Introduction

In the almost 2 years since the emergence of the COVID-19 pandemic, the global medical and scientific community has had to reevaluate models and assumptions regarding transmissible respiratory viral infection. Questions immediately arose about whether SARS-CoV-2 infection would mirror the features of other formidable etiologic agents, such as influenza group viruses. For better or worse, scientists quickly understood from early data that the novel viral infection would carry very significant personal and socioeconomic burdens. The so-called 'Spanish' influenza pandemic (1918–1920) was a similar milestone event of the early twentieth century. The influenza pandemic had a beginning and an end [1], yet the COVID-19 pandemic is ongoing. Alternating episodes featuring decreases and increases in the numbers infected are occurring, with similar dynamics being observed in most countries globally. As of September 11, 2021, 225,241,173 human SARS-CoV-2 infections were registered worldwide, which is 28.8 thousand people per 1 million population. Of those, 4,641,113 people (2%) have died [2].

The Kyrgyz Republic (Kyrgyzstan) is a Central Asian country located in the western and central areas of the Tian Shan mountain system. The population of Kyrgyzstan was 6,522,811 people as of 01.01.2020. The average population density is about 33 per km2 [3]. In regional terms, the population is distributed unevenly. The bulk is concentrated in the two largest cities, Bishkek and Osh, where about 22% of the Republic's population lives. Bishkek City (the capital of the country) is home to 6371 people per km2. The population density in Osh City is 2134 per km2. In the rest of the Republic, it is 32 per km2 [IQR 17.4–40.6]. In the Issyk-Kul and Naryn regions, the population density is 12 per km2 and 6 per km2, respectively. Such an uneven population distribution inevitably affects infectious disease levels in the country.

As of September 12, 2021, 177,158 cases of SARS-CoV-2 infection had been registered in Kyrgyzstan (19,268 per 1 million population). Of those, 2573 people died (1.45%). COVID-19 incidence in 2020–2021 manifested itself in three waves. The first wave began in the last 10 days of June 2020 and ended by mid-August 2020. The beginning of the second wave was in the last 10 days of September 2020 and ended only at the end of December 2020. The third wave formed in the last 10 days of April 2021. Peak incidence was noted in mid-July, and the end fell at the end of August 2021. The incidence during these periods ranged from 500 to 1600 people per day. As of September 13, 2021, the number of newly infected people in the Republic was 92 over the previous 7 days [4]. In other words, SARS-CoV-2 epidemic process activity was not high during this period. The largest numbers of cases were registered in the cities of Bishkek and Osh, as well as in the Osh, Jalal-Abad, and Chui regions.

Vaccination of the population against COVID-19 can affect conditions in terms of morbidity. Three vaccines are used in Kyrgyzstan: Gam-COVID-Vac ('Sputnik V', Russia); EpiVacCorona (Russia); and Sinopharm (PRC). As of 09/13/2021, 219,809 people (3.3%) had been vaccinated with one dose of vaccine, and 530,448 people (8.1%) had been completely vaccinated [5]. Thus, only 11.4% of the Kyrgyz population is protected from COVID-19 to some extent [6]. A certain contribution to protection is also made by post-infectious immunity that develops as a result of transmitted COVID-19 [9,10,11]. Thus, the real seroprevalence level of the population consists of post-infectious and post-vaccination components of seropositivity. The purpose of this study was to assess the seroprevalence of antibodies (Abs) to SARS-CoV-2 in the Kyrgyz population formed as a result of previous infection and/or vaccination. Similar studies have not been conducted in Kyrgyzstan before.

Materials and methods

Formation and randomization of the volunteer cohort

The study was organized and carried out as part of activities under a program of scientific cooperation between countries of Eastern Europe, the Caucasus, and Central Asia in accordance with: a Rospotrebnadzor Order (dated 09.09.2021 No. 512) on "The Procedure For Implementing Orders Of The Government Of The Russian Federation" (dated 18.06.2021, No. 1658-r) on the one hand; and Order of the Minister of Health and Social Development of the Kyrgyz Republic (dated July 23, 2021 No. 839) on the other.

A cross-sectional, randomized study of SARS-CoV-2 seroprevalence was carried out according to a program developed by Rospotrebnadzor, with the participation of the Saint Petersburg Pasteur Institute, taking into account WHO recommendations [12]. All blood samples were taken in a short time period (28.06.2021 to 03.07.2021), followed by analysis. In all stages of the study (collection, organization, and analysis of data), cloud (internet) technologies were used (Fig. 1).

