Abstract
Interest in research on malnutrition is decreasing due to thoughts that the problem of malnutrition has been solved in an age of over-nourishment or obesity and defining malnutrition is not uniform. This study aimed to critically appraise the prevalence of malnutrition according to various diagnostic tools and proportion of severity used in previous studies. A literature review was performed using a total of 16 studies published between 1980 and 2020 regarding malnutrition in patients with chronic liver disease. Most of the analyzed studies were conducted before 2010, and only a few studies were conducted after 2010. Nutrition assessment tool (NAT) and nutrition screening tool (NST) to explain malnutrition were distinguished; however, there was no clear distinction between them. NST often used questionnaires while NST used various malnutrition measuring tools. Our results show that, in the age of over-nourishment, reduction in malnutrition in chronic liver disease still hasn't been significant. Malnutrition prevalence in studies published prior to 2,000 ranged between 13.3% and 85% (mean, 37.6%), whereas that in studies published after 2,000 ranged between 13.3% and 78.5% (mean, 35.2%). Malnutrition prevalence largely depends on the diagnostic tool and proportion of disease severity in the target population. The prevalence of malnutrition in patients with chronic liver diseases varies widely. This big difference is related to various diagnostic tools, mixed etiologies, and different disease severity in different studies. The prevalence of malnutrition was 36.4% (10-80.3%) in all patients with liver disease, 39.9% (13.3-80.3%) in compensated liver disease, and 44.1% (26.7-93.6%) in decompensated cirrhosis. Malnutrition prevalence was 38.2% and 23.7% in alcoholism-related and hepatitis C virus (HCV)-related diseases, respectively. Malnutrition also largely depended on the judgement tool. Malnutrition prevalence according to the diagnostic tool was approximately 28-85% for subjective global assessment (SGA), 30.8-78.5% for anthropometric approach, and 21-80.3% for clinical judgment. It became similar over time.