Cardiovascular disease is the leading cause of morbidity and mortality in developed countries. Atherosclerotic-associated events such as myocardial infarction and ischemic stroke are increasing annually worldwide,1 and heart failure (HF) is also a growing cause of cardiovascular morbidity and mortality. In some regions, the prevalence of HF is predicted to increase by more than 50% by 2030.2

Consequently, individual cardiovascular risk stratification is recommended and typically assessed using classic lipid profile components such as total cholesterol, triglycerides, low-density lipoprotein (LDL) cholesterol (LDL-C), and high-density lipoprotein (HDL) cholesterol (HDL-C).3 However, this classic risk stratification accounts for atherosclerotic-mediated diseases but not for HF. Moreover, although high levels of LDL-C and low levels of HDL-C are known risk factors for future cardiovascular events, statin therapy does not abolish the total cardiovascular risk, even when LDL-C is normalized. In addition, attempts to reduce cardiovascular risk by increasing circulating levels of HDL-C have been unsuccessful.4–7 As a result, interest has focused on studying the detailed lipoprotein profile rather than total cholesterol, triglycerides, or LDL-C and HDL-C alone. Lipoprotein tests based on 2-dimensional (2D) diffusion-ordered 1H-magnetic resonance (MR) spectroscopy have emerged as novel tests for analyzing lipoprotein parameters beyond LDL-C, including particle concentration (-P), composition, and size.8,9 Indeed, LDL-P, HDL-P, and their subfractions have been described as better predictors of cardiovascular risk than LDL-C or HDL-C.10–13 However, studies so far have focused mainly on atherosclerotic risk and not in the HF population, and the few that have examined changes in the lipoprotein profile in HF have largely involved the HDL-P and HDL subfractions, globally suggesting that they are involved in risk of HF, especially when patients exhibit a decrease in small HDL particle subfraction.14–17 To date, knowledge is lacking on how the entire lipoprotein profile is affected in ambulatory patients with chronic HF.

The aim of our study was to describe the profile of a complete lipoprotein analysis using 2D diffusion-ordered MR spectroscopy in a population with chronic HF and to compare these findings with those in a clinically matched cohort without HF. To determine whether differences could also be detected with disease progression, we analyzed the correlation of the observed results with New York Heart Association (NYHA) class and classic cardiac biomarkers of inflammation, fibrosis and remodeling (interleukin-1 receptor-like 1 [ST2]), injury (high-sensitivity troponin T [hs-TnT]), and cardiac stretch (N-terminal pro-brain natriuretic peptide [NT-proBNP]).

This study provides a detailed description of a lipoprotein profile analyzed by 2D diffusion-ordered MR spectroscopy in a real-life cohort of outpatients with chronic HF compared with a cohort matched for age, sex, diabetes, and statin therapy. Our study indicates that significant derangements in lipoprotein profile are present in chronic HF. In this disease, serum total cholesterol concentrations were significantly lower than in controls, mainly because of lower LDL-C and HDL-C. Patients with chronic HF also had higher values of all 3 lipoprotein particle sizes (VLDL-Z, LDL-Z, and HDL-Z) and lower overall LDL-P and HDL-P, the latter because of a lower small HDL subfraction. Finally, the observed disruption was consistent regardless of NYHA functional class, with a trend to worsening with increasing classification.

Serum total cholesterol concentrations, LDL-C, and LDL-P were lower in the cohort with chronic HF. A possible explanation is that total cholesterol and LDL may have declined because of advanced chronic HF. The progressive decrease in total cholesterol concentrations as NYHA functional class worsened could align with this hypothesis. Moreover, low total and LDL cholesterol levels have been previously associated with a worse prognosis and higher overall mortality in patients with HF, regardless of etiology.23 However, this last explanation, known as the “cholesterol paradox,” although plausible, is usually associated with cachexia.24,25 In contrast, we found differences in cholesterol concentrations between patients with HF and controls in the overweight and obesity ranges of BMI but not with BMI <25. Alternatively, in vitro studies have suggested that circulating cytokines, acting as causative agents of increased mortality, may decrease lipoprotein particle and cholesterol levels by dampening hepatic lipoprotein production and increasing LDL receptor activity. If so, LDL would be considered an independent predictor of mortality and not only a marker of nutritional status.26,27

The overall lower LDL-P and HDL-P observed in the HF cohort, with the latter attributable to fewer small HDL-Ps, was also reported by Potočnjak et al. and Hunter et al.14–16 Although both authors used different methods to analyze small HDL particles, they correlated this finding with increase mortality in patients with acute HF and in a cath–lab-based HF population, respectively. Likewise, we have recently reported an association of lower levels of the small HDL subfraction with cardiovascular death in patients with HF.17 Here we observed a progressive change in these parameters with worsening NYHA functional class and a significant positive correlation of total HDL-P and small HDL subfraction with left ventricular ejection fraction. These findings, along with a negative correlation of these disruptions with NT-proBNP and ST2 levels, suggest a probable association not only with HF itself but also with the severity of the disease. This observation aligns with previous reports.15,28 Overall, these results strongly imply the importance of HDL in HF. Whether these derangements in HDL are simply a marker in HF or are relevant in the pathophysiological process remains to be elucidated. It should be considered that HDL particles may exert beneficial effects on overall myocardial function. Experimental reports have demonstrated that HDL may protect against ischemia and reduce myocardial remodeling.29 Moreover, small dense HDL particles may exert anti-inflammatory properties in endothelial cells and inhibit oxidized LDL-induced apoptosis.30

