To the Editor:
We read with great interest the letter by Ji et al.
1
Obesity is a well-recognized risk factor for the development of non-alcoholic fatty
liver disease (NAFLD) or metabolic dysfunction-associated liver disease (MAFLD) and
is associated with adverse outcomes in COVID-19 patients.
2
,
3
Qatar's population has a high prevalence of obesity
4
and also has one of the highest rates of COVID-19 cases per million population, with
one of the lowest mortality rates.
5
We hypothesized that NAFLD is an independent risk factor for worse outcomes in hospitalized
COVID-19 patients in our population.
Methods
We studied 589 patients with confirmed symptomatic COVID-19 who were hospitalized
from May 2020 to June 2020 to COVID-19 facilities in the state of Qatar. We categorized
them into 2 groups; NAFLD and no NAFLD, based on the hepatic steatosis index (HSI).
6
People with an HSI index of 36 and above were considered as having NAFLD. The lowest
aspartate aminotransferase (AST) and alanine aminotransferase (ALT) values within
the last 1 year before the COVID-19 diagnosis, and, when unavailable, the lowest values
at the time of recovery were used for calculation of HSI index. The primary outcome
was mortality, and secondary outcomes were disease severity on presentation, and disease
progression, and liver injury. Development of acute respiratory distress syndrome
(ARDS), requirement for intensive care unit (ICU) admission, and mechanical ventilation
were regarded as markers of disease progression. Disease severity was defined using
the WHO classification into severe and non-severe.
7
Liver injury was classified as borderline if ALT or AST were less than twice the upper
limit of normal (ULN), mild if elevated 2–5×, moderate if 5–15×, and severe if more
than 15× the ULN.
8
There were no patients with excessive use of alcohol in our study. The Medical Research
Center of Hamad Medical Corporation approved the study (MRC-01-20-631).
Results
Univariate regression analysis showed that age, gender, diabetes mellitus, and increased
BMI were significantly associated with mortality (Table 1
). NAFLD was significantly associated with increased disease severity on admission,
ICU admission, and requirement for mechanical ventilation (Table 1). We performed
multivariate regression analysis using all the statistically significant variables.
Age above 50 years was the only predictor of increased mortality (odds ratio [OR]
8.88; 95% CI 3.61–21.84; p <0.000∗). For disease severity on presentation, only the
presence of diabetes mellitus was a statistically significant predictor (OR 2.2; 95%
CI 1.5–3.19; p <0.000). Age above 50 years and a BMI above 25 kg/m2 were the only
2 significant predictors of all 3 markers of disease progression in our study: development
of ARDS, ICU admission, and mechanical ventilation. Age was associated with increased
risk of ARDS (OR 2.57; 95% CI 1.80–3.66; p <0.000∗), ICU admission (OR 2.36; 95% CI
1.67–3.33; p <0.000∗) and mechanical ventilation (OR 2.03; 95% CI 1.42–2.90; p <0.000∗).
Obesity was associated with increased risk of ARDS (OR 1.97; 95% CI 1.32–2.94; p =
0.001), ICU admission (OR 1.83; 95% CI 1.26–2.65; p = 0.001) and mechanical ventilation
(OR 1.89; 95% CI 1.25–2.87; p = 0.002). The presence of NAFLD was not an independent
predictor of increased mortality, disease severity on presentation, or disease progression.
However, the presence of NAFLD was a predictor of the development of mild liver injury
(OR 2.99; 95% CI 1.62–4.37; p = 0.000) and moderate liver injury (OR 5.104; 95% CI
3.21–6.99; p = 0.000).
Table 1
Comparative results of patients with and without NAFLD on univariate analysis.
