The Sydney Diet Heart Study published earlier this year brewed fresh debate on Monday at a symposium in Boston organized by the American Society for Nutrition at Experimental Biology 2013 surrounding dietary recommendations to substitute saturated fats with polyunsaturated fatty acids (PUFAs) and, in particular, omega-6 linoleic acid (n-6 LA). An evaluation of recovered data from the randomized controlled trial that was conducted from 1955 to 1973 showed no evidence of cardiovascular benefit from n-6 LA.
PUFAs and their relationship with coronary heart disease and cardiovascular disease has been an active area of discussion among nutrition and medical researchers since the 1950s and ’60s. It was around that time that Ancel Keys and others began to emphasize substitution of saturated fatty acids (SFAs) with PUFAs because of evidence that it would lead to reductions in total and LDL cholesterol. But at that time, all PUFAs were treated similarly in the scientific literature and it wasn’t until far later that studies began to distinguish between n-6s and n-3 PUFAs. That has had some researchers calling for a deeper review of the literature.
In the Sidney Diet Heart Study, the intervention group was instructed to reduce saturated fats to less than 10 percent of energy intake and increase safflower oil, a concentrated source of n-6 LA, from 6 percent to more than 15 percent of energy intake. The re-evaluation showed that the n-6 LA group had increased rates of deaths from all causes, including from cardiovascular and coronary heart disease as compared to the control group.
Chris Ramsden, MD, of the U.S. Public Health Service, who was the re-evaluation’s clinical investigator, said that n-6 LA did lower cholesterol as predicted. However, the significant increased risk of cardiovascular and coronary heart disease calls into question whether guidelines to increase n-6 LA intake to between 5 and 10 percent, according to American Heart Association recommendations, are ideal or if they cause premature death.
Two “extreme views” about n-6 LA prevail, Dr. Ramsden warned: The first is that increases of n-6 LA lead to increases of arachidonic acid, increasing inflammation. There is no evidence to support that common belief because tight regulation of the rate-limiting step involving delta-6 desaturase. The second extreme view is that because n-6 lowers total and LDL cholesterol, “the more the merrier.” Dr. Ramsden is calling for more research suggesting that the truth may lie somewhere in the middle.
The proposed model for linking n-6 LA oxidation to cardiovascular disease that Dr. Ramsden supports is through a mechanism related to oxidized LA metabolites (OXLAMs). These include 9- and 13-hydroperoxy-octadecadienoic acid (9- and 13 HODEs). When found abundantly in oxidized low density lipoprotein, he said, these are potentially atherogenic. As shown in a figure published as part of the Sydney Heart Diet Study, OXLAMs are linked to formation of macrophage foam cells, endothelial cell activation, migration, proliferation, and formation of foam cell of vascular smooth muscle cells, and inhibition of hydrolysis of LDL cholesteryl esters that are all involved in cardiovascular pathogenesis. OXLAMs, caused by free radical-mediated oxidation, helps explain why smoking and alcohol are major risk factors of cardiovascular disease.
While randomized controlled trials showed that lowering n-6 LA didn’t reduce arachidonic acid, Ramsden said, evidence from randomized controlled trials supports that it does reduce n-6 LA and arachidonic oxidation after 12 weeks. In addition, he said “linoleic acid’s effects may be tissue specific extending beyond inflammation.” For these reasons, the relationship between n-6 LA and inflammation or cardiovascular risk deserved more attention from researchers.
A contrasting view
Professor of epidemiology and nutrition Walter Willett, MD, of Harvard School of Public Health, contended with Dr. Ramsden’s view, saying that it was a “hypothesis without fuel.” For one thing, he said, n-6 LA contributes to several kinds of both pro- and anti-inflammatory pathways in the body. “We can’t draw conclusions from pathways. It’s oversimplification,” Dr. Willett said.
Dr. Willett challenged the findings of the re-evaluation of the Sydney Heart Diet Study claiming that the study was small. He also said it had a major flaws and that the use of margarine in the study was a large confounding variable. The margarine was likely high in trans fat, he said, in fact, up to 18:2 ratio trans fat to n-6 LA. “No one in the U.S. eats that way,” Dr. Willett said.
Dr. Willett continued by offering several examples of studies in the “substantial literature” showing that n-6 LA was not associated with inflammation or increased risk of cardiovascular disease. One of the most recent studies he cited was a systematic review of 15 randomized controlled trials by Johnson and Fritsche (2012, J Acad Nutr Diet) that found “virtually no evidence” that increased n-6 LA in the diet increased the concentration of inflammatory markers (including C-reactive protein, fibrinogen, and tumor necrosis factor-α).
Countering the argument from Dr. Ramsden that people may not see a benefit from reductions of n-6 LA until its energy contribution in the diet reaches less than 2 percent, Willett responded that older studies (including Keys et al) did study populations with low intakes below 4 percent. He added that the notion of an ideal n-3 to n-6 ratio (like that of our predecessors) is utter nonsense and completely unsupported by evidence.
Any change to reduce n-6 LA could lead to greater risk of cardiovascular risk, Dr. Willett warned. As n-6 LA increased in the diet since the 1980s, he said deaths from coronary heart disease have “massively declined.” Then, with the intent of burying the notion that greater intake of n-6 LA causes any harm, he added that there is no evidence of higher intake related to cancer, weight change, suicide, or a host of other adverse effects.
After Willett’s talk, Dr. Ramsden responded by making several points: He said that some of the studies Willett cited failed to consider amounts of n-3s and that it was important to concede that food frequency questionnaires used in observational studies have serious limitations that are likely to confound results. “If you ask how often you eat certain foods, it’s then hard to disentangle linoleic acid and alpha-linolenic acid,” he said.