Fig. 1
figure 1

Cohort formation algorithm and sequence of serological studies using cloud technologies [13]

Full size image

Four days before the start of the study, a wide information campaign (on state television channels, state news sites, the social network Facebook) was carried out among the population to explain the goals and objectives of the study. Those who expressed a desire to participate in the study filled out a special questionnaire posted through the Internet to the cloud service. The next stage of the study was analysis of the questionnaires by the cloud service. Groups of volunteers were formed from those who met the inclusion criteria. In accordance with the specific equation of de Moivre Laplace's theorem [14], the total number of volunteers in the cohort was 9,471. The cloud service featured quotas (maximum number of volunteers) for each age group in each region. After reaching this quota, the cloud service stopped registering volunteers. In that way, volunteers were randomized by age and place of residence, taking into account the proportional representation of each region of the Republic (Tables 1, 2, 3). It should be emphasized that in accordance with the conditions of the study, all information was obtained only from the questionnaires. An additional survey of volunteers was not conducted.

Table 1 Number of surveyed residents living in various regions of the Kyrgyz Republic
Full size table
Table 2 Distribution of volunteers by occupational group
Full size table
Table 3 Age structure of the surveyed volunteer cohort
Full size table

Each volunteer, or their legal representative, was acquainted with the goals and conditions of the upcoming study, followed by the signing of an informed consent. The study was organized in accordance with the provisions of the Declaration of Helsinki and approved by the ethics committees of the Scientific and Production Association "Preventive Medicine" (protocol No. 7, ref. No. 01-288, dated December 9, 2020) and the St. Petersburg Pasteur Institute (protocol No. 64, dated May 26, 2020).

In total, 0.15% (95% CI 0.14–0.15) of the overall population was selected for the cohort from regions of the Republic. The smallest share of volunteers was in the City of Osh, as well as in the Osh and Chui regions. The largest shares were from the Talas region, Naryn region, and the City of Bishkek. The cohort included 2868 men (30.3%) and 6603 women (69.7%). The distribution of volunteers by professional activity is given in Table. 2.

In addition to regional randomization, volunteers were stratified into seven age groups (years old): 1–17; 18–29; 30–39; 40–49; 50–59; 60–69; and 70 or more (70+). Taking into account the peculiarities of maturation of the immune system in children [15, 16], individuals in the 1–17 years old group were divided into three subgroups (years old): 1–6; 7–13; and 14–17 (Table 3).

Vaccination against SARS-CoV-2

A history of SARS-CoV-2 vaccination with one of the three vaccine preparations was noted in the questionnaires of 2439 volunteers. During the study period in Kyrgyzstan, the Gam-COVID-Vac (Sputnik V) vaccine, a two-component vector vaccine officially approved in the country, was used (Gamaleya Research Institute of Epidemiology and Microbiology, Russia) [17]. Two other vaccines were permitted for use in accordance with the Decree of the Government of the Kyrgyz Republic No. 30 dated January 29, 2020 in a situation of acute shortage of vaccines in the context of a worsening epidemiological situation: EpiVacCorona, a peptide vaccine (VECTOR State Scientific Center for Virology and Biotechnology, Russia) [18]; and BBIBP-CorV inactivated vaccine (Sinopharm Group Co., Ltd., Shanghai, China) [19, 20]. Vaccination was carried out in accordance with their instructions for use. As follows logically from their composition: the immune response to Sputnik V vaccination is mainly aimed at binding the RBD [21]; the inactivated BBIBP-CorV vaccine induces antibodies against all viral antigens [22]; and the EpiVacCorona induces Abs to a complex of chemically synthesized peptide immunogens and SARS-CoV-2 S protein [18].

Testing of volunteers for SARS-CoV-2 antibodies

For the study, 3 ml of blood was taken from all volunteers from the cubital vein into vacutainers containing EDTA solution. Following centrifugation, blood plasma was separated from cell components and used for quantitative determinations (Thermo Scientific Multiskan FC). Antibodies to SARS-CoV-2 were determined by enzyme immunoassay (indirect two-phase ELISA variant) using the following reagents. For quantitative determination of specific IgG Abs to SARS-CoV-2 nucleocapsid protein (N antigen), the 'Reagent set for quantitative determination of human IgG to SARS-CoV-2N protein (N-Cov-2-IgG PS), series 001' (St. Petersburg Pasteur Institute, Russia) was used. For qualitative analysis of the presence of anti-RBD Abs in vaccinated volunteers (except EpiVac Corona), the 'SARS-CoV-2 IgG ELISA screen' kit (LabPack, Russia) was used. The results were registered by spectrophotometry (450 nm with a reference at 620 nm).

