Soft drink consumption is associated with fatty liver disease independent of metabolic syndrome

Background/Aims: The independent role of soft drink consumption in non-alcoholic fatty liver disease (NAFLD) patients remains unclear. We aimed to assess the association between consumption of soft drinks and fatty liver in patients with or without metabolic syndrome. Methods: We recruited 31 patients (age: 43 ± 12 years) with NAFLD and risk factors for metabolic syndrome, 29 patients with NAFLD and without risk factors for metabolic syndrome, and 30 gender- and age-matched individuals without NAFLD. The degree of fatty infiltration was measured by ultrasound. Data on physical activity and intake of food and soft drinks were collected during two 7-day periods over 6 months using a food questionnaire. Insulin resistance, inflammation, and oxidant-antioxidant markers were measured. Results: We found that 80% of patients with NAFLD had excessive intake of soft drink beverages (>500 cm³/day) compared to 17% of healthy controls (p < 0.001). The NAFLD group consumed five times more carbohydrates from soft drinks compared to healthy controls (40% vs. 8%, p < 0.001). Seven percent of patients consumed one soft drink per day, 55% consumed two or three soft drinks per day, and 38% consumed more than four soft drinks per day for most days and for the 6-month period. The most common soft drinks were Coca-Cola (regular: 32%; diet: 21%) followed by fruit juices (47%). Patients with NAFLD with metabolic syndrome had similar malonyldialdehyde, paraoxonase, and C-reactive protein (CRP) levels but higher homeostasis model assessment (HOMA) and higher ferritin than NAFLD patients without metabolic syndrome (HOMA: 8.3 ± 8 vs. 3.7 ± 3.7 mg/dL, p < 0.001; ferritin: 186 ± 192 vs. 87 ± 84 mg/dL, p < 0.01). Logistic regression analysis showed that soft drink consumption is a strong predictor of fatty liver (odds ratio: 2.0; p < 0.04) independent of metabolic syndrome and CRP level. Conclusions: NAFLD patients display higher soft drink consumption independent of metabolic syndrome diagnosis. These findings might optimize NAFLD risk stratification.