What it is:A skeletal disease where bone resorption outpaces bone formation for long enough that the bone loses strength and density, making fracture the presenting problem instead of pain. Nearly a third of women and one in six men over 50 already meet diagnostic criteria.
The core problem:Bone is constantly being torn down by osteoclasts and rebuilt by osteoblasts. Resorption takes 2-3 weeks, formation takes 3-4 months. That mismatch in speed is why anything that tips the balance toward resorption (menopause, steroids, aging) wins the race and net bone is lost. Nobody feels this happening until a hip, spine, or wrist fracture announces it.
What you do about it:Everyone gets calcium, vitamin D, weight-bearing exercise, and fall prevention. Above a fracture-risk threshold (T-score, FRAX, or an actual fragility fracture), you add a drug. Almost every drug in this chapter is either an antiresorptive(blocks the osteoclast) or a bone-forming/anabolicagent (stimulates the osteoblast), and matching the drug to that mechanism is most of the exam.
Think of it as a seesaw between osteoclasts and osteoblasts. Antiresorptives (bisphosphonates, denosumab, SERMs, calcitonin, estrogen) sit on the osteoclast side and slow the losing side down. Anabolics (teriparatide, abaloparatide, romosozumab) push the osteoblast side up. Romosozumab is unique: it does both at once, which is why it's saved for the highest-risk patients and only used for a year.
Bone remodeling runs through 5 phases in a Bone Multicellular Unit: quiescence, activation/resorption, reversal, formation, and mineralization. Resorption is fast (2-3 weeks), formation is slow (3-4 months). Any process that accelerates the cycle produces more immature, poorly mineralized bone in circulation at any given time, and that immature bone is structurally weaker even before you factor in the density loss.
| Type | Mechanism | Typical fractures |
|---|---|---|
| Primary, Type 1 (sex-hormone deficiency) | Estrogen loss at menopause removes its brake on osteoclast activity, so resorption accelerates well past formation. In men, rising sex-hormone-binding globulin inactivates testosterone and estrogen with age. | Wrist, spine |
| Primary, Type 2 (age-related, >70) | Reduced osteoblast formation plus falling calcium/vitamin D intake and less exercise. Men lose bone this way too, just later and slower because they start with more peak bone mass and larger bones. | Hip, spine |
| Secondary / drug-induced | Corticosteroids, excess thyroid hormone, antiepileptics (phenytoin, phenobarbital), depot medroxyprogesterone, and diseases like CKD, hyperthyroidism, and malabsorption disorders. Can hit any age, either sex equally. | Any site |
Once you see resorption and formation as a race, the drug classes explain themselves. Bisphosphonates and denosumabslow the osteoclasts down directly. Estrogen and SERMsrestore the hormonal brake that menopause removed. Teriparatide and abaloparatidehijack PTH signaling to push osteoblasts into overdrive. Romosozumabblocks sclerostin, the molecule osteocytes use to put the brakes on Wnt-driven bone formation, so removing sclerostin's brake speeds up formation while also slowing resorption.
Men are naturally protected relative to women: larger bone size, higher peak bone mass, bones that widen with age instead of just thinning, fewer falls, and a shorter life expectancy. That's why male osteoporosis is disproportionately secondary (steroids, hypogonadism, alcohol) rather than purely age-related.
This disease is famous for being silent until it isn't. Most patients have zero idea they have low bone mass until they fracture, sometimes from something as unremarkable as bending over or lifting a bag of groceries.
| Fracture site | Typical picture |
|---|---|
| Vertebral (most common) | Can be silent, or present with moderate to severe back pain radiating down a leg that usually settles in 2-4 weeks. Multiple vertebral fractures compound: they lose height and cause kyphosis or lordosis over time. |
| Hip (proximal femur) | Severe pain, swelling, loss of function at the fracture site. This is the fracture that kills: roughly 30% one-year mortality after a fragility hip fracture. |
| Wrist (distal radius / Colles) | Usually from an outstretched-hand fall. Often the first fracture in a woman who goes on to fracture her hip or spine years later. |
Height loss and kyphosis are physical exam findings of established disease, not early warning signs. By the time you see >1.5 in (3.8 cm) of height loss or a dowager's hump, the patient has already sustained silent vertebral fractures. This is why screening happens on age and risk factors, not on symptoms.
