Ethno-demographic Disparities in humoral responses to the COVID-19 Vaccine among Healthcare Workers

Abstract

Introduction:

The COVID-19 pandemic had a profound impact on global health, but rapid vaccine administration resulted in a significant decline in morbidity and mortality rates worldwide. In this study, we sought to explore the temporal changes in the humoral immune response against SARS-CoV-2 healthcare workers (HCWs) in Augusta, Georgia, USA, and investigate any potential associations with ethno-demographic features. Specifically, we aimed to compare the naturally infected (NI) individuals with naïve individuals to understand the immune response dynamics over time.

Methods:

A total of 290 HCWs were included and assessed prospectively in this study. COVID status was determined using a saliva-based COVID assay. Neutralizing antibody (NAb) levels were quantified using a chemiluminescent immunoassay system, and IgG levels were measured using an enzyme-linked immunosorbent assay (ELISA) method. We examined the changes in antibody levels among participants using multiple statistical tests including logistic regression and MCA analysis.

Results:

Our findings revealed a significant decline in NAb and IgG levels at 8–12 months after SARS-CoV-2 vaccination. Furthermore, a multivariable analysis indicated that this decline was more pronounced in white HCWs (odds ratio [O.R]=2.1, 95% confidence interval [C.I]=1.07–4.08, p=0.02) and IgG (O.R=2.07, 95% C.I=1.04–4.11, p=0.03) among the whole cohort. Booster doses significantly increased IgG and NAb levels, while a decline in antibody levels was observed in participants without booster doses at 12 months post-vaccination.

Conclusion:

Our results highlight the importance of understanding the dynamics of immune response and the potential influence of demographic factors on waning immunity to SARS-CoV-2. In addition, our findings emphasize the value of booster doses to ensure durable immunity.

Introduction

The global coronavirus pandemic has infected over 676 million people and caused more than 6.8 million deaths since its emergence in November 2019 ¹. The unprecedented spread and magnitude of the pandemic led to the rapid deployment of resources to develop a vaccine. As a result, Pfizer-BioNTech BNT162b2 received FDA Emergency Use Authorization (EUA) on December 11, 2020, which was soon followed by FDA EUA of Moderna vaccine (mRNA-1273) ². As of December 2022, 50 vaccines have been approved or authorized for limited use to protect against serious illness, hospitalizations, and death induced by SARS-CoV-2 ³. In addition to mRNA vaccines, there are five other types of vaccines that have been approved by regulatory bodies in various countries against SARS-CoV-2. These vaccines use a range of biological entities, including adenoviral vectors, DNA, protein subunits, live viruses, and attenuated viruses ⁴. In addition, over 230 candidate vaccines are under various stages of testing and trials ³. As of March 2023, nearly 5.5 billion people worldwide, or nearly 72.3% of the world’s population, have received at least one dose of a SARS-CoV-2 vaccine ⁵. Extensive vaccination campaigns contributed to a significant decline in COVID-19 cases, serious hospitalizations, and deaths in the general population. Though critical in curbing the COVID-19 pandemic, little data was available on the safety and immune response triggered by these new-generation mRNA vaccines when first introduced to the general population.

Since then, studies have reported that exposure to SARS-CoV-2 or its components, either through natural infection or vaccination, stimulates both the humoral and cellular components of the immune system. The humoral response can be detected in the bloodstream and is linked to protection against future infection or reduced severity of disease ⁶. IgM antibodies are produced shortly within days after exposure to antigen, usually followed by a rise in IgG levels. This increase can be observed within roughly two weeks of the initial exposure, and the levels of IgG antibodies usually peak at around 25 days post-exposure. Following this peak, IgG antibody levels remain high for several weeks, indicating the role of long-term plasma cells in continuing to generate an immune response and sustaining the body’s protection long after the initial exposure ^7,8.

Neutralizing antibodies are a specific subset of antibodies with the capacity to neutralize or reduce the infectivity of pathogens or viruses by diverse mechanisms such as blocking virus uptake or inhibiting the uncoating of the viral genome ^9,10. Neutralizing antibodies are essential for reducing disease severity and have been identified as a critical correlate of protection against COVID-19 ^11–13.

