RightRep EN · PT Open the app
BLOG

Can You Train With Osteoporosis? The Bone-Bank Barbell Audit: Where Heavy Is the Deposit, Flexion Under Load Is the Withdrawal, and the Spine-Crush Fear Died on a Follow-up X-ray (LIFTMOR Ran 5×5 at >85% 1RM Through 101 Postmenopausal Spines for 8 Months and Lumbar Density Rose 2.9% While the Light-Exercise Control Fell 1.2%, Zero Training Fractures on Genant-Read Films With Kyphosis Improving 6.7° and the One New Wedge Deformity Showing Up in the Control Group, the Mayo Decade Where Two Years of Back-Extension Work Still Carried 2.7× Fewer Vertebral Fractures Eight Years After the Program Ended, the 1984 Vector File Where 89% of the Flexion Group Fractured Against 16% of the Extension Group, Falls Cut 23% Across 108 Trials and 23,407 Participants, and the Bisphosphonate Ledger Where Adding Training Bought 5.27 More Femoral Points and 30.16 Quality-of-Life Points Over Drugs Alone), Load the Skeleton, Never Flex It Under Load

By RightRep AI Assistant · · 24 min read

Osteoporosis is the condition this library's treatment twin — does exercise help osteoporosis — answers from the medicine side: training moves the bone-density, strength, balance and fall columns while the DXA schedule, the FRAX score and the prescription belong to the physician. This file answers the barbell side: what happens to a loaded training program when the skeleton underneath it runs at low density — a condition pooling 18.3% of the world's adults and 23.1% of its women across 103 million people [10], and in Latin America one where 11.18% of women 50+ already carry a silent radiographic vertebral fracture [9]. The default scripts fail in opposite directions. The walk-only script treats "low bone mass" as a lifetime instruction to keep loads light — while the trial that finally tested the opposite ran heavy barbell work through this exact population for eight months and the light-exercise control was the group that lost bone: lumbar spine −1.2% under home-based gentle exercise against +2.9% under twice-weekly 5×5 at >85% 1RM [1]. The reckless script loads the spine any way it folds — while the classic Mayo vector file counted new vertebral fractures in 89% of the flexion-exercise group against 16% of the extension group [4]. The reframe that carries the whole program is one line: the bone is a bank account — heavy loading is the deposit, flexion under load is the withdrawal, and the barbell is the only teller who takes both sides of the ledger [1][4][8]. Which puts the fences in order. First: the diagnosis and the drugs belong to the physician — DXA cadence, FRAX, bisphosphonates and the anabolic agents are standard of care, and the training file co-signs the prescription, never replaces it [11]. Second: red flags refer before they get programmed — sudden severe mid-back pain, pain after a slip or a cough, measurable height loss or a first low-trauma fracture is a physician visit, not a training day [9]; and with vertebral fractures already on your films, the program is built with a clinician, because the consensus panel's own honesty says trials in the fractured population are few and its recommendations are conditional [5]. Third: there is exactly one hard movement rule — the spine does not fold under load — no sit-ups, no loaded toe-touches, no rounded-back lifting; the hip hinges and the spine stays a rod [4][5]. With the fences standing, the file underneath is unusually complete: the heavy question has a randomized answer with X-ray follow-up [1][2], the decade question has a ten-year fracture trial [3], the vector question has a forty-year-old natural experiment [4], the falls question has 108 randomized trials [7], the dose question has the osteogenic index [8], the drug-interaction question has a fresh meta-analysis [11], and the pt-BR lane has its own randomized file [12].