Regarding lipoprotein particle size, we observed a larger size of all 3 (VLDL, LDL, and HDL) particles in patients with chronic HF compared with the controls. Triglycerides usually drive changes in lipoprotein diameter, but the 2 groups had fairly similar plasma triglycerides, suggesting that the larger size is a specific feature of chronic HF. Overall lipoprotein particle size has been associated with atherosclerotic cardiovascular risk,31 but described as mainly mediated by more large HDL-P and small dense LDL-P,31,32 which does not concur with our observations. In our cohort with chronic HF, the larger amounts of LDL-Z and HDL-Z were secondary to the lower values for the small LDL and HDL subfractions, respectively. As for the HDL particle count, HDL-P size has previously been positively associated with overall cardiovascular risk,32 all-cause mortality,16 and cardiovascular death in HF.17 As with our findings for small HDL, we also observed a positive gradation of VLDL-Z, LDL-Z, and HDL-Z with NYHA functional class. Hunter et al.16 reported a similar positive association of HDL-Z with degree of left ventricular dysfunction, finding the smallest HDL-Z in patients without HF and the greatest for those with HF with reduced ejection fraction, while those with HF and preserved ejection fraction fell in between. Moreover, in this study, we observed a positive association of HDL-Z with cardiac biomarkers, emphasizing a probable relationship of HDL-Z with both the presence and severity of HF.

Previous studies have explored the HDL-C/P ratio as a surrogate of the cholesterol content.17,22,33 Evidence demonstrates that cholesterol-overloaded HDL-P is associated with the progression of carotid atherosclerosis in a disease-free population and with cardiovascular mortality in a chronic HF population. In the present study, we found that patients with chronic HF had cholesterol-overloaded HDL-P compared with controls. As for the HDL-P concentration, in the small subfraction of HDL and HDL-Z, the HDL-C/P ratio appeared to be important in terms of both the presence and the severity of the HF process. The cholesterol overloading of HDL-P in HF is not well understood, and whether it is secondary to an alteration in HDL remodeling, cholesterol transport/efflux of HDL or to an anomalous clearance of large HDL needs to be further explored. The detailed pathophysiological mechanism of this finding in HF should be investigated.

Prior evidence is controversial regarding the influence of the cholesterol content of the HDL particle and the particle number of HDL-C with cardiovascular risk in long-term follow-up and it could be explained by the complex physiopathological mechanisms involved in its metabolism.34 The question of whether all differences observed in our study in the HF population have a true impact on long-term cardiovascular outcomes deserves more attention in future research.

In an outpatient population with chronic HF, significant derangements in serum concentrations of total cholesterol and several lipoprotein particles were found in comparison with a matched control group. A detailed description of each variable has been presented so that the findings could serve in future as reference values in chronic HF. Overall, the population with chronic HF had lower serum total cholesterol concentrations and lower LDL-P and HDL-P concentrations, with the HDL-P findings being mediated by a significantly lower small HDL subfraction. Lipoprotein particle size was significantly larger in chronic HF, and these patients also had an elevated HDL-C/P ratio. The final causes for the differences observed in lipoprotein profile in HF are probably multifactorial, and should be elucidated in future research.

Of note, all HDL derangements detected in the patients with chronic HF were consistent regardless of NYHA functional class or BMI, although with a significant trend to worsening with worsening NYHA functional class. This pattern suggests that HDL particles should be further interrogated in the setting of HF to better understand their pathogenic role.

This work was supported by grants from Fundació La MARATÓ de TV3 (201502 and 201516 to A. Bayés-Genís., 201602-30-31 to N. Alonso and J. Julve); the Ministerio de Educación y Ciencia (SAF2014-59892 to A. Bayés-Genís); AdvanceCat (2014-2020 to A. Bayés-Genís); the Ministerio de Economía y Competitividad (MINECO) - Instituto de Salud Carlos III (ISCIII) (PI17-00232 to J. Julve, PI17-01362 to N. Alonso, PI15-00625 to D. Mauricio and RED2018-102799-T to J. Julve); Centro de Investigación Biomédica en Red Enfermedades Cardiovaculares. (CIBERCV, CB16/11/00403 Centro de Investigación Biomédica en Red de Diabetes y Enfermedades Metabólicas Asociadas, (CIBERDEM, CB15/00071 and CB07/08/0016), an initiative from ISCIII, Spain, with cofunding from the European Regional Development Fund (ERDF) and ISCIII Miguel Servet II Program (CPII18/00004 to J. Julve).

A. Teis, Germán Cediel, N. Amigó, J. Julve, D. Mauricio, N. Alonso, and A. Bayés-Genís contributed to the conception and design of the work. A. Teis, N. Amigó, E. Castelblanco, J. Ribalta, M. Guardiola, J. Franch, M. Bermúdez-López, P. Codina and J. Lupón contributed to data acquisition for the study. A. Teis and G. Cediel performed the analysis and interpretation of the data. All authors participated actively in drafting the manuscript and revising it critically and approved and agreed on the final version.

A. Teis and E. Castelblanco contributed equally to this article.

N. Amigó owns stock in Biosfer Teslab and has a patent related to the lipoprotein profiling described in the present manuscript. The remaining authors have nothing to disclose. The funders had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.