The production of antibodies post-natural infection with SARS-CoV-2 can last several months for up to 16 months in most individuals ^14,15. Additionally, studies have indicated that individuals with severe symptoms tend to exhibit stronger immune responses compared to those with mild symptoms ¹⁶. It is therefore crucial to further explore the immune response elicited by SARS-CoV-2 and to examine the durability of this response in different populations. Moreover, differences in immune response between individuals with natural exposure to SARS-CoV-2 and those who are naive to the virus warrant investigation.

Due to their occupational exposure to COVID-19, healthcare workers (HCWs) were considered a priority group for vaccination and hence were used as a representative cohort to investigate the longitudinal dynamics of humoral immunity in a real-world scenario. The study involved both naturally infected (NI) individuals and SARS-CoV-2-naive HCWs, who were followed up to 18 months from the start of the study. The objective of the study was to investigate the immune response as measured by antibody titer of IgG and Neutralizing antibodies following SARS-CoV-2 vaccination and booster administration, as well as any potential associations with ethno-demographic features.

Material and Methods

Study design

In this prospective study within Augusta, Georgia, 625 HCWs were enrolled in the SeroPrevalence And Respiratory Tract Assessment (SPARTA) study at Augusta University, GA. The study was approved by the institutional review board of Augusta University (AU-IRB #1858403 and WIRB #1663744).

All the participants provided informed consent before the start of the study. Eligible individuals were HCWs from Augusta University and the surrounding area, between the ages of 18 and 90 years, weighed at least 110 lbs, were not pregnant, and did not have any conditions such as anemia or bloodborne infections like HIV or Hepatitis C. Of the 625 participants enrolled in the study, 290 were included in data analysis as they provided at least more than one blood sample during the study (Fig 1). Sequential samples could be obtained at six-time points (enrollment, post-vaccination (<3 months), 8–12 months, 12–14 months, 14–16 months, and 16–18 months). A total of 1,351 serological tests were conducted: 736 NAb and 615 IgG at various time points. Of the participants, 175 (60.34%) gave blood at enrollment, 162 (55.86%) gave blood post-vaccination, 172 (59.31%) between 8–12 months, 146 (50.34%) between 12–14 months, 63 (21.73%) between 14–16 months, and 21 (7.24%) between 16–18 months (Figure 1).

Figure 1:

Flow chart of the study.

Data collection

The Research Electronic Data Capture (REDCap) software was employed to securely store the study data, including baseline information collected at the initiation of the study and subsequent updates entered after each visit of the participating HCWs. Several clinical variables were collected at enrollment for the study. Patients with a positive COVID-19 test were included in the Natural Infection (NI) group, while those without a positive test were assigned to the naïve group. During the longitudinal assessment, only participants without breakthrough infections were included. Demographics collected were sex, age, race, body mass index (BMI), and preexisting medical conditions. Additionally, vaccination status for influenza and MMR vaccines was also documented.

Sample processing

De-identified pre-COVID samples were accessed from the Georgia Esoteric and Molecular (GEM) laboratory (Augusta University) and comprised of patient specimens that had been previously tested for cystic fibrosis and were determined to be of wild-type genotype. These samples were treated as a control to assess the specificity and threshold limit of our detection kits. For the prospective participants, blood was collected at multiple time points, centrifuged to collect serum, and stored at −80°C for further analysis.

SARS-CoV-2 saliva assay

For SARS-CoV-2 detection, we used a saliva-based PCR method standardized in our laboratory ¹⁷. Briefly, saliva was collected and transported to the GEM lab for further analysis. The SARS-CoV-2 detection assay employed is based on nucleic acid extraction followed by a TaqMan-based real-time PCR assay to conduct in vitro transcription of SARS-CoV-2 RNA, DNA amplification, and fluorescence detection (PerkinElmer Inc., Waltham, MA). The assay targets specific genomic regions of SARS-CoV-2: nucleocapsid (N) and ORF1ab genes. The TaqMan probes for the two amplicons are labeled with FAM and ROX fluorescent dyes, respectively, to generate target-specific signals. The assay includes an RNA internal control (IC; bacteriophage MS2) to monitor the processes from nucleic acid extraction to fluorescence detection. The IC probe is labeled with VIC fluorescent dye to differentiate its fluorescent signal from SARS-CoV-2 targets. The samples were classified as positive or negative based on the cycle threshold (Ct) values specified by the manufacturer.

