There is no single dietary template that works for everyone.
Different individuals have particular needs and goals, and nutrition should be adjusted accordingly.
It can be useful to compare how your diet stacks up against other dietary approaches to identify where you might further refine and improve your diet.
The insulin load of your diet should not exceed the capacity of your pancreas to keep up and achieve normal blood glucose levels.
Once excellent blood glucose levels are achieved, reviewing your nutrient density, calorie density and fibre intake can be a useful refinement to optimise your diet to achieve your goals.
keeping it simple…
Eating should not be complex!
I could understand how some people might find all this discussion about nutrition a bit daunting and/or irrelevant.
If all this talk about food is confusing and leaves you a bit perplexed then I apologise.
The reality is that if most people ate a range of whole foods from natural sources they would be pretty much OK. Most people will not turn into ripped physique competitors with bulging six packs without meticulously tracking food intake, however, most would be able to achieve good health and vitality.
A good diet is something that people will stick to and enjoy without tracking and obsessing over too much. At the point it becomes enjoyable, effortless and normal it stops being a diet and becomes just a way of eating!
To this end, I’ve created a list of optimal foods and optimal meals to suit different goals that people can just run with without too much hassle or conscious thought.
finding your path in all the noise
Unfortunately, many of us have spent too long either eating poorly and need some more intense intervention, or have more specific dietary needs (e.g. diabetes, cancer, epilepsy etc) that require more targeted approaches.
My key aim in all of this is to demonstrate how a nutrient dense diet with a managed glucose load can be optimal and modified to suit a person’s goals. Rather than giving general platitudes, I hope I can add something to the discussion and help people tweak their diet to achieve specific goals.
There is a lot of sometimes-conflicting dietary guidance out there that makes it hard to synthesise it all into a coherent plan that’s right for you. You may be ketogenic, vegetarian, LCHF, LCHP, Paleo, ‘peagan’, or zero carb, or just do ‘everything in moderation’ (whatever that means).
taking it to the next level
No one eats optimally all the time, however, it’s useful to know what can be done to improve things if you want to move further in a particular direction.
If you’re not getting the results from your current approach, you might be motivated enough to refine your current regimen to move further forward in the desired direction. Unfortunately, a lot of people fall into this category and are left looking for a bit more specific guidance on what they can do to reach their goals.
Some people are already putting a lot of thought and effort into what they eat, but sometimes not getting everything they hoped for, whether it be athletic performance, blood sugar control or weight loss.
Quantification of insulin load, nutrient density and fibre are powerful tools to further manipulate your diet in the pursuit of specific goals. In this article, we look at how we can review and refine our current diet to suit specific goals.
I hope that reviewing the application of a number of tools to real life examples can demonstrate how we can manage the glucose load of our diet while at the same time optimising the vitamins, minerals and amino acids.
I think it’s useful to review, refine and tweak someone’s existing diet rather than trying to get them to adopt a whole new way of eating that might be hard to stick to in the long term. So to this end, I’ve reviewed a number of people’s actual food diaries to see what the system would tell us about how they can move forward in a particular direction.
The plan is to profile a dietary analysis every couple of weeks to see how we might apply the tools detailed throughout this blog to optimise their nutrition while keeping in mind their insulin load.
so let’s meet Wendy…
Wendy describes herself as being obese for her entire adult life, except for a few short-lived diet-induced periods (using “eat less, move more” diet templates). Despite obesity, Wendy had good blood lipids and fasting glucose until early 2014, when fasting glucose crept into the pre-diabetic range. She had been hypertensive since her mid-20’s, diagnosed as essential or idiopathic hypertension (i.e. no known cause), but was advised it was caused by her obesity.
In early 2014, at age 44, Wendy began exercising regularly and then changed her diet first to lower-carb (inspired by Robert Lustig) and then ultimately low-carb. She lost over 70 pounds over eighteen months, going from a BMI of 39 to 27. Wendy says she is comfortable with low-carb eating as a life-long proposition, along with regular exercise (mostly yoga and strength training, with some HIIT).
Wendy had some insulin resistance before making these changes, but now has blood sugars that are “typically in the 70s, 80s or 90s” and a fasting insulin of 5.1mIU/ml (i.e. upper end of the excellent range).
Wendy’s most recent HbA1C was 5.2% which is pretty good (see Diabetes 102 for more details of target blood sugar and HbA1c levels), so it appears that her insulin resistance has improved with her recent weight loss or at least is in remission with her improved diet and exercise regimen. Wendy has also been able to drop all hypertension medications and now regularly home-tests blood pressure at slightly-below-normal levels.
Wendy has been very active on the Optimising Nutrition, Managing Insulin blog with a heap of insightful questions and comments. Lately, she’s also been helping me with reviewing and editing of some of the posts on the blog. She’s certainly an educated, motivated and seasoned nutrition nerd. You can also read more about Wendy’s journey on her own blog https://fitteratfortyish.wordpress.com/.
If you look at her food diaries you’ll see that the reason she’s lost so much weight already is that she’s eating pretty darn well! But let’s look at what she could do to further move towards her weight loss goals.
review of glucose load
First up I’ve analysed Wendy’s macro nutrients to determine the insulin load, percentage insulinogenic calories and macronutrient split as shown in the table below.
With less than 20g net carbohydrates per day, an average of 5% calories from carbohydrates and 23% insulinogenic calories, Wendy’s diet definitely qualifies as low carb!
