When I am at a restaurant, I often tell the waiter that I am eating lowcarb as a convenient way of explaining what I am going for. But that is not really true. What I am doing is following a low-insulin-index diet as a complement to the weight-loss power of time-restricted eating. I introduce these two key ideas in my posts
I discuss the theory behind these ideas in
Of these two ideas, time-restricted eating is by far the most important idea. But eating foods low on the insulin index when you do eat makes it much, much easier to go for periods of time without eating at all—whether those periods without eating are the last few hours before bed and then skipping breakfast the next day, or going without eating for more extended periods of time. So it is worth understanding what low-insulin-index eating is.
I found it fascinating to study the available information on which foods generate a large insulin response to write "Forget Calorie Counting; It's the Insulin Index, Stupid." There are many surprises. I distilled six general principles:
Avoid foods labeled "lowfat"
Avoid cold cereal
Avoid sweet beverages, including fruit juice
Avoid things with added sugar
Avoid starchy foods
If you drink alcohol, lean towards white wine instead of beer
But there are many other things to learn from the evidence on which foods have a big insulin kick. For example, in addition to avoiding pancakes, bread and potatoes, and yogurt marketed as healthy that has a lot of sugar added, certain fruits, such as honeydew melon, seem to have a powerful insulin kick, with a reported 95% confidence interval of 63 to 123. (I think the insulin index of other types of melon is simply unknown because no one has done the experiment yet.)
Not All Types of Meat Are Freebies
One of the most important things to know is that meat is moderately high in its insulin kick because certain proteins can cause a substantial insulin response. Consider these numbers from among those I retail in "Forget Calorie Counting; It's the Insulin Index, Stupid," selected here to make a point:
vanilla ice cream: 65
chocolate cake brownie with chocolate frosting: 60
white rice: 58
chocolate chip cookies: 57
McDonald's french fries: 57
doughnuts with cinnamon sugar: 54
battered fish fillets: 54
white fish: 43
cinnamon swirl pastry: 42
honey-raisin bran muffin: 37
snickers bar: 37
beef steak: 37
white corn tortillas: 36
chocolate milk: 34
Australian brown pasta (I assume with no sugar added): 29
Australian white pasta (I assume with no sugar added): 29
The point is that white fish and steak are in the 30-50 insulin index category I label "Portion Sizes Should Be Kept Small Except on Special Occasions." They are not foods to make a regular part of your diet if you want to eat low on the insulin index, any more than chocolate milk, snickers bars or cinnamon swirl pastry.
Because the insulin index has not been measured for everything, particularly in all the shrouded variations that occur in processed food, an important test is to think of how hungry you feel an hour or two or three after eating something. I do eat meat in restaurants, but I feel just a bit hungry afterwards. I deal with that by eating a mix of almonds and cashews afterwards. So white fish and steak (and presumably a range of other types of meat that we don't have a measured insulin index for) are not freebies; they are medium-high on the insulin index—not so high that you shouldn't ever eat them, but high enough that you should eat them sparingly.
Some Types of Meat Are Low on the Insulin Index. There are types of meat that are much lower on the insulin index. Here are some for which we have data, along with some comparison items:
In the "Go-To Staples for a Low-Insulin Approach" Category:
tuna canned in water: 26
7% fat cheddar cheese: 20
In the "Especially Good Foods" Category:
chicken fried in olive oil with skin: 19 (95% confidence interval from 11 to 27)
roast chicken without skin: 17 (95% confidence interval from 9 to 25)
tuna canned in oil: 16
Australian hot dog (I assume with no sugar added): 16
Australian bologna (I assume with no sugar added): 11
In the "Suitable for Eating and Drinking Even on an Extended "Modified Fast" Category:
full-fat bacon (with no added sugar): 9
Bad Carbs and Good Carbs
The bottom line is that a low-insulin-index diet is not exactly a "lowcarb diet" in the usual sense of a "lowcarb diet" in our culture. Even less is a low-insulin-index diet a "lowcarb diet" in the sense that researchers often use. As far as the insulin index is concerned, the problem is not really carbohydrates, but easily-digestible carbohydrates such as all the many types of sugar and the most obviously starchy-seeming kinds of vegetables. Technically, even non-starchy vegetables and nuts have a lot of "carbohydrates," but these are carbohydrates that are digested slowly, so they don't cause much of an insulin spike. To repeat, it isn't "carbohydrates" that are the problem, but easily-digestible carbohydrates.