Anti-spike IgG antibody measurement assay

The enzyme-linked immunosorbent assay (ELISA) was used for the determination of serum IgG antibodies against SARS-CoV-2 (Anti-SARS-CoV-2 QuantiVac ELISA (IgG); EUROIMMUN). Briefly, serum was diluted to 1:100 with dilution buffer and 100μl was incubated for 1 h in a microplate strip well coated with recombinant S1 domain (spike protein) of SARS-CoV-2. The wells were washed three times with wash buffer prepared by 10X dilution and incubated with secondary anti-human IgG antibodies for 30 minutes. The wells were washed again and were treated subsequently with TMB and stop solution and were read at 450 nm on the ELISA reader. All experiments were performed in triplicates and the readings were interpreted as per the manufacturer’s guidelines ¹⁸.

Anti-SARS-CoV-2 neutralizing antibody assay

Anti-SARS-CoV-2 NAb levels were determined using the SuperFlex^™ Anti-SARS-CoV-2 Neutralizing Antibody Kit and SuperFlex^™ Chemiluminescent Immunoassay System (PerkinElmer Inc. Waltham, USA). Briefly, the assay utilizes superparamagnetic microparticles together with direct chemiluminescence technology to detect anti-SARS-CoV-2 NAbs in serum. The sample is automatically added to the sample well, mixed, and incubated with the acridinium ester labeled RBD (receptor-binding domain) antigen for 15 minutes. The magnetic particles coated with human ACE2 protein were added to form a competitive model. The anti-SARS-CoV-2 NAb in the sample and the human ACE2 (angiotensin-converting enzyme-2) protein coated on the magnetic particles compete for binding to the RBD domain-containing antigen labeled with acridinium ester. After incubating for 15 minutes, the unbound substance was removed by washing, and the luminescence value of the chemiluminescence reaction was measured. The luminous intensity was negatively correlated with the concentration of the anti-SARS-CoV-2 NAb in the sample.

Statistical analysis

Descriptive statistics for the continuous variables in the study were presented as the median and interquartile range (IQR). To evaluate the distribution of the categorical variables, we presented relative frequency with percentages. Non-normally distributed continuous variables were analyzed using the Wilcoxon rank-sum test. Categorical variables were tested using either the chi-square test or Fisher’s exact test (for categorical variables with expected cell sizes less than 5). To compare the differences between different time points, the Wilcoxon signed-rank test was utilized. The relationship between independent clinicopathological variables and antibody levels was analyzed using Spearman’s correlation test for distribution in antibody levels. In this study, a decline in antibodies was defined as a 2-fold decrease between the post-vaccine (<3 months post-vaccine) and 8–12 months after receiving two doses of the COVID-19 vaccine. Univariate and Multivariate Logistic regression models were developed based on dichotomized antibody decline variables. Multiple Correspondence Analysis (MCA) was performed to investigate the relationships between seven categorical variables. The eigenvalues obtained from the singular value decomposition of the correspondence matrix in MCA represented the amount of variation captured in each dimension. The top-ranked dimensions, which explained the majority of data variation, were plotted using eigenvalues. MCA has been widely used in various studies, including cancer and COVID-19 studies ^19–23. The standardized score or z-score was calculated at each time point using the formula z = (x-μ)/σ, where x represents the raw value, μ represents the mean, and σ is the standard deviation. The statistical analyses were conducted using R (version 4.2, R Foundation for Statistical Computing, Vienna, Austria), JMP (version 16.0, SAS Institute, Cary, USA), and GraphPad Prism (version 9, GraphPad Software, La Jolla California USA).

Results

Ethno-demographic distribution of the participants in the study

This study included 290 participants, 172 participants (59%) were naïve, and 118 participants (41%) were exposed to SARS-CoV-2 before enrollment. Of all enrolled individuals, 194 (67%) were female and 96 (33%) were male. The distribution of participants according to age in different ranges was as follows: 20–30 years, with 70 participants (24.1%); 30–40 years, with 46 participants (15.8%); 40–50 years, with 62 participants (21.3%); 50–60 years, with 56 participants (19.3%); 60–70 years, with 34 participants (11.7%); 70–80 years, with 16 participants (5.5%); and 80–90 years, with 4 participants (1.3%). In this study, 230 participants (79%) reported no chronic underlying conditions, while 60 (21%) had one or more underlying conditions. The most common chronic conditions were hypertension in 13 (4.5%) and allergies in 8 (2.8%). In addition to these variables, Table 1 presents the distribution based on race, vaccine dose, flu vaccine status, and BMI, providing complete ethnical and demographic features.

