Vitamin E, as alpha-tocopherol has been widely investigated in liver diseases earning the status of most promising treatment at a daily dosage of 533.3 mg (or 800 IU) in non-diabetics with biopsy-proven NASH . The present study is the first to consider the hepatoprotective effects of tocotrienols. Their effectiveness in ameliorating NAFLD could in part be explained by the fact that the tocotrienols have been shown to distribute preferentially to the liver [6, 7] where alpha-tocotrienol has been found to exert a 40-60 times more potent antioxidant activity in microsomal lipid peroxidation . Oxidative stress has been recognised amongst the contributing factors in the progression of steatosis [2, 17] with involvement of nuclear factor-κΒ (NF-κΒ) activation and increased release of proinflammatory cytokines, such as tumour necrosis factor alpha (TNFα) and the interleukin-6 (IL-6) representing the second insult in the “two hits” model [18, 19]. Tocotrienols were reported to inhibit TNFα and to “prevent reactive oxygen species-induced NF-κΒ expression” to a higher extent compared to alpha-tocopherol in streptozotocin-treated diabetic rats . Recently, biological mechanisms for the activity of tocotrienols against hepatic triglycerides accumulation have been reported. It has been shown that gamma-tocotrienol, but not alpha-tocopherol, was able to attenuate triglycerides accumulation by regulating fatty acid syntase and carnitine palmitoyltransferase leading to a reduction of hepatic inflammation and endoplasmic reticulum stress [21, 22]. A 16-week study on rats fed high fat and high carbohydrate diet, showed a reduction of the hepatic fatty changes with “normalised portal inflammatory cells” . The authors suggested that interference with the inflammatory processes to be one of the possible mechanisms for hepatoprotection .
Clinical trials employing the same mixed tocotrienols preparation used in the present study has been shown to be effective in improving arterial compliance  and in reducing Model for End-stage Liver Disease (MELD) score . Moreover, the lipid lowering activity of tocotrienols has been previously reported  and a modest reduction of lipid profile was recorded also in the present study. Blood lipids biochemistry results showed that tocotrienols significantly reduced TC, LDL and TG compared to baseline, but not compared to placebo. However, the values for TC and LDL were still above the normality range.
In our study, baseline levels of alpha-tocopherol did not differ between the two groups. The plasma alpha-tocopherol concentrations of subjects in the tocotrienols treated group at conclusion of the study were found to be slightly higher than baseline, although not statistically significant. The presence of 61.1 mg of alpha-tocopherol in the mixed tocotrienols preparation administered may have contributed to slight levels increase of alpha-tocopherol. Nevertheless, mean plasma concentrations of alpha-tocopherol in the participating subjects were still within the normal range of 5.5-18.0 μg/ml reported by the National Reference Laboratory  throughout the study, irrespective of the treatment received. The presence of alpha-tocopherol in the mixed tocotrienols capsules was unlikely to have any influence on the outcomes of the present study since the amount present in the capsules was far below those used in earlier studies with negative outcomes [26–29].
We acknowledge that histology examination is the gold standard for evaluation of liver diseases and the use of USG might represent an intrinsic limitation. However, USG examination has been shown to have adequate sensitivity and specificity, in particular as front line evaluation of non-advanced fatty liver.
Diagnostic reports showed that USG has specificity range of 60-100% and sensitivity range of 80-100% [30–32]. Obesity with BMI above 40 might affect USG evaluation, reducing specificity and sensitivity when fat infiltration is below 33% . In the present study, only one morbid obese was enrolled whose baseline diagnosis of NAFLD was unchanged at conclusion.
A gross advantage of USG is that the liver is seen in full, while a needle biopsy only samples about one 50,000th of the organ and area sampling error could be an issue in non-advanced cases [31, 34]. The choice of investigating hepatoprotective activity of a nutritional supplement, such as mixed tocotrienols, in adults drawn from a population-based study  with the use of non-invasive USG has to be seen in this light, being the majority of patients classified as mild steatosis with normal to slightly elevated liver aminotransferases, thus a non-advanced stage of NAFLD.
A possible weakness of our study may lie in the use of USG versus other non-ionising imaging methodologies such as proton magnetic resonance spectroscopy (1H MRS). 1H MRS is a powerful and objective tool for assessing fat fraction in the liver, even though it lacks of “information about regional fat distribution” .
A recent study  aimed at correlating 1H MRS with the echodiscrepancy ratio between kidney and liver as indicator of steatosis, on subjects with essentially similar anthropometric and blood biochemical values to our present study, concluded that the hepatorenal echodiscrepancy ratio evaluated with USG is in accord with the quantitative data obtained from 1H MRS for hepatic lipid infiltration in the range 1%-30% for comparable severity of steatosis, albeit what Mancini  considered as moderate in their study, is categorised as mild in ours.
We acknowledge that our patients were a low-risk cohort. However, patients with similar characteristics as those considered in our study, including normal liver transaminases levels, have been shown to still be at risk with respect to progression of NAFLD [37, 38].
In view of the not fully elucidated reasons for steatosis to progress to more advanced stages of liver disease, it might be beneficial to manage the condition during the early stages, such as at the steatotic stage. Hence, in view of our findings, mixed tocotrienols might be considered for the treatment of steatosis.
In summary, the present study showed that patients assigned to tocotrienols at the current dosage, have significantly benefited in terms of normalisation of the hepatic echogenic response in NAFLD compared to placebo during the 1-year treatment. We acknowledge that this is a pilot study and therefore larger cohort studies should be warranted to further confirm the positive results of tocotrienols in NAFLD of the present study. Furthermore, at the current combined dosage of 400 mg/d used in the present study, the mixed tocotrienols were well tolerated and no adverse reactions were reported.