Sage — Fewer Double Bonds
Astaxanthin has its own Kit page: Sage — Astaxanthin. The Fuel pillar it sits under: Sage — Fuel.
FUEL · MEMBRANES
Fewer Double Bonds
Why migratory birds, astaxanthin, and cardiolipin rewrite the fish-oil story
Val Zudans, MD · EnSurf · October 2026
A migrating bird burns fat for hours at a metabolic rate most mammals never reach, and it does not arrive oxidized. That fact sat at the center of the migration essay. One of the adaptations underneath it is easy to miss, because it runs against the supplement aisle: the membranes doing the work are built to leak fewer reactive oxygen species, and a large part of that resistance is fewer double bonds in the fatty acids that make the membrane.
The modern reading of fish oil runs the other direction. More omega-3, more fluidity, more benefit. The bird, the inner mitochondrial membrane, and the molecule that actually folds cristae suggest a tighter claim. Most of what we call an omega-3 benefit is displacement of omega-6 and of oxidized seed-oil residue. Some of what we call a fish benefit is not the oil at all. And the surface area of the mitochondrion is a cardiolipin story, not a DHA story.
The membrane as a peroxidation surface
Polyunsaturated fatty acids are useful and fragile for the same reason. Each double bond is a site a peroxyl radical can abstract a hydrogen from. The more double bonds, the faster the chain reaction, and the closer that chain sits to the electron transport system, the more the leak feeds itself. A. J. Hulbert’s membrane-pacemaker work put a number on this. Combine the fatty-acid composition of a membrane with the relative oxidizability of each chain and you get a peroxidation index. Across mammals and birds, that index runs inverse to maximum lifespan.
Birds are the clean comparison. They have higher mass-specific metabolic rates than similar-sized mammals and, inconveniently for a simple wear-and-tear theory, lower free-radical leak from isolated mitochondria and more peroxidation-resistant membranes. Long-lived seabirds push this further: more monounsaturates, fewer polyunsaturates, a lower peroxidation index in heart membrane lipids, and a lifespan several-fold longer than the fowl they are compared against. The n-6 to n-3 balance matters inside that pattern. Linoleic acid has two double bonds. DHA has six. A membrane shifted toward n-6 and monounsaturates is simply harder to set on fire than one loaded with docosahexaenoate.
This is not a claim that saturated fat is a mitochondrial building material. Storage fat and structural fat are regulated separately. Membrane classes — saturated, monounsaturated, polyunsaturated — are held in a fairly tight range despite diet. What varies, and what the bird appears to have been selected on, is the number of double bonds inside the polyunsaturated fraction, and how close those double bonds sit to the machinery that makes radicals.
The flight trade-off
Migration does not abolish the trade-off. It makes it visible. Unsaturated chains move through water faster and can lower the energy cost of a long flight. In one controlled study, a migratory bird whose fuel stores were richer in omega-6 polyunsaturates spent about 11 percent less energy over six hours of flight, with no immediate rise in oxidative damage. The bill came later. Over the migratory season, the same composition carried higher oxidative damage. Nocturnal long-haul migrants, the birds making the longest single bouts, run lower plasma polyunsaturates than diurnal migrants — mostly less linoleic acid — and a lower calculated susceptibility of those fats to peroxidation. Starlings and white-throated sparrows fed more polyunsaturated fat show higher circulating lipid-peroxidation markers.
The bird solves an engineering problem, not a nutrition-label problem. Enough unsaturation to move fuel. Not so much that the membrane becomes the fuel. Antioxidant upregulation is the other half of the solution. Fatty-acid composition is the half that does not have to be remade every night.
Omega-3 as displacement, not upgrade
None of this makes EPA and DHA useless. It relocates the benefit. On a diet high in linoleic acid from seed oils, arachidonic acid accumulates in membrane phospholipids and feeds the series-2 and series-4 eicosanoids. EPA and DHA compete for the same positions, shift that balance, and give rise to resolvins and protectins. If the background diet is oxidized soybean, corn, and sunflower oil, that displacement is a real repair. It is also most of the signal.