Table 1:

Ethno-demographic features of participants in this study.

Characteristic	N = 2901
Group
Naive	172 (59.31%)
Naturally infected (NI)	118 (40.69%)
Sex
Female	194 (66.90%)
Male	96 (33.10%)
Age
20 — 30 years	70 (24.13%)
30 — 40 years	46 (15.86%)
40 — 50 years	62 (21.38%)
50 — 60 years	56 (19.32%)
60 — 70 years	34 (11.72%)
70 — 80 years	16 (5.52%)
80 — 90 years	4 (1.38%)
Missing	2 (0.69%)
Race
White	174 (60.00%)
Black	57 (19.65%)
Asian	29 (10.00%)
Hispanic or Latino	15 (5.17%)
native Hawaiian	1 (0.34%)
American Indian	7 (2.42%)
Missing	7 (2.42%)
Flu vaccine
Received	263 (90.68%)
Not received	22 (7.59%)
Missing	5 (1.73%)
MMR vaccine
Received	187 (64.48%)
Not received	35 (12.07%)
Missing	68 (23.45%)
BMI
Normal	106 (36.56%)
Obese	90 (31.03%)
Over-weight	90 (31.03%)
Missing	4 (1.38%)
Chronic conditions
Present	60 (20.68%)
Absent	230 (79.32%)
Conditions
Allergies	8 (2.75%)
GERD	4 (1.37%)
Hypertension	13 (4.49%)
T2D	6 (2.07%)
None	230 (79.32%)
others	29 (10.00%)

¹n (%)

Variables associated with baseline

The distribution of COVID-19-associated symptoms was assessed in the 118 NI participants. The most typical symptoms of COVID-19, including fever, runny nose, loss or decrease in sense of taste, muscle or body aches, fatigue, and headache, were also analyzed. Among 113 NI participants for which the symptoms information was available, 28% had ≤ 2 symptoms, 26% had 3–4 symptoms, and 46% reported 5–6 symptoms. The median duration of major symptoms was 11 days (Figure 2c). Those that reported having had a longer symptom duration experienced a greater number of symptoms, with shortness of breath and loss of taste being the most commonly reported, in 17 and 21 individuals, respectively. Conversely, sore throat was the most prevalent symptom in patients with shorter symptom duration. (Figure 2a). At baseline, participants who reported more symptoms had significantly higher levels of neutralizing antibodies (20.08 vs. 57.33 vs. 99.9 ng/ul, p < 0.05, Kruskal–Wallis test) compared to those with fewer symptoms. However, there was no significant difference in IgG levels (31.07 vs. 63.80 vs. 54.66 ng/ul, p > 0.05, Kruskal–Wallis test) among the groups (Figure 2d and 2f). The median time between the onset of COVID-19 and the baseline visit was 16 weeks (Figure 2e). The pre-COVID-19 control samples had a median of 0.19 ng/ul of NAb and 0 RU/ml of IgG (Supplementary Figure 1). At baseline, there was a significant difference in neutralizing antibody and IgG levels between the naïve and NI groups (Figure 2g, 2h). Participants with NI had higher NAb levels (median = 35 ng/ul) compared to those who were naïve (median = 0.25 ng/ul). Furthermore, IgG antibodies (54 RU/ml) were present in the NI group but absent in the naïve group at baseline (Table 2).

Figure 2:

Analysis of ethno-demographic features, symptoms and Antibodies in Naturally Infected Participants. (a) Distribution of symptoms and their duration in naturally infected (NI) participants (b) Prevalence of major COVID symptoms in the NI group (c) Length of symptoms duration in the NI group (d) Distribution of NAb antibodies in participants with symptoms (e) Time interval between COVID illness and baseline visit in participants with symptoms (f) Distribution of IgG antibodies in participants with symptoms (g) Comparison of NAb levels in the NI and naïve groups (h) Comparison of IgG levels between the NI and naïve groups (i) Correlation of antibody levels with other factors. The solid horizontal lines represent the median value.