I’ve plotted the macro nutrients from her daily diaries in the chart below (labelled 8, 9 and 10 June). Two of the three days sit just outside the threshold of Steve Phinney’s well formulated ketogenic diet triangle (i.e. the orange line in the figure below). This analysis would indicate that Wendy is eating a good diet that would be great for normalisation of blood glucose.
If Wendy was still having blood sugar issues (say an average of greater than 97mg/dL or 5.3mmol/L) she might benefit from reducing her protein to bring her macros back towards the bottom left corner of this chart in order to reduce the total glucose load of the diet and make it more ketogenic.
This is not the case for Wendy though as her blood sugars are now reasonably well controlled, with her insulin resistance improved after her already significant weight loss. So reducing the insulin load of her diet is not the primary issue.
nutritional analysis – diabetes weighting
I’ve written before about the theory of balancing glucose load while maximising nutrition. I thought it would be interesting to look at how we could apply this in practice with Wendy’s food diary data.
A score of 100 in either the “nutrient balance” or “protein quality” scores means that you’re covering the recommended daily intake of each of the nutrients with 1000 calories. What this means in Wendy’s case is that she’s doing pretty well with her protein score at 140. However, there are some possible deficiencies with her vitamins and minerals which are only at 52.
The table below shows a comparison of scores for the 9 June 2015 food diary with the suggested refinements discussed below. Based on the diabetes weighting (which prioritises low insulin load) Wendy’s diet from 9 June ranks at #77 of 175 which is fairly solid, however, there is some room for improvement.
A score of zero would mean that it is average in comparison to the other meals, greater than zero (blue) means that it is better, while less than zero (red) means that it is worse than average.
The areas that the food from the 9 June 2015 food diary doesn’t do so well is calorie density, vitamins and minerals, and fibre.
so why is fibre and calorie density important?
Every individual is different. What will work for one person may not work so well for another. For example, simply applying a body builder or a diabetes dietary approach to Wendy’s situation may not be appropriate given that she wants to continue on her weight loss journey.
The comparison of Wendy’s base diet to the other diets analysed indicates that she could benefit from adding fibre and reducing her calorie density. But why would increasing fibre and reducing calorie density be useful for someone trying to lose weight, particularly when the low carbohydrate approaches advocated by Westman and Eades may likely recommend continuing to focus on low total carbohydrates (i.e. not net carbohydrates) until goal weight is achieved.
I think the “magic” of the low carbohydrate dietary approach is that it normalises blood glucose levels and reduces insulin so that you can release rather than store fat. For more detail on the importance of insulin in fat loss check out the article why we get fat and what to do about it v2.
Without a reduction in insulin levels, it is difficult to unlock the fat stores for energy. I get frustrated when I hear ‘experts’ in the body building scene just saying that fat people need to achieve a calorie deficit and they’ll be all good, apparently not understanding the effect that a highly insulinogenic diet can have on fat loss and appetite for people who are insulin resistant.
Chris Gardner’s A to Z trial  identified that people who are insulin resistant typically only lose weight on a low carbohydrate diet. The reduced insulin load allows energy to be released from fat stores which in turn leads to increased satiety and decreased calories. However, people who don’t have insulin resistance issues can lose weight with reduced calories regardless of the carbohydrate level as long as they create a calorie deficit.
However, once you’ve normalised your blood glucose levels and insulin I think all bets are off. There may be a limit to how far a diet with liberal quantities of added fat will take you when it comes to weight loss. Somehow you do need to work out a sustainable way to burn more calories than you consume. However, I don’t think this is as simple as just counting calories and maintaining a deficit.
While a low carbohydrate / ketogenic diet in maintenance mode involves high levels of dietary fat, you don’t necessarily need to be adding extra dietary fat while you’re trying to lose weight. If you’re trying to lose weight ideally the energy from fat can come from your body fat stores as shown diagrammatically below in Steve Phinney’s four phases of a ketogenic diet diagram.
I’ve got a lot of time for Jonathan Bailor who isn’t a big fan of calorie counting. He prefers rather to manipulate diet so that you naturally feel satiated with fewer calories. As noted in this video he advocates for nutrient dense foods that have plenty of water, fibre and protein to naturally feel full so you don’t need to manually track calories. This video gives a good overview of Bailor’s philosophy which I think makes sense for most people.
While this meal might have 24g of total carbohydrate there is 16g of fibre, meaning that there is only 8g of net carbs. So if we abandon the concept of total carbohydrate and focus on net carbohydrates when it comes to real unprocessed whole foods, we can keep our insulin load low and achieve satiety naturally by eating a larger volume of nutrient dense high fibre foods. This will ideally allow us stop worrying about counting calories while still keeping our insulin load fairly low.
One of the criticisms of Terry Wahls’ diet approach is that there is just so much food and it’s hard to eat it all to get your calories. If you do not want to lose weight Wahls recommends adding MCT or coconut oil to increase the calorie density. However, you can see how not being able to fit in enough calories (as opposed to just counting calories) would help you to sustain a lower calorie approach without as much conscious effort.
Similarly, Dave Asprey criticises Joel Fuhrman’s ANDI score as a recipe for starvation, which is sort of what we’re after if we are trying to lose weight.  So sure, maybe you’re not going to feel full eating parsley and watercress, but that doesn’t mean you shouldn’t try to eat as many non-starchy veggies as you possibly can as your first priority, and then fill up on the other foods.
