Serum albumin levels are associated with cardioembolic and cryptogenic ischemic strokes: Northern Manhattan Study

Abstract

Background and purpose: Low serum albumin concentrations have been associated with increased stroke risk, but the underlying mechanisms are less well studied. We aimed to investigate the association between serum albumin levels and ischemic stroke etiologies in a large, population-based, multiethnic, prospective, cohort study.

Methods: Participants from the Northern Manhattan Study (NOMAS; n=2986; mean age, 69±10 years) free of stroke at baseline were followed for incident stroke (a median follow-up of 12 years). Cox proportional hazard models were used to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for baseline serum albumin levels and risk of ischemic stroke and ischemic stroke subtypes after adjusting for vascular risk factors.

Results: The mean baseline serum albumin level was 4.42±0.33 g/dL. There were 271 ischemic strokes during follow-up. Participants with serum albumin levels of 2.7 to 4.2 g/dL (the lowest tertile) had increased risk of all stroke (HR, 1.76; 95% CI, 1.32-2.35), ischemic stroke (HR, 1.67; 95% CI, 1.21-2.29), cardioembolic stroke (HR, 1.92; 95% CI, 1.10-3.34), and cryptogenic stroke (HR, 2.59; 95% CI, 1.21-5.53), compared with those with levels of 4.6 to 5.5 g/dL (the top tertile; reference). Low albumin levels (2.7-4.2 g/dL) were not associated with large vessel or lacunar stroke.

Conclusions: Our study shows an association between low serum albumin levels and ischemic stroke, particularly cardioembolic and cryptogenic subtypes. These results suggest the potential shared pathophysiological relationship between low serum albumin levels, cardiac embolism, and cryptogenic infarction, which warrants further investigation.

Keywords: prospective studies; serum albumin; stroke.

Figure 1

Multivariable-adjusted hazards ratio for per 1 g/dL serum albumin decrease (A) and for the lowest tertile compared to the top tertile (B) by ischemic stroke subtype. The model was adjusted for age, sex, race-ethnicity, body mass index, moderate alcohol drinking, diabetes, hypercholesterolemia, white blood cell count, estimated Glomerular Filtration Rate, and history of atrial fibrillation

Figure 1

Multivariable-adjusted hazards ratio for per 1 g/dL serum albumin decrease (A) and for the lowest tertile compared to the top tertile (B) by ischemic stroke subtype. The model was adjusted for age, sex, race-ethnicity, body mass index, moderate alcohol drinking, diabetes, hypercholesterolemia, white blood cell count, estimated Glomerular Filtration Rate, and history of atrial fibrillation