Gastrointestinal symptoms in coronavirus disease 2019: a cross-sectional study in Wuhan, China

Background To investigate the clinical symptoms of coronavirus disease 2019 (COVID-19), particularly the prevalence, time of symptom onset, and duration of gastrointestinal (GI) symptoms. Methods This was a cross-sectional study using paper questionnaires. COVID-19 patients in a temporary hospital in Wuhan voluntarily completed surveys collecting data on COVID-19 symptoms and investigation results.


Abstract Background
To investigate the clinical symptoms of coronavirus disease 2019 , particularly the prevalence, time of symptom onset, and duration of gastrointestinal (GI) symptoms.

Methods
This was a cross-sectional study using paper questionnaires. COVID-19 patients in a temporary hospital in Wuhan voluntarily completed surveys collecting data on COVID-19 symptoms and investigation results.

Conclusions
In our cohort of patients, GI symptoms were common in COVID-19, occurred mostly during the middle stage of the disease, and lasted for a short duration. GI symptoms may not be associated with COVID-19 related treatment.

Background
Coronavirus disease 2019 (COVID-19) has recently become a global pandemic [1]. The disease was rst identi ed in the Wuhan City of China in December 2019 [2]. At the time of this writing on 11 March 2020, a total of 113,702 con rmed cases and 4,012 deaths have been reported globally [3]. Fever and cough are the most common clinical symptom [4][5][6]. Gastrointestinal symptoms (GI symptoms) such as nausea, vomiting, and diarrhoea are often seen in coronavirus infections such as severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS) [7][8]. However, early studies reported low prevalence of typical GI symptoms, with the prevalence of diarrhoea in the range of 1-3.8% [4][5][6]. On the contrary, with current growing attention to GI symptoms, the latest published studies have shown that GI symptoms are also common in COVID-19, and the prevalence of diarrhoea is between 12.0-29.2% [9][10][11][12]. However, the time of symptom onset and duration of GI symptoms have not been commonly investigated in recent studies [9,12]. We recently worked in a temporary hospital and conducted a detailed cross-sectional study of the clinical symptoms of patients with COVID-19 to further recognize the clinical characteristics of GI symptoms.

Methodlogy Ethical considerations
This study was reviewed by the ethics committee of Renmin Hospital of Wuhan University (Ethics Approval No.: WDRY2020-K033) and written informed consent from patients was waived. Only a verbal explanation was required, and verbal informed consent was obtained from patients.

Study participants
From February 18 to March 2, 2020, patients with con rmed COVID-19 according to WHO interim guidelines [13]. and who were ≥ 16 years of age, were recruited from Wuchang mobile cabin hospital. Wuchang mobile cabin hospital was a temporary hospital facility designed speci cally for the care of patients with con rmed COVID-19. Our exclusion criteria were (i) incomplete data, (ii) cases with GI symptoms resulted from new medication and (iii) individuals below the stipulated age limit. A total of 240 patients were invited, and 222 responses were received (response rate 92.5%). Among the patients, 10 responses met the exclusion criteria (6 cases with GI symptoms resulted from new medication and 4 cases below the stipulated age limit) and these were not included in the data analysis. The patients themselves completed a form that collected general demographic information, current medical history, past medical history, date of rst positive severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) nucleic acid test results, chest computed tomography (CT) results, and COVID-19 contact history. COVID-19-related clinical symptoms were listed with tick boxes in the form. The study participants were asked to check the boxes of their recent abnormal symptoms and write down the speci c date of symptom onset and duration of each symptom. On handing over the questionnaire (supplementary material pp1-2), we checked the data completion. If there were incomplete data, we required the patients to ll in the blank.
So, there was no missing data. All patients in this temporary mobile hospital were long-time residents of Wuhan and tested positive for SARS-CoV-2 nucleic acid by reverse transcription polymerase chain reaction (RT-PCR) from a nasopharyngeal swab before admission. After patients completed the hardcopy forms, photographs were taken, and the data were uploaded onto the network. Before the temporary mobile hospital was closed, the hardcopy forms were destroyed as medical waste.
Patient and Public Involvement It was not appropriate or possible to involve patients or the public in the design, or conduct, or reporting, or dissemination plans of our research.

