Use of Self-Collected Dried Blood Spots and a Multiplex Microsphere Immunoassay to Measure IgG Antibody Response to COVID-19 Vaccines

ABSTRACT Serosurveys can determine the extent and spread of a pathogen in populations. However, collection of venous blood requires trained medical staff. Dried blood spots (DBS) are a suitable alternative because they can be self-collected and stored/shipped at room temperature. As COVID-19 vaccine deployment began in early 2021, we rapidly enrolled laboratory employees in a study to evaluate IgG antibody levels following vaccination. Participants received a DBS collection kit, self-collection instructions, and a brief questionnaire. Three DBS were collected by each of 168 participants pre- and/or postvaccination and tested with a multiplex microsphere immunoassay (MIA) that separately measures IgG antibodies to SARS-CoV-2 spike-S1 and nucleocapsid antigens. Most DBS (99.6%, 507/509) were suitable for testing. Participants with prior SARS-CoV-2 infection (n = 7) generated high S antibody levels after the first vaccine dose. Naïve individuals (n = 161) attained high S antibody levels after the second dose. Similar antibody levels were seen among those vaccinated with Moderna (n = 29) and Pfizer-BioNTech (n = 137). For those receiving either mRNA vaccine, local side effects were more common after the first vaccine dose, whereas systemic side effects were more common after the second dose. Individuals with the highest antibody levels in the week prior to the second vaccine dose experienced more side effects from the second dose. Our study demonstrated that combining self-collected DBS and a multiplex MIA is a convenient and effective way to assess antibody levels to vaccination and could easily be used for population serosurveys of SARS-CoV-2 or other emerging pathogens. IMPORTANCE Serosurveys are an essential tool for assessing immunity in a population (1, 2). However, common barriers to effective serosurveys, particularly during a pandemic, include high-costs, resources required to collect venous blood samples, lack of trained laboratory technicians, and time required to perform the assay. By utilizing self-collected dried blood spots (DBS) and our previously developed high-throughput microsphere immunoassay, we were able to significantly reduce many of these common challenges. Participants were asked to self-collect three DBS before and/or after they received their COVID-19 vaccines to measure antibody levels following vaccination. Participants successfully collected 507 DBS that were tested for IgG antibodies to the spike and nucleocapsid proteins of SARS-CoV-2. When used with self-collected DBS, our relatively low-cost assay significantly reduced common barriers to collecting serological data from a population and was able to effectively assess antibody response to vaccination.

with low seroprevalence to future outbreaks (1,2). Early in the SARS-CoV-2 pandemic, our laboratory developed a high-throughput microsphere immunoassay (MIA) to detect IgG antibodies to the SARS-CoV-2 nucleocapsid and spike S1 antigens in dried blood spots (DBS). In April 2020, we used this assay to test >15,000 DBS collected by trained staff at grocery stores throughout New York State (NYS) and reported an estimated cumulative incidence of SARS-CoV-2 infection of 22.7% in New York City and 14% statewide (3). This assay was then employed for additional NYS serosurveys through June 2020 (4) and a large-scale serosurvey of DBS collected from all newborns born in NYS from November 2019 through November 2021 (5).
With the introduction of COVID-19 vaccines, we realized a need for DBS collected from individuals before and after vaccination to evaluate the ability of our MIA to measure antibody levels after vaccination. The first two COVID-19 vaccines, Pfizer-BioNTech and Moderna, received U.S. Food and Drug Administration Emergency Use Authorization (EUA) in December 2020 (6,7), and a third vaccine, Janssen, was authorized in February 2021 (8). All three vaccines induce antibodies to the SARS-CoV-2 spike antigen. The Pfizer-BioNTech and Moderna vaccines are both mRNA-based vaccines while the Janssen vaccine is an adenovirus vector vaccine. In order to rapidly deploy our study without burdening health care staff, we chose to use self-collected DBS, a technique which has previously been used successfully for serosurveys (9)(10)(11). Therefore, we developed a DBS self-collection kit, instructions, and a brief questionnaire to collect demographic and COVID-19 history data. Wadsworth Center laboratory employees were recruited to selfcollect DBS before, when possible, and after COVID-19 vaccination. DBS were tested using our multiplex MIA to measure antibody levels following vaccination in our study population.

