The association between haemoglobin level and risk of developing dementia

Background

Dementia is a major global health challenge with rapidly increasing prevalence in aging societies. Despite extensive research, effective disease-modifying treatments remain limited, highlighting the importance of identifying modifiable risk factors. Hemoglobin levels, as indicators of systemic health and oxygen-carrying capacity, have been associated with various adverse outcomes, including cardiovascular disease and stroke. Both low and high hemoglobin levels may influence cerebral perfusion and vascular integrity. However, previous studies examining the relationship between hemoglobin levels and dementia risk have yielded inconsistent findings, and large-scale population-based evidence remains limited. In addition, potential sex differences in this association have not been fully clarified.

Objectives

This study aimed to investigate the association between hemoglobin levels and the incidence of all-cause dementia and its subtypes in a nationwide Korean cohort, and to evaluate whether these associations differ according to sex.

Methods

We conducted a retrospective cohort study using data from the National Health Insurance Service–National Health Screening Cohort (NHIS-HEALS). A total of 316,542 adults aged ≥40 years who underwent health screening in 2004–2005 were included and followed until 2019. Hemoglobin levels were categorized into sex-specific quintiles after excluding extreme values (top and bottom 1%). Dementia was defined based on ICD-10 diagnostic codes combined with prescriptions for anti-dementia medications to improve diagnostic validity. Outcomes included all-cause dementia, Alzheimer’s dementia, and vascular dementia. Multivariable Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs), adjusting for demographic characteristics, socioeconomic status, lifestyle factors, and comorbid conditions. Restricted cubic spline analyses were performed to assess potential nonlinear associations, and sex-stratified analyses were conducted to explore effect modification.

Results

During follow-up, both low and high hemoglobin levels were associated with an increased risk of all-cause dementia compared to the reference (fourth quintile). The adjusted HRs were 1.13 (95% CI, 1.04–1.24) for the lowest quintile and 1.08 (95% CI, 1.02–1.15) for the highest quintile. For Alzheimer’s dementia, increased risk was observed primarily in the lower hemoglobin quintiles, showing a graded association. In contrast, vascular dementia risk was elevated at both low and high hemoglobin levels, suggesting distinct underlying mechanisms. Restricted cubic spline analyses demonstrated a nonlinear association, with a more pronounced increase in risk at lower hemoglobin levels. Notably, sex-stratified analyses revealed that these associations were stronger and more consistent in women, whereas no clear significant associations were observed in men.

Conclusions

Abnormal hemoglobin levels, particularly low levels, are associated with an increased risk of dementia, with distinct patterns across dementia subtypes. The stronger associations observed in women suggest potential sex-specific biological mechanisms, possibly related to hormonal or vascular differences. Given that hemoglobin levels are modifiable through clinical management and lifestyle interventions, these findings highlight a potential target for dementia prevention and risk stratification in population-based settings. These results may also support the integration of routine hematologic parameters into population-level screening strategies for early identification of individuals at elevated risk of dementia.

References

No reference