We congratulate Berzigotti et al.1 for performing an excellent study that raises more questions, just as many post hoc analyses of observational studies do. It is intriguing that obesity may predict future decompensation in patients with compensated cirrhosis and portal hypertension.
The potential mechanism is unexplained. The presence of a proinflammatory and proangiogenic state driven by extrahepatic adipose tissue may accelerate existing liver disease. Furthermore, hemodynamic changes linked to metabolic syndrome may affect the development of portal hypertension. Age is another well-recognized determinant of outcomes for patients with liver disease. Obesity and smoking have been shown to accelerate the biological aging process.2 Therefore, it would be interesting to determine whether smoking was also associated with accelerated decompensation in this cohort. Finally, the presence of a dual hepatic pathology must be considered as an explanation for the reported association. Hence, a review of biopsy findings for cirrhosis and earlier stages of fibrosis (where available) could further clarify the current findings.
The translation of the findings of this study into an effective intervention could be challenging. The prognosis of patients with cirrhosis is influenced by different macronutrient intake patterns3 and is worsened by protein-energy malnutrition.4 In addition, weight-loss targets have been difficult to achieve in the setting of nonalcoholic fatty liver disease without cirrhosis, even with motivated patients and intensive community support.5 In comparison with calorie-matched daytime supplementation, nocturnal nutritional supplements improve total body protein,6 and this suggests that prolonged fasting is probably detrimental for patients with cirrhosis. Very low-calorie diets and carbohydrate restriction may pose risks to patients with cirrhosis by inducing hepatic stressors (i.e., an increased demand for endogenous gluconeogenesis and an increased requirement for catabolizing ketone bodies). We have limited data about a cirrhotic liver's ability to cope with such insults. Increased exercise is ordinarily an attractive weight-loss strategy, but impaired endogenous gluconeogenesis in patients with cirrhosis results in the potential risk of exercise-induced hypoglycemia during aerobic exercise.7
Therefore, caution needs to be applied when weight loss is being recommended as an intervention for these patients. We reinforce the authors' comment that dedicated studies on the effects of obesity and weight loss on the prognosis of patients with compensated cirrhosis are needed. These studies should be performed in parallel with studies evaluating the methods, safety, and efficacy of lifestyle interventions.