RESULTS
A total of 175 individuals were enrolled in this study. Four participants did not submit any samples, three submitted only one sample, and 168 (96%) submitted a full series of three DBS. A total of 509 DBS collected either at the participant's home or at the worksite were received from 171 individuals. Two samples were deemed unsuitable for testing because they were received more than 14 days after being collected, leaving 507 (99.6%) suitable for testing. Individuals reported having some difficulty collecting a sample for 71 (14.0%) of the 509 specimens received. Most of these issues (11.8%) were associated with lack of blood flow or fast clotting and were commonly resolved by pricking a different finger, exercising, or taking a warm shower before collecting the sample. Other reported issues included messy collection (1.2%), difficulty working lancets (,1%), and not following instructions properly (,1%). Other issues related to not following instructions included DBS cards that were not fully dry upon receipt in the laboratory (1.4%), overlapping drops of blood (1.2%), missing paperwork (,1%), and names written on cards which should have only been identifiable by barcodes (,1%). None of these issues prevented the samples from being tested.
Participant characteristics. Participant ages ranged from 20 to over 65 years old with most participants falling between 40 and 59 years of age ( Table 1). Most of the participants were female, White, and not Hispanic or Latino. Most participants lived in NYS with the majority residing in Albany County and the surrounding counties of Rensselaer, Saratoga, and Schenectady. All three FDA-authorized or approved vaccines are represented in our sample (Pfizer-BioNTech, Moderna, and Janssen). Most participants received the Pfizer-BioNTech vaccine while only three individuals received Janssen. Individuals received their first vaccine dose between December 23, 2020 and May 7, 2021 with a median date of March 4, 2021. Seven individuals in the study were infected with SARS-CoV-2 prior to beginning the study (4.1%) and two became infected during the study (1.1%). Of these nine individuals, six became infected in December 2020 while the others were infected in February and March 2021. All infected individuals received a mRNA vaccine. Eight of these individuals experienced COVID-19 symptoms and received a COVID-19 diagnosis (Table 2). One individual did not receive a COVID-19 diagnosis but had an initial nonreactive sample followed by a sample with IgG index values >1 for both N bead sets and was therefore classified as having been infected. For this study, recovered individuals are defined as those who were infected with COVID-19 prior to beginning the study. Naïve individuals are defined as individuals who were not infected prior to beginning the study, including the two individuals who became infected during the study. All three COVID-19 vaccines in this study induced spike (S) antibodies while natural infection with SARS-CoV-2 typically induced both nucleocapsid (N) and S antibodies in immunocompetent individuals. At their initial sample, recovered individuals had an IgG index value between 0.34 and 12.18 (mean index value = 3.61) for the N-SB bead set and between 1.48 and 5.84 (mean index value = 3.10) for the N-NA bead set. Prevaccine samples from naive individuals had IgG index values between 0.02 and 3.22 (mean index value = 0.19) for the N-SB bead set and between 0.04 and 0.91 (mean index value = 0.28) for the N-NA bead set. The one individual with a reactive result for N-SB (index value = 3.22) had nonreactive results for N-NA (index value = 0.79) and S (index value = 0.04) and thus was classified as naive. a Age and sex were collected at time of enrollment and are reported for all enrolled individuals (n = 175). Race, ethnicity, and county were collected with first sample submission and are reported for individuals who submitted at least one sample (n = 171). Vaccine manufacturer was reported for all individuals who submitted at least one sample post 1st vaccine dose (n = 169).
Effect of previous infection on IgG S antibody levels. All naive individuals were nonreactive for IgG S antibodies prior to receiving a vaccine dose (mean index = 0.05; range 0.01 to 0.49) while all recovered individuals were reactive (mean index = 3.40; range 1.07 to 6.29, P , 0.01) (Fig. 1). Recovered individuals showed an abrupt increase in S antibody levels after their first dose compared to naive individuals who showed a gradual increase between their first and second dose (15 to 21 days post dose 1, naive mean index = 3.90, recovered mean index = 21.50, U = 0, P = 0.0008). Naïve individuals' S antibody levels increased rapidly after a second dose of either mRNA vaccine to reach levels similar to those of the recovered individuals (naive versus recovered:36 to 42 days post dose 1, U = 68 P = 0.28; 42 to 49 days post dose 1, U = 42.5, P = 0.65). S antibody levels in participants peaked several weeks after the second dose of either mRNA vaccine before gradually declining.
Recovered individuals submitted their first DBS prior to vaccination and between 60 and 120 days post symptom onset. In order to compare antibody levels induced by mRNA vaccines to those induced by natural infection, the initial samples submitted by all recovered individuals were compared to naive individuals who submitted samples between 60 and 120 days after their second vaccine dose. The S index values of recovered individuals induced by natural infection alone were significantly lower than S index values induced by vaccination in naive individuals in the same time frame (recovered mean = 3.40, naive mean = 9.53, U = 159, P = 0.0004).
