Demographics of the study subjects

In this study, 90 hospitalized symptomatic COVID-19 patients and 90 asymptomatic SARS-CoV-2 carriers were analyzed. The symptomatic group comprised 40 males and 50 females (mean age 53 ± 19), while in the asymptomatic group there were 42 males and 48 females (mean age 49 ± 15). No difference was found for gender and age between the two groups (P > 0.05). Among symptomatic patients, 30 had received no vaccine (unvaccinated), 24 had received one-dose vaccine and 36 had received the second vaccine dose. In comparison, in the asymptomatic group, there were 7 unvaccinated subjects (statistically significant from the symptomatic group, P < 0.0001), 36 one-dose vaccinees (P = 0.08) and 47 two-dose vaccinees (P = 0.17). Median time passed since symptom onset to sample collection was 25 [23–28] days for the symptomatic group. Median times passed since the last vaccination to sample collection for symptomatic and asymptomatic groups were 49 [30–73] days and 38 [28–57] days, respectively. In the asymptomatic group, 83 patients were primarily infected and 7 patients experienced secondary infection with SARS-CoV-2, whereas in the symptomatic group the number of primarily infected cases was 69 and reinfected cases 21. Thus, the two groups were significantly different in terms of reinfection cases (P = 0.007). Reinfection was defined based on patients' records showing a PCR-confirmed SARS-CoV-2 infection in the past. Median times elapsed since the first SARS-CoV-2 infection to the second infection were 150 [60–212] days and 190 [148–247] days for the symptomatic and asymptomatic groups, respectively. Among the vaccinated symptomatic patients, 48 patients had received Sinopharm, 6 patients had received Barekat, 4 had received AstraZeneka, 1 patient had received Sputnik V and 1 had been injected with Pastocovac. In the asymptomatic group, the number of cases receiving Sinopharm, Barekat, AstraZeneka and Sputnik V were 48, 9, 13 and 10, respectively. No one had received Pastocovac and, instead, three cases had received Pfizer vaccine. In the symptomatic group, 47 patients had no underlying disease (comorbidity), while other patients were suffering from diabetes (23), hypertension (17) and other cardiovascular disorders (2). Besides, one patient had chronic lung disease. Among hospitalized patients, nine patients had a fatal outcome because of COVID-19 during their hospitalization, of which four had hypertension, three had diabetes, and two had no comorbidity. Table 1 represents the demographic data of the symptomatic and asymptomatic cases including some clinical laboratory data, paraclinical findings and vaccination statuses.

Table 1 Demographic data of symptomatic and asymptomatic subjectsDetermining the levels of IgG antibodies produced against S and N antigens of SARS-CoV-2

As stated in the methods, median times between confirmation of SARS-CoV-2 infection (by RT-PCR) and serum sample collection from symptomatic and asymptomatic cases were 23 [22–26] days and 22 [21–23] days, respectively, to give the patients enough time for antibody production. Also, some patients had been previously exposed to SARS-CoV-2 and there were cases receiving the first and even second vaccine doses. Median time elapsed since the last vaccination to sample collection were 49 [30–73] days and 38 [28–57] days for the symptomatic and asymptomatic groups, respectively. Accordingly, all the study subjects had enough time to generate specific anti-SARS-CoV-2 IgG antibodies. According to our findings, of the 90 hospitalized symptomatic patients, 12 were seronegative for anti-S IgG, while only 5 out of 90 asymptomatic carriers were seronegative for this antibody (P = 0.12). Regarding anti-N IgG antibody, 71 out of 90 symptomatic patients tested positive, 11 patients were seronegative and 8 patients were considered suspicious for COVID-19, as their COI values fell within 0.9–1.1. Also, among 90 asymptomatic cases, 18 individuals were seronegative and 2 were detected as suspicious for anti-N IgG antibody (P > 0.05). Seronegative and suspicious cases were excluded from avidity measurement experiments. Serum levels of anti-S IgG and anti-N IgG antibodies were significantly higher in the symptomatic than asymptomatic patients (P < 0.01 and P < 0.001 respectively, Fig. 1). For anti-S IgG, the symptomatic group showed a median value of 157.7 [47.07–160.5] RU/ml, while this value in the asymptomatic group was 149.0 [115.1–151.4] RU/ml. Median COI values for anti-N IgG in the symptomatic and asymptomatic groups were 6.078 [1.905–9.072] and 3.676 [1.706–5.937], respectively. It is worth noting that regarding anti-S IgG titers following the instructions, for some serum samples the obtained OD values were outside the linear range of the ELISA kit. For these samples, we further diluted the sample so that the obtained OD fell within the linear range of the kit and accordingly measured the antibody titer (multiplying by the respected dilution factor). Herein, based on our experiences with further increase in dilution (e.g., 1/2, 1/4, 1/8 and…), the obtained OD values reduced in a roughly linear manner.

