Stopping Osteoporosis Medication: What the Trials Show, Class by Class

Why is "drug holiday" the wrong phrase for most of these drugs?

The term gets used as though every osteoporosis medicine behaves the same way when you stop it. They do not, and the differences are large enough to change how the decision should be made.

What happens after your last dose depends entirely on which class you were on. One class leaves a reservoir in the skeleton. One rebounds above baseline. One loses everything it built unless something follows. Treating them interchangeably is how avoidable fractures happen.

Bisphosphonates: what does the trial evidence actually show?

Alendronate, risedronate, ibandronate and zoledronic acid bind to the mineral surface of bone and stay there. When you stop, the drug already in the skeleton keeps exerting an effect that fades over months to years rather than switching off.

Two extension trials define the evidence base.

FLEX randomised women who had taken alendronate for five years either to continue to ten years or to switch to placebo. Clinically recognised vertebral fractures occurred in 5.3% of the placebo group against 2.4% of those who continued. There was no significant difference in morphometric vertebral fractures, and no difference in non-vertebral fractures. Women with a hip T-score between −2 and −2.5 benefited most from continuing.

HORIZON-PFT Extension did the same for zoledronic acid, comparing six annual infusions with three followed by placebo. Six years produced a greater reduction in morphometric vertebral fractures; the difference in clinical fractures was not statistically significant.

Read together, these say something more nuanced than "you can stop after five years." Bone loss after discontinuation was modest compared with continued treatment, supporting a genuinely persistent effect. But women at high risk of vertebral fracture specifically do better continuing.

Denosumab: why is there no holiday at all?

Denosumab does not bind to bone mineral. It clears, and when it does, bone resorption rebounds above baseline. Density gained is lost within roughly 6 to 12 months, and spontaneous vertebral fractures, frequently multiple, have been reported in women who stopped without follow-on treatment.

So there is no denosumab drug holiday. Coming off it requires a planned handover to another agent, normally a bisphosphonate, and the evidence shows bridging reduces but does not entirely eliminate the loss. Raloxifene has been tested for this purpose specifically and does not work. Our full page on denosumab covers the mechanism and the transition.

Anabolic drugs: what is lost if nothing follows?

Teriparatide and abaloparatide build bone rather than only slowing its breakdown, and they are given for a limited course. The gains are lost if nothing follows, because the newly built bone is not protected.

Standard practice is to move to an antiresorptive afterwards to lock in what was built. Finishing the course and stopping is not a neutral option, it is the option that wastes the treatment. Romosozumab sits in a related position and carries its own cardiovascular boxed warning, which is why cardiovascular history is weighed before it is started.

Which class are you on?

Hormone therapy is worth a note. It protects bone while taken and stops protecting when stopped, without a rebound overshoot. That is one reason the bone benefit alone is not usually a reason to continue indefinitely, and why the conversation about hormone therapy covers more than the skeleton.

Who should not pause?

A pause is a tool for the lower-risk end of the spectrum. The groups usually kept on treatment are:

• Anyone who has already had a fragility fracture, meaning a break from a fall at standing height or less
• Very low bone density, particularly a hip T-score below −2.5
• Ongoing glucocorticoid treatment
• Continuing high falls risk
• Women on aromatase inhibitors for breast cancer

The rationale for pausing at all is the small long-term risks that accumulate with continuous antiresorptive use, atypical femoral fracture and osteonecrosis of the jaw. Both are rare and both are more associated with longer duration. In a woman at high fracture risk those risks remain much smaller than the fracture being prevented, which is the whole reason high-risk women continue.

What does a well-run pause look like?

A pause without monitoring is not a pause, it is simply stopping. A reasonable structure has four parts:

• A documented reason. How long you were treated, what your density did, whether you fractured.
• A monitoring interval. Bone density repeated at a stated point, commonly around two years, sooner if risk is higher. See bone density testing.
• Height measured at each review. Loss of height can be the first sign of a vertebral fracture that caused no pain at all.
• Defined restart triggers, agreed in advance rather than improvised later.

Restart triggers worth writing down: any new fragility fracture, a meaningful fall in density on repeat testing, starting a medicine that accelerates bone loss, or height loss and new back pain suggesting a vertebral fracture.

What continues during a pause?

Everything that is not the prescription. Calcium and vitamin D intake, resistance and weight-bearing exercise, falls prevention, and the alcohol and smoking conversation all carry on unchanged.

This matters because a pause is often heard as "we are done with your bones." Our guides to exercise for bone density, calcium-rich foods and vitamin D cover the parts that do not stop. If your osteopenia was managed without drugs in the first place, those are the whole intervention.

What should you ask?

• "Which class am I on, and does it leave a reservoir in bone?" The answer determines everything else.
• "If we pause, when do we retest, and what result would make us restart?"
• "If I am on denosumab, what exactly is the handover plan and its timing?"
• "Given my fracture history and T-score, am I in the group that should continue?"
• "What am I still supposed to be doing while I am off the drug?"

How long should a pause last?

There is no fixed number, and anyone quoting one is going beyond the evidence. The trials tested continuing versus stopping, not how long a pause should run before restarting.

In practice, pauses of two to three years are common for oral bisphosphonates and are guided by repeat testing rather than by the calendar. The reservoir effect fades gradually rather than expiring on a date, so the question is not "has enough time passed" but "has anything changed."

What changes the answer: a new fracture, a meaningful fall in density, a new medicine that accelerates bone loss, a new diagnosis, or significant unintended weight loss. Any of those ends the pause regardless of how long it has run.

What about the risks that motivate pausing in the first place?

Two rare complications drive the entire drug holiday concept, and both deserve their actual scale rather than the vague dread they usually generate.

Atypical femoral fracture is a stress fracture of the thigh bone with a characteristic appearance. It is associated with longer antiresorptive use. The warning sign is new thigh or groin pain that builds over weeks rather than appearing suddenly, and reporting it early is what prevents a complete fracture.

Osteonecrosis of the jaw is associated with invasive dental procedures during treatment. It is markedly rarer at osteoporosis doses than at the much higher doses used in cancer care, which is a distinction frequently lost when the risk is described.

Both are rare. In a woman at genuinely high fracture risk, the arithmetic favours continuing treatment, because hip fracture carries consequences that these complications do not approach in frequency. The pause exists for women whose risk has fallen, not as a universal endpoint.

What if you stopped without a plan already?

Common, and not a disaster for most classes. If you were on an oral bisphosphonate and stopped a year ago, the reservoir effect means you have likely lost less than you fear, and the sensible step is a density scan to establish where you actually are.

The exception is denosumab. If you stopped that without a handover, contact a prescriber promptly rather than waiting for a routine appointment, because the rebound is time-dependent. Our page on denosumab explains why.

If you were on an anabolic drug and nothing followed, the gains have probably been lost, but that is a reason to restart a plan rather than to conclude treatment failed.