Central (hip and spine) DXA is the gold standard and the only measurement that should drive therapy decisions. Peripheral DXA or quantitative ultrasound (forearm, heel, finger) is screening-only, it tells you who needs the real test, not who needs treatment.
The T-scorecompares the patient's BMD to a healthy 20-29 year old of the same sex (peak bone mass reference). It is standard deviations from that peak, not from an age-matched peer, which is exactly why it's usable for diagnosis: it measures how far someone has drifted from their theoretical best bone, not how they compare to equally-aged people who may also be losing bone. Z-score (age/gender matched control) is reserved for patients under 40, mainly to hunt for secondary causes.
| Category | T-score |
|---|---|
| Normal | Above -1.0 |
| Low bone mass (osteopenia) | -1.0 to -2.4 |
| Osteoporosis | -2.5 or below |
| Severe osteoporosis | -2.5 or below witha fragility fracture |
A low-trauma (fragility) fracture of the spine or hipis diagnostic of osteoporosis on its own, T-score irrelevant. Falling from standing height or less and breaking a hip means the patient has osteoporosis by definition, regardless of what a scan would show.
The WHO FRAX tool estimates 10-year probability of hip fracture and 10-year probability of any major osteoporotic fracture using age, sex, race, BMI, prior fracture, parental hip fracture, smoking, alcohol, glucocorticoid use, rheumatoid arthritis, and secondary causes, with femoral neck BMD as an optional input. It's validated for postmenopausal women and men ≥40, and only applies to untreatedpatients, so it's not a tool for reassessing someone already on therapy. The Garvan calculator is an alternative that adds falls and fracture count but uses fewer other risk factors.
CBC, SCr, calcium, phosphorus, electrolytes, alkaline phosphatase, albumin, TSH, total testosterone (men), 25-hydroxyvitamin D, and 24-hour urine calcium/phosphorus. The point of this workup is to catch a secondary cause (hyperparathyroidism, hyperthyroidism, malabsorption) before you commit someone to a bone drug for a problem that's actually endocrine.
Prevention is the real goal, maximizing peak bone mass while young heads off the whole problem. Once low bone mass or osteoporosis exists, the goals shift to stabilizing or improving bone mass, and once a fracture has happened, to reducing pain and deformity, restoring function, and preventing the next fracture.
The decision tree from the DiPiro algorithm and the ACP guideline lands on medication for:
Above that, risk stratifies further into osteoporosis, severe osteoporosis (T ≤ -3.0, or -2.5 to -2.9 with a new fracture), and very high/severe risk (any fragility fracture within the year, multiple fractures, fracture on treatment, high fall risk, or frailty). That stratification is what decides which classof drug you reach for first.
Standard risk:alendronate, risedronate, zoledronic acid, or denosumab, since these four are the only agents proven to cut both hip and vertebral fracture risk. Very high / severe risk:sequential therapy starting with an anabolic agent (abaloparatide, romosozumab, or teriparatide) first, followed by an antiresorptive, because starting antiresorptive-first and switching to an anabolic later gives worse BMD outcomes than the reverse order.
Vertebroplasty/kyphoplasty (cement injection into a fractured vertebra) sounds like a fix for compression fractures but the evidence only supports short-term benefit, with no major pain relief long-term and real procedural complication risk. Don't reach for it as a default.
| Group | Elemental Ca RDA | Ca upper limit | Vitamin D RDA | Vit D upper limit |
|---|---|---|---|---|
| Adults 19-50 | 1000 mg | 2500 mg | 600 units | 4000 units |
| Men 51-70 | 1000 mg | 2000 mg | 600 units | 4000 units |
| Women 51-70 | 1200 mg | 2000 mg | 600 units | 4000 units |
| >70 years | 1200 mg | 2000 mg | 800 units | 4000 units |
Guidelines actually recommend 800-1000 units, or in some cases 1000-2000 units, for adults who already have osteoporosis, higher than the bare RDA. Target a 25-hydroxyvitamin D concentration of at least 20-30 ng/mL, and recheck after about 3 months since vitamin D's half-life is roughly a month.
