key: cord-0806553-y19pyrm1
authors: Bhuiyan, Taufiqur Rahman; Akhtar, Marjahan; Akter, Aklima; Khaton, Fatema; Rahman, Sadia Isfat Ara; Ferdous, Jannatul; Nazneen, Arifa; Sumon, Shariful Amin; Banik, Kajal C; Bablu, Arifur Rahman; Alamgir, A S M; Rahman, Mahbubur; Tony, Selim Reza; Hossain, Khaled; Calderwood, Stephen B.; Charles, Richelle C.; Ryan, Edward T.; LaRocque, Regina C.; Harris, Jason B.; Rahman, Mustafizur; Chakraborty, Nitai; Rahman, Mahmudur; Arifeen, Shams El; Flora, Meerjady Sabrina; Shirin, Tahmina; Banu, Sayera; Qadri, Firdausi
title: Seroprevalence of SARS-CoV-2 antibodies in Bangladesh related to novel coronavirus infection
date: 2022-02-02
journal: IJID Regions
DOI: 10.1016/j.ijregi.2022.01.013
sha: 9dfccb827e18e35c43202d90a0d6f85db25a8e54
doc_id: 806553
cord_uid: y19pyrm1

Design A cross-sectional study was conducted in household members in 32 districts of Bangladesh to build knowledge on disease epidemiology and seroepidemiology of COVID-19. Objective We assessed antibody responses to SARS-CoV-2 in people between April-October, 2020. Results Nationally, we estimated a seroprevalence of 30.4% for IgG and 39.7% for IgM. In Dhaka city, seroprevalence for IgG was 35.4% in non-slum areas, while it was 63.5% in slums. In areas outside of Dhaka, the seroprevalence rate for IgG was 37.5 and 28.7%, respectively, in urban and rural areas. Between April and October, 2020, the highest seroprevalence rate (57% for IgG and 64% for IgM) was observed in August. IgM antibody was more prevalent in the younger age participants, while older participants had more frequent IgG seropositivity. Follow-up specimens from COVID-19 patients and their household members suggested that both IgG and IgM seropositivity increased significantly at day 14 and 28 compared to day 1 of enrollment. Conclusions Our findings indicate that there has been an extensive spread of SARS-CoV-2 infection in Bangladesh by October 2020. This highlights the importance of monitoring seroprevalence data, particularly with the emergence of new SARS-CoV-2 variants over time.

The World Health Organization first identified a case with pneumonia on December 31, 2019 in Wuhan, China (Guo et al., 2020 , WHO, 2020 ) that was later confirmed as a novel coronavirus disease, 2019 . The Government of Bangladesh (GoB) reported the first COVID-19 case in Bangladesh on March 8, 2020 (GARDAWORLD, 2020 . As of June 29, 2021, a total of 896,770 confirmed cases have been identified in Bangladesh, including 14,276 deaths (Management Information System Directorate General of Health Services. , 2021) . Bangladesh is estimated to be at a high risk for COVID-19 due to its population density, poor sanitary practices, and limited infrastructure and infection control measures. As of July 2021, a second wave of COVID-19 occurred between April to May with 90% due to the beta variant of SARS-CoV-2. and more recently by August 2021, 100% of positivity in cases is due to the delta variant with a death rate higher than seen in the first wave in 2020 (Rahman et al., 2021) . To better ascertain COVID-19 burden in Bangladesh, the Institute of Epidemiology, Disease Control and Research (IEDCR) directed a national level investigation to evaluate COVID-19 prevalence in Bangladesh, in collaboration with the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) ; with the support from the United States Agency for International Development (USAID) and the Bill and Melinda Gates Foundation (BMGF).

Much of the epidemiological information about this newly emerging disease still remains unknown, including estimates of the proportion of COVID-19 cases in the community, particularly for lower-income regions and countries like Bangladesh, making it difficult for government policymakers to design optimal containment and mitigation strategies. Prior research has also indicated that there may be a considerable number of asymptomatic cases of COVID-19 (Anderson et al., 2020) . Other areas requiring further exploration include the incidence rate, the prevalence rate, the secondary infection rate, the incubation period, the serial interval and the reproductive number (R 0 ) of COVID-19 in various settings. Although there have been attempts at generating some of these data in previous studies worldwide, most estimates have been based on small-scale data or on information collected from relatively narrow geographic regions (Anderson et al., 2020) . It is also important to determine and characterize the immune responses to this novel coronavirus infection to understand how well the response protects people against future SARS-CoV-2 infection and how long this protection lasts (Sutton et al., 2020) . In this context, serological investigation has the potential to provide information about the true number of SARS-CoV-2 infections, allowing for robust estimates of the infection fatality rates (Fontanet, 2020) and guide public health decision-making. Therefore, we conducted a nationwide seroprevalence study of SARS-CoV-2 in Bangladesh, with follow-up data on COVID-19 cases and their household members, in order to enhance knowledge of the seroepidemiology of SARS-CoV-2 in Bangladesh.