With the N-Cov-2-IgG PS kit, the quantitative content of Abs in the test sample is calculated from a calibration curve obtained from seven calibration samples. These are serial twofold dilutions of a blood serum pool of donors containing IgG Abs to the SARS-CoV-2N protein and a zero sample (K0). The obtained concentrations of the studied samples were expressed in conventional units: arbitrary units (AU/ml); and the international 'binding antibody units' (BAU/ml). The obtained concentration values of the studied samples were multiplied by a factor of 100 in accordance with the dilution of the serum/plasma (1:100). The concentration of IgG Abs to SARS-CoV-2N protein in BAU/ml was determined by comparing the calibration curve obtained from three reference calibrator samples [First WHO International Standard for anti-SARS-CoV-2 Human Immunoglobulin (NIBSC code 20/136)]. To convert AU/ml values to BAU/ml, the BAU/ml values of the calibration samples were determined from the WHO standard curve, followed by comparison between the quantitative values of the calibration curve in AU/ml and BAU/ml. By comparison, it was determined that 1 BAU/ml = 5.97 AU/ml. The sensitivity of analysis, determined by the average value of blank samples (n = 8) + 2 σ, was 0.2 AU/ml (0.03 BAU/ml). The specificity of analysis was confirmed by an absence of cross-reaction when testing plasma/serum samples obtained from individuals with various infections (HBV, HCV, HIV-1, HIV-2, adenovirus, influenza A/B virus, parainfluenza virus, respiratory syncytial virus, EBV, Chlamydia pneumoniae, Streptococcus pneumoniae, Mycobacterium tuberculosis, Mycoplasma pneumonia) before 2018. The threshold of seropositivity was determined by the method of linear regression according to the formula LoQcalibration = 10*(Sd/S), where Sd is the standard deviation for the mean values of the K7 calibration sample points, and S is the logit-log slope of the calibration curve. The limit of quantitation was 17 BAU/ml or 102 AU/ml, taking into account the dilution factor k = 100.

Vaccinated volunteers (except those who received EpiVacCorona) were qualitatively analyzed for the presence of anti-RBD Abs using the 'SARS-CoV-2 IgG ELISA screen' kit (LabPack, Russia). The sensitivity of the kit was determined by analyzing 54 serum/plasma samples from patients with PCR-confirmed COVID-19 taken 30–50 days after diagnosis; the sensitivity was 100% (95% CI 98.5–102.5). The specificity of analysis, confirmed by absence of cross-reaction when testing 100 plasma/serum samples obtained from individuals with the aforementioned viral infections, was 98.0% (95% CI 95.2–100.8). Evaluation of results was carried out by seropositivity coefficient (SC), calculated as SC = OD (sample)av/ODc, where: SC—seropositivity coefficient; OD (sample)av—the average value of sample optical densities; and ODc—the critical optical density (average OD of the negative control sample + the offset value set by the manufacturer). Above a threshold, the sample was considered positive (SC > 1.1).

Statistical analysis

Data processing was performed using the statistics functions in Excel 2010. Confidence intervals (95% CI) were calculated by the method of Wald and Wolfowitz [23], with correction as described by Agresti and Coull [24]. Correlation analysis was performed according to Spearman's method. The statistical significance of differences was calculated by the z-test using the appropriate online calculator [25]. Distribution normality was checked using the Kolmogorov–Smirnov test in the Excel 2010 package. Multivariate statistical analysis was performed using regression analysis for parametric data and correspondence analysis for nonparametric (nominal) data. Multivariate analysis was conducted using standard tools in the program Statistica version 12 (StatSoft). Unless otherwise indicated, differences were designated as significant when p ≤ 0.05.

Results

Gender distribution of SARS-CoV-2 Nag Ab seroprevalence

The study cohort included 2858 men and 6603 women. Thus, women were 2.3-fold more active in the study. As determined by serological testing, the share of SARS-CoV-2-seropositive men was 45.4% (95% CI 43.5–47.5). The share of seropositive women was 50.1 (95% CI 48.6–51.5).

Distribution of anti-Nag Abs among volunteers, by age

Twenty-one months after the onset of the pandemic, the proportion of volunteers with Abs to SARS-CoV-2 Nag was 48.7% (95% CI 47.7–49.7). The highest seroprevalence levels were recorded among volunteers in the older age groups (50–59, 60–69, 70+ years old); the lowest levels were seen in the age group 1–17 years old, or its subgroups (Table 4). The differences from the population mean were statistically significant (p < 0.001).

