hip fracture and 128,600 controls. Measures included:
- hip axis length (HAL),
- femoral neck axis length (FNAL),
- femoral neck width (FNW),
- neck-shaft angle (NSA),
- other conventional geometric measures obtained from DXA, radiographs or other imaging.
Random-effects models were used, with subgroup analyses by sex, fracture type and imaging method. Models that adjusted for BMD were examined separately.
Main results
Compared with people without fracture, those with hip fracture had, on average, a hip axis length greater by 0.16 cm, a femoral neck axis length greater by 0.07 cm, a femoral neck width greater by 0.08 cm and a neck-shaft angle greater by 1.74°.
The more important analysis was the one that accounted for BMD. In those models, fracture risk remained significantly associated with:
- hip axis length: OR 1.31; 95% CI 1.20–1.42,
- femoral neck width: OR 1.32; 95% CI 1.10–1.59.
For femoral neck axis length and neck-shaft angle, the association after BMD adjustment was no longer convincing. Associations were more apparent in women and in analyses of femoral neck fractures.
The odds ratios should not be converted directly into an individual’s percentage fracture risk. The included studies differed in how the geometric variables were measured and modelled, so the pooled estimates are not equivalent to a bedside risk calculator.
Clinical interpretation for an orthopaedic surgeon
The findings support a biologically plausible concept: hip-fracture risk reflects not only the material quality of bone but also the geometry of the structure. A longer hip axis may increase the lever arm of force during a fall, while femoral neck width may interact with load distribution and structural properties.
From a clinical perspective, the most interesting result is that HAL and FNW remained associated with fracture in analyses adjusted for BMD. This suggests that at least some structural information is not merely a surrogate for lower bone mineral density.
That does not mean these measurements should now be added routinely to every DXA report or used alone to make treatment decisions. The next step would be to show that adding geometric variables to existing risk models actually improves fracture prediction and leads to better clinical decisions.
Important limitations
- Substantial heterogeneity: populations, imaging techniques, variable definitions and statistical models differed across studies.
- Prediction intervals included no effect: for all continuous geometric measures, indicating that the association in a new population may be considerably smaller than the pooled mean effect.
- Observational evidence: the meta-analysis demonstrates associations rather than causation.
- No single measurement standard: values obtained from DXA and radiography are not necessarily interchangeable, limiting the ability to define a universal clinical threshold.
- No evidence of improved treatment decisions: the study did not test whether adding geometry to BMD or existing risk calculators improves treatment selection or reduces fractures.
What does this change in practice?
For now, the practical message is that BMD is not a complete description of the mechanical risk of hip fracture. Proximal femoral geometry is a reasonable candidate for future refinement of fracture-risk models, particularly when clinical risk and BMD do not tell a fully coherent story.
The study does not show that HAL or FNW should replace BMD, does not establish a universal cutoff for either measure, and does not demonstrate that routine measurement improves outcomes. It also does not justify changing osteoporosis pharmacotherapy solely because of hip geometry. These parameters are best viewed as potentially complementary rather than standalone risk markers.
Reference
- Jokar R, Sharifzadeh Y, Sarabi Javid SR, et al. Proximal femoral geometry and low-trauma hip fracture risk beyond bone mineral density: an updated and expanded systematic review and meta-analysis. BMC Musculoskelet Disord. Published 9 September 2026. DOI: 10.1186/s12891-026-10451-6.
← Research Updates, Literature and Evidence-Based Medicine
Content prepared by: Tomasz Poboży - orthopaedic surgeon.