Ammonia for Hepatic Encephalopathy from Cirrhosis

Ammonia Levels Do Not Guide Clinical Management of Patients With Hepatic Encephalopathy Caused by Cirrhosis.

Haj M, Rockey DC.

Am J Gastroenterol. 2020 May;115(5):723-728. [Full text]

Hepatic encephalopathy (HE) results from an accumulation of neurotoxins, such as ammonia [2]. Ammonia is a protein breakdown product largely carried out by enterocytes and colonic microflora. Normally, ammonia is converted to urea in the liver (urea cycle) and is then cleared by the kidneys and skeletal muscle. In cirrhosis, portosystemic collaterals, sarcopenia, and renal dysfunction, if present, decreases ammonia metabolism and excretion. Subsequently some of the the increased ammonia can cross the blood-brain barrier and cause astrocyte dysfunction and cerebral edema [1] with severity graded from minimal, or if symptomatic from grade I to grade IV, or coma [3].

It follows that ammonia could be a marker of HE; however, accurate ammonia levels in practice require rapid processing of samples [4] and “normal” results vary from person to person and patients with cirrhosis can have a higher “baseline” ammonia level [1, 4]. Given this uncertainty, the authors of the current study hypothesized that patients with a measured ammonia level that was elevated would receive more aggressive lactulose administration than those with a normal ammonia level or without an ammonia level checked.

Patient population and Design

This was a retrospective cohort study of consecutive patients admitted to a single academic hospital in South South Carolina with cirrhosis and HE from 2005 to 2015 identified by ICD code and all records were reviewed to verify the diagnosis of HE.

The authors excluded patients with an obvious alternative cause of altered mental status, patients younger than 18 years, HE secondary to acute liver failure, or with a liver transplant, or already on se therapy when the ammonia was drawn.

The primary outcome was the amount of oral lactulose given in the first 48 hours. A secondary outcome was the time until resolution of HE symptoms, or an HE grade of zero (back to baseline). At the study institution, hyperammonemia was considered at an ammonia level >72 ╬╝mol/L.


A total of 1,202 cirrhotic patients with HE were identified, with most of the cirrhosis secondary to nonalcoholic steatohepatitis (38%) or alcohol (30%). Of these, 551 patients (46%) had an ammonia level drawn at admission and 651 patients (54%) did not. The MELD and Child-Pugh scores were similar between these groups. Of those with an ammonia level checked, 328/551 (60%) had elevated ammonia levels. Interestingly, in patients with a normal ammonia, infection was a more common precipitant (42%) of the HE compared those with an elevated ammonia (26%, P < 0.001).

Each patient in the study cohort received lactulose with and average dose of 172 mL in the first 48 hours. Those with an ammonia checked had a similar amount of lactulose administered in the first 48 hrs as those that did not, 167 versus 171 mL (P = 0.08) and there was no correlation between the ammonia level and lactulose dose received (R2 = 0.0026).

The time to resolution of HE was similar in those without ammonia level and those with ammonia levels drawn, 3.6 days for each (P = 0.15). And the subset of patients with an elevated ammonia was also similar to those with a normal ammonia, 3.4 days vs. 3.8 days, (P = 0.07).


Ammonia does seem to play a role in the pathogenesis of HE, and as a diagnostic test when the cause of encephalopathy is unknown checking ammonia may have a role (i.e. valproic acid toxicity). This study shows that lactulose therapy was administered and dosed independent of ammonia level and that time to resolution of HE did not depend on the ammonia level. In other words,these findings suggest that HE is a clinical diagnosis and the management of HE should not be influenced by the ammonia level.

In a prior, much smaller cohort study of 121 patients with HE (vs. 1,202 in this study) it was found that “ammonia levels correlate with the severity of HE” [5]. But it should be noted the correlation, a Spearman rank correlation of 0.61 for arterial ammonia, is in fact not very strong. Furthermore, clinical outcomes like time to HE resolution or length of stay were not reported, nor was the management strategy and if it changed based on severity.

What this study then re-emphasizes is that HE is a clinical diagnosis. Checking an ammonia level may have a role when there is some diagnostic uncertainty. Otherwise the management of HE from cirrhosis is not influenced by the ammonia level and checking ammonia remains one of the “Things we do for no Reason” [6].

F: Follow upThrough index hospital stay
R: RandomizationN/A
I: Intention to treatN/A
S: Similar at baselineYes
B: BlindingN/A
E: Equal treatmentYes
S: Source (funding)Internally funded

  1. Haj M, Rockey DC. Ammonia Levels Do Not Guide Clinical Management of Patients With Hepatic Encephalopathy Caused by Cirrhosis. Am J Gastroenterol. 2020 May;115(5):723-728.
  2. Parekh PJ, Balart LA. Ammonia and Its Role in the Pathogenesis of Hepatic Encephalopathy. Clin Liver Dis. 2015 Aug;19(3):529-37.
  3. Vilstrup H, Amodio P, et al. Hepatic encephalopathy in chronic liver disease: 2014 Practice Guideline by the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver. Hepatology. 2014 Aug;60(2):715-35.
  4. Vierling, JM, Mokhtarani, M, Brown, RS, et al. Fasting Blood Ammonia Predicts Risk and Frequency of Hepatic Encephalopathy Episodes in Patients With Cirrhosis. Clinical Gastroenterology & Hepatology, 2016; 14 (6), 903-906e1.
  5. Ong JP, Aggarwal A, Krieger D, et al. Correlation between ammonia levels and the severity of hepatic encephalopathy. Am J Med. 2003 Feb 15;114(3):188-93. doi: 10.1016/s0002-9343(02)01477-8. PMID: 12637132.
  6. Ninan J, Feldman L. Ammonia Levels and Hepatic Encephalopathy in Patients with Known Chronic Liver Disease. J Hosp Med. 2017 Aug;12(8):659-661.