Table 4 SARS-CoV-2 Nag Ab seroprevalence in different age groups of Kyrgyz residents, July 2021
Full size table

Interesting results were obtained when quantifying anti-Nag Ab levels (Fig. 2, Table 1S). Volunteers in the first serogroup with minimal Ab levels (16.8–31.2 BAU/ml) were relatively evenly distributed among all age groups. There was some predominance in two groups (30–39, 40–49 years old) and significantly less among children 1–17 years old (p < 0.0001). The trend line is shown in blue. The age distribution of seroprevalence was described by a second-order polynomial equation of the form: y = − 0.4143x2 + 3.8929x + 5.3, with R2 = 0.9482. The angle coefficient (tgα) was 0.70, which indicates a weak dependence of Ab level on volunteer age.

Fig. 2
figure 2

Anti-Nag Ab serological intervals plotted by age group. Serogroups reflect serum anti-Nag Ab levels expressed in BAU/ml. Colored points with black confidence interval bars—the proportion of seropositive volunteers having anti-Nag Abs at the corresponding levels. Solid, colored lines are trend lines for the corresponding serogroup. Regression equations, determination coefficients (R2), and angle coefficients (tgα) are shown (colors correspond to serogroup). The numerical values are presented in Table 2S

Full size image

In the second serogroup (31.3–125.6 BAU/ml), the dependence was more complex. The smallest proportion of seropositive individuals was noted in the groups '1–17 years old' and '18–29 years old'. Starting from 60 to 69 years, the trend showed a slight decrease (all differences insignificant). Such dynamics are described by the third-order polynomial equation of the form: y = − 0.3778x3 + 4.0798x2 − 9.5782x + 22.229, with R2 = 0.9414. With further increasing Ab levels in the three remaining serogroups, the tendency was still preserved, but age differences gradually smoothed out, manifesting as a simplification of the curve's shape.

In the third serogroup (125.8–251.2 BAU/ml), the curve was described by the third-order polynomial equation: y = − 0.1167x3 + 1.4976x − 4.7786x + 8.6714, with R2 = 0.9, while tgα decreased to 0.94. Starting from the fourth serogroup group (251.4–502.5 BAU/ml), the polynomial dependence degenerated into a linear form: y = 0.7x + 0.2429, with R2 = 0.97, while tgα decreased to 0.68. The dependence of Ab level on age was still preserved, but its magnitude decreased.

Finally, in the fifth serogroup (> 502.5 BAU/ml), the age dynamics of antibodies fully corresponded to the upward linear equation of the form: y = 0.35x + 1.3, with R2 = 0.87, while tgα decreased to 0.40. There were no statistically significant differences between Ab levels among volunteers of any age in this serogroup.

Regional distribution of SARS-CoV-2 Nag seropositivity

In Kyrgyz regions, the proportion of seropositive volunteers varied from 38.1% (95% CI 32.6–43.9) in the Osh region to 53.3% (95% CI 49.8–56.8) in the Naryn region (Table 5). The differences are statistically significant (p < 0.0001).

Table 5 SARS-CoV-2 Nag Ab seroprevalence among residents of different Kyrgyz regions
Full size table

It was not possible to reveal any correlation dependencies for pairs: seropositivity with population density; morbidity with seropositivity; or morbidity with population density.

Application of a correspondence analysis statistical methodology (Fig. 3) in assessing regional SARS-CoV-2 seroprevalence also showed: a high degree of association between seropositivity and Osh region, Issyk-Kul region, Naryn region, and Chui region; and relatively low association with Osh City. The total Chi-square of the performed correspondence analysis was 64.1549 (p < 0.001).

Fig. 3
figure 3

Relationship between the dichotomous variable ‘presence or absence of SARS-CoV-2 antibodies (1—present, 0—absent)’ and different Kyrgyz regions, assessed by correspondence analysis

Full size image

SARS-CoV-2 Nag Ab seroprevalence in volunteers from different professional groups

To analyze the role of risk factors, the proportion of seropositive volunteers in different social and professional groups was assessed (Table 6). The highest seroprevalence level, significantly higher than the average for the surveyed population, was found in the health-care worker group [57.1% (95% CI 55.5–58.6) p < 0.0001]. The lowest seroprevalence was observed in artists and office workers (p < 0.001).

Table 6 SARS-CoV-2 Nag Ab seroprevalence among volunteers from different professional groups
Full size table

Significantly lower seroprevalence values were noted in business workers, those in manufacturing, the unemployed, and in persons in the classification 'Other' (all differences statistically significant, p < 0.05). Application of correspondence analysis statistical methodology (Fig. 4) in assessing the relationship between SARS-CoV-2 seroprevalence and professional activity revealed: a higher degree of association between the parameter ‘Abs present’ and such fields of activity as medicine and government service; and a relatively low seroprevalence in students and children. The total Chi-square of the performed correspondence analysis was 272.043 (p < 0.001).

