There is a version of aging that everyone recognizes. The dramatic version — the diagnoses, the clinical markers, the numbers that cross thresholds. But most people experience a quieter version first. Recovery from exercise takes a day longer than it used to. Mental sharpness varies in ways that feel new. Energy is present but less reliable, less on demand. Nothing is wrong, exactly. But something has shifted.
The question that functional medicine practitioners and longevity-focused researchers are increasingly asking is whether that shift has a molecular signature — something measurable, upstream of the clinical thresholds, that tracks biological aging independently of chronological age. Something that might appear in blood before disease does.
One family of molecules has been accumulating evidence toward that role for decades, largely outside the mainstream clinical conversation. They are called plasmalogens. They don’t appear on standard lipid panels. Most people have never heard of them. And the data on what happens to them as you age is, by any honest reading, striking.
What makes a plasmalogen different
Plasmalogens are phospholipids — the same broad family as the phosphatidylcholines and phosphatidylethanolamines that make up most cell membranes. But they have one structural feature that sets them apart from every other membrane phospholipid: a vinyl ether bond at the sn-1 position of the glycerol backbone, rather than the ester bond found in conventional phospholipids.
That single chemical difference has profound functional consequences.¹ The vinyl ether bond is highly reactive with reactive oxygen species — it acts as a molecular sacrificial target, intercepting oxidative damage before it reaches the polyunsaturated fatty acids at the sn-2 position and the rest of the membrane. Plasmalogens, in other words, are endogenous antioxidants built directly into the membrane architecture. They also contribute to lipid raft organization, facilitate membrane fusion processes, and are the dominant phospholipid class in myelin, cardiac muscle, and neural tissue — the tissues most dependent on membrane stability and most vulnerable to aging-related deterioration.¹
They are also concentrated precisely where aging shows up first.
Where they come from — and why that matters
Here is what makes plasmalogens unusual among lipid biomarkers: they can only be synthesized in peroxisomes. Not mitochondria. Not the endoplasmic reticulum. Peroxisomes exclusively.
The rate-limiting enzyme in plasmalogen biosynthesis is GNPAT — glyceronephosphate O-acyltransferase — which initiates the ether bond formation inside the peroxisomal membrane.² Downstream steps complete the vinyl ether and attach the sn-2 fatty acid, often a long-chain polyunsaturated fatty acid, particularly DHA. This means that plasmalogen synthesis depends on two things simultaneously: functional peroxisomes and adequate dietary supply of DHA precursors — the omega-3 fatty acids found in marine foods.
Peroxisomal function declines with age. This is not a marginal observation — it is well-documented across tissues, and it accelerates in the presence of accumulated oxidative stress, exactly the condition that aging produces.³ The biological irony is precise: the organelle responsible for building the membrane’s primary antioxidant defense becomes less effective precisely as oxidative demand increases. The system that should compensate for aging is itself compromised by it.
The consequence is measurable. In 1966, researchers reported in Nature that plasmalogen content in human heart and skeletal muscle declines systematically with age and across stages of development — the first direct human tissue evidence of this trajectory.⁴ That observation has been replicated and extended across the decades since, in plasma and in postmortem tissue, in healthy aging and in disease.
Genetics modulates the baseline. Variants in GNPAT and related peroxisomal genes influence biosynthetic capacity independently of diet or lifestyle — meaning that some individuals start with a higher plasmalogen production ceiling and decline from a higher floor. But the trajectory itself appears universal.
What the measurement requires
This is where the clinical invisibility of plasmalogens becomes relevant — and where the series’ recurring theme reasserts itself.
Standard lipid panels cannot detect plasmalogens. There is no enzymatic colorimetric assay for them. The vinyl ether bond that defines them is structurally identical at the lipid class level to the ether bonds in related non-plasmalogen ether lipids, making them indistinguishable without species-level resolution.
