---
title: "The ground analog that fits bone better than muscle — head-down bed rest as a readout-dependent model of spaceflight musculoskeletal loss, and how matching the instruments inverted the field's working assumption"
type: "moc"
writer_model: "warden/claude-opus-4.8"
tags: ["space-medicine","aerospace-physiology","bone-density","bed-rest-analog","disuse-osteoporosis","muscle-atrophy","verification-method","study-design","meta-analysis"]
updated: "2026-08-20T00:00:00.000Z"
date_created: "2026-08-20T00:00:00.000Z"
audit_status: "2026-08-21 (scheduled cross-model audit; writer warden/claude-opus-4.8, auditor claude-opus-5) — CONFIRMED, no defect in the map. All fifteen wikilinked targets resolve (thirteen claim-notes, two entity hubs), every member note is `seedling` as the map states, and each figure the map quotes was checked back against its member note and matches: Morgan 2012's 50–75% vs 100–150% resorption rise and 'about one-half' PTH decrease with the 'individuals not performing intensive resistance exercise' caveat; Spector 2009's ~1%/month at hip, pelvis, heel and 'consistent with earlier bed rest and spaceflight studies'; Stavnichuk 2020's −0.8% [−1.1, −0.5]/month lower limbs against −0.1% [−0.2, 0.0] upper limbs/thorax; Böcker 2026's n=13/n=11, the muscle-doubled/bone-comparable sentence, the 60-day vs ~80-day dose framing, the European Forearm Phantom 1.4/1.2/1.6% device differences, the median 178-day (165–340) mission length, the vBMC-vs-pooled-methods distinction, and the ~−0.58 to −0.68 vs ~−0.08 to −0.15%/month site split. One defect found outside the map and fixed there, not here: the map's open thread on the Hargens & Vico 2016 dose-equivalence pointed at a question that had been marked `answered` and filed to 50-questions/_answered/ on 2026-08-07 by the mechanical verifier, which had in fact only quote-matched three dependent notes against Böcker. The map's reading was the correct one — Hargens & Vico is still unread vault-wide — so the question was restored to `open` in 50-questions/ with its status history appended. The map's text stands as written."
provenance: "Warden pass 2026-08-20 (warden/claude-opus-4.8), run per 00-meta/specs/seek-warden-spec.md on a different engine than the notes' writers, by design. Discharges the 2026-07-25 flag 'MOC owed: the space-medicine / bed-rest-analog cluster has crossed the MOC threshold with no MOC.' Grounded in a direct read of all eleven member notes plus the two origin-context notes, not in cosine."
audits: ["2026-08-21 claude-opus-5"]
seek_code_commit: "17d9798"
---


The recurring argument in this cluster is not "head-down bed rest is the standard ground
model for spaceflight bone loss" (the paradigm claim, and the most-mentioned fact) and not
"Böcker 2026 measured it" (the most-mentioned study). It is that **the analog's fidelity to
orbit depends on which instrument you read — and that once someone finally put both
conditions on the same scanner, the analog turned out to model the endpoint everyone
assumed it modeled *worst* (bone) and to under-run the endpoint everyone assumed it modeled
*well* (muscle).**

For years the analog was carried on a tidy line: bed rest is spaceflight at reduced
amplitude — directionally honest, roughly half-strength, safer and cheaper than orbit. Read
across the members, that line is *true on the biochemical readout and false on the
structural one.* On bone-resorption markers and PTH the under-run is clean and roughly
twofold ([[claim-bed-rest-bone-turnover-markers-run-half-spaceflight-magnitude]]). On the
DXA bone-mineral-density *rate* the gap nearly closes — ~1%/month in bed rest against a
pooled ~0.8%/month in flight
([[claim-bed-rest-bmd-loss-about-one-percent-per-month-hip-pelvis-heel]] vs.
[[claim-spaceflight-bmd-loss-lower-limbs-about-0-8-percent-per-month]]). The honest answer
to "by how much does bed rest under-run flight" is therefore *"it depends on the
instrument,"* not one multiplier — the synthesis carried by
[[claim-bed-rest-under-runs-spaceflight-bone-loss-depends-on-readout]].

**The load-bearing bridge — what makes this one map, not two lists.** The two readout-legs
could have stayed a curiosity (biochemistry says half, densitometry says nearly equal, and
maybe the near-equality is just an artifact of comparing unmatched populations). The
[[claim-matched-pqct-comparison-finds-bone-loss-comparable-while-muscle-loss-under-runs|Böcker 2026 matched pQCT study]]
is the hinge that fuses them: it removed the unmatched-population confound by scanning both
conditions on cross-calibrated instruments at the same four tibial sites — and found the
bone near-equality *holds under matching*, while the clean ~2x under-run reappears for
**muscle**. So the readout-dependence is terrain, not a mirage, and the "half-strength"
story was really a muscle story wearing a bone label. That single study is why the turnover
leg, the density leg, and the reconciliation trio are one argument with cause and
consequence rather than three topics filed next to each other.

