The prevalence and prognostic significance of vitamin C deficiency in patients with cirrhosis: a prospective observational cohort study

ABSTRACT

Background and aims: Malnutrition is a common comorbidity in cirrhotic patients and confers a poorer prognosis. Vitamin C (VC) is a micronutrient essential for human health. Vitamin C deficiency (VCD) can lead to scurvy and may impair immune and liver functions. Although previously thought to be rare in developed countries, VCD is now well described in patients with pneumonia, COVID19 and upper gastrointestinal bleeding (UGIB). The prevalence and clinical significance of VCD in cirrhosis remains poorly studied. Method: Patients with cirrhosis admitted to 3 metropolitan tertiary centres in Australia were prospectively included over a 10-month period in 2021. Fasting VC levels were collected on admission and we recorded demographic data and clinical outcomes. The primary outcomes were the prevalence of VCD (defined as VC level <23 mcmol/L) and severe VCD (SVCD), defined as <11 mcmol/L. Secondary outcomes included mortality, intensive care admission, length of stay (LOS) and rate of infection. Results: 117 patients were included. Mean age was 57.1 ± 13.9 years, 59.0% were male and 23.9% belonged to the lowest socioeconomic decile. The most common aetiologies of cirrhosis were alcohol (62.4%), viral hepatitis (24.0%) and non-alcoholic fatty liver disease (18.8%). Median MELD scorewas 29 (IQR 22–36) and Child Pugh (CP) grades were 12.8% A, 46.2% B and 41.0% C. Most patients (74.4%) were hospitalised with complications of decompensated cirrhosis, including ascites (59.0%), encephalopathy (31.6%) and variceal bleeding (11.1%). Median VC level was 34mcmol/L (IQR 16–55) and did not differ with age, gender, or aetiology of cirrhosis. Increasing CP grade correlated with significantly lower median VC levels (CP-A 46.0 mcmol/L vs. CPB 36.5 mcmol/L and CP-C 20.5 mcmol/L, p = 0.026). The prevalence of VCD and SVCD were 39.3% and 17.1% respectively. SVCD was more prevalent in patients with a body mass index <25 (28.3% vs 13.0%, p = 0.036). In-hospital mortality was 12.8% and did not differ by VCD status, however in the subgroup of patients presenting with UGIB, SVCD correlated with significantly higher mortality (50% vs 4.1%, p = 0.045). Bacteraemia was more frequent in patients with VCD (13.3% vs. 1.4%, p = 0.014) and SVCD (26.3% vs 2.1%, p < 0.001), which remained significant at multivariate analysis (OR for every 1mcmol/L increase in VC, 0.91 (95% CI 0.83–0.99), p = 0.037). Overall infection rateswere higher in patients with SVCD (40.0% vs. 27.8%) although thiswas nonsignificant (p = 0.279). Median hospital LOS was 10 (IQR 6–18) days and did not differ by VCD status. (Figure Presented) Conclusion: VCD is common in hospitalised cirrhotic patients and prevalence increases with severity of liver disease. VCD increases the risk of infective complications and higher mortality was observed in patients with UGIB and SVCD. Further studies are required to assess the significance of VCD in cirrhosis and the impacts of VC replacement.