Astaxanthin

Astaxanthin is a red, fat-soluble xanthophyll carotenoid. Fish and crustaceans cannot make it; they get it from algae and other microorganisms in their diet. That is why salmon, shrimp and crayfish are pink or red [2]. The microalga Haematococcus pluvialis can accumulate it up to 3.8% of its dry weight and is the usual source of supplements. Commercial astaxanthin also comes from Phaffia yeast and from chemical synthesis, and the US FDA approves it as a colourant in animal and fish feed [1]. Unlike beta-carotene, it has no provitamin A activity [1].

The human trial literature is large but made of small studies, many run in Japan or Iran and several funded by manufacturers [4][10]. When those trials are pooled, most of the headline claims shrink.

Oxidative stress and inflammation. A meta-analysis of 12 randomised trials (380 people) found that astaxanthin lowered blood malondialdehyde (SMD −0.95), with some evidence of higher superoxide dismutase and lower isoprostanes in overweight people. It lowered IL-6 in type 2 diabetes but had no significant effect on CRP or TNF-α [3]. Individual trials go both ways. In overweight adults, 12 weeks lowered malondialdehyde and isoprostanes and raised total antioxidant capacity compared with placebo [11]. In middle-aged and older adults, 6 or 12 mg a day lowered phospholipid hydroperoxides in red blood cells [12]. In young Finnish men, 8 mg a day for 3 months lowered plasma hydroxy fatty acids within the astaxanthin group, but the difference from placebo just missed significance (p = 0.056) [13]. And 20 mg a day for 4 weeks in trained cyclists raised plasma astaxanthin without changing antioxidant capacity or malondialdehyde [14]. In the XANTHIN trial, a year of 12 mg a day in 61 kidney transplant recipients changed neither F2-isoprostanes nor arterial stiffness nor inflammation [15].

Blood lipids, glucose and blood pressure. In general populations the pooled evidence is mostly null. A 2015 meta-analysis of 7 trials (280 people) found no effect on total, LDL or HDL cholesterol or on triglycerides, and only a non-significant trend for glucose [5]. A 2020 meta-analysis of 14 trials found only a small rise in HDL (+1.5 mg/dL) and no effect on glucose, HbA1c, LDL, triglycerides, weight or blood pressure. Its overall CRP result was null, with 98% heterogeneity, and the CRP benefit it reports is confined to subgroups [6]. In adults at risk of metabolic syndrome, systolic pressure fell by 4.2 mmHg, a result at the edge of significance (p = 0.05) [16].

The picture is better in diabetes and prediabetes. A 2026 meta-analysis of 9 trials (403 patients) found lower fasting glucose (−16 mg/dL), HbA1c (−0.34 points), LDL (−9 mg/dL) and total cholesterol, and higher HDL. Body weight and HOMA-IR did not change. The caveats are that most of the trials came from Iran, the fasting-glucose result had 98% heterogeneity, and the HbA1c result was rated low certainty [10]. The most thorough single trial, 24 weeks of 12 mg a day in 34 people with prediabetes, missed its primary outcome (insulin sensitivity measured with a glucose clamp). It found falls from baseline in LDL (−0.33 mmol/L) and in fibrinogen, fetuin-A and L-selectin [17].

Skin. Pooling 8 oral trials, a 2021 meta-analysis found higher skin moisture (SMD 0.53) and elasticity (SMD 0.77) but no significant reduction in wrinkle depth. It rated the evidence preliminary because the trials were small and several were run or funded by commercial entities [4]. In one of those trials, run by a manufacturer, 6 mg a day for 6 weeks in 36 men improved crow's-feet wrinkles, elasticity and water loss through the skin [18]. In a 9-week trial of 4 mg a day, the dose of UV needed to redden the skin rose from baseline more than on placebo, but the post-treatment values themselves did not differ significantly between groups (22 people analysed; FUJIFILM sponsored) [19].

Brain. Astaxanthin is marketed for memory, but no trial has shown a significant difference from placebo. In the largest (96 middle-aged adults, 6 or 12 mg a day for 12 weeks), some scores improved within the astaxanthin groups but none differed from placebo [20]. A small trial in mild cognitive impairment combined astaxanthin with sesamin, so it cannot isolate astaxanthin [21]. A 2024 meta-analysis found at most a marginal effect on cognitive accuracy and none on reaction time [22]. A review of trials in older adults also found no statistically supported benefit [23].

Exercise. Two cycling trials were positive: a 20 km time trial improved by 121 seconds (7 completers per group) [24], and a 40 km time trial by 1.2% [25]. A larger trial of 32 trained cyclists on 20 mg a day found no change in fat oxidation or performance [14]. Two trials in resistance-trained men found no effect on muscle damage or soreness [26][27]. A 2026 meta-analysis of 24 trials found no improvement in VO2max, time-trial performance or maximal power (low certainty). Creatine kinase after exercise was lower (SMD −0.45, moderate certainty), which points to a possible effect on recovery rather than on performance [7].

Fertility. In a 30-man pilot trial of 16 mg a day, sperm linear velocity and pregnancy rates were better with astaxanthin, but only 11 men were in the astaxanthin group [28]. A larger trial of 80 men with oligo-astheno-teratozoospermia, on the same dose, found no effect on any semen parameter [29]. A 2026 meta-analysis of the human trials found no significant improvement in semen quality [30]. In women undergoing assisted reproduction, four small Iranian trials pooled to a higher oocyte maturation rate (+8.4 percentage points), but no pregnancy outcome changed [31].