Fig. 4
figure 4

Relationship between the dichotomous variable ‘presence or absence of SARS-CoV-2 antibodies (1—present, 0—absent)’ and various types of professional activity, assessed by the correspondence analysis

Full size image

SARS-CoV-2 Nag Ab seroprevalence in COVID-19 convalescents and those in contact with them

Information about past COVID-19 and contacts with patients was obtained from the questionnaire. Among those who experienced manifest COVID-19, the proportion seropositive was 64.5% (95% CI 61.9–67.1) (Table 7). As expected, this indicator was significantly higher than the surveyed cohort average [48.7% (95% CI 47.7–49.7) p < 0.0001].

Table 7 SARS-CoV-2 Nag Ab seroprevalence in COVID-19 convalescents and contacts
Full size table

Given the fact that convalescents play a significant role in the spread of infection, we assessed seroprevalence in a group of volunteers (2212 people) who had indicated contact with a sick or recovering COVID-19 individual in their questionnaire. Seroprevalence in these individuals was 49.7% (95% CI 47.6–51.8) and did not differ from the cohort average (Table 6).

Assessment of the proportions of asymptomatic COVID-19 among SARS-CoV-2 Nag-seropositive volunteers

Insofar as a high proportion of asymptomatic forms is a characteristic feature of SARS-CoV-2, we estimated the proportion of volunteers with probable asymptomatic infection. In this study, individuals were classified as asymptomatic, who, in addition to Abs to Nag, did not show any symptoms of COVID-19. This study identified 2440 asymptomatic cases in 3165 unvaccinated seropositive volunteers [77.1% (95% CI 75.6–78.5)] (Table 8).

Table 8 Share of asymptomatic infections, relative to overall unvaccinated seropositive volunteers, by age group
Full size table

The largest number of asymptomatic, yet SARS-CoV-2 seropositive, individuals was found among volunteers aged 1–17 and 19–29 years (p < 0.0001). In the remaining groups, the number of such individuals was relatively homogeneous. There were no statistically significant differences, in comparison with the average population, in these groups.

Assessment of seroprevalence in vaccinated volunteers

Vaccination was noted in the questionnaire by 2439 out of 9471 volunteers, which represents 25.8% (95% CI 24.9–26.7). Of these, the structure was: Sputnik V in 405 people [16.6% (95% CI 15.9–18.1)]; EpiVacCorona in 134 people [5.5% (95% CI 4.6–6.5)]; and Sinopharm in 1900 people [77.9% (95% CI 76.2–79.5)]. Hence, the Sinopharm vaccine was used in the Republic 3.5-fold more often than the two others. At the same time, the percentage of those vaccinated among volunteers showed a clear age dependence (Fig. 3). The largest proportion vaccinated was found among volunteers in the age groups of 40–49 and 50–59 years. The smallest proportion of those vaccinated was noted in the group ‘children aged 1–17 years (p < 0.001) (Fig. 5).

Fig. 5
figure 5

Distribution of vaccinated volunteers, by age group. Note sectors of the diagram—age groups (years old); vertical axis (radial)—share vaccinated (%); inner corners of the polygon—share vaccinated (%)

Full size image

In terms of region, the largest number of vaccinated volunteers lived in the Osh region and in the Republic’s capital, Bishkek. The fewest were in Osh (Fig. 6). The range of variation was 8.3-fold. We were unable to identify any significant climatic and geographical factors that could explain the observed regional distribution of vaccinated volunteers. It is possible that in some regions, a proportion of those vaccinated could be associated with a representation of health-care workers. Distribution of health-care workers by regions was as follows: Osh region—17.2% (95% CI 15.9–18.5) of all healthcare workers; Bishkek City—5.9% (95% CI 5.1–6.7); Talas region—16.2% (95% CI 15.0–17.5); Jalal-Abad—17.5% (95% CI 16.2–18.8); Naryn—12.2% (95% CI 11.1–13.3); Issyk-Kul—9.4% (95% CI 8.4–10.4); Chui—6.7% (95% CI 5.9–7.6); Batken—12.8% (95% CI 11.6–13.9); and Osh City—2.3% (95% CI 1.8–2.9).

Fig. 6
figure 6

Distribution of vaccinated volunteers by region

Full size image

When analyzing the distribution of vaccinated volunteers by occupation, it was found that the largest number of vaccinated were among health-care workers, and the vast majority of them were vaccinated with Sinopharm. The shares of health-care workers vaccinated with Sputnik V and EpiVacCorona were distributed almost equally (6.5% and 6.1%, respectively) (Table 9). Health-care workers were followed by pensioners by a wide margin [9.3% (95% 8.1–10.4) of 2439 vaccinated]. Among them, however, 59.3% (95% CI 52.6–65.8) preferred Sputnik V and only 3.5% (95% CI 1.5–6.9) received EpiVacCorona. The smallest proportion of the total number of vaccinated was noted among those in the arts [0.4% (95% CI 0.2–0.8)].