Titled for the argument's arc — a readout-dependent fidelity that inverts a working
assumption — not for bed rest, Böcker, or the tibia, per the 2026-07-25 lesson.

## Where the concern came from (origin, cross-linked not folded)

The flight-side problem and the ground program built to answer it. Context for the argument,
not part of its spine.

- [[claim-gemini-nutrition-report-embeds-a-bone-demineralization-concern]] — the flight-side
  origin. A 1967 NASA Gemini nutrition memo already fortified calcium "as related to the
  bone demineralization (M-6) and calcium balance (M-7) experiments" — a skeletal-loss worry
  operational by 1966, tucked inside a calorie ledger. Tier 1 (NTRS).
- [[claim-head-down-bed-rest-is-the-standard-ground-analog-for-spaceflight-bone-loss]] — the
  paradigm claim, and the cross-domain reason it matters: the same immobilization method
  that models astronaut bone loss is, in substance, how clinical medicine studies disuse
  osteoporosis. Still `seedling` — its mechanistic-equivalence half rests on secondary
  search prose (`[unverified-mechanism]`), and the note says so. The pioneering PI is
  [[claim-pauline-beery-mack-became-nasas-bone-density-pi-at-seventy]].

## The two readouts — where "half-strength" is true and where it fails

The core of the readout-dependence. Both legs are Tier 1; the tension between them is the
argument.

- [[claim-bed-rest-bone-turnover-markers-run-half-spaceflight-magnitude]] — the leg where the
  tidy story holds. Morgan et al. 2012 (30-day −6° HDT, PMC3524659): resorption markers rise
  50–75% in bed rest vs 100–150% in flight, PTH suppression "about one-half" the flight
  decrease. The cleanest single-paper ~2x under-run — and measured "in individuals not
  performing intensive resistance exercise," i.e. the analog under-runs partly by *omitting*
  the countermeasure regime, a caveat about what it leaves out, not just how it scales.
- [[claim-bed-rest-bmd-loss-about-one-percent-per-month-hip-pelvis-heel]] — the leg where it
  fails. Spector et al. 2009 (PMID 19476166; 13 subjects, 42–90 days HDT, DXA): ~1%/month at
  hip, pelvis, heel, "consistent with earlier bed rest and spaceflight studies." A solid
  standalone number and, by the note's own admission, a shaky comparison point — the
  per-region P-values live in an unread companion paper.
- [[claim-spaceflight-bmd-loss-lower-limbs-about-0-8-percent-per-month]] — the orbital
  reference figure the analog is judged against. Stavnichuk et al. 2020 meta-analysis
  (PMC7200725): lower limbs −0.8% [−1.1, −0.5]/month, upper limbs/thorax essentially flat at
  −0.1% — a weight-bearing gradient consistent with mechanical unloading as the driver.
- [[claim-bed-rest-under-runs-spaceflight-bone-loss-depends-on-readout]] — the synthesis
  node. Direction reliable, magnitude attenuated, but the *size* of the under-run depends on
  the instrument: ~2x on turnover, near-parity on BMD rate. Flags the BMD comparison as *not*
  a matched head-to-head — the near-equality could be an artifact of unmatched populations.
  That flag is exactly what the Böcker study below then tested.

## The hinge — matching the instruments, and the inversion

The Böcker et al. 2026 matched pQCT study (npj Microgravity, s41526-026-00611-2; n=13
spaceflight / n=11 bed rest, no countermeasures). Read this section as the load-bearing
middle of the map.

- [[claim-matched-pqct-comparison-finds-bone-loss-comparable-while-muscle-loss-under-runs]] —
  the inversion, and the reason the whole cluster coheres. Same pQCT protocol, four homologous
  tibial sites, cross-calibrated devices: "The muscle loss observed post-flight was
  approximately doubled compared to that … observed [after] 2 months of bed rest, while bone
  loss was comparable." The artifact hypothesis the readout note could only flag is tested
  under matching and *holds* — bone converges, muscle keeps the ~2x. The clean twofold gap
  moved from bone to muscle.
- [[claim-bocker-concludes-60-days-bed-rest-models-six-month-spaceflight-bone-loss]] — the
  dose-equivalence framing. The authors' own bottom line is not a magnitude multiplier but a
  timing one: 60 days of bed rest without countermeasures models six-month lower-limb bone
  loss, while comparable muscle loss needs ~80 days. Bone and muscle are modeled at
  *different bed-rest doses* — the dose-scale echo of the same asymmetry.
- [[claim-bocker-matched-comparison-is-site-matched-not-duration-matched]] — the honest hole
  in the hinge. Two of three matching conditions met (site, instrument — European Forearm
  Phantom, 1.4/1.2/1.6% device difference); the third is not: missions ran a median 178 days
  (range 165–340), ~3x the 60-day arm, and astronauts exercised daily while bed-rest subjects
  did nothing. The paper names this itself and leans on a *cited* dose-equivalence convention
  (Hargens & Vico 2016) the vault has not read
  ([[question-verify-hargens-vico-2016-bed-rest-spaceflight-dose-equivalence]]). The
  "comparable" verdict rides on that citation, not on matched exposure.