Here, straight from the list of nonstarchy vegetables on Wikipedia, are a list of vegetables I highly recommend as part of your diet (given what I know), even though they are full of carbohydrates as a matter of chemistry:
Beans (green, Italian, yellow or wax)
Raw vs. Cooked; Intact vs. Pulverized. One of the intriguing facts pointing to the importance of whether a type of cabohydrate is easily-digestible or not is one I discussed in "The Keto Food Pyramid": cooked carrots have a higher glycemic index than raw carrots. The glycemic index isn't the same thing as the insulin index, but within the same food group it is highly enough correlated with the insulin index that I use the glycemic index to guess the insulin index when direct data on the insulin index is not available. What this means is that you have to think not only about the processing of food by big food companies, but the processing of food that you do at home! In addition to what food you eat, you need to think about what you do to it before you eat it. Cooking carrots makes them easier to digest, so they cause a bigger spike in blood sugar.
I don't know of anyone having done this experiment, but I'd love to see someone measure the insulin index of intact veggies as compared to veggies that have been run through a blender to make a veggie smoothie. I am betting that the veggie smoothie will have a higher insulin index than the very same ingredients if they are eaten intact.
This doesn't mean that you shouldn't cook your food. Making food delicious matters as well as weight loss. Fortunately, it is easy to make delicious food that is low on the insulin index. For gustatory delight, the fact that dietary fat is OK from the standpoint of the insulin index makes it easy to compensate for sugar being banned, for the most part. (On banning sugar, I personally make an exception by eating a few squares of chocolate with 88% or more cocoa, even if it has some sugar, as I talk about in "Intense Dark Chocolate: A Review." Also, I bend the no-sugar rule by using commercial full-fat ranch dressing despite its having a bit of sugar listed in the ingredients, and not always making it to Whole Foods to get sugar-free "Paleo" bacon when we want to have bacon.)
The DIETFITS Randomized Intervention Trial
Let me use this distinction between a low-insulin-index diet and a lowcarb diet to discuss the recently reported results of the DIETFITS Randomized Trial flagged at the top of this post. The video of Christopher Gardner, in particular, is good at explaining what they did.
First, both diets tried to get people to avoid sugar, refined grains and processed foods. These are all great for a low-insulin-index approach. This similarity between the two diets tested was much, much more important than the modest differences between the two diets that got into the headline. Although the two approaches were not directed in any big way toward time-restricted eating, and so couldn't test the most important idea very well, I see the results of this study as a vindication for the idea of a low-insulin-index diet. These two diets, which did the key things to lower the insulin index of food eaten, resulted in 11.6 pounds and 13.2 pounds of weight lost on average, once I convert from kilograms to pounds. Let me excerpt the key passages from the abstract of the study:
The Diet Intervention Examining The Factors Interacting with Treatment Success (DIETFITS) randomized clinical trial included 609 adults aged 18 to 50 years without diabetes with a body mass index between 28 and 40. ...
Results Among 609 participants randomized (mean age, 40 [SD, 7] years; 57% women; mean body mass index, 33 [SD, 3]; 244 [40%] had a low-fat genotype; 180 [30%] had a low-carbohydrate genotype; mean baseline INS-30, 93 μIU/mL), 481 (79%) completed the trial. In the HLF [Healthy Low Fat] vs HLC [Healthy Low Carb] diets, respectively, the mean 12-month macronutrient distributions were 48% vs 30% for carbohydrates, 29% vs 45% for fat, and 21% vs 23% for protein. Weight change at 12 months was −5.3 kg for the HLF diet vs −6.0 kg for the HLC diet (mean between-group difference, 0.7 kg [95% CI, −0.2 to 1.6 kg]). There was no significant diet-genotype pattern interaction (P = .20) or diet-insulin secretion (INS-30) interaction (P = .47) with 12-month weight loss. There were 18 adverse events or serious adverse events that were evenly distributed across the 2 diet groups.
The key points are
(a) This was a relatively large intervention study, with 609 participants
(b) The difference between the two diets is stated in terms of differences in proportions of calories from the three macronutrient categories of fat, protein and carbohydrates, which lumps both the unhealthy easily-digestible carbs and the healthy carbs in things such as nuts and nonstarchy vegetables together under the heading of carbohydrates .
(c) Viewed from the perspective of the insulin index, the difference between the two diets was relatively modest, which makes it less surprising that there was little difference in the results. If you are already avoiding sugar, refined grains, and processed foods, then cutting back further on what are technically carbohydrates in a scattershot way that doesn't pay attention to the insulin index is not that valuable. On the other side, trying to cut back on fat is not particularly valuable.
(d) The study did not have a formal control group that continued to eat as they would have normally. But if one is willing to think of people in the population at large as if they were a control group (technically "adults aged 18 to 50 years without diabetes with a body mass index between 28 and 40") then the weight-loss relative to that "control group" is impressive. So the elements the two diets had in common seemed helpful: avoid sugar, avoid refined grains and avoid processed foods. These are all things that can dramatically lower the insulin index of the food one eats, even if one is not thinking explicitly about the insulin index.