Table 9 Distribution of vaccinated volunteers by occupation and vaccine type
Full size table

Although the overall share vaccinated was small, it would nevertheless be expected to have some effect on overall SARS-CoV-2 Ab prevalence. The significance of this increase can be assessed by analysis of the proportion of Nag and RBDag seropositivity. For this phase of the study, the presence of Abs to N and RBD antigens in plasma was determined in all vaccinated volunteers. Since the number of volunteers vaccinated with EpiVacCorona was small (not representative), they were not included in the analysis. Plasma samples from 44 examined individuals were found to be invalid; further analysis included data from 2262 plasma samples from vaccinated volunteers (Tables 10, 11, 12). Overall, RBDag Ab seropositivity was observed in 91.4% (95% CI 90.2–92.6) of those vaccinated. There were no statistically significant differences by vaccine type. In 7.3% (95% CI 6.3–8.4) of vaccinated volunteers, neither Nag nor RBDag Abs were detected.

Table 10 Presence of IgG Abs to SARS-CoV-2 antigens in volunteers immunized with the Sinopharm vaccine (1869 people)
Full size table
Table 11 Presence of IgG Abs to SARS-CoV-2 antigens in volunteers immunized with the Sputnik V vaccine (393 people)
Full size table
Table 12 Seroprevalence in vaccinated volunteers
Full size table

Analysis of the data obtained for those vaccinated with Sinopharm (Tables 10, 12) showed that at least 91.6% (95% CI 90.3–92.8) of volunteers seroconverted after vaccination (1713 volunteers with Abs to RBDag). Perhaps, this figure is higher [92.7% (95% CI 91.5–93.9)] since the 21 people who were not found to have anti-RBD Abs had anti-Nag Abs; these could be due to either previous infection or vaccination. Only 7.2% (95% CI 6.1–8.5) of vaccinated volunteers lacked both Ab types (anti-Nag, anti-RBDag).

Analysis of data obtained from volunteers vaccinated with Sputnik V (Tables 11, 12) showed that 90.6% (95% CI 87.3–93.3) of volunteers seroconverted after vaccination (gained Abs to RBDag). One hundred and eighty volunteers [45.8% (95% CI 40.8–50.9)] had likely experienced an asymptomatic SARS-CoV-2 infection (they did not indicate a history of infection in the questionnaire). In these persons, the response to vaccination was more pronounced: 173 seropositive out of 180 patients [96.1% (95% CI 92.2–98.4)]. Among those not previously infected, there were 183 seropositive out of 213 [85.9% (95% CI 80.5–90.3)]. In 7.6% (95% CI 5.2–10.7) of volunteers, there was no response to vaccination (did not have Abs to RBDag).

Discussion

Based on the available official data, the incidence in the Republic can be regarded as consistently low. In the global ranking of countries, in terms of the number of infected, Kyrgyzstan occupies the 92nd place. The total number infected was 177,158 (19,268 per 1 million population) as of September 12, 2021. In the last week of September 2021, the number of infections registered daily did not exceed 100 (0.015‰). More than 96.7% of those infected had recovered as of September 16. Over the entire pandemic period, 1.4% of those officially diagnosed died. Such favorable data is probably a consequence of low population density not exceeding 33 per km2. In a number of areas, such as the Chui region, it is as low as approximately 6 per km2. In addition, the low official incidence could be due to the fact that PCR screening in Kyrgyzstan was insufficient: only those ill with COVID-19 who themselves sought medical help were subject to PCR testing and official registration. According to official Ministry of Health data, no more than 24% of the ill made contact for medical assistance. Thus, official data on population incidence may distort the real level of incidence.

The statistically significant gender difference identified in SARS-CoV-2 seroprevalence (men 45.4%, women 50.1%) is likely due to a larger number of household contacts for women, a characteristic of the Republic. Analysis of age stratification also showed patterns. Unlike other regional countries [26,27,28,29], higher seroprevalence in the age group 1–17 years old, or its subgroups, was not seen in the Kyrgyz Republic. On the contrary, the age structure of seroprevalence was characterized by significantly higher indicators among persons in older groups (50 years and above). The differences were statistically significant at p < 0.0001. When quantifying anti-Nag Abs, it was shown that minimal Ab levels prevailed among young and middle-aged individuals, with significantly higher levels among older volunteers (Fig. 2, Table 2S). This relationship between age and anti-Nag Ab level can be considered a feature of SARS-CoV-2, and it has been established in other studies [30, 31].