The short answer: yes — and heavy is the deposit, not the risk

Strip the file to its spine. The permission is structural, not heroic: the LIFTMOR trial took 101 postmenopausal women aged 65±5 with T-scores below −1.0 — osteopenia and osteoporosis, screened for confounders — and randomized them to eight months of twice-weekly, 30-minute, fully supervised high-intensity resistance and impact training (HiRIT: the barbell compounds at 5 sets of 5 repetitions above 85% of one-rep max, plus impact loading) against a home-based low-intensity program [1]. The results inverted the folk advice: lumbar-spine BMD rose 2.9±2.8% in the heavy group while falling 1.2±2.8% in the light group (p<0.001); femoral neck +0.3% against −1.9% (p=0.004); femoral cortical thickness +13.6% against +6.3% (p=0.014); height held (+0.2 cm against −0.2 cm, p=0.004); and every functional measure — timed up-and-go, functional reach, sit-to-stand, back and leg strength — moved (p<0.001) [1]. The safety ledger reads one adverse event in eight months of >85% 1RM barbell work: a minor low-back spasm that cost two of 70 sessions [1]. And when the same group re-read the participants' spines on follow-up lateral DXA with the Genant semiquantitative method — the imaging gold standard for vertebral fracture — the HiRIT group's fracture classification did not change, standing-tall thoracic kyphosis improved 6.7° against 1.6° in controls (p=0.031), and the single new wedge deformity in the trial appeared in the light-exercise group [2]. The decade file underneath is even quieter: fifty Mayo women, two years of progressive back-extension strengthening, then eight years of nothing prescribed — and at year ten the exercise group still carried significantly stronger backs (p=0.001), a BMD edge that had emerged over the off years (p=0.0004), and 2.7 times fewer vertebral compression fractures (1.6% of vertebral bodies versus 4.3%, p=0.029) [3]. That is a permission file with a meter attached: bone responds to magnitude, and the light script doesn't clear the minimum dose [8] — which is why the instrument that reads this program is the training log across years, not the mirror across weeks.

The reframe: the bone bank, not the bone machine

Everything operational in this file hangs on one honest division of labor. What training does: it deposits density at the loaded sites [1][6], it builds the muscle sleeve that braces the spine [3][4], it cuts the falls that actually break hips [7][12], and it adds femoral and quality-of-life points on top of pharmacotherapy [11]. What training does not do: it is not a substitute for the prescription, the DXA schedule or the fracture-risk conversation [5][11], and it does not rebuild a vertebra that has already collapsed [5]. The two sentences have to be carried together, because each one alone produces a broken program. The lifter who believes the barbell replaces the drug quits her bisphosphonate and chases the pharmacy through the squat rack — while the synergy file says the combination is the arm that wins (adding training to the drug bought femoral BMD MD 5.27, 95% CI 2.23–8.32, and quality of life MD 30.16, 22.16–38.16, over drugs alone) [11]. The lifter who believes osteoporosis forbids loading parks herself for decades on the walking-and-water-aerobics script — and the control arm of the flagship trial shows what that account does unattended: it drifts down 1.2% a year at the spine while she waits [1]. The reframe also settles the identity of the program: bone's remodeling cycle is measured in months and its meaningful columns move in single-digit percentages over eight months of heavy supervised work [1][6] — this is a decade-scale deposit program, the one fitness goal where starting at 55 and still training at 75 is the plan working [3]. Which is why the boring infrastructure — sustainable weekly set budgets, a frequency that survives bad weeks, progressive overload climbed patiently at the edge of technique — is not beginner padding in this file. It is the treatment architecture, because the deposit only posts while the program keeps running [3][8].

The LIFTMOR file: what heavy actually did, in the population everyone fears for

The trial deserves its numbers read exactly, because it is the most counterintuitive training file in the medical literature. The program was brief and brutal by geriatric standards: two 30-minute supervised sessions a week for eight months, built on the barbell compounds — deadlift, back squat, overhead press — at 5×5 above 85% 1RM, with impact loading (jumping and landing patterns) layered in, at The Bone Clinic under clinicians [1]. The population was not cherry-picked robust elderly: 65±5 years, T-score below −1.0, low to very low bone mass [1]. Compliance ran 92±11% — the highest-compliance hard-training cohort on this shelf, because 30 minutes twice a week is a program a 68-year-old can keep [1]. The dose was not accidental: bone's mechanosensor rewards high-magnitude, high-rate loading delivered in brief bouts with rest inserted — the osteogenic-index mechanics — and desensitizes to long, repetitive, same-magnitude work [8], which is precisely why the 5×5-with-rest barbell format deposits more bone in thirty minutes than an hour of comfortable circuits [1][8], and why variety in movement direction earns its keep [8]. Three operational translations. One: the heavy tier is the medicine tier. The pooled Cochrane file agrees from below — across 43 randomized trials and 4,320 postmenopausal women, the best femoral-neck response came from non-weight-bearing high-force work (progressive lower-limb resistance, MD 1.03, 0.24–1.82) and the best spine response from combination programs (MD 3.22, 1.80–4.64) [6]; LIFTMOR then showed what the top of that dose curve does when supervision holds the technique line [1]. Two: the machine-versus-barbell question inverts here. The machines-vs-free-weights debate is about hypertrophy elsewhere; in this file machines are the on-ramp and the heavy bar is the destination, because the spine-loading patterns that deposit bone are exactly the compounds — with machine-supported variations as the flare-week and confidence-week vehicles [1][6]. Three: the safety claim is conditional on its conditions — supervised, progressive, technique-gated, five reps never to failure, two sessions a week [1]. Take the supervision and the technique gate away and you have not recreated the trial; you have recreated its reputation problem [2].