Effect of vaccine manufacturer on IgG S antibody levels. Pfizer-BioNTech vaccine doses are administered 21 days apart while Moderna doses are administered 28 days apart  . Antibody levels in both groups slowly declined after this peak.
Effect of age and sex on S antibody levels. In order to assess the effect of age and sex on S antibody response to mRNA vaccination, samples were grouped into 7-day periods based on days after the first vaccine dose. Due to the difference in timing between vaccines and the low number of individuals who received the Moderna vaccine in our study, only data from naive individuals who received the Pfizer-BioNTech vaccine were included in this analysis. No consistent trend among age groups was found across time periods; however, 15 to 21 days after the first vaccine dose, individuals 20 to 29 years old had significantly higher S antibody index values than individuals aged 60 years and Self-Collected DBS to Measure Ab to COVID-19 Vaccines Microbiology Spectrum older (U = 76, P = 0.003) (Fig. 3). Similarly, no consistent difference between males and females was found across time periods. Females had significantly higher S antibody levels 29 to 35 days after their first vaccine dose, or 8 to 14 days after their second dose, compared to males (U = 99, P = 0.03) (Fig. 4). mRNA vaccine side effects. We also analyzed the side effects reported by study participants. Local side effects include all side effects near the injection site, including injection site pain, sore arm, redness, and swelling. Systemic side effects include side effects that impacted the whole body or areas of the body not near the injection site. Of all individuals receiving either mRNA vaccine, local side effects were experienced more frequently after the first dose (54%) than after the second (40%, P = 0.02) and systemic side effects were experienced more frequently after the second dose (69%) than after the first (31%, P , 0.0001) ( Table 3). For dose 1, significantly more local side effects were reported than systemic side effects (P , 0.0001) while for dose 2, significantly more systemic side effects were reported than local side effects (P , 0.0001). Fatigue, muscle or joint pain, and headache were the most commonly reported side effects to either dose.
Although differences in local and systemic side effects were not significant between the naive and recovered groups, recovered individuals reported systemic side effects from dose 1 at a higher frequency than naive individuals and at a similar frequency to what both groups reported for dose 2 (Fig. 5A). There was no significant difference in side effects reported from either dose between naive individuals receiving the Moderna vaccine and naive individuals receiving Pfizer-BioNTech, although individuals receiving the Moderna vaccine reported slightly higher systemic side effects to both doses than those receiving Pfizer-BioNTech (Fig. 5B). Due to the low number of recovered individuals and individuals who received the Moderna vaccine in our study, the influence of age and sex on vaccine side effects were only assessed for naive individuals who received the Pfizer-BioNTech vaccine. Slight differences in local side effect reporting were observed between males and females (dose 1 P , 0.05, dose 2 P , 0.01) (Fig. 5C). Females reported slightly more systemic side effects than males for both doses, although this difference is not significant. A significant trend was found across age groups for local side effects from dose 1 (P , 0.01) where the 20-to 29-year-old and 30-to 39-year-old groups tended to report more side effects than other age groups (Fig. 5D). A similar trend was found for systemic effects experienced after dose 2 (P , 0.01) where individuals 30 to 39 years old reported side effects at the highest frequency followed by individuals 40 to 49 years old and 20 to 29 years old. Finally, we wanted to assess if having a high level of S antibodies after the first vaccine dose was associated with a change in side effects experienced from the second vaccine dose. To do this, we utilized data from all naive individuals who received the Pfizer-BioNTech vaccine and submitted a sample within 1 week prior to receiving their second dose (n = 80). We then used Mann-Whitney U Tests to compare the mean MFI index values between individuals who reported side effects and those who did not. Individuals who reported experiencing any side effects had a higher mean S index than individuals not reporting side effects (4.32 versus 2.62, respectively, U = 351, P = 0.006). Individuals who reported experiencing muscle or joint pain, fever, nausea or vomiting, and lymph node pain or swelling had significantly higher mean S index values than individuals who did not report those side effects (Table 4).

