Hip fracture mortality rates climb steeply with age, and the risk concentrates in the first 90 days, when 30-day mortality runs roughly 5 to 10 percent across health systems and one-year mortality reaches about 20 to 30 percent, with the highest rates in the oldest and frailest patients. Age, prefracture frailty, delayed surgery, and medical complications combine to drive that curve. This explainer covers the biology behind each factor, the numbers, and what the evidence says about changing the trajectory.

The fracture-to-death mechanism

A hip fracture rarely kills directly. It starts a chain of physiologic events that, in a vulnerable patient, is difficult to interrupt. Tissue injury triggers a systemic inflammatory response, and the resulting immobility accelerates muscle loss, venous stasis, and atelectasis. Inflammatory cytokines and catecholamines rise together and destabilize an already marginal cardiovascular system.

The operation itself adds a second hit. Anesthesia, blood loss, and surgical trauma provoke further inflammatory and coagulative activation. Patients who cannot mount a reserve response decompensate, which is why cardiac complications, pneumonia, delirium, and thromboembolism cluster in the first two weeks.

Bone biology sets the stage long before the fall. Age-related loss of trabecular bone reduces both bone mineral density and microarchitecture quality, so a low-energy fall produces a fracture that a younger skeleton would absorb. The same processes that thin bone, including sarcopenia and reduced mobility, also thin the physiologic reserve needed to survive treatment. The fracture is therefore a marker of systemic vulnerability, not just a local injury.

One deeper mechanism deserves emphasis: postoperative delirium. It affects a substantial share of older hip fracture patients and is associated with higher mortality and loss of independence. The mechanism involves neuroinflammation plus sleep disruption and sensory deprivation in an unfamiliar ward, and it is partly modifiable through pain control, orientation, and early mobilization. A patient who develops delirium often stops eating, stops moving, and loses the recovery window.

Plain language: the break is a stress test the body did not choose, and the body's reserve decides much of the outcome.

Quantified context and the age gradient

Mortality follows age in two distinct steps. The first is the acute event: 30-day mortality is highest in patients over 85 and in those with severe comorbidity, and it is lower in patients who walk independently before the fall. The second is the residual risk: many survivors never regain prefracture function, and that functional loss predicts death over the following year.

Excess mortality, the difference between observed deaths and expected deaths in a matched population, is large in the first six months and then narrows. This pattern matters for interpretation. A report of "one-year mortality of 25 percent" is not a statement that the fracture causes all 25 percent of deaths; some patients were already near the end of life. Studies that compare with matched controls consistently find the excess concentrated early.

The age gradient is steep but not absolute. A robust 90-year-old who walks without aid and has no cognitive impairment can do better than a frail 70-year-old with heart failure and dementia. Age is a proxy for reserve, and the proxies around it, including gait speed, grip strength, and cognitive status, carry much of the predictive power.

Sex modifies the picture. Women sustain hip fractures more often because of postmenopausal bone loss, but men who fracture have higher mortality at every age. The mechanism is partly later presentation, partly higher comorbidity burden, and partly differences in treatment and rehabilitation engagement. Any comparison of rates across populations must account for this.

The timing pattern follows a predictable course:

Time after fracture Typical mortality pattern Dominant mechanism
In-hospital 3 to 6 percent Surgical complications, cardiac events, sepsis
30 days 5 to 10 percent Pneumonia, delirium, thromboembolism, wound issues
90 days 12 to 20 percent Functional decline, deconditioning, nosocomial complications
1 year 20 to 30 percent Loss of independence, recurrent falls, comorbidity progression
2 years Plateau near matched-population excess Chronic frailty trajectory
5 years Most survivors regain or stabilize function; mortality gap narrows Recovery or persistent disability

These figures are ranges from large registry and cohort data, and they vary by country, case mix, and whether patients are treated in dedicated orthogeriatric services. A center that reports 30-day mortality near 5 percent is usually one with early surgery and coordinated geriatric care, not one with healthier patients.

Edge cases that shift the curve

Pathologic fractures from metastatic disease carry far higher mortality than osteoporotic fractures, and they require a different treatment pathway. A patient with known cancer and a new hip fracture should be evaluated for metastasis before assuming a fragility mechanism.

Periprosthetic fractures around a prior hip replacement are increasing in frequency as the population with implants ages. These patients are older and have more complex surgical needs, and their mortality is comparable to or higher than native fragility fractures.

