Osteoporosis
The most common bone disease, and the most undertreated: fewer than a quarter of hip-fracture patients ever get therapy for the disease that broke the hip. A fragility fracture is to osteoporosis what an MI is to coronary disease. Screen by age and risk, treat by risk tier, and never stop denosumab without a plan.
🔍 Overview
Overview
Osteoporosis is skeletal fragility from loss of bone mass and microarchitecture: bone resorption (osteoclasts, driven by RANKL) outpaces formation (osteoblasts), accelerating sharply after menopause as estrogen's brake on RANKL is lost. It is silent until it breaks something. About 1 in 2 women and up to 1 in 4 men over 50 will suffer a fragility fracture, and a hip fracture carries 20-30% one-year mortality with half of survivors losing independent ambulation. That is why the disease belongs to internists, not just endocrinologists.
The Diagnosis (any ONE of these)
- T-score ≤ -2.5 at the femoral neck, total hip, or lumbar spine on DEXA. The lowest site governs. Osteopenia is T-score -1.0 to -2.5.
- A fragility fracture of the hip or vertebra, regardless of the T-score -a fracture from standing height or less IS the diagnosis, because the bone has already proven it fails under physiologic load. A normal DEXA does not un-diagnose it.
- Osteopenia plus elevated FRAX risk (10-year hip fracture probability ≥ 3% or major osteoporotic fracture ≥ 20%, the US treatment thresholds).
T-Score vs Z-Score (a distinction that changes the workup)
- T-score compares to a healthy 30-year-old and defines osteoporosis in postmenopausal women and men ≥ 50.
- Z-score compares to age-matched peers and is the right scale for premenopausal women and men < 50. A Z-score ≤ -2.0 is "below expected range for age" and demands a hunt for secondary causes, because young people should not have old bones without a reason.
Risk Factors and Secondary Causes
- Non-modifiable: age, female sex, prior fragility fracture (the strongest predictor of the next one), parental hip fracture, low body weight.
- Drugs (review the list before blaming biology): glucocorticoids (≥ 2.5 mg prednisone daily for ≥ 3 months erodes bone fastest in the first year), aromatase inhibitors, androgen deprivation therapy, long-term PPIs, anticonvulsants, SSRIs, thiazolidinediones, excess thyroid hormone.
- Diseases: hyperparathyroidism, hyperthyroidism, celiac disease and other malabsorption, CKD, multiple myeloma, hypogonadism, anorexia nervosa, rheumatoid arthritis, alcohol overuse, smoking.
The treatment gap is the headline problem. After a hip fracture, fewer than 25% of patients are started on osteoporosis therapy, though the next fracture is precisely what treatment prevents (subsequent fracture risk roughly doubles after the first). Discharging a hip-fracture patient without a bone plan is the osteoporosis equivalent of discharging an MI without aspirin or a statin.
🧪 Screening & Workup
Who Gets Screened (USPSTF, January 2025)
- All women ≥ 65: DEXA, Grade B. Screening exists because the first symptom of osteoporosis is often a broken hip, and treating what you find prevents fractures.
- Postmenopausal women < 65 with at least one risk factor: run a clinical risk tool (FRAX without BMD is acceptable) and screen if elevated (Grade B).
- Men: evidence rated insufficient by USPSTF, but specialty societies support DEXA for men with a fragility fracture, glucocorticoid use, androgen deprivation therapy, or hypogonadism -the disease is underdiagnosed in men and their post-hip-fracture mortality is higher than women's.
- Anyone with a fragility fracture, regardless of age: that is not screening, that is established disease -work up and treat.
- Repeat interval: if treating or T-score is close to threshold, repeat DEXA in 1-2 years on the same machine -inter-machine precision error is large enough to fake a change that never happened.
FRAX (what it adds and where it misleads)
- FRAX turns clinical risk factors ± femoral neck BMD into 10-year probabilities of hip and major osteoporotic fracture. Its job is to find the osteopenic patient who still warrants treatment: hip ≥ 3% or major osteoporotic ≥ 20% crosses the US treatment threshold.
- It underestimates risk with recent fractures, multiple vertebral fractures, high-dose steroids, and falls -none are fully captured, so do not let a reassuring FRAX overrule a fracturing skeleton.
Vertebral Imaging (the fractures nobody reports)
- Two-thirds of vertebral fractures are silent, yet each one multiplies future fracture risk and can reclassify a patient to very-high-risk (changing drug choice). Get lateral spine imaging or DEXA-based vertebral fracture assessment when there is historical height loss > 1.5 inches, prospective loss > 0.8 inches, new kyphosis, unexplained back pain, or long-term glucocorticoids.
Secondary-Cause Labs (before or with starting therapy)
- CBC (cytopenias hint at myeloma or malabsorption), CMP with calcium and creatinine (hypercalcemia points to hyperparathyroidism, and renal function chooses between bisphosphonate and denosumab), 25-OH vitamin D (replete before antiresorptives or you provoke hypocalcemia), TSH, testosterone in men.
- Targeted when suspected: PTH, celiac serologies (tTG-IgA), SPEP/free light chains (unexplained anemia, high ESR, lytic lesions, or Z-score ≤ -2.0), 24-hour urine calcium (hypercalciuria and malabsorption both show up here), cortisol testing if Cushingoid.
🚨 Management
Foundation for Everyone (drugs fail without it)
- Calcium 1,000-1,200 mg/day, food first -dietary calcium achieves the same skeletal goal without the possible stone and cardiovascular signals attached to high-dose supplements; supplement only the dietary shortfall.
- Vitamin D 800-1,000 IU/day targeting 25-OH ≥ 30 ng/mL -antiresorptives given to a vitamin D-deficient patient cause hypocalcemia, which is why you replete first.
- Weight-bearing and resistance exercise plus balance training -bone responds to load, and balance work is the part that prevents the fall that breaks the hip. Exercise is the single strongest non-drug fall intervention.
- Deprescribe fall-makers and bone-thieves: sedatives, orthostatic agents, and the drug list above. Stop smoking, keep alcohol under 2 drinks/day.
Who Gets Pharmacotherapy
- T-score ≤ -2.5, OR a hip/vertebral fragility fracture at any T-score, OR osteopenia with FRAX above threshold (hip ≥ 3% / MOF ≥ 20%).
- Glucocorticoid users (≥ 2.5 mg prednisone ≥ 3 months): assess everyone; most patients on ≥ 7.5 mg with moderate or higher fracture risk warrant an oral bisphosphonate per ACR guidance, because steroid bone loss is fastest in the first 3-6 months -waiting for the DEXA to look bad wastes the window.
Choose by Risk Tier (the sequencing insight)
- High risk (most patients): oral bisphosphonate first-line -alendronate or risedronate. Decades of fracture data, generic cost, and a safety profile dominated by rare events. Zoledronic acid IV yearly if the gut cannot tolerate or adherence is failing.
- Very high risk (T ≤ -3.0 with fracture, multiple or recent vertebral fractures, fracture on therapy): anabolic FIRST, then antiresorptive. Teriparatide/abaloparatide for 18-24 months or romosozumab for 12, then lock in the gains with a bisphosphonate or denosumab. Why the order matters: starting the antiresorptive first blunts the anabolic response, and stopping an anabolic without consolidation lets the new bone drain away -build first, then lock.
- Denosumab when bisphosphonates are contraindicated (CrCl < 30-35) or failing -with the standing rule that it is a commitment, not a course (below).
- Raloxifene is niche: vertebral protection only, adds VTE risk, subtracts breast cancer risk -occasionally the right trade for a younger postmenopausal woman with high breast-cancer and low hip-fracture risk.
Duration, Holidays, and the Denosumab Rule
- Bisphosphonate holiday: after 5 years oral (3 years IV), reassess. If risk is now low-moderate, pause -bisphosphonate stays bound in bone for years, so protection persists while the (duration-driven) risks of atypical femoral fracture and ONJ recede. High-risk patients continue to 10 years oral / 6 IV.
- Keep the rare risks in proportion: ONJ occurs in roughly 1 in 10,000 to 100,000 patient-years and atypical femoral fracture is similarly rare, while the drug prevents fractures in 1 of every 10-20 treated -patients stopping therapy over these headlines trade a rare risk for a common one. New thigh or groin ache on long-term therapy gets an X-ray (AFF prodrome).
