The search for novel biomarkers with greater specificity, sensitivity, and predictive power has gained considerable momentum over recent decades, driven by technological advances and the promise of more precise medical care. Yet before any molecule can be established as a meaningful biomarker, fundamental questions must be addressed: how does it vary with age, sex, hormonal status, diet, physical activity, or fitness? Only by defining what is expected in healthy individuals can we meaningfully interpret whether a given value is truly “high” or “low”.
This question was at the heart of our study on serum sphingolipids across the adult lifespan. Recent advances in lipidomics have positioned sphingolipids, and ceramides in particular, as promising markers of cardiometabolic risk. Specific ceramide species have already been integrated into clinical risk scores that may improve cardiovascular risk prediction beyond conventional measures such as total cholesterol. Beyond ceramides, some sphingomyelins have also been associated with potentially favourable cardiometabolic profiles. However, interpreting these emerging lipid biomarkers requires understanding how they vary in clinically healthy individuals, not only between people with and without disease, but also across age, sex, and fitness levels.
Our study emerged from the COmPLETE-Health project, a population-based study conducted at the University of Basel (Switzerland). COmPLETE stands for CardiOPulmonary Exercise TEsting, and the broader project was designed to investigate how cardiorespiratory fitness relates to vascular health, muscle function, blood biomarkers, and healthy ageing (https://www.complete-project.ch/english/). This made it an ideal setting to ask whether sphingolipids, increasingly discussed as cardiometabolic biomarkers, also follow predictable patterns across adulthood in people without overt disease.
For this study, we analysed serum sphingolipids in 522 clinically healthy adults aged 20 to 91 years from the COmPLETE-Health cohort. Using targeted liquid chromatography-tandem mass spectrometry, we measured a panel of sphingolipids and examined how their concentrations varied with age and sex, while accounting for cardiorespiratory fitness and other relevant factors. We also modelled age- and sex-specific percentile curves across adulthood (https://www.nature.com/articles/s43856-026-01565-y).
The central finding was clear: many serum sphingolipids increase with age. Fourteen of the 21 detected sphingolipids were positively associated with age in both sexes. This included all four ceramides used in cardiovascular event risk tests: Cer16:0, Cer18:0, Cer24:0, and Cer24:1. In contrast, the corresponding ceramide ratios were not significantly associated with age, suggesting that absolute concentrations and ratios may capture different biological information. We also observed sex differences for several sphingomyelins, ceramides, HexCer18:0, and the Cer16:0/Cer24:0 ratio.
One of the motivations for this work came from a simple translational concern. In clinical practice, many measurements are interpreted in context. Blood pressure, lung function, bone density, and cardiorespiratory fitness are not assessed in isolation; they are usually considered in relation to age, sex, body size, or other individual characteristics. We felt that emerging lipid biomarkers should be held to the same standard. Sphingolipid concentrations measured in a 30-year-old man and a 75-year-old woman may not carry the same meaning, even if the analytical values are identical.
This is why we believe that precision medicine requires more than increasingly sophisticated molecular measurements. It also requires precision context. Without understanding the expected biological variability of a biomarker in healthy individuals, we risk mistaking normal ageing for pathology or overlooking meaningful deviations in younger individuals.
The study was also an opportunity to bring together different perspectives. COmPLETE-Health was an interdisciplinary project involving exercise physiologists, physicians, and statisticians. For this lipidomics work, expertise in sport and exercise medicine, cardiometabolic health, analytical chemistry, and biostatistics had to come together. The biological question was not simply whether sphingolipids change with age, but how such changes should be understood in relation to health, fitness, and future clinical interpretation.
A particularly interesting aspect was the role of cardiorespiratory fitness. Fitness is one of the most powerful markers of overall health, but its relationship with sphingolipid levels was not uniform. In our analysis, associations varied by lipid species, suggesting that the links between exercise capacity, lipid metabolism, and cardiometabolic risk are unlikely to be captured by a single molecule or score. Instead, lipidomics may eventually help us understand more subtle metabolic signatures of health and ageing.
Our study does not provide definitive clinical cut-offs, nor does it claim that sphingolipid measurements are ready for routine clinical use in every setting. Rather, it offers a first map of how selected serum sphingolipids vary across adulthood in a clinically healthy Swiss population. We hope these data will help future studies, risk scores, and clinical applications interpret sphingolipid biomarkers with greater nuance.
For us, the broader lesson is that precision medicine is not only about measuring more molecules. It is also about understanding the person behind the measurement: their age, sex, fitness, and biological context. Before biomarkers can become reliable clinical tools, we need to know what healthy variability looks like. In other words, before precision medicine, we need precision context.