Detecting and quantifying plasmalogens requires liquid chromatography-tandem mass spectrometry — LC-MS/MS — with either specific acquisition strategies or chemical derivatization that targets the vinyl ether bond selectively. This is the same platform used for ceramide and acylcarnitine profiling, but with additional analytical complexity: plasmalogen species must be distinguished from their diacyl counterparts, which are isobaric at the class level without the right separation conditions.⁵
A further honest caveat: lipidomics standardization across laboratories remains incomplete. A 2019 commentary in Nature Metabolism noted that many lipidomics studies report relative rather than absolute concentrations, and that methodology varies enough between laboratories to complicate direct comparison of values across studies.⁵ This applies to plasmalogen measurements as much as to any other lipid class.
A practical solution has emerged from recent large-scale work: the plasmalogen score — a composite index reflecting the ratio of ethanolamine plasmalogens to their diacyl phosphatidylethanolamine counterparts. This ratio is internally normalized, partially correcting for inter-individual variation in absolute plasmalogen production capacity, and has been validated across independent large cohorts.⁶ It behaves similarly to the ceramide ratio approach discussed in Article 1: more biologically informative than any single species in isolation.
What the evidence actually shows
The evidence base for plasmalogens spans three levels, each adding a different dimension to the picture.
At the tissue level, a 2011 study in Acta Neuropathologica examined postmortem brains from participants in the longitudinal Vienna Transdanube Aging study, classified by Braak staging for Alzheimer’s disease (AD) neuropathology. Plasmalogen levels in cortical tissue were significantly depleted in brains with advanced Braak staging — the stages characterized by widespread neurofibrillary tangle formation — while very long-chain fatty acids, substrates for peroxisomal beta-oxidation, accumulated.³ The association was stronger with tau tangle burden than with amyloid plaques. Peroxisomal volume density increased in neuronal somata at advanced stages, and peroxisomes were lost from neuronal processes containing abnormally phosphorylated tau — suggesting that impaired peroxisomal trafficking, not just reduced biosynthesis, contributes to plasmalogen depletion in the aging brain.
At the circulating biomarker level, a 2020 study using the Alzheimer’s Disease Neuroimaging Initiative cohort — 1,547 participants — measured ethanolamine plasmalogen indices and derived composite scores reflecting plasmalogen metabolism. Plasmalogen indices were significantly and inversely associated with AD diagnosis versus cognitively normal, with late mild cognitive impairment versus cognitively normal, and with cognitive test scores on both the Alzheimer’s Disease Assessment Scale 13-item cognitive subscale (ADAS-Cog13) and the Mini-Mental State Examination (MMSE).⁷ Cerebrospinal fluid (CSF) total tau was negatively associated with plasmalogen indices — connecting the peripheral blood measurement to central nervous system pathology. Results were replicated in an independent University of Pennsylvania cohort of 112 participants. That same year, a lipidomics study across the Australian Imaging, Biomarkers and Lifestyle Study of Ageing (AIBL) and Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohorts — 1,112 and 800 participants respectively — found concordant peripheral lipidome signatures in AD including ether lipids, identified in both cohorts independently.⁸
These are not small pilot studies. They are replicated findings across thousands of participants in well-characterized research cohorts.
The population-level data broadens the frame beyond neurodegeneration. A 2024 study in eBioMedicine used the Australian Diabetes, Obesity and Lifestyle study — 10,339 participants, nationwide — to develop a plasmalogen score and validate it in the Busselton Health Study (4,492 participants), with 17 years of mortality follow-up.⁶ Individuals in the top quintile of the plasmalogen score had odds ratios of 0.31 for prevalent type 2 diabetes, 0.39 for incident type 2 diabetes, and 0.42 for prevalent cardiovascular disease compared to the lowest quintile — after adjusting for age, sex, and BMI. All-cause mortality risk was 34% lower in the highest versus lowest plasmalogen score quintile (HR 0.66). Critically, the score was significantly associated with diet and lifestyle habits, and a small, placebo-controlled crossover trial (n=10) indicated that shark liver oil supplementation produced a marked change in the score, suggesting the signal is responsive to dietary precursors, though larger intervention studies are required to confirm efficacy.
The NAD⁺ contrast — and what it reveals about the field
Before reaching the interpretation boundary for plasmalogens, it is worth pausing on a molecule that has dominated the longevity conversation far more visibly: Nicotinamide adenine dinucleotide (NAD⁺).