When analyzing any association between share of seropositive persons and areas of the Republic, no connection with region of residence was noted (Fig. 7). In the capital of the Republic, Bishkek, the proportion seropositive was 43.7% (95% CI 41.8–45.8). It was somewhat less in the City of Osh, 38.1% (95% CI 32.6–43.9). In contrast, the highest shares of seropositive volunteers were found in regions with the lowest population densities: the Naryn region with 53.3% seropositive (95% CI 49.8–56.8); and the Chui region with 51.3% seropositive (95% CI 47.9–54.8). We were unable to identify correlations between population density, seroprevalence, and morbidity. To a certain extent, this is probably due to the very uneven distribution of the population over the Republic's regions. About 20% of the total population is concentrated in the two largest cities (Bishkek, Osh); only 8.3 to 8.6% live in the Issyk-Kul, Naryn, or Chui regions (Table 1). Such a significant range in numbers suggests that in most of the country, morbidity and seroprevalence dynamics mainly develop in a random manner, without forming patterns. A significant proportion of rural residents permanently work in cities, where the likelihood of contact is higher. After work, they return to rural areas, potentially being passive carriers. This lifestyle may be one of the reasons for the heterogeneity of SARS-CoV-2 Ab seroprevalence.

Fig. 7
figure 7

Heat map of the regional distribution of SARS-CoV-2 Nag Ab seroprevalence in the Kyrgyz Republic. Seroprevalence levels (%) are shown on the tags. Color intensities correspond to percent seropositive individuals

Full size image

When assessing the influence of occupational factors on seroprevalence, the largest proportion of seropositive individuals was found among health-care workers. The results obtained are in good agreement with data from similar studies. In a significant number of publications, increased seroprevalence has often been noted in health-care workers who have frequent contact with COVID-19 patients in infectious disease departments and hospitals; they are classified as at risk regarding nosocomial SARS-CoV-2 infection [32,33,34, 36, 37]. In addition, high Nag Ab seroprevalence among health-care workers may be associated with high usage of the Sinopharm vaccine, which induces anti-Nag Ab production.

Significantly lower seroprevalence values were noted in business workers, those in manufacturing, the unemployed, and in persons in the classification 'Other' (all differences statistically significant, p < 0.01). Unfortunately, it was not possible to accurately analyze seroprevalence among agricultural workers, although this area of activity plays a significant role in Kyrgyzstan. Land in the Republic is exploited privately, and people engaged in agriculture are the owners of the land. As such, they do not consider themselves to be officially employed citizens. In our study, agricultural workers classified themselves as 'Unemployed' and 'Other' in the survey, with a lower seroprevalence than the cohort average. In addition, transhumance (pastoral) workers were unable to participate in the study.

Convalescents play a significant role in infectious spread and perpetuation of the epidemic process. In the clinical and post-clinical phases, individuals are able to transmit the pathogen to others by shedding of viral particles through saliva when coughing, sneezing, or talking. Among volunteers who indicated contact with a sick or recovering COVID-19 individual, the seroprevalence rate did not differ from the cohort average. It is clear that a known convalescent contact increases the risk of seropositivity. Whether this seropositivity is the result of an asymptomatic infection or increased vaccination readiness (due to awareness of increased personal risk) awaits clarification.

Seroprevalence among those who experienced manifest COVID-19 was significant [64.5% (95% CI 61.9–67.1)], but still not as high as one might expect. There are at least three explanations for this situation. Firstly, the data on infection are self-reported, as obtained from the questionnaire completed by the volunteer only. Secondly, even manifest COVID-19 does not always leave a trace in the form of detectable anti-Nag Ab levels. It is believed that about 35% of COVID-19 survivors produce Abs at concentrations below the minimum threshold of detection [35]. Thirdly and finally, there is a certain time period of Ab circulation after which concentrations decrease below methodological sensitivity thresholds. According to some reports, SARS-CoV-2 specific Ab levels begin to decline in the blood 2–3 months post-infection [36].

A characteristic feature of SARS-CoV-2 is a high proportion of asymptomatic forms of infection; their share can reach 90% or more. In this study, the share of asymptomatic SARS-CoV-2 infection among seropositive volunteers was 77.1% (95% CI 75.6–78.5), with a significant predominance of such forms among children 1–17 years old [92.2% (95% CI 88.2–95.3)] and adults 18–29 years old [81.8% (95% CI 77.7–85.4)]. In general, the level of such individuals (asymptomatic seropositive) is close to that of the neighboring Eurasian country Russia [26, 28, 38, 39].

