The multi-center cohort study (K2-study of the Canton of St. Gallen, Switzerland) as well as the control cohort (G2-study of the Canton of St. Gallen, Switzerland) were registered in the Swiss COVID-19 database¹ and approved by the regional ethics committee (ID: 2020-00941 and ID: 2020-01063). We reached out to potential participants through the distribution of flyers to all general practitioners in the Canton of St. Gallen, through a call for participation in social media and through the identification through the public health department database. The criteria for participation in the study were a positive SARS-CoV-2 PCR result (or, for the control group, a negative PCR result in addition to a negative SARS-CoV-2 serology), informed consent and the ability to contact and visit outpatient test centers on their own. Upon inclusion into the study, a questionnaire was to be completed. The control cohort (n = 115) was composed of participants with COVID-19 like symptoms but with a negative PCR result from a nasopharyngeal swab and with negative SARS-CoV-2 serology [anti-SARS-CoV-2 (spike) IgA/IgG]. PCR positive participants than had multiple study visits with blood draws. The first cycle of serial blood draws (blood draws every week in the first month and then one blood draw after another 4 weeks in the second month, a total of 5 visits per cycle) in the PCR-positive cohort (n = 278, entire cohort) took place between April 2020 and July 2020. Participants were offered to repeat cycles of repeated blood draws later on. Fifty-four of the individuals repeatedly participated in up to three study visit cycles of blood draws at various time points (long-term sub cohort), effectively providing detailed longitudinal data over a duration of more than 1 year. Detailed cohort characteristics as well as co-morbidities are described in Table 1.

At ZLM, swabs for PCR (Copan ESwap^®) were collected from the nasopharynx. Samples were extracted using Molgen PurePrep Extraction Kit and the IDEAL96 extraction robot, followed by RT-PCR targeting sequences of the SARS-CoV-2 E-gene, based on published protocols (18, 19). Blood draws were obtained from an antecubital vein into vacuum tubes (BD 367955). Serum was immediately used or stored at −80°C until further use. High sensitive troponin I was measured with the (Beckman Coulter Switzerland, Nyon, Switzerland) Access high sensitive troponin I assay (limit of detection and limit of quantification both at 2.3 ng/L) on a UniCel DxI 800 (Beckman Coulter Switzerland, Nyon, Switzerland). All determinations were performed with the same reagent batch. Creatinine and alanine aminotransferase (ALT) levels were measured on an AU 5800 (Beckman Coulter Switzerland, Nyon, Switzerland). D-Dimer (D-Dimer HS 500, IL) was measured on an ACL TOP 750 (II, AxonLab, Baden, Switzerland) and brain natriuretic peptide (Quidel Ireland Ltd., Galway Ireland) was measured on a UniCel DxI 800 (Beckman Coulter Switzerland, Nyon, Switzerland). All assays were performed according to the recommendations of the manufacturer in an accredited laboratory (ZLM St. Gallen, ISO 17025). Quality control was performed following the manufacturer’s instructions and on each day of testing.

P-values have been calculated using an unpaired, two-sided t-test with independent variances, or the Wilcoxon rank sum, respectively the Mann–Whitney test. P-values of <0.05 were considered statistically significant. Statistical definitions, analysis and visualization were based on or performed with software R, using the implemented statistical tests and the packages “tidyverse” and “ggplot2” (20). Data are presented as median with interquartile range (IQR).

Two hundred and seventy-eight participants (entire cohort), aged 12.0–91.2 years (Median = 51.2, IQR = 25.8) who had a positive SARS-CoV-2 RT-PCR diagnosis were enrolled (Table 1); 26 individuals that had COVID-19 and a need for hospitalization were excluded. The entire cohort (n = 278) contained 166 (59.5%) women and 112 (40.5%) men. The first visit was conducted at a median of 6 weeks (range 1–40, IQR = 4) after positive PCR diagnosis. In total, 2016 sera were collected for further analysis. Among the 278 participants, 54 participants aged 24.8–91.2 (Median = 55.8, IQR = 13.9) repeated the study course three times. This long-term sub cohort (n = 54) consisted of 22 (40.7%) women and 32 (59.3%) men, whereof 20 (37%) men were older than 54 years. If individuals were vaccinated during the study course, only data from visits before the first vaccination took place were included (66 sera of 24 participants were excluded).

