Friday, December 9, 2011

Some Things Fishy: Oxidized Fish Oil Totally Benign!?Plus: The Inflammatory Side of EPA and Peroxide & Alkenal Levels in Commercial Fish and Vegetable Oils.

Image 1: Surströmming, a Swedish delicates is essentially rancid fish and it stinks exactly like that. Now, the results of a recent study show that the rancidity does probably not compromise the health benefits of the fish... so if you like it, go for it!
You know that whenever something is so (over-)hyped like fish oil or vitamin D that rings an alarm with me and when I hear "experts" on popular podcast say things along the lines of "as long as you take your fish oil that can compensate for a whacky diet", this is totally burning me up. Yes, there is conclusive evidence that for someone who has damaged his/her body by years and years of omega-6 over-consumption the inclusion of even "high" dose (I consider 5-6g high!) fish oil supplements can make sense, but NO, it will neither allow you to keep eating the same crap that has brought you to where you are at now, nor (and I think this is even more important for most of the SuppVersity readers) is there conclusive evidence that a healthy, active and lean human being is not way better off by limiting his total PUFA intake instead of popping grams of highly oxidizable n-3 fatty acids from fish oil caps.

Highly oxidizable? Yes! Dangerous? Surprisingly not!

A pros pos "highly oxidizable", the argument that polyunsaturated fatty acids (PUFAs) are readily oxidized not only in your body, but even at the shelves of your nutrition store, is one of the few possible caveats of fish oils supplementation even fish oil enthusiasts will acknowledge. After all previous animal studies have shown that diets rich (5%) in rancid (=oxidized) fish oils lead to increases in thiobarbituric acid-reactive substances (TBARS) levels and elevate liver specific transaminases, as well as the alkaline phosphatase (ALP) levels in the plasma of rats (detrimental effects which can by the way be ameliorated by taurine supplementation, cf. Hwang. 2000). The results of a recent study by Inger Ottestad and colleagues from Norway may thusly surprise the "pro-fish oil"-faction about as much as they surprised me (Ottestad. 2011): The ingestion of 8g of oxidized (peroxide value: 18mEq/kg; ansidine value: 9) fish oil (1.6g EPA+DHA) did not have any unfavorable short term-effects in previously healthy individuals.
Figure 1: Serum (left, 8-iso on secondary axes was measured in urine) and erythrocyte (right, GPx on secondary axes) markers of oxidative stress in 68 healthy subjects who were randomly assigned to ingest 8g of "fresh" fish oil, oxidized fish oil or high oleic-acid sunflower oil per day for before (pre) and after (post) the 7 week intervention (data adapted from Ottestad. 2011)
If you take a closer look at the measured levels of serum (4-hydroxy-2-hexenal: 4-HHE, 4-hydroxy-2-nonenal: 4-HNE, alpha-tocopherol,  high-sensitive C-reactive protein: hsCRP and 8-iso-PFG2a, the latter in urine) as well as erythrocyte (total GSH, 4-hydroxy-2-nonena: GR, CAT and glutathione peroxidase: GPx) markers of oxidation before and after the 7-week intervention (cf. figure 1), it is quite obvious that there were no statistically significant oxidation-related changes in the concentrations of the measured markers of oxidative stress, of which the scientists state that they are the current, yet debatable, "gold standard" for in vivo studies.
Figure 2: Changes in n-3 and n-6 levels and the n-6/n-3 ratio (small graph) in the course of the study period (data calculated based on Ottestad. 2011)
It is thus not really surprising that both fish oil groups experienced virtually identical (and highly favorable) -50% reductions in the ratio of omega-6 (n-6) to omega-3 (n-3) fatty acids. Moreover, ...
[a]fter 3 and 7 weeks of intervention, the plasma level of EPA, docosapentaenoic acid and DHA were significantly increased in both fish oil groups compared to the HOSO group, but no significant difference in EPA, doc-osapentaenoic acid and DHA between the FO and oxFO groups was observed.
The scientists are thusly right to conclude that their results do not support the often-heard hypothesis that higher intakes n-3 long-chain fatty acids could increase in vivo lipid peroxidation and more importantly, that ...
[...] the content of hydroperoxides in fish oil supplements, even with a PV that exceeds the European Pharmacopeia for marine n-3 oils, does not apparently influence the plasma level of n-3 FA.
With regards to the obvious differences to previous animal studies, the scientists state that secondary oxidation of hydroperoxides, which are then absorbed in the intestine has until now been observed in animal and cell studies. In view of the relative short duration of the study and the reliance on healthy subjects, it is also questionable whether identical results would have been achieved, when sick patients (the usual customer group at least for the pharma-grade n-3 supplements) had been treated with the same product for years.

