In recent years, among those aiming to eat a limited carbohydrate diet, the concept of ‘net carbs’ has taken root, viewed by some as a boon and others as a detriment to healthy eating. Although I know quite a lot about the concept and its origins, when I set out to write this piece about net carbs, I decided to begin by asking the wizard bots at Google a quick question, framed thus: When did the concept of ‘net carbs’ begin? And Google answered:

I found the answer surprising because it’s quasi true, but not exactly so. In the realm of fact checking, I’d have to rate it mostly false. Why?

Because I happen to have firsthand knowledge of how it began. Dr. Michael Eades, my husband and medical practice partner, first developed the concept in the late 1980s, which he presented in his first book, Thin So Fast (Warner Books 1989). And we amplified the concept in Protein Power, published in 1996, which I co-authored with him.

Here’s the BTS of how it came about.

Back in 1987-88, when Mike was writing Thin So Fast—a low carb structure of three protein shakes a day and a whole food protein, fat, and low carb meal once a day—my job was to come up with a tasty protein shake recipe (since believe it or not protein shakes didn’t really exist then) and a batch of healthful, whole foods low-carb recipes for the book.

In thinking about how to give people an appealing mixed diet, maximize the amount of various foods a person could consume in a day, and keep their carbohydrate grams within a range that would keep their blood sugar and insulin levels low and let them stay at least in light nutritional ketosis, we realized that when people were counting their carbohydrate grams, fiber grams were included in that total.

But fiber—both soluble and insoluble—is not digestible by the human GI tract.

Fiber, of course, is nothing but long chains of glucose molecules hooked together by chemical bonds, like starch in a way, except that the human GI tract doesn’t have the ability to break the bonds in fiber. So, unlike starches, the sugar (the carbs) in fiber can’t be liberated as glucose in the small intestine to be absorbed as such. So no glucose rise and therefore no insulin rise as a consequence of eating fiber.

(To be completely accurate, fiber does get absorbed, but only downstream after it’s acted upon by the bacteria—the microbiome—in the colon, which digest it for their own benefit and turn it into acetic, propionic, and butyric acid, short chain fatty acids that are good for the colonocytes and that we can absorb but that still have no impact on blood sugar or insulin levels.)

So since what Mike was after physiologically in his patients was keeping their blood sugar and insulin low to promote weight loss and resolve other facets of the metabolic syndrome, he figured he didn’t need to worry about the fiber content of food. We speculated based on that knowledge that we could tell people they could subtract the fiber grams from the total carbohydrate content of a given food to arrive at only those carb grams that actually had an impact or an effect on blood sugar and insulin.

He called this concept the ‘effective carbohydrate content’ of food. The ECC. And I developed a table for the Appendix of the book of giving the ECCs for 5 to 15 gram carb portions of a long list of foods.

The purpose of the concept at its inception was to gently steer people’s carb intake away from concentrated sources of carbs and toward whole foods with the most nutrient and fiber content for the least number of carbs. So an easy way of pointing them toward foods such as asparagus, broccoli, cauliflower, and raspberries, toward green and colorful fruits and vegetables, where they could get more bang for their carb buck rather than concentrated starches such as wheat, potatoes, rice, or corn.

For instance, there are about 5 grams of total carbohydrate in a cup of cooked cauliflower, and 2.1 of those are fiber. Whereas there are 45 grams of carbohydrate in a cup of cooked rice and only 0.6 of those are fiber. So a person could enjoy a whole cup of cauliflower for 3 effective carb grams of their meal’s carb total and get 2 grams of fiber in the bargain, or they could opt for a tablespoon of cooked white rice for that same 3 grams and get zippo fiber. The choice is pretty easy.

The intent of the ECC when Mike published it in 1989 was to improve reader’s nutrition, to encourage them to eat more fresh green and colorful fruits and vegetables alongside their rich intake of quality protein and good fats. To let them eat more of what was good for them, making their allotted 7 to 10 grams of effective carb per meal more like 15 to 20 grams of total carb, if they chose well.

