Sugar contributes to heart disease, what is the mechanism

I’ve read some info saying extra sugar in the diet contributes to heart disease.

http://www.heart.org/HEARTORG/GettingHealthy/NutritionCenter/HealthyEating/Added-Sugars-Add-to-Your-Risk-of-Dying-from-Heart-Disease_UCM_460319_Article.jsp

Fair enough, but why?

I get the impression they are only talking about ‘added sugar’, not the simple sugars you get from fruits or dairy.

I’m also assuming that long chain carbs like those found in staple crops or vegetables do not cause illness.

Is it a high GI diet that causes the problem, not the sugar itself? I fail to see why simple sugars added are bad but simple sugars from other sources are ok.

My questions are:

Why are ‘added sugars’ worse than sugars found naturally in food?

Why are simple carbs according to this study worse than complex carbs? Vegan diets are very high in carbs, and those are good for your heart.

Is it the GI/GL that is the problem, not the sugar itself? if so, why didn’t they just say so?

I know America gets into moral panics about food from time to time, so I don’t know if this is an extension of that or if there is real consistent evidence that those looking to avoid heart disease should avoid excess sugar.

Maybe people who eat a lot of sugar, exercise less, smoke more, weigh more, like the taste of sugar because it relieves their stress, can’t afford to ‘eat well’ (with lower socioeconomic status also linked to heart disease), . . .

Association does not equal causation.

Excess sugar, as well as large amounts of simple sugars (as opposed to complex carbohydrates), are believed to contribute to what is called De Novo Lipogenesis (ie making lipids from scratch), and DNL is associated with the formation of saturated fats, including palmitoleic acid, believed by at least some to contribute to the processes involved in the development of cardiovascular disease, when excess carbohydrate is consumed…I think…PLOS One isn’t responding, so I’m pulling this from memory. I’ll try to link the study I’m thinking of later if I can.

Sugar in fruit, veggies etc. is accompanied with nutrients, soluble fiber, and antioxidants, some of which may have beneficial effects on heart health, among other things. Eating refined, or added sugar, is void of this.

There’s a lot of controversy back-and-forth between the Paleo diet and the plant-based diets (don’t call them vegan). They each start out saying what you SHOULD eat (meat or vegetables) and then say what you SHOULDN’T (sugar). The plant-based diets got a lot of traction when Fuhrman published his diet based on his study of heart disease in Norway during WWII: the Germans took over in September of 1941 and in October confiscated all the barnyard animals. Starting in 1941, deaths by heart attack for heart disease patients dropped by a third, and continued to drop by a third of the remainder for the rest of the war. Once the war was over, the numbers shot back up to where they had been before.

“Ah hah!” said Fuhrman. “Meat causes heart disease!” So, he got himself a passel of heart disease patients and put them on a diet that removed all meat and junk food and they started getting better.

“Wait a minute…” said the Paleo people. They then took a closer look at his data, and at other data, and found that in 1940, due to those pesky U-boats disturbing shipping, the Norwegian government imposed a sugar ration: 2 pounds per adult per year, none for kids. Now, the Norwegians love them some pastries, but unleavened bread is synonymous with “not dessert”. If they wanted their treats, they had to sweeten them with something else. So, they grabbed their berry buckets and put on their walking shoes and hiked an hour up to the top of the fjord, picked berries for a while, and hiked back down again.

Also, the idea “meat causes heart disease” only works if you don’t think that fish are made of meat. By the end of the war, the average Norwegian was eating more meat than at the beginning, because fish was the only food they could get in bulk. Also, catching fish in huge nets and hauling them ashore is considerably more exercise than watching cows grow, so they were getting even more exercise.

Here’s a rebuttal of the Fuhrman diet, which includes more detail on the above and more on the effects of junk food in our diet:

A high GI is a problem because foods with a low GI take longer to digest and the sugar is broken down to glucose less quickly. This prevents huge spikes in the insulin levels. Other factors already mentioned are foods with a lower GI are associated with many nutrient and with more fiber.

Well, sure, there are always confounding variables. Good studies will account for them and separate the effect of a variable of interest independent of known confounders.

To quote the paper that was mentioned by the the link in the OP:

After adjustment for all of those factors, they still found that people in the top quintile of sugar consumption had approximately double the risk of death by cardiovascular disease. There may be other unidentified confounders, but they seem to have covered a lot of the obvious confounding variables.

(my underline)

That’s the problem with any study that’s looking for associations. The authors try to identify and adjust for confounding variables, BUT . . .

. . . were all relevant confounders identified and adjusted for?

Obviously, you can never be sure. So, are there ways of assessing whether a study did, in fact, identify and adjust for all the relevant confounding variables?

One way is to see what happens to an ostensible association (or hazard ratio or risk ratio) as you go from unadjusted, to partially adjusted, to more fully adjusted, and so forth, is to see what happens to the strength of the association as you do more adjustments. What happens to the strength of the purported association?

In the study in question (an almost impenetrable statistical jungle), we aren’t given a lot of such ‘progressive’ adjustments/comparisons. However, to take two examples, the hazard ratio of Q5 to Q1 when only adjusted for sex, age, and race was 2.43 but decreased to 2.03 when fully adjusted. Likewise, when comparing the >25% to <10% dietary calories from sugar groups, the respective HRs are 3.55 and 2.75. In other words, for both HRs, as more confounders were adjusted, the strength of the association decreased.