Statistical analyses
Symptom duration was de ned as the interval from the onset of the symptom to resolution or investigation. Time of symptom onset refers to the duration from onset of the initial symptom of COVID-19. High fever was de ned as a temperature ≥ 39°C. The de nition of diarrhoea was the passing of loose stools ≥ 3 times per day, and the two types of abnormal loose stools included mushy stool (type 6) and watery stool (type 7) according to Bristol stool chart. A history of chronic disease was de ned as other diseases that persisted for three months before COVID-19, such as hypertension, diabetes mellitus, liver disease, and respiratory disease. The Kolmogorov-Smimov test was used to determine whether the data were normally distributed. Normally distributed variables were expressed as the mean and standard deviation, while non-normally distributed variables were expressed as the median and interquartile range (IQR). Categorical variables were summarized as the counts and percentages in each category. We grouped patients into diarrhoea and non-diarrhoea groups. Wilcoxon rank-sum tests were applied to nonnormally continuous variables, t tests were applied to normally distributed variables, and chi-square tests were used for categorical variables as appropriate. Fisher's chi-square tests were applied when the expected values in any of the cells of a contingency table were below 5. A two-sided P of < 0.05 was considered statistically signi cant. SPSS (V.26.0) was used for all analyses.
Additionally, 66.3% (140/212) patients had GI symptoms, including anorexia in 88, diarrhoea in 93, and nausea and vomiting in 44. When anorexia was excluded from the analysis (since it is less speci c), there were 111 total cases (52.3%) in which patients presented with a speci c GI symptom. Furthermore, 3.3% (7/212) of the patients developed conjunctival congestion. In the patients with diarrhoea, 86.0% (80/93) had mushy stools and 14% (13/93) had watery stools. There were 42.3%(90/212) patients who developed all types of the three common COVID-19 symptoms such as, u-like symptoms, pneumonia and GI symptoms.
Cough was the most persistent symptom, whereas conjunctival congestion had the shortest duration. Patients experienced a median duration of 4.00 (IQR: 2.00-8.75) days with diarrhoea, 5.00 (IQR: 3.00-10.00) days with anorexia, and 6.00 (IQR: 4.00-10.00) days with nausea and vomiting, which were shorter than the durations of fever and pneumonia symptoms (Fig. 3).
The patients with diarrhoea were younger (45.85 ± 13.28 years vs 50.61 ± 12.82 years, OR 0.964, 95%CI (0.942-0.987), P = 0.009) and were more likely to have abnormal chest CT results (95.7% vs 82.4%, OR 8.729, 95%CI (2.376-32.064), P = 0.001). There was no signi cant difference in in-hospital stay, duration from onset to hospitalization and antibiotic use between patients with and without diarrhoea. The patients with diarrhoea had a higher percentage to have the symptoms of fever, cough, sputum, shortness of breath, headache, myalgia, fatigue, bitter taste in mouth, dry mouth, excessive sweating, and nausea or vomiting than those without diarrhoea(P < 0.05) ( Table 1).