But once the low-carb craze began to really boom again, which it did following publication of our New York Times bestseller Protein Power in 1996 (which coincidentally resurrected the sales of Dr. Atkins’ most recent book languishing at the time) food manufacturers began to take notice. Just as they had done with the low-fat diet mania of the ‘80s and ‘90s that gave us fat-free cookies, candy, ice cream, and salad dressing and low-fat and non-fat milk, yogurt and cheese, they hopped on the low-carb band wagon and began cranking out every sort of low-carb junk food, by replacing sugar with emulsifiers, artificial sweeteners, and sugar alcohols.

Sugar alcohols are compounds similar to a sugar but fermented (usually) or chemically altered to make the molecule an alcohol. So the simple sugar glucose becomes glucitol (better known as sorbitol), mannose becomes mannitol, maltose becomes maltitol, xylose becomes xylitol, etc. Some are more readily absorbed than others, but most of them are at least partly non-absorbable, and it is this characteristic that accounts for their claiming ‘not-an-effective or net-carbohydrate’ cred. But that same non-absorbability is also responsible for their unpleasant GI side effects. When it was a gram of xylitol here or there in a piece or two of sugar free chewing gum, sugar alcohols didn’t cause any issues, but when it became a matter of 8 or 10 or more grams of the stuff in a serving of sugar free ice cream, problems ensued.

We spoke in 2004 at the Low Carb Biz Summit in Denver, which was a huge program mainly for small business food purveyors, food manufacturers, journalists, and people interested in health. My keynote speech warned that it would be treading in deeply dangerous waters to rush into replacing low-fat junk food with low-carb junk food and think that a victory. And that doing so could derail the low carb movement by derailing the health goals of its practitioners. My red flags went unheeded and food manufacturers crammed sugar alcohols into every sort of food, labeling them ‘low net carb’ and ‘zero net carb’, which wasn’t really the case.

One of the sponsors of that conference was Blimpie, the fast food franchise, and they’d just created and launched a new ‘Zero Net Carb Brownie’ to their menu. The CEO of a large Blimpie’s chain was there and asked me to sample their new low-carb dessert. When I saw it had nearly 30 grams of sugar alcohols in a single serving I told him to be sure that all their store bathrooms were in good working order, because with that much sugar alcohol they would need them. Plus, the first ingredient was almond flour, with 10 effective grams of carb per cup. So even if a single large brownie had but ¼ cup, there were still 2.5 effective (net) carbs that should have been on the label. Plus there’s some carb in the cocoa powder, and at least some portion of the sugar alcohols ought to be counted as effective carb grams as well. So the moral to that story is “if it seems too good to be true, it probably is.”

And one last interesting bit that brings it all full circle. I recall the afternoon at a medical conference in probably 1999 or 2000, when we ran into Dr. Atkins himself on the conference floor. He asked Mike if what we’d published in Protein Power about fiber not counting as a carbohydrate was correct, and what it was based on. (It’s probably not a coincidence that he was just coming out with his Atkins Foods product line.) Mike laid out the concept for him and said he’d checked with the people at USDA about how products were officially labeled to be sure that fiber was counted in the total carb count even if it was listed separately, which he affirmed was the case. And that, yes, it was true that fiber was not absorbed as a carb though it was counted as one. And voilà! Before you could blink Dr. Atkins introduced his concept of the ‘Net Atkins Carb Count’ which became’ Atkins Net Carbs’ which became just ‘net carbs.’

What began as an effort to steer people toward better nutrition had suddenly been co-opted and put to evil nutritional purposes.

As was the case when we first developed the concept, we still firmly believe that it’s a reasonable stance to subtract the naturally occurring fiber in a whole food from the carbohydrate total to get to the actual grams of sugars and starches that can have an adverse impact on health. But when you start to game the system with sugar alcohols and other processed, chemically-altered Frankenfoods, all bets are off. As they say, ‘It’s not nice to fool Mother Nature.’