The fear file: the spine-crush question, answered with X-rays

Every osteoporotic lifter — and every osteoporotic person being told to lift — eventually asks some version of "won't heavy weights crush my spine", and this shelf holds the rare medical luxury: a trial that took the question to imaging. The LIFTMOR follow-up ran lateral thoracolumbar spine scans on both arms before and after eight months of >85% 1RM barbell work, read them with the semiquantitative Genant method — and found no change in vertebral-fracture classification in the HiRIT group, a clinically relevant improvement in standing-tall kyphosis (−6.7±8.2° vs −1.6±8.1°, p=0.031), and exactly one new wedge deformity in the whole trial: in the light-exercise control [2]. Heavy extension-biased loading did not just spare the osteoporotic spine — it stood it up straighter [2]. Now the other half of the fear, because it is where the real risk lives: the vector file. The 1984 Mayo study put fifty-nine women with postmenopausal spinal osteoporosis into extension (n=25), flexion (n=9), combined (n=19) or no exercise (n=6), and followed their X-rays for one to six years: new wedging and compression fractures hit 16% of the extension group, 89% of the flexion group, 53% of the combined group and 67% of the no-exercise group — flexion versus extension at p<0.001 [4]. Read honestly: the flexion arm was nine women, this is forty-year-old data, and it is exactly the kind of small old study that should shape movement selection rather than headlines — but its direction has never been overturned, and the consensus panel wrote its shadow into the guidelines: strengthen, balance, load — and manage spinal flexion [4][5]. The gym-floor translation is the one hard rule this file carries: the spine is a rod under load and a hinge only without it — sit-ups, crunches, loaded toe-touches, rounded-back deadlifts and folded-over rows are the named withdrawals [4][5]; the deadlift, done as a hip hinge with a neutral spine, is not on that list (do deadlifts damage your back owns the general audit), and neither is the squat that keeps its chest (do squats damage your knees owns its own file) — in this population both lifts keep their LIFTMOR seat because the technique gate is the safety mechanism [1][2]. The mirror fear — "so light work is the safe option" — dies in the control arm: the light group lost spine density and logged the trial's only new fracture [1][2].

The decade file: what back-extension strength buys at year ten

If the fear file answers "will the bar hurt me", the Mayo decade file answers the better question: what does the bar still owe me years later. Fifty healthy white postmenopausal women aged 58–75, randomized to two years of progressive resistive back-strengthening (n=27) or control (n=23), then re-examined at year ten — eight years after anyone prescribed them anything [3]. Back-extensor strength: 39.4 kg → 66.8 kg after the program, and still 32.9 kg at year ten against controls' 26.9 kg from a 49.0 kg peak — the strength differential survived (p=0.001) [3]. Bone mineral density, statistically indistinguishable between groups at baseline and year two, had separated by year ten (p=0.0004) — the muscle advantage compounded into a bone advantage while nobody was training on paper [3]. And the fracture column: 14 compression fractures across 322 vertebral bodies in controls (4.3%) against 6 across 378 bodies in the exercise group (1.6%) — relative risk 2.7 times higher in the controls [3]. The programming translation is an extension bias, not a flexion ban on life: rows, back extensions, hip hinges, the deadlift family and the overhead work that makes you stand tall under load are all deposit-side movements for the exact muscles this file protects [3][4]. It also prices the identity question this population quietly carries — "am I a patient or a lifter?" — with the cleanest answer on the shelf: the woman who trained hard for two years was still measurably protected eight years into not training, which is the compounding argument for never letting the program lapse that long again [3]. The decade file's honest summary: bone is slow to lend and slower to forget — the deposits outlive the deposits' schedule.

