The pathogenesis of hepatic encephalopathy(HE) is unclear. However gut flora changes, inflammation and neuro-glial injury have been implicated. The aim was to evaluate factors that were associated with HE recurrence after lactulose withdrawal by analyzing the clinical phenotype, stool microbiome and systemic metabolome longitudinally. HE patients on a standard diet who were adherent on lactulose underwent characterization of their phenotype [cognition, inflammatory cytokines, in-vivo brain MR spectroscopy(MRS)], gut microbiome (stool Multitag Pyrosequencing) and metabolome (urine/serum ex-vivo MRS) analysis while on lactulose and on days 2, 14 and 30 post-withdrawal. Patients whose HE recurred post-withdrawal were compared to those without recurrence. We included seven men (53 ± 8 years) who were adherent on lactulose after a precipitated HE episode were included. HE recurred in three men 32 ± 6 days post-withdrawal. In-vivo brain MRS showed increased glutamine+glutamate (Glx) and decreased myoinositol with a reduction in stool Faecalibacterium spp., post-withdrawal. HE recurrence was predicted by poor baseline inhibitory control and block design performance and was associated with a shift of choline metabolism from tri-methylamine oxide formation towards the development of di-methylglycine, glycine and creatinine. This was accompanied by a mixed effect on the immune response (suppressed IL-10 and Th1/Th2/Th17 response). The correlation network showed Prevotella to be linked to improved cognition and decreased inflammation in patients without HE recurrence. We conclude that lactulose withdrawal results in worsening cognition, mixed inflammatory response effect, lowered stool Faecalibacterium and increase in MR-measurable brain Glx. HE recurrence post-lactulose withdrawal can be predicted by baseline cognitive performance and is accompanied by disrupted choline metabolism.