The role of statins in the primary and secondary prevention of CVD has been well established. Patients with NAFLD are believed to have increased cardiovascular risk  and are logical candidates for their long-term use. In accordance with their higher cardiovascular risk, the use of statins should be encouraged in patients with NAFLD but their use has remained controversial in such patients, particularly in the setting of elevated liver enzymes . Recent practice guidelines on statin use in patients with NAFLD have clearly established that they are overall safe and that they do not carry a higher risk of liver toxicity [79,80].
In addition to their apparent safety in NAFLD, several small studies have suggested that statins may improve histology in NASH by means of their anti-inflammatory, antioxidant, or other pleiotropic properties . Several studies reported that statins may decrease plasma aminotransferase concentration or hepatic steatosis by ultrasound [5,81,82]. However, it should be emphasized that these studies were overall of inadequate quality due to small sample size, uncontrolled design, short duration and/or lack of gold-standard endpoints, such as liver MRS or histology. In many of these trials effects were difficult to separate from weight loss, dietary changes or other lifestyle modifications during the study. In studies in which a liver biopsy was performed before and after statin treatment, Kimura et al.  (n=43) and Rallidis et al.  (n=5) found that liver inflammation improved, in contrast to reports from Georgescu et al.  (n=10) and Ekstedt et al.  (n=17). Beyond these discrepancies, there is agreement that hepatocyte ballooning and fibrosis do not improve with statin therapy. In the only randomized controlled trial in patients with biopsy-proven NASH, Nelson et al.  reported a nonsignificant trend towards a reduction in plasma aminotransferase concentration but no changes in liver histology. However, the study was rather small (n=16) and of relatively short duration (12 months).
Taken together, the evidence showing a beneficial effect of statins in NAFLD or NASH is weak and awaits well-designed long-term studies. Their use should be promoted to prevent cardiovascular disease, but not as a means for histologic improvement.
Recent large randomized controlled fibrate trials in patients with T2DM have failed to meet the primary outcome (reduction of a major cardiovascular or coronary event) when combined with statins [88,89]. However, post-hoc analyses have shown that patients with high plasma triglycerides (>200 mg dL−1) and low HDL-C concentration may benefit from combined statin and fibrate therapy . Patients with NAFLD are characterized by this type of dyslipidemia, and therefore, are good candidates for combination therapy.
As discussed earlier for statins, there has been substantial interest on the role of fibrates in NAFLD. Fibrates have many biologic effects on hepatic lipid metabolism (most mediated by PPAR-γ effects) that appear to support their potential role in NAFLD. Fibrates have been reported to improve plasma aminotransferase levels or decrease hepatic steatosis, but overall results have been inconsistent and histology unchanged . In a study by Athyros et al. , atorvastatin plus fenofibrate in patients with NAFLD had no effect beyond that of atorvastatin alone on plasma aminotransferase concentration or liver fat measured by ultrasound. In a well-designed RCT, two months of fenofibrate therapy had no effect on liver triglyceride accumulation when assessed by MRS .
In summary, fibrates are considered safe when needed for the prevention of cardiovascular disease in patients with NAFLD, but according to available evidence, they do not improve liver histology.
Ezetimibe has been evaluated in a few small studies in patients withNAFLD and atherogenic dyslipidemia. It has been reported to improve liver enzymes, inflammatory markers, and some histologic parameters [93,94]. In a randomized, single-blinded trial ezetimibe decreased liver fat by MRS but histologic outcomes were not examined . However, these studies were small, open-label, or uncontrolled. Therefore, it is difficult to interpret whether these changes were the effect of ezetimibe therapy or just the consequence of weight reduction or lifestyle modification.
Pentoxifylline is a nonselective phosphodiesterase inhibitor that reduces TNF-α and may affect other inflammatory pathways. Due to these anti-inflammatory effects, it has been considered as a potential treatment for NASH . Although several studies have assessed the role of this drug in patients with NASH, just a couple of these were RCTs . In one of such studies, Zein et al.  reported benefit on steatosis, inflammation, and fibrosis in 55 patients (46 with biopsy at the end of the study) randomized to pentoxifylline or placebo for one year. However, at baseline the two groups were not completely well matched, with patients in the pentoxifylline group having higher levels of AST/ALT and a trend towards worse histology. In another 12-month trial in 30 patients with NASH , pentoxifylline did not improve histology compared to placebo but decreased steatosis and ballooning against baseline. Once again, patients randomized to pentoxifylline had a worse baseline liver histology. Others studies assessing the role of this drug in NASH were either open-label, had a small sample size, or did not include histology or liver fat by MRS as endpoints . Although the results with pentoxifylline are encouraging (especially the results of Zein et al. showing an improvement in fibrosis), much more work remains to be done to clearly understand the true role of this drug in NAFLD.
Omega-3 polyunsaturated fatty acids
Omega-3 polyunsaturated fatty acids (PUFAs) have been proposed as a potential therapy for liver steatosis . They have been shown to activate PPARα receptors, which upregulate several genes involved in fatty acid oxidation and at the same time downregulate pro-inflammatory genes . Moreover, PUFAs activate PPARγ, which also increases fat oxidation. Other important effects of PUFAs include reduction of endogenous lipid production and inhibition of hepatic glycolysis. Animal models have shown not only that low-PUFA diets promote steatosis and insulin resistance, but also that supplementing diet with PUFAs can prevent and even reverse hepatic steatosis . Although several studies have examined the role of these compounds in humans, most of these studies have significant study design limitations [98,99], such as being open-label and lack of proper controls, or use as the primary outcome improvement of liver enzymes or liver fat by ultrasound. In the only RCT with PUFAs, Cussons et al.  reported a small, but statistically significant, reduction in hepatic steatosis by MRS. However, as discussed earlier, the meaning from a clinical standpoint of a small reduction in liver triglyceride content (from 10.2% to 8.4%) remains uncertain. Also, just one study assessed the effect of PUFAs on liver histology, but was small (n=23) and uncontrolled . Therefore, based on the available evidence it is difficult to make a final conclusion in regard to the role of PUFAs in NAFLD. Larger, well-designed RCTs are needed to understand their true effects in this disease.
Vitamin E is an antioxidant believed to reduce hepatocyte oxidative stress in patients with NASH [102–105]. Several small studies of short duration in patients with NASH have reported inconsistent results [102–104,106–108]. In nondiabetic patients with biopsy-proven NASH  vitamin E led to a 43% response in the primary histologic endpoint compared to 19% in the placebo group (p=0.001). The primary histologic endpoint was an improvement in ≥2 grades in the NAS with at least one point improvement in hepatocellular ballooning and one point in either the steatosis or lobular inflammation score, with no worsening of fibrosis.
In summary, vitamin E may be beneficial in patients with NASH without T2DM because it appears safe and is relatively inexpensive, but its long-term efficacy (beyond 2 years) has not been established.Moreover, its use in patients with diabetes and NASH has never been studied and remains to be established.