In a randomized crossover study in healthy volunteers, we observed the highest incorporation of EPA, DHA and total n-3 FA into plasma PL after the ingestion of krill oil (PL and FFA), followed by rTAG derived from fish oil, again followed by EE. Due to high standard deviation values, the differences were not significant. However, a trend was observed for EPA.
Few studies investigated differences in EPA+DHA bioavailability between krill oil and fish oil TAG. One randomized, double-blind, placebo-controlled trial over 4 weeks compared differences in EPA+DHA uptake between krill oil and fish oil TAG by analyzing total plasma FA composition . While subjects on krill oil received 216 mg/d EPA and 90 mg/d DHA, subjects on menhaden oil received 212 mg/d EPA and 178 mg/d DHA. Total plasma EPA and DHA levels in both n-3 FA groups increased similarly. Consistent with our results, the EPA uptake from krill oil was higher compared to menhaden oil. On the other hand, the DHA uptake from menhaden oil was greater than in the krill oil group, probably due to the higher dose given. In contrast to the present study, doses of n-3 FA were not well matched, obviating strong conclusions about bioavailability. Of note, systolic blood pressure declined only in the menhaden oil group, possibly reflecting the fact that DHA, but not EPA, lowers blood pressure .
More recently, Ulven et al. , compared the uptake of n-3 LC PUFA in the form of TAG from fish oil to krill oil in a randomized, parallel group study over 7 weeks, and used total plasma FA composition as an endpoint. Although the EPA+DHA doses used were different (543 mg in krill oil, 864 mg in fish oil), changes in total plasma FA and in a plasma lipid panel were similar. Although, again, differences in dose obviate clear conclusions on bioavailability, the data indicates that absorption of EPA+DHA from krill oil was superior to fish oil TAG, which would be in keeping with our results.
While the studies from Maki et al.  and Ulven et al.  analyzed total plasma FA composition, we used changes in plasma PL FA composition after a single dose treatment as a proxy for bioavailability. There is some disagreement as to what parameter reflects the best bioavailability of LC n-3 FA. In terms of LC n-3 FA, red blood cells reflect tissue composition, at least of the heart, but probably also of other organs . It remains to be demonstrated whether our findings can be reproduced in a longer study focusing on red blood cell FA composition, e.g. by using the omega-3 index.
The mechanisms underlying the larger incorporation of EPA+DHA from krill oil into plasma PL remain unknown and might be due to several reasons. First of all, it was rather surprising that the used krill oil contained a remarkable amount of EPA (22% of total EPA amount) and DHA (21% of total DHA amount) as FFA. It is assumed that the rest of EPA (78%) and DHA (79%) in the krill oil sample is bound to phospholipids, while only negligible contents are bound in TAGs. So far it was generally accepted that EPA and DHA from krill oil is predominantly bound in PL. The unexpected high content of FFA in krill oil might improve the bioavailability of EPA+DHA from krill oil. A greater EPA or DHA bioavailability of FFA compared to rTAG or EE, respectively, has been shown in earlier studies [8, 9, 19], whereas Dyerberg et al.  found the EPA+DHA bioavailability from FFA to be equivalent to natural TAGs. Likewise, a recent study demonstrated a markedly enhanced bioavailability of n-3 FFA over EE in overweight patients on a low-fat diet, which is recommended for patients with hypertriglyceridemia .
On the other hand, the major proportion of EPA+DHA in krill oil is bound in PL, and the intestinal absorption of PL bound FA might be more efficient compared to rTAG and EE. The digestion of FA esterified as PL is carried out mainly by pancreatic phospholipase A2 (pPLA2) and other pancreatic lipases. pPLA2 interacts with PL at the sn-2 position yielding FFA and lyso-phosphatidylcholine, which are absorbed by the enterocytes as parts of mixed micelles . PL exhibit an amphiphilic character and therefore emulsification properties. As a result, PL influence the surface composition of fat droplets, which possibly facilitates the binding of hydrolyzing enzymes and hence the digestion . Similarly, their presence is essential for the formation of mixed micelles. It is possible that higher contents of PL support this formation process, leading to an enhanced absorption of lipids. The activities of other PL hydrolyzing lipases are likely to play a meaningful role in the digestion of PL. Studies with pPLA2-KO mice indicates that pPLA2 deficiency does not affect PL hydrolysis and absorption in contrast to TAG , possibly because its activity is compensated by other PL hydrolyzing enzymes . Thus the hydrolyzing capacity of these lipases might also contribute to an efficient digestion of dietary LC n-3 FA PL.
After absorption into enterocytes, the metabolism of LC PUFA involves re-esterification into TAG (2-monoglyceride pathway) and PL (α-glycerophosphate pathway), as well as the formation of chylomicrons for further transport [25, 26]. It can only be speculated if the higher EPA+DHA plasma PL responses after krill oil ingestion is a result of an intensified incorporation of EPA+DHA into PL in consequence of an increased presence of lyso-phosphatidylcholine. Furthermore PL are required for the formation of chylomicrons, which could be another pathway facilitating the transport of EPA+DHA in the circulating blood. A better understanding of the re-esterification LC PUFA to form TAG and PL, as well as the following integration into lipoproteins requires further investigation.
Strengths and Limitations
Strengths: Almost identical doses of EPA+DHA were used among treatments and the study had a straightforward design. Recently, we demonstrated substantial inter-individual variability in bioavailability of a TAG form of EPA+DHA in a convenience drink . The effect of inter-individual variability is minimized by our randomized crossover study design, while previous studies largely used randomized parallel designs. Limitations: Despite the study design, we observed high standard deviation values and hence no significant differences in plasma PL FA compositions between groups. Furthermore, the study was a single-dose trial yielding no data on safety and tolerability. Further studies with a larger sample size carried out over a longer period are necessary to substantiate our findings. To match the EPA, DHA and total n-3 FA intake, it was necessary to increase the krill oil capsule intake. Hence the subjects treated with krill oil ingested slightly more fat compared to rTAG and EE (approximately one tenth more in total), which could potentially bias our results. Finally, plasma PL is not representative for tissue, a disadvantage that has been already discussed.