Oxidized fats in fish oil and beyond

It is also worth mentioning that Ottestad et al. are not sure, whether their "aritifically oxidized" fish oil (oxidation was achieved by sparkling pure oxygen through the oil for 20 min twice a day for 21 d) was an appropriate model for commercially available (oxidized) fish oils. After all, there could be major differences in the composition of the oxidation products, when the oils go rancid over months or get damaged by heat etc. While I obviously cannot answer this question without setting up my own lab, I can however tell you that another recent study by Halvorsen et al. who examined the peroxide and alkenal (one of the major products of secondary oxidation) content of fish and vegetable oils, found average peroxide levels in 33 commercially available fish oil products (mean PV: 3.61mEq/kg) that were ~500% below the ones of the oxidized fish oil (18mEq/kg) in the Ottestad study.
Figure 3: Mean peroxide and alkenal values of 33 commercially available fish and 35 vegetable oils (Halvorsen. 2011).
In this regards, fresh vegetable oils, obviously are way in front, as the data from a study by Bente Lise Halvorsen and Rune Blomhoff clearly shows, that they have lower peroxide and much lower alkenal levels than fish oils (cf. figure 3). Interestingly, vegetable oils are also less prone to being oxidized during storage, something Halvorsen and Blomhoff conclude based on the absence of the "negative correlation (r=−0.557, p<0.001) [...] between the number of days until expiry and the PV [peroxide value]" they observed in the marine omega-3 oils.
Figure 4: Peroxide (PV in mEq/kg) and alkenal (in nM/ml) levels in fresh vegetable oils and after being heated for 25 minutes at 225°C in an oven (data adapted from Halvorsen. 2011); solid red line - maximal peroxide value for olive oils, dotted red line - maximal peroxide values for fish oils as suggested by Turner et al. (Turner. 2006)
Contrary to fish oils, which are usually taken "fresh" and in a capped form, the main fate of vegetable is however to be (ab-)used as cooking / frying oils. During the heating process, the amount of secondary lipid oxidation products, the alkenals, doubles or quadruples depending on the type of oil (cf. figure 4). In that, it may at first seem counterintuitive that, when the scientists heated the samples for 25 minutes at 225°C in an oven, the amount of primary oxidation products was slightly reduced in most, but not all (e.g. soy bean oil) of the 11 vegetable oils. If you do yet take into consideration that the latter are the "raw material" for the secondary oxidation products, it becomes quite clear that this is not a desirable process ;-)
Image 3: Extra virgin olive oils (EVOOs) have generally higher peroxide values than the cheap refined stuff, and yet, EVOOs and not refined oils have been shown to exhibit numerous health benefits.
Putting peroxide values (PV) into perspective: All potential health hazards aside, it may be interesting to know that the general "rule of thumb" says that a fat is rancid when the PV is about 10 meq/kg (the fish oil in the study with PV=18 was thusly "rancid"). A fresh and refined product on the other hand should have PV below 1 meq/kg (Gunstone. 1996). That being said, it may surprise you that for high quality extra virgin olive oils, the PV limit is 20meq/kg, while for "regular" olive oil it is only 10 meq/kg. If you know look at studies related to the health benefits of refined vs. extra virgin olive oil, you will have to admit that - quite obviously - fish oil apparently is not the only oil, where increased peroxide levels do not negate the beneficial health effects of the oil.
So, if pure vegetable oils are generally "fresher" than fish oils does that mean that as long as you do not heat them, they are the better choice? No, they are not! I mean, look at the research that is out there... the abundance of n-6 fatty acids in the "healthy" vegetable oils that are getting pimped especially by the US government, is at the heart of an epidemic of which the authorities still claim that it was caused by high cholesterol levels. Instead banning all saturated fats from YourPlate (which should never look like the governments MyPlate ;-), you should rather incorporate more coconut oil and saturated fats from butter, beef etc. into your diet. Select (vegetable) oils that are relatively high in mono-unsaturated fatty acids, like extra virgin olive oil (don't care about its high peroxide value, cf. red box above) and try to reduce the amount of n-6 fats you ingest - you will get more than enough even from grass-fed meats, olive and other oils and any processed foods that may still be part of your diet, anyway.
Image 4: Not all Omega-3 are created equal. We know for some time that DHA (not EPA) is what your brain needs and a recent study from Norway suggest that eicosapentaenoic acid (EPA) is actually pro- not anti-inflammatory at a cellular level. It may yet well be that this in turn triggers a beneficial hormetic response which would support my "fish oil = exercise in a pill hypothesis"
Although this is not directly related to the topic of oxidation I still want to add that another study appears to confirms my long-cherished skepticism towards EPA (most fish oils have a 2:1 EPA to DHA ratio), which, as a recent study from Norwegian scientists shows (Myhrstad. 2011), is not really "beneficial super-antioxidant" people are led to believe. In their trial the scientists fed 14 healthy female volunteers test meals. The cakes the participants ate were enriched with either flaxseed, cod liver or coconut oil and the intention of the study was to elucidate differential effects of meal fatty acid composition on inflammatory markers. Not to my, but probably to the scientists surprise the "evil" saturated fat from the coconut oil turned out to be similarly benign as the flaxseed cake. Only the EPA-laden cod liver oil cake produced a statistically significant increase in IL-8 mRNA levels 6h post ingestion. Similarly, incubation of peripheral blood mononuclear cells with EPA, yet not ALA lead to >3x increase in IL-8 and >2x increases in IL-6 mRNA expression.

