Mechanism of Differential Bone Response by Skeletal Site (Spine vs Femoral Neck/Hip)

What the evidence says

Across three independent bodies of evidence — a retrospective human cohort on romosozumab, three meta-analytic comparisons covering abaloparatide and sequential anabolic-then-antiresorptive regimens, and a mouse experiment with parathyroid hormone (PTH) — bone mineral density (BMD) gains at the lumbar spine consistently exceed gains at the femoral neck or total hip during osteoanabolic therapy in humans. In a 12-month romosozumab cohort, median lumbar spine BMD rose 6.7% versus 2.3% at the femoral neck and 2.9% at the total hip. A meta-analysis reported pooled effect sizes of standardized mean difference (SMD) 1.28 at the lumbar spine against 0.70 at the femoral neck and 0.86 at the total hip. The same meta-analysis found abaloparatide increased lumbar spine BMD by 11.29% versus 3.91% at the total hip, and sequential therapy produced an SMD of 1.64 at the spine versus 0.57 at the femoral neck — in both cases the spine effect is roughly two to three times the hip or femoral neck effect. None of these sources establish why the spine responds more; the meta-analysis flags high heterogeneity (I2 96–99%) in the sequential-therapy estimates, meaning the size of the gap varies by study population or method even though its direction is consistent.

A separate line of evidence describes a candidate structural explanation without testing it against the BMD data above. A review of comparative bone anatomy states that blood supply is anatomically and functionally site-specific: arteries feeding the greater trochanter form a vascular network separated from the supply to the femoral neck and shaft. The same review reports that blood flow conditions in bone are site-specific and change with trauma, metabolic demand, and aging, and links this to differences in healing rate — for example, stress fractures in the fifth metatarsal heal slowly because of poor local vascular support. These are anatomical and physiological observations, not measurements tied to the spine-versus-hip BMD gap; no source on this page connects the two directly.

One finding points the opposite way. In mice, intermittent PTH treatment increased hindlimb BMD more than spine BMD, the reverse of the site pattern seen in human osteoanabolic therapy. The authors describe this only qualitatively ("not shown"), so no magnitude is available, and confidence in the magnitude is low.

A weaker, associational thread links spine BMD to vascular stiffness rather than to skeletal site: a decrease in spine BMD was observed alongside increased carotid arterial stiffness, but the source describes this as a "trend," not a significant finding, and confidence in both direction and magnitude is correspondingly limited.

The evidence on this page never measures whether the larger spine BMD response translates into a larger reduction in spine fracture risk relative to hip fracture risk — the Alshamsi cohort explicitly notes this was not assessed.

Where the evidence splits

The mouse PTH finding (hindlimb BMD increases more than spine BMD) disagrees with the human romosozumab, abaloparatide, and sequential-therapy findings (spine BMD increases more than femoral neck or total hip BMD). The claims themselves identify one explanatory difference: the reversed pattern comes from a mouse experiment with a vehicle control, while the spine-dominant pattern comes from human retrospective cohort and meta-analytic data. Species is the only design difference the sources state; they do not say whether drug class (intermittent PTH versus romosozumab or abaloparatide) or anatomical scope (hindlimb, which spans multiple appendicular bones, versus the femoral neck specifically) also contributes, so the split beyond species is unexplained by the sources on this page.

Thresholds and dose-response

No threshold established in the current sources.

Open empirical questions

• Whether the larger spine BMD response to osteoanabolic therapy translates into a larger reduction in spine fracture risk relative to hip fracture risk — not measured in the sources here.
• What accounts for the reversed site pattern in the mouse PTH study (hindlimb > spine) versus the human osteoanabolic studies (spine > femoral neck/hip) beyond species.
• Whether the anatomical vascular separation described for bone sites causally contributes to the measured spine-vs-hip BMD gap in osteoanabolic therapy — no source tests this link directly.
• Whether the spine BMD / arterial stiffness association holds up as a significant finding, since the only source reporting it calls it a trend.