Statistical significance was defined at a p-value of 0.05, **** denotes p ≤ 0.0001.

Table 2:

The comparison of NAB and IgG levels in participants at three-time points, baseline, post-vaccine, and 8–12 months pos-vaccine.

Variable	N	naive, N = 1721	NI, N = 1181	p-value2
NAB baseline	174	0 (0, 0)	35 (10, 120)	<0.001
NAB post vaccine	162	1,819 (1,090, 2,081)	2,112 (1,611, 2,395)	0.002
NAB 8–12 months	172	836 (253, 1,495)	1,474 (568, 2,059)	0.002
IgG baseline	103	0 (0, 0)	54 (16, 89)	<0.001
IgG post vaccine	159	188 (176, 202)	192 (178, 216)	0.11
IgG 8–12 months	160	125 (86, 157)	150 (122, 174)	0.008

¹Median (IQR)

²Wilcoxon rank sum test

A correlation analysis was conducted to examine the relationship between clinicopathological variables and antibody levels at baseline in NI participants (Figure 2i). NI participants had higher levels of IgG and Nab than naïve participants. Furthermore, a significant negative correlation was found between the number of post-infection days and baseline IgG levels (r=−0.38, p=0.002). Additionally, there was a positive correlation between the number of symptoms reported and higher neutralizing antibody levels at baseline (r=0.25, p=0.02).

Variables associated with post-vaccine decline.

The antibody analysis identified that the administration of two doses of the COVID-19 vaccine resulted in a significant increase in the production of both NAb and IgG antibodies (Table 2 and Figure 3). We found that after vaccination, participants with prior exposure to SARS-CoV-2 (NI) exhibited significantly higher median levels of neutralizing antibodies compared to naïve participants (1819 ng/ μl and 2112 ng/ μl, respectively; p<0.001, Wilcoxon test) (Figure 3a). Compared to baseline levels, there was a significant increase in NAb levels in both the prior exposed and naïve participant groups after the vaccination regimen (Figure 3b and 3c). However, at 8–12 months post-vaccination, there was a significant decline in neutralizing antibody levels in both the NI and naïve participants (Figure 3d, 3e, and 3f). The decline was more pronounced in the naïve group compared to the NI group.

Figure 3:

Analysis of Antibody Responses after 2-doses of COVID-19 at post vac (<3 months)and decline (8–12 months) (a) Distribution of NAb levels post-vaccination (b) Increase in NAb levels in naturally infected (NI) participants post-vaccination (c) Increase in NAb levels in naïve participants post-vaccination (d) Distribution of NAb levels 8–12 months after vaccination (e) Decrease in NAb levels in NI participants (f) Decrease in NAb levels in naïve participants (g) Distribution of IgG levels post-vaccination (h) Increase in IgG levels in NI participants post-vaccination (i) Distribution of IgG levels 8–12 months after vaccination (j) Decrease in IgG levels in NI participants (k) Decrease in IgG levels in naïve participants. The Wilcoxon signed-rank test was used to assess differences between paired samples. The solid horizontal lines represent the median value. Statistical significance was defined at a p-value of 0.05, where p > 0.05 indicates non-significant (ns), ** denotes p ≤ 0.01, and **** denotes p ≤ 0.0001.

The post-vaccination IgG levels did not differ significantly between the NI and naïve groups (188 RU/ μl and 192 RU/ μl, respectively; p>0.05, Wilcoxon test) (Figure 3g). However, there was a significant increase in IgG levels in the NI group from baseline to post-vaccination (Figure 3h). At 8–12 months post-vaccination, there was a significant decline in IgG antibody levels in both the NI and naïve groups (Figure 3i, 3j, and 3k). However, the decline was much more pronounced in naïve group compared to the NI. Further, other than previous infection status and ethnicity, none of the other ethno-demographic features showed an association with antibody levels (Supplementary Table 3-6).