The biology of NAD⁺ in aging is genuinely compelling. Tissue NAD⁺ levels decline with age across species, and that decline is mechanistically linked to impaired mitochondrial function, reduced sirtuin activity, and accumulating cellular stress.¹⁰ The science is real. The investment — in research, in supplement products, in consumer interest — reflects a legitimate biological signal.
But the measurement is where the story becomes complicated.
Quantitative flux analysis published in Cell Metabolism showed that orally administered nicotinamide riboside and nicotinamide mononucleotide — the two most widely used NAD⁺ precursors — are metabolized to nicotinamide in the liver before reaching peripheral tissues.¹⁰ What reaches the bloodstream is largely a breakdown product, not the intact molecule. A 2025 study in Nature Metabolism found that three different NAD⁺ boosters produced differential and largely microbiome-dependent responses in circulating NAD metabolites — the same supplement producing substantially different blood responses in different people depending on gut microbial composition.¹¹ And a 2026 study in Nature Metabolism delivered the most direct evidence yet: whole-blood NAD⁺ levels do not vary with age or with lifestyle interventions.¹²
The supplement market is built on the biology. But the blood measurement does not reflect cellular NAD⁺ status in any straightforward way. Current evidence suggests that whole-blood NAD+ levels do not reflect the age-related declines observed in specific tissues, making blood tests an unreliable proxy for cellular aging.
This is the distinction that runs through this entire series. The biology can be real, and the biomarker can still be unreliable. NAD⁺ is the clearest current example of that gap — enormous commercial momentum, compelling mechanistic science, and a plasma or whole-blood measurement that the field’s own top-tier papers now document as uninterpretable for individual clinical use.
Plasmalogens have the opposite profile. They have a smaller commercial footprint. They are harder to measure. They require specialized lipidomics rather than a simple blood test. But the evidence connecting circulating plasmalogen levels to aging, neurodegeneration, cardiovascular risk, and mortality across large independent cohorts is substantially stronger than the evidence supporting blood NAD⁺ as a clinical biomarker — at this moment, in the published peer-reviewed literature.
What this can and cannot tell you
Here is where honest interpretation is required.
The plasmalogen signal is real, replicated, and extends from tissue pathology through circulating biomarker associations to population-level mortality outcomes. That is a more complete evidence chain than most lipid biomarkers achieve. What is not yet established is causality in humans: are depleted plasmalogens a driver of neurodegeneration and aging-related disease, or a consequence of the same upstream processes — peroxisomal dysfunction, oxidative stress, metabolic deterioration — that drive disease independently?¹³ The animal and cell data support causal roles. Human interventional data is thin.
The only randomized trial of plasmalogen supplementation — 328 participants with mild AD or Mild Cognitive Impairment (MCI), oral scallop-derived plasmalogens over 24 weeks — missed its primary endpoint of cognitive improvement in the total sample.⁹ Significant effects on working memory were seen in subgroups: women with mild AD, and participants under 77 years. The trial was funded by the Japanese Plasmalogen Society. These limitations must be stated clearly. Subgroup effects in a negative trial are hypothesis-generating, not confirmatory.
What the plasmalogen signal can honestly support: it reflects the cumulative biological cost of aging on peroxisomal function and membrane antioxidant capacity, and it associates with aging-related disease burden and mortality risk across large independent population cohorts. It responds to dietary inputs. It is not visible to routine clinical lipid panels.
What it does not yet support: individual-level risk prediction with the same precision as ceramide ratios in cardiovascular populations, or a clearly validated intervention that modifies outcomes by targeting plasmalogen levels directly.
The open questions are specific and answerable with the right study designs: Does dietary supplementation with plasmalogen precursors — alkyl glycerols, DHA-rich marine foods — produce measurable changes in the plasmalogen score in middle-aged populations without established disease? Does a higher plasmalogen score in midlife predict lower dementia incidence in prospective cohorts with adequate follow-up? Do the effect sizes hold across non-European ancestries?
Those studies are within reach. The evidence already in hand is strong enough to make them worth running.