Eyes and immunity. In a manufacturer-run trial in 64 children with digital eye strain, 4 mg a day for 12 weeks improved a symptom questionnaire score more than placebo (−4.0 vs −1.7 points). Visual acuity and accommodation did not differ between groups [32]. In 42 young women, 2 mg a day lowered CRP and a DNA-damage marker and raised several immune responses, while 8 mg a day raised IL-6 and interferon-γ [33].

Astaxanthin has no known receptor. It is a C40 carotenoid: a long chain of conjugated double bonds capped at each end by a six-membered ring carrying a 3-hydroxyl and a 4-keto group [1][2]. That layout makes it polar at both ends and non-polar in the middle, so it can insert into a cell membrane with its polar ends at the surfaces [1][4]. Most of the antioxidant story follows from that position.

In the membrane. In model membranes rich in polyunsaturated fatty acids, astaxanthin preserved bilayer structure and cut lipid hydroperoxide formation by 40%. Lycopene and beta-carotene, which lack polar end groups, disordered the bilayer and increased lipid hydroperoxides by more than 85% [34]. In that experiment, where the carotenoid sat in the membrane decided whether it protected the lipids or harmed them.

Singlet oxygen and radicals. The polyene chain takes up the excess energy of singlet oxygen, and the molecule also reacts with peroxyl and hydroxyl radicals. In vitro, astaxanthin scavenged peroxyl radicals 16% more effectively than alpha-tocopherol. The paired hydroxyl and keto groups can also bind iron and copper and block the Fenton reaction that makes hydroxyl radicals [2]. The claim that it is the strongest carotenoid antioxidant depends on where it is measured. Inside liposome membranes, astaxanthin quenched singlet oxygen more slowly than lycopene or beta-carotene [35], and its rate constants in membranes are far lower than those measured in organic solvents [2].

Signalling. In cell and rodent studies, astaxanthin activates the Nrf2 antioxidant-response pathway. It raises Nrf2 and its targets HO-1, NQO1, SOD and catalase and lowers Keap1, the protein that normally holds Nrf2 in check. It also inhibits NF-κB, the main switch for inflammatory genes, possibly acting on the IKK kinase complex. The review that collects this work says the mechanism is still being worked out [36]. In human trials the anti-inflammatory signal is weak: IL-6 fell in type 2 diabetes, but CRP and TNF-α did not change on pooling [3].

Isomers. The ring carbons at positions 3 and 3′ are chiral, so astaxanthin exists as (3S,3′S), (3R,3′R) and the meso (3R,3′S) form. Each can also form cis (Z) isomers along the chain [2]. Absorption is selective. In men, cis isomers made up a larger share of plasma astaxanthin than of the dose. When a mixture of stereoisomers was given, the (3R,3′R) form rose from 31% of the dose to 54% of plasma astaxanthin [37][38].

Source and stereochemistry. Haematococcus algae make (3S,3′S)-astaxanthin, mostly as mono- and di-esters of fatty acids. Phaffia yeast makes the (3R,3′R) form, and synthetic astaxanthin is a mixture of all three stereoisomers [1][2]. People who ate wild salmon for four weeks carried mostly the (3S,3′S) form in plasma (80%), while those who ate farmed salmon carried mostly the meso form (48%) [2]. EFSA's acceptable intake was derived from a rat study of synthetic astaxanthin and applied to algal astaxanthin too [8]. A safety review argues that the two are chemically different and that synthetic astaxanthin has few human trials of its own [9].

Esters. Esters are split in the gut before absorption; none appear in plasma [38]. In a three-man study of esterified astaxanthin, a tenfold higher dose gave only about 3.5 times the peak level [38].

Delivery systems. Because astaxanthin is very fat-soluble, the vehicle matters. In a study of 32 men given 40 mg, three lipid-based formulations containing polysorbate 80 gave 1.7 to 3.7 times the exposure of a commercial supplement [40]. A micellar product (8 mg, 12 men, crossover) roughly doubled the peak concentration and reached it in 3.7 rather than 8.5 hours. Total exposure (AUC) and half-life were not different, so the micelles sped up absorption rather than increasing it [41]. That study was co-authored by employees of the product's maker, and its plasma levels after 8 mg were higher than those reported elsewhere after 100 mg [37][41].

No safety concern has emerged from the human trials. A 2019 review of 87 human studies of natural astaxanthin found none, including 35 at 12 mg a day or more [9]. In an 8-week placebo-controlled safety trial of 6 mg a day, blood pressure, a full metabolic panel and blood count showed no clinically meaningful change [42]. Trials at 6–16 mg a day for 12 weeks to 6 months reported no treatment-related adverse events [17][20][29].

The regulatory limit comes from rat data. EFSA set an acceptable daily intake of 0.2 mg/kg body weight by applying a 200-fold safety factor to the lowest dose (40 mg/kg a day) that raised the number of multinucleated liver cells in a 2-year rat study; that study used synthetic astaxanthin [8][9]. EFSA found astaxanthin neither mutagenic nor carcinogenic. It concluded that 8 mg a day from supplements is safe for adults even on top of high seafood intake, reaches the limit for 14–17-year-olds, and exceeds it in younger children, by 28% at 10–13 years and up to 524% in young infants [8]. Earlier EFSA opinions used a limit about six times lower [8], and doses approved in different countries range from 2 to 24 mg a day [9].

None of the studies cited here tested a drug interaction in people. In cultured human liver cells astaxanthin induced CYP3A4 and CYP2B6, the enzymes that clear many drugs, but not CYP1A or CYP2C enzymes [39]. In enzyme assays it only weakly inhibited CYP2C19 (IC₅₀ 16.2 µM), far above the concentrations seen in people, and the authors judged a clinical interaction unlikely [45]. Whether the induction seen in liver cells matters at supplement doses has not been tested.