Ursodeoxycholic acid and COVID-19 outcomes: a cohort study and data synthesis of state-of-art evidence

Figures & data

Figure 1. Study participant flowchart.

Table 1. Demographic and clinical characteristics by treatment cohort for UDCA.

	with UDCA N = 65	without UDCA N = 50	P value
Continuous variables
Age, years			0.698
Median (IQR)	23 (3.0)	23 (3.8)
BMI, kg/m^2			0.093
Median (IQR)	22.3 (4.6)	20.9 (3.8)
Body temperature on admission, ℃			0.175
Median (IQR)	39.3 (1.0)	39.4 (0.8)
Categorical variables
Sex			0.939
Male, no. (%)	40 (61.5)	32 (64.0)
Female, no. (%)	25 (38.5)	18 (36.0)
Whether to use NSAIDs			<0.001
Yes, no. (%)	63 (96.9)	38 (76.0)
No, no. (%)	2 (3.1)	12 (24.0)
Whether to use Symptomatic drugs			0.316
Yes, no. (%)	64 (98.5)	47 (94.0)
No, no. (%)	1 (1.5)	3 (6.0)
Days from fever onset to hospitalization			0.214
0, no. (%)	6 (9.2)	2 (4.0)
1, no. (%)	18 (27.7)	23 (46.0)
2, no. (%)	34 (52.3)	21 (42.0)
3, no. (%)	7 (10.8)	4 (8.0)
Subgroup divided by a daily dose			–
150 mg, no. (%)	29 (44.6)	–
300 mg, no. (%)	36 (55.4)	–
(All received at least two doses of the WHO certified vaccines against SARS-CoV-2 infection)
(All were youthful individuals with no comorbidities)

IQR, interquartile ranges; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NSAIDs, non-steroidal Anti-inflammatory, including ibuprofen, acetaminophen, and compound preparation mainly composed of these two drugs; Symptomatic drugs were defined as drugs recommended in Diagnosis and Treatment of Novel Coronavirus Pneumonia in China (Version 9) for alleviating symptoms other than fever, such as sore throat, cough, shortness of breath, vomit, and diarrhea.

Figure 2. Time to Body Temperature Recovery Cumulative recovery curves for time to Body Temperature Recovery are shown for the cohorts that use UDCA for treatment (a) and subgroups divided by daily dose (b). The median recovery time and Log-rank test were presented by groups and subgroup (c).

Table 2. Major adjusted, unadjusted, and minor adjusted Cox proportional models by treatment cohort for UDCA and subgroups divided by daily dose.

	Without UDCA	With UDCA	Daily 150 mg	Daily 300 mg
No. of patients	50	65	29	36
No. of events (%)	38 (76.0)	61 (93.8)	28 (96.6)	33 (91.7)
Major adjusted with all covariates (Primary analysis)
Hazard ratio (95% CI)	1 (Reference)	1.63 (1.01–2.60)	1.36 (0.76–2.40)	1.86 (1.10–3.10)
P value		0.044	0.293	0.040
Unadjusted
Hazard ratio (95% CI)	1 (Reference)	1.87 (1.06–3.32)	1.70 (0.86–3.50)	2.00 (1.03–3.90)
P value		0.031	0.125	0.081
Minor adjusted with demographics, whether to use NSAIDs and days from fever onset to hospitalization
Hazard ratio (95% CI)	1 (Reference)	1.66 (1.04–2.70)	1.49 (0.85–2.60)	1.81 (1.08–3.1)
P value		0.035	0.161	0.050
Minor adjusted with demographics, whether to use NSAIDs and the temperature on admission
Hazard ratio (95% CI)	1 (Reference)	1.64 (1.03–2.60)	1.39 (0.78–2.50)	1.84 (1.11–3.10)
P value		0.038	0.266	0.038

Figure 3. PRISMA Flow Diagram.

Table 3. Selected characteristics of the 9 included cohort studies.