(e) The researchers looked for interactions between individual differences in insulin-secretion across people and which diet worked better for that individual and didn't find anything. Here I think there was a big missed opportunity. If you are looking for interactions between differences between two diets and the insulin-secretion patterns of different individuals, to find anything you need to make the two diets quite different in their food insulin indexes. Despite being interested in differences in insulin secretion patterns across people, these researchers did not do the logical thing of focusing on a high food-insulin-index diet and a low food-insulin-index diet. I hope they focus on this in their next big study!
Anahad O'Connor, writing in the New York Times, has a take on the results of the study similar to mine. Here is Anahad's description of key elements of the study:
... the low-fat group was told to avoid those things and eat foods like brown rice, barley, steel-cut oats, lentils, lean meats, low-fat dairy products, quinoa, fresh fruit and legumes. The low-carb group was trained to choose nutritious foods like olive oil, salmon, avocados, hard cheeses, vegetables, nut butters, nuts and seeds, and grass-fed and pasture-raised animal foods.
The participants were encouraged to meet the federal guidelines for physical activity but did not generally increase their exercise levels ...
The new study stands apart from many previous weight-loss trials because it did not set extremely restrictive carbohydrate, fat or caloric limits on people and emphasized that they focus on eating whole or “real” foods — as much as they needed to avoid feeling hungry. ...
“We really stressed to both groups again and again that we wanted them to eat high-quality foods,” Dr. Gardner said. “We told them all that we wanted them to minimize added sugar and refined grains and eat more vegetables and whole foods. We said, ‘Don’t go out and buy a low-fat brownie just because it says low fat. And those low-carb chips — don’t buy them, because they’re still chips and that’s gaming the system.’”
Dr. Gardner said many of the people in the study were surprised — and relieved — that they did not have to restrict or even think about calories. ...
... both groups ultimately ended up consuming fewer calories on average by the end of the study, even though they were not conscious of it. The point is that they did this by focusing on nutritious whole foods that satisfied their hunger.
A low-insulin-index diet is not the same thing as a "lowcarb diet" either as the phrase "lowcarb diet" is used in common speech or as "low-carbohydrate-diet" is used in nutrition research. All of these involve avoiding sugar, refined grains and processed foods, which is a big deal. But beyond that, they are different.
It is unfortunate that the results of the DIETFITS study were reported as showing that a lowcarb diet and a lowfat diet were equally good, when in the nontechnical sense of what most people mean when they say "lowcarb diet" both diets were "lowcarb" because they both involved avoiding sugar, refined grains and processed foods. (You can see another example of such reporting here in this otherwise excellent article.) Note here that the researchers who did the DIETFITS study did not question the idea that sugar is bad, despite the dispute about just how bad sugar is that I discuss in "The Case Against Sugar: Stephan Guyenet vs. Gary Taubes" and "The Case Against the Case Against Sugar: Seth Yoder vs. Gary Taubes." (See also "How Sugar Makes People Hangry" and "Sugar as a Slow Poison.") In the language of the team who did the DIETFITS study, both a "Healthy Low-Carbohydrate Diet" and a "Healthy Low-Fat Diet" are low in sugar, refined grains and processed foods.
At a deeper level, here is the important point: It isn't eating lowcarb that is important. The most important element of an effective diet is time-restricted eating. The second most important element of an effective diet is food and drink low on the insulin index.
Don’t miss my other posts on diet and health:
I. The Basics
II. Sugar as a Slow Poison
III. Anti-Cancer Eating
IV. Eating Tips
V. Calories In/Calories Out
VI. Other Health Issues
VIII. Debates about Particular Foods and about Exercise
Julia Belluz and Javier Zarracina: Why You'll Be Disappointed If You Are Exercising to Lose Weight, Explained with 60+ Studies (my retitling of the article this links to)
IX. Gary Taubes
X. Twitter Discussions
XI. On My Interest in Diet and Health
See the last section of "Five Books That Have Changed My Life" and the podcast "Miles Kimball Explains to Tracy Alloway and Joe Weisenthal Why Losing Weight Is Like Defeating Inflation." If you want to know how I got interested in diet and health and fighting obesity and a little more about my own experience with weight gain and weight loss, see “Diana Kimball: Listening Creates Possibilities” and my post "A Barycentric Autobiography. I defend the ability of economists like me to make a contribution to understanding diet and health in “On the Epistemology of Diet and Health: Miles Refuses to `Stay in His Lane’.”