Lazarus Took Fish Oil

Tales of the prowess and health effects of fish oil (or, more generically, long-chain omega-3 fatty acids) have flopped around more over the years than…well, you know what. We have been told that fish oil is all but a panacea; and we have heard it is entirely useless. But this flipping and flopping have suddenly probed whole new depths of absurdity. Within literal days of one another, two headlines appeared on Medscape, arguably the premier information portal for health care professionals, reaching diametrically opposing conclusions. 

On September 18, we got: “Omega 3s: Is This the Final Word?” and a commentary about the ASCEND trial telling us “the supplements had no effect on serious vascular events, cancer, or mortality.” Then, on September 25, we got: “REDUCE-IT: 25% Reduction in MACE with High-Dose EPA.” EPA stands for eicosapentaenoic acid, one of the two major omega-3s in fish and other marine sources (the other being docosahexaenoic acid, or DHA), while MACE stands for “major adverse cardiac events.” In other words, a week after being found useless in ASCEND, omega-3s were busy preventing heart attacks, strokes, and sudden cardiac death in REDUCE-IT.

Ordinarily for this column, I would analyze the studies in question and explain how the competing conclusions issued from them. But that’s been done already by others, and quite frankly- I think we have bigger fish to fry. 

The conflicting conclusions of these two trials, one presented at the European Society for Cardiology 2018 Congress, the other at the American Heart Association Scientific Sessions 2018, are, after all, only the most recent of many. The whole, long lineage of flip-flopping conclusions warrants analysis and explanation, and that in turn requires a view from altitude.

Here, I think, is what we really want to know: Is there some truth about omega-3s we can rely on for longer than the duration of a news cycle, and why does keeping pace with the “science” on this particular topic feel like watching PingPong?

Regarding that first question, and the basic health effects of omega 3s, the answer has long been and remains “yes,” for three reasons. First, omega-3 fatty acids are called “essential” because they are. These compounds are required for the integrity of our cell membranes, and in the manufacture of various hormones. They are essential nutrients for the same reason that vitamins, minerals, and “essential amino acids” are essential: they are construction material we need, and cannot produce. We consume them, or we lack a required component of ourselves.

In the case of the long-chain omega-3s found in fish, we can actually make them from the shorter-chain omega-3s found in plants, notably alpha-linolenic acid (ALA). However, we do that fairly inefficiently and quite variably; that pathway can be blocked with a high intake of omega-6 fatty acids, which is common; and besides, most people eating modern diets have low intake of plant as well as marine omega-3s. 

To understand why a balanced inventory of construction materials matters to the human body, imagine an actual construction site for a house. Now, imagine the construction crew has a large surfeit of plywood, but a serious deficiency of 2X4s for framing. Can you build a house that won’t blow over in a storm from such unbalanced materials? No, you cannot. Much the same pertains to a vital human body.

Second, we are adapted to a certain intake of omega-3s, and modern eating tends to provide us much less than that. That problem is likely compounded by the high level of competing omega-6 fatty acids most modern diets provide. Whenever the science regarding a given nutrient leaves us all in doubt about how much is optimal, I turn to what we can infer about our native levels. Adaptation does not tell us everything, but it is why zebras eat grass and lions eat zebras. It explains a lot.

Finally, no matter how important a nutrient, the effects it has on any given outcome over the dose range from adequate to optimal is apt to be very small. I can illustrate this. Imagine if, instead of just accepting the evidence that daily walking is good for health (which it certainly is), we decided we needed to know the active ingredient in a daily, say, 3-mile walk. Is it step 187? Step 352?

Now imagine we design trials to answer this particular question: what is the net contribution to health of step 352 on a 3-mile walk of roughly 5,300 steps? I am confident I don’t need to explain the futility of this endeavor; isolating the health contributions of step 352 would be practically impossible. But does that mean there is no health effect of step 352? Of course not! If any one step has “zero” value, then the entire walk has zero value, because (5,300 X 0) = 0. (Remember that old riddle about accepting a dollar bill with a tiny piece torn out of one corner?) And the same would be true for a ten-mile walk. So, in fact, each step MUST contribute something to the net health benefit of the walk. But isolating that contribution does not seem to tempt us. In the case of nutrition, for whatever reason, it not only tempts us, it actually prevails.

Because of what we may henceforth call “the step 352 problem,” I have much preferred looking at omega-3 effects across as large an expanse of evidence as possible, rather than focusing on any one trial. Doing that, as I have been obligated to while writing editions of a nutrition textbook, I have been impressed by the diversity of such effects. While truly decisive evidence of benefit related to any one health condition has been elusive, there are strong suggestions of benefit related to cardiovascular disease, diabetes, dementia, traumatic brain injury, autoimmune disease, and more. In the aggregate, these many suggestions make a case for net health benefits from omega-3s beyond a reasonable doubt.

Accordingly, I take a daily omega-3 supplement derived from algae. Algae can provide the EPA and DHA found in fish, but is far more sustainable. Since it is possible now to eat our fish oil and have our fish (in the seas) too, why do it any other way?

We can now explain the long litany of conflicting research findings. For starters, omega-3s are not unique in this regard. The problem pertains in general to nutrient supplements. There is much less money, less profit, and less patent protection in nutrient supplements than in pharmaceuticals. Accordingly, the studies tend to be smaller and shorter, and often- inconclusive as a result.

In addition, while a new drug can be tested directly against an established drug when a placebo is unethical in a population with a malady, nutrient supplements are typically added to “state of the art” pharmacotherapy. At that point, they can only show an effect by improving on the best we can achieve with medication. The less is left “broken,” the harder it is to show that you are fixing anything.

The only way to avoid consigning nutrients to the residual fixes left by optimal drug therapy is to test them in people with no need to take medication in the first place, namely: healthy people. Here, we encounter another problem.

Imagine that (I am making up these numbers, but they serve to illustrate a basic point) omega-3 supplements prevent one heart attack per year in a population of 10,000 middle-aged adults without known heart disease and not on medication. The trial it would require to show this effect is absolutely massive. If we enrolled 40,000 people, half in the treatment group, half in the control group getting placebo, for a year- we would have two fewer heart attacks in the omega-3 group, and that would be statistically invisible.  If we enrolled 200,000 people, our study would rival the largest randomized trial in history. But we would only have 20 fewer heart attacks in our treatment group, and this, too, might be statistically invisible. So, the largest randomized trial in the history of science might well not be big enough to discern such an effect.

Doesn’t that mean the effect in question must be trivial? Actually, no. There are nearly 750,000 myocardial infarctions in the U.S. every year. If omega-3s did prevent one in just 10,000, that would still mean a lot to the 75 families each year not needing to rush to the bedside of a loved one in the CCU. 

The recent trials- ASCEND, and REDUCE-IT- used different doses of omega-3, enrolled different populations, and differed in other ways as well. We can account for the opposing conclusions accordingly.

More importantly, however, we must account for the context in which nutrient trials are conducted, and why that context is conducive to just such conflicts. Until or unless we do so, and remediate our mishandling of science, we may expect to mourn the demise of omega-3s one week, celebrate their triumph the week following. In other words, the one reliable scientific conclusion to date appears to be: Lazarus must have been taking fish oil.