In our study, about 25% of volunteers indicated a history of vaccination in the questionnaire, which is more than twofold higher than the official figure: 11.4% of the population as of 09/13/2021 [5]. This could be due to the fact that a significant proportion of the volunteers were healthcare workers vaccinated as a risk group. In addition, vaccinated residents were more motivated to participate in the study, as it allowed them to evaluate the effectiveness of vaccination. Among the vaccines permitted in Kyrgyzstan, Sinovac is the most widely used. At the same time, the distribution of vaccinated volunteers was uneven. Among the population of different ages, the largest share vaccinated was noted among volunteers of middle and older ages (30–69 years old), who accounted for 82.6% (95% CI 81.1–82.8). Children were least vaccinated [0.9% (95% CI 0.5–1.3)], which is explained by the later commencement of immunization. In regional terms, volunteers from the Osh region and the capital Bishkek, located in the Chui region, were most actively vaccinated, with health-care workers, followed closely by retirees, being well represented. The greatest compliance of health-care workers (64.3% vaccinated) is likely associated with an understanding of the need to develop SARS-CoV-2 immunity, which is in line with the global trend [40]. Thus, professional field had a significant impact on SARS-CoV-2 vaccination compliance in general and vaccine choice (among those approved for use in Kyrgyzstan). Health-care workers were predominantly vaccinated with Sinopharm in connection with state policy on vaccination of at-risk groups.

Assessment of seroprevalence in vaccinated volunteers (who did not indicate a history of COVID-19 in the questionnaire) showed that Sputnik V and Sinopharm produced comparable Ab seroprevalence: at least 91.6% (95% CI 90.3–92.8) of volunteers seroconverted with Sinopharm vaccination; and at least 90.6% (95% CI 87.3–93.3) seroconverted with Sputnik V. The fact that Sputnik V induces only anti-RBD Abs led to the conclusion that about half of those vaccinated had had asymptomatic infection (based on no indication of illness in the questionnaire). In these individuals, the response to vaccination was more pronounced [96.1% (95% CI 92.2–98.4)] relative to those not previously infected [85.9% (95% CI 80.5–90.3)].

The proven efficacy of the vaccines used convincingly indicates that there is no alternative to mass vaccination in order to achieve the threshold level of Ab seroprevalence, which ranges from 60 to 85% of the population [7, 8]. Vaccination of the Kyrgyz population at the time of writing was still well below the required level, which poses a risk of further increases in COVID-19 incidence.

Conclusion

Seropositivity for SARS-CoV-2 Nag, excluding the contribution of anti-RBD antibodies, was 48.7% (95% CI 47.7–49.7). The largest share of seropositive persons was noted among volunteers aged 50–70+ years. Statistically significant regional differences were noted in the Republic. Specifically, seroprevalence was significantly lower in Osh City, and significantly higher in the Naryn region, than the national average (p < 0.0001).

The highest level of seroprevalence, detected in the health-care worker group [57.1% (95% 55.4–58.8)], significantly exceeded the national average [50.1% (95% CI 49.0–51.2) p < 0.0001]. The share of asymptomatic individuals among those who were seropositive was 77.1% (95% CI 75.6–78.5). Maximums were observed in children [92.2% (95% CI 88.2–95.3)] and young people 18–29 years old [81.8% (95% CI 77.7–85.4)]. The differences were statistically significant at p < 0.0001.

Comprehensive analysis shows a satisfactory level SARS-CoV-2 Ab seroprevalence in the Kyrgyz population in the context of a relatively low COVID-19 incidence. Mass vaccination is undoubtedly having a positive impact on Ab seroprevalence while somewhat hindering transmission. Nevertheless, the pace of this process is still insufficient, which represents a risk of increased COVID-19 incidence.

Limitations of the study

The authors would like to note several factors that might affect sample representativeness or conclusions reached through data analysis. Despite the fact that an information campaign was carried out as widely as possible for the population (state television channels, news sites, Facebook), limited Internet access is a problem for parts of the rural population. Furthermore, residents who are more involved with their health and that of their loved ones (primarily women and health-care workers) are more likely to take part in studies of this kind. Health-care workers are well motivated, partially due to frequent contacts with COVID-19 patients. Since the study included an assessment of post-vaccination seropositivity, vaccinated residents were also more interested in participating. These factors may form a bias toward greater representation of women, health-care workers, and vaccinated residents. The fact that health-care workers were predominantly vaccinated with Sinopharm, which induces anti-Nag Ab production (similar to COVID-19 recovery), may also have affected estimated seroprevalence structure in the occupational group.