Effects of Dietary Fat and Saturated Fat Content on Liver Fat and Markers of Oxidative Stress in Overweight/Obese Men and Women under Weight-Stable Conditions

Abstract: Dietary fat and oxidative stress are hypothesized to contribute to non-alcoholic fatty liver disease and progression to steatohepatitis. To determine the effects of dietary fat content on hepatic triglyceride, body fat distribution and markers of inflammation and oxidative stress, overweight/obese subjects with normal glucose tolerance consumed a control diet (CONT: 35% fat/12% saturated fat/47% carbohydrate) for ten days, followed by four weeks on a low fat (LFD (n = 10): 20% fat/8% saturated fat/62% carbohydrate) or high fat diet (HFD (n = 10): 55% fat/25% saturated fat/27% carbohydrate). Hepatic triglyceride content was quantified by MRS and abdominal fat distribution by MRI. Fasting biomarkers of inflammation (plasma hsCRP, IL-6, IL-12, TNFα, IFN-γ) and oxidative stress (urinary F2-α isoprostanes) were measured. Body weight remained stable. Compared to the CONT, hepatic triglyceride decreased on the LFD (mean (95% CI): change −2.13% (−3.74%, −0.52%)), but did not change on the HFD and there was no significant difference between the LFD and HFD. Intra-abdominal fat did not change significantly on either diet, but subcutaneous abdominal fat increased on the HFD. There were no significant changes in fasting metabolic markers, inflammatory markers and urinary F2-α isoprostanes. We conclude that in otherwise healthy overweight/obese adults under weight-neutral conditions, a diet low in fat and saturated fat has modest effects to decrease liver fat and may be beneficial. On the other hand, a diet very high in fat and saturated fat had no effect on hepatic triglyceride or markers of metabolism, inflammation and oxidative stress.

Free Full-text

Alex’s Notes: Liver fat is no fun. The liver is such an essential organ for optimal health that it even possesses a limited ability to regenerate when injured. The same cannot be said for less overlooked organs such as the heart and brain. Nonalcoholic fatty liver disease (NAFLD) can be the result of many things, probably one of the best known being an unrealistic amount of pure fructose intake. But fat is in the name, so a valid question is what role does dietary fat play?

A previous rodent study found that a high fat/high saturated fat diet increases liver fat and injury, and human trials demonstrate that two to three weeks of a low-fat/low-saturated fat diet reduce liver fat relative to a high-fat diet in overweight non-diabetic subjects. In order to better understand the relationship between dietary fat and NAFLD, the current study underwent a randomized controlled dietary feeding with the following layout.

Control diet

Intervention diet


Control diet

Intervention diet

10 days

4 weeks

6 weeks

10 days

4 weeks

The subjects included ten men and three women (n=13) with an average age of 36 years and BMI of 33.6kg/m2. They had to be healthy, nonsmokers, with normal liver and renal function, and drink less than two alcoholic drinks per day by self-report. Additionally, they did not have any food allergies or intolerances, medications affecting inflammation, insulin sensitivity, or liver fat, and were normoglycemic (average fasting glucose <100 mg/dL & 2-hour oral glucose tolerance test <140 mg/dL). Despite all this, NAFLD (liver fat >5%) was present in 7/13 subjects.

As mentioned earlier, this study was controlled, so that the subjects did not have control, per se, over their food intake. All food was prepared in the University kitchen and the subjects picked up their food and were weighed twice per week. Compliance was determined by having subjects record all food consumed each day with a checklist returned to the nutritionist – which of course does guarantee the subjects actually ate what they checked off.

Caloric needs were estimated using the average of the Mifflin-St. Jeor and Harris-Benedict equations, adjusted for physical activity, and food intake was adjusted weekly to achieve weight stability. This was a eucaloric intervention. Subjects were instructed to maintain regular physical activity and to eat all of the food provided, not to eat any non-study food, and to report any deviations from the diet. Physical activity was assessed at the end of each control diet and each intervention diet by use of the short form seven-day international physical activity questionnaire (IPAQ). Body fat and its distribution were determined with DXA scans and abdominal MRI, while liver fat was quantified with MRS.

The intervention diets were a low-fat diet (LFD; 20% fat, 8% saturated fat) or high-fat diet (HFD; 55% fat, 25% saturated fat), with the control diet (35% energy from fat, 12% saturated fat) being intermediate between the two. Of course planned diets don’t always turn out as planned, and the study diets’ actual compositions appear in the table below. For amusement, they are compared to the baseline diets of the study participants estimated by a 3-day food log. The menus were designed by a research nutritionist and major sources of fats in all three diets included butter and high oleic safflower oil. Vegetable content was matched between groups, and because fructose was limited on the HFD due to the low carbohydrate content, fructose was limited in all diets to <15 g/day/1000kcal.






Daily kcal





Fat (% energy intake)





Saturated fat

























Total fiber (g/day)





Fructose (g/day)





First and foremost, I have to address the huge discrepancy of the study diets compared to the baseline diet. All participants except two reported consuming all the food provided, there were no significant changes in physical activity, and weight remained stable throughout the interventions. Yet the participants reported an energy intake 28% lower than their prescribed diets – strongly suggesting that they underreported their intake and providing yet more evidence that overweight and obese persons don’t know how much they are truly eating.

Low-fat, High-fat, liver-fat, all-fat

The LFD resulted in a statistically significant reduction in liver fat of 2.13% with no changes in any other anthropometric, metabolic, or inflammatory markers. Similarly, the HFD demonstrated no significant changes in liver fat or any other variable, except that subcutaneous fat increased by about 6% despite no change in weight. Actually, upon inspection of data table it appears that the authors forgot to mention that the HFD also significantly reduced gamma-glutamyl transferase (GGT) by about 20% compared to the LFD. There was also some interesting individual responses to the diets. For example, looking a visceral fat, there was no change for all but one person who shot up by about 50%. Same deal for one person on the LFD when looking at subcutaneous fat – it increased by about 10%. Finally, one subject on the LFD had a dramatic decrease in liver fat from 13.9% to 1.3%, and removal of this subject decreased the mean change from the aforementioned 2.13% to 1.18%, but it remained significant.

What did we learn?

In a nutshell, on a weight-stable diet, a high-fat and high-saturated fat diet does not increase liver fat, although it may increase subcutaneous fat. Also, neither diet impacted markers of inflammation or oxidative stress, although it could be that 4 weeks is too short of a timeframe to detect changes. Regardless, as the authors point out,

“The fact that hepatic triglyceride decreased on the LFD but did not change on the HFD suggests that either (a) the relationship between dietary fat/SFA intake and hepatic triglyceride is not linear and/or (b) factors other than total fat and saturated fat content of the HFD contribute to hepatic triglyceride content. Differences in fat and carbohydrate composition, age and underlying glucose metabolism status may have contributed to the discrepant findings between studies.”

We must also acknowledge that the differences in fructose content of the diets may have influenced the results, and it is possible that the lower fructose in the HFD that might have counterbalanced negative effects of the high fat content on liver fat. The small sample size is another limitation that may have limited statistical power to detect differences, and of course these results cannot be extended to other populations such as type-2 diabetics. 


Network Affiliates

Quick Links I

Our Location

2908 Brownsboro Rd
Suite 103
Louisville, KY 40206
(502) 690-2200

SHR Newsletter

Subscribe to our FREE newsletter
to receive the latest updates in your inbox!
SHR Newsletter
Internet Radio