Saturday, July 25, 2015

Refined Sugar Worsens Blood Lipid Markers of Cardiovascular Disease

Blood lipids such as LDL and HDL cholesterol are markers of the biological processes that impact cardiovascular disease, and they are commonly measured to assess cardiovascular risk.  When we think about the impact of food on blood lipids, dietary fat typically comes to mind.  Yet a new study shows that dietary carbohydrate, specifically high-fructose corn syrup, can have a large impact on blood lipid markers of cardiovascular disease risk.


Dietary fats have well-established impacts on blood lipids.  For example, in short-term feeding trials, saturated fat tends to increase total cholesterol, increase LDL ("bad") cholesterol, and increase HDL ("good") cholesterol, while the omega-6 polyunsaturated fat linoleic acid decreases total cholesterol and decreases LDL cholesterol.  For this reason, dietary advice to reduce cardiovascular risk tends to focus on dietary fat.

The hypothesis that refined dietary sugar is harmful to the cardiovascular system isn't new.  In 1972, British physiologist and nutrition researcher John Yudkin published a classic book called Pure, White, and Deadly, which argued, among other things, that refined sugar is harmful to the cardiovascular system.  Yet at the time, the supporting data were weak, and the hypothesis was never taken very seriously by the scientific community.

Peter Havel and his group at UC Davis have begun to breathe new life into this hypothesis with their rigorous work on the cardiovascular effects of dietary sugars.

The study

Havel's team recruited 85 lean to obese volunteers that didn't have diabetes, kidney or liver disease, or high blood pressure (1).  They divided the volunteers into four groups, each of which received a different amount of high-fructose corn syrup (HFCS) beverage daily for two weeks:
  1. 0 percent of calorie requirements as HFCS
  2. 10 percent of calorie requirements as HFCS
  3. 17.5 percent of calorie requirements as HFCS
  4. 25 percent of calorie requirements as HFCS
A non-caloric sweetener (aspartame) was added to the 0 and 10 percent beverages so that they had a similar sweetness level as the 17.5 and 25 percent beverages.

After the two-week period, Havel's team measured the volunteers' body weight, blood lipids, and uric acid level (a marker of gout risk and also possibly metabolic health).


After two weeks of drinking their assigned beverages daily, body weight increased with increasing HFCS dose.  This became non-significant after additional statistical adjustment (adjustment for multiple comparisons, which is rigorous), but the trend is still clear.  People in group 4 ended up gaining 1.8 lbs (0.8 kg), and the amount of weight gained was directly proportional to the dose of HFCS. 
Effect of HFCS dose on body weight.  Adapted from Stanhope et al., AJCN, 2015.
This is called a "dose-response relationship" because stepwise increases of the HFCS dose led to stepwise increasese of body weight.  A dose-response relationship provides very strong evidence that an effect is real.  

Havel's group also found dose-response relationships between HFCS dose and blood lipid levels.  The higher the HFCS dose, the higher the LDL climbed over the two-week period.  While LDL decreased slightly in group 1 receiving no HFCS, it increased by 18 percent in group 4.  ApoB concentration, a marker of LDL particle number, increased by 19 percent.

Effect of HFCS dose on fasting LDL cholesterol.  Adapted from Stanhope et al., AJCN, 2015. 
Uric acid levels also increased in a dose-response fashion with increasing HFCS.


This new paper from Peter Havel's group suggests that supplementing the diet with HFCS leads to weight gain and detrimental changes in blood lipids.  The increases in LDL cholesterol and ApoB concentration they observed would be expected to lead to a significant increase in cardiovascular risk.  The higher the dose of HFCS, the worse the outcomes.

This study is certainly interesting, but it requires some interpretation.  The first thing to note is that calories weren't held constant, and volunteers were given calorie-rich beverages to drink in the context of their habitual eating patterns.  Usually in trials like this, when you ask people to add a calorie-rich food to their normal regimen, they don't fully compensate for the excess calories and end up overeating.  Judging by the increase in body weight, that appears to have happened here.  

Many studies have evaluated the impact of refined sugar on body weight, metabolism, and cardiovascular risk markers.  John Sievenpiper, a nutrition researcher at the University of Toronto, and his colleagues have conducted systematic meta-analyses of the studies on fructose-containing sugars, and their results show that there is a large difference between studies in which calories were held constant and those in which sugar-eating groups increased their calorie intake.  Basically, when calories are held constant, fructose has little unique impact on weight, metabolism, or cardiovascular risk factors (2, 3, 4, 5, 6).  When sugar intake and calorie intake rise in parallel, sugar increases body weight and begins to cause harmful metabolic and cardiovascular changes.  My interpretation of the evidence is that refined sugar exacerbates the harmful effects of overeating.  

