Basal Ganglia Hemorrhage Prognosis by Age: Survival Curves, Outcome Shifts, and a 30-Day Timeline

Basal ganglia hemorrhage prognosis by age is driven less by age itself than by what age does to the brain's reserve and to the hemorrhage's volume: patients under 60 have roughly 40 to 55 percent 30-day survival, those 60 to 79 fall to about 25 to 40 percent, and patients 80 and older drop to roughly 10 to 20 percent in most series. Age is a genuine independent risk factor, but in every age band the two variables that move the needle most are hematoma volume and the Glasgow Coma Scale score on arrival. In plain terms: age sets your starting odds, but clot size and consciousness level decide the actual outcome.

Why age affects prognosis: reserve, atrophy, and clot mechanics

The mechanism behind the age gradient is not simply "older bodies are weaker." Four distinct processes compound.

First, brain atrophy. Older brains have larger cerebrospinal fluid spaces and more slack inside the skull. That sounds protective, and up to a point it is: an atrophic brain can absorb a 30-mL hemorrhage with less midline shift than a young, tight brain. This is why some octogenarians with moderate clots do better than expected. But the same atrophy means the hemorrhage tracks along white matter pathways more freely, and functional recovery is slower because there is less healthy tissue to recruit.

Second, amyloid and small-vessel burden. Cerebral amyloid angiopathy prevalence roughly doubles each decade after 60, and hypertensive lipohyalinosis of the lenticulostriate arteries is near-universal in patients with long-standing hypertension. Older patients are more likely to have both, which raises the risk of hemorrhage expansion in the first 24 hours — the single most modifiable driver of early deterioration.

Third, comorbidity load. Anticoagulant use, chronic kidney disease, and atrial fibrillation cluster in older age groups. Warfarin-associated hemorrhage carries a mortality penalty of roughly 10 to 20 percentage points over spontaneous hypertensive hemorrhage in matched volume and GCS, largely because reversal takes hours, not minutes.

Fourth, physiologic reserve. The brain's ability to tolerate perihematomal edema and secondary inflammation declines. Edema peaks between days 3 and 7 and can add an effective 20 to 30 percent to lesion volume; a 70-year-old has less capacity to buffer that than a 45-year-old.

In plain terms: age matters, but mostly because of what accumulates with age — fragile vessels, blood thinners, and less spare brain.

Quantified context: what the mortality numbers actually look like by decade

Mortality from spontaneous intracerebral hemorrhage overall sits at roughly 40 percent at 30 days across all ages and locations, per the 2015 American Heart Association/American Stroke Association scientific statement on intracerebral hemorrhage management. Basal ganglia hemorrhage is specifically among the higher-mortality locations, alongside thalamic bleeding, because of its proximity to the internal capsule and third ventricle.

An often-cited analysis of the INTERACT2 trial cohort (Anderson et al., 2013) and its follow-on observational data found a clear stepwise age effect. The table below synthesizes the ranges most consistently reported across large cohorts; individual series vary with case mix, withdrawal-of-care practices, and whether comatose patients were included.

Two caveats matter. These ranges are broad because age is co-linear with hematoma volume and with do-not-resuscitate orders. And the 80-plus band is the most distorted by treatment bias: some studies show that when elderly patients receive equivalent aggressive care, the mortality gap narrows by 10 to 15 percentage points.

The 30-day timeline: when deterioration actually happens

Prognosis is not a single number; it is a curve with inflection points. Most deaths from basal ganglia hemorrhage occur in three distinct windows, and the cause of death differs in each.

The third row is where age interacts most strongly with physiology. Edema that peaks on day 3 produces a smaller absolute volume increase in older patients (less inflammation) but a larger functional hit (less reserve), so the clinical trajectory can look deceptively stable on imaging while the patient declines.

In plain terms: the first two days decide whether the patient lives; the first two weeks decide how well.

Age versus other prognostic variables: what actually predicts outcome

Age is a useful prior, but three variables routinely outrank it in predictive models. The ICH score (Hemphill et al., 2001) assigns points for GCS, hematoma volume, intraventricular extension, infratentorial location, and age 80 or older. Age above 80 contributes just 1 of a possible 6 points — but it is the point that most often pushes a patient from a "moderate" to a "poor" prognostic category.

Notice the ordering. A 55-year-old with a 60-mL clot and GCS 6 is in far worse shape than an 82-year-old with a 12-mL clot and GCS 14. Age models are misused when they are treated as the primary decision variable rather than a modest additive one.