- NEVER simply stop denosumab. RANKL inhibition wears off in ~7 months and the rebound osteoclast surge can cause multiple spontaneous vertebral fractures. Every missed dose is a hazard; discontinuation requires a transition to a bisphosphonate. There is no denosumab "holiday."
The Hip-Fracture Inpatient (where the treatment gap closes or doesn't)
- Start therapy before discharge, or hand off a dated plan. "Too acute to treat" becomes "never treated" -that is how the < 25% number happens. Ortho-geriatric comanagement and fracture liaison services exist precisely to close this loop.
- Zoledronic acid within 90 days after hip-fracture repair is evidence-backed: in HORIZON-RFT, 2007 it cut new clinical fractures by ~35% and all-cause mortality by 28%. Replete vitamin D first (hypocalcemia risk), check CrCl ≥ 35, and expect a flu-like acute-phase reaction after the first infusion (warn the patient; acetaminophen blunts it).
- Acute vertebral fracture: analgesia, early mobilization (bed rest accelerates bone loss), short-term calcitonin only as an analgesic bridge, and treat the osteoporosis -the fracture is the indication, no DEXA needed.
- On-therapy fracture or falling BMD: first question is adherence (half of oral bisphosphonate patients stop by 1 year), second is missed secondary causes, third is escalation to anabolic-first sequencing.
💊 Medications
Medications
| Drug | Role | Dose | Why / Watch For |
|---|---|---|---|
| Alendronate (Fosamax) FIRST-LINE | Oral bisphosphonate | 70 mg weekly | Empty stomach, full glass of water, upright 30 min: oral absorption is under 1% and food abolishes it; the upright rule prevents pill esophagitis. Avoid CrCl < 35. |
| Risedronate (Actonel) | Oral bisphosphonate | 35 mg weekly | Same rules as alendronate; slightly gentler on the esophagus. Avoid CrCl < 30. |
| Zoledronic acid (Reclast) | IV bisphosphonate | 5 mg IV once yearly | Solves adherence and GI intolerance in one infusion. Acute-phase reaction after first dose (fever, myalgia, 1-3 days). Replete vitamin D first; avoid CrCl < 35. The post-hip-fracture agent (HORIZON-RFT mortality benefit). |
| Denosumab (Prolia) NO ABRUPT STOP | RANKL inhibitor | 60 mg SC every 6 months | Not renally cleared, so usable in advanced CKD, but watch hypocalcemia there. Effect fully reverses in ~7 months; stopping without a bisphosphonate transition risks rebound multiple vertebral fractures. |
| Teriparatide (Forteo) | Anabolic (PTH analog) | 20 mcg SC daily, 18-24 months | For very-high-risk. Intermittent PTH builds bone (continuous PTH destroys it -the paradox worth knowing). Follow with an antiresorptive or gains drain away. Avoid in Paget disease, prior skeletal radiation, unexplained high ALP. |
| Abaloparatide (Tymlos) | Anabolic (PTHrP analog) | 80 mcg SC daily, 18-24 months | Same class logic as teriparatide with less hypercalcemia. |
| Romosozumab (Evenity) | Anabolic (sclerostin inhibitor) | 210 mg SC monthly, 12 months | Builds AND blocks resorption; beat alendronate head-to-head (ARCH, 2017). Boxed warning: avoid within 1 year of MI or stroke. Follow with an antiresorptive. |
| Raloxifene (Evista) | SERM | 60 mg daily | Vertebral fractures only; raises VTE risk, lowers ER+ breast cancer risk. Niche trade-off drug. |
| Calcium + Vitamin D | Foundation | Ca 1,000-1,200 mg/day (diet first), D 800-1,000 IU/day | The substrate every drug above depends on; antiresorptives without vitamin D cause hypocalcemia. Target 25-OH D ≥ 30. |
| Calcitonin | Acute vertebral fracture analgesia | 200 IU intranasal daily, short course | Modest analgesic effect only; not a long-term osteoporosis drug (weak antifracture data, possible malignancy signal with years of use). |
The denosumab discontinuation trap is the most dangerous prescription decision in this disease. A patient who misses doses or whose refills lapse (hospitalization, insurance change, moving) is exposed to rebound vertebral fractures within months. If denosumab is stopped for any reason, bridge with a bisphosphonate. Flag it on the med list like an anticoagulant.