Anticoagulation changes management and risk. A patient on a direct oral anticoagulant may face surgical delay while the drug clears, and delay is independently associated with worse outcomes. Balancing bleeding risk against the harm of waiting is a genuine clinical tension, not a checklist item.

Cognitive impairment is a powerful modifier. Patients with dementia present later, are more likely to be undertreated for pain, and have higher rates of delirium and institutionalization. They also have higher mortality, and the effect persists after adjusting for age and comorbidity.

Counterintuitive detail: the strongest single predictor in many models is not the fracture type or the surgical approach. It is the patient's prefracture walking ability and cognitive status. Two patients with identical radiographs and identical operations can have very different trajectories based on reserve alone. This is why risk scores built purely on surgical variables underperform.

Professional caveat: mortality statistics are almost always reported as all-cause mortality. They cannot tell you which deaths were preventable, which were attributable to the fracture, and which would have occurred anyway. Read any rate with the comparison group in mind.

Misconceptions that distort risk perception

The first misconception is that the operation is the main danger. In reality, the harm of prolonged immobility generally exceeds the risk of a carefully timed operation. Early surgery, typically within 48 hours when medically feasible, is associated with lower mortality and fewer complications in most registry analyses. Delaying surgery to "optimize" a patient who is not acutely unstable often causes more harm than it prevents.

The second is that surviving the hospital discharge is the finish line. Most excess deaths and much of the functional loss occur after discharge. The post-acute period, with its transitions between hospital, rehabilitation, and home, is where continuity breaks down.

The third is that mortality rates are destiny. They are averages across heterogeneous populations, and they should inform goals-of-care conversations rather than replace them. A specific patient is not a statistic.

The fourth is that age alone should drive treatment decisions. Frailty and function predict better than chronologic age, and older adults with good reserve frequently benefit from the same aggressive pathway as younger patients.

Practical actions and thresholds

The evidence supports a structured, time-sensitive approach. The following steps are grounded in orthogeriatric consensus and should be applied by the treating team, not as self-directed instructions.

Assess frailty and cognitive status on admission using a validated tool, and document prefracture walking ability, because these predict mortality and discharge destination. Target surgery within 48 hours when the patient is medically stable; treat reversible problems (volume, pain, electrolytes) without unnecessary delay. Provide multimodal pain control and early mobilization, ideally within 24 hours of surgery when safe, since immobility drives complications. Run a delirium prevention bundle: reorientation, sleep protection, vision and hearing aids, and family presence where possible. Start venous thromboembolism prophylaxis and pressure-injury prevention according to local protocol, and review anticoagulation timing with the surgical team. Use a dedicated orthogeriatric co-management model where available; these services are associated with lower mortality and better function.

Reconciling thresholds: different guidelines use different cutoffs for "early" surgery, with 24, 36, and 48 hours all appearing in the literature. The 48-hour threshold has the broadest support and the clearest association with mortality in registry data. The stricter 24-hour threshold may benefit selected patients but is not universally achievable and should not be treated as a failure when clinically justified delay occurs. For frailty, there is no single universal score; the Clinical Frailty Scale is widely used and stratifies risk reasonably well. For cognitive screening, a brief validated instrument is adequate for risk stratification.

One plain-language rule: the two levers with the most consistent evidence are not waiting when you do not have to, and getting the patient moving as soon as it is safe.

Limitations of the evidence

Mortality data come largely from registries and observational cohorts, and they are subject to confounding by indication and case mix. Centers that operate early may also have better geriatric care, so the independent effect of timing is hard to isolate. Randomized trials of surgical timing are limited and ethically constrained.

Reported rates also vary by how mortality is defined and captured. In-hospital mortality, 30-day mortality, and 90-day mortality measure different things and can move in different directions. Cross-country comparisons are further complicated by differences in admission thresholds, do-not-resuscitate practices, and post-acute care capacity.

Prediction models exist, but most have modest discrimination and limited external validation. No model should replace clinical judgment or substitute for a goals-of-care conversation with the patient and family.

Finally, most studies do not report cause-specific mortality, and few report patient-reported outcomes such as pain, function, and return to home. A mortality rate alone cannot capture what matters most to patients. Better data on functional recovery and quality of life would change how these risks are discussed, and until then, the numbers should be treated as a guide rather than a verdict.