Take the seed-oil excess away and the same molecules look different. DHA incorporates into cardiolipin at the expense of linoleic acid once it becomes a substantial fraction of dietary fat. In mouse heart, replacement of tetralinoleoyl-cardiolipin with a DHA-rich species lowered the activity of respiratory complexes I, IV, and V, not by dissolving supercomplexes but by disrupting the phospholipid domains those complexes use. Putting linoleic acid back rescued the activity. Highly unsaturated n-3 chains also raise the peroxidation index more, per molecule, than the linoleate they replace. In a membrane that was not inflamed by oxidized seed oil to begin with, extra long-chain omega-3 is closer to a cost than to an upgrade.
The supplement trials are consistent with this narrower claim, once oxidation of the capsule is admitted as a confounder. A softgel of ethyl-ester or triglyceride fish oil is a polyunsaturated oil stored at room temperature, often for months, often without the antioxidants that were in the fish. An oxidized omega-3 is not a failed version of the benefit. It is a lipid-peroxidation product you swallowed.
The color in the fish
Whole fish keeps beating isolated fish oil in the gap between observation and trial, and the usual explanations — dose, compliance, the food matrix — are incomplete. Wild salmon, trout, krill, and the crustaceans they eat carry astaxanthin. Standard fish oil does not, unless a formulator adds it back.
Astaxanthin is a ketocarotenoid that spans the bilayer. One end sits toward the aqueous surface, the chain runs through the core, and it quenches singlet oxygen and peroxyl radicals from both faces of the membrane. In membrane models it is far more effective at this than vitamin E or the carotenoids that sit only in the hydrophobic middle. It is also, unlike DHA, not itself a peroxidation substrate of any importance. The practical implication is awkward for the oil industry. Some of the cardiovascular and anti-inflammatory signal attributed to fish may be a membrane antioxidant that never survived extraction, sitting on top of whatever EPA and DHA were doing. Krill oil keeps a little of this, and keeps its omega-3s in phospholipid form. A purified ethyl ester keeps neither.
This does not demote EPA and DHA to zero. It demotes them from the whole explanation. Fish is a tissue. Fish oil is a fraction, and often an oxidized one.
Cardiolipin folds the cristae
The inner mitochondrial membrane is not a generic bilayer that omega-3s loosen into higher performance. Its surface area is a shape problem. Cristae are folds of extreme curvature, packed with protein, and the phospholipid that makes that curvature is cardiolipin.
Cardiolipin is a dimer: four acyl chains, a small headgroup, a cone rather than a cylinder. Cones generate negative curvature. Cardiolipin binds the MICOS contact-site machinery, stabilizes respiratory supercomplexes, and holds the inner membrane in the geometry that gives it its area. Tafazzin mutations, the lesion in Barth syndrome, produce fragmented mitochondria and dysmorphic cristae. You do not get that phenotype from a low-omega-3 diet. You get it from a failure to remodel cardiolipin.
The remodeling is specific, and it is not toward DHA. Nascent cardiolipin is made on the inner membrane from phosphatidic acid, CDP-diacylglycerol, and phosphatidylglycerol, then deacylated and rebuilt. Tafazzin, the main remodeler in heart, preferentially moves linoleoyl chains onto monolysocardiolipin. In human, dog, and rat heart, about 80 percent of cardiolipin is tetralinoleoyl-cardiolipin — four chains of 18:2. Liver is the same pattern, a little less pure. Brain is the exception, richer in oleate with some arachidonate and DHA. The alternate remodeler ALCAT1, which loads DHA and arachidonate onto cardiolipin, produces a more oxidizable species and is associated with mitochondrial dysfunction, not with better cristae.
So the suspicion is the right one. Inner-membrane area tracks cardiolipin abundance and tafazzin remodeling. The preferred acyl chain in the tissues that live and die by respiration is linoleic acid, in controlled amounts. DHA can occupy the same positions, and when it does, it does so at linoleate’s expense.
