Low hemoglobin levels in older adults were associated with elevated blood biomarkers of Alzheimer’s disease (AD) pathology and a significantly higher risk of developing dementia years later, a large cohort study found.

In a cohort of more than 2200 dementia-free adults aged 60 years or older, those with anemia had a 66% higher risk of developing dementia over a 16-year follow-up than those with normal hemoglobin levels.

Anemia was also associated with higher blood concentrations of phosphorylated tau 217 (p-tau217), neurofilament light chain (NfL), and glial fibrillary acidic protein (GFAP) — biomarkers that reflect AD pathology, neurodegeneration, and glial activation, respectively.

Dementia was highest among individuals with both anemia and elevated biomarker levels, with the risk for dementia increasing nearly fourfold among those with anemia and high NfL.

The findings underscore that the brain cannot be understood in isolation from the rest of the body, said lead investigator Martina Valletta, MD, neurologist at the Aging Research Center at Karolinska Institutet in Stockholm, Sweden.

“We tend to think of the brain as something completely separate from the rest of the body, but this is not the case,” Valletta told Medscape Medical News. “If you’re vulnerable due to anemia and Alzheimer’s pathology, you’re more likely to develop dementia — and develop it faster.”

The study was published online on April 17 in JAMA Network Open.

Taking a Deeper Look

Anemia affects an estimated 25% of the global population and roughly 10% of adults aged 65 years or older in the US, with prevalence rising with age.

Previous research has linked anemia to increased dementia risk. Several mechanisms have been proposed, including chronic cerebral hypoxia, oxidative stress, and systemic inflammation. However, the biological pathway connecting anemia to dementia has remained poorly characterized.

Direct evidence linking hemoglobin levels to AD biomarkers has been limited and conflicting. The CABLE study in 2021 reported cross-sectional associations between anemia and low cerebrospinal fluid amyloid-beta 42, but not tau. Another study found no link between hemoglobin and amyloid or tau deposition on brain imaging.

The new study grew out of an earlier analysis by Valletta and colleagues on the link between chronic anemia and elevated AD blood biomarker levels.

“We saw this very strong effect of anemia on AD blood biomarker levels, and we wanted to understand what was going on,” she said.

The researchers used data on 2282 dementia-free individuals (median age, 72.2 years; 61.6% female) with available hemoglobin and biomarker levels from the SNAC-K, a longitudinal population-based study of randomly selected adults aged 60 years or older in Stockholm enrolled between 2001 and 2004.

At baseline, anemia was reported in 8.7% of participants, defined by World Health Organization criteria (hemoglobin ≤ 12 g/dL for female and ≤ 13 g/dL for male). Notably, most cases of anemia were mild and normocytic, the kind commonly seen but often unaddressed in older patients.

Dementia was assessed at follow-up visits every 3 years for participants aged 78 years or older and every 6 years for younger participants, using Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria.

Anemia-Dementia Link

During a mean follow-up of 9.3 years, 362 participants (15.9%) developed dementia. A nonlinear dose-response relationship emerged, with dementia risk increasing progressively as hemoglobin levels declined. In fully adjusted models, anemia was associated with a 66% higher risk for dementia (adjusted hazard ratio [aHR], 1.66; 95% CI, 1.21-2.28).

Valletta noted that no participants had severe anemia. “This tells us that even mild anemia may have a negative impact on brain health,” she said.

The central finding emerged from the joint exposure analysis. When anemia and elevated biomarkers coexisted, the risk for dementia increased more than what each alone would predict.

Among those with both anemia and high NfL, the risk for dementia was nearly fourfold higher than in non-anemic individuals with low NfL (aHR, 3.64; 95% CI, 2.39-5.56). A statistically significant additive interaction was detected for NfL, but not for p-tau217 or GFAP.

The findings lend themselves to two possible interpretations: That anemia drives general neurodegeneration that leads to dementia, or that anemia directly triggers AD-specific pathology.

Valletta said she finds the first explanation more plausible. NfL captures the kind of broad neuronal and vascular damage anemia would be expected to produce, she noted, whereas p-tau217 reflects a highly specific disease process that anemia is unlikely to directly influence.

Because none of the participants had severe anemia, it’s possible that even mild cases may negatively affect brain health, she added.

Association Strongest in Men

The associations were more pronounced in men than in women, a surprising finding. Valletta suggested that because women naturally have lower hemoglobin levels than men throughout life, they may develop a kind of physiologic adaptation. Men tend to develop anemia later, usually in the presence of chronic disease.

“It’s not that having low hemoglobin is protective, but perhaps females are more accustomed to it,” she said. “If you are used to having high hemoglobin levels your whole life, and then those levels drop in old age, you might suffer more for it.”

On the clinical implications, Valletta emphasized that AD blood biomarkers should not be interpreted in isolation.

For example, impaired kidney function caused by chronic kidney disease can lead to elevated AD biomarker levels because the kidneys cannot clear these proteins efficiently. In that case, elevations may reflect poor renal clearance rather than actual brain pathology.

“These biomarkers must be interpreted in the context of the whole person’s health,” she said. “Having high p-tau is an indication of Alzheimer’s pathology, but it does not exclude that there is also something else going on.”

Filling a Gap

The biomarker observations provide compelling evidence linking hemoglobin directly to AD pathology, offering “strong leads for unraveling the role of anemia in dementia,” Frank J. Wolters, MD, PhD, of Erasmus MC University Medical Centre Rotterdam, Rotterdam, Netherlands, wrote in an accompanying editorial.

The direct binding of amyloid-beta to iron, along with recent evidence that cortical iron decreases after antiamyloid antibody treatment, suggests disrupted iron metabolism may be a shared mechanism linking anemia and AD, he noted.

The study’s long-term follow-up limited attrition and reduced the risk for reverse causation, strengthening the results, Wolters noted. But moving from observation to intervention will require mechanistic studies and validation in geographically diverse populations, he added.

Valletta said the critical next step is confirming that blood biomarker findings correspond to actual brain pathology.

“We are measuring something in the blood that we know is highly associated with brain pathology, but we are not directly measuring the relationship between anemia and brain pathology,” she said. “I would like to see that gap filled.”

The study was supported by the Swedish Research Council, the Swedish Ministry of Health and Social Affairs, and other funders. Disclosure information for study authors is available in the original study publication. Wolters reported having no relevant financial relationships.