While counting calories can be useful and effective as an educational tool, your appetite will probably win out in the long run unless you find a way of eating that will naturally keep you satiated. I believe a high fibre, nutrient dense, lower calorie density approach can be helpful to achieve this goal.
So in Wendy’s case, in order to improve the vitamin and mineral score, I’ve reduced the chicken from lunch and added some spinach and mushroom. This has raised the “Nutrient Balance” score from 52 to 75. The chicken is quite a calorie dense in comparison to the spinach and mushroom, and you can eat a lot of these veggies without increasing calories very much.
This updated food diary then ranks at #50 of 175 meals analysed (previously #77) based on the diabetes weighting as shown above. The overall fibre increases from a fairly low 13g to a more reasonable 21g. This is closer to the recommended daily minimum fibre intake of 25g for women and 30g for men ).
This change does increase the net carbohydrates from 11g to 32g and the insulin load from 60g to 77g, so someone who did have serious insulin resistance or diabetes should monitor their blood sugars to make sure they were not adversely affected by shifting to this approach.
For most people, this increase in net carbohydrates would not be a major concern, particularly as the increase in carbohydrates is from low glycemic whole foods which tend to raise blood sugars much less than manufactured products. You might also find that you end up naturally eating fewer calories because of the high volume of food and the high nutrient density which might leave you satisfied with fewer calories.
reducing calorie density for weight loss
For Wendy, though, blood glucose/insulin resistance is not the primary issue. Her current priority is to move forward with her weight loss which has now stalled, more than a year into continuous weight loss.
Eliminating processed carbohydrates is critical to the success of the low carbohydrate approach but what do you replace them with? As per Terry Wahls’ approach, I would recommend trying to maximise nutrient dense non-starchy veggies in the first instance and then supplementing with added fats (if required) to make sure you’re satiated. If we focus on eating as much high fibre, nutrient dense, low-calorie density foods as we can we no longer have to worry about limiting how much we eat!
So if we want to tweak Wendy’s meal plan more towards the weight loss goal by decreasing calorie density we can:
use a whole avocado rather than half an avocado to increase the fibre,
drop the added olive oil (to the extent that is practical for cooking),
drop the “half and half” cream to one tablespoon rather than four in the coffee, and
drop the calorie dense macadamias.
If we sort the meal revisions based on the weight loss weighting (which emphasises high fibre and low-calorie density) we can see that the revised diabetes diet has a ranking of 43.
With these changes, we’ve nearly doubled the weight of the food for the day while keeping the total calories the same. Fibre has gone up from 13 to 26g which meets the minimum recommended minimum fibre intake. This approach will be a lot more filling, which is useful if weight loss is the goal.
Net carbs have gone from 10g in the original scenario to 50g per day. This is still considered low carbohydrate; however, Wendy should keep an eye on her blood sugars as her HbA1c is good but not yet in the excellent range. If they go outside the normal range (see criteria here), she should revert to the nutrient dense diabetes approach (see criteria and foods here).
Different people will have different carbohydrate tolerances, and these can change as your body heals and releases fat from your belly, liver and pancreas. Most reasonably healthy people would be able to deal with this level of carbohydrate, particularly given that it is from low calorie density, low GI carbohydrates from vegetables.
As shown in the updated nutritional analysis below, the protein quality score is still pretty high at 135 (down from 140) and the quantity of protein is still quite high at 26% of calories (down from 27%). The only way to increase the protein quality score without increasing calories further would be to incorporate organ meats, which is not everyone’s cup of tea. The nutritional completeness score has increased to 88 which is a significant change from the base diet that had 52!
Reducing excess insulin (as indicated by poor blood sugar control and high body fat levels) is the first priority, however once blood glucose and insulin are stabilised, targeting high fibre nutrient dense foods (while still keeping the insulin load as low as possible) is likely to be the next step when it comes to weight loss.
If there were any nutritional issues that were causing the body to hold onto weight, these may be improved with the highly nutrient dense diet and possibly help to break through the weight stall.
The body requires somewhere between 160 and 600 calories per day from glucose.
This glucose can be sourced both from ingested carbohydrates as well as the glucogenic portion of protein not used for growth and repair.
Rather than raising blood glucose immediately, amino acids from protein circulate in the blood until they are required.
Excessive glucose from either carbs or protein will lead to increased insulin requirement, insulin resistance, diabetes, obesity and a range of other issues associated with hyperinsulinemia and metabolic syndrome.
Someone who is insulin resistant and/or has diminished pancreatic function does not produce adequate insulin to maintain normal blood glucose. Rather than using diabetes medications or exogenous insulin, the alternative option is to decrease one’s dietary insulin load to a point that the body’s natural insulin production can keep up.
We can manage our dietary glucose to achieve normal blood sugars by considering the total insulin load from carbohydrate plus the glucogenic portion of protein.
Rather than simply focusing on the ideal macronutrient split, this article endeavours to take the discussion one step further to look at how we can optimise the split between dietary glucose and fat given that glucose can be obtained from both carbohydrates, and the glucogenic portion of protein in excess of the body’s requirement for growth and maintenance.
the Goldilocks glucose zone
This article outlines a basis upon which to determine the optimum balance between what are often polar extremes.