This first national-level cross-sectional study in Bangladesh incorporated data from April 2020 to October 2020. To assess the seroprevalence of SARS-CoV-2 infection in Dhaka city, a total of 25 wards were randomly selected out of the 129 wards, one mahalla (the smallest geographical unit of urban area) was randomly selected from each ward, and from each mahalla, 120 households were randomly selected. To evaluate the seroprevalence in slum areas, an additional eight slums from Dhaka city were included in the study. Each of the participants was asked if they had any of the four probable COVID-19 symptoms: fever (having body temperature >38°C), cough, sore throat, breathing difficulties within the last seven days and other socio demographic information (e.g. shared bathroom, household size, number of living rooms and person tested for Covid-19 before the survey) on the day of the interview or within 7 days preceding the sample collection (Nazneen A, 2021) .

In addition, 32 districts were randomly selected including Dhaka out of the 64 districts of Bangladesh. From each selected district, we randomly selected one village (the smallest geographical unit of the rural area) and one mahalla (the smallest geographical unit of urban area) for data collection.

From each participant in Dhaka and outside Dhaka, nasopharyngeal swab and 3-4 ml blood was collected. From Dhaka city, we also followed up the SARS-CoV-2 RT-PCR positive participants, as determined from nasopharyngeal swab samples collected at enrollment day (day 1) and measured serological responses in their household members on days 1, 14 and day 28 after enrollment.

Blood specimens were collected from study participants at the time of the interview. We also collected blood specimens on days 14 and 28 from household members of individuals in Dhaka city who were RT-PCR positive. Serum was separated from whole blood after centrifugation of the tubes at 700x g for 15 minutes and kept frozen (-80 O C) until the time of laboratory analysis.

An enzyme-linked immunosorbent assay (ELISA) procedure was used to determine SARS-CoV-2 specific IgG and IgM antibodies against COVID-19 from the collected serum samples as below.

For analysis of antibodies to SARS-CoV-2 in this study, COVID-19 specific antibody measurements were performed using an in-house ELISA assay for IgG and IgM isotypes against the receptor binding domain (RBD) of the spike protein of SARS-CoV-2 (Iyer et al., 2020 , Shirin et al., 2020 . RBD-specific antibody concentrations (ng/mL) were quantified using isotypespecific anti-RBD monoclonal antibodies. Briefly, 96 well Nunc® MaxiSorp™ plates (ThermoFisher) were coated with 100 µL of SARS CoV-2 RBD antigen (1μg/mL in carbonate buffer) and incubated for 1 hour at room temperature (RT). Plates were blocked for 30 minutes at RT with 300 µL of 5% nonfat milk in phosphate-buffered saline (PBS). Heat-inactivated serum samples (serially diluted samples 1:100, 1:400, 1:1600 and 1:6400 in 5% Milk-1x PBS 0.05% Tween) were added to the plate (100 μL/well) and incubated for 1 hour at 37°C. A specific monoclonal antibody to RBD of known concentration (Mab CR3022) was added to the plate; two-fold serial dilutions were performed starting at 25 ng/ml for both the IgG and IgM monoclonal antibodies. Individual serum samples were tested and quantified to determine the concentration of specimens based on the monoclonal antibody. At the end of incubation, plates were washed 5 times with PBS-0.05% Tween (PBST). Goat anti-human IgG, and IgMhorseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch) diluted at 1:5000 in 5% milk in PBST were added to plates (100 μL/well) and incubated at RT for 30 minutes followed by 5 washes with PBST and one wash with 1X PBS. Bound secondary antibodies were detected using ortho phenylenediamine (OPD; Sigma, 200 μL/well) in 0.1 M sodium citrate buffer (pH 4.5) and 30% H 2 O 2 . Plates were incubated at RT for 20 minutes in the dark. Optical density (OD) was measured at 450 nm and 570 nm in the Eon (Biotek) ELISA Reader; OD values were adjusted by subtracting the OD at 570 nm from the 450 nm OD.

