Gastrointestinal symptoms in 609 Japanese patients with COVID-19: a single-center retrospective study

Abstract Background GI symptoms are common in acute COVID-19 patients. This study aimed to characterize the GI symptoms occurring in Japanese COVID-19 patients. Methods This retrospective single-center cohort study included 751 hospitalized acute COVID-19 patients. The primary outcomes were the frequency and severity of GI symptoms. The secondary outcomes included the association between COVID-19 severity and GI symptoms and the timing of GI symptom onset. Results After exclusion, the data of 609 patients were analyzed. The median age was 62 years, and 55% were male. The median time from initial symptom onset to admission was five days. On admission, 92% of the patients had fever, 35.1% had fatigue, 75% had respiratory symptoms, and 75% had pneumonia. The sample included patients with mild (19%), moderate (59%), and severe COVID-19 (22%). A total of 218 patients (36%) had GI symptoms, of which 93% were classified as grade 1/2; 170 patients had both respiratory and GI symptoms. Diarrhea was the most frequent GI symptom, occurring in 170 patients, followed by anorexia in 73 patients and nausea/vomiting in 36 patients, and abdominal pain in 8 patients. There was no significant relationship between COVID-19 severity and GI symptoms. Among COVID-19 patients with both GI and respiratory symptoms, 48% had respiratory symptoms preceding GI symptoms, 25% had GI symptoms preceding respiratory symptoms and 27% had a simultaneous onset of respiratory and GI symptoms. Conclusion Thirty-six percent of the Japanese COVID-19 patients had GI symptoms; diarrhea was the most frequent GI symptom but did not predict severe COVID-19.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), similar to the SARS-CoV-1 virus that spread in 2003, has been reported to infect cells via angiotensin-converting enzyme (ACE) 2 receptors [1].The ACE2 receptor is most abundantly expressed in the small intestine; however, its association with gastrointestinal (GI) symptoms is unclear.GI symptoms (e.g.diarrhea, anorexia, nausea/vomiting, and abdominal pain) are common in coronavirus disease 2019 .Therefore, it is possible that the high expression of ACE-2 receptors in the gastrointestinal tract is responsible for the GI symptoms observed in COVID-19 patients [2].To date, however, there have been no details on GI symptoms in Japanese patients with acute COVID-19 because of insufficient data on Japanese COVID-19 cases.
In Japan, pneumonia caused by COVID-19 was first reported in Kanagawa Prefecture on January 15, 2020.The first wave of the COVID-19 pandemic occurred in April 2020.Subsequently, COVID-19 peaked in early August (second wave), followed by a peak in early January 2021 (third wave), a peak in early May (fourth wave caused mainly by α strains), a peak in late August (fifth wave caused mostly by δ strains), and a peak in January 2022 (6th wave), when the Omicron strains rapidly replaced the δ strains.As of August 2022, the seventh wave of Omicron strains recorded the highest number of patients [1].The mutant strains differed depending on the epidemic period; in the seventh wave, the BA.5 strain, which is more infectious, is believed to have replaced the previous strain.However, the number of patients with severe disease requiring mechanical ventilation has been relatively small since the sixth wave [1].
GI symptoms are now well recognized as a potential feature of acute COVID-19 [3,4].However, active research is still in progress to clarify the details of these GI symptoms, including their severity and the relationship between their severity or onset timing and the severity of acute COVID-19.The clinical manifestations of acute COVID-19 may differ among different populations of different ethnicities [5].To date, however, there have been few reports on GI symptoms in Japanese patients with acute COVID-19.In this study, we clarified the characteristics of GI symptoms in Japanese patients with acute COVID-19, including the relationship of COVID-19 severity with GI symptoms and the timing of such symptoms in the Japanese population.

