You have probably heard already that fructose is commonly portrayed as the fat sugar. Even Dr Lustig once said that 30% of fructose is converted to fat and the media keep repeating it. I would like to focus on this in this article because I believe the things are not so simple and we still do not exactly know enough about the lipogenic (fat forming) properties of fructose to make such statements. Make yourself comfortable, this will be long.
Dr Lustig presented results of ONE study (Hudgins et al, 2011) that examined the de-novo lipogenesis (DNL) after fructose ingestion:
I have done my own independent research on metabolic consequences of fructose intake given in various study settings for my final year project at the uni. And I know that the figures vary from study to study. Therefore - using only one study like this and demonstrate its results in figures that should be representative for everybody is not quite right. But there is far more to be criticized here. Get ready for an interesting surprise.
From my research I have learned that there are many methodologies of the assessment of DNL. All these have their strengths and limitations. I am not going to bore you with details, but I would like to say that most of these (including the one presented here) rely on isotope traced acetate and/or palmitate, as an indirect method when the effect of fructose intake on DNL is assessed. Other studies followed the metabolic pathway of the examined sugar as such: when examining the metabolic fate of fructose, they also labelled the fructose with an isotope and then assessed to which extent the fructose contributed to the DNL. The reason why I have brought these two examples to your attention was the comparison of the findings of two such studies of fructose effect on DNL within as similar conditions as possible.
Dr Lustig presented the results of the study mentioned above (Hudgins et al) and said that in a fasting state fructose increased DNL by 5%. I have read that study myself. For achieving this result they fed their overweight to obese participants with 0.5 g of fructose per kg of body weight, corresponding to about 35 g of fructose for a 70 kg man. Heavier participants received proportionally more, of course. FIY: one 330 ml tin of Coca Cola has exactly 35 g of sugar, of which it is around half fructose, the other half is glucose.
In comparison, Chong et al (2007) also performed similar assessment on fasting participants with similar amount of people, similar profile of people but tracing also the labelled fructose as such, which the previous study did not. They fed the participants with 0.75 g of fructose per kg of body weight, which would be 52.5 g of fructose for a 70 kg man, therefore more than the previous study. What they have found was that fructose contributed only with 0.4% to the DNL. The remaining increased appearance of lipoproteins and triglycerides above this amount was explained as re-esterification of free fatty acids already present in the system (their concentration is always reported to decrease in this kind of studies) and by reduced lipids clearance into the fat cells because fructose does not trigger such high insulin response as glucose does. Now, do you not think that the 33% higher dose of fructose resulting in 12.5 times lower DNL outcome is interesting?
If you were shocked by the substantially increased effect of fructose in a fed state that Dr Lustig discussed later (those 25% increase), let me tell you what amounts of fructose/sugar it was based on. The team of Hudgins et al fed the participants with the following amounts of sugars:
- 35/35 g of fructose and glucose = 70g of sugar = 280 kcal for a 70 kg heavy man (0.5/0.5 g of fructose and glucose per kg of body weight)
- 70/70 g fructose and glucose = 140g of sugar per 70 kg man = 560 kcal. If the overweight or obese person weighing 100 kg (they did not include people of healthy weight in this second protocol) consumed the adequate proportion of sugar for their weight, it would be 200g of sugar = 800 kcal, because this group received 1 g of fructose and 1 g of glucose per body weight.
Any wonder why the results were so bad after all? Dr Lustig did not give you these numbers, he only shocked you with the results. 800 kcal of sugar given in the morning in a form of drink - do you think it is real for most of us? This amount would correspond to the consumption of mere 3% of the US population.
Putting this in numbers, Mariott et al 2010 reported that the 97th centile of the U.S. population consumed 35% of total energy intake in a form of sugar between years 2003-2006. I have calculated that this corresponds to about 14% of daily calories from fructose or 350 kcal for a person consuming 2500 kcal. In normal diet the approximate ratio of fructose to glucose is 0.8, not half on half as it is for sucrose. Even the top 10% of highest sugar consumers (90th centile) had 25% of total calories from sugars (10% from fructose), meaning that 90% of the U.S. population had lower sugar intake. And these amounts are consumed during the whole day in a liquid and solid form, not only in a drink within 15 minutes on an empty stomach as the study participants had to. This is the reality check. And remember that since year 2000 the sugar consumption has been decreasing.
