Combined dietary intervention and exercise
The available information on the role of combined lifestyle interventions (hypocaloric diets plus exercise) in patients with NAFLD is much more extensive, with several randomized, controlled trials showing significant benefit with this comprehensive approach [28,32,33,38,56–58]. Again, the overall reduction in liver fat reported in most of these trials has been strongly correlated with the amount of weight loss. Table 20.4 summarizes the studies that have examined the effect of lifestyle intervention on liver fat by MRS. As expected, those studies that achieved a lesser degree of weight loss [28,31] had a smaller reduction in liver fat content. Of note, studies that have compared hypocaloric diet versus hypocaloric diet plus exercise have failed to report a significant difference in the improvement of liver fat content with the combined strategy, but in both studies weight loss was about the same with both interventions [28,56]. Keating et al.  reached a similar conclusion in a meta-analysis of all available studies, in which exercise plus diet had no significant pooled effect size (ES) when compared to diet alone (ES: −0.05; 95% CI −0.38–0.27, p=0.76).
An unresolved issue about many of these trials is whether small absolute changes in liver fat (even if of statistical significance) truly have clinical relevance. For instance, Lazo et al.  and Larson-Meyer et al.  reported apparently important reductions in liver fat content by MRS of 51% and ∼35%, respectively, but that reflected only a slight absolute change in hepatic steatosis (4.2% to 1.9% and 1.5% to 0.6%, respectively). It remains to be established if small absolute changes of intrahepatic triglycerides within the normal range (note: NAFLD by
MRS is defined as being >5.5%) translate into a true meaningful clinical benefit, or liver histologic improvement, in patients with NAFLD/NASH. A few studies have performed liver biopsies before and after a weight loss intervention. As discussed earlier, histologic outcomes could be predicted by the magnitude of weight loss.
In a 12-month uncontrolled study  using a standardized nutritional supplement combined with exercise counseling, failure to reduce weight was associated with no improvement in liver histology in the 15 (out of 23 enrolled) patients who completed the study. Consistent with the above, Eckard et al.  examined three different lifestyle interventions (low-fat diet or low-carbohydrate with moderate exercise or moderate exercise alone) compared to a standard care control group. Small study groups (11 to 12 per arm) and a high dropout rate (27%) limited the study conclusions. As in the previous study, failure to lose weight with diet and exercise (∼1% weight loss) translated into no major histologic improvement compared to the control group. In contrast, Promrat et al.  found that 48 weeks of lifestyle intervention (diet plus moderate exercise), that significantly reduced weight when compared to controls (9.3% vs. 0.2%, p<0.01), markedly improved the NAFLD activity score (NAS) (−2.4 vs. −1.4, p=0.05). Of note, this was mainly due to an improvement in steatosis (−1.1 vs. −0.3, p=0.02), as lobular inflammation and ballooning were similarly improved in both groups, while fibrosis was unchanged. However, when the data was re-analyzed for subjects who had lost ≥7% of body weight, lobular inflammation and ballooning were significantly improved (as well as steatosis), underscoring the importance of the magnitude of weight loss to histologic benefit.
Although there are no large, long-term studies to assess the role of lifestyle intervention in patients with NAFLD/NASH, the importance of this strategy in the management of patients with NAFLD cannot be overemphasized.Weight loss achieved through different strategies appears to be the most important factor to improve plasma aminotransferases, liver fat content, and histology. The role of lifestyle changes in the absence of weight reduction is less clear and needs further assessment in well-designed studies.
Dietary intervention and weight-loss medications
Because long-term compliance with lifestyle interventions remains a major challenge, several investigators have examined the role of pharmacologic interventions to enhance weight loss in patients with NAFLD/NASH. In a double-blind, randomized, controlled trial (RCT) by Zelber-Sagi et al. , 52 patients were randomized to lifestyle intervention±orlistat for 6 months. After treatment there was no difference in weight loss between the two groups (−8% vs. −6%, p=0.26). Histologic changes were similar in both groups.
In another RCT in 50 obese patients with NASH, adding orlistat for 36 weeks to lifestyle intervention and vitamin E (800 UI daily) did not enhance weight loss (−8.3% vs. −6.0%, p=NS) or impact liver histology (primary endpoint) . However, subjects who lost ≥5% of body weight markedly improved hepatic steatosis, while amelioration of liver necrosis and lobular inflammation required a body weight loss of at least 9%, suggesting that there is a lower threshold for steatosis resolution compared to other histologic parameters.
In summary, the role of the weight-loss medications tested to date in NAFLD is marginal. Large RCTs are much needed to assess the effectiveness of the recently FDA-approved weight management agents lorcaserin (Belviq®) and phentermine/ topiramate extended-release (Qsymia®) in patients with NAFLD.
Surgically induced dietary intervention
The use of bariatric surgery as an alternative method to lose weight has become more popular in recent decades given the challenge of long-term weight loss. Different procedures are currently available and can be divided into three types: restrictive (reduction of gastric capacity), malabsortive (bypasses a segment of the small bowel), and hybrid procedures (combination of the above) . Currently, the most popular procedures are Roux-en-Y gastric bypass (RYGB; hybrid) and laparoscopic adjustable gastric banding (LAGB; restrictive). There is large evidence on the effects of bariatric surgery on glucose control , insulin sensitivity , and overall improvement of related comorbidities (hypertension, dyslipidemia, and so on)  as well as positive effects on long-term survival . However, the impact of these procedures on liver histology in patients with NAFLD has not been so extensively addressed. Moreover, most of the existing studies are retrospective, uncontrolled or have a small sample size [67–78].
Prospective studies [69–74,78] specifically assessing the role of RYGB have reported overall positive results with improvement in steatosis, inflammation, ballooning, and to a lesser extent fibrosis, after 12–24 months of follow-up. In some reports, histologic improvement led to complete NASH resolution [69,71,73]. Of note, in all of these studies there was a very significant body weight reduction after surgery. The main drawbacks of these studies, however, have been their sample size (7 to 19 patients, except for one study  that included 70 patients) and the lack of a control group. In addition, a few studies have reported that some patients develop new fibrosis after the surgical intervention [68,71,72]. As a consequence of these findings, there have been some concerns regarding the effect of bariatric surgery on hepatic fibrosis. Several hypotheses have been suggested to explain the potential mechanisms responsible for worsening fibrosis in a subgroup of patients: rapid weight loss, lack of macro/micronutrients replacement or the direct effect of the procedure in a liver with pre-existing fibrosis. Of note, other series have not found any worsening of fibrosis in patients with NASH after bariatric surgery [69,73,74,78].
Regarding LAGB, although fewer studies have assessed its effect on histology in patients with NAFLD, studies are larger (n=36 to 381) and many of longer duration (25–50 months) [75–77]. Similarly to RYGB, after major weight loss there is usually a significant improvement in steatosis, inflammation, and fibrosis. However, Mathurin et al. , in a large long-term follow-up study (381 patients followed for up to 5 years), found no change in hepatic inflammation and even a significant increase in fibrosis after one year, but without any further deterioration over time.
At present, there are no RCTs that have prospectively evaluated the best bariatric surgery approach for patients with NAFLD. It is also not well understood if the histologic changes after bariatric surgery in NASH are just the consequence of weight reduction or are secondary to other metabolic effects, such as changes in the gut microbiota or intestinal hormones. Further investigation is needed to determine the real long-term benefit of bariatric surgery in patients with NASH.