high glycemic index diet causes obesity: a calorie is not just a calorie

David Ludwig, long-time promoter of the low glycemic index diet, has just published an update (see glycemic index carbohydrate insulin model jamaintmed2018 in dropbox, or doi:10.1001/jamainternmed.2018.2933), with more convincing data suggesting that the high glycemic index diet is associated with weight gain even in the absence of excessive calories.

Details:
-- recent commentaries have suggested that there is an inherent setpoint in people’s weight, that if they lose weight they tend to regain their lost weight, etc. However there are several lines of argument against this, especially if the setpoint argument is a genetic one.
    -- There clearly has been a secular shift to increasing weight since the 1970s, paralleling the emphasis on low fat diets (read: high carb diets). The average body weight in the United States and Western Europe is 25 to 30 pounds greater than it was 50 years ago.
    -- There was also an interesting article suggesting that those with the highest genetic predisposition to gaining weight actually were able to lose the most weight by diet. See http://gmodestmedblogs.blogspot.com/2018/01/dietary-effect-strong-when-high-genetic.html
-- there has been a recent update in both the understanding of glucose and insulin in increasing weight, as well as the role of a low glycemic index (GI) diet. The glycemic index is a quantitative measurement of the blood glucose 1-2 hours after eating, with sugar or white bread considered to be the standard high-GI food. the glycemic index of foods does not necessarily coincide with how sweet they seem to taste, but more with the specific carbohydrate composition (most refined starchy foods have high GIs vs nonstarchy veges, fruit, legumes). For example: corn's GI is very high (increasingly so as it gets more processed: corn meal or pasta, corn flakes, chips), bread is high (but sourdough bread is much better), rice is bad (but basmati rice is okay), potatoes (but the sweeter-tasting sweet potatoes/yams are better than white…), watermelon and tropical fruits are bad (but the sweeter-tasting grapes are okay).  And, the glycemic index of a food is pretty consistent independent of the other foods consumed with it (though it will be somewhat lower if consumed with fats, which delay the glucose absorption: for example, the GI of a Mars bar candy is much lower than sugar). Also, the GI of foods does not differ much between diabetics and nondiabetics
--glycemic load is calculated by multiplying the grams of available carbohydrate in the different foods consumed times the glycemic index and then dividing by 100: i.e. the composite glycemic index for the meal.  the glycemic load of a meal is the “best single predictor of postprandial blood glucose levels, explaining up to 90% of the variance”
-- the original glycemic index paper by Ludwig detailed a purported mechanistic explanation of the effects of a low glycemic index diet: within 2 hours of eating a high glycemic load meal there is relative hyperglycemia, leading to GLP-1 and glucose-dependent insulinotropic polypeptide to stimulate pancreatic insulin release, which then stimulates gluconeogenesis and lipolysis. This high insulin response leads to relative hypoglycemia several hours later, then to suppression of free fatty acids (FFAs), low circulating concentration of metabolic fuels, and then to a counterregulatory hormone response to return to normoglycemia, and increase FFAs
-- the new and improved model is an extension of this last one, but includes a few additions:
    -- in its “carbohydrate-insulin model”, high glycemic index foods lead to postprandial hyperinsulinemia which directly promotes the deposition of calories into fat cells, instead of oxidation in lean tissues and a quick energy source when consuming a lower GI index diet (ie, there is a partitioning of calories to those being deposited in fat tissue vs accessible energy sources)
        -- in effect, insulin’s role in shifting the energy balance from carbohydrates into fat storage ultimately restricts energy available in the blood, triggering a starvation response with increasing hunger, lowered metabolic rate, and elevated stress hormones
    -- protein also stimulates insulin secretion, but also elicits a secretion of glucagon which antagonizes the effects of insulin. Dietary fat has little effect on insulin secretion (which may explain the weight loss of the very high fat Atkin's diet)

Data supporting the above:
-- injecting insulin in animals promotes fat deposition, increases hunger, and leads to weight gain
-- even with restricted calories, rodents on high GI diets have increased adipocyte diameter, greater adiposity, lower energy expenditure, and increased hunger (i.e., a significantly calorie-restricted high GI diet still led to increased adiposity)
-- an interesting genetic analysis, using Mendelian randomization, assessed the relationship between insulin secretion and body mass index, finding that people with genetically-determined high insulin secretion tended to have a high BMI, whereas those with genetically determined high BMI did not necessarily have high insulin secretion: i.e., it’s the insulin secretion that seems to cause obesity and not vice versa
    -- also, variants in insulin promoter gene associated with insulin hypersecretion predicted weight gain in human adolescents
-- high carbohydrate diets also seem to be associated with lower resting and total energy expenditure, similar to what is found in starvation, whereas low glycemic index diets do not lower the metabolic rate (see glycemic index dec energy expend jama2012 in dropbox, or JAMA. 2012;307(24):2627-2634), further supporting the carbohydrate-insulin model
-- several studies have also suggested that high carb diets stimulate the appetite, for example finding similar total calorie intake in severely obese patients who have either 500 cal/day restricted low fat diet vs unrestricted low glycemic index Atkin’s diets (ie, less need to limit calories in a low-GI diet vs a low-fat diet)
-- http://gmodestmedblogs.blogspot.com/2013/07/glycemic-index.htmlreviews some of the older studies, along with Ludwig’s original article (see glycemic index Ludwig in dropbox, Ludwig DS. JAMA 2002; 287:2414), including a study showing that a high-GI diet was associated with a lower plasma glucose in the late postprandial period, and increased hunger and stimulation of brain regions associated with reward and craving  (perhaps affecting eating behavior at the next meal). See glycemic index food craving amj jclin nutr 2013 in dropbox, or Am J Clin Nutr doi: 10.3945/ajcn.113.064113​
-- though not mentioned in these articles is the inherent assumption that there is a continuum here, from a very low-GI to a very high-GI diet: ie, lowering the glycemic load of a high-GI diet should (presumably) gradually shift the partitioning of calories from largely going into fat cells to becoming more available as an immediate energy source (and even the resting brain needs lots of accessible energy, requiring >100g/d). and i have not seen any data suggesting what this relationship might be: is it a linear and predictable one??

so, i do think that a low GI diet is particularly important in several groups of people, where there have been documented benefit: diabetics, prediabetics, obese, and those with high triglycerides/low HDL. and it probably does make sense for the general population to eat more veges overall and decrease the more processed, high-GI alternatives.  the studies are mixed on the benefits of one diet over another for weight loss: the issue seems to be maintenance of the diet in the long-run (hard to do in our society with the enticing ads for cheap, very accessible, calorie dense, high carb fast foods...). but the data on the high-GI diet are pretty clear: the high insulin state seems to lead more directly to fat storage, fewer calories available directly for energy, more hunger, and lowering the resting metabolism (a "hunger" response). a pretty powerful argument against it....

and, though a recent scientific statement from the Endocrine Society commented that a calorie is just a calorie, this article suggests that a calorie is not just a calorie is not just a calorie...

 ​

geoff​

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

to get access to blogs since 8/15/17:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just type in a name in the search box and get all the blogs with that name in them

to access older blogs from the BMJ website, from October 2013 until 8/15/17: go to http://blogs.bmj.com/bmjebmspotlight/category/archive/ 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list