Patients and methods
This retrospective single-center cohort study examined 751 consecutive hospitalized patients with laboratory-confirmed COVID-19 between February 20, 2020, and August 10, 2021.Our public hospital is located in the northern area of Osaka Prefecture, specifically in Toyonaka City.The city has a population of approximately 400,000.Our hospital has been designated primarily as a general hospital for treating patients with moderate to severe COVID-19.We have admitted acute COVID-19 patients since the first wave of the pandemic in Japan [2,[6][7][8][9][10].
We judged the worst severity during the hospital stay and followed a treatment strategy for hospitalized patients with acute COVID-19 based on the guidance of the Japanese Ministry of Health, Labor, and Welfare available on the COVID-19 website, which had been available since March 7, 2020; the guidance has been updated since then, and the latest version (version 8.0), was released in July 2022 (in Japanese) [1].This guidance divides severity into four groups: mild disease without respiratory symptoms (SpO 2 ≥96%), moderate I disease with breathing difficulties, pneumonia without respiratory failure (93%<SpO 2 <96%), moderate II disease requiring oxygen (SpO 2 ≤93%), and severe disease requiring intensive care unit treatment or mechanical ventilation.This Japanese severity classification is generally equivalent to the WHO severity definitions: mild, moderate, severe, and critical diseases, respectively [11], although there are some differences in the definitions.The treatment strategies have been updated according to advancements in therapeutics [1].We treated hospitalized patients with acute COVID-19 by consulting with the respiratory specialists at our hospital (Dr.Morimura O, and Dr. Abe K and others).
We collected electronic medical records from our hospital (MegaOak Online Imaging System, NEC, Japan) to review the clinical characteristics of the acute COVID-19 patients with GI symptoms.Clinical and laboratory data were collected from all the patients at hospital admission.We treated patients using the clinical pathway for acute COVID-19 patients.Specifically, treatment was fundamentally conservative, but the following therapeutic agents were also used in our hospital: chloroquine, favipiravir, ciclesonide, and Decadron in waves 1 to 3 and favipiravir, dexamethasone, heparin, and remdesivir in waves 4 and 5, depending on the severity of the case [1].We prospectively used medical templates to assess all the hospitalized acute COVID-19 patients in our infectious disease unit to avoid missing data.Each patient's height and body weight data were obtained through medical interviews to reduce patient contact.We performed a portable chest radiograph for the initial evaluation of pulmonary complications at admission.
This study was conducted in accordance with the Declaration of Helsinki, and approval was obtained from the Institutional Review Board of Toyonaka Municipal Hospital (No. 2022-03-05).The requirement for informed consent was waived using the opt-out method on our hospital website.

Outcomes
The primary outcomes were the frequency and maximum severity of GI symptoms noted throughout the clinical course.GI symptoms were classified into four groups: diarrhea, anorexia, nausea/vomiting, and abdominal pain.The secondary outcomes were the associations of the severity of acute COVID-19 with GI symptoms and the onset timing of such symptoms.If the patient was taking a medication that was incontrovertibly associated with GI symptoms prior to the onset of GI symptoms, the medication was discontinued and the causation of the symptoms was evaluated.In addition, patients with diarrhea symptoms were tested for Clostridioides difficile infection.
We evaluated the severity of GI symptoms in the patients according to the National Cancer Institute Common Toxicity Criteria (NCI-CTCAE version 5.0) [12].The CTCAE defines grades 1-5 according to the following principles, with a separate description of the severity of each symptom.Grade 1 is mild and does not require treatment.Grade 2 is moderate, requiring minimal/local/noninvasive treatment.Grade 3 is severe or medically significant but not immediately life-threatening.Grade 4 is life-threatening and requires emergency treatment.Grade 5 is death.

Statistical analysis
The medians and interquartile ranges (IQRs) are reported for continuous variables.Categorical variables are summarized as frequencies (percentages) with 95% confidential intervals (CIs).We used the values of the potential risk factors affecting disease progression measured at admission; however, there were some missing data.All reported p values were two-sided, and statistical significance was set at p < 0.05.Statistical analyses were performed using JMP statistical software (ver.14.3, SAS Institute, Inc., Cary, NC, USA).

Results
We evaluated a total of 751 hospitalized patients with acute COVID-19 during the study period.Among them, we excluded 19 asymptomatic patients, 29 pediatric patients and 13 infants, 16 non-Japanese patients, and 65 patients who were readmitted or transferred from other hospitals.A total of 609 patients with symptomatic acute COVID-19 were eligible for enrollment (Figure 1).
The clinical characteristics of the enrolled acute COVID-19 patients are summarized in Table 1.The median age was 62 years (IQR 45.5, 78), and 336 patients (55.1%) were male.
During the hospital stay, oxygen was administered to 371 (60.9%) patients, mechanical ventilation was required in 87 (14.3%) patients, and the mortality rate was 5.3%.Of the 32 deaths, all were caused by respiratory failure due to COVID-19 pneumonia, except for one death caused by sepsis.The severity of the patients with COVID-19 when considering their most aggravated condition after admission was as follows: mild in 19% of the patients, moderate I in 20%, moderate II in 39%, and severe in 22% according to the COVID-19 severity classification of the Japanese Ministry of Health, Labor, and Welfare.The median time from onset to admission was five days (IQR: 2, 8 days), and the median length of hospital stay was ten days (IQR 7,15) (Table 2).There were no significant differences in clinical background between the patients with and without GI symptoms (Supplementary Table ).

