Controlled human studies examine how fructose affects liver fat and insulin sensitivity, including when calories and body weight are held stable.
There is a tendency in nutrition to reduce everything to calories. If two foods contain the same amount of energy, the assumption is that their metabolic effects should ultimately be the same.
That is not how metabolism works.
Fructose is a particularly good example. Human intervention studies show that fructose can change liver metabolism even when calories are held constant and people do not gain weight. It can increase hepatic de novo lipogenesis, increase liver fat, and impair hepatic insulin sensitivity without requiring caloric excess.
One of the cleanest demonstrations came from Schwarz and colleagues.
Eight healthy men were admitted to an inpatient clinical research center for 18 days. Every meal was provided. Nonstudy food was prohibited. Their weight was measured daily.
For nine days they consumed one diet, and for nine days they consumed the other. The diets contained the same calories and the same proportions of carbohydrate, fat, and protein. The major difference was that complex carbohydrate was replaced with fructose.
Their weight remained stable.
Their liver metabolism did not.
When they ate the high-fructose diet, hepatic de novo lipogenesis increased from 11.0% to 18.6%. Liver fat was a median 137% higher. Every participant had more liver fat after the fructose diet.
Fructose also impaired insulin’s ability to suppress glucose production by the liver, while peripheral insulin-mediated glucose disposal did not significantly change.
Nothing about this required weight gain. The calories were matched. The macronutrients were matched. The subjects were living in a research center eating the food investigators gave them.
The carbohydrate changed, and their liver metabolism changed with it.
A larger randomized trial published in the Journal of Hepatology found something similar.
Geidl-Flueck and colleagues randomized 94 healthy men to beverages containing fructose, sucrose, glucose, or no sugar-sweetened beverages for seven weeks.
Despite consuming the sweetened beverages, participants compensated elsewhere in their diets, and total energy intake did not significantly increase.
But what happened in the liver depended on which sugar they drank.
Basal hepatic de novo lipogenesis was 9.1% in the control group and 11.0% in the glucose group. With fructose it was 19.7%. With sucrose it was 20.8%.
Fructose and sucrose approximately doubled hepatic DNL. Glucose did not.
That is difficult to reconcile with the idea that the liver simply sees carbohydrate calories as carbohydrate calories. Glucose and fructose provide the same amount of energy per gram, but they do not follow the same metabolic pathway and did not produce the same hepatic response.
And this isn’t just one or two trials.
Ter Horst and colleagues pooled 29 intervention studies involving 1,005 participants. When fructose was substituted for other carbohydrates without adding calories, fructose significantly worsened hepatic insulin resistance.
Interestingly, the effect was not seen in peripheral glucose disposal. Again, the signal was particularly apparent in the liver.
So we have controlled intervention data showing that changing the type of carbohydrate, without necessarily changing the calories, changes hepatic insulin action.
There is a biological reason for this.
After we eat fructose, a substantial amount is metabolized by the intestine and liver. Isotope-tracing experiments have shown that the small intestine can metabolize much of a small fructose load before it ever reaches the liver.
But that capacity is limited.
Jang and colleagues showed experimentally that at lower exposures the small intestine cleared approximately 90% of dietary fructose. As the fructose load increased, intestinal clearance became overwhelmed and more fructose escaped intestinal metabolism, exposing the liver and colonic microbiota to more fructose.
Follow-up work made the point even more interesting.
When researchers altered intestinal fructose metabolism, they could change how much fructose reached the liver and how much hepatic lipogenesis and steatosis developed. The speed at which fructose was delivered mattered as well. A large rapid exposure was metabolically different from spreading the same amount over time.
In other words, the body doesn’t just count the calories and move on. It has actual metabolic machinery for handling fructose, and that machinery has a finite capacity.
There is also an interesting human trial involving fruit.
Researchers in Iran randomized 80 people with nonalcoholic fatty liver disease to eat at least four servings of fruit per day or fewer than two servings per day for six months.
The group eating more fruit finished the study with worse hepatic steatosis, insulin resistance, fasting glucose, ALT, AST, ALP, GGT, and several lipid measurements.
HOMA-IR was 7.36 in the higher-fruit group compared with 2.66 in the lower-fruit group. ALT was 89.1 versus 32.0 U/L, and AST was 74.5 versus 24.0 U/L.
The investigators reported that adjustment for BMI and calorie intake did not change their findings.
That doesn’t mean fruit and purified fructose are the same thing. They aren’t. But it does challenge the assumption that the effects of fructose-containing foods can always be dismissed as simply a matter of excess calories.
There is an even broader lesson here.
A 2024 Nature study examined fructose metabolism in cancer models. Dietary fructose increased tumor growth without producing weight gain or insulin resistance. Interestingly, the cancer cells weren’t particularly good at metabolizing fructose themselves.
The liver was doing the work.
The liver metabolized fructose and released lipid molecules, including lysophosphatidylcholines, into the circulation. Those lipids could then be used by tumor cells for growth. When researchers interfered with hepatic fructose metabolism, the effect was reduced.
This was an animal study and isn’t evidence that eating fructose causes cancer in humans. But it demonstrates a simpler point: a nutrient can have biological effects through its metabolism that have nothing to do with whether it caused weight gain.
Calories obviously matter. Chronic energy excess matters. And when researchers deliberately overfeed people fructose, the metabolic consequences can become even more pronounced.
But that isn’t the question.
The question is whether fructose only becomes metabolically relevant because it supplies excess calories and causes weight gain.
The controlled human data tell us that it doesn’t.
In an inpatient feeding experiment, investigators kept calories constant and kept body weight stable, changed complex carbohydrate to fructose, and watched liver fat, hepatic DNL, and hepatic insulin resistance change.
In another randomized trial, fructose approximately doubled hepatic DNL without a significant increase in total energy intake, while an equivalent glucose intervention did not.
Across controlled intervention studies, replacing other carbohydrates with fructose without adding calories worsened hepatic insulin sensitivity.
Calories tell us how much energy a molecule contains.
They don’t tell us what the body does with that molecule.
References
Schwarz JM, et al. Effect of a High-Fructose Weight-Maintaining Diet on Lipogenesis and Liver Fat. Journal of Clinical Endocrinology & Metabolism. 2015;100:2434-2442. PMID: 25825943.
Geidl-Flueck B, et al. Fructose- and sucrose- but not glucose-sweetened beverages promote hepatic de novo lipogenesis: A randomized controlled trial. Journal of Hepatology. 2021;75:46-54. PMID: 33684506.
Ter Horst KW, et al. Effect of fructose consumption on insulin sensitivity in nondiabetic subjects: a systematic review and meta-analysis of diet-intervention trials. American Journal of Clinical Nutrition. 2016. PMID: 27935520.
Alami F, Alizadeh M, Shateri K. The effect of a fruit-rich diet on liver biomarkers, insulin resistance, and lipid profile in patients with non-alcoholic fatty liver disease: a randomized clinical trial. Scandinavian Journal of Gastroenterology. 2022;57:1238-1249. PMID: 35710164.
Jang C, et al. The Small Intestine Converts Dietary Fructose into Glucose and Organic Acids. Cell Metabolism. 2018. PMID: 29414685.
Jang C, et al. The small intestine shields the liver from fructose-induced steatosis. Nature Metabolism. 2020. PMID: 32694791.
Fowle-Grider R, et al. Dietary fructose enhances tumour growth indirectly via interorgan lipid transfer. Nature. 2024. PMID: 39633044.
PMID
25825943
33684506
27935520
35710164
29414685
32694791
39633044