On the high glucose end of the argument we are faced with the following issues:
high insulin levels,
obesity and excess fat accumulation,
high blood glucose levels,
heart diseases risk, and
the plethora of issues that accompany metabolic syndrome and hyperinsulinemia.
At the ketogenic extreme, we have concerns about a range of issues including:
Somewhere in the middle, there must be an optimal balance of fuel for each individual, a balance between the extremes.
But how do we find this balance point? Then what do we monitor to ensure we stay there?
Not too hot. Not too cold.
Not too hard. Not too soft.
What we are searching for is the “Goldilocks glucose zone”.
the safe starches debate
The ‘safe starches debate’ has been intriguing and has informed my thinking on this controversial issue.
The discussion started at the 2012 Ancestral Health Symposium with a panel hosted by Jimmy Moore.  It continued on the blogs of the two lead representatives of each side of the argument, Paul Jaminet  and Ron Rosedale .
the case for limiting carbohydrates
On the low carb end of the debate, we have Ron Rosedale who argues that:
1. Non-fibre carbohydrates are:
detrimental as they lead to increased insulin levels, oxidation and accelerated ageing, and
unnecessary as we can obtain our glucose needs via gluconeogenesis from protein.
2. Glucose can be manufactured from glycerol or from lactate and pyruvate recycling. In some respects, this is even better than making glucose from protein. 
natural glucose utilisation level
On the not so low carb end of the argument, Paul Jaminet argues that the human body runs on a fuel mix of about 30 to 35% of calories from carbohydrates (say 600 calories per day). The remaining 70% or so of our fuel comes from fat.
Jaminet recommends that people follow a ‘low carb’ diet, however, Jaminet’s version of low carb is a carbohydrate intake somewhere less than the body’s 30% requirement for glucose. This forces some proportion of the glucose needs to come from gluconeogenesis.
When you look at this in the context of the fact that the typical western diet has 40 to 50% of calories coming from carbohydrates, we are really arguing over whether a low carb diet or a very low carb diet is best for our metabolic health.
Jaminet’s glucose flux has a lot of similarities with Mark Sisson’s Primal Blueprint Carbohydrate Curve.  Jaminet’s 600 calories equates to 150g of carbohydrates which aligns with the top end of Sisson’s ‘effortless weight maintenance zone’.
But what if limiting carbohydrates to less than 150g per day is not working for you (e.g. your blood sugars are not in normal range or you are not achieving weight loss)?
What can we learn from the food insulin index data to help us build on standard carbohydrate counting?
How can we determine the optimum fuel mix for our individual situation, body and goals?
minimum carbohydrate requirement
One of the concerns about a low carbohydrate diet centres on the understanding that the brain needs carbohydrates.
This seems to stem from Institute of Medicine’s advice that the brain needs about 400 calories per day from glucose. This equates to 100g of carbs which most people wind up to 130g to provide a safety factor.
The IOM, however, notes that a person who is fat adapted can run on lower amounts of carbohydrates as their brain is fuelled by ketones and there is no minimum requirement for carbohydrates, only glucose which can also be obtained from gluconeogenesis.  In spite of this, nutritionists still recommend a minimum carbohydrate intake.
Jaminet makes a similar differentiation that a typical sedentary person requires about 600 calories for glucose per day, however, this may decrease to 300 calories per day for someone on a ketogenic diet.
The understanding of the absolute minimum glucose requirement comes from research by George Cahill who undertook extreme starvation experiments and found that people could survive on as little as 40g of glucose per day (i.e. 160 calories). 
In the fed state the body will rely on glucose from ingested carbohydrates. After a period of fasting, it transitions to using glucose from the glycogen stores in the liver and muscles. Once the glycogen stores are exhausted the body will obtain glucose via gluconeogenesis from cannibalising muscle.
At this point however the brain and the rest of the body have largely transitioned to being fuelled by fat so it only needs to obtain 40g of glucose per day from protein via gluconeogenesis. This would equate to around 5% of calories from glucose (not necessarily from carbohydrates).
I am not suggesting that starvation ketosis is optimal for most people. The point is that the body can survive on very little glucose if it needs to for quite a long time.
The longevity crowd will tell you that this is an evolutionary advantage so you can prolong life until a time when there is enough nutrition to reproduce and thrive. People who could use their fat and muscle for fuel survived to be your ancestors, and those that couldn’t didn’t.
what is the minimum protein requirement?
According to Nuttall and Gannon  the body requires between 32 and 46g of high-quality dietary protein to maintain protein balance.
This equates to around 6 to 7% of calories in a 2000 to 2500 calorie diet being taken “off the top” for growth and maintenance, with everything else potentially available as excess.
The same paper notes that the American diet typically consists of between 65 and 100g of protein per day (i.e. 13 to 16% of calories).
three macros or two fuel sources?
Something that has been very interesting to me that I had not understood until recently was that protein is made up of glucogenic and ketogenic amino acids. Some amino acids can turn into either glucose or fat. 
The table below shows the differentiation of amino acids into different categories.
I will be discussing this concept in more detail in a separate article (The Insulin Index v2), however in essence, what this means is that there are really only two fuel sources for the body, glucose and fat, with “excess” protein being turned into one or the other.
the “well formulated ketogenic diet”
Steve Phinney is probably the most well respected authority on the ketogenic diet. This figure shows a comparison of what Phinney calls the “well formulated ketogenic diet” (WFKD) as a triangle with a number of possible dietary approaches shown for comparison. 
