The overall aim of this doctoral thesis was to investigate the role of circulating, liver and dietary fatty acids in non-alcoholic fatty liver disease (NAFLD). For circulating and liver fatty acids, special emphasis was given to fatty acids reflecting diet.

In paper I, circulating cholesteryl ester (CE) linoleic acid (18:2n-6), which is considered a good biomarker of dietary intake of 18:2n-6, was cross-sectionally inversely associated with liver fat in n=308 50-year old men and women. Several fatty acids reflecting both exogenous intake and endogenous metabolism were associated with liver fat, basal fat oxidation and resting energy expenditure (REE). No association between fatty acids and liver fat, except for docosahexaenoic acid (22:6n-3) and liver fat, were attenuated after adjusting for REE. 

In paper II, phospholipid (PL) 22:6n-3 in liver tissue, a potential biomarker of dietary intake of 22:6n-3, was cross-sectionally inversely associated with liver fibrosis in n=60 men and women with biopsy-verified NAFLD. This finding was not replicated in plasma. Several other fatty acids reflecting both exogenous intake and endogenous metabolism were associated with fibrosis. Pooled saturated fatty acids (SFA) were generally positively associated whereas monounsaturated (MUFA) and polyunsaturated fatty acids (PUFA) were inversely associated with fibrosis. 

In paper III, no clear (i.e. imprecise) associations were observed for any of the nutrient or food substitutions with incidence NAFLD cirrhosis or hepatocellular carcinoma (HCC), over a median follow-up of 24 years in n>77 000 middle-aged to elderly men and women. 

In paper IV, a 12-month randomized controlled trial (RCT) was conducted to investigate the effects of a low-carbohydrate high PUFA (LCPUFA) diet and a healthy Nordic diet (HND) on liver fat in men and women with type 2 diabetes (T2D) or prediabetes. The comparator diet (usual care (UC)) aligned with the Nordic Nutrition Recommendations. Liver fat decreased more in the LCPUFA diet and the HND versus UC. No difference in liver fat was observed between LCPUFA and HND. The LCPUFA diet and the HND improved several other cardiometabolic markers compared to UC, with more favorable improvements in the HND group.

In conclusion, findings from this thesis suggest that higher intakes of dietary unsaturated fatty acids (in particular PUFA) and lower intakes of SFA may be of importance for the prevention and treatment of NAFLD (at least for liver fat and fibrosis). Findings from this thesis also suggest that fatty acids reflecting both diet and endogenous metabolism may play a role in NAFLD.