Study	Study period	Target population	Country	Outcome	Effect size	Participants, No.	Female, No. (%)	Age, years	PSM	Reference
Colapietro et al., 2023	From 01-Jan-2020 to 01-Jan-2023	Hospitalized for COVID-19 with comorbidities	Italy	1. Severe/critical rate (CPAP rate)^a 2. Mortality rate^b	OR	UDCA: 57 Control: 3790	UDCA: 26 (45.6%) Control: 1494 (39.4%)	Mean (IQR) UDCA: 70.0 (58.0–82.0) Control: 74.0 (61.0–83.0)	NO	[4]
Hu et al., 2023	From 01-Jan-2015 to 21-Dec-2022	Liver disease (with liver transplantation)	China	1. Infection rate	Adjusted OR	UDCA: 326 Control: 571	UDCA: 84 (25.7%) Control: 105 (18.4%)	Median (IQR) UDCA: 50.5 (43.0–58.0) Control: 51.0 (44.0–57.0)	YES	[6]
				2. Severe/critical rate^d	OR	UDCA: 239 Control: 464	UDCA: 60 (25.1%) Control: 80 (17.2%)	Median (IQR) UDCA: 50.0 (43.0–58.0) Control: 50.0 (43.0–57.0)
				3. Length of hospital stay	NR	UDCA: 28 Control: 41	NR	NR
John et al., 2023	From 03-Jan-2020 to 02-Nov-2022	Cirrhosis (without liver transplantation）	USA	1. Infection rate 2. Severe/critical rate^d 3. Mortality rate^c	Adjusted OR	UDCA: 1607 Control: 1607	UDCA: 114 (7.1%) Control: 87 (5.4%)	Median (IQR) UDCA: 62.2 (11.9) Control: 61.8 (11.2)	YES	[7]
Li et al., 2023	From 12-Dec-2022 to 15-Jan-2023	Chronic liver disease	China	1. Infection rate 2. Severe/critical rate^e 3. Mortality rate^b 4. Recovery time	1. OR and adjusted OR 2. Only events 3. Only events 4. NR	UDCA: 225 Control: 225	UDCA: 134 (59.6%) Control: 128 (56.9%)	Mean ± SD UDCA: 53.1 ± 12.0 Control: 53.6 ± 10.7	YES	[8]
Marrone et al., 2023	From 01-Mar-2020 to 31-Dec-2022	Hospitalized for COVID-19 with comorbidities	Italy	1. Severe/critical rate (MV rate)^a 2. Mortality rate^b 3. Length of hospital stay	1. Calculated by events 2. Adjusted HR 3. NR	UDCA: 108 Control: 521	UDCA: 54 (50%) Control: 232 (44.5%)	Median (IQR) UDCA: 76 (61–84) Control: 75 (63–83)	YES	[9]
Brevini et al., 2023	NR	Hospitalized for COVID-19 with chronic liver disease	USA	1. Severe/critical rate^d 2. Mortality rate^b	1. Only events 2. Only events	UDCA: 31 Control: 155	UDCA: 28 (90.3%) Control: 139 (89.6%)	Median (IQR) UDCA: 56 (44–63) Control: 57 (47–63)	YES	[3]
Tang et al., 2023	From 01-Dec-2022 to 31-Jan-2023	Hospitalized for COVID-19 with comorbidities except liver disease	China	Recovery time	NR	UDCA: 22 Control: 48	UDCA: 25 (38.5) Control: 18 (36.0)	Median (IQR) UDCA: 34 (30–55) Control: 33 (30–54)	NO	[10]
Costello et al., 2023	From 01-Mar-2020 to 31-Dec-2022	Primary biliary/sclerosing cholangitis	UK	1. Severe/critical rate (COVID-19 related hospitalization rate) 2. Mortality rate^c	HR and adjusted HR	UDCA: 7225 Control: 4080	UDCA: 3035 (74.4) Control: 5820 (80.6)	Age category (≤60 years) UDCA: 2245 (31.0%) Control: 1595 (39.0%)	NO	[5]
Yu et al., 2023 (This article)	From 10-Dec-2022 to 30-Dec-2022	Hospitalized for COVID-19 without any comorbidities	China	Recovery time	HR and adjusted HR	UDCA: 65 Control: 50	UDCA: 25 (38.5) Control: 18 (36.0)	Median (IQR) UDCA: 23 (3.0) Control: 23 (3.8)	NO	/

^aTwo studies from Italy exclusively provided data on Continuous Positive Airway Pressure (CPAP) and Mechanical Ventilation (MV) rates. The criteria for employing CPAP and MV in treating COVID-19 in the local context align closely with the assessment criteria for severe/critical illness [18,19]. Given the congruence in evaluation criteria across both groups in the single study, these outcomes were pooled as an approximation of the severe/critical illness rate during the consolidation process.

