In partnership with
Replace 16 supplements in one daily drink.
David Beckham. Giannis Antetokounmpo. Aryna Sabalenka. Inter Miami CF. The world's best choose IM8 because performance starts with what you do every day.
Daily Ultimate Essentials and Longevity deliver a complete, clinically formulated system for energy, digestion, sleep, focus and healthy aging. In a 12-week randomized controlled trial, 95% felt a boost in daily energy, 85% felt less bloated with better digestion, 80% reported better sleep and 75% reported sharper focus and mental clarity.
IM8 is NSF Certified for Sport, tested against 280+ banned substances and recommended by 300+ clinicians. With 50 million servings sold across 46 countries, discover the IM8 routine trusted at the highest levels of sport—and built for everyday life. See the clinical results.
LONGEVITY LATEST ISSUE 30 · 30 SEPTEMBER 2026
LONGEVITY LATEST · DEEP DIVE
What a T-Score Can’t See
Why weight loss thins bone, when the scan misleads, and how to tell whether your own result has really changed.
By Christian Thomsen · Companion to Issue 30 · 30 September 2026 · ~7-minute read
I expected the bones of dieting older adults to come out weaker. In LIMB-Q, a one-year trial of diet plus aerobic and resistance training in adults aged 65 to 85 with obesity, hip density on DXA fell, as it usually does with weight loss. But CT measures of cortical thickness, and CT-based estimates of hip strength, held up. Less density, same estimated strength.
Now the uncomfortable half. Look AHEAD randomised 5,145 adults with type 2 diabetes to intensive lifestyle change or diabetes support and education. Over a median of 9.6 years, the intervention group lost 6.0% of their weight against 3.5%. Total and hip fractures didn’t differ. Frailty fractures, the kind that follow minor trauma, were 39% more common (HR 1.39).
The scan can overstate the damage, and it can miss it. Both happen.
Issue 30 set out three numbers: density, fracture risk and falls. This piece looks underneath the first: what the scan measures, why weight loss changes it, what evidence is still missing, and how to read your own results over time.
Gregori et al. LIMB-Q. Lancet Healthy Longevity (2025) · Johnson et al. Look AHEAD fractures. JBMR (2017)
What the T-score measures, and what it doesn’t
DXA passes two low-dose X-ray beams through the hip or spine and reports areal density: grams of mineral per square centimetre of a flat projection. The T-score compares that figure with a healthy young adult’s, in standard deviations. At −2.5 or lower the WHO calls it osteoporosis; between −1 and −2.5, osteopenia. Four things fall outside that picture.
Size: a flat projection can’t see depth, so larger bones read denser than smaller bones of the same quality.
Architecture: the thickness of the outer shell and the lattice of struts inside change strength without changing density much. CT and the DXA-derived trabecular bone score capture some of this; routine reports usually don’t.
Spine artefacts: arthritis and compressed vertebrae can push spine readings up, which is why I trust the hip more in older adults.
Soft tissue: large changes in weight or fat can affect the precision and accuracy of the measurement, which is exactly the situation during weight loss.
Why most fractures happen above the line
Risk climbs as density falls, but far more people sit in the osteopenia band, so that’s where most fractures happen in absolute numbers. NORA, which used peripheral scanners, put a starker figure on it: 82% of women who fractured within a year had T-scores better than −2.5.
The more useful detail is who those women were. Among the non-osteoporotic women who broke a hip in the Study of Osteoporotic Fractures, older age, no exercise in the past year, poor contrast vision, recent falls and an existing spinal fracture all raised risk. Three of those five aren’t captured by a density scan at all.
Why weight loss thins bone
Four mechanisms overlap.
Less load: bone cells buried in the matrix sense strain. When load falls they release more sclerostin, a brake on bone formation. In a one-year trial of 107 older adults with obesity, sclerostin rose about 10% with diet alone and stayed flat when exercise was added.
Fewer hormonal signals: fat tissue is a source of oestrogen and leptin. In 160 dieting older adults, falls in body mass and leptin independently predicted hip density loss.
Faster turnover: breakdown speeds up and formation doesn’t keep pace. In a placebo-controlled semaglutide trial, a breakdown marker rose while the formation marker stayed flat.
Measurement: some apparent loss may reflect the shifting soft tissue around the bone rather than the bone itself.
A caveat worth noting: a lighter body needs less bone, so part of the loss may be sensible adjustment rather than damage. LIMB-Q supports that reading. Look AHEAD’s frailty fractures say it isn’t the whole story.
Armamento-Villareal et al. JBMR (2012) · Armamento-Villareal et al. JBMR (2020) · Hansen et al. eClinicalMedicine (2024)
What the weight-loss drug data show
Four studies, four slightly different answers.