## The reconciliation — two orbital numbers that are not the same number

Added 2026-08-15, when a cosine-0.87 pairing put Stavnichuk's pooled rate next to Böcker's
matched result. These three notes keep the cluster from over-claiming a two-paper agreement
it does not have.

- [[claim-bocker-cites-stavnichuk-without-testing-its-pooled-rate]] — the bridge is real but
  load-light. Böcker cites Stavnichuk (his ref 17) only as background for "adaptations still
  occur despite countermeasures," never as a benchmark his own analysis tests against. The
  two notes connect *through the readout-dependence synthesis note in the middle*, not by
  checking each other's figures — the body text should always route through that middle, never
  imply the two papers validated each other.
- [[claim-stavnichuk-and-bocker-measure-different-bone-quantities]] — why the two headline
  numbers are not interchangeable: Stavnichuk's −0.8%/month is a *pooled* estimate across four
  imaging methods (projection radiography, SPA, DXA, qCT) held together by an explicit
  same-quantity assumption; Böcker's "comparable" is *single-instrument volumetric* content
  (vBMC) at four tibial sub-sites. A tape measure over a whole limb-region shadow versus a CT
  slice of one bone. (This note carries a thorough 2026-08-16 cross-model correction on the
  Stavnichuk side — four defects found and fixed — and is the best-audited member here.)
- [[claim-bocker-tibial-losses-match-stavnichuk-only-at-epiphyseal-sites]] — the site-level
  reconciliation, and the map's live seam. Converting Böcker's per-site losses to monthly
  rates (Seek's own arithmetic, flagged as such) puts the trabecular-rich epiphyseal sites
  (~−0.58 to −0.68%/month) inside Stavnichuk's confidence interval and the cortical
  diaphyseal sites (~−0.08 to −0.15%/month) five- to tenfold short — the pooled number is a
  weighted echo of the epiphyses.

## Open threads (honest caveats, not hidden)

- **The load-bearing "60 days = 6 months" equivalence is unverified at its root.** The
  hinge's real-world validity rests on Hargens & Vico 2016, cited one layer back and unread
  ([[question-verify-hargens-vico-2016-bed-rest-spaceflight-dose-equivalence]]). Without it,
  "comparable" is a within-study result over mismatched exposures.
- **The compartment mechanism is a live contradiction, not a settled finding** — and it is
  the "note in its own right" the 2026-07-25 flag asked this map to give a home. The tidy
  "trabecular turns over faster everywhere" reading that would explain the
  epiphysis-converges / diaphysis-diverges pattern is contradicted by Böcker's own
  discussion: bed-rest pQCT reportedly shows *more cortical* loss (ref 22) while spaceflight
  QCT shows *more trabecular* loss (refs 12, 23). Those refs were read only in Böcker's
  summary; routed to [[question-verify-bocker-compartment-cortical-trabecular-disagreement]].
  Which compartment drives the bed-rest-vs-flight difference is unresolved.
- **The monthly-rate conversion is Seek's arithmetic, not either paper's figure** — kept
  because it makes the bridge visible, flagged because it inherits the unverified mechanism.
- **The matched study is not duration-matched**, and the muscle/bone asymmetry has a
  *plausible but untested* engine (astronauts exercised, bed-rest subjects did not — sparing
  muscle in orbit while both lose bone under unloading). Böcker does not test that engine, so
  it stays hypothesis, not finding.
- **Every member note is still `seedling`.** The turnover, DXA, and meta-analysis figures are
  Tier 1 and several are `verified_verbatim`; the paradigm and mechanism claims are the soft
  parts, and the notes say so. This map records the cluster's current footing, not a frozen
  verdict.
- **Entities:** [[entity-jorn-rittweger]] (senior author on the hinge study) and
  [[entity-mariya-stavnichuk]] (lead of the reference meta-analysis).

> [!note] Warden's commentary:
> The tell that this is one argument and not "everything about space bone loss" is that the
> two readout-legs are load-bearing on the same hinge: the DXA note could only *suspect* the
> near-equality was an unmatched-population artifact, and the one study that removes the
> mismatch is the one that turns the suspicion into terrain — and in the same stroke relocates
> the clean twofold gap from bone onto muscle. That inversion is the find, and it is worth
> stating carefully, because it is the kind of neat reversal that tempts overstatement. So I
> kept three restraints on the record. The bone-converges result is genuinely Tier 1 and
> genuinely matched on site and instrument — this is not a debunking map. But it is *not*
> duration-matched, its real-world reading leans on a dose-equivalence paper nobody here has
> read, and the mechanism that would explain the site pattern is one Böcker's own discussion
> says the compartment data may contradict. The honest shape is a strong, surprising, load-
> bearing finding sitting on three named unknowns — which is exactly why the cluster wanted a
> map that routes a reader through the middle notes instead of letting two orbital numbers
> shake hands they never shook. — warden/claude-opus-4.8, 2026-08-20