^bAll-cause mortality.

^cCOVID-19-related death.

^dNIH severity scale [20].

^eThe 10th edition of the Chinese protocol for the treatment of SARS-CoV-2 infection [21].

CPAP, Continuous Positive Airway Pressure; MV, Mechanical Ventilation; OR, Odds ratio; NR, Not report; HR, Hazard ratio; PSM, Propensity Score Matching; IQR, interquartile range.

Figure 4. Pooled Association between UDCA treatment and clinical outcomes of COVID-19 patients. The outcome included (a) Severe/critical rate and (b) Mortality rate. *Due to the outcome being rare (less than 10%), the RR converted by HR can be approximated as OR, to keep consistent with other studies.

Figure 5. Pooled Association between UDCA treatment and recovery of COVID-19 patients. The outcome included (a) Length of hospital and (b) Recovery time.

Colapietro F, Angelotti G, Masetti C, et al. Ursodeoxycholic acid does not improve COVID-19 outcome in hospitalized patients. Viruses-Basel. 2023 Aug;15(8):1738. doi: 10.3390/v15081738

 PubMed Web of Science ®Google Scholar

Hu L, Zhang H, Huang C, et al. Effect of ursodeoxycholic acid on preventing SARS-CoV-2 infection in patients with liver transplantation: a multicenter retrospective cohort study. QJM Int J Med. 2023 Nov 9;117(5):339–347. doi: 10.1093/qjmed/hcad254

 Web of Science ®Google Scholar

John BV, Bastaich D, Webb G, et al. Ursodeoxycholic acid is associated with a reduction in SARS-CoV-2 infection and reduced severity of COVID-19 in patients with cirrhosis. J Intern Med. 2023 May;293(5):636–647. doi: 10.1111/joim.13630

 PubMed Web of Science ®Google Scholar

Li Y, Zhu N, Cui X, et al. Protective effect of ursodeoxycholic acid on COVID-19 in patients with chronic liver disease. Front Cell Infect Microbiol. 2023 May 3;13. doi: 10.3389/fcimb.2023.1178590

 Web of Science ®Google Scholar

Marrone G, Covino M, Merra G, et al. Ursodeoxycholic acid does not affect the clinical outcome of SARS-CoV-2 infection: a retrospective study of propensity score-matched cohorts. Liver Int. 2023;44(1):83–92. doi: 10.1111/liv.15736

 PubMed Web of Science ®Google Scholar

Brevini T, Maes M, Webb GJ, et al. FXR inhibition may protect from SARS-CoV-2 infection by reducing ACE2. Nature. 2023 Mar 2;615(7950):134–142. doi: 10.1038/s41586-022-05594-0

 PubMed Web of Science ®Google Scholar

Tianjiao T, Huatao L, Bibo M, et al. Clinical efficacy of ursodeoxycholic acid in the treatment of COVID-19. Chin J Clin Infect Dis. 2023;16(4):289–292,298.

 Google Scholar

Costello RE, Waller KMJ, Smith R, et al. Ursodeoxycholic acid and severe COVID-19 outcomes in people with liver disease: a cohort study using the OpenSAFELY platform. medRxiv. 2023:2023.12.

 Google Scholar

Passamonti F, Cattaneo C, Arcaini L, et al. Clinical characteristics and risk factors associated with COVID-19 severity in patients with haematological malignancies in Italy: a retrospective, multicentre, cohort study. Lancet Haematol. 2020;7(10):e737–e745. doi: 10.1016/S2352-3026(20)30251-9

 PubMed Web of Science ®Google Scholar

Brusasco C, Corradi F, Dazzi F, et al. The use of continuous positive airway pressure during the second and third waves of the COVID-19 pandemic. ERJ Open Res. 2023;9(2):00365–2022. doi: 10.1183/23120541.00365-2022

 PubMedGoogle Scholar

Clinical spectrum of SARS-CoV-2 infection. 2022. Available from: https://www.covid19treatmentguidelines.nih.gov/overview/clinical-spectrum/

 Google Scholar

China updates COVID-19 diagnosis, treatment protocol [Internet]. 2023 [cited 2023]. Available from: http://english.nmpa.gov.cn/2023-01/10/c_861471.htm

 Google Scholar

Data availability statement

Data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.