What you can actually eat
Intact cardiolipin is not a supplement. Phospholipids are hydrolyzed in the gut. Cardiolipin is assembled inside the mitochondrion that needs it, then remodeled from the acyl chains on hand. There is no oral cardiolipin with evidence of delivery to the inner membrane. The levers are the substrate pool and the signal to build more inner membrane.
Heart and liver. Cardiolipin is roughly 12 to 15 percent of heart phospholipid, and about 5 to 6 percent of liver phospholipid, mostly linoleate. Beef heart is the densest ordinary food source, which is why it has been the starting material for mitochondrial preparations for decades. Other organ meats and dark muscle follow. Eggs and fish are rich in phospholipids and poor in cardiolipin.
Linoleic acid, from food, not from a fried oil. Dietary 18:2 tracks the linoleate content of cardiac cardiolipin up to about 20 percent of dietary fat, and is regulated after that. The chain the heart wants is the same chain seed oils flooded the food supply with. The difference is the package. Linoleate in a ruminant tissue or an egg yolk arrives with the phospholipids and antioxidants of that tissue. Linoleate in a heated, repeatedly used seed oil arrives partly peroxidized, and the peroxidation products are the inflammatory signal omega-3s are then asked to clean up.
Whole fish, occasionally, for the pigment. If fish is in the diet, the argument here favors the animal over the capsule: EPA and DHA in phospholipid form, plus astaxanthin, plus the protein matrix, minus the months in a softgel. Astaxanthin itself, from Haematococcus, is the rational isolate if the goal is the membrane antioxidant rather than the fatty acid.
The signal to build cristae. Cardiolipin content rises because inner-membrane area rises. Resistance work, circadian feeding, adequate protein, and the other biogenesis inputs are how you ask for more of the phospholipid. A capsule cannot substitute for the fold.
The surface
EnSurf treats the mitochondrion as the engine and the membrane as the surface the engine runs on. The migration essay was about how a bird fuels a flight without setting that surface on fire. The sequel is narrower. Do not confuse a displacement therapy with a structural upgrade. Do not confuse the oil with the fish. Do not confuse DHA with the molecule that folds the cristae.
Fewer double bonds next to the respiratory chain. Linoleate, in modest whole-food amounts, for the cardiolipin the heart already knows how to build. A membrane-spanning antioxidant if you want what the salmon had and the softgel left behind. The rest of the omega-3 story is mostly the repair of a seed-oil error. The bird never made that error, and its membrane shows it.
Notes
Hulbert and colleagues on membrane peroxidation index and the longevity of birds relative to mammals. McWilliams, Cooper-Mullin, and colleagues on PUFA stores, flight economy, and oxidative cost in migratory birds; Eikenaar and colleagues on lower plasma PUFA and lower peroxidation susceptibility in nocturnal versus diurnal migrants. Cortie and Else, and the 2018 Journal of Biological Chemistry paper on DHA replacing linoleate in cardiac cardiolipin and lowering complex I, IV, and V activity. Schlame and others on tafazzin, tetralinoleoyl-cardiolipin, and cristae morphology in Barth syndrome. Clinical and membrane work on astaxanthin as a bilayer-spanning quencher, present in wild salmon and krill and absent from ordinary fish oil unless added.
A working hypothesis, open to correction and revised as new evidence arrives. Last revised October 17, 2026.
Related: Fuel · Animal fats · Astaxanthin (Kit) · Biogenesis · Coach — Fewer Double Bonds · Surfer — Kit
Related: Sage — A Metabolic Hypothesis of Ocular Surface Disease (Wipeouts · Ophthalmology).
See the notice at the top of this page. Working hypothesis, not medical advice and not a dosing recommendation. No supplement dosing appears here. This page does not tell anyone to stop a fish oil or omega-3 product a clinician prescribed; that decision stays with the prescriber.