No NAFLD (n = 269)
NAFLD (n = 320)
p value
Age (years) [±SD]
44.5 [13.85]
47.78 [13.4]
0.0073
Gender (male percentage)
242, (Male 89.6%)
257, (Male 80.8%)
0.016
Diabetes mellitus
75 (27.78%)
160 (50.3%)
0.000
Hypertension
70 (25.93)
135 (42.45%)
0.000
Smoking
31 (11.56%)
37 (11.86%)
0.764
BMI [±SD]
25.61 [7.47]
30.76 [4.9]
0.000
Coronary artery disease
19 (7.04%)
30 (9.43%)
0.224
Chronic kidney disease
17 (6.32%)
33 (10.38%)
0.062
Cirrhosis
7 (2.59%)
3 (0.95 %)
0.516
Malignancy
12 (4.46%)
4 (1.25%)
0.067
Lung disease
15 (5.56%)
25 (7.85%)
0.252
Use of Immunosuppressive drugs
11 (4.07%)
14 (4.39%)
0.556
Severe disease on presentation
60 (22.3%)
95 (29.87%)
0.041
ARDS
77 (28.9%)
128 (40.3%)
0.004
ICU admission
102 (38.64%)
157 (49.68%)
0.008
Mortality
15 (5.68%)
19 (6.01%)
0.866
Mechanical ventilation
70 (26.62%)
114 (36.31%)
0.013
Multiorgan failure
19 (7.04%)
29 (9.21%)
0.382
Liver failure
1 (0.38%)
6 (1.9%)
0.095
Liver injury
132 (49.4%)
186 (59%)
0.018
Borderline elevation (<2× ULN)
95 (35%)
121 (38.4%)
Mild (2–5×)
30 (11.3%)
55 (17.4%)
Moderate(5–15×)
5 (1.87%)
10 (3.1%)
Severe (>15×)
2 (0.7%)
0 (0%)
The continuous variables are normally distributed and expressed as mean ±SD, and were
compared using Student’s t test. normally distributed. The categorical variables were
presented as numbers (percentage) and compared by the chi-square test or Fisher’s
exact test. ARDS, acute respiratory distress syndrome; BMI, body mass index; ICU,
intensive care unit; NAFLD, non alcoholic fatty liver disease; ULN, upper limit of
normal.
Discussion
The presence of NAFLD has been associated with poor outcomes in patients with COVID-19.
1
,
3
,
9
However, our results conflict with these reports. We used similar inclusion criteria
and used the HSI index as a surrogate marker for the presence of NAFLD like Ji et al.
1
We found that NAFLD is an independent predictor of the development of mild to moderate
liver injury, like the authors described. However, when controlled for covariates
in multivariate analysis, NAFLD was not a predictor of mortality, disease severity,
or markers of disease progression in our study. We used the development of ARDS, ICU
admission requirements, and the need for mechanical ventilation as surrogate markers
of disease progression while the authors used the development of tachypnea and the
requirement for supplemental oxygenation as surrogates of disease progression . Our
observation was that in hospitalized patients using tachypnea and need for supplemental
oxygen even once was quite common, so they are probably not the best markers for disease
progression.
The difference in our results could be due to a much larger sample size of patients
with NAFLD from heterogeneous ethnic populations in our study. Although the HSI index
has a high specificity (93%) and positive predictive value (99%), it is affected by
inflammation and potentially could overestimate the prevalence of NAFLD.
10
There is a need to study the outcomes in large scale studies with histologically confirmed
cases of NAFLD and COVID-19.
Conclusion
The presence of NAFLD is an independent predictor of mild to moderate liver injury
in hospitalized patients with COVID-19. However, NAFLD was not an independent predictor
of mortality, disease severity on presentation, or disease progression in patients
with COVID-19.
Financial support
Funding for publications costs are covered by Hamad Medical Corporation.
Authors' contribution
KM and MUK conceived and designed the study, did data analysis, literature review
and manuscript writing. FI, DHA, HSC, FA, PI, KEN GB, MA, KA, SA, YMK did data collection,
data analysis, manuscript writing and literature review. All authors verified the
final version of the study.
Conflict of interest
The authors declare no conflicts of interest that pertain to this work.
Please refer to the accompanying ICMJE disclosure forms for further details.