The effect of dietary patterns on non-alcoholic fatty liver disease diagnosed by biopsy or magnetic resonance in adults: a systematic review of randomised controlled trials

Adhering to specific dietary patterns might hold promise as a lifestyle modification treatment of non-alcoholic fatty liver disease (NAFLD). The aim of this systematic review was to examine the effect of dietary patterns on changes in hepatic fat content, liver enzymes and metabolic syndrome components. We searched Pubmed, Embase, CINAHL and Web of Science for randomised controlled trials published in English until April 2020, comparing a specific dietary pattern with no treatment, usual care, or a different diet in adults with NAFLD. Studies were included if NAFLD had been diagnosed using biopsy, magnetic resonance imaging, or proton magnetic resonance spectroscopy. Data from three trials in adults with NAFLD but without diabetes (n = 128; mean age 49.9 ± 5.0 years, range 42-55 years) were included in the qualitative synthesis; across them, risk of bias was considered low, unclear and high for 33%, 38% and 29% of domains, respectively. There was moderate evidence that a low-carbohydrate, compared to a low-calorie diet (-27%, P = 0.008, one study, n = 18) and the Mediterranean, compared to a low-fat, high-carbohydrate diet (-4.4%, P = 0.030, one study, n = 12) result in greater reductions in hepatic fat content, but no such evidence was found for the Fatty Liver in Obesity dietary pattern (based on the principles of the Mediterranean diet), compared to the American Heart Association diet (-0.6%, P = 0.706, one study, n = 98). No between-group differences were reported for other outcomes across studies. A post hoc analysis, including two eligible studies assessing the effect of the Mediterranean, compared to a low-fat diet, irrespective of baseline presence of diabetes, showed strong evidence that the Mediterranean diet reduces hepatic fat content (-4.1%, 95% CI = -5.8 to -2.3, P < 0.001; I2 = 0%) and triglyceride concentrations (-16.9 mg/dL, 95% CI = -26.3 to -7.7, P < 0.001; I2 = 0%). Well-designed, adequately powered and rigorous randomised controlled trials are needed to provide robust evidence on the effect of these dietary patterns, but also other whole dietary approaches, on NAFLD progression.