Misconceptions and the counterintuitive mechanism

The most common correction: "Age 80 means surgery is futile." The STICH II trial (Mendelow et al., 2013) excluded patients over 95 and found no overall survival benefit from early surgery for superficial lobar hemorrhage, but the subgroup analyses did not show age as a modifier of treatment effect — the problem was the intervention, not the patient's birthday. For basal ganglia hemorrhage specifically, the evidence for open surgery is even weaker, and minimally invasive approaches such as MISTIE III (Hanley et al., 2019) also failed their primary endpoint while showing a signal in patients with residual clot under 15 mL. The takeaway is that age is not the reason to decline intervention; clot characteristics are.

The second, less obvious mechanism concerns midline shift. A young patient with a tight brain develops midline shift early at a smaller clot volume, which triggers herniation at lower hematoma size. An older patient with atrophy can accumulate 40 to 50 mL before significant shift appears. Paradoxically, this means the older brain sometimes tolerates the initial insult better but pays later in functional recovery, and the young brain sometimes dies earlier at a smaller absolute volume.

The third misconception is about "prognosis" itself. Published mortality figures are contaminated by self-fulfilling prophecy: physicians withdraw care early in elderly patients at roughly twice the rate of younger patients with matched ICH scores. Some estimates suggest that 15 to 25 percent of the apparent age effect on mortality is attributable to differential withdrawal of life-sustaining therapy, not biology.

What to actually do: concrete actions that change the trajectory

The actions below are either established standard of care or supported by strong observational data. Where evidence is weaker, it is labeled.

• Reverse anticoagulation within the first hour if the patient is on warfarin: target an INR below 1.4 using vitamin K plus prothrombin complex concentrate, per 2015 AHA/ASA guidance. Rapid reversal is the single fastest way to reduce expansion risk in older patients.
• Control blood pressure aggressively but without overshoot: the INTERACT2 protocol targeted a systolic of 140 mmHg within one hour. Overly rapid drops below 120 mmHg in elderly patients with chronic hypertension carry a risk of renal and cerebral hypoperfusion; a 140 mmHg target is the safer floor.
• Obtain a repeat CT at 6 to 12 hours if the initial scan shows any expansion sign (blend sign, spot sign, or irregular margin). Expansion occurs in about 30 percent of patients and is the leading cause of early deterioration.
• Do not make early do-not-resuscitate decisions in the first 48 hours unless the patient's premorbid wishes are known. Withdrawal within 24 to 48 hours is the strongest documented source of age-related mortality distortion.
• Screen aggressively for aspiration and start swallow evaluation within 24 hours. Aspiration pneumonia is the most common non-neurological cause of death in the second week, and it is disproportionately common in patients over 70.
• Consider an external ventricular drain in any patient with intraventricular hemorrhage and declining consciousness; hydrocephalus is potentially reversible and its treatment effect does not appear to be age-dependent (observational data, moderate certainty).

Red flags that change prognosis on the spot

These findings reliably reset the expected outcome, and they must be documented early:

• Hematoma volume above 30 mL on the initial CT with any decline in GCS within 6 hours.
• A "spot sign" on CT angiography, indicating active contrast extravasation and a higher expansion risk.
• Intraventricular extension reaching the third or fourth ventricle, raising hydrocephalus risk.
• Systolic blood pressure above 200 mmHg on arrival or failure to control below 160 mmHg within 6 hours.
• International normalized ratio above 1.4 at presentation with anticoagulant use.
• Any new midline shift above 5 mm on repeat imaging.

Limits of the evidence

Three caveats should temper every number above. First, most prognostic studies are retrospective or post hoc analyses of trial cohorts, and trial populations systematically exclude the sickest and the oldest. Second, age bands in these studies are not standardized — some use 80 as the cutoff, some 70, some 75 — so the ranges above are synthesized, not drawn from a single validated nomogram. Third, the strongest predictor of outcome in individual cases is not on any of these tables: the patient's premorbid functional status. An 85-year-old living independently has a materially better expected trajectory than a 65-year-old with advanced dementia, and no age-based model captures that.

The honest summary is that basal ganglia hemorrhage prognosis by age is a probabilistic prior, not a verdict. Age gives you the starting distribution. Hematoma volume, consciousness level, anticoagulation status, and — critically — the treatment intensity the patient actually receives determine where within that distribution they land.