🏥 Rounds
Pimp Questions
❓ Why can't denosumab simply be stopped the way a bisphosphonate can?
Bisphosphonate binds hydroxyapatite and persists in bone for years after stopping, tapering its own effect -which is what makes a drug holiday possible. Denosumab is an antibody with no skeletal residence: RANKL inhibition vanishes in ~7 months and the suppressed osteoclast pool rebounds above baseline, producing rapid bone loss and multiple spontaneous vertebral fractures. Discontinuation therefore requires a bisphosphonate transition, never an unplanned stop.
❓ Why is the anabolic given BEFORE the antiresorptive in very-high-risk patients, and not the reverse?
Two reasons. Prior antiresorptive therapy blunts the anabolic response (teriparatide works partly through remodeling, which the bisphosphonate has suppressed). And anabolic gains are transient unless consolidated: stop teriparatide or romosozumab without follow-on antiresorptive and the new bone is resorbed within a year or two. So the sequence is build first, lock second -and trials of the reverse order showed inferior BMD trajectories.
❓ A patient with a hip fracture has a T-score of -1.8. Does the "osteopenic" DEXA change the diagnosis or the treatment?
No. A fragility fracture of the hip or vertebra IS osteoporosis regardless of BMD -the bone has demonstrated mechanical failure under physiologic load, which is more informative than any densitometric estimate. The patient warrants pharmacotherapy (and counts as very-high-risk if the fracture is recent). Waiting for the T-score to reach -2.5 is how the treatment gap happens.
❓ Why does teriparatide build bone when hyperparathyroidism destroys it -same hormone?
Kinetics, not chemistry. Continuous PTH exposure (hyperparathyroidism) keeps RANKL-driven resorption switched on and catabolizes bone. Intermittent once-daily pulses preferentially stimulate osteoblasts before the resorptive machinery ramps up, so formation outruns resorption and bone accrues. Same receptor, opposite outcome, decided entirely by the exposure pattern.
❓ When is a bisphosphonate drug holiday appropriate, and why is it safe?
After 5 years of oral (3 years of IV) therapy in a patient whose risk is now low-to-moderate: no recent fractures, T-score improved above roughly -2.5. It is safe because bisphosphonate remains bound in the skeleton and keeps releasing for years, so antifracture protection persists while the duration-linked risks (atypical femoral fracture, ONJ) decline. High-risk patients skip the holiday and continue to 10 years oral / 6 IV. The holiday logic applies ONLY to bisphosphonates -denosumab has no skeletal depot.
Sample Presentation
Ms. Okafor is a 78-year-old woman admitted after a fall from standing with a left femoral neck fracture, now post-op day 2 from hemiarthroplasty. History: hypertension, GERD on omeprazole for 8 years, no prior DEXA, no prior fractures documented, though she reports losing 2 inches of height. Labs: Ca 9.1, Cr 0.9 (CrCl ~52), 25-OH vitamin D 16 ng/mL, TSH normal, CBC normal. She has never been on osteoporosis therapy.
Key Points: (1) The fracture makes the diagnosis -no DEXA needed to justify treatment, and the height loss warrants lateral spine imaging for silent vertebral fractures that would confirm very-high-risk status. (2) Replete vitamin D now (deficient at 16), then zoledronic acid 5 mg IV before or shortly after discharge, within 90 days of repair -HORIZON-RFT showed fewer refractures and a 28% mortality reduction in exactly this population, and an infusion solves the adherence problem her 8-year PPI course hints at. (3) Close the loop: calcium/D, PT for balance work, medication review for fall-makers, and a discharge summary line that names the diagnosis "osteoporosis with fragility fracture," because unnamed diseases don't get followed up.
Clinic & Discharge Checklist
- Name the diagnosis in the chart after any fragility fracture -"mechanical fall with fracture" buries the disease that caused it.