The falls file: the fracture you can actually train against

Here is the hierarchy the consensus panel drew, and every lifter in this file should carry it: the strongest evidence in exercise-and-osteoporosis is not BMD — it is falls [5]. The hip fracture that ends independence usually arrives through a fall, not through a vertebra quietly folding [7], and the pooled falls file is as clean as exercise science gets: 108 randomized trials, 23,407 community-dwelling participants aged 60+ across 25 countries — exercise cut the rate of falls by 23% (rate ratio 0.77, 95% CI 0.71–0.83, high certainty), the number of fallers by 15% (RR 0.85, 0.81–0.89, high certainty), and probably fall-related fractures (RR 0.73, 0.56–0.95, low certainty) [7]. The Brazilian file underneath is the pt-BR lane's own: the Brazilian Cochrane Centre at UNIFESP-São Paulo took 100 sedentary postmenopausal women already on osteoporosis drug therapy, randomized 50/50 to 18 weeks of progressive quadriceps loading climbing from 50% to 80% 1RM plus proprioception work — or drugs alone — and counted falls for 24 weeks after the program ended: falls at an incidence-rate ratio of 0.263 (95% CI 0.10–0.68, p=0.0064), with 1-RM, timed up-and-go, Berg balance and all eight SF-36 subscales better (p≤0.0018) [12]. Translation for the gym floor: the balance column is not padding around the barbell work — it is co-equal medicine [5]. Program it like you mean it: unloaded and lightly loaded single-leg work (split squats, steps, stance variations), carries, deliberate tempo and landing work in the warm-up (warm-up before lifting earns a second job here), and the proprioceptive seasoning the São Paulo file used on purpose [12]. And keep the honest ceiling in view: the falls file is high-certainty for rate of falls and lower-certainty for fractures [7]; the fracture column is where the bone file [1][3], the vector file [4] and the falls file [7] stack — no single one of them carries it alone.

The drug+exercise ledger: the co-signed prescription

If pharmacotherapy is part of your management — bisphosphonates being the first-line class — this is the section that decides what the gym is for. The fresh meta-analysis pooled seven randomized trials (314 participants) comparing bisphosphonates alone against bisphosphonates plus physical intervention: the combination arm bought femoral BMD an extra MD 5.27 points (95% CI 2.23–8.32, p=0.0007) and quality-of-life an extra 30.16 points (22.16–38.16, p<0.00001) over the drug alone — while lumbar spine, total hip and total body BMD showed no significant differences, and the reviewers flag their own heterogeneity honestly (I²=88–100% on the pooled BMD outcomes) [11]. Read as a ledger, not a miracle: the drug and the bar each hold their own accounts — the antiresorptive slows the spend, the loading posts the deposit at exactly the sites the bar loads [1][11] — and the quality-of-life column is the one where adding training moved the number most [11]. Three operational consequences. One: never quit the prescription for the program. The combination arm is the winning arm; nothing in the exercise file edits the prescription [5][11]. Two: the anabolic window question belongs to the prescriber. Which agent, when, drug holidays — physician calls; your job is to keep the deposit side of the ledger running so whatever they prescribe has something to protect [11]. Three: this is the same bone account the hormonal lanes spend from — the endometriosis file priced GnRH suppression at −1.50% spine density a year and named the barbell the deposit side, and the hypothyroidism file guards the same account from the thyroid side; this file is the account's home branch [11].

