Serum albumin and white matter hyperintensities

Abstract

People living with HIV and those diagnosed with alcohol use disorders (AUD) relative to healthy individuals commonly have low levels of serum albumin, substantiated as an independent predictor of cardiovascular events. White matter hyperintensities (WMH)-a neuroimaging feature of cerebral small vessel disease-are also related to cardiovascular disease. Despite consensus regarding associations between high levels of urine albumin and WMH prevalence, and low serum albumin levels and impaired cognitive functioning, relations between serum albumin and WMH burdens have rarely been evaluated. Here, a sample including 160 individuals with AUD, 142 living with HIV, and 102 healthy controls was used to test the hypothesis that serum albumin would be inversely related to WMH volumes and directly related to cognitive performance in the two diagnostic groups. Although serum albumin and periventricular WMH volumes showed an inverse relationship in both AUD and HIV groups, this relationship persisted only in the HIV group after consideration of traditional cardiovascular (i.e., age, sex, body mass index (BMI), nicotine use, hypertension, diabetes), study-relevant (i.e., race, socioeconomic status, hepatitis C virus status), and disease-specific (i.e., CD4 nadir, HIV viral load, HIV duration) factors. Further, serum albumin contributed more significantly than periventricular WMH volume to variance in performance on a verbal learning and memory composite score in the HIV group only. Relations in both HIV and AUD groups between albumin and hematological red blood cell markers (e.g., hemoglobin, hematocrit) suggest that in this sample, serum albumin reflects hematological abnormalities. Albumin, a simple serum biomarker available in most clinical settings, may therefore help identify periventricular WMH burden and performance levels in specific cognitive domains in people living with HIV. Whether serum albumin contributes mechanistically to periventricular WMH in HIV will require additional investigation.

Fig. 1. Serum albumin and white matter hyperintensity (WMH) volumes.

a Serum albumin levels in individuals with Alcohol Use Disorders (AUD), infected with HIV, and healthy controls. b Serum albumin levels relative to periventricular (pv)WMH volumes in the 3 study groups. c Serum albumin levels relative to deep WMH (dWMH) volumes in the 3 study groups.

Fig. 2. Variables relevant to white matter hyperintensity (WHM) volumes in HIV.

a Variables in the HIV group contributing significantly to periventricular (pv) WMH volumes included age, sex, serum albumin, and body mass index (BMI). b List of these variables by their relative false discovery rate (FDR) contribution to pvWMH volume in the HIV group.

Fig. 3. Variables relevant to white matter hyperintensity (WHM) volumes in AUD.

a List of variables contributing with false discovery rate (FDR) significance to periventricular (pv) WMH volumes in the Alcohol Use Disorders (AUD) group. b List of variables contributing with FDR significance to deep WMH (dWMH) volumes in the AUD group.

Fig. 4

Relations in the HIV group between serum albumin levels and the visual and verbal memory (VVM) cognitive composite score.

Fig. 5. Result of a principal component analysis (PCA) including 9 variables associated with periventricular white matter hyperintensity (pvWMH) volumes the combined Alcohol Use Disorders (AUD) and HIV groups.

AUD alcohol use disorder, BMI body mass index, VVM visual and verbal memory.

Fig. 6. Result of a principal component analysis (PCA) including 14 blood markers associated with serum albumin levels or HCV.

AST aspartate aminotransferase, GGT γ-glutamyl transferase, IP10 Interferon gamma-induced protein 10, MCHC mean corpuscular hemoglobin concentration, RDW red cell distribution width, TNFA tumor necrosis factor-α.