- Vitamin D repleted before the antiresorptive? Deficiency plus a potent antiresorptive is the recipe for symptomatic hypocalcemia.
- Renal function checked? CrCl < 30-35 removes bisphosphonates and points to denosumab (with the stop-rule documented).
- Denosumab patients: next-dose date visible and a bridge plan if it will lapse. Treat it like an anticoagulant on the med list.
- DEXA follow-up ordered (1-2 years, same machine) and FRAX documented for the osteopenic.
- Falls addressed: orthostatics checked, sedatives deprescribed, PT/balance referral -the drug protects the bone, but not from the floor.
- Steroid patients: anyone on ≥ 2.5 mg prednisone ≥ 3 months has been risk-assessed, not just noted.
📋 Summary
Summary
A Fracture IS the Diagnosis
Hip or vertebral fragility fracture = osteoporosis at any T-score, because mechanical failure under physiologic load outranks densitometry. The T-score criteria (≤ -2.5) and FRAX thresholds (hip ≥ 3%, MOF ≥ 20% in osteopenia) exist to catch the disease before it declares itself. Subsequent fracture risk roughly doubles after the first, which is why the first one is the treatment trigger.
Screen by the 2025 Rules
All women 65 and older get DEXA (USPSTF Grade B, January 2025); younger postmenopausal women screen if a risk tool says so. Men lack a USPSTF mandate but fracture, steroids, ADT, and hypogonadism justify scanning them anyway, since their post-hip-fracture mortality exceeds women's. Repeat scans go on the same machine or the "change" may be machine noise.
Hunt Secondary Causes When the Story Is Wrong
Z-score ≤ -2.0, fractures on therapy, or bone loss in the young demand a cause: hyperparathyroidism, celiac disease, myeloma, hypogonadism, hyperthyroidism, and the drug list (steroids, PPIs, aromatase inhibitors, ADT, anticonvulsants). Check 25-OH vitamin D, calcium, creatinine, TSH in everyone before treating -partly for causes, partly because antiresorptives in a vitamin D-deficient patient cause hypocalcemia.
Foundation First, Then Match Drug to Risk
Calcium 1,000-1,200 mg (diet first), vitamin D 800-1,000 IU, resistance plus balance exercise, deprescribe fall-makers. Then: oral bisphosphonate for most; anabolic-first for the very-high-risk (T ≤ -3.0 with fracture, multiple/recent vertebral fractures) because building then locking beats locking then trying to build.
Bisphosphonate Rules That Prevent the Common Failures
Empty stomach, water only, upright 30 minutes -absorption is under 1% and food erases it, so "taking it with breakfast" mimics treatment failure. Avoid below CrCl 30-35. After 5 years oral / 3 IV, reassess for a holiday, which is safe only because the drug lives on in bone. Keep AFF and ONJ in proportion: rare, duration-linked, and vastly outnumbered by fractures prevented -but X-ray the thigh that aches.
Denosumab Is a Commitment
Every 6 months, forever or until a planned bisphosphonate transition. Its effect disappears in ~7 months and the osteoclast rebound causes multiple spontaneous vertebral fractures. Useful in CKD where bisphosphonates are barred (watch hypocalcemia), but a lapsed refill is a patient-safety event, not a missed appointment.
The Hip-Fracture Admission Is the Intervention Point
Zoledronic acid within 90 days of repair cut refractures ~35% and mortality 28% in HORIZON-RFT -few drugs on the wards carry a mortality benefit this clean. Replete vitamin D, check CrCl, warn about the first-infusion flu-like reaction, and start before discharge or hand off a dated plan. Fewer than a quarter of these patients ever get treated; be the reason yours does.
Anabolics: Build, Then Lock
Teriparatide/abaloparatide 18-24 months or romosozumab 12 months, always followed by an antiresorptive, because unconsolidated gains drain away. Intermittent PTH builds bone while continuous PTH destroys it (the kinetics paradox). Romosozumab carries a boxed warning: not within a year of MI or stroke. Antiresorptive-first blunts the anabolic, which is why sequence, not just selection, is the decision.