Fig. 1

Serum levels of anti-S IgG (A) and anti-N IgG (B) in both groups of the study. Symptomatic patients had higher IgG antibodies compared to the asymptomatic group. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. Mann–Whitney test was used for statistical comparison between the two groups. **P < 0.01 and ***P < 0.001

Comparison of anti-SARS-CoV-2 IgG avidity between the symptomatic and asymptomatic groups

After determining the anti-S and anti-N IgG levels in the symptomatic and asymptomatic groups, in the next step we analyzed the total binding affinities of these antibodies (avidity) using a modified ELISA method (urea dissociation assay) for evaluation of IgG avidity. In this method, low avidity antibodies are removed after adding urea, leaving antibodies with higher avidities bound to the antigen of interest. By dividing OD obtained from urea-treated well by OD of the intact well, the proportion of high avidity antibody/total antibody, which is defined as avidity index (AI), is acquired. The higher the AI value, the more strong is the binding of antibody to the antigen being assayed. Our findings revealed that, in total, the asymptomatic group had serum anti-S and anti-N IgG antibodies of higher avidities compared to the hospitalized symptomatic group (P < 0.0001 for both comparisons) as illustrated in Fig. 2. Regarding anti-S IgG, the symptomatic and asymptomatic groups showed median AI values of 57.10% [13.02–95.70%] and 93.78% [51.51–98.45%], respectively. Median anti-N IgG AI values were 50.81% [26.98–76.07%] for the symptomatic group and 73.52% [52.84–88.43%] for the asymptomatic cases.

Fig. 2

Comparison of the anti-S IgG (A) and anti-N IgG (B) AI values between the symptomatic and asymptomatic groups. Regarding both antibodies, AI values in the asymptomatic carriers were higher compared to symptomatic patients. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. Mann–Whitney test was used for statistical comparison between the two groups. AI: avidity index. ***P < 0.001

Anti-S and anti-N IgG antibodies were considered to be of “low’, “intermediate” and “high” avidity as per explanations given in the methods section. Of the 78 anti-S IgG seropositive symptomatic patients, the total number of patients with anti-S IgG antibodies of “low”, “intermediate” and “high” avidity was 32, 4 and 42, respectively. Comparatively, in the asymptomatic group, 13 out of 85 seropositive cases had “low” avidity anti-S IgG (P = 0.0004), 7 cases showed intermediate avidity (P = 0.63) and 65 cases showed high avidity (P = 0.004). Therefore, the two groups were statistically different in terms of total numbers of cases with “low”, and “high” avidity serum IgG specific to SARS-CoV-2 S antigen. When statistically comparing anti-S IgG AI values of each AI grade between the two groups, it was found that among “low avidity” cases, asymptomatic carriers had a statistically higher median AI value (27.22% [19.34–34.55%]) compared to the respective symptomatic subgroup (11.16% [5.31–17.77%], P < 0.0001). Interestingly, comparison of AI values within the “intermediate avidity” subgroups yielded a different result, so that the anti-S IgG AI value was significantly higher in the symptomatic as compared to the asymptomatic subgroup (47.92% [47.08–48.94%] vs 44.82% [42.37–44.82%], respectively, P < 0.01). However, the two groups showed no statistically significant differences with regard to “high avidity” anti-S IgG (95.42% [88.50–98.36%] for symptomatic vs 97.29% [81.39–98.88%] for asymptomatic, Fig. 3A). Considering serum anti-N IgG, the symptomatic group comprised 27, 7 and 37 patients with “low”, “intermediate” and “high” avidity antibody, respectively, in comparison to 8, 4 and 58 subjects in the asymptomatic group (P = 0.0005, P = 0.54 and P = 0.0002, respectively). However, no statistically significant difference was found between the two groups for neither of the anti-N IgG avidity grades as demonstrated in Fig. 3B.