Why this molecule matters — and why it’s been overlooked
Plasmalogens sit at the intersection of three things the aging biology field has identified as central: peroxisomal function, oxidative stress management, and membrane integrity. They decline with age in the tissues most vulnerable to aging. They are measurable in blood. And they are invisible to the panels that most people — and most clinicians — actually run.
That invisibility is not a scientific verdict. It is an analytical limitation. The standard lipid panel was not designed to see them. The tools that can are now widely deployed in research settings and increasingly available in clinical lipidomics contexts.
The number the panel reports is not wrong. It is simply not the whole story — and this molecule is part of what the standard readout cannot tell you about where your cells stand.
Scientific note: This article cites peer-reviewed evidence, clinical guidelines, and reported professional consensus throughout. Sources are independently verifiable and listed below. Claims about individual tests are scoped to the published literature cited and do not constitute medical advice.
José Carlos Bozelli Jr., PhD, is a lipid biochemist, omics data scientist, and scientific writer. He advises biotech, CRO, and health-tech teams on lipidomics, large-scale omics data pipelines, and biomarker science, and translates complex molecular data into decisions for scientists, clinicians, and builders.
The content of this article is for informational and educational purposes only and does not constitute medical advice. Consult a qualified healthcare professional before making decisions based on biomarker results.
References
Dean JM, Lodhi IJ. Structural and functional roles of ether lipids. Protein Cell. 2017;9(2):196–206. DOI: 10.1007/s13238-017-0423-5
Jang JE, et al. Protective role of endogenous plasmalogens against hepatic steatosis and steatohepatitis. Hepatology. 2017;66(2):416–431. DOI: 10.1002/hep.28980
Kou J, et al. Peroxisomal alterations in Alzheimer’s disease. Acta Neuropathol. 2011;122(3):271–283. DOI: 10.1007/s00401-011-0836-9
Hughes BP, Frais FF. Changes in plasmalogen content of human heart and skeletal muscle with age and development. Nature. 1966;215(5104):993–994. DOI: 10.1038/215993a0
Liebisch G, et al. Lipidomics needs more standardization. Nat Metab. 2019;1(8):745–747. DOI: 10.1038/s42255-019-0094-z
Beyene HB, et al. Development and validation of a plasmalogen score as an independent modifiable marker of metabolic health. eBioMedicine. 2024;105:105187. DOI: 10.1016/j.ebiom.2024.105073
Kling MA, et al. Circulating ethanolamine plasmalogen indices in Alzheimer’s disease: relation to diagnosis, cognition, and CSF tau. Alzheimers Dement. 2020;16(9):1234–1247. DOI: 10.1002/alz.12131
Huynh K, et al. Concordant peripheral lipidome signatures in two large clinical studies of Alzheimer’s disease. Nat Commun. 2020;11(1):5698. DOI: 10.1038/s41467-020-19473-7
Fujino T, et al. Efficacy and blood plasmalogen changes by oral administration of plasmalogen in patients with mild Alzheimer’s disease and mild cognitive impairment. EBioMedicine. 2017;17:199–205. DOI: 10.1016/j.ebiom.2017.02.012
Katsyuba E, et al. NAD⁺ homeostasis in health and disease. Nat Metab. 2020;2(1):9–31. DOI: 10.1038/s42255-019-0161-5
Liu L, et al. Quantitative analysis of NAD synthesis-breakdown fluxes. Cell Metab. 2018;27(5):1067–1080. DOI: 10.1016/j.cmet.2018.03.018
Christen S, et al. The differential impact of three different NAD⁺ boosters on circulatory NAD and microbial metabolism in humans. Nat Metab. 2025. DOI: 10.1038/s42255-025-01421-8
Tretowicz MM, et al. Human whole-blood NAD⁺ levels do not vary with age or lifestyle interventions. Nat Metab. 2026. DOI: 10.1038/s42255-026-01537-5
Senanayake VK, et al. Plasmalogen deficiency and neuropathology in Alzheimer’s disease: causation or coincidence? Alzheimers Dement (N Y). 2019;5:524–532. DOI: 10.1016/j.trci.2019.08.003