Association Between Hemoglobin Levels and Hemorrhagic Stroke Severity and In-Hospital Outcomes: A Multi-Centre...

To evaluate the association between admission hemoglobin levels and in-hospital functional outcomes in patients with acute hemorrhagic stroke in Pakistan.

Abnormal hemoglobin levels have an impact on cerebral perfusion and oxygen delivery to the brain, potentially affecting the recovery in stroke patients. Both anemia and polycythemia can impact neurological outcomes, yet evidence is not yet conclusive. This study examined the relationship between Hb levels, hemorrhagic stroke severity, and clinical outcomes.

A retrospective, multi-center cross-sectional study was conducted among adults admitted with acute hemorrhagic stroke. Hemoglobin levels measured within 24 h of admission were divided into three groups that are low (<12 g/dL in females, <13 g/dL in males), normal, and high. Patients with high hemoglobin [n = 1] were excluded due to insufficient sample size. The sample size was calculated using the standard formula for cross-sectional studies at a 95% confidence interval and 5% margin of error. The minimum required sample size was 384 participants; however, 400-450 participants were included to compensate for non-response and incomplete data. Stroke severity was assessed using the NIH Stroke Scale (NIHSS) and functional outcome by the Modified Rankin Scale (mRS) at the time of discharge. Baseline characteristics were summarized descriptively. Group comparisons used Pearson χ² with Cramér's V for categorical variables and Mann-Whitney U or Kruskal-Wallis tests for continuous or ordinal data. Multivariable logistic regression identified independent predictors of favorable outcome (mRS 0-2) and mortality, adjusting for age, sex, NIHSS, and comorbidities. Model calibration was evaluated using the Hosmer-Lemeshow test and discrimination with Nagelkerke R². Non-linearity was assessed using restricted cubic splines. NIHSS was retained as a core adjustment variable in all models.

We included 445 stroke patients at discharge only (Hb-low 330 [74.2%], Hb-normal 114 [25.6%], and Hb-high 1). Baseline severity and outcomes were similar by Hb group (NIHSS 9.74 ± 7.02 vs. 10.64 ± 6.25; mRS 2.15 ± 2.20 vs. 2.25 ± 1.78), with deaths 32/330 (9.7%) versus 4/114 (3.5%). Later hospital arrival was associated with higher mRS (χ² p < 0.0001) and mortality (χ² p = 0.013). Diabetes (p < 0.001), hypertension (p = 0.0001), and older age (p = 0.0049) were linked to poorer outcomes, whereas Hb category showed no difference in mRS (p = 0.257) or NIHSS (p = 0.195). In multivariable models, Hb (low vs. normal) was not independently associated with favorable outcomes (aOR 0.77, 95% CI 0.43-1.38, p = 0.378) or mortality (aOR 2.15, 95% CI 0.78-5.91, p = 0.139). NIHSS remained the dominant predictor (favorable outcome: aOR 0.77 per point, p < 0.001; mortality: aOR 1.07 per point, p = 0.019), with diabetes also increasing death risk (aOR 2.58, p = 0.014). No evidence of non-linearity was found [p = 0.41].

Admission hemoglobin levels were not independently associated with in-hospital functional outcome or mortality in this cohort. Stroke severity and diabetes were the strongest predictors of poor prognosis. Routine hemoglobin assessment remains clinically relevant, but targeted correction of hemoglobin may not directly improve outcomes in patients. Findings are limited to the in-hospital period and should not be extrapolated to long-term recovery.