Randomised, higher-risk adults: in a Danish trial of 64 adults with increased fracture risk, a year of semaglutide 1 mg brought 6.8 kg more weight loss than placebo, along with lower hip and spine density.
Randomised, with exercise: the Danish liraglutide trial from Issue 30, where adding exercise preserved hip and spine density that the drug alone lowered.
Real-world, matched: a July 2026 JCEM study paired 255 semaglutide or tirzepatide users with 255 non-users, all with repeat DXA scans. Over a median of 17 months, users without diabetes lost hip density faster; users with diabetes didn’t differ from controls. Observational, and people sent for repeat scans aren’t typical.
Older adults, short term: a 20-week pilot analysis found no clear change in density or bone markers, though the numbers leaned the wrong way.
Modest density loss, largely tracking weight loss, in trials too small and short to count fractures. That’s the honest summary.
None of these studies was designed or large enough to detect a difference in fractures. Too few fractures to judge isn’t the same as reassurance.
Hansen et al. (2024) · Jensen et al. JAMA Netw Open (2024) · Liu et al. JCEM (2026) · Frontiers in Aging pilot (2025)
What evidence is still missing
Each option graded in Issue 30 is waiting on one kind of study.
Loading: a trial during drug-induced weight loss that measures CT bone strength and follows fractures, using loading as heavy as LIFTMOR’s supervised lifts above 85% of maximum.
Treating by risk: fracture trials in men, and in people losing weight on the new drugs.
Weighted vests: a trial designed around women that tests whether the subgroup signal holds.
What would change my view: a trial showing fewer fractures, not just more density, from any bone-protection strategy during drug-induced weight loss.
Five checks on a bone headline
Measure: density, CT bone strength, bone markers or actual fractures?
Site: hip, spine or forearm? Spine results in older adults can be inflated by arthritis.
Noise: is the change bigger than the scanner’s least significant change?
Comparator: did a control group losing similar weight lose similar bone?
Population: older or younger, men or women, already osteopenic, already on bone drugs?
Run those five before letting a result change what you do.
Reading your own scan
Like the chair test in Issue 29, a DXA result wobbles. A change smaller than the wobble may be real, but it can’t be confidently told apart from measurement error. Five rules help.
Two numbers, two jobs: the femoral neck result is the one FRAX uses. For tracking change, the total hip is usually better, because it’s measured more precisely: reported precision errors run about 0.8–1.2% at the total hip against 1.4–1.6% at the femoral neck.
Compare density, not T-scores: the least significant change is calculated on the density itself, in g/cm², so that’s the figure to compare between scans.
Know your centre’s threshold: each centre measures its own precision error, and the change needed for 95% confidence is 2.77 times that error. A centre with a 1.2% error needs a change of about 3.3%; a 2% drop there is unproven, not absent. The ISCD’s 5% at the total hip and 5.3% at the spine are the most it considers acceptable for a centre, not a universal threshold. Ask for your centre’s own figure.
Same machine, sensible timing: results from different scanners need cross-calibration before they’re comparable. On most standard bone drugs, real change is rarely detectable within three years. Rapid weight loss moves density faster, but that’s also when shifting soft tissue muddies the reading most.
Fracture beats T-score: a low-trauma hip or spinal fracture counts as osteoporosis whatever the scan says.
For FRAX, run it without density first, then with the femoral neck result. If your ten-year risk sits near your country’s treatment threshold, that’s when the scan earns its place. FRAX isn’t built to show whether a treatment is working. If you’re already on one, follow-up draws on repeat scans on the same machine, any new fractures and a clinical review, with FRAX still useful when a treatment change is being considered.
ISCD precision FAQ · DXA precision. J Clin Densitom (2019) · Updated DXA practice guideline (2025) · FRAX calculator
If I were designing a protocol from this evidence, it would be boringly simple: a fracture-risk calculation before any big weight loss after 60, lifting twice a week with some impact if your joints and physio allow, calcium and protein from food, a standard vitamin D dose only where it’s indicated, and repeat scans on the same machine read against their noise. The most expensive item in it is the scan.
Density is a clue. The fracture is the outcome.
Sources and further reading
BMD precision for lumbar spine configurations, with hip precision benchmarks. J Clin Densitom (2019)
Evidence reviewed through 25 September 2026. Fracture findings here come from two randomised trials (Look AHEAD, Auckland zoledronate) and two observational cohorts (SOF, NORA); the GLP-1 and exercise findings rely on density, CT estimates and bone markers. Least-significant-change figures vary by centre. Educational content only; do not start, stop or change treatment without a qualified clinician.
© 2026 FrontWave Media Ltd · Longevity Latest