So the increase in calorie intake that presumably occurred in Havel's study is highly relevant to the outcome he observed.  Yet, we need to remember that refined sugar tends to promote overeating, so in real life, it's very common for a high sugar intake and overeating to co-exist.  At the same time, when a person makes a major dietary change such as adding a large quantity of sweetened beverages to his diet all of a sudden, there is an adaptation period during which he gradually adjusts the rest of his diet to the change.  My suspicion is that the calorie intake of group 4 went way up at first, but that it would have gradually stabilized at a more normal (but perhaps still a bit too high) level as they began to compensate for the excess calories.  In other words, I would expect the long-term effects on weight and blood lipids to be somewhat attenuated relative to what was reported in this two-week trial.

Elevated uric acid is a marker of impaired metabolic function and increased gout risk.  The increased uric acid in the HFCS-consuming groups is not good, and probably reflects a combination of overeating and high fructose intake.  

Although Havel's team used HFCS for this study, the evidence overall suggests that HFCS is pretty much the same as sucrose (table sugar).  For me, the result is relevant to all forms of refined sugar that are about half fructose and half glucose.

In the end, I think this new study does raise serious questions about the health impacts of eating refined sugar-- particularly for cardiovascular health.  As we collectively recover from our 'sugar mania' phase, in which every ailment under the sun was blamed on sugar, I think we can settle into a more nuanced view of refined sugar as a dietary factor that interacts with excess calorie intake to adversely affect body weight, metabolic health, and the risk of cardiovascular disease.

And as a reminder, there is still no evidence whatsoever that sugar from fresh fruit is fattening or harmful to health.  The evidence currently suggests that fresh fruit, if anything, is slimming and promotes good health.  This is probably because fresh fruit doesn't favor overeating, and it also delivers an array of other beneficial substances such as polyphenols and fiber.


Reijo said...

One more way how fructose and HFCS may increase the risk of CVD in presence of overeating might be the fructose's ability to increase absorption of iron. Animal studies and preliminary human studies suggest that fructose increase iron absorption which in turn may lead to physiologically too high iron stores. Elevated iron stores are a risk factor for type 2 diabetes and CHD as you have written in meat series. Iron intake tend to be high among those who consume high amount of calories and red/processed meat. One more mechanism behind the known association between Western junk food diet and cardiometabolic diseases?

Christides T, Sharp P (2013) Sugars Increase Non-Heme Iron Bioavailability in Human Epithelial Intestinal and Liver Cells. PLoS ONE 8(12): e8303

Stimpaka said...

It would be nice if they had used glucose to balance the calorie contents. Also the remarkable change to my eyes is the difference between the 17.5 and the 25% groups.

Dr. L. Lee Coyne said...

Too much time spent on "lipid profiles" as risk factors when they are really of little consequence. Inflammatory markers are much more interesting.

aluchko said...

Agreed on the calorie contents, from the weight perspective the conclusion seems to be that if you ingest a bunch of extra calories you gain weight (neat to see though not really shocking). I guess the one interesting thing is that if artificial sweeteners do cause weight gain through some mechanism then that mechanism takes more than a couple weeks to do it.

As for the LDL, I don't understand the biology that well, but would that be related to the weight gain? The fat their gaining has to come from somewhere, is the LDL the thing transporting it into the cells?

karl said...

Welcome to the 1960's

First, consider these
links to the forgotten science.

Now it becomes interesting that when people are not eating monocotyledon derived foods (which only entered human food about 5000 years ago) - TRYGL levels are around 50mg/dL.

The effects of unnatural TRYGL levels is becomming known:

High insulin causes Leptin resistance

High triglycerides also nearly instantly block leptin transport across the blood brain barrier

Note that BOTH of the sugars - Glucose and Fructose mess with leptin sensitivity.

Genetic adaption to tolerate high carbohydrates diets is new and quite varied - Lactose tolerance has evolved twice - and tolerence to the high monocotyledon diets of today are similarly variable by origin.

Exposure to BG levels above 110 does not represent simply a linear risk as assumed by many - it is exponential - thus we have a process that glycosylation lipids(and fructose is thought to be a magnitude more reactive) - increasing their oxidation - creating oxLDL - confused by our immune systems for dying bacteria - that recruit macrophages - which engulf them forming foam-cells in artery walls.

IMO the normal level of both BG and Trygl need to be redefined. There are two simple ways to bring these levels down to something that is healthy - diet and exercise:

The key is it must be high stress exercise - weight lifting - sprints. (Personally when I started doing squats - my fasting BG went from 95 to 85 and my tolerance for carbs greatly improved). Telling people to go for walks - not to over do it is thus poor advice. (low stress exercise leads to repetitive stress injury - and takes a magnitude more time to achieve the same effect). Simply looking at bones of people from 100 years ago tells us that our modern lack of physical stress is quite unnatural - leg bone about half the diameter - reduced muscle mass..

Suza said...

