Osteoporosis Drugs, Fracture Reduction, and the Drug-Holiday Question

For a woman at high fracture risk, a bisphosphonate prevents many more hip fractures than it will ever cause. The side effect that frightens people most, the atypical femur fracture, is real but rare, and the trade is heavily one-sided. In a 2020 New England Journal of Medicine analysis of women treated in a large California health system, roughly 149 hip fractures were prevented for every two bisphosphonate-associated atypical femur fractures over three years among White women. That imbalance, plus the fact that the rare risk grows with each additional year of treatment and eases after stopping, is the whole reason a drug holiday exists. It is a scheduled point to weigh benefit against harm again, not a rule that fits everyone.

Key points#

• Bisphosphonates reduce vertebral and hip fractures, the breaks that most threaten mobility and survival.
• Atypical femur fractures are rare; their relative risk sounds large but the absolute count stays small.
• The atypical-fracture risk climbs with years of continuous use and falls within roughly a year after stopping.
• A drug holiday is an individualized reassessment after about five years of oral or three years of intravenous therapy, not an automatic stop.
• Screening, not treatment, is where this whole conversation begins.

Two numbers, pointing in opposite directions#

Most treatment decisions in medicine come down to holding two numbers side by side. Here the first is the benefit: bisphosphonates slow the cells that break bone down, which lets bone density stabilize and cuts the fractures that carry the heaviest consequences. Large randomized trials established reductions in both spine and hip fractures, and the 2016 task force report of the American Society for Bone and Mineral Research treats that evidence as the case for starting people at high fracture risk. A hip fracture in an older adult can end independent living and carries real excess mortality in the following year, so a drug that reliably prevents one clears a high bar.

The second number is the one you are most likely to fixate on, and it deserves to be taken seriously rather than waved away. An atypical femur fracture is not an ordinary break. It occurs in the shaft of the thigh bone below the hip joint, often with little or no fall behind it, and it can be preceded by weeks of dull thigh pain. It feels like a betrayal of the drug's purpose, which is part of why it looms so large in your mind.

How the rare risk actually behaves#

The useful thing about the 2020 analysis is that it measured how the atypical-fracture risk changes over time rather than reporting a single average. Compared with less than three months of use, the hazard climbed steeply with duration: roughly nine-fold at three to five years and more than forty-fold beyond eight years of continuous treatment. Read alone, that trajectory is alarming.

The other half of the same finding is reassuring, and it is the part you rarely hear. Once the medication stopped, the excess risk dropped quickly, by about half within roughly fifteen months and by about three-quarters after that. The risk tracks with continued, prolonged treatment, and it recedes when treatment ends. That single fact, a harm that both accumulates with time on the drug and unwinds after coming off it, is what turns duration into something you and a clinician can actively manage.

Why a big relative risk can still be a small problem#

A forty-fold hazard ratio only means something once it is anchored to how common the event is to begin with. Atypical femur fractures are uncommon, so multiplying a very small baseline by a large factor still produces a small number of actual fractures. That is why the head-to-head count is the figure to remember rather than the ratio. Over three years, the White women in the study saw about 149 hip fractures and 541 total clinical fractures prevented against two atypical fractures caused. Prevention swamps harm.

The balance was not the same for every group, which is exactly why a single blanket policy fails. Women of Asian ancestry in the same analysis carried a higher atypical-fracture risk, so their trade was tighter: about 91 hip fractures prevented against eight atypical fractures over three years. The benefit still favored treatment, but by a narrower margin. A frightening relative risk and a small absolute risk can describe the very same fact, and only the absolute one tells you what to do.

The holiday as a checkpoint, not an exit#

If harm builds with cumulative years and fades after stopping, then time on the drug becomes a variable to steer rather than accept. That is the logic of the drug holiday. Give a bisphosphonate long enough to bank its fracture protection, then pause and look again instead of continuing on autopilot.

What that reassessment finds should decide what happens next. The ASBMR task force offered a practical checkpoint: after roughly five years of oral therapy or three years of the intravenous form, revisit the person's risk. Those who remain high risk, for example someone with a very low hip bone-density score or a fracture that happened during treatment, generally do better continuing, a pattern borne out in the long-term extensions of the alendronate and zoledronic acid trials. Those whose risk has drifted lower may reasonably pause, since oral bisphosphonates linger in bone and offer some carryover protection after the last dose. A holiday is not a permanent goodbye; it is a monitored interval during which treatment can resume if risk climbs again.

Where the conversation begins#

None of this reaches a person who was never flagged as being at risk, which is the job of screening. In early 2025 the US Preventive Services Task Force reaffirmed a Grade B recommendation for bone-density testing in women 65 and older, and in younger postmenopausal women whose risk is raised by other factors. Screening is what identifies the people for whom this benefit-versus-harm arithmetic is worth doing at all. The Task Force also found the evidence insufficient to make the same recommendation in men, a reminder that guidance follows data and stops where the data stop.

Reading it as a patient#

The lasting takeaway is a way of asking questions. Request absolute numbers, not only relative risk. Ask how a benefit and a harm each move over time instead of treating either as fixed. And expect that the same evidence can support starting, continuing, or pausing depending on your fracture risk, ancestry, and response to treatment. Choices about osteoporosis treatment belong to each person and their own clinician.