Post-booster dose antibody dynamics

The administration of a booster dose significantly increased the production of NAb in participants, compared to those who did not receive a booster dose during the 12–14-month post-vaccination period. The NAb levels of participants who received the booster dose were higher with a median of 2064 ng/μl, compared to those without a booster with a median of 475 ng/μl (p < 0.001) (Figure 4a & 4b, Table 3). There was a non-significant relationship between the booster and non-boosted participants, probably due to the low sample size of the without booster sub-group (Figure 4c). However, there was a significant increase in IgG levels in all participants after the administration of the booster dose, compared to the levels seen at 8–12 months post-vaccination (that represented a drop from post-vaccine levels) (Figure 4d). Moreover, NAb levels were significantly lower in non-boosted participants until 16–18 months (Table 3). In contrast, the IgG levels did not show statistical significance in the 14–16 month and 16–18-month periods (Table 3).

Figure 4:

Comparison of Antibody Responses in Participants with and without Booster Doses. (a) Difference in Neutralizing Antibody (NAb) levels between participants with booster and without booster. (b) Increase in NAb levels in participants receiving booster doses. (c) Comparison of Immunoglobulin G (IgG) levels in participants with and without booster doses. (d) Distribution of IgG levels in participants receiving booster doses. The horizontal lines represent the median value. Statistical significance indicated by ns (P > 0.05), and **** (P ≤ 0.0001).

Table 3:

The comparison of levels of NAB and IgG in participants with booster compared to without booster dose.

Variable	N	vac_only, N = 1751	vac+booster, N = 1151	p-value2
NAb 12–14 months	144	475 (196, 938)	2,064 (1,897, 2,175)	<0.001
NAb 14–16 months	63	473 (169, 829)	1,987 (1,829, 2,098)	<0.001
NAb 16–18 months	21	583 (297, 852)	1,956 (1,667, 2,058)	0.001
IgG 12–14 months	121	166 (138, 201)	185 (164, 213)	0.060
IgG 14–16 months	51	194 (169, 204)	183 (153, 201)	0.6
IgG 16–18 months	21	169 (126, 179)	168 (136, 189)	0.8

¹Median (IQR)

²Wilcoxon rank sum test; Wilcoxon rank sum exact test

Variables associated with a decline in humoral immunity against SARS-CoV-2

The longitudinal analysis of 65 participants at four-time points (pre-vaccine, post-vaccine, pre-booster, post-booster) revealed that individuals with natural immunity and vaccination maintained higher neutralizing antibodies at 8–12 months post-vaccination compared to naïve and vaccinated (Figure 5a). However, the levels of neutralizing antibodies were not statistically different at 12–14 months post-booster. Further, participants without booster doses showed a significant reduction in antibody levels at the same time point.

Figure 5:

Analysis of variables associated with decline in antibody (a) Distribution of antibody levels at four distinct time points (n=65). (b) Odds ratio plot of decline in NAb at 8–12 months post vaccine. (c) Odds ratio plot of decline in IgG levels at 8–12 months post vaccine. (d) Ethnicity-based z-score distribution of antibody levels at different time points. (e) Ethnicity based z-score distribution of antibody levels at post booster stage. (f) Multivariate Correspondence Analysis (MCA) of categorical variables and their relationship to the decline in IgG and NAb levels at 8–12 months post-vaccination.

The results of the logistic regression analysis at the univariate level indicated that a 2-fold drop after vaccination in NAb antibodies was associated with a SARS-CoV-2 naïve group (OR: 2.17, 95% C.I: 1.06–4.41, p=0.03) and white race (OR: 2.1, 95% C.I: 1.07–4.08, p=0.02) (Supplementary Table 7). The decrease in IgG antibodies was associated with a SARS-CoV-2 naïve group (O.R: 2.52, 95% C.I: 0.96–6.59, p=0.059) and white ethnicity (O.R: 2.69, 95% C.I: 1.07–6.73, p=0.03) in univariate models (Supplementary Table 8). After adjusting for other variables at the multivariate level, a significant association between the decrease in NAb and IgG antibodies and white race was maintained (OR 2.07, 95% CI 1.04–4.11, p=0.03 and OR 2.54, 95% CI 1.004–6.46, p=0.04, respectively) (Figure 5b & 5c). Further comparison of race was assessed using z-scores, revealing higher NAb & IgG antibody levels in participants with non-white races (Figure 5d & 5e). MCA analysis revealed stable antibody response in the 8–12 months post-vaccination period was associated with non-white ethnicity, female gender, and previous infection (Figure 5e).