DISCUSSION
As COVID-19 vaccines became authorized in the United States, establishing a mechanism to capture blood samples before and after vaccination for serological assessment became an urgent need for our public health laboratory. Our ability to initiate a study within the relatively brief window between vaccine authorization and distribution was facilitated by recruiting laboratory staff to self-collect DBS specimens before and after receiving vaccine doses. We enrolled 175 individuals who submitted 509 DBS, 507 (99.6%) of which were suitable for testing. While 14% of individuals reported issues with DBS collection, none of these issues prevented the DBS from being tested. This success supports the potential for self-collected DBS to be used as an effective serosurvey strategy to assess response to vaccination or general antibody prevalence. However, our study participants were primarily laboratory personnel and included many individuals who are familiar with DBS. Therefore, this high rate of success in DBS collection may not occur in the general population. Similar studies using at-home self-collected DBS reported lower return rates than this study (10,(14)(15)(16) but the proportion of samples that were unsuitable for testing upon arriving at the lab were comparable (14)(15)(16). It is of note that while our study demonstrates an effective use of self-collected DBS as a sample type, rather than more invasive and labor-intensive plasma and serum samples, the test still requires trained laboratory staff, specialized equipment, and time to perform. However, the test used in this study can be performed in less than a day and is designed to accommodate high-throughput testing, resulting in a short turnaround time in the lab. Additionally, not including labor and equipment costs, this assay has a relatively low reagent cost of approximately $1.00 per sample (4,5).
Using DBS, we saw that vaccinated individuals' spike IgG antibody levels increased until several weeks after their last dose before beginning to slowly decline, matching results from other studies that used serum/plasma (17)(18)(19). Vaccinated individuals' spike IgG also reached higher levels than were seen in those who were infected but not yet vaccinated (18,(20)(21)(22)(23)(24)(25)(26). Also, like previous studies, we found that the seven recovered individuals in our study showed an abrupt increase in antibodies after the first dose of vaccine while naive individuals showed a more gradual increase in antibody levels (18, 21-23, 25-29). However, due to the small number of recovered individuals in our study, FIG 4 Spike index values in 7-day groups post first vaccination for naive individuals who received the Pfizer-BioNTech vaccine grouped by sex (n = 133). *, P , 0.05.

Self-Collected DBS to Measure Ab to COVID-19 Vaccines
Microbiology Spectrum we cannot generalize our results regarding the difference in final antibody levels or rate of antibody waning between naive and previously infected individuals. Naïve individuals who received the Moderna or Pfizer-BioNTech vaccine reached their peak antibody levels at different times which may be partially associated with the difference in dosage timing. Our study found that both groups receiving an mRNA vaccine reached similar peak antibody levels following their second dose. Data from some studies support this finding (20,30); however, other studies have found that individuals who received the Moderna vaccine reached higher antibody levels than those who received the Pfizer-BioNTech vaccine (19,31,32). These different findings may be due to sample size effects and differences between sampled populations.
Our study did not find any consistent trend of S antibody levels induced by the Pfizer-BioNTech vaccine by age group. However, our study and several others found that younger age groups reached higher antibody levels between the first and second doses (28,30,31,33,34). Other studies have also found that younger individuals reach higher antibody levels than older individuals following their second dose (18,24,32,35). Although Amodio et al. found that females reached higher IgG concentrations than males (24), in agreement with other work, we did not find a consistent correlation between sex and S antibody levels (30,(33)(34)(35). Due to a relatively small male sample size in our study, it is difficult to draw conclusions regarding sex differences in S IgG concentrations.
Our investigation of the number of side effects reported by individuals who received either mRNA vaccine did not reveal a significant difference between individuals who received the Moderna or Pfizer-BioNTech vaccine. However, other studies have shown that individuals who received the Moderna vaccine reported more side effects than those who received the Pfizer-BioNTech vaccine, suggesting that our data may have approached statistical significance if our study population included more individuals who received the Moderna vaccine (19,36,37). We found that recovered individuals reported more systemic side effects after receiving their first dose of the vaccine than naive individuals did and although these results did not reach statistical significance, several other studies have found a similar pattern (23,25,26,38,39).