GI symptoms: severity and timing
In total, 218 patients (35.8%) with 287 GI symptoms (diarrhea, anorexia, nausea/vomiting, and abdominal pain) were included in this study.Some patients had more than one of these symptoms.Table 3 shows the frequency and severity of the GI symptoms evaluated by the CTCAE version 5.In most of the patients, the GI symptoms were grade 1 or 2 except for two patients with grade 3 diarrhea and 18 patients with grade 3 anorexia; however, none of the patients had grade 4 disease or higher.Diarrhea was the most common GI symptom, occurring in 170 patients (28%, 95% CI: 24.5-31.7%),followed by anorexia in 73 (12%, 95% CI: 9.6-14.8%),nausea/vomiting in 36 (5.9%, 95% CI: 4.3-8.1%),and abdominal pain in 8 patients (1.3%, 95% CI: 0.7-2.6%).
Comparing the 475 patients with mild to moderate disease and the 134 with severe disease during their hospital stay, there was no correlation between the presence of diarrhea, including all grades, and the severity of COVID-19 (131 patients vs. 39, p = 0.994), and the same was true for other GI symptoms: anorexia (54 vs. 19, p = 0.419), nausea/vomiting (29 vs. 7, p = 0.184), and abdominal pain (8 vs. 0, p = 0.127).Furthermore, we analyzed the associations between the severity of COVID-19 and the grade of each symptom and found that there was no correlation between the severity of GI symptoms and that of COVID-19: diarrhea (p = 0.400), anorexia (p = 0.419), nausea/vomiting (p = 0.237), and abdominal pain (p = 1.00) (Table 4).
Table 5 shows days from any initial symptom to the development of each symptom.Respiratory symptoms had the earliest onset, and gastrointestinal symptoms occurred on a median of 3 days.Table 6 shows the timing of GI and respiratory symptom onset among patients with both types of symptoms.A total of 170 patients experienced both respiratory and GI symptoms simultaneously or at different times during the clinical course.Among them, the respiratory symptoms preceded the onset of the GI symptoms in 81 patients (48%), and the GI symptoms preceded the respiratory symptoms in 43 patients (25%).In 46 patients (27%), the respiratory and GI symptoms appeared simultaneously.We found no correlation between the timing of GI and respiratory symptom onset and the severity of COVID-19 (Table 6).