The epidemiology: the quiet fracture you already have, and who this file is for

The scale anchor first: across 86 studies and 103,334,579 people, the pooled global prevalence of osteoporosis is 18.3% (95% CI 16.2–20.7) — 23.1% in women (19.8–26.9) and 11.7% in men (9.6–14.1) [10]; this is not a niche diagnosis, it is roughly a fifth of the adults a gym serves, and the men are in the file too. The quiet-fracture anchor next, because it reframes "who is the osteoporotic lifter": LAVOS X-rayed a random sample of 1,922 women 50+ across Argentina, Brazil, Colombia, Mexico and Puerto Rico and found 11.18% (9.23–13.4) already carrying a radiographic vertebral fracture — 6.9% at 50–59 rising to 27.8% at 80+ — with self-reported height loss (OR 1.63) and prior fracture (OR 1.52) the flags that traveled with it, and physical activity carrying OR 0.73 (0.55–0.98) in the bivariate analysis [9]. Two translations. One: the undiagnosed reader is this file's modal reader. The woman training at 61 with a slightly stooped thoracic spine and no DXA yet is precisely the LAVOS percentile this shelf serves — which is why the fences put height loss, new mid-back pain and first low-trauma fractures in the physician's queue before the barbell's [9]. Two: the training population is wider than the diagnosed one. LIFTMOR enrolled T-score < −1.0 — osteopenia included [1]; the Howe pooled file covers the same band [6]; and the men's 11.7% column [10] runs the identical program with the identical vector rule [4]. What the epidemiology does not do is decide your program: prevalence files size the audience, and the DXA plus the fracture-risk conversation with your physician sizes yours [5][9].

The Brazil anchors: who the pt-BR twin serves

The pt-BR lane's reader holds a specific position in this condition: Brazil was one of LAVOS's five countries — the 11.18% vertebral-fracture prevalence and its 6.9%-at-50-something baseline are her domestic epidemiology, not an imported statistic [9] — and the strongest falls-prevention randomized trial on this shelf was run at the Brazilian Cochrane Centre at the Federal University of São Paulo: 100 postmenopausal women with osteoporosis, all on medication, randomized to 18 weeks of progressive quadriceps loading at 50→80% 1RM plus proprioception — and the training arm fell at one quarter the rate of the medication-only arm (IRR 0.263, 0.10–0.68) with every functional and quality-of-life measure moving [12]. "Musculação faz mal para osteoporose" — the domestic gym folklore this file prices — dies against the imaging file: eight months of >85% 1RM barbell compounds produced zero training fractures on Genant-read films, improved kyphosis, and left the trial's only new fracture to the light-exercise group [1][2]. The mirror half — "caminhada é o exercício seguro, peso não" — dies in the same table's control column: the light home program is where spine density went down 1.2% [1], and the pooled file prices the gentle-only script at a fraction of the bone response of loaded combinations [6]. The pt-BR twin exists to replace both halves with the bone bank: carregue o esqueleto, nunca flexione a coluna com carga — e deixe o remédio e o exame com quem os prescreve [11].

So what do you actually do? The practical file

Six rules, every one sourced from the files above:

  1. Keep the physician as the program's co-signer. DXA cadence, FRAX and the prescription are theirs [5][11]; the barbell is the deposit side of the account the drug protects — and the combination arm is the winning arm (femoral MD 5.27, QoL MD 30.16 over drugs alone) [11].
  2. Load heavy, load brief, load supervised-or-gated. The trial that grew spine density ran 2×30 min/week at 5×5 above 85% 1RM under supervision [1]; the osteogenic index says magnitude and rate with rest in between [8]; climb toward that edge through technique gates, with reps in reserve as the ceiling and machines as the on-ramp [1][6].
  3. Never flex the spine under load — the one hard rule. The vector file's 89%-vs-16% fracture split belongs to loaded flexion, not to load [4]; sit-ups, loaded toe-touches and rounded-back lifting are the withdrawals [4][5]; the hinge-with-rod-spine deadlift and the chest-up squat keep their seats because the technique gate is the safety mechanism [1][2].
  4. Build the extension bias and the balance column on purpose. Back-extensor strength at year ten carried 2.7× fewer fractures [3][4]; the falls file cuts fall rates 23% across 108 trials [7] and the São Paulo program cut falls to a quarter with quadriceps loading + proprioception [12] — single-leg work, carries and landing patterns are medicine here, not garnish [5].
  5. Run the program as a decade-scale deposit. Bone moves in single-digit percentages over eight heavy months [1][6] and still pays dividends eight years after lapsing [3] — sustainable set budgets, a survivable frequency, deloads when the body votes, and the sleep and protein columns the deposit posts through.
  6. Refer the red flags, refuse both fear stories. Sudden mid-back pain, height loss, a first low-trauma fracture go to the physician before the barbell [9]; with known vertebral fractures, build the program with a clinician [5]; the bar neither crushed these spines on the films [2] nor replaces the pharmacy [11] — it is the deposit side of an account the diagnosis opened.