If this study had examined the effect of sugar on DNL using the method of Chong et al (isotope labelled fructose as well) are we sure we would achieve the same results? You could already see how different the results were when compared the two similar settings with different methodology earlier.
Moreover, the high sugar effect (those 25% increase from the diagram) was found in people already overweight and obese, suggesting they already might have had their metabolism affected, not only by the high sugar feeding but also by their body weight. This was not told to you either.
Even more: while our bodies can vary enormously in their size due to the fat or muscle content, our liver size does not. And since you know that fructose is primarily metabolized by the liver, do you not think that such immense overload of sugar will show up with negative consequences after all? Whether the person is 70 kg or 120 kg, their liver capacity for metabolism of fructose does not vary that much, but the second person receives almost double the dose of the first one. Bear this in mind.
Something I found fascinating in the study (among few other things). Look at the bit I copied for you:
Have you noticed something? Fifteen mild to moderately overweight participants, of which the BMI ranged from 25.8 to 35.6. For those who are familiar with the grades of BMI this must be a bad dream. Those moderately overweight were not only obese, but beyond BMI 34.9 it is even Class II or severe obesity. Even the mean BMI was in the range of obesity (31.0). It took me a while to realize that their term 'overweight' was probably just a politically correct one since I remember a suggestion by someone to avoid using a word 'obese' because it could traumatize the obese person. Take it as a curiosity.
The truth is that no person had a weight below 80 kg in the second protocol, including women, so my reference man weighing 70 kg (once a true reference man for the Europe, 60 kg for women) was a different league. They intentionally tested these high doses of sugars on people with already altered energy metabolism. Moreover, while consuming the mega doses of sugars they were only allowed to remain passive, such as reading or watching the TV. In contrast, most of us move around daily. We take care of ourselves, we go to work, chase the bus, some walk the dog regularly or pick up children from school and do not normally consume such amounts of sugar anyway.
I have also found one part of the study discussion a bit manipulating. They compared the highest dose of sugars they gave to the obese participants (if weighing 100 kg -> 200g = 800 kcal) to the "amount of sugar in a commercial "supersize" soft drink consumed by 60 kg person". I tried to find out how big this supersize soft drink was but I could not find a specific figure for this term. Instead I calculated the amount of sugar the 60 kg person would consume in this study and it came out that it would be 120g of sugar or 480 kcal. That is far more than a 500 ml coke (53 g sugar, 210 kcal) or 16 ounces, above which the sale of soft drink volume was banned by the major of New York. While I agree that some extra large people may gulp down the 1300 ml of mega size soft drink in one sitting, I have doubts that an average 60 kg person would manage it.
To make you more amazed, I will continue with debunking the blame put on fructose when combined sugars were consumed. To make the diagram from above more legible, I will copy it from the actual study:
I want you to focus on the two bars in the middle, the F and F:G ones. From the description you see that the first group received 0.5 g of fructose per kg of body weight and the second one received the same plus the same amount of glucose. The figures above them (0.49 and 1.50) refer to the magnitude of change from the baseline. The rationale of added glucose to he fructose was to mimic the consumption of sugar or HFCS because we rarely consume fructose alone and also that the fructose has poor absorption rate on its own whereas it has the highest absorption rate when accompanied with the same amount of glucose.
How much more fructose do you think we absorb in the presence of glucose so it could increase the DNL three times as much? Do you guess it TRIPLED? WRONG. The same study reported the increased blood concentration of fructose in the presence of glucose from the mean 57 micrograms per mililitre (fructose alone) to only 72 micrograms (when fructose and glucose consumed together in the same amounts). The difference was not 2x57, but only mere 15. Even if one would argue that this was the systemic blood circulation level, whereas the liver captures most of the ingested fructose and may not release the extra in the proportional manner to the influx, this high jump is unlikely to be only due to the fructose alone. So the two-fold rise from the first test you can see above could have been largely accountable to the glucose, not fructose, since the dose of fructose has not changed and even at its highest absorption rate it is not 100% absorbed.