While I am not quite sure what to make of these observations, these results stand in line with previous studies reporting differential effects of EPA vs. DHA rich fish-oils, where across the board, the DHA appeared to be the major driving force of the beneficial health effects people hope to be getting from their fish oil caps (e.g. brain health, Engström. 2009).
Fish oil caps can be a good addition to this regimen specifically for those who are just about to start out on a low omega-6 diet to offset the skewed n-6 to n-3 ratio (something that takes its time). They are yet by no means obligatory for someone who eats fish on a regular basis and invests the extra bucks into grass-fed beef and eggs from pastured chicken. If you thusly satisfy your (anyway low) dietary DHA requirements... and most importantly, taking fish oil will not compensate for eating shitloads of processed foods and lack of exercise, even if the aforementioned pro-inflammatory effects of EPA support my previously uttered hypothesis that fish oil has some resemblance to "exercise in a pill".

7 comments:

Anonymous said...

very interesting topic...thank you for posting

Tim said...

The result of the Myhrstad study were to be expected. But not because EPA is pro-inflammatory per se (there is plenty of evidence to the contrary) but becaused they BAKED(!!!) it into a cake. I have no idea what they were smoking when they came up with this idotic idea.

Edward McBride said...

great post! this is very interesting.. hope to see more post about this. red diesel

jeff said...

You are selling fish oil short. Fish oil is an amazing compound that can treat inflammatory disease and depression. I used ultra-high dose fish oil to treat my chronic eczema: 15 grams per day! It worked far better than any corticosteroid ever did and had none of the skin thinning effects.

Google: Michael Maes Inflammation

Maes is a Dr. with tons of research on inflammation and fish oil. Ultra high doses like 15-20 grams are required for therapy.

Anonymous said...

it's pointless to talk about a "fish oil" study that uses cod liver oil. It's garbage only useful for preventing rickets. It can't correct a serious Vitamin D deficiency like acute Vitamin D supplementation can since the amount of preformed Vitamin A is going to be toxic to get more than 2000IU of Vitamin D. Isn't this a confound itself for talking about EPA content?

I eat canned herring with 3-4g of EPA and 1.6g of DHA per can (1 meal), so the ratio does not seem to problematic coming from a traditional dietary source.

Belfast Biker said...

Interesting. I can't stand oily fish, why do you say cod liver oil is garbage?

Belfast Biker said...

Interesting. I can't stand oily fish, why do you say cod liver oil is garbage?

(Added another post (I ALWAYS forget to click Notify Me....)