Fig. 3

Comparison of anti-S (A) and anti-N (B) IgG AI values between the two groups in relation to AI grades. Concerning anti-S IgG, “low avidity” asymptomatic cases showed a higher AI value compared to the respective subgroup of symptomatic patients. In the “intermediate avidity” subgroups, however, AI value of symptomatic patients was higher in comparison to that of asymptomatic SARS-CoV-2 carriers. Comparison between “high avidity” subgroups showed no statistically significant difference. For anti-N IgG, neither of the three subgroups showed any significant difference of AI. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. To statistically compare AI values between the two groups for each AI grade, unpaired t test (where data distribution was normal) or Mann–Whitney test (where data were non-normally distributed) was used. AI: ****P < 0.0001 and **P < 0.01

Comparison of IgG AIs in relation to vaccination doses

Regardless of vaccine type administered, we tried to find the link between vaccination doses (first or second dose) with IgG AI in the two groups, to see if COVID-19 vaccination is associated with improved AI. Our statistical analysis showed that anti-S IgG AI was correlated with vaccination status. COVID-19 symptomatic patients previously receiving one-dose and two-dose vaccines exhibited higher anti-S IgG AIs in comparison to unvaccinated patients (median 90.43% [14.67–95.95%], P < 0.05 and 87.24% [43.23–98.15%], P < 0.001, respectively, vs 12.83% [5.263–48.12%] of the unvaccinated subgroup; Fig. 4A). Nonetheless, anti-S IgG AI was not significantly different between one-dose and two-dose vaccinees. In the asymptomatic group, we could not find any statistically significant differences for anti-S IgG AIs among unvaccinated, one-dose vaccinated and two-dose vaccinated individuals (Fig. 4C), though vaccinated cases showed increased AI values. Besides, we compared anti-N IgG AI values in relation to vaccination status in symptomatic and asymptomatic groups, as well. The results revealed that anti-N IgG AIs did not significantly differ among unvaccinated, one-dose and two-dose recipients of SARS-CoV-2 vaccines in the symptomatic and in the asymptomatic groups (Fig. 4B and D).

Fig. 4

Comparison of IgG AI values among different subgroups based on vaccination status. A, B Represent anti-S IgG AI and anti-N IgG AI, respectively, among different subgroups in the symptomatic patients. C, D AI values related to anti-S IgG and anti-N IgG antibodies in the asymptomatic group, respectively. Significant differences were only found for anti-S IgG in symptomatic subgroups. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. Statistical comparisons between subgroups in all cases were performed by Kruskal–Wallis test except for anti-N IgG in asymptomatic subgroups for which one-way ANOVA was used. AI: avidity index; no vaccine: unvaccinated individuals; *P < 0.05 and ***P < 0.001

Comparison of IgG AIs in relation to vaccine type

Anti-S and anti-N IgG AIs were also compared between unvaccinated and recipients of different types of vaccines, regardless of vaccine dose. Considering anti-S IgG AI in the hospitalized symptomatic COVID-19 patients, a statistically significant difference was only found between recipients of Sinopharm with unvaccinated patients (90.87% [43.49–97.56%] vs 12.78% [4.922–26.04%], P < 0.0001). Although recipients of other vaccines also showed increased AI values, the differences were not statistically significant (Fig. 5A). In the asymptomatic carriers, the highest level of anti-S IgG AI was observed in recipients of Sputnik (97.85% [78.78–99.19%]) and AstraZeneka (96.59% [73.05–99.65%]). Also, three cases received Pfizer with a median of 94.69% [91.86–97.58%]. However, no statistically significant difference was detected among different subgroups (Fig. 5C). Intriguingly, asymptomatic individuals receiving Barekat vaccine even had a lower median anti-S IgG AI value (60.70% [31.15–97.44%]) compared to the unvaccinated subgroup (with median AI value of 82.67% [32.16–97.55%]). However, this difference was not statistically significant, either. Evaluation of anti-N IgG AIs among different vaccine recipients in symptomatic and asymptomatic individuals revealed no statistically significant difference as depicted in Fig. 5B, D. It should be noted that as two out of three Pfizer-receiving cases in the asymptomatic group were seronegative for anti-N IgG, this subgroup comprised only one AI value and was therefore excluded from statistical analysis (Fig. 5D).

Fig. 5

Comparison of IgG AI values among different subgroups within each group based on vaccine type. A, B Anti-S IgG AI and anti-N IgG AI values in different subgroups of symptomatic patients, respectively. C, D AI values of anti-S IgG and anti-N IgG antibodies in the asymptomatic group, respectively. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. Kruskal–Wallis test was used for statistical comparisons between subgroups. AI: avidity index; no vaccine: unvaccinated individuals. ****P < 0.0001

