Abstract: Dietary sugar consumption, in particular sugar-sweetened beverages and the monosaccharide fructose, has been linked to the incidence and severity of non-alcoholic fatty liver disease (NAFLD). Intervention studies in both animals and humans have shown large doses of fructose to be particularly lipogenic. While fructose does stimulate de novo lipogenesis (DNL), stable isotope tracer studies in humans demonstrate quantitatively that the lipogenic effect of fructose is not mediated exclusively by its provision of excess substrates for DNL. The deleterious metabolic effects of high fructose loads appear to be a consequence of altered transcriptional regulatory networks impacting intracellular macronutrient metabolism and altering signaling and inflammatory processes. Uric acid generated by fructose metabolism may also contribute to or exacerbate these effects. Here we review data from human and animal intervention and stable isotope tracer studies relevant to the role of dietary sugars on NAFLD development and progression, in the context of typical sugar consumption patterns and dietary recommendations worldwide. We conclude that the use of hypercaloric, supra-physiological doses in intervention trials has been a major confounding factor and whether or not dietary sugars, including fructose, at typically consumed population levels, effect hepatic lipogenesis and NAFLD pathogenesis in humans independently of excess energy remains unresolved.
Alex’s Notes: It was only a matter of time before another study on fructose was published. This time, however, things are different. The “evil” fructose has been bashed time and time again as an enemy of our health, and there is no denying that consuming bolus amounts of fructose in quantities unrealistic for most individuals alters hepatic insulin sensitivity, increases lipogenesis, and increases liver fat accumulation. But let’s be realistic, and that’s why things are different.
Fructose consumption has risen over the past 40 years, but so has total carbohydrate and caloric intake. Actually, between 1999 and 2008 the consumption of added sugars in the U.S. has decreased by about 33%. Ignoring the studies performed on rodents, conclusions about fructose are equivocal and influenced largely by the amount of fructose and total energy intake of the person. For instance, a very recent meta-analysis observed no effect on post-prandial triglycerides when fructose was swapped with another carbohydrate under isocaloric conditions. The trials that failed to find an effect had an average fructose intake of 20% the total energy intake, which translates to about 100g of pure fructose for a 2000 kcal diet. Conversely, when fructose was provided under hypercaloric conditions, providing 25% of energy intake on top of usual intake, there was a significant effect. It is hard to blame fructose for the observed effects when confounded with the overfeeding condition.
Unfortunately, many persons who frown upon fructose rely on short-term intervention trials (<7 days). One such intervention fed healthy young males a weight-maintenance control diet, or a diet that provided 3.5g/kg/day of fructose or glucose and a 35% caloric surplus for seven days. They found that compared to the control, both the fructose and glucose interventions resulted in a significantly greater increase in liver fat that were not significantly different from one another. How can we blame fructose when glucose does the same thing? Conversely, another week-long study compared overfeeding 3g/kg/day of fructose or glucose with a weight-maintenance control diet and found that the fructose intervention led to liver fat accumulation almost double that of the glucose intervention. The common theme is the excess calories, and whether the increased liver fat is a result of fructose metabolism itself or simply increased energy availability is up for debate.
Thus, we should look to longer term clinical trials. When fructose was added to the diet of healthy males over four weeks, representing an 18% caloric surplus, there was no change in liver fat accumulation despite increased concentrations of fasting triglycerides and (surprisingly) leptin. Similarly, supplying 150g of fructose or glucose as a surplus of calories had no effect on liver fat, visceral fat, subcutaneous fat, or intramuscular fat in healthy males over four weeks. Yet, in overweight and obese men, providing an additional 25% of caloric needs as fructose or glucose has shown to increase liver fat after two weeks. The punch-line is that there was no difference between the sugars. So again, is it the fructose or the excess calories?
Four weeks is still pretty short in my opinion. Fortunately, a ten week long intervention confirms all the above with healthy adults showing no difference in liver fat after consuming high-fructose corn syrup as 8, 18, or 30% of energy intake when compared to sucrose or baseline levels. Interestingly, some researchers have taken the opposite approach to the problem and looked at how the reduction of fructose has affected liver fat. Reducing fructose intake by 50% for six months resulted in reduced liver fat in persons with NAFLD. Of course, the persons also had significantly reduced caloric intake and weight loss after the intervention. So was it the fructose or the caloric deficit and weight loss?
Lustig must be blushing by now.
Yah, probably. But we aren’t done yet. What about all the talk of this de novo lipogenesis? This is the term used for the process whereby fatty acids are created from non-lipid precursors such as fructose, glucose, or amino acids. Using isotopic glucose or fructose, which lets us “follow” it through the body and see where it ends up, it was demonstrated that consuming 0.75g/kg bodyweight of fructose or glucose didn’t live up to Lustig’s expectations. Specifically, only 0.05% and 0.15% of fructose were converted to fatty acids and glycerol, respectively, while values after glucose consumption were close to 0%. These results have been confirmed by a systemic review of isotopic tracer studies that concluded the immediate metabolic fate of ingested fructose was not into triglycerides (<1%), but towards oxidation (45%) and conversion to glucose (41%).
Blast from the past
This isn’t the first post to set the record straight on fructose. I have previously written about the metabolic fat of fructose when ingested with and without glucose, as well as the normal roles for dietary fructose in carbohydrate metabolism. There was also a meta-analysis that provocatively suggested that when in caloric surplus, glucose and fructose are on equal grounds. An observational study of older Finnish adults actually found an inverse association between NAFLD and fructose intake, and moderate amounts of fructose- or glucose-sweetened beverages (65g total sugar daily) did not affect metabolic parameters in adolescents.
Finally, if you are still worried about fructose for some weird dogmatic reason, then go for a walk. One of the first studies I wrote about demonstrated that merely walking about 13,000 steps per day significantly attenuated the adverse effects of adding 75 grams of fructose to the diet in the form of 40 ounces of soda per day.