A WKFD can contain 30% protein and 5% carbs or 20% carbs and 10% protein. A WKFD, however, cannot, however, contain 30% protein and 20% carbs because we would get too much glucose which would increase insulin and suppress ketosis.
As shown in the WFKD figure above the protein content of a ketogenic diet can range between 0.8 and 2.4g/kg lean body mass. However, if we are running higher levels of protein we will only achieve ketosis if we also limit carbohydrates.
Listen to Steve Phinney discuss this concept from 2:51 in this video.
Interestingly, the slope of the line along the face of the WKFD triangle corresponds with the assumption that 7% of protein goes off to muscle growth and repair with 75% of the remaining ‘excess’ protein being glucogenic. This also aligns nicely with the observation from the food insulin index data and the theoretical proportion of glucogenic amino acids in protein.
the Goldilocks glucose zone
Listed below are the various levels of glucose requirement in terms of calories discussed above along with the equivalent carbohydrates and the percent of glucogenic calories in a 2250 calorie diet.
insulin load (g)
glucose utilisation (Jaminet)
ketogenic threshold (Phinney)
ketogenic maintenance (Jaminet)
The glucose utilisation is Jaminet’s approximation of the glucose calories used by a non-ketogenic person each day. If we run above this level our glycogen stores will become overfull, with excess glucose spilling into the blood, requiring insulin and being stored as fat. Below this level, we need to obtain some of our glucose from protein via gluconeogenesis.
The ketogenic threshold represents the theoretical boundary between the WFKD and the rest of the world according to Phinney’s protein vs carbohydrates plot. Below this point, our glycogen stores will become depleted to a point that we be forced to rely on our protein and fat stores for energy rather than carbohydrate. After a period of consuming fewer carbs than required to keep our glycogen stores topped off, we will start to show ketones in our blood and rely on ketones and fat more than glucose. This level is about 500 calories per day which is about 22% of a 2250 calorie per day diet.
The ketogenic maintenance level is based on the 300 calories per day that Jaminet says we need from glucose if we are fat adapted. With a greater proportion of energy coming from fat in the form of ketones we require less glucose for brain function.
The starvation level represents what people can survive on as an absolute minimum. In this extreme starvation state, the body is cannibalising muscle via gluconeogenesis to convert to glucose to survive. This is not something I recommend you try at home. However, it is useful to know that the body can survive (but not necessarily thrive) at very low levels of glucose for a significant period of time.
The chart below shows these glucose levels superimposed on a plot of protein versus carbohydrate. The points on the left-hand side of the chart labelled with calorie values represent the point at which all glucogenic calories come from carbohydrates with only the minimum 7% protein for maintenance ingested (i.e. no “excess” protein).
As we move to the right we have increasing levels of protein and decreasing levels of carbohydrates to maintain the same total number of glucogenic calories (assuming that 75% of “excess” protein converts to glucose).
The only thing we can be certain of here is that the concepts shown graphically in this figure will not be accurate due to the fact that it is built on a number of layers of theory. And everyone’s body is different. However, this chart gives us a conceptual framework with which to manipulate our diet to achieve our goals.
The take home message is that, if we are trying to reduce the glucose load of our diet to the point at which our own pancreas can keep up, we need to think, not just in terms of carbohydrates, but in terms of total glucose (or insulin load) from carbohydrates plus excess protein.
I don’t think the body minds that much whether it gets glucose from carbohydrates or protein.  My view is that it is better to maximise vitamins (generally from carbohydrate containing foods) and amino acids (from protein containing foods) as far as possible while at the same time keeping our glucose load within our own pancreas’ ability to keep our blood sugars at normal levels. What this means is that some people may need to restrict their carbohydrates and their protein more than others to achieve normal blood sugars.
what about the Kitavans?
When faced with the hormonal theory of obesity many people are quick to point to hunter gatherer populations such as the Kitavans that do quite well on high levels of carbohydrates.
Some people seem to tolerate high levels of carbohydrate from whole food sources. Perhaps they are metabolically flexible such that they can store carbohydrates as fat and quickly use them again, or they are very active and hence using up their glycogen stores regularly, and are very insulin sensitive and adapted to handle significantly more than 600 carbohydrate calories per day from whole food sources.
It may also be that people eating predominantly unprocessed high fibre foods are less likely to be in a caloric excess meaning that they do not have a lot of left over calories to store as fat or to require excess insulin.
Dr Jason Fung points out in this video that in spite of a higher glucose load the Kitavans managed to keep low insulin levels, which seems to be the critical factor.
If you are highly active with great insulin sensitivity and you can consume high levels of carbohydrates while maintaining normal blood glucose and staying lean then good luck to you. I’m jealous. Enjoy, at least while it lasts!
It is worth noting that a number of the champions of the low carbohydrate movement such as Tim Noakes,  Ben Greenfield  and Sami Inkenen  found that they had or were becoming diabetic after decades of extreme exercise on a high carbohydrate diet, hence transitioned to a low carbohydrate approach to manage their blood sugars.
comparison of dietary approaches
To help make more sense of this concept I have shown a number of dietary approaches from the article Diet Wars… Which One is Optimal? on the protein vs carbohydrate chart below.