This blog is a great resource to learn about the addictiveness food. Do you have any advice about how to restrain from highly-rewarding food without having a 'dieting' mindset and inevitably compulsive binge-eating? Intuitive eating and putting junk food on the bad foods lists seem to be clashing actions.


Chris Davies said...

It would be interesting to compare the effects of refined sugar, saturated fat, polyunsaturated fat (in the form of refined vegetable oils), and then palatable mixtures (sugar+fat) on weight gain and overeating.

Great Blog, I think we're starting to get the picture as a culture that saturated fat and animal proteins in general aren't inherently bad but that doesn't mean demonizing sugar is the answer. Low-fat and low-carb are two sides of the same coin and somewhere in the middle lies an omnivorous diet that can include any range of macronutrients or calories as long as those foods are whole foods, prepared naturally and there is at least some meat or dairy included.

Karin said...

The full paper is behind a paywall, but from the abstract and what I've read here, I'm think the effects are conflated.

Without calorie control you can't be sure if it's the HFCS, other sugars, or overeating in general causing the weight gain. One weekend of partying will put me up 3-4 lbs, due to water retention after too much salt. Without diet control you can't be sure if it's the HFCS, other sugars, any other food or even alcohol consumption driving the blood markers. What about exercise and other lifestyle differences?

I'm not sure how you can draw any conclusions with so many uncontrolled factors during the outpatient days.

Glycogenica S said...

This study was done on HFCS so it would be interesting to see a study like this that used cane sugar instead of HFCS.

Glycogenica S said...

How does this relate to your post from 2012?

Stephan Guyenet said...

Hi Glycogenica,

I'd say it fits well with my previous piece. My conclusion there was that refined sugar is not uniquely fattening, but it can be fattening if it increases calorie intake, which it often does due to its palatability and energy density. I also concluded that in excess, refined sugar can promote the metabolic syndrome (which includes detrimental blood lipid changes).

rusticj said...

Given political will, the sugar habit should not be impossible to break, extrinsic sugar having played virtually no part in human nutrition up until 150 or so years ago, as you, Stephan, have shown. Increasing pressure for additional taxes on sugar-sweetened soft drinks, which we are now seeing in several countries, may be a portend. Let's hope something will be done before healthcare systems buckle under the strain of coping with the avoidable ill-health that poor nutrition and lack of exercise are causing.

If a visitor from Mars were to ask for a simple guide to healthy eating on this planet, I would say switch on a television and watch the food and soft-drink advertising. Then avoid like the plague all that you see there, including the so-called restaurants. Any food which bears no resemblance whatsoever to the things you see advertised on television is probably OK. Learn to cook, using basic ingredients.Try to make your diet as varied as possible. Sample and enjoy our planet's traditional food cultures in all their richness and diversity. But don't overdo it. You should never feel completely full.

I was interested during a recent visit to France, where foodway trends are as concerning as anywhere (it's even difficult these days to find a drinkable cup of coffee in that country), to see that health warnings now accompany all food advertising on television. Intriguing when a snack is advertised with a streaming ticker urging viewers, for the sake of their health, not to eat between meals. Or to see products which are maybe high in two of the three accompanied by a warning against eating too much salt, sugar or fat. Could it be that where tobacco reluctantly led, commercially-processed food will be forced, equally grudgingly, to follow?

Viewers of this advertising are referred to an attractive and informative website,
which has been set up by the PNNS, a government-supported public health agency. It suggests nine markers for a good life and a healthy one:-

1) At least five portions of fruit and vegetables a day
2) Three dairy products per day (four for children and adolescents)
3) Starchy foodstuffs (bread, cereals or pulses) with each meal, whole-grains where possible
4) Meat, fish or egg once or twice a day
5) Restrict fats
6) Restrict consumption of anything to which sugar has been added
7) Restrict salt
8) Consume water freely during and between meals
9) At least 30 minutes of vigorous walking or its equivalent each day (an hour for children and adolescents)

Suggested dairy products and portion sizes are yoghurt (125g), an individual fromage blanc (100g ), two petits suisses cheeses (60g), 30g of other cheese, or one glass of milk. Crème fraîche and butter, being high in fat but low in calcium, are not included.

Ah, I dream about those wonderful cheeses, especially the ones from artisan producers herding a few cows in the mountains, made with raw milk, and savored after they have been a day or more at room temperature! If consumed in moderation (and given their price, they are unlikely to be consumed other than in moderation), how could they ever have been thought bad for us?

That's what awakens the reward sensors in my brain, Stephan, not the slickly-marketed, sugar-laden, over-processed, synthetic junk you talk of elsewhere in your blog as 'high reward'. Perhaps your use of the expression in this context is ironic. But I don't think so.

I enjoy your blog even if some of it makes little sense to me. I will persist in trying to understand. The subject fascinates. Keep up the good work.