In defense of the Sydney Heart Diet Study’s results, Dr. Ramsden said that the safflower-based margarine wasn’t likely to be high in trans fat. “If you hydrogenate linoleic acid, you don’t produce trans-linoleic acid,” he said. You’re adding hydrogens, reducing double bonds that produce oleic acid, or mono-unsaturated fats and saturated fats of which were recorded, he said.
Martha Belury of Ohio State, who chaired the event, reminded that the “premise of the dogma” behind n-6’s relationship to inflammation rests on the hypothesis of arachidonic acid-derived eicosanoids (not on OXLAMs). Hammering down the point that there was no relationship between n-6 LA and arachidonic acid, she discussed one of her own studies in mice. She gave the mice 0.8, 4.9, and 8.4 percent n-6 LA. The results were that there was an expected increase of n-6 LA in tissues, but an actual reduction in arachidonic acid.
How much n-6 LA a person should consume should depend on “personal context” and “duration and dose need to be worked out”, she said. However, she added, until we know more, it was important to continue to keep n-6 LA in the diet at 5 to 10 of energy intake percent in line with American Heart Association recommendations.
Penn State professor of nutrition Penny Kris-Etherton spoke last about the health effects of new “designer oils” that are currently being introduced to the market. Many of these oils are higher in monounsaturated fats as opposed to PUFAs n-6 and n-3 to increase their stability. That has Kris-Etherton concerned because health benefits from increased n-6 LA in the diet could be lost with any dramatic changes over time.
She emphasized that it is arachidonic acid, not n-6 LA, that is the only n-6 that accounts for an inflammatory association. According to evidence from the available literature, including Johnson and Fritsche, higher TNF-a and IL-6 is associated with higher arachidonic acid and not with higher n-6 LA. To reduce intake of arachidonic acid (already low at around .13g/d in the Western-style diet), she said one must reduce animal foods such as chicken, eggs, beef, and sausage. In comparison, the other vegetable oils including monounsaturated fats and n-3 alpha-linolenic acid (ALA) have no effect on inflammation. Docosahexanoic acid (DHA), in particular, had a marked effect on reducing IL-6 as compared with ALA.
“When arachidonic acid changes, it’s not related to n-6 LA,” she said. “With designer oils moving to monounsaturated fats, we shouldn’t throw the baby out with the bathwater,” she said. “Omega-6s are good!”
Dr. Willett agreed that swapping out designer oils could lead to broad repercussions, such as increasing risk of cardiovascular disease. He added that loss of n-3 ALA in oils could also be detrimental. Loss of PUFAs in the diet may also mean loss of intake of tocopherols, added Belury. Currently, highest sources of vitamin E tocopherol intake comes from vegetable oils. She also noted that vitamin E tocopherols may account for another variable not looked at yet in the Sydney Heart Diet Study as it may have a role in prevention of OXLAMs.
Where does this leave n-6 LA’s relationship to inflammation and OXLAMs as a proposed mechanism of cardiovascular pathogenesis? In addition, what about substituting some n-6 LA with n-3s such as DHA or stearidonic acid as a strategy for reducing n-6 LA-derived OXLAMs without reduced intake of PUFAs? When I asked Dr. Willett about these concepts, he answered, “there are too few data available” to be able to offer a good answer to the question.
Without a doubt, the conversation, the controversy, and the debate will continue. In a final statement to me, Dr. Willett rehashed his recommendations for reducing inflammation and risk of cardiovascular disease of which include healthy weight management, a diet higher in vegetable PUFAs and lower in animal-derived saturated fats and trans fats, and, more specifically, a reduced dietary intake of red meat.
Interestingly, a few weeks after the Experimental Biology conference I received an email from retired nutritional biochemist Bill Lands about my report above on the linoleic acid debate. He wrote, “Your effort to transfer knowledge from the Boston session about linoleic acid is commendable. Some facts that were not presented clearly enough might broaden your understanding of what you heard. I want to attach two papers that may help you learn what you need to know. This abbreviated venue doesn’t do that.”
I had actually met Dr. Lands at the conference and tried to follow up with him for an interview in person and by phone after the event. However, amidst all of what was going on at the conference, we never were able to meet up.
“Please consider these bits of evidence,” Dr. Lands wrote to me. He had these two papers attached: Lands B, Lamoreaux E. Describing essential fatty acid balance as 3 – 6 differences rather than 3/6 ratios. Nutrition & Metabolism 2012, 9: 46-54. and Lands B. Consequences of Essential Fatty Acids. Nutrients 2012, 4: 1338-1357.
In summary, the two papers argue against the use of n-6/n-3 ratio in favor of a “food balance score” using an “arithmetic difference” between n-3 and n-6 in food items as a way to influence tissue concentrations of the two PUFAs. The reason is, according to the papers, that the shorter-chained n-6/n-3 PUFAs are not able to influence tissue. However, the longer-chain n-3/n-6 fatty acids do influence tissues and production of eicosanoids involved in inflammatory pathways.
In other words, to truly influence tissue balance and inflammation, Dr. Lands has proposed using a scoring system that accounts for particular foods that will directly affect whether tissues increase levels of n-3s or n-6s, regardless of ratio. It’s a system, he says, that could help guide personal food choices in a way that can help the public seriously improve their health.In fact, there is already a website with searchable pdfs (http://www.fastlearner. org/Omega3-6Balance.htm) and a mobile phone app created to facilitate the food choices found here: http://www.fastlearner.org/Omega3-6BalanceApp.htm
Dr. Lands declined to directly comment specifically about the event, simply ending with: “As you absorb the information in the papers that I sent, you will likely come to realize that my helping you move forward to better understandings is more productive than re-working imprecise comments from the small collection of dissonant over-simplified presentations in Boston. They are now history, and the public needs to prepare for their future.”