Nasopharyngeal SARS-CoV-2 viral RNA shedding in patients with diabetes mellitus.

Dear Editor

The paper by Buetti et al. [1] regarding the prolonged SARS-CoV-2 viral shedding in lower respiratory tract samples of critically ill patients presented a fascinating finding. In this Swiss cohort, the presence of diabetes mellitus (DM) in critically ill patients proved to be a risk factor for prolonged SARS-CoV-2 RNA. In a recent work, we assessed clinical and demographic risk factors for prolonged viral RNA shedding and time to achieve the cessation of viral RNA shedding (CVS) in 251 patients with COVID-19 based on nasopharyngeal samples only [2].

We identified 34 (14%) patients with DM (Table 1). These patients were older (60 years [IQR 15.5] vs. 50 years [IQR 28]; p < 0.001) and had a higher frequency of comorbidities such as obesity, coronary artery disease, chronic obstructive pulmonary disease, and heart failure (p < 0.01) when compared to patients without DM. Patients with DM were more frequently hospitalized (61.8 vs. 18.9%; p < 0.001) than patients without DM. Dyspnea, diarrhea, and dizziness were more frequently seen in patients with DM (p < 0.01) at the time of diagnosis. Despite these differences, there was no difference in time to achieve CVS when compared to patients with and without DM (24 days (IQR 11.8) vs. 23 days (IQR 12); p = 0.591). This is best observed in Fig. 1. Interestingly, dizziness was more commonly reported at the time of CVS in patients with DM. DM was not associated with prolonged viral RNA shedding in a univariate regression model (p = 0.49).

Table 1

Clinical characteristics, timing of PCR test, symptoms at diagnosis, and symptoms after cessation of viral shedding of COVID-19 patients with and without diabetes mellitus

Clinical characteristics	Diabetes mellitus (N = 34)	No diabetes mellitus (N = 217)	Total (N = 251)	p value
Demographics
Age, years	60 (15.5)	50 (28)	53 (27)	<0.001a
Male	20 (58.8%)	128 (59%)	148 (59%)	0.986b
Hospitalized	21 (61.8%)	41 (18.9%)	62 (24.7%)	<0.001b
Comorbidities
Coronary artery disease	25 (73.5%)	45 (20.7%)	70 (27.9%)	<0.001b
Obesity	19 (55.9%)	56 (25.8%)	75 (29.9%)	<0.001b
Chronic obstructive pulmonary disease	9 (26.5%)	13 (6.0%)	22 (8.8%)	<0.001b
Chronic kidney disease	5 (14.7%)	15 (6.9%)	20 (8.0%)	0.119b
Heart failure	4 (11.8%)	5 (2.3%)	9 (3.6%)	0.006b
Asthma	3 (8.8%)	43 (19.8%)	46 (18.3%)	0.123b
Use of anti-neoplastic chemotherapy, immunomodulators, or immunosuppressive drugs	3 (8.8%)	13 (6.0%)	16 (6.4%)	0.53b
Initial symptoms
Cough	30 (88.2%)	181 (83.4%)	211 (84.1%)	0.475b
Dyspnea	26 (76.5%)	107 (49.3%)	133 (53.0%)	0.003b
Fever (T > 38.5 °C)	16 (47.1%)	110 (50.7%)	126 (50.2%)	0.694b
Subjective fever	8 (23.5%)	43 (19.8%)	51 (20.3%)	0.617b
Chills	19 (55.9%)	104 (47.9%)	123 (49.0%)	0.388b
Shivering	0 (0.0%)	1 (0.5%)	1 (0.4%)	0.692b
Muscle pain	27 (79.4%)	139 (64.1%)	166 (66.1%)	0.079b
Diarrhea	17 (50.0%)	55 (25.3%)	72 (28.7%)	0.003b
Headache	12 (35.3%)	82 (37.8%)	94 (37.5%)	0.78b
Sore throat	15 (44.1%)	94 (43.3%)	109 (43.4%)	0.93b
Ageusia	7 (20.6%)	44 (20.3%)	51 (20.3%)	0.966b
Anosmia	7 (20.6%)	48 (22.1%)	55 (21.9%)	0.841b
Dizziness	14 (41.2%)	43 (19.8%)	57 (22.7%)	0.006b
Timing to PCR tests
Days from symptoms to positive PCR (IQR)	2 (5.5)	3 (6)	3 (6)	0.509a
Days from symptoms to negative PCR (IQR)	24 (11.8)	23 (12)	23 (12)	0.591a
Symptoms at time of negative PCR
Cough	7 (20.6%)	55 (25.3%)	62 (24.7%)	0.55b
Dyspnea	0 (0.0%)	2 (0.9%)	2 (0.8%)	0.574b
Chills	0 (0.0%)	1 (0.5%)	1 (0.4%)	0.692b
Muscle pain	11 (32.4%)	65 (30.0%)	76 (30.3%)	0.777b
Diarrhea	1 (2.9%)	1 (0.5%)	2 (0.8%)	0.13b
Headache	1 (2.9%)	13 (6.0%)	14 (5.6%)	0.471b
Sore throat	7 (20.6%)	42 (19.4%)	49 (19.5%)	0.866b
Ageusia	6 (17.6%)	41 (18.9%)	47 (18.7%)	0.862b
Anosmia	6 (17.6%)	45 (20.7%)	51 (20.3%)	0.677b
None	9 (26.5%)	56 (25.8%)	65 (25.9%)	0.934b
Unknown	1 (2.9%)	10 (4.6%)	11 (4.4%)	0.659b
Dizziness	6 (17.6%)	12 (5.5%)	18 (7.2%)	0.011b

aKruskal–Wallis rank sum test

bPearson’s Chi-squared test

Bold values indicate statistical significance

Fig. 1

Time to cessation of SARS-CoV-2 viral RNA shedding in people with and without diabetes mellitus

In contrast to Buetti et al., our findings suggest that patients with DM do not show prolonged RNA viral shedding based on nasopharyngeal sampling despite older age and a higher frequency of comorbidities. DM has been described by Wang et al. [3] as a risk factor for viral RNA detection in both nasopharyngeal and sputum samples. Nevertheless, the presence of DM was not associated with prolonged shedding in their sample either. The determination of whether prolonged viral RNA shedding in lower respiratory samples represents a particular phenomenon of patients with DM or other immunocompromising states requires additional research and collaboration that combines different types of samples, timelines, and clinical characteristics.