Yet the scientists still focused on fructose as the main villain, similarly as Dr Lustig has been all the time. And nobody should now have doubts about the flawed argument that when the glucose tolerance test alone caused the reduction of the blood lipids (the very first column marked OGTT) that the enormous rise of new palmitate after added glucose to fructose was only due to the fructose alone. The truth is that glucose not only enhances absorption of fructose, but the body also alters the metabolism of both sugars when they are in the system both at once. Unless the labelled fructose AND glucose are traced and only fructose detected in the newly formed palmitate while the glucose is not, we cannot be 100% confident about the findings. The metabolism of glucose and fructose is very much inter-connected and that is why these specific tracer studies came into place: to distinguish between the resultant effects of these two sugars.
The best for the last. Look at some information from the slide once more:
What Dr Lustig said was that at fasted state the DNL after fructose was only 5% higher from the baseline whereas in fed, obese and insulin resistant state it was 25%. WRONG again. All these results from the diagram above were on the same people under the same conditions, only the composition and amounts of sugar changed. So all of them received the fructose in fasting state, just in different amounts and the diagram presents the effect of different doses of fructose, not the difference between fasting and fed state. All of these participants were overweight or obese (Protocol 2) and all of them were tested for the response to those three different doses of either fructose or fructose and glucose in the mentioned amounts. Do you understand? That 5% was of 0.5 g per kg of body weight of fructose alone, those 25% was of 1 g of fructose and 1 g of glucose together, for the same participants under the same conditions, just spaced in time because it was the cross-over study. I have found nothing in the study saying that this amount was given before eating and this was after eating or that they would come up with these different responses before or after eating.
The Protocol 1 was slightly different. Instead of giving the whole volume of drink at once to be consumed within 15 minutes, it gave the participants regular doses of fructose sweetened drink, which made total 1.4 g per kg of body weight within 6 hours. To give you an idea, this amount would represent 98 g of fructose for a 70 kg man. 100kg person would receive 140g of fructose only. Too much of fructose, hardly representative of the general population. To match this amount of fructose with the glucose (as we normally consume them together), it would have to be mind blowing 280g of sugar = 1120 kcal consumed just within 6 hours and on empty stomach (if taking 50/50, not 0.8 fructose to glucose ratio). Luckily for the participants, the scientists did not bring this idea into practice. Note, this last protocol was not pictured in the diagrams above.
Instead, the link illustrates how the measured concentrations of lipids shot up during the first 8 hours in the Protocol 1 and then steadily dropped towards the end of the 24-hour monitoring period when the participants consumed food, therefore were in a fed state. And when you look at the bottom picture, you will see that at the beginning of the curve the concentration of palmitate started to rise even before time zero, which means even before fructose was given! This is just normal because after the 12-h fasting there was not enough insulin secreted, which was a signal to the fat cells to release fat for energy metabolism because even the glycogen stores were at their minimum after so many hours without a food intake. The participants simply switched to the keto-metabolism. Once you eat carbohydrates again, insulin rises and opens the gates for the glucose to enter into the cells (and for fat into the fat cells, too) and the fat burning mode is reduced again. And because it was a 'fructose challenge', the same applies what was discussed in case of Chong et al: lower level of insulin secreted after fructose then after glucose, hence lower clearance of lipids into the adipose tissue and although fructose is converted to fats in high doses, it may not be as high as we think, especially in the fasting state as both Protocols administered it.
Once again: nothing was mentioned in that study about fasted and fed state in terms of assessed biochemistry for the Protocol 2 and from the Protocol 1 you can see that the DNL steadily decreased after the participants no more received fructose and started receiving food (containing glucose and some fructose).
This slide presented by Dr Lustig was just a nonsense as were the claims related to it.
However, I am not saying that consumed fats along sugars (as per fed state) do not worsen the situation when it is overdone. I am saying that this video was full of crap.
Remember all these flaws and selectively presented information when you will listen to other Dr Lustig's horror conclusions about the sugar again. The studies like those two introduced here aim to assess the actual mechanism of the metabolism of the examined substrates. They usually do not aim to assess the real life situation of you or me. In fact, I consume very little added sugar, so I am not a good representative for this. Therefore I have no reason for trying to deny these results to curb the guilt or something. And I am definitely not sponsored by the sugar industry. These are the scientific findings that contradict each other to some extent, followed with my critical evaluation of how inappropriately these materials have been presented to you. And that was the point of this article, not to deny that fructose or sugar can be dangerous when consumed in excess.
No comments:
Post a Comment