| Class / Drug | Dose | Comments |
|---|---|---|
| Bisphosphonates (fracture-reducing, hip + vertebral) | ||
| Alendronate (Fosamax) | Treatment 10 mg daily or 70 mg weekly; prevention 5 mg daily or 35 mg weekly | Cheapest agent in the class. Also comes combined with vitamin D3. |
| Risedronate (Actonel, Atelvia DR) | 5 mg daily, 35 mg weekly, or 150 mg monthly | Delayed-release Atelvia taken right after breakfast instead of fasting. |
| Zoledronic acid (Reclast) | Treatment 5 mg IV yearly; prevention 5 mg IV every 2 years | Contraindicated CrCl <35. Premedicate with acetaminophen for infusion reaction. |
| Bisphosphonate (vertebral only, not first-line) | ||
| Ibandronate (Boniva) | 150 mg orally monthly, or 3 mg IV quarterly | No hip fracture data, so it's the one you skip when a first-line agent is an option. |
| RANKL Inhibitor | ||
| Denosumab (Prolia) | 60 mg SC every 6 months | Given in office. Correct hypocalcemia first. Safe across renal function including CrCl 10-35. |
| Anabolic (bone-forming) agents | ||
| Teriparatide (Forteo) | 20 mcg SC daily, max 2 years lifetime | Refrigerate; discard pen after 28 days. |
| Abaloparatide (Tymlos) | 80 mcg SC daily, max 2 years lifetime | Less hypercalcemia than teriparatide (3.4% vs 6.4%). Discard after 30 days. |
| Formation + Antiresorptive | ||
| Romosozumab (Evenity) | 210 mg SC monthly (two 105 mg injections) for 12 doses total | Boxed warning: MI, stroke, CV death. Avoid within 1 year of either. |
| Estrogen agonist/antagonist and tissue-selective complex | ||
| Raloxifene (Evista) | 60 mg daily | Reduces vertebral, not hip, fractures. Also cuts invasive breast cancer risk. |
| Bazedoxifene/CEE (Duavee) | 20 mg / 0.45 mg daily | For women with a uterus; no added progestogen needed. |
| Last-line | ||
| Calcitonin-salmon (Fortical, Miacalcin) | 200 units (1 spray) intranasal daily, alternating nostrils | Vertebral fracture reduction only; weakest agent in the chapter. |
| Nutritional support (everyone) | ||
| Calcium carbonate | Max 600 mg elemental per single dose, take with meals | 40% elemental calcium, cheapest, needs stomach acid to absorb. |
| Calcium citrate | Max 600 mg elemental per single dose, any time | 21% elemental calcium, acid-independent absorption, fewer GI effects. |
| Cholecalciferol (D3) | 800-2000 units daily typical maintenance | OTC, preferred over D2 for raising 25-OH-D. |
Bisphosphonates are synthetic analogs of pyrophosphate, an endogenous inhibitor of bone resorption, but with a nonhydrolyzable backbone that resists breakdown. They bind hydroxyapatite in bone and get taken up by osteoclasts during resorption, where they poison osteoclast maturation, recruitment, and lifespan. Because they get incorporated directly into bone matrix, their biologic half-life runs up to 10 years, which is exactly why "drug holidays" are even a possible concept for this class: the drug keeps working long after the last dose.
BMD gains are dose-dependent and biggest in the first 12 months, and effects persist for a while after stopping, though duration varies by agent. Alendronate, risedronate, and IV zoledronic acid all reduce hip andvertebral fracture risk and are FDA-approved for postmenopausal, male, and glucocorticoid-induced osteoporosis. Ibandronate only has spine data, which is why it's not first-line.
Oral bisphosphonates must be taken first thing in the morning, fasting, with 6-8 oz of plain water(not coffee, juice, mineral water, or milk), then nothing by mouth (food, supplements, or other meds) for at least 30 minutes (60 for ibandronate). Stay upright(sitting or standing) for that same window to prevent esophageal irritation and ulceration. The one exception is delayed-release risedronate (Atelvia), taken immediately after breakfast. Get this wrong and you get chemical esophagitis, not better bones.
Missed dose rule:weekly dose missed by less than a day, take it the next day; more than a day gone, skip it entirely. Monthly dose can be taken up to 7 days before the next scheduled dose.