Comparison of IgG AIs in primarily and secondarily SARS-CoV-2 infected cases

As a previous history of infection with a microbe can affect humoral immune response quantity and quality to the current re-exposure to the same microbe, we analyzed IgG AIs with respect to infection history of individuals. To this aim, first anti-S and anti-N IgG AIs were compared between primarily infected and reinfected cases of each group separately. No statistically significant differences in anti-S and anti-N IgG AIs were found between primarily and secondarily infected cases within each group (data not shown). In the symptomatic group, anti-S IgG AIs of primarily infected and reinfected patients were determined to be 57.04% [13.08–95.52%] and 87.16% [12.73–98.11%], respectively. Also, anti-N AIs were 50.91% [26.57–76.59%] and 42.19% [22.71–83.24%] in these symptomatic patients. In asymptomatic SARS-CoV-2 carriers, anti-S IgG AIs were found to be 93.75% [47.68–98.64%] and 95.44% [62.93–98.09%] in primarily infected and reinfected cases, respectively. The levels of AI for IgG antibodies against SARS-CoV-2 N antigen were relatively lower and determined to be 72.15% [53.96–87.95%] and 77.46% [42.84–93.77%], respectively, in the first infection and reinfection cases.

The next step was to make comparisons on serum IgG AIs in primarily infected cases of symptomatic and asymptomatic groups as well as in reinfected cases of the two groups, respectively. According to the acquired data illustrated in Fig. 6, only primarily infected cases of the two groups showed statistically significant differences in terms of serum anti-S and anti-N IgG AI values, while the differences between secondarily infected cases were not statistically significant. In symptomatic patients infected with SARS-CoV-2 for the first time, anti-S and anti-N IgG AI values were calculated to be 57.04% [13.08–95.52%] and 50.91% [26.57–76.59%], respectively, while the corresponding AI values in asymptomatic cases were found to be 93.75% [47.68–98.64%] (P < 0.001) and 72.15% [53.96–87.95%] (P < 0.001) as shown in Fig. 6A, B. With regard to SARS-CoV-2 reinfected cases, although anti-S and anti-N AI values were higher in the asymptomatic group (95.44% [62.93–98.09%] and 77.46% [42.84–93.77%], respectively) in comparison to symptomatic patients (87.16% [12.73–98.11%] and 42.19% [22.71–83.24%], respectively), the differences were not statistically significant (Fig. 6C, D).

Fig. 6

Comparison of IgG AI values between the symptomatic and asymptomatic groups with regard to infection count (primary infection or reinfection). Anti-S IgG AI (A) and anti-N IgG AI (B) values of primarily infected cases were higher in the asymptomatic SARS-CoV-2 carriers in comparison to corresponding values in the symptomatic group. C, D AI values of anti-S IgG and anti-N IgG antibodies, respectively, of secondarily infected cases compared between the symptomatic and asymptomatic groups. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. In all cases, statistical comparisons between the two groups were performed by Mann–Whitney test except for anti-N IgG in secondarily infected cases for which unpaired t test was used. AI: avidity index. ***P < 0.001 and ns: not significant

Comparison of IgG AIs in relation to the number of exposures to the SARS-CoV-2 S and N antigens

The AI values of anti-SARS-CoV-2 antibodies were also compared between the two groups and within each group based on the number of exposures to the respective SARS-CoV-2 antigens, regardless whether these exposures were to vaccinations or infections. Herein, an individual who received two vaccinations and experienced a single breakthrough infection would count as three exposures to the SARS-CoV-2 S antigen. The same is true for a one-time vaccinated individual who has been infected twice. However, depending on the vaccines received, these two individuals might have different numbers of exposures to the SARS-CoV-2 N antigen. Taking these into account, further statistical analyses were performed. In some cases, very few out-of-range data were excluded to have a better comparison between the subgroups. First, the anti-S IgG and anti-N IgG AI values were compared between different subgroups within each group (symptomatic and asymptomatic). As illustrated in Fig. 7, statistically significant differences were observed only for the anti-S IgG within symptomatic subgroups. Patients who had two (S2) and three (S3) exposures to the SARS-CoV-2 S antigen showed significantly higher anti-S IgG AIs as compared to one-time exposed (S1) cases (P < 0.001 and P < 0.0001, respectively). Median anti-S IgG AI values for the S2 and S3 subgroups were determined to be 91.21% [17.01–95.29%] and 87.32% [31.19–98.28%], respectively, while that of the S1 subgroup was 7.82% [3.40–15.36%]. In addition, antibody avidity was compared between the two symptomatic and asymptomatic groups for each subgroup based on the number of exposures to the respective antigens. According to the obtained results, presented in Fig. 8, concerning anti-S IgG, the two groups exhibited statistically significant differences only in the S1 and S3 subgroups (P < 0.0001 and P < 0.05, respectively). Symptomatic COVID-19 patients with one-time exposure to SARS-CoV-2 S antigen showed a median anti-S IgG AI value of 7.82% [3.40–15.36%], while the same subgroup of asymptomatic SARS-CoV-2 carriers had a median AI of 82.67% [31.86–99.02%]. Also, S3 subgroups of symptomatic and asymptomatic cases showed AI values of 87.32% [31.19–98.28%] and 96.64% [61.74–98.41%], respectively. Although asymptomatic S2 and S4 subgroups also showed higher avidity of anti-S IgG, the differences were not statistically significant (Fig. 8A). Similarly, asymptomatic individuals with one (N1) and four (N4) exposures to SARS-CoV-2 N antigen had statistically non-significant higher AI values for this antibody compared to the corresponding subgroups within the hospitalized symptomatic group (P > 0.05, Fig. 8B). However, significant differences were observed for N2 (asymptomatic: 72.19% [60.21–91.12%] vs symptomatic: 49.06% [24.44–60.03%]; P < 0.001) and N3 subgroups (asymptomatic: 73.87% [55.08–95.12%] vs symptomatic: 42.43% [25.33–79.20%]; P < 0.05) as demonstrated in Fig. 8B.