Discussion
GI symptoms are common in COVID-2019, but the underlying mechanisms are not fully understood.A meta-analysis of 60 studies with 4243 patients, excluding the Japanese population, showed that GI symptoms were seen in 17.6% of patients (95% CI, 12.3-24.5).Patients with severe disease were more likely to have GI disease than those without (17.1% vs. 11.8%)[3].However, significant heterogeneity was noted among the studies because only three studies included more than 200 patients with COVID-19, and among the 60 studies, 53 studies (88%) were performed in China.Another large systematic review, including 78 studies with 12,797 patients with COVID-19 [4] and 104 Japanese individuals (0.8%) [13], reported that 17% of patients had GI symptoms, 12.4% had diarrhea, 9.0% had nausea/vomiting, 22.3% had anorexia, and 6.2% had abdominal pain.There were no differences in the mortality rates between the patients with and without GI symptoms.The percent of patients with GI symptoms of the two meta-analyses were similar at approximately 17%.In this study, we found that among the 609 Japanese patients with mild-to-severe COVID-19, 36% had GI symptoms, which was approximately twice as prevalent in  13) 0 (0) Grade 3 0 (0) 0 (0) 0 (0) the Japanese patients as in non-Japanese patients.This is consistent with the fact that a comparison of studies from China and non-Chinese countries showed that GI symptoms are more prevalent outside of China [14].In the present study, diarrhea was the most common GI symptom (28%), followed by anorexia (12%), nausea/vomiting (5.9%), and abdominal pain (1.3%).In addition, GI symptoms did not predict severe COVID-19.Compared with a previous large systematic review, the Japanese population seemed more susceptible to GI symptoms from COVID-19 but experienced less abdominal pain.Most COVID-19 patients with GI symptoms were found to have grade 1 or 2 disease, and few reports have evaluated the severity of GI symptoms in patients with COVID-19.Our study may have included patients with milder symptoms than those in the previous meta-analysis.Acute COVID-19 causes thrombotic tendencies due to injury to vascular endothelial cells.Thus, microthrombosis/ embolization may cause ischemic enterocolitis, resulting in abdominal pain.The complication rate of thrombosis/embolism increases with the severity of the disease; therefore, abdominal pain is reported to be an indicator of the severity of the disease [15,16].No significant relationship between the severity of COVID-19 and the degree of GI symptoms was observed in this study.The incidence of abdominal pain was relatively low (1.3%) in this study compared with a previous systematic review in which 6.2% of the 12,797 patients had abdominal pain [4].Our cohort showed that approximately 80% of the patients had mild to moderate COVID-19, and most of the GI symptoms were mild (grade 1/2: 93%).No patients with bloody diarrhea were observed in this Japanese cohort.A rapid review by the American Gastroenterological Association in 2020 showed that diarrhea occurs slightly more frequently in hospitalized patients than in outpatients [14].Since we included only hospitalized patients, the incidence of GI symptoms may have been higher.Additionally, heparin was administered to patients with elevated D-dimer levels in the early phase; this treatment may have prevented thrombus formation and intestinal ischemia, thereby reducing the severity of GI symptoms such as abdominal pain.Another possible cause is the difference in ethnicity between this and other reported populations .
Regarding the timing of GI symptoms, diarrhea was reported to be one of the initial symptoms of acute COVID-19 and appeared 1-2 days before the respiratory symptoms and fever in 14 (10%) of 138 hospitalized patients [17].To date, there are, however, few reports evaluating the timing of GI symptoms compared to respiratory symptoms, and this is the strength of this study.Our results revealed that 28% (170/609) of the hospitalized patients had respiratory and GI symptoms simultaneously or at different times.Among these 170 patients, approximately half of them (52%) initially had GI symptoms, and there was a similar frequency of patients who had GI symptoms preceding respiratory symptoms (25%) and respiratory and GI symptoms simultaneously (27%) (Table 3).
The current study has several limitations.First, this was a single-center retrospective study.However, this study included 609 Japanese acute COVID-19 patients, which was a relatively high number of patients.Second, we did not evaluate outpatients with mild COVID-19 symptoms.In this study, we analyzed only patients who required hospitalization, which may cause an overestimation of the severity of GI symptoms because diarrhea has been reported to occur slightly more frequently in hospitalized patients than in outpatients [14].Third, it is possible that the diarrhea that appeared in patients after hospitalization could have been caused by treatment or medication, including antibiotics, PPIs and ACE inhibitors/ARBs [18].Gastric acid has defense mechanisms against pathogens such as viruses, preventing their entry into the small intestine.PPI increases the pH of gastric acid and inhibits viral inactivation [19,20].ACE2 receptors are widely expressed throughout the intestinal tract.SARS-CoV-2 binds to ACE2 receptors to invade and proliferate within cells, causing inflammation of the intestinal tract and GI symptoms [21,22].ACE2 inhibitors prevent the interaction between receptor connection domains and viruses [23].Moreover, ACE2 is considered to play an important role in controlling inflammation and diarrhea [24], and ACE2 inhibitors/ARBs could affect these functions.However, there was no differences in GI symptoms between the patients with and without the use of PPIs or ACE2 inhibitors/ARBs in this study (Supplementary Table 1).In addition, we performed univariate and multivariate logistic analyses, but we did not find a significant relationship between GI symptoms and patient characteristics, including comorbidities and medications, except for age group (elderly or non-elderly) (Supplementary Table 2).Although we did not exclude the effects of medication on COVID-19 in the present study, the median time from the initial onset to admission was longer than that of GI symptom onset (5 days IQR: 2, 8 vs. 3 days IQR: 0,3); therefore, most of the symptoms appeared before admission and were considered to be caused by COVID-19 itself.Fourth, we did not evaluate the duration of GI symptom with long COVID-19 because of not enough available data after discharge.Finally, which strain infected the patients was not determined in this study.It was also unknown how many times the patients were vaccinated or which vaccine the patients had, including which company manufactured the vaccine for COVID-19.The possibility that the frequency and severity of symptoms could differ depending on the type of virus and vaccination status cannot be ruled out.
In conclusion, 36% of Japanese hospitalized patients with acute COVID-19 had GI symptoms, with diarrhea being the most prevalent of these symptoms.However, approximately 90% of the patients with GI symptoms showed them at only a mild level, and these symptoms did not predict severe COVID-19.

Table 1 .
characteristics and clinical course of 609 patients with covid-19.

Table 3 .
frequency and severity of Gi symptoms according to the national cancer institute common toxicity criteria.

Table 4 .
relationship between the severity of covid-19 and the degree of Gi symptoms.

Table 5 .
timing of Gi and respiratory symptom onset.

Table 6 .
timing of Gi and respiratory symptom onset and the severity of covid-19.