When that happens, i.e. when the HRs decrease as more confounders are adjusted for, it is suggestive that there may remain persistent but unidentified confounders. On the other hand, if adjustments for confounders leads to a persistently stable HR, it suggests that there are no more to be found (or so I have been taught; sorry, no cite).

tl’dr: I think there may well be residual confounders. So, at best, the results are interesting and ‘hypothesis generating’ but definitely inconclusive. More research is needed.

My lack of faith in this type of study became entrenched about twenty years ago by two separate and seemingly watertight epidemiological associations.

The first was the highly convincing, and I thought virtually certain, link between consumption of vitamin E and protection from coronary heart disease. The inverse association seen in the epidemiological studies was striking and the biologic plausibility very enticing. But, when the randomized, prospective, controlled trial was finally done - nothing. No protection.

The second epidemiologically strong and extremely biologically plausible association was between use of postmenopausal estrogen and, again, protection from coronary disease. The epidemiological studies were totally convincing in my opinion, so when you coupled them with a biological mechanism to ‘explain’ the link (and there were a number of possible mechanisms out forth, all of which made a lot of sense), it seemed to be a slam dunk. Fast forward to the RCT and, once more - nothing, no association. Well, maybe even worse than no association. Estrogens seemed to increase the risk according to the RCTs (e.g. Women’s Health Initiative Study).

So, you can see where I’m coming from. Caveat emptor when buying epidemiological research results seeming to show a cause and effect.

Perhaps vitamin E is only protective against heart disease in the presence of one or more other specific nutrients. Was the controlled study on vitamin E done with supplements? I assume so. Was the epidemiological study based on supplements, or consumption in food?

The best of the vitamin E epidemiological studies (Nurses’ Health Study) took into account ALL sources of vitamin E (both dietary and supplemental).

In terms of the possible mechanism …

The speculations vary but one popular one is Lustig’s blaming the fructose.

Fructose is present pretty much equally in both table sugar and high-fructose corn syrup (HFCS), so in this regard HFCS is only a specific villian because it has been cheap and thus added more.

The idea is that fructose is unlike glucose:

How is the fructose in real foods, like fruit, different (in this way of thinking)? First because usually there is less of it there. But per Lustig mainly because in fruit and real foods fructose comes packaged with what he thinks of as “the antidote,” fiber.

That said one does have to remember that one of two things are happening as added sugar intake goes up (either alone or in some mixed fashion):

  1. Other foods are replaced and intake otherwise decreases enough that there is no fat gain. In this case (and to the degree that other foods are replaced) the impact might not be by what is eaten but by what is not eaten: less vegetables, fruits, beans, whole grain, fish, less real food in general.

  2. Added sugar may represent additional calories. In this case the impact might be the impact of extra fat mass.

Of course all of the above - a direct negative impact of large loads of fructose suddenly delivered to the liver, with its limited ability to maetabilize fructose both quickly and safely, after a single Big Gulp; the impact of of what it replaces in the diet; and the impact of the additional calories consumed - may all contribute. And the added sugar usually comes in foods that have a whole host of other crap in it - maybe the mechanism is the other crap too.

Teasing out which plays what role may be difficult but easy to say that eating a diet high in the foods that are high in added sugar is very very likely very very bad.

Last time I checked the only, and I mean only, ailment that can be scientifically proven to be caused by sugar is tooth decay. Not hyperactivity nor diabetes nor cancer. Not even obesity.

Even that hasn’t really been proven yet:

http://www.jdentaled.org/content/65/10/1017.full.pdf

Surreal you made that cite before and again I need to point out to you that the paper you cite does NOT claim that sugar does not cause tooth decay. It merely states that in the modern era of floridation the relationship is weaker than it was. To quote from the article:

Their conclusion

Why you repetitively cite an article claiming that it says something that it does not say after your mistake has been pointed out to you before is hard to fathom.

Like I was saying earlier, America gets into moral panics about gluttony and sloth so any medical advice related to those two vices needs strong evidence.

Not only that, but what kind of diet provides 25%+ of calories from added sugar? Even baked goods and desserts usually top out at about 25% calories from added sugar since so much of them is fat and intrinsic carbs from wheat. You’d have to drink tons of soda a day to get to that leve.

Isn’t that probably what’s happening? (Not just in the U.S., but in Mexico too…)

Not just ‘what kind of diet’ but what kind of individual? What other unusual traits do they have? And, is it those other traits that are associated with the heart disease? That is the point I was making in my very first post above (and tried to justify in my subsequent posts about the pitfalls of epidemiologic research results). So, I’ll say again: association does not equal causation.

Wesley, per the article referred to in your op the diet of about 10% of US adults is the kind of diet that does.

The strength of the article is that they did control for many confounders including obesity, physical activity level, healthy eating index, and smoking, and found that CV disease risk goes up once added sugar surpasses 15% of daily calorie intake (and increases exponentially from there). That significantly reduces (not eliminates) the likelihood of my alternate hypotheses. (And leaves the unidentified confounder possibility.) Given that the adult mean is about 13%, 15% or more of intake is an amount that a sizable fraction of Americans consume. According to the editorial that accompanied the article the added sugar issue is a particularly significant issue for Black Americans as a demographic group: at the peak 25.9% of black Americans were consuming over 25% of daily calorie intake in the form of added sugars; this has since decreased to 16.9% but is still much higher than White populations.

Not that unusual of individuals with highy unusual traits.

As an academic discussion the question of whether the impact is a direct causation from the sugar or as an indirect consequence is interesting but the health related bottom-line is still the obvious one.

My mistake, I misread the abstract.

Thanks for the correction.