Discussion

The COVID-19 pandemic has caused an unprecedented global health crisis. Given the rapid administration of COVID-19 vaccination and the likely continued potential for exposure to SARS-CoV-2 for years to come, it is essential to track the durability of the immunity generated by vaccination and identify the factors that correlate with its protection. This is essential for ongoing monitoring of the response and effective management of the pandemic and is particularly important for HCWs who may have a higher risk of exposure.

The activation of the humoral and cellular components of the immune system plays a critical role in the protection against SARS-CoV-2. The production and maintenance of steady antibody levels have been linked to protection against SARS-CoV-2. The generation of IgM and IgG antibodies are among the first measurable immune responses to occur after exposure to SARS-CoV-2 and have been extensively studied. The half-life of IgG has been estimated to be approximately 21 days. However, sustained antibody titers have been observed to persist for months after the initial peak, likely due to the presence of long-lived plasma cells ⁸. In severe SARS-CoV-2 infections, a more robust IgG response has been reported compared to those with milder symptoms ²⁴. Further, higher levels of neutralizing antibodies have been reported post-infection with SARS-CoV-2 ²⁵. In our study, we found this correlation between the severity of infection and antibody response. The participants in our study with a history of more symptoms during a prior COVID-19 infection had significantly higher neutralizing antibodies. Such increased levels of antibodies have been linked to the severity of the infection ²⁶. Higher antibody levels have been reported to correlate with male sex, age, and serious complications ²⁷. In this study, we did not observe significantly elevated levels of IgG at baseline visits in participants with prior infection. This can be partially attributed to the relatively long interval between the onset of infection and the baseline visit (approximately 16 weeks) ²⁸. Interestingly, in our study, 11 NI participants showed lower levels of IgG antibodies (<10 RU/ml) and corresponding lower NAb levels. Other studies have reported a similar rate of 9.6% of individuals with no antibody response after COVID-19 ²⁷. Additionally, we detected low levels of neutralizing antibodies against SARS-CoV-2 in pre-covid and naïve samples. This could potentially be explained by partial cross-reactivity between pre-existing antibodies from evolutionarily related viruses. A recent study identified a total of 15 such SARS-CoV-2 spike protein epitopes that cross-react with antibodies generated against various acute and chronic viral antigens ²⁹. Several other studies have also reported the presence of cross-reactive antibodies in naïve or pre-covid samples ^30–33.

While several studies have assessed antibody variations in HCWs, few have investigated the association of sociodemographic factors. In a 10-month study of HCWs, IgG response was detected up to 10 months post-vaccination. The determinants of higher antibody levels were found to be female sex and prior COVID infection ³⁴. A large study of 1407 HCWs in the UK, found that natural infection and two doses of vaccination were associated with stronger protection against symptomatic infection compared to unvaccinated HCWs ³⁵. These findings highlight the importance of monitoring antibody levels and administering booster doses to ensure sustained protection against COVID-19.

Initial studies on individuals with natural infection have reported contrasting observations regarding the stability of IgG response. While stable levels have been observed at 4 months post symptom onset, a decline in antibodies has also been reported ^8,36. In our study, we observed significant differences in antibody levels at baseline, post-vaccination, and 8–12 months post-vaccination between NI and naïve participants. However, after receiving a booster dose, both the NI and vaccinated groups showed similar peak antibody levels, suggesting that both groups generated equivalent immune responses. A recent meta-analysis of 26 studies concluded that individuals with hybrid immunity (i.e., natural infection and vaccination) have higher levels of antibodies and more durable protection against COVID-19 compared to those who are naive. The analysis further suggested that the time between vaccination and booster dose can be extended in NI individuals compared to naive ones³⁷.

Multiple studies have explored the correlations between ethnicity and socio-economic factors with seroprevalence, immune response, hospitalization risk, and long-term antibody maintenance ^38–42. In our study, we applied MCA to reduce the dimensionality of clinical variables and to identify underlying trends. MCA revealed a closer association between a drop in IgG and NAb, whites, previous infection status, and females. It is known that there is a differential immune response to infection in males and females and they respond differently to COVID-19 along with race-based variations in the immune response.

