Association Between Hemoglobin Levels and Hemorrhagic Stroke Severity and In-Hospital Outcomes: A Multi-Centre Cross-Sectional Study on Recovery and Prognosis in Pakistan.
Johar A, Ahmad A, et al. • Brain and behavior • 2026
Admission hemoglobin levels were not independently associated with in-hospital functional outcome or mortality in acute hemorrhagic stroke patients, with stroke severity (NIHSS) and diabetes emerging as the strongest predictors of poor prognosis.
Key Findings
Results
Hemoglobin category was not associated with functional outcome or stroke severity in unadjusted analyses.
445 acute hemorrhagic stroke patients were included, with 330 (74.2%) in the low Hb group and 114 (25.6%) in the normal Hb group.
Mean NIHSS scores were similar between low and normal Hb groups (9.74 ± 7.02 vs. 10.64 ± 6.25).
Mean mRS scores at discharge were similar between groups (2.15 ± 2.20 vs. 2.25 ± 1.78).
Hb category showed no statistically significant difference in mRS (p = 0.257) or NIHSS (p = 0.195).
Results
In multivariable logistic regression, low hemoglobin was not independently associated with favorable functional outcome or mortality.
Adjusted odds ratio for favorable outcome (mRS 0–2) for low vs. normal Hb was aOR 0.77 (95% CI 0.43–1.38, p = 0.378).
Adjusted odds ratio for mortality for low vs. normal Hb was aOR 2.15 (95% CI 0.78–5.91, p = 0.139).
Models were adjusted for age, sex, NIHSS, and comorbidities.
No evidence of non-linearity in the hemoglobin–outcome relationship was found (p = 0.41), assessed using restricted cubic splines.
Results
NIHSS score was the dominant independent predictor of both favorable outcome and mortality.
Each one-point increase in NIHSS was associated with an aOR of 0.77 for favorable outcome (p < 0.001).
Each one-point increase in NIHSS was associated with an aOR of 1.07 for mortality (p = 0.019).
NIHSS was retained as a core adjustment variable in all multivariable models.
Results
Diabetes was independently associated with increased mortality risk in multivariable analysis.
Diabetes was associated with an aOR of 2.58 for mortality (p = 0.014).
Diabetes was also linked to poorer outcomes in unadjusted group comparisons (p < 0.001).
Hypertension (p = 0.0001) and older age (p = 0.0049) were also linked to poorer outcomes in unadjusted analyses.
Results
Later hospital arrival was associated with worse functional outcomes and higher mortality.
Later hospital arrival was associated with higher mRS scores (χ² p < 0.0001).
Later hospital arrival was associated with higher mortality (χ² p = 0.013).
This finding highlights the importance of timely presentation in hemorrhagic stroke.
Results
Mortality rates differed numerically between low and normal hemoglobin groups, though this difference was not statistically significant after adjustment.
Deaths occurred in 32 of 330 (9.7%) patients in the low Hb group versus 4 of 114 (3.5%) in the normal Hb group.
Despite the numerical difference, adjusted analysis showed no independent association between Hb category and mortality (aOR 2.15, 95% CI 0.78–5.91, p = 0.139).
The single patient with high hemoglobin was excluded due to insufficient sample size.
Methods
The study was a retrospective multi-center cross-sectional study conducted in Pakistan with a predominantly low-hemoglobin patient population.
Low hemoglobin was defined as <12 g/dL in females and <13 g/dL in males.
74.2% of the 445 included patients had low hemoglobin at admission, reflecting a high burden of anemia in this population.
Hemoglobin levels were measured within 24 hours of admission.
Functional outcome was assessed by the Modified Rankin Scale (mRS) and stroke severity by NIHSS at discharge.
Minimum required sample size was 384; 400–450 participants were included to compensate for non-response and incomplete data.
What This Means
This research suggests that the level of hemoglobin (the protein in red blood cells that carries oxygen) at the time of hospital admission does not independently predict how well patients with bleeding strokes (hemorrhagic stroke) will recover or whether they will survive during their hospital stay. The study looked at 445 stroke patients across multiple hospitals in Pakistan and found that while most patients (74%) had low hemoglobin, the hemoglobin level itself—whether low or normal—did not significantly change the odds of a good recovery or death once other factors were taken into account.
Instead, the strongest predictor of both recovery and death was stroke severity at admission, measured by the NIH Stroke Scale (NIHSS). Patients with more severe strokes on admission were much less likely to have good functional outcomes and more likely to die in hospital. Diabetes was also independently linked to a higher risk of death, and arriving at the hospital later, having hypertension, and being older were all associated with worse outcomes in preliminary analyses. These findings suggest that how severe the stroke is at the time of arrival—and whether the patient has diabetes—matters more for short-term outcomes than their blood hemoglobin level.
This research suggests that while routine hemoglobin testing remains a clinically relevant part of stroke care, simply correcting low hemoglobin levels may not directly improve short-term outcomes for hemorrhagic stroke patients. The study's authors caution that these results only apply to the in-hospital period and should not be used to draw conclusions about longer-term recovery. The findings also highlight the importance of rapid hospital arrival after stroke onset and aggressive management of conditions like diabetes in this patient population.
Johar A, Ahmad A, Akhter J, Aftab W, Ali A, Jamal M, et al.. (2026). Association Between Hemoglobin Levels and Hemorrhagic Stroke Severity and In-Hospital Outcomes: A Multi-Centre Cross-Sectional Study on Recovery and Prognosis in Pakistan.. Brain and behavior. https://doi.org/10.1002/brb3.71715