Discussion
Coronavirus is a large virus family that can cause various conditions, ranging from the common cold to severe infectious diseases such as SARS and MERS [14]. COVID-19 is a newly discovered coronavirus infection in humans, and is currently a pandemic [1][2]. The most common clinical symptoms associated with coronavirus infection are fever and cough [8,14]. The results of this study have further supported previous published data that fever and cough are the most common clinical symptoms of COVID-19 [4][5][6]. Studies have found that GI symptoms are one of the common clinical symptoms in SARS and MERS [7][8]14]. With growing attention to GI symptoms of COVID-19, the latest studies have found that diarrhoea is also common in COVID-19, with estimates ranging 12-29.2% [9][10][11]. Pan et al. reported that aside from respiratory symptoms, diarrhoea may be an initial or only symptom of COVID-19 [11]. In this study, we also found that a minority of patients, approximately 10.4%, had an initial symptom of diarrhoea. In contrast to the results of previous studies, this study found that GI symptoms including anorexia, diarrhoea, nausea, and vomiting are more common in COVID-19; especially diarrhoea, with a prevalence of 43.8%. This leads to the following question: why are there major differences between the results of this study and previous clinical investigations? First, the data sources of these studies vary. The reported information in the other studies were obtained from medical records [4][5][6][9][10][11]. whereas the data in this study was provided by the patients themselves. Second, the record time of the data are different. The results of this study found that diarrhoea mostly occurred 5 days after the development of initial symptoms; therefore, patients may not have developed GI symptoms such as diarrhoea, nausea, and vomiting in the other studies. In this study, the average time from symptom onset to admission was 16.00 ± 9.87 days, which far exceeds the duration, 7-8.1 days, in the other studies [11,15]. In addition, 79.6%(74/93) of patients experienced diarrhoea prior to admission. Diarrhoea disappeared at the time of investigation in 81.7% (75/93) of patients.
Although some studies revealed that GI symptoms, other than respiratory symptoms, may be the initial symptom in a minority of patients with COVID-19 [10][11]16], there were no studies with a large population that showed when GI symptoms occurred. In this study, we found that most patients developed anorexia at 2.00 (IQR: 0-6.00) days, nausea and vomiting at 2.00 (IQR: 0-9.00) days, and diarrhoea at 5.00 (IQR: 0.25-11.00) days after the onset of initial symptoms. The time of symptom onset of diarrhoea in this study was the same as the rst US case reported [17]. Furthermore, we found that ulike symptoms, such as fever, myalgia, headache, and cough were the initial symptoms of COVID-19 in most patients.
Furthermore, we found that, in contrast to the longer median duration of cough and sputum, diarrhoea in COVID-19 patients only lasted for a median duration of 4.00 (IQR: 2.00-8.75) days, which was the same as the study from Jin et al [12]. Additionally, most diarrhoea occurred as mushy stools instead of watery diarrhoea in SARS [7]. Cholankeril et al [9]. reported that the median duration of GI symptoms was 1 day (IQR: 0-4), which was signi cantly shorter than the duration of respiratory symptoms. Therefore, diarrhoea may be overlooked by clinicians and patients during COVID-19.
Pan et al [11]. reported that the patients with GI symptoms had a longer time from symptom onset to hospitalization compared to those without GI symptoms. However, our study did not nd this difference, which was probably related to the fact that disease severity of recruited patients was mild in this study. In addition, we found that the patients with diarrhoea were more likely to have u-like symptoms, pneumonia, and other GI symptoms, and there were 42.3%(90/212) patients who developed all types of the three common COVID-19 symptoms such as, u-like symptoms, pneumonia and GI symptoms in this study, which further supported developing knowledge that u-like symptoms, pneumonia, and GI symptoms are common symptoms of COVID-19. Pan et al [11] also found 45% patients with COVID-19 developed both pneumonia and GI symptoms. As for why the patients with diarrhoea were younger, it was not clear. It is possible that there were unknown confounding factors that in uenced the recent results.
Why is it important to pay attention to GI symptoms such as diarrhoea, nausea, and vomiting? First, vomiting and diarrhoea affect nutrient absorption in patients and therefore may affect disease outcomes.
Our study found that diarrhoea is signi cantly correlated with fatigue(P < 0.001), suggesting that diarrhoea may decrease patients' quality of life. Some studies have stated the patients with GI symptoms had a more critical type than those without GI symptoms [10][11]18]. In addition, two studies isolated viable SARS-CoV-2 virus from stool samples of COVID-19 patients and detected SARS-CoV-2 RNA in the stomach, duodenum, and rectum specimens [10,19]. Xiao et al. [19] also observed that viral host receptor angiotensin-converting enzyme receptor 2 (ACE2) stained positive primarily in the cytoplasm of gastrointestinal epithelial cells, which highly supports the possibility that SARS-CoV-2 can enter host cells in the digestive tract. Gu et al. [20] and Ng et al. [21] proposed that the GI tract may be one of the target organs of SARS-CoV-2, and fecal-oral transmission may occur in COVID-19. Our study showed that GI symptoms such as diarrhoea, nausea, and vomiting are rather common in COVID-19, suggesting that the possibility of fecal-oral transmission is high. Finally, understanding that GI symptoms typically occur in the early and middle phase of COVID-19 infection, and that they are self-limiting and short-lived, may reduce unnecessary invasive investigations such as diagnostic endoscopy and decrease the infection risk of health providers.
There are several limitations in this study. First, as the subjects of this study were patients with mild cases in a temporary hospital, we were unable to determine whether the same GI characteristics are present in severe COVID-19 patients. Second, there could have been selection bias because this was a voluntary survey. Third, because most patients in this study took medications before diarrhoea in the study, we cannot exclude the in uence of medications on GI symptoms. However, we think the possibility that medications caused GI symptoms is small. This is because more patients took medications, in particular antibiotic, during hospitalization than before hospitalization, yet the prevalence of diarrhoea didn't markedly increase after hospitalization. At the same time, Wan Y et al. [18] stated they didn't observe a correlation between diarrhoea and use of antibiotics or antiviral medicine. Forth, limited by the lack of medical facilities in the temporary hospital, there were no laboratory data to reveal the detailed clinical characteristics of COVID-19, and we also failed to exclude the possibility that diarrhoea resulted from other germs. Besides the above, among the patients with diarrhoea, we failed to test positive SARS-CoV-2 RNA in stool, so the cause of diarrhoea in COVID-19 is probable multifactorial [22]..

Conclusions
In summary, fever and cough are the most common initial symptoms of COVID-19, but GI symptoms are also common, primarily occur during the middle stage of the disease, and lasts for a short period of time. GI symptoms may not be associated with COVID-19 related treatment.
Declarations Figure 1 Prevalence of clinical symptoms among COVID-19 patients Figure 2