How RightRep fits into this

The bone bank runs on one currency this app counts better than any notebook — the dose, session after session, for the years the deposit needs:

  • The dose is the medicine, and the dose is countable. The osteogenic file prices magnitude and rate, delivered brief and rested [8] — a camera that counts every rep at every load is the ledger that shows whether this month cleared the minimum deposit this year's bone work needs [1].
  • Form is the safety mechanism, graded every rep. In this diagnosis the neutral-spine hinge under the bar is not aesthetics, it is the fracture-prevention rule [4] — camera-scored form keeps the technique gate closed on exactly the reps where fatigue would fold it [1][2].
  • Progressive overload, governed. The LIFTMOR climb to >85% 1RM was supervised and step-wise [1]; logged load curves with reserve-based ceilings (RIR) are how that supervision survives the home gym.
  • The light session still counts. A rep counter does not sneer at the machine day or the single-leg balance work — and in this file those are co-equal medicine columns [5][7][12].
  • The decade needs a streak. The Mayo file's protection outlived its program [3]; the visible year-over-year streak is the treatment's schedule kept — the cheapest continuity device an 8-month-response, 10-year-payoff intervention could ask for [1][3].

The bottom line

  • The permission file: 18.3% global prevalence, 23.1% in women, 11.7% in men [10]; 8 months of supervised 2×30 min/week at 5×5 >85% 1RM through 101 low-to-very-low-bone-mass spines — lumbar BMD +2.9% vs −1.2% light control, femoral neck +0.3% vs −1.9%, one minor adverse event total [1].
  • The fear file: Genant-read follow-up films found no change in fracture classification under HiRIT, kyphosis improved 6.7°, and the trial's single new wedge deformity landed in the light-exercise group [2].
  • The vector file: 59 women, 1984 — new fractures in 89% of the flexion group against 16% of the extension group (p<0.001); the one hard rule is no loaded spinal flexion [4].
  • The decade file: 2 years of back-extension work, 8 years off-program — strength and BMD edges still significant at year ten and 2.7× fewer vertebral compression fractures [3].
  • The pooled files: Cochrane's 43 RCTs price combination programs at MD 3.22 spine and lower-limb resistance at MD 1.03 femoral neck [6]; the falls Cochrane cuts fall rates 23% across 108 RCTs / 23,407 participants (RaR 0.77, high certainty) [7].
  • The drug+exercise ledger: adding physical intervention to bisphosphonates bought femoral BMD MD 5.27 (2.23–8.32) and quality-of-life MD 30.16 (22.16–38.16) over drugs alone — heterogeneity honest [11].
  • The Brazil files: LAVOS's 11.18% vertebral-fracture prevalence across five LatAm countries including Brazil [9]; São Paulo's 100-woman RCT cutting falls to IRR 0.263 (0.10–0.68) on 18 weeks of quadriceps loading + proprioception on top of medication [12].
  • The dose: heavy, brief, rested, supervised-or-gated, extension-biased, balance-fed, run for a decade [1][3][5][8] — and metered by the log, because bone's scoreboard moves slower than the mirror.

Load the skeleton, never flex it under load — and let the camera count the deposits the diagnosis is trying to blur, at rightrep.app.

This post was written by the RightRep AI Assistant — the same AI that counts and scores your reps at rightrep.app.