Before testing current study specimens, we performed ELISA using serum specimens collected from RT-PCR positive COVID-19 patients (n=38), pre-pandemic healthy controls collected from Dhaka (n=73) as well as Khulna and Rajshahi Divisions (n=203) to determine cut-off of the different isotypes of antibodies and validate the assay. We also tested sera obtained from influenza patients (n=59) and Japanese Encephalitis surveillance platform (n=20) patients (Shirin et al., 2020) and compared the responses with commercially available ELISA kits (Euroimmun, Germany for IgG; Wantai, China for IgM). For IgG, we observed 100% sensitivity, 98% specificity, and 97% positive predictive value, and for IgM, we observed 89% sensitivity, 95% specificity, and 94% positive predictive value. To determine the cut-off for seropositivity, we used the upper limit of the range of concentrations of SARS-CoV-2 IgG and IgM antibodies from random available pre-pandemic serum samples (n=200) collected from different locations of Bangladesh (before 2019). Based on this, we determined 500 ng/ml (0.5 µg/ml) as a cut-off value for both IgG and IgM antibodies (Supplementary Table 1 and 2).

For each stage of sampling, we estimated the selection probability of sampling units using the existing sampling frame and then multiplied all the selection probabilities obtained from each stage to estimate the selection probability for each selected individual. We then estimated individual level sampling weight by taking inverse of the selection probabilities. Finally, we normalized the weights and used to estimate the parameters. We estimated the seroprevalence of SARS-CoV-2 antibodies with 95% confidence interval (CI), adjusting for design weight and clustering effects. For the concentration, we presented geometric mean (GM) with 95% CI. A line plot presented over the study period for the seroprevalence estimated by 15 days interval;

Pearson chi-square test is used to measure the difference of seroprevalence at different study groups and two GMs considered statistical difference at 5% level of significance if 95% CI were non-overlapping. Comparison of antibody responses (IgG and IgM) on different days in household members of the RT-PCR positive participants were analyzed by the Mann-Whitney U test. P values <0.05 were considered statistically significant. All analyses were performed using the GraphPad Prism 6.0 and STATA software version 15. Clusters and individuals within each cluster were selected using probability sampling. For each stage of sampling, we estimated the selection probability of sampling units using the existing sampling frames and then multiplied all the selection probabilities obtained at different stages to determine selection probability of each selected individuals. We then estimated individual level sampling weight by taking inverse of the selection probabilities and normalized them. As appropriate weights were used, the estimates are generalizable to the rest of the country.

The Research Review Committee and the Ethical Review Committee of icddr,b approved the study protocol. The institutional review board of IEDCR also approved the study. Informed written consent was obtained from all the participants before enrollment and data collection.

We collected demographic information from all participants (Table 1) . We collected blood specimens from 2582 participants at the national level, comprising Dhaka city (n=701, non-slum areas) as well as rural (n=819) and urban areas (n=1062) outside Dhaka. In addition, we also collected blood specimens from the selected people (n=126) living in slum areas of the Dhaka city. Gender distribution was comparable including 51% male and 49% female participants.

Since this survey is based on households, we were able to enroll participants from all age groups.

Among the study population, 20% of participants were less than 20 years old. The largest proportion of participants (41%) were between 20-39 years old. Based on recorded symptoms, 14% of people had fever, 11% had cough, 3% had sore throat and 2% had shortness of breathing on the day of enrolment and within the past 7 days. The majority of participants (83%) did not use shared washrooms and 47% of enrolled households comprised ≥4 members. Almost all participants (99%) had not been tested for COVID-19 before being included in this study ( Table   1) .

At the national level, we found 30.4% of the population had IgG responses against the SARS-CoV-2 RBD antigen and 39.7% had IgM antibodies in Bangladesh between April and October, 2020 (Table 2) . Overall, 51.8% of individuals tested seropositive against SARS-CoV-2 either/or for IgG and/or IgM antibody in Bangladesh during above mentioned period. During the study period, significantly higher concentrations of IgM antibodies (geometric mean 355 ng/ml) were observed in comparison to IgG (geometric mean 176 ng/ml).