Bernstein’s approach is designed to be high protein, low carb, to provide diabetics with their glucose needs from protein which releases glucose more slowly than carbohydrate.
This version of the Atkins diet is unlikely to be ketogenic due to the high levels of protein. Reducing carbohydrates and/or protein is likely to be necessary to achieve ketosis, and possibly the weight loss that is typically the aim of the Atkins diet.
The Zone and Mediterranean diets, though generally thought to be moderate carbohydrate dietary approaches, are still well above Jaminet’s glucose utilisation threshold.
Terry Whals’ Paleo Plus approach achieves a good balance between maximising nutrition through the use of high fibre vegetables and MCT oil without excess protein.
The 80% fat diet approach is below the ketogenic maintenance level of 300 glucogenic calories per day but still above starvation ketosis. Personally, I think it would be hard for most people to get optimal levels of vitamins, minerals, fibre and possibly protein at these levels without supplementation or focussing on nutrient dense organ meats. However it may be desirable for someone using ketosis therapeutically for something like cancer or epilepsy.
The typical western diet contains between 40 to 50% carbohydrates, 35 to 40% fat and 15 to 20% protein.  The figure below shows that between 1970 and 2000 carbohydrate intake increased from around 42% to around 49% for men while protein intake has largely stayed constant. During this period obesity increased from 14.5% to 30.9%. 
It’s fair to say that macronutrient composition is only part of the story, but perhaps if we moved the carbohydrate intake back towards the ketogenic corner (along with a shift to more whole unprocessed foods) this trend would turn around again?
what is our light on the horizon?
So how do you decide what dietary approach is optimal for each individual? What is right for you? What is the lighthouse on the horizon that you can guide your boat of metabolic health towards?
Back in the Diabetes 102 article we reviewed a number of risk factors that appear to be related to blood sugar control such as the heart disease risks shown in the chart below. 
Building on this I developed this table showing the relationship between HbA1c, average blood sugar and ketone values for different heart disease risk categories.
average blood sugar
Everyone should be striving for optimal blood sugar control in order to manage their overall health and reduce a plethora of risks.
The point where you achieve excellent blood sugar control (i.e. average blood glucose less than 5.4mmol/L) is about where most people will start to show low levels of ketones in their blood. This is likely to be somewhere around Phinney’s ketogenic threshold (orange line in the protein / carb plot).
People with more severe issues such as extreme insulin resistance, epilepsy, morbid obesity or cancer may choose to push deeper into ketosis beyond the point of simply achieving normal blood sugars and normal HbA1c. This may require more discipline, intentional supplementation and limitation of food selection than most people are willing to invest.
what gauges do we use to steer the boat?
The most successful diets are the ones that people can stick to.
To this end I have developed a list of optimal foods that prioritises low insulin load, high fibre, nutrient dense foods based on your personal goals (e.g. weight loss, blood sugar control, nutritional ketosis, athletic performance or therapeutic ketosis). I have also developed this database of optimal meals that will enable you to easily choose simple everyday meals that will provide high levels of nutrition while achieving a low insulin load.
If you have diabetes or insulin resistance then I recommend that you track your blood sugars and ‘eat to your meter’. You will quickly learn what meals raise your blood sugars and hence what to avoid.
With the understanding that non-fibre carbohydrates plus excess protein raise blood sugar and require insulin you can work to manage your diet until you achieve the excellent blood sugar levels with a reduced or ideally eliminated reliance on medications.
Many people benefit from journaling or tracking food intake on an app such as MyFitnessPal or Cronometre. Rather than looking at calories or carbohydrates I encourage you to consider insulin load which can be calculated using this formula.
As shown in the table above, you will likely need to get below an insulin load of 150g per day to be under the blue line and under 125g per day to be ketogenic.
While I don’t think it is healthy, natural or normal to consciously monitor everything you eat for extended periods, many people find it useful for a period of time to retrain their habits or to help guide them toward a short term goal.
As a worked example I have calculated the insulin load, % insulinogenic calories as well as the % carbs and % protein for Deshanta from the Optimising Nutrition Facebook group who provided her MyFitnessPal food diary which is summarised in the table below.
insulin load (g)
I’ve also plotted this on the chart below indicating that her diet puts her just outside the realm of a ‘well formulated ketogenic diet’. If she wanted to improve her blood glucose control further she could consider moving back towards the more ketogenic bottom left of the chart by reducing carbohydrates and / or protein.
If you’re interested in seeing how you can refine your diet to balance your blood sugars with consideration of your blood sugars and glucose load as well as your vitamins and amino acid you could join this closed Facebook group.
what are the levers we can use to steer the boat?
In order to reduce the insulin load of our diet we should do the following:
Increase fibre from non-starchy vegetables (e.g. spinach, mushrooms, peppers, broccoli etc). These will provide vitamins and minerals as well as indigestible fibre that will feed the gut which will also improve insulin resistance.  Increasing fibre in our diet will increase the bulk and the weight of our food without increasing calories or insulin and will tend to decrease our cravings for processed carbohydrates.
Reduce carbohydrates, particularly ones that come in packages with a bar code. Enough said.
If you are not getting the desired results, look to reduce your protein intake until you are achieving excellent blood sugar control and/or your target HbA1c.
If you are still not getting the results you want then look at some form of intermittent fasting to improve your insulin sensitivity and to kick-start ketosis. 