In a recent study, among 13,343 participants, non-Hispanic Blacks were 2.7 times more likely to test positive for IgG antibodies compared to non-Hispanic Whites. Physician assistants and therapists were significantly more likely to have IgG antibodies compared to those in administration ³⁹. In another comprehensive cross-sectional analysis of 28,503 participants, residents of non-Hispanic Black and Hispanic neighborhoods demonstrated significantly higher odds of seropositivity compared to residents of predominantly non-Hispanic White neighborhoods ⁴². Further studies have identified higher seropositivity among black individuals and an increased risk of COVID-19 infection among hospital staff of Black Asian Minority Ethnicity (BAME) ^43,44. Other research has also shown higher antibody titers among non-white individuals compared to whites ^45,46. Also, the presence of comorbidity is linked with higher hospitalization risk in white patients compared to others ⁴⁰.

In our study, we observed elevated antibody levels in non-white participants at both post-vaccination and during the 8–12 months period. These findings are supported by previous studies, which have also shown higher persistence of antibodies in individuals of black or Asian ethnicity ⁴¹. In a different study, participants of non-white ethnicity displayed higher peak levels of antibodies, while IgG half-life was estimated to be of longer duration in white participants ⁴⁷. Additionally, non-white ethnicity along with age, hypertension, and COVID-19 symptoms were all found to be associated with higher antibody levels ⁴¹. In another study, individuals of Asian origin exhibited a higher peak of antibodies compared to those of other ethnicities, indicating a difference in antibody levels based on ethnicity. ⁴⁸.

The association between race, immune response, and disease severity highlights the importance of assessing demographic features for protection against SARS-CoV-2 ^43,49. Further, demographic-based risk assessment and monitoring of HCWs (HCWs) to enhance their protection and ensure a safe work environment is critical. Additionally, maintenance of the vaccine schedule and adherence to booster doses have been shown to maintain humoral immunity over a sustained period ⁵⁰. Furthermore, continued monitoring of community-level immune responses to COVID-19 is crucial to address potential declines in immunity among individuals and allocate proper resources for its address. The insights obtained from these monitoring efforts can inform public health strategies and intervention measures to promote immunity against COVID-19 or design strategies to counter potential vaccine escape variants of SARS-CoV-2.

Strengths and limitations

Our study has several strengths. First, it was conducted during the initial phase of the pandemic when HCWs were prioritized for COVID-19 vaccination. Second, we analyzed prospectively variables of humoral immunity, including neutralizing and IgG titers, over 18 months. Third, we performed logistic regression and multivariate regression to identify the role of multiple demographic and clinical variables in the decline of antibodies.

Several limitations to our study should be taken into consideration. Our study cohort included a higher number of females compared to males, reflecting the higher proportion of females in the healthcare workforce at our institution. Secondly, the median gap of 4 months between the natural infection of participants and their baseline visit may have impacted the antibody levels. Thirdly, due to unforeseeable constraints on the manufacturing of IgG kits during the COVID-19 Pandemic, we did not receive delivery of our kits as originally planned. As a result, we could not quantify the IgG antibodies for all patients in the study. However, we still obtained sufficient data at the critical time points necessary to draw conclusions from our study. Lastly, due to a high attrition rate of participants, we divided our time frames at >3 months post-vaccination and 8–12 months, which may not accurately reflect the immune response at more precise time points. Future studies with a larger sample size with a longer follow-up period could provide more robust insights into the dynamics of immune response to SARS-CoV-2 and the efficacy of booster doses.

Conclusion

In conclusion, our study sheds light on the dynamic nature of the humoral immune response to SARS-CoV-2 in HCWs in Augusta, GA. The results demonstrate a decline in NAb and IgG levels and highlight the importance of demographic factors such as ethnicity on waning immunity. Importantly, our findings highlight the importance of booster doses in maintaining durable immunity against COVID-19. As the global community continues to navigate the challenges posed by the pandemic and the persistence of SARS-CoV-2, understanding the dynamics of the humoral immune response is essential for devising effective strategies to manage the immune response to COVID-19.

Supplementary Material

Fig S1

Supplementary Figure 1: Distribution of NAb and IgG Antibodies in Pre-COVID Samples (a) Distribution of NAb in pre-covid samples (b) Distribution of IgG in pre covid samples. The median value is represented by a solid line.

Data availability:

The datasets used and/or analyzed in the current study are available from the corresponding author upon reasonable request.