Association of homocysteine level with biopsy-proven non-alcoholic fatty liver disease: a meta-analysis

Previous studies have reported inconsistent findings regarding the association between plasmatic higher of homocysteine level and non-alcoholic fatty liver disease. We aimed to investigate this association by conducting a meta-analysis. Literature was searched on PubMed from inception to January 2015. Eight studies evaluating plasma level of homocysteine in biopsy-proven non-alcoholic fatty liver disease subjects compared to healthy controls were included. Compared with the controls, non-alcoholic fatty liver disease patients witnessed a higher level of homocysteine [standard mean difference (SMD): 0.66 µmol/L, 95% CI: 0.41, 0.92 µmol/L], and were associated with a significant increased risk for hyperhomocysteinemia [odds ratio (OR) 5.09, 95% CI: 1.69, 15.32]. In addition, patients with non-alcoholic fatty liver presented 0.45 µmol/L higher levels of homocysteine compared to healthy controls (95% CI: 0.09, 0.82 µmol/L), whereas non-alcoholic steatohepatitis patients had 1.02 µmol/L higher levels of homocysteine (95% CI: 0.28, 1.76 µmol/L). There was neither difference of folate level nor vitamin B12 level between non-alcoholic fatty liver disease subjects and healthy controls. This study revealed that non-alcoholic fatty liver disease patients presented an increased serum concentration of homocysteine, and were associated with an increased risk of hyperhomocysteinemia. Further studies are needed to demonstrate a causal role of hyperhomocysteinemia in non-alcoholic fatty liver disease.