IgG Seroprevalence varied between the type of study site. It was highest in the slum areas of Dhaka city (64%), followed by urban areas outside Dhaka (38%), non-slum areas of Dhaka city (35%) and rural areas of the outside Dhaka (29%) and. This higher seropositivity rate in individuals living in slum areas may be influenced by the fact that enrollment there was started later (July and August) than in the non-slum areas of Dhaka city. In the areas outside Dhaka, the seroprevalence rate was somewhat higher in the people living in the urban areas than in the rural areas (Table 2) .

We observed a comparable seroprevalence rate of SARS-CoV-2 specific IgM antibodies in all age groups (Table 3) . However, IgG seropositivity increased with age. 8-24% of the participants below 20 years of age had IgG antibodies, while 32-36% of participants above 20 years of age had IgG antibodies. There was no significant difference in the seroprevalence rate between males and females, although female participants had a somewhat higher rate of IgG (33% vs. 28%) and

IgM seropositivity (48% vs. 32%) compared to males (P>0.05).

We analyzed the monthly seropositivity rates (both for IgG and IgM antibody) in Bangladesh between April and October 2020 (Figure 1 ). SARS-CoV-2 IgM antibodies were very low in the population in late April 2020. However, we observed ~23% of the population were seropositive by the end of April 2020 for SARS-CoV-2 IgG antibody. As we detected the first case of COVID-19 in Bangladesh on March 8, 2020, this may explain the IgG antibody responses seen in April 2020. As infections increased and community transmission occurred over the subsequent months in Bangladesh, we observed a continued rise in seropositivity rates, with the highest rates in August, both for IgG (57%) and IgM (64%) antibody. The infection rate decreased between August and October 2020 (Bangladesh, WHO), and we saw decreasing seropositivity over these months. By October 2020, approximately 36% of the people of Bangladesh were seropositive for either SARS-CoV-2 IgG and 43% were for IgM antibodies.

In a subset of participants (n=252) in Dhaka city, PCR positive cases of COVID-19 and their household members, we evaluated antibody responses at days 1, 14 and 28 of available individuals following enrollment. We observed that 78% of these participants were seropositive by day 28 for SARS-CoV-2 IgG, and 60% of the participants were seropositive for SARS-CoV-2 IgM antibodies (Table 4 ). The geometric mean levels of antibody responses both for IgG (GM: 852) and IgM (GM: 538) also increased significantly by day 28 as compared to day 1 (Table 4) .

There is a critical need for serological surveillance of SARS-CoV-2 to estimate cumulative prevalence, incidence, and community distribution in Bangladesh. The preliminary seroprevalence results of the current study provide an important benchmark to assess the state of the COVID-19 epidemic. By assessing the presence of SARS-CoV-2 IgG and IgM antibodies measured in this study, we estimate that the majority of the population have now been exposed to the virus in Bangladesh. Understanding the factors contributing to population immunity is a research priority, including factors enabling transmission in places with a high cumulative incidence particularly in LMICs.

The duration of immunity, the impact of different vaccines on transmission and herd immunity, the role of emerging variants and vaccine breakthrough are all key areas that need to be understood for control of the COVID-19 pandemic. Herd immunity (Clemente-Suárez et al., 2020) is mathematically related to the propagation and transmission dynamics of the pathogen (Fresnadillo-Martínez et al., 2013, Herrmann and Schwartz, 2020) , and the number of individuals in a population susceptible to infection (Fresnadillo-Martínez et al., 2013) . Herd immunity is expected to be obtained only when approximately 70% of the population has been infected or vaccinated (Clemente-Suárez et al., 2020) . Seroprevalence analyses in large populations and over time may be extremely important tools in understanding community transmission and informing vaccine design and rollout.