Who shouldn't get an oral bisphosphonate:anyone who can't stay upright 30+ minutes, has an esophageal or swallowing disorder, or won't follow the instructions. Neither oral norinjectable bisphosphonates should go to a patient with hypocalcemia or renal insufficiency (CrCl <30-35 depending on agent).
Adverse effects:nausea, abdominal pain, dyspepsia, esophageal/gastric irritation or ulceration with oral agents; fever and flu-like symptoms with IV. Rare but feared: osteonecrosis of the jaw (ONJ, mostly seen with high-dose IV bisphosphonates in cancer patients, not typical osteoporosis dosing) and atypical subtrochanteric femoral fractures.
Drug holidays:considered after 5 years of oral or 3 years of IV bisphosphonates in patients without a significant fracture history, hip T-score above -2.5, and fracture risk not high. Very-high-risk patients instead continue to 10 years oral / 6 years IV. Restart if BMD drops below T ≤ -2.5 or a fracture occurs during the holiday.
Denosumab is a monoclonal antibody against RANK ligand, the signal osteoblasts use to activate osteoclasts. Block RANKL and you inhibit osteoclast formation, function, and survival all at once, more potent resorption suppression than bisphosphonates achieve (roughly 90% versus 60-70%). Unlike bisphosphonates, it isn't incorporated into bone, so its effect reverses relatively quickly after stopping, which is exactly why discontinuation needs a plan.
It's FDA-approved for postmenopausal and male osteoporosis at high fracture risk, glucocorticoid-induced osteoporosis, men on androgen-deprivation therapy for prostate cancer, and women on aromatase inhibitors for breast cancer. It's the go-to when bisphosphonates are off the table for renal reasons: no dose adjustment needed for renal impairment, and it can be used down to CrCl 10-35, though stage 5 CKD/dialysis patients need extra caution for hypocalcemia.
Stopping denosumab causes a rapid rebound in bone turnoverand can produce fast BMD loss, including multiple vertebral fractures in a short window. This is the opposite of the bisphosphonate holiday concept: with denosumab, you plan the next drug (usually a bisphosphonate) beforeyou stop, you don't just walk away.
Adverse effects:hypocalcemia (correct before dosing), serious infections (UTI, cellulitis, endocarditis), dermatologic reactions, and rare ONJ/atypical femoral fracture. Most common: back pain, musculoskeletal pain, hypercholesterolemia.
Teriparatide is recombinant human PTH(1-34); abaloparatide is a PTHrP analog. Endogenous PTH, given continuously, actually increasesresorption, that's the paradox: given once daily as an intermittent pulse instead, it preferentially drives osteoblast activity, a net anabolic effect. Abaloparatide has a greater anabolic-to-resorptive ratio than teriparatide, meaning less activation of the resorption side for a similar bone-building effect.
Both cut vertebral and nonvertebral fracture risk over 2 years in postmenopausal women; teriparatide has observational data suggesting similar benefit in men, abaloparatide doesn't have male data. Neither has been shown to specifically reduce hip fractures.
Anabolic-then-antiresorptive beats antiresorptive-then-anabolic. Starting teriparatide aftera patient has already been on an antiresorptive gives a smaller BMD gain than starting teriparatide first. That's why severe/very-high-risk patients get the anabolic agent up front, then transition to a bisphosphonate or denosumab to lock in the gains, since stopping a PTH analog without a following antiresorptive lets BMD drift back down.
Limits:both are capped at 2 years lifetimeuse, driven by a boxed warning for osteosarcoma seen in rat studies (not confirmed in humans, but the caution stands). Avoid in hypercalcemia, Paget's disease, prior skeletal radiation, unexplained elevated alkaline phosphatase, metastatic or skeletal cancers, and premenopausal women who could become pregnant. Adverse effects: nausea, dizziness, leg cramps, transient orthostatic hypotension, and hypercalcemia (less common with abaloparatide).
Romosozumab targets sclerostin, a protein osteocytes normally release to put the brakes on Wnt/β-catenin signaling, the pathway that drives bone formation. Neutralize sclerostin and Wnt signaling proceeds unopposed: bone formation goes up andresorption goes down at the same time, a genuinely different mechanism from anything else in the chapter, and part of why it can outperform teriparatide on BMD gains at the hip and spine.
After 1 year, vertebral fractures dropped 73% in trial data with a smaller, non-significant drop in nonvertebral fractures. It's approved for postmenopausal women at high fracture risk (multiple risk factors, prior osteoporotic fracture, or failure/intolerance of other therapies); data in men is more limited but reassuring.
Avoid in anyone with an MI or stroke in the past year, and weigh benefit versus risk carefully in anyone else with cardiovascular disease. Correct hypocalcemia before starting.
Duration and follow-through:limited to 12 monthly doses (12 months total). After that year, transition to an antiresorptive, denosumab or alendronate, to preserve the BMD gains; stopping without a follow-on antiresorptive lets the gains erode just like with the PTH analogs.
Estrogen(with or without a progestogen if the uterus is intact) reliably reduces fracture risk and bone loss, dose-dependently, but it is FDA-approved only for preventionof postmenopausal osteoporosis, not treatment, and carries endometrial cancer, breast cancer, thromboembolism, and cardiovascular risk. Reserve it for women near the menopause transition who also need treatment for vasomotor symptoms; if a patient is already on estrogen for hot flashes, they get an incidental bone benefit, but you wouldn't start it for bone alone. Benefit disappears once estrogen stops, and bone loss accelerates back to baseline.
Raloxifeneis a SERM: an estrogen agonist at bone (helps BMD) but antagonist at breast tissue (reduces invasive breast cancer risk), with minimal uterine effect so no progestogen is needed. It only reduces vertebral fractures, not hip, and the benefit vanishes on discontinuation. Bazedoxifene/CEE (Duavee)pairs a SERM with estrogen: agonist at bone, antagonist at both uterus and breast, so no progestogen needed, though breast cancer risk reduction hasn't been shown the way it has for raloxifene alone.
Calcitonininhibits osteoclast activity directly but is the weakest agent here, only shown to reduce vertebral fractures, with benefit lost 1-2 years after stopping. It's reserved for women at least 5 years post-menopause who can't use anything more effective, and its one other niche is short-term (≤4 weeks) pain relief after an acute vertebral fracture via calcitonin-stimulated endorphin release, not as a substitute for definitive osteoporosis therapy.
Steroids attack bone from every angle at once: they decrease osteoblast proliferation and differentiation while increasing osteoblast apoptosis, increase the number and activity of osteoclasts, decrease intestinal calcium absorption, and increase renal calcium excretion. That's resorption up, formation down, and less calcium available to compensate, all simultaneously.
Bone loss is fastest in the first 3-6 monthsof glucocorticoid therapy (up to 12-15% in year one), then continues at a slower 2-3%/year afterward. This is why baseline BMD assessment happens early, prior to or within 6 months of starting steroids, not a year or two down the road once the damage is already done.
Oral bisphosphonates (alendronate, risedronate) are first-line for glucocorticoid-induced osteoporosis; IV zoledronic acid is the alternative for nonadherent patients or those who can't take oral. Teriparatide is next if bisphosphonates can't be used, denosumab after that (not first-line here due to more limited safety data in this population). Typical durations: oral bisphosphonate 5 years, IV bisphosphonate 3-5 years, denosumab 5 years, teriparatide 2 years, same cap as elsewhere.
| Parameter | When | Watching for |
|---|---|---|
| Central DXA (hip/spine) | 1-2 years after starting therapy, then every 2 years until stable, then can space out further | Response to treatment. More frequent (every 1-2 years) if on chronic glucocorticoids or other high-turnover risk factors |
| Fracture symptoms | Every visit | New back pain, height loss, disability signaling a new or worsening fracture |
| Serum calcium | Before denosumab or romosozumab dosing; periodically on bisphosphonates | Hypocalcemia, must correct before RANKL inhibitor or sclerostin inhibitor dosing |
| Renal function (SCr, CrCl) | Before starting and periodically on IV bisphosphonates | CrCl <35 (zoledronic acid) or <30-35 (other bisphosphonates) is a contraindication |
| 25-hydroxyvitamin D | About 3 months after starting or adjusting supplementation | Target 20-30+ ng/mL; half-life is ~1 month so earlier recheck is too soon |
| Medication adherence | Every visit | Bisphosphonate prescribing has fallen over 50% since 2007 largely from adherence and fear of rare adverse events; nonadherence is the most fixable failure point |