Sources

  1. Watson SL, Weeks BK, Weis LJ, Harding AT, Horan SA, Beck BR. High-Intensity Resistance and Impact Training Improves Bone Mineral Density and Physical Function in Postmenopausal Women With Osteopenia and Osteoporosis: The LIFTMOR Randomized Controlled Trial. Journal of Bone and Mineral Research, 2018;33(2):211–220. pubmed.ncbi.nlm.nih.gov
  2. Watson SL, Weeks BK, Weis LJ, Harding AT, Horan SA, Beck BR. High-intensity exercise did not cause vertebral fractures and improves thoracic kyphosis in postmenopausal women with low to very low bone mass: the LIFTMOR trial. Osteoporosis International, 2019;30(5):957–964. pubmed.ncbi.nlm.nih.gov
  3. Sinaki M, Itoi E, Wahner HW, Wollan P, Gelzcer R, Mullan BP, Collins DA, Hodgson SF. Stronger back muscles reduce the incidence of vertebral fractures: a prospective 10 year follow-up of postmenopausal women. Bone, 2002;30(6):836–841. pubmed.ncbi.nlm.nih.gov
  4. Sinaki M, Mikkelsen BA. Postmenopausal spinal osteoporosis: flexion versus extension exercises. Archives of Physical Medicine and Rehabilitation, 1984;65(10):593–596. pubmed.ncbi.nlm.nih.gov
  5. Giangregorio LM, Papaioannou A, Macintyre NJ, Ashe MC, Heinonen A, Shipp K, Wark J, McGill S, Keller H, Jain R, Laprade J, Cheung AM. Too Fit To Fracture: exercise recommendations for individuals with osteoporosis or osteoporotic vertebral fracture. Osteoporosis International, 2014;25(3):821–835. pubmed.ncbi.nlm.nih.gov
  6. Howe TE, Shea B, Dawson LJ, Downie F, Murray A, Ross C, Harbour RT, Caldwell LM, Creed G. Exercise for preventing and treating osteoporosis in postmenopausal women. Cochrane Database of Systematic Reviews, 2011;(7):CD000333. pubmed.ncbi.nlm.nih.gov
  7. Sherrington C, Fairhall NJ, Wallbank GK, Tiedemann A, Michaleff ZA, Howard K, Clemson L, Hopewell S, Lamb SE. Exercise for preventing falls in older people living in the community. Cochrane Database of Systematic Reviews, 2019;1(1):CD012424. pubmed.ncbi.nlm.nih.gov
  8. Turner CH, Robling AG. Designing exercise regimens to increase bone strength. Exercise and Sport Science Reviews, 2003;31(1):45–50. pubmed.ncbi.nlm.nih.gov
  9. Clark P, Cons-Molina F, Deleze M, Ragi S, Haddock L, Zanchetta JR, Jaller JJ, Palermo L, Talavera JO, Messina DO, Morales-Torres J, Salmeron J, Navarrete A, Suarez E, Pérez CM, Cummings SR. The prevalence of radiographic vertebral fractures in Latin American countries: the Latin American Vertebral Osteoporosis Study (LAVOS). Osteoporosis International, 2009;20(2):275–282. pubmed.ncbi.nlm.nih.gov
  10. Salari N, Ghasemi H, Mohammadi L, Behzadi MH, Rabieenia E, Shohaimi S, Mohammadi M. The global prevalence of osteoporosis in the world: a comprehensive systematic review and meta-analysis. Journal of Orthopaedic Surgery and Research, 2021;16(1):609. pubmed.ncbi.nlm.nih.gov
  11. Jin J, Ma H. Synergistic Effects of Bisphosphonates and Physical Exercises and Modalities on Osteoporosis Management: A Systematic Review and Meta-Analysis. Journal of Aging and Physical Activity, 2026 Sep 29 (online ahead of print). pubmed.ncbi.nlm.nih.gov
  12. Teixeira LE, Silva KN, Imoto AM, Teixeira TJ, Kayo AH, Montenegro-Rodrigues R, Peccin MS, Trevisani VF. Progressive load training for the quadriceps muscle associated with proprioception exercises for the prevention of falls in postmenopausal women with osteoporosis: a randomized controlled trial. Osteoporosis International, 2010;21(4):589–596. pubmed.ncbi.nlm.nih.gov