Cardiovascular risk in adolescents: why cholesterol does not tell the whole story
A study involving Biosfer Teslab reveals markedly different cardiometabolic profiles in adolescents with anorexia nervosa, normal weight and obesity. Advanced lipoprotein profiling by nuclear magnetic resonance provides information beyond conventional cholesterol measurements.
Two nutritional extremes, different cardiovascular profiles
Anorexia nervosa and obesity are very different clinical conditions, yet both can have important cardiovascular consequences during adolescence. Assessing them solely through total cholesterol or LDL cholesterol (LDL-C) can overlook relevant information about lipoprotein particles, inflammation and cardiac structure.
A study published in Clínica e Investigación en Arteriosclerosis compared the clinical, lipid and metabolomic profiles of adolescents across three weight categories. The research is the result of a multidisciplinary collaboration between hospitals, research institutions and Biosfer Teslab.
Publication authors and affiliations
Authors: Jose M. Siuranaa,b, Neus Martínez-Micaeloc, Sara Saminoc, Cristina Marimond, Rosa Collelld, Cristina Girbaua,b, Paula Sol Venturae, Enric Ortigosaa,b, Zelmira Boscha,b, Nuria Sanza,b, Anna Sabaté-Rotésf, Diego Yesteg, Silvia Fernándeza,h, Vanessa Sanchez-Gistaui,j, Anna Bonadak, Núria Planal, Albert Feliuk and Núria Amigóc,m.
Affiliations (as listed in the original article):
- a. Hospital HM Nens, HM Hospitales, Barcelona, Spain.
- b. Instituto de Investigación Sanitaria, HM Hospitales UCJC, Spain.
- c. Biosfer Teslab, Reus, Tarragona, Spain.
- d. Departamento de Cardiología Pediátrica, Hospital Sant Joan, Reus, Tarragona, Spain.
- e. Hospital Arnau de Vilanova, Lleida, Spain.
- f. Departamento de Cardiología Pediátrica, Hospital Universitari Vall d’Hebron, Universitat Autònoma de Barcelona, Barcelona, Spain.
- g. Departamento de Endocrinología Pediátrica, Hospital Universitari Vall d’Hebron, Universitat Autònoma de Barcelona, Barcelona; CIBERER, ISCIII, Madrid, Spain.
- h. Centros Ita, Barcelona, Spain.
- i. Unitat referencia Trastorns Conducta Alimentaria, Institut Pere Mata, Tarragona; Institut d’Investigació Sanitària Pere Virgili-CERCA, Reus, Tarragona, Spain.
- j. CIBERSAM, Centro Investigación Biomédica en Red Salud Mental, Instituto de Salud Carlos III, Madrid, Spain.
- k. Departamento de Endocrinología Pediátrica, Hospital Sant Joan, Reus, Tarragona, Spain.
- l. Unitat de Medicina Vascular i Metabolisme (UVASMET), Hospital Universitari Sant Joan, IISPV, CIBERDEM, Universitat Rovira i Virgili, Reus, Tarragona, Spain.
- m. Basic Medical Sciences, Faculty of Medicine, Universitat Rovira i Virgili, Institut Investigació Sanitaria Pere Virgili, Reus, Tarragona, Spain.
Biosfer Teslab contributors: Neus Martínez-Micaelo, Sara Samino and Núria Amigó.
This is a cross-sectional observational study: it describes associations and differences between groups but cannot establish causality or predict future cardiovascular events.
What does nuclear magnetic resonance add?
In addition to routine laboratory tests, the researchers analysed blood samples using proton nuclear magnetic resonance spectroscopy (¹H-NMR). Advanced lipoprotein profiling, performed using LipoScale® technology, enabled the characterisation of VLDL, LDL and HDL particles by concentration, size and composition.
The study also examined several lipid families, including omega-3 fatty acids and DHA, as well as inflammation-related plasma glycoprotein biomarkers such as GlycA. These data were integrated with clinical and echocardiographic measurements.
Advanced lipoprotein profile
Number, size and distribution of LDL, HDL and VLDL particles, beyond cholesterol concentration.
Metabolism and inflammation
Fatty acids, lipid species and markers such as GlycA to contextualise metabolic differences.
The findings: similar LDL-C does not mean the same lipoprotein profile
In the anorexia nervosa or underweight group
Participants in this group had higher total cholesterol and LDL-C than normal-weight participants. However, NMR analysis revealed a larger mean LDL particle diameter, along with more medium-sized HDL particles and higher omega-3 and DHA concentrations. This pattern differs from the typically atherogenic profile observed in obesity.
Cardiac findings in this group included a lower heart rate, lower systolic blood pressure and a lower indexed left ventricular mass.
In the moderate-to-severe obesity group
The researchers observed a pattern of higher triglycerides, more large VLDL particles and more small LDL particles, together with lower concentrations of HDL particles and omega-3 fatty acids. Higher GlycA and other inflammatory markers were also identified, alongside increased left ventricular mass.
| Parameter (median) | Underweight / AN | Normal weight | Obesity |
|---|---|---|---|
| LDL-C (mg/dl) | 114 | 102 | 107 |
| LDL diameter (nm) | 21.3 | 21.1 | 21.0 |
| Small LDL particles (nmol/L) | 569 | 560 | 623 |
| GlycA (µmol/L) | 547 | 612 | 806 |
| Indexed left ventricular mass (g/m²·⁷) | 22.1 | 30.8 | 46.4 |
Descriptive data taken from Tables 1–3 of the original article. Between-group comparisons should be interpreted with reference to the reported significance values and clinical context.
An essential distinction: although the lipoprotein profile in the anorexia nervosa group displays less atherogenic characteristics, this does not mean anorexia nervosa is cardiovascularly safe. Bradycardia, hypotension and other cardiac complications still require specific clinical assessment.
Cholesterol concentration is only part of the picture
Understanding how many lipoprotein particles circulate, their size and their distribution can help interpret metabolic profiles that conventional blood tests may not adequately distinguish.
Integrating clinical and metabolomic data
Using multivariate PLS-DA models, the researchers observed separation among the three groups. Body mass index (BMI) was one of the most influential variables, alongside left ventricular mass, several lipid components and omega-3 fatty acids.
The model integrating all data blocks achieved a balanced classification error of 0.11 in repeated cross-validation, lower than the errors of models based on individual data blocks. This finding supports the value of data integration in research, but does not constitute validation as a diagnostic test for routine clinical use.
What does this mean for paediatric cardiovascular research?
The findings show that two adolescents with similar LDL-C values may have very different particle distributions and inflammatory profiles. Advanced lipoprotein profiling can therefore complement clinical information and support a more detailed characterisation of cardiometabolic risk.
These findings should be interpreted cautiously: the sample was relatively small, the design was cross-sectional and the sex distribution differed across groups. Longitudinal studies and further validation are needed to establish prognostic value and potential clinical applications.
Biosfer Teslab’s contribution
In this study, Biosfer Teslab contributed advanced sample analyses using NMR-based technologies, including lipoprotein characterisation and profiling of lipids and inflammation-related biomarkers.
This research illustrates the potential of metabolomic tools to connect molecular data with clinical phenotypes and advance a more precise understanding of cardiovascular health from an early age.
Scientific reference
Siurana JM, Martínez-Micaelo N, Samino S, Marimon C, Collell R, Girbau C, et al. Cardiovascular risk and metabolomic profile in adolescents: A comparative study between anorexia nervosa, normal weight, and obesity. Clínica e Investigación en Arteriosclerosis. 2026. DOI: 10.1016/j.arteri.2026.500955.
Read the scientific article ↗