Younger adults with intracerebral hemorrhage have better outcomes but distinct risk profile: INTERACT3 sub-analysis

• Approximately 19% of ICH patients in the INTERACT3 trial were ≤50 years old.

• Younger patients had larger hematoma volumes but significantly lower 6-month mortality (9.1% vs 16.6%; adjusted OR 0.42) and reduced disability.

• A significant age-by-treatment interaction suggested younger patients benefited more from the bundled care intervention.

• Younger patients reported better quality of life across mobility, self-care, and usual activities.

• Distinct risk profile: higher rates of smoking, alcohol use, male predominance, and higher BMI in younger group.

Study snapshot

What the study asked

Intracerebral hemorrhage (ICH) accounts for about 10–20% of all strokes and carries disproportionately high morbidity and mortality. Although ICH is more common in older adults, a substantial proportion occurs in younger individuals, often with distinct etiology and risk factors. The INTERACT3 trial, a large stepped-wedge cluster-randomised study conducted across 122 hospitals in 10 countries, previously demonstrated that a bundled care intervention improved outcomes for patients with acute ICH. This secondary analysis aimed to compare baseline characteristics, in-hospital management, and 6-month outcomes between younger (≤50 years) and older adults enrolled in the trial.

Why age matters in intracerebral hemorrhage

ICH in younger adults is often attributed to vascular malformations, illicit drug use, or coagulopathies, but hypertensive and lifestyle-related risk factors are increasingly recognized. Understanding whether younger patients respond differently to acute interventions is crucial for tailoring treatment protocols. The INTERACT3 dataset offered a unique opportunity to explore age-related differences within a controlled, multinational framework.

How the analysis was conducted

The INTERACT3 trial enrolled patients with acute ICH from 2017 to 2021. For this pre-specified secondary analysis, investigators categorized patients into two age groups: young adults (≤50 years) and older adults (>50 years). Baseline demographics, imaging characteristics, in-hospital treatments, and clinical outcomes at 6 months were compared. The primary outcome for the sub-study was functional status measured by the modified Rankin Scale (mRS), mortality, and quality of life assessed with EQ-5D-3L. Statistical analyses used generalized linear mixed models accounting for clustering by hospital, with adjustment for pre-specified covariates including sex, baseline hematoma volume, and time from onset to treatment.

Key findings

Baseline characteristics

A total of 7031 patients were included in this secondary analysis, with 19.2% aged 50 years or younger. Compared to older patients, younger individuals were more often male (70.8% vs 62.4%, P < .0001), had higher BMI (25.3 vs 23.8, P < .0001) and were more likely to smoke and consume alcohol (both P < .0001). They had fewer comorbidities, including hypertension, previous stroke, coronary artery disease, atrial fibrillation and diabetes (all P < .01) and demonstrated better pre-stroke functional status (mRS 0 in 84.9% vs 75.4%, P < .0001). These younger patients were also less likely to be on antihypertensives, lipid-lowering therapy, antiplatelets or glucose-lowering agents at admission. While systolic BP was comparable, diastolic BP was significantly higher among younger patients (105.7 vs 97.8 mmHg, P < .0001). Stroke severity at baseline, assessed by NIHSS and Glasgow Coma Scale (GCS) scores, was similar across age groups. Randomisation between intervention and control groups was balanced (P = .4252).

Mortality and disability

Of the 7031 participants, 1351 (19.2%) were aged 50 years or younger. Younger patients were more likely to be male (70.8% vs 62.4%), have higher body mass index (25.3 vs 23.8 kg/m²), and report current smoking (36.1% vs 21.4%) and alcohol consumption (33.2% vs 13.3%). Despite having slightly larger median hematoma volumes (18.0 vs 15.0 mL), younger patients experienced significantly better outcomes. Six-month mortality was 9.1% in the young group versus 16.6% in older adults (adjusted odds ratio [OR] 0.42; 95% confidence interval [CI] 0.33–0.54). The rate of death or disability (mRS 3–6) was 46.5% vs 57.8% (adjusted OR 0.56; 95% CI 0.48–0.65). Younger patients were also more likely to achieve symptom-free recovery (mRS 0: 14.0% vs 8.3%; adjusted OR 0.51; 95% CI 0.45–0.58).

Median haematoma volumes in millilitre with interquartile range at baseline, 24 h and day 7 for each age group.

Quality of life

Across EQ-5D-3L domains, younger patients reported fewer problems with mobility (73.4% reporting problems vs 62.8%; adjusted OR 0.53; 95% CI 0.45–0.62), self-care (61.4% vs 54.1% with no problems; adjusted OR 0.61; 95% CI 0.52–0.70), and usual activities (48.2% vs 38.9%; adjusted OR 0.59; 95% CI 0.51–0.68). No significant differences were seen for pain/discomfort (P=0.82) or anxiety/depression (P=0.56). The mean overall health utility score favored younger patients (0.7 vs 0.6; adjusted mean difference 0.11; 95% CI 0.09–0.14).

Age and treatment interaction

A statistically significant interaction between age group and treatment effect was observed for the primary composite of death or disability (mRS 3–6; P for interaction=0.0251) and for disability alone (mRS 3–5; P=0.0085). Stratified analyses suggested that the bundled care intervention had a more pronounced effect in younger patients, while the effect was attenuated in older adults. This finding was consistent across adjusted models and indicates that younger age may be a treatment-effect modifier. No significant interactions were seen for mortality alone, recurrent events, or other secondary outcomes.

Important limitations

As a secondary analysis of a cluster-randomised trial, this study has inherent limitations. The age comparison is observational, and unmeasured confounders may influence outcomes. The INTERACT3 protocol was not designed to test age-specific hypotheses, and the subgroup analyses were exploratory. The definition of young as ≤50 years is arbitrary, and results may vary with different cutoffs. Furthermore, the trial excluded patients with certain comorbidities (e.g., advanced dementia, terminal illness), which may limit generalizability. Lastly, the stepped-wedge design and variable implementation of the bundle across sites could introduce temporal biases.

Implications for clinicians and patients

These findings reinforce that younger ICH patients have a distinct risk profile—marked by smoking, alcohol use, and male sex—and that aggressive lifestyle modification may be especially important for primary prevention. Acutely, younger patients appear to derive greater benefit from early bundled care, including prompt blood pressure lowering and integrated management. The results highlight the need for age-stratified approaches in future ICH trials and may inform tailored protocols. For patients and families, the data provide evidence that younger age is associated with better survival and functional recovery after ICH, but the burden of disability remains substantial.

References

1. Khan M, Ouyang M, Wasay M, et al. Clinical characteristics and outcomes of intracerebral haemorrhage in young vs older adults: insights from the INTERACT3 trial. Eur Stroke J. 2026;11(6):aakag040. doi:10.1093/esj/aakag040. PMID: 42378499.