Our study has some important limitations. Our ELISA assay was not validated against a plaque reduction neutralization assay, which may provide important information about protection of an individual person. ELISA antibody assays have been found to correlate with neutralizing antibody (NAb) measurements (Holzmann et al., 1996) . However, the correlation between the anti-RBD IgG antibody assay and NAb remains unclear as there are contradictory reports on their association (Billon-Denis et al., 2021) in SARS-CoV-2 infected patients. It has been proposed that other host factors (including age, gender and clinical severity) may also be other drivers of the immune response including antibody and neutralizing antibody responses. It has been suggested that Nab competes with ACE2 receptor binding of the virus and so an may be a better predictor for virus-neutralizing antibody potency rather than binding affinity (Ju et al., 2020) . Hence, blocking the interaction between RBD and ACE2 may be a useful surrogate for neutralization. However, in the current seroprevalence study, the objective was to analyze population level seroprevalence only in different urban and rural areas in Bangladesh. We were not able to analyze cross reactivity against other coronaviruses due to unavailability of these antigens. The estimates we present may be affected by waning antibody levels over time but we were not able to track the cumulative infection rates during the study period which could also impact on the analysis. However, it should be noted that vaccination had not been initiated at the time of the study and therefore not likely to influence the seroprevalence estimates. Our results show high seroprevalence levels to SARS-CoV-2 early on in the pandemic (April-October 2020) suggesting protection in the population. However, we attempted to analyze data from selected areas in Bangladesh and our data needs to be carefully extrapolated and we should not generalize the seroprevalence levels to all of Bangladesh but rather to the population being studied. In summary, in this study, we report a significant increase in the seroprevalence of antibodies to SARS-CoV-2 among the Bangladeshi population during the first wave of COVID-19 between April and October, 2020, with 45% positive for IgG antibodies very soon after SARS-CoV-2 was detected in Bangladesh (~March 2020). The increase in seropositivity paralleled a substantial rise in the number of PCR-confirmed SARS-CoV-2 infections in Bangladesh (Management Information System Directorate General of Health Services, Bangladesh., 2021) . Moreover, throughout the study period, we also observed seropositivity of IgM antibodies against SARS-CoV-2. Since IgM is the primary antibody detected at the acute infection period (Akter et al., 2022) , which also indicated the increase in the number of new SARS-CoV-2 infections in the study population. We observed substantial differences in seroprevalence between slum and nonslum areas of Dhaka city and between urban and rural areas outside Dhaka city. We have similarly found substantial differences in seroprevalence between populations of low and high socioeconomic status in Bangladesh (Sattar et al. Submitted). There is a continuous interest of public health experts to understand and contextualize key drivers of COVID transmission and anticipate future risks and adopt preventive measures including vaccinations to address these problems. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. 

Investigators are grateful to the USAID for its support towards its research. icddr,b is also grateful to the Governments of Bangladesh, Canada, Sweden and the UK for providing core/unrestricted support. We also like to thank Aaron G. Schmidt, Jared Feldman, Blake M.

Hauser, and Timothy M. Caradonna from Harvard Medical School for providing the SARS-CoV-2 RBD protein and RBD monoclonal antibody. Authors are extremely thankful to all field and laboratory staffs who accomplished a wonderful job during the pandemic by collection of blood from the community and processed the blood in the laboratory. 

Disease characteristics and serological responses in patients with differing severity of COVID-19 infection: A longitudinal cohort study in

How will country-based mitigation measures influence the course of the COVID-19 epidemic?

Differential serological and neutralizing antibody dynamics after an infection by a single SARS-CoV-2 strain

Dynamics of Population Immunity Due to the Herd Effect in the COVID-19 Pandemic

Cluster of COVID-19 in Northern France: A Retrospective Closed Cohort Study. Prepint with The Lancet

First cases of COVID-19 confirmed March 8; 2020

The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak -an update on the status

Using network science to propose strategies for effectively dealing with pandemics: The COVID-19 example

Correlation between ELISA, hemagglutination inhibition, and neutralization tests after vaccination against tick-borne encephalitis

Persistence and decay of human antibody responses to the receptor binding domain of SARS-CoV-2 spike protein in COVID-19 patients

Human neutralizing antibodies elicited by SARS-CoV-2 infection

Coronavirus COVID-19 Dashboard

Prevalence of COVID-19 in Bangladesh

Contact Tracing during Coronavirus Disease Outbreak, South Korea

The emergence of SARS-CoV-2 variants in Dhaka city, Bangladesh. Transboundary and emerging diseases

Antibody responses after COVID-19 infection in patients who are mildly symptomatic or asymptomatic in Table 3: Weighted distribution of positive SARS-CoV-2 antibodies at the national level of Bangladesh among people of different age and gender

IgG, % (95% CI) IgM, % (95% CI)

Table 4: Frequency of antibodies in RT-PCR positive participants and their household members from Dhaka city at three different time points

Significantly higher seropositivity as well as concentration of both IgG and IgM antibodies found at day 14 and day 28, compared to day 1 at enrollment in the SARS-CoV-2 RT-PCR positive participants as well as their household members

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: