Urinary metabolic signatures reflect cardiovascular risk in the young, middle-aged, and elderly populations

Abstract

The predictive value of traditional cardiovascular risk estimators is limited, and young and elderly populations are particularly underrepresented. We aimed to investigate the urine metabolome and its association with cardiovascular risk to identify novel markers that might complement current estimators based on age. Urine samples were collected from 234 subjects categorized into three age-grouped cohorts: 30-50 years (cohort I, young), 50-70 years (cohort II, middle-aged), and > 70 years (cohort III, elderly). Each cohort was further classified into three groups: (a) control, (b) individuals with cardiovascular risk factors, and (c) those who had a previous cardiovascular event. Novel urinary metabolites linked to cardiovascular risk were identified by nuclear magnetic resonance in cohort I and then evaluated by target mass spectrometry quantification in all cohorts. A previously identified metabolic fingerprint associated with atherosclerosis was also analyzed and its potential risk estimation investigated in the three aged cohorts. Three different metabolic signatures were identified according to age: 2-hydroxybutyrate, gamma-aminobutyric acid, hypoxanthine, guanidoacetate, oxaloacetate, and serine in young adults; citrate, cyclohexanol, glutamine, lysine, pantothenate, pipecolate, threonine, and tyramine shared by middle-aged and elderly adults; and trimethylamine N-oxide and glucuronate associated with cardiovascular risk in all three cohorts. The urinary metabolome contains a metabolic signature of cardiovascular risk that differs across age groups. These signatures might serve to complement existing algorithms and improve the accuracy of cardiovascular risk prediction for personalized prevention. KEY MESSAGES: • Cardiovascular risk in the young and elderly is underestimated. • The urinary metabolome reflects cardiovascular risk across all age groups. • Six metabolites constitute a metabolic signature of cardiovascular risk in young adults. • Middle-aged and elderly adults share a cardiovascular risk metabolic signature. • TMAO and glucuronate levels reflect cardiovascular risk across all age groups.

Keywords: Biomarkers; Cardiovascular risk; Early prevention; Elderly; Lifetime risk; Metabolomics.

Fig. 1

Urinary TMAO and glucuronate reflect cardiovascular risk in young adults, middle-aged, and elderly adults. Variation in metabolite abundance variation in the three age-based cohorts investigated (30–50 years, 50–70 years, and > 70 years) is shown. Differences in abundance between control (C), cardiovascular risk factor (RF), and cardiovascular event (CVE) groups are represented as mean ± SEM. *p value < 0.05; **p value < 0.01; ***p value < 0.001; ****p value < 0.0001 (Online Resource 2)

Fig. 2

Cardiovascular risk metabolic signature in young adults. Panel a shows the variation in metabolite abundance between control (C), cardiovascular risk factor (RF), or cardiovascular event (CVE) groups in young adults (30–50 years) represented as mean ± SEM. *p value < 0.05; **p value < 0.01; ****p value < 0.0001 (Online Resource 3). Panel b shows receiver operating characteristic curves including area under the curve (AUC) values

Fig. 3

Correlation between metabolite abundance and lifetime risk. Lifetime risk was estimated using LTR QRISK®, and Spearman correlation was performed. *p value < 0.05; **p value < 0.01; ****p value < 0.0001

Fig. 4

Cardiovascular risk metabolic signature shared by middle-aged and elderly populations. Panel a shows variation in metabolite abundance between control (C), cardiovascular risk factor (RF), or cardiovascular event (CVE) groups in middle-aged (50–70 years) and elderly (> 70 years) cohorts, represented as mean ± SEM (a). *p value < 0.05; **p value < 0.01; ***p value < 0.001; ****p value < 0.0001 (Online Resource 4). Panels b and c show receiver operating characteristic (ROC) curves including area under the curve (AUC) values for cohorts II and III, respectively

Fig. 5

Cardiovascular risk is reflected in urine by metabolic regulation under oxidative stress conditions in young adults. Purine metabolites precursors of uric acid (hypoxanthine and GUAD), intermediates in glutathione synthesis (2-hydroxybutyrate and serine), and molecules involved in counteracting oxidative stress, endothelial dysfunction, or cardiomyocytes apoptosis (guanidoacetate, GABA, GDF15, and ECP) are shown as the main molecular players reflecting cardiovascular risk in urine. Bold letter represents identified urinary metabolites and proteins showing altered levels associated with cardiovascular risk. Arrows represent higher (↑) or lower (↓) variation. ECP eosinophil cationic protein, GABA gamma-aminobutyric acid, GDF15 growth differentiation factor 15, GUAD guanine deaminase, NO nitric oxide