Gastrointestinal symptoms and RT-PCR in adults with COVID-19: а post-hoc analysis

Introduction

Coronavirus disease (COVID-19) is caused by SARS-CoV-2 and has spread globally since December 2019 [1]. To date, over 6.8 million COVID-19 cases have been reported in Indonesia, with an estimated 162,063 associated deaths¹. COVID-19 manifests with respiratory symptoms (e.g., dyspnea) and gastrointestinal (GI) symptoms such as nausea, abdominal pain, and diarrhea [2][3]. The diagnostic gold standard is reverse transcription polymerase chain reaction (RT-PCR) from respiratory specimens – typically nasopharyngeal swabs – but alternative specimens can be informative [2][3]. Wang et al. reported SARS-CoV-2 positivity in fecal specimens of up to 53.4%, and a higher specificity for gastrointestinal symptoms when using lower GI specimens (anal swabs) at approximately 67.5%, compared to nasopharyngeal swabs [4]. Additional studies show high positivity and viral loads in lower GI specimens, detection among asymptomatic cases, and potential reduction in false negatives; anal swabbing may also limit exposure risk for examiners and help pediatric sampling [5][6]. One analytic parameter, the cycle threshold (Ct), inversely reflects viral load and is often used to contextualize RT-PCR results [5][7].

Materials and methods

A secondary analysis was conducted using de-identified records from three Indonesian hospitals (RSCM, Mitra Keluarga Depok, and Mitra Keluarga Kelapa Gading) collected between July and November 2020, originally included in Abdullah et al. [3]. Among 136 screened cases, 45 had positive RT-PCR results from lower GI specimens (anal swabs). After excluding 8 records with missing Ct values, 37 participants with complete data were included in the final analysis. This dataset represents all eligible positive cases with Ct information from the three participating centers in the Jakarta region.

GI symptoms captured included diarrhea, constipation, nausea, vomiting, and abdominal pain. Ct values were dichotomized: low <25 vs high ≥25. Categorical variables are presented as the number of patients (n) and the corresponding percentage (%). Age is presented both as a mean with standard deviation and as categorical age groups. Association between Ct category and GI symptom presence was assessed with Fisher’s exact test; effect size is presented as prevalence ratio (PR) with 95% confidence interval (CI). Analyses used SPSS v26. (IBM, the USA).

Results

Participant characteristics (Table 1) showed a near-even sex distribution (43.2% male; 56.8% female) with the majority aged 40–60 years (51.4%), followed by <40 years (37.8%) and >60 years (10.8%). Most patients had no recorded comorbidity (67.6%), while 24.3% had fewer than two and 8.1% had two or more. Obesity (body mass index ≥25²) was present in 32.4%. Lifestyle factors included current smoking in 16.2% and former smoking in 8.1%; almost half reported low physical activity and 51.4% reported rarely exercising. 

Ct values were predominantly high: 36/37 (97.3%) fell in the high (≥25) category and only 1/37 (2.7%) in the low (<25) category.

GI symptom prevalence was 59.5% (22/37), with 40.5% (15/37) reporting none (Fig). Among symptomatic patients, nausea was most common (12/22), followed by vomiting (4/22), abdominal pain (4/22), and diarrhea (3/22); constipation was not recorded. 

In the primary analysis, the association between Ct category and the presence of any gastrointestinal symptom was evaluated. Due to sparse expected counts, Fisher’s exact test was applied and revealed no statistically significant association (p = 0.595). The prevalence ratio (PR) comparing high versus low Ct values was 1.048 (95% CI: 0.956–1.148), consistent with a null effect estimate (Table 2). 

Discussion

Testing of lower GI specimens serves as a practical complement to nasopharyngeal sampling and has demonstrated higher detection rates in patients with gastrointestinal manifestations, consistent with previous findings from both Indonesian and international studies [3–5][8][9]. Our descriptive data similarly show that GI symptoms were common among adults with positive lower GI specimens RT-PCR.

Despite biologic plausibility that lower Ct (higher viral load) might align with symptomatology, we did not find an association between Ct category and GI symptoms. Ct values are influenced by numerous pre-analytical and analytical factors – including sampling technique, timing relative to illness, transport/storage, and assay efficiency – so crude categorization may obscure clinically meaningful relationships [5][7]. The very low number of low-Ct observations (n = 1) further limits power to detect differences.

Observed heterogeneity in comorbidities and lifestyle risks such as obesity and smoking could also confound relationships between Ct and symptoms, as these factors are associated with COVID-19 outcomes and severity [10]. Future studies should incorporate multivariable adjustments and standardized collection protocols to reduce noise from these sources.

Comparative evidence suggests average Ct values around the high-20s among patients with GI involvement and variable links to severity, yet these patterns originate largely from nasopharyngeal or fecal sampling rather than lower GI specimens [7][9][11]. Standardized, prospective designs that analyze Ct as a continuous measure and align sampling with symptom timing are likely needed to clarify any true association.

Conclusion

In this secondary analysis of 37 adults with positive RT-PCR results from lower GI specimens at three Indonesian hospitals, GI symptoms were common; however, the Ct category (low <25 vs high ≥25) showed no statistically significant association with the presence of GI symptoms (Fisher’s exact p = 0.595; PR 1.048, 95% CI 0.956–1.148). Given substantial pre-analytical variability and potential confounding by comorbidities and lifestyle factors, Ct values from lower GI specimens should be interpreted cautiously for symptom prognostication. Larger, prospective studies with standardized protocols are warranted.