Fig. 7

Comparison of anti-S and anti-N IgG AI values among different subgroups within each group in relation to the number of exposures to the respective SARS-CoV-2 antigens. A, B Anti-S IgG AI and anti-N IgG AI values in different subgroups of the symptomatic group, respectively. C, D AI values of anti-S IgG and anti-N IgG antibodies in different subgroups of the asymptomatic group, respectively. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. Kruskal–Wallis test was used for statistically comparing anti-S IgG AI values among different subgroups. For anti-N IgG, statistical comparisons among the subgroups were performed by one-way ANOVA. AI: avidity index. ***P < 0.001, ****P < 0.0001 and ns: not significant

Fig. 8

Comparison of anti-S (A) and anti-N (B) IgG AI values between the two groups in relation to the number of exposures to the respective SARS-CoV-2 antigens. S1, S2, S3 and S4 represent, respectively, the subgroups with one, two, three and four times exposure to the SARS-CoV-2 S antigen. N1, N2, N3 and N4 correspond to the subgroups with one, two, three and four times exposure to the N antigen, respectively. Data are depicted as box plots with medians, boundaries between the interquartile ranges, and whiskers between the minimum and maximum. For each subgroup, Mann–Whitney test was used for statistical comparison. AI: avidity index, *P < 0.05, ***P < 0.001 and ****P < 0.0001

Correlation analyses

We analyzed the existence of potential correlations between IgG titers and IgG AI values with different parameters in the two groups including laboratory findings of hospitalized patients. In the symptomatic group, no correlation was found between anti-S IgG titer, anti-S IgG AI, anti-N IgG titer and anti-N IgG AI with age, gender, fever, presence of comorbidity or different laboratory findings including WBC count, CRP, ESR, D-dimer, creatinine, urea, ALT, AST, ALP and LDH. As previously stated, 43 out of 90 symptomatic patients had some kind of comorbidity. No statistically significant difference in anti-S avidity or anti-N avidity was found between different subgroups with regard to comorbidity (data not shown). Totally, when considering AI values between patients without any kind of comorbidity with those having comorbidity (irrespective of its type), anti-S AI values were estimated to be 89.71% [12.96–96.89%] and 48.12% [14.04–93.92%], respectively. Considering anti-N AI values, the corresponding values were 51.08% [27.33–79.72%] and 43.69% [23.66–74.39%], respectively. But, no statistically significant difference was detected regarding neither of the antibodies evaluated. However, in the symptomatic group, significant positive correlations were found between anti-S antibody titer with anti-S antibody avidity (r = 0.74, P < 0.001) as well as between anti-N antibody titer with anti-N antibody avidity (r = 0.63, P < 0.001). In comparison, in the asymptomatic group, more significant correlations were observed between different antibody parameters. Not only anti-S IgG and anti-N IgG levels were positively correlated with anti-S avidity (r = 0.584, P < 0.001) and anti-N avidity (r = 0.622, P < 0.001), respectively, but also we found correlations between antibody parameters of different antigens. For example, anti-S IgG avidity showed a positive correlation with anti-N IgG avidity (r = 0.413, P < 0.001), while such a correlation was not found in symptomatic patients. In addition, anti-S IgG titer showed a weak positive correlation with anti-N avidity (r = 0.278, P < 0.05). Interestingly, correlation between anti-N IgG titer with anti-S IgG avidity was rather stronger (r = 0.428, P < 0.001) in the asymptomatic carriers.