12. NNT — The Most Important Number in Medical Arithmetic

12. NNT — The Most Important Number in Medical Arithmetic

Another review which Anti-Saturated Fat devotees fall back on, most without ever having seen the actual results or numbers, is the Cochrane Systematic Review (2020) which surveyed 15 individual studies involving almost 60,000 people. Its intent? Prove how diets with typical saturated-fat intake were less healthy than lower-fat diets.

Actual Result? Like PREDIMED, NO DIFFERENCE in Heart Attacks, Cardiac Deaths, or All-Cause Deaths, but as a consolation prize 17% fewer “cardiac events”.

We don’t want to experience a cardiac event, do we, so some doctors will only focus on the 17%. Sounds impressive, but when we dig further we uncover the NNT.

Let me explain NNT, Number Needed to Treat, which asks “how many persons need to do this before one person, by statistical expectation, benefits?”

Consider rabies. If you were bitten by a rabid dog, you would be almost 100% sure to die, unless you were given a course of Anti-Rabies Serum. But if you did get the serum, now you would be almost 100% sure to survive. The benefit of the serum will be evident immediately, because you didn’t die. The Number Needed to Treat for Anti-Rabies Serum is 1. Almost every single rabies bite victim who gets it will survive.

Now consider alcoholism, and think of Alcoholics Anonymous (AA) as a kind of cure for alcoholism. Though statistics are fuzzy, let’s go with an estimate that out of every person who signs up for AA and attends a few sessions, actually only 20% get cured. That’s a Number Needed to Treat of 5 — for every 5 persons who join, only 1 will eventually be cured of alcoholism. (No offense meant to AA, which is a noble endeavor, but that’s the current prevailing higher-end estimate of its success rate.)

Now we get to dietary or non-acute drug interventions, all of which take a long time to show results (positive or negative). This is not a person-by-person matter like the rabies serum. You can die even if you took the drug; you may live even if you didn’t. The best that scientific researchers can do is construct a study, then see what percent of each group died or had an adverse cardiac event.

Most clinical studies will show you table after table of numbers, and throw out terms like “Hazard Ratio”, a range of possibilities around that ratio figure at a “confidence level” which is always 95%, and a mysterious number called “p-value”. As a lay person, you don’t need to understand any of these numbers. The only number you need to understand is Number Needed to Treat, because it’s the number you can relate it to your real-world life.

To return to the Cochrane Systematic Review, what was its Number Needed to Treat? Based on those 60,000 people, statistically speaking, how many people need to follow the “desirable diet” (unsaturated fats in lieu of saturated fats, total calorie and fat intake held steady), for ONE PERSON to benefit?

The answer for the Cochrane Systematic Review’s very-carefully chosen 15 studies (out of thousands to choose from) was an NNT of 56. FIFTY-SIX people need to replace “bad” saturated fats with “good” unsaturated fats to create a statistical expectation that ONE of them will avoid a non-fatal stroke thereby. How many will avoid dying? Statistical expectation: NONE (their diet will not affect their risk of dying one way or the other).

If you get bitten by a rabid dog, you KNOW you’d better get the serum. As for shifting to unsaturated fats and giving up steak, pork belly, and ice cream, the choice is not so clear, and can be likened to joining a lottery where — according to Cochrane — your chance of “winning” (by not dying or not having a heart attack) is 1 in 56. (Other studies show NO difference, no chance to win healthwise by cutting saturated fats.)

Always find out the Number Needed to Treat.

There must be thousands of studies in this regard, and like PREDIMED they’re all not really “Clinical”, since no one watches over participants’ shoulders to make sure what they eat over a 5-year period. So, despite Cochrane’s high reputation for meta-analysis of many individual clinical studies, this particular Review lacks the strict scientific foundation of true “clinical” studies, in which it has to be known with reasonable certainty that the participants took an actual drug or followed an actual diet over time.

Therefore, the Cochrane Systematic Review of dietary-fat studies would be better described as “observational”, and I haven’t seen a really convincing explanation of how Cochrane justified picking those exact 15 studies to aggregate, instead of the thousands of others. Were all the others conducted by charlatans, or did Cochrane choose only studies that “kind of” proved their point? In a later article, we’ll examine a prominently disseminated study that was carried out by researchers who were, imho, either charlatans or nincompoops.

By the way, by far the best and most comprehensive “observational study” on this subject is the BILLIONS of people in the Westernized world, who for 70 years have been strenuously advised to limit all fats except MUFA and PUFA.

Compliance? Excellent, as can be seen in what the food industry manufactures and sells in supermarkets — saturated fats widely suppressed, hogs bred to be skinny, inches of beef fat trimmed by butchers and thrown away, low-fat versions preferred to original, but compensated by much higher starch and sugar contents. (We see families of doctors dutifully trimming every last bit of fat off pork ribs, then happily eating Cheerios, bread, and jam for breakfast.)

Result? A worldwide pandemic of obesity, chronic kidney disease, colon cancer, hypertension, and diabetes.

But the members of the Medical Establishment aren’t interested in the most obvious observational study possible, because it shows them to be tragically, fatally wrong.