Once you are achieving normal blood sugars you may want to occasionally test your blood ketones to confirm you have achieved nutritional ketosis; however tracking your blood sugars will be adequate for most people.
Once you have achieved your desired level of blood sugars, weight and metabolic health you can drop back to monitoring less frequently, just to make sure you are not regressing and then ramp up the efforts again if required.
Then, go outside. Move. Have fun. Find a hobby. Enjoy life! And stop thinking so much about food!
Looking at the foods sitting above the trend line the chart below it appears that the foods with the greatest insulin response compared to what would be predicted by carbohydrate and protein tend to be the ones that are more processed such as ice-cream, baked beans, pancakes and Jelly Beans.
On the lower side of the trend line we have less processed foods (e.g. full cream milk, navy beans, porridge and All Bran with added fibre).
Processed foods tend to contain less fibre, while carbohydrates in their original state typically contain more fibre. Fibre is indigestible carbohydrate and hence does not raise blood sugars or require insulin.
Fibre is also important for the health of our gut and feeds the good bacteria in our digestive tract.
Could it be possible to also use fibre as a proxy for the level of processing to help refine the prediction of insulin demand by different foods?
In order to test whether fibre is useful to predict insulin demand I tested the relationship between carbohydrates plus different amounts of the fibre in the various foods.
The best correlation was achieved by removing all of the fibre. Using net carbs gives an increased correlation compared to the carbohydrates alone (i.e. R2 = 0.435 compared to R2 = 0.482).
Considering the carbohydrates, protein and fibre in a food enables us to more accurately predict insulin demand.
This concept is known in the low carb community as “net carbs”. If you’re trying count carbohydrates to manage insulin people are often advised to consider the total carbohydrates minus the fibre as fibre cannot be digested but is rather digested by our intestinal bacteria.
This aligns with the understanding that carbohydrates consumed with the packaging that they came with (i.e. fibre) do not have as big an effect on insulin.
This relationship might be part of the reason why many populations have maintained good health on a higher level of carbohydrate consumed in their raw natural state compared to when they come from the supermarket in boxes with barcodes.
There is some disagreement on how to deal with fibre on a restricted carbohydrate diet:
Some people say you should ignore fibre because “net carbs” is just a marketing ploy.
Some people choose to count half the fibre as carbs as a middle ground.
Experienced type 1 diabetics who monitor their blood sugars using continuous glucose meters will tell you that the fibre in their veggies will not raise their blood sugars however they ignore the fibre in packaged foods because it does raise their blood sugar.
My interpretation of the food insulin index data indicates that real fibre in foods is indigestible and hence does not raise blood sugar and require insulin. However excess cooking and processing will soften this fibre and make it digestible.
It appears that the fibre in a food is a useful proxy for the level of processing and helps us to better predict the insulin demand of a food.
If you want to reduce your insulin load you can increase the amount of non-starchy veggies such as spinach, broccoli, mushrooms, Brussel sprouts and kale in your diet.
If you’re using packaged diet products then I suggest you ignore the fibre content, or maybe don’t buy them in the first place.
The major problem I see with encouraging people to consider total carbohydrates rather than net carbohydrates is that it will encourage people to avoid vegetables which not be optimal for health in the long term. The reality is, when you take fibre into account, you can eat a lot of non-starchy vegetables without significantly impacting your net carbohydrates.
Prioritising foods that provide adequate nutrition with minimal calories increases your chances of achieving health, satiety and weight loss.
Weight loss can be achieved by eating high fibre, nutrient dense, low calorie density, low carbohydrate foods.
Eating more on days when you are active and less on low activity days will be more effective in the long term than monotonous calorie restriction.
Eating more fat than required for satiety you may not be giving your body a chance to burn body fat.
how to lose weight
Looking for a sure fire diet to lose weight, guaranteed? Try eating this every day:
mushroom – 200g
spinach – 4 cups
artichoke – 120g
raspberries – 200g
pepper – 6g
parsley – 1 cup
collard greens – 2 cups
Swiss chard – 1 cup
turnip – 200g
steamed broccoli – 5 cups
Brussel sprouts – 16 oz
mung beans – 0.5 cups
lentils – 3 cups
asparagus – 200g
mushroom – 200g
Will this meal plan lead to fat loss? Yes.
Could most people do this in long term? Probably not.
On first glance it doesn’t look like a ketogenic diet, however given that you probably couldn’t actually eat all that food in a day and you’d end up using so much of your own body fat it would probably be ketogenic.
This high fibre, high nutrient density low calorie density would require you to eat a massive four kilograms (nine pounds) of food a day to get 2000 calories.
The positives of this approach are:
extremely low calorie density,
extremely high fibre (150g per day compared to the average western intake of 17g per day),
extremely high nutrient density,
extremely filling, and
although 70% carbohydrates, the massive amount of fibre means the insulin load is only moderate, making it better for a diabetic than the typical western diet.
The negatives of this approach are:
without any fat in the diet you may not be able to actually absorb all the nutrition from the fat soluble vitamins A, E and K,
vitamin B, vitamin D, cholesterol and saturated fat are non-existent,
protein quality is only moderate without any animal protein, and
it may be hard to cook many of these foods without any added fats.