A control cohort of 115 participants (control cohort) aged 18.0–81.0 (Median = 51.0, IQR = 23.5) was enrolled in parallel (Table 1); 11 participants were excluded (6 because of missing SARS-CoV-2 PCR data and 5 because of missing troponin values). The control cohort (n = 115) consisted of 82 (71.3%) women and 33 (28.7%) men. Due to the recruitment design, all control individuals were also symptomatic. First visit for blood draws was conducted at a median of 9 weeks (range 0–41, IQR = 6) after negative PCR diagnosis.

Median high sensitive troponin I concentration (cTnI) in the entire cohort (n = 278) over the whole study duration of 14 months was 2.7 ng/L (95% CI 2.0–3.4 ng/L). Longitudinal gender-specific analysis of the cTnI results showed a significant increase from week 3 (2.65 ng/L, 95% CI 1.3–4.7 ng/L) to week 9 (4.65 ng/L, 95% CI 4.0–5.4 ng/L) after infection in men (Figure 1, p < 0.0001). The cTnI level of men at week 9 is significantly higher compared to men with a negative SARS-CoV-2 PCR and serology (median cTnI 2.8 ng/L, 95% CI 2.47–3.17 ng/l, p < 0.0001) and as well-compared to the baseline value of the entire positive cohort (2.7 ng/L, 95% CI 2.0–3.4 ng/L). In contrast, no significant increase of cTnI was detected in females during the first 9 weeks after SARS-CoV-2 infection (Figure 1).

Extending the observation period to 57 weeks (long-term sub cohort, n = 54) revealed also no significant change in cTnI in women. However, median cTnI in men remained persistently increased (after the initial significant cTnI increase) over the entire observation period (Figure 2A). Further analysis disclosed a gender and age dependency of the changes observed (Figure 2B). Only men older than 54 years showed significant increased cTnI values over the entire period of more than 1 year. Men younger than 54 years did not present elevated cTnI values (Figure 2B).

As recently shown in an animal model (21), cardiac troponins are renally and hepatically cleared. We thus evaluated markers of kidney and liver function in the cohort study. Although alanine aminotransferase (ALT) levels showed the expected gender difference (22) in men (median 26.4 U/l) and women (median 17.6 U/l), ALT values did not change over time in neither gender (p for difference > 0.12 in men and p > 0.33 in women). Also, ALT values in neither women (p > 0.23) nor men (p > 0.12) were different from those observed in the control cohort. The same circumstances were true for creatinine: values were expectedly (23) higher in men (median 82.9 μmol/l) than women (median 65.6 μmol/l); but they were again unchanged over time in both genders (men p > 0.37, women p > 0.24). Again, no significant differences to the control cohort were seen (men p > 0.19, women p > 0.15). The same was true for D-Dimer and brain natriuretic peptide (BNP): D-Dimer values (median in men 0.456 mg/l; median in women 0.466 mg/l) were unchanged over time in comparison to the control cohort (men p > 0.65, women p > 0.72); BNP values (median in men 25.4 ng/l; median in women 33.4 ng/l) showed a similar behavior and were not significantly altered in comparison to the control cohort (men p > 0.83, women p > 0.98).

Potential associations between various significant conditions of the patients (cancer, diabetes, pulmonary disease, cardiovascular disease, hypertension, immune suppression, smoking, and the intake of medications) and cTnI levels were evaluated. Only cardiovascular disease and hypertension were associated with cTnI (Figure 3), all other conditions did not associate with cTnI.