If you’re interested in a ketogenic diet you’re probably not going want to follow this sort of extreme vegetarian-style diet. However there are a few things that we can learn from this approach that we could incorporate into a ketogenic approach.
high fibre, low calorie density
Eating high fibre, low calorie density foods will help to keep you full. Non-starchy vegetables are bulky, contain a lot of water, fibre as well as lot of nutrients.
While counting calories will work over the short term, your body will win out over your mind and your iPhone app in the long term if you’re not giving it the nutrients it needs.
Recent research  suggests that we will keep eating until we get enough protein and eating foods low in protein leads people to eat more calories than they need.
Ensuring that you’re getting adequate protein (say 15 to 30% of calories) will cause you to be satiated with less calories.
In a similar way, if you’re not giving your body the vitamins and minerals it needs it will keep on seeking out more food.
In his Perfect Health Diet Paul Jaminet notes that a nourishing, balanced diet that provides all the required nutrients in the right proportions is the key to eliminating hunger and minimising appetite and eliminating hunger at minimal caloric intake.
If you keep your calorie intake consistently low for an extended period of time your body will sense an impending famine and slow down your metabolism, leaving you tired, cold, depressed and miserable.
Don’t be afraid to mix it up a bit with restricted calories a few days a week by missing a few meals on low activity days and then eating to satiate your hunger on higher activity days.
In our current food environment we don’t give our body any time when it’s not awash with calories and insulin than enable your bodies to use our stored body fat for energy.
Eat when you’re hungry. But conversely, don’t be afraid to not eat when you’re not hungry.
“Break-fast” is an important meal, even if it occurs at 3pm in the afternoon!
eat fat to lose fat?
The reason that eating a high fat diet leads to increased satiety is that your body can access your stored body fat.
In most people eating a ketogenic diet leads to greater satiety because you’re using body fat for fuel, which leads to a reduction in food intake.
Conversely if you are eating a diet full of simple carbohydrates your insulin levels will stay high and your body fat will be locked away.
When you lose fat, your body burns the saturated fat on your body. If at first you don’t succeed by reducing your insulinogenic load and intermittent fasting consider cutting back your dietary fat intake to create a caloric deficit which will be filled by your body fat. 
Some people can eat massive amounts of fat while keeping carbs low and lose weight,  however others can lose their way on a LCHF or ketogenic diet by eating too much dietary fat and end up not getting the results they hoped for.
Jimmy Moore emphasis that you need to eat fat to satiety.  If you mainline dietary fat and are not hearing your natural satiety signals you’re not going to give your body the best chance to burn body fat.
One of the most famous diet studies looking at low carb diets is Dr Chris Gardner’s A to Z Study.  Gardner, a practicing vegan, was surprised to find that it was the Atkins dieters who lost the most weight in his study.
More interestingly though were the results of a follow-up analysis where he assessed peoples’ insulin resistance. He found was that people who were insulin resistant lost the most weight on the low carb diet while the insulin resistant lost nothing on the higher carbohydrate diets. 
How do you know if you’re insulin resistant? Your weight and waist line are pretty good indicators, but your average blood sugar is even better. If you want to know what diet is right for you, pick up a blood sugar metre from your local chemist and do some testing.
If your average blood sugars are in the excellent range according to the values below then focussing on carbohydrates as your primary goal may not be ideal.
average blood sugar
food choices for weight loss
We can use the food prioritisation system  to identify foods that align with these goals by prioritising nutrient density (20% weighting), fibre (10% weighting), and low calorie density (30% weighting).
ND / calorie
fibre / calorie
ND / $
ND / weight
calorie / 100g
$ / calorie
The resultant foods are listed below, in order of priority, using these weightings.
A few items that you would not generally expect to see on a ketogenic diet come to the top of the list such as lentils and mung beans due to their low calorie density, high fibre content and low cost.
This weighting system does not give a high priority to fats and oils as they are coming from the body fat stores. The list of nuts and seeds is also quite short in view of their high calorie density.
I’ve also developed this ‘cheat sheet’ using this approach to highlight optimal food choices depending, wither they be reducing insulin, weight loss or athletic performance. Why not print it out and stick it to your fridge as a helpful reminder or when you’re looking for some inspiration for your next shopping expedition?
vegetables & spices
nuts, seeds & legumes
fats and oils
example daily diet
Below is an example daily meal plan for someone wanting to lose weight by reducing calorie density and maximise nutrition using the prioritised list of foods above. There’s nothing radical or objectionable here other than the high amounts of nutrient dense green veggies you need to eat in a day. Some added fat is used for cooking. There are no snacks and no calorie dense nuts and seeds.
Using this approach we achieve great nutrition and protein scores along with an impressive 36g of fibre per day.
This approach involves eating nearly two kilograms of food which would leave you feeling quite full.
Although this diet is full of veggies it still has 60% of the dietary calories coming from fat. If we ran a 1/3 calorie deficit in the early stages of a weight loss program we would have 73% of the calories coming from fat when your body fat is included. This would very likely be ketogenic.
With the high amount of fibre, the net carbs are quite low at 44g per day which would still qualify as low carb diet.
If you find your blood sugars are unacceptably high you should consider backing off on the carbohydrate containing foods. On the other hand if your blood sugars were excellent you could even consider increasing the non-starchy veggies to increase satiety and reduce the calorie density.
In our next article we’ll look at nutrient dense foods options that might work for you if your blood sugars are excellent and you’re doing intense exercise.