Our analysis of naive individuals who received either mRNA vaccine showed that the most commonly reported side effects in our study, injection site pain, fatigue, muscle or joint pain, and headache, are similar to those reported in other studies and clinical trial data (12,13,18,20,23,37,40). Our study found similar rates of both local and systemic side effects to several studies (23,26,38); however, clinical trials and other studies have reported higher frequencies of both systemic and local side effects (12,13,36,37,39). Similar to some studies, we found that the occurrence of local side effects decreased from the first to the second dose while the occurrence of systemic  Self-Collected DBS to Measure Ab to COVID-19 Vaccines Microbiology Spectrum side effects increased (13,39). However, other studies have found that the rate of local side effects increases from dose 1 to dose 2 (12,18,25). Our questionnaire used a free response write-in section for participants to describe their side effects. This free response strategy may have resulted in individuals forgetting to describe certain symptoms because they were not prompted or did not deem them significant enough to list resulting in skewed data. Within the Pfizer vaccinated group, we saw significant differences in side effects experienced by age group. Younger age groups were more likely to report local side effects after dose 1 than older groups. After dose 2, individuals 30 to 39 years old reported systemic side effects at the highest rates, followed by individuals 40 to 49 years old and 20 to 29 years old. Clinical trials and similar studies have found that younger individuals tend to experience more side effects from COVID-19 vaccination than older individuals (13,(39)(40)(41). We also found that females were more likely to experience side effects than males where the difference in local side effects was statistically significant and the difference in systemic side effects was not. Other authors have also found that females are more likely to experience side effects than males (39,40); although, Klugar et al. reported that both sexes experienced similar levels of side effects (41). We found that individuals with higher antibody levels in the week prior to dose 2 reported more side effects after dose 2 than those with lower antibody titers. Michos et al. found a similar result when comparing antibody levels after the second dose while Röltgen et al. did not find a correlation between antibody levels and side effects experienced (18,33).
The major limitations of our study are related to the use of a convenience sample which does not reflect the general population. Study participants were highly motivated due to their involvement in public health work, highly educated, actively employed, presumed healthy, and lacked diversity. We also do not have data from recruited individuals who decided not to participate. The study population was mostly female, between the ages of 40 and 59, White, not Hispanic or Latino, and had not been previously infected with SARS-CoV-2. Additionally, the majority of participants received the Pfizer-BioNTech vaccine and data on those who received the Moderna or Janssen vaccine is limited. We also relied on self-reports and a nonstandard, write-in method to collect information on side effects from the vaccine which required an individual to manually categorize side effects and leaves room for misinterpretation of responses. Individuals were enrolled in the study at various times before and after vaccination, so this study does not have a standard sample follow-up period. Finally, this study did not include measurements of neutralizing antibodies or other indicators of humoral immune response. While antibody levels have been correlated with neutralizing antibody levels (33,42,43), antibody levels alone should not be used to make predictions about the ability of different vaccines to protect against SARS-CoV-2 infection or disease. Conclusion. We demonstrate that self-collected DBS and a multiplex MIA are an effective way to measure antibody responses to COVID-19 vaccines. Our results using DBS were similar to previous vaccine response studies that used serum/plasma samples. Self-collected DBS are a convenient option for obtaining blood specimens to monitor antibody response to vaccinations and could also be used to perform large-scale population serosurveys for SARS-CoV-2 or other emerging infections.

MATERIALS AND METHODS
Recruitment and sample collection. Starting in February 2021, we enrolled a convenience sample of current employees, students, or fellows at the Wadsworth Center, NYS Department of Health (DOH) who were at least 18 years old and planned on receiving, or had recently received, a COVID-19 vaccine. Participants were recruited using fliers emailed to their DOH email address. Approval for human subject research was obtained from the New York State Department of Health Institutional Review Board (IRB) .
Eligible individuals who signed and returned an informed consent form received a DBS collection kit which contained instructions for collecting DBS, three preaddressed envelopes, three postage stamps, six safety lancets, six gauze pads, six alcohol wipes, and six bandages. The kits also contained three questionnaires for collecting demographic information, COVID-19 infection history, previous COVID-19 antibody test history, vaccination details and side effects, and three 903 five-spot blood collection cards (Eastern Business Forums, Maudlin, SC) each labeled with a barcoded and deidentified participant ID. Participants completed and returned a questionnaire with each DBS card. The instructions included photographs and text describing precollection set-up, collection, and postcollection procedures (Fig. S1).
Participants collected a total of three DBS and followed one of two collection schedules based on their vaccination status at enrollment. Individuals who were not vaccinated prior to enrollment collected samples 1 to 7 days before the first dose of vaccine, 1 to 7 days before the second dose of vaccine, or 3 to 4 weeks after the first DBS collection, and 3 to 4 weeks after the second DBS collection. Individuals who were vaccinated prior to enrollment collected a sample upon receiving the kit, 3 to 4 weeks after the first DBS collection, and 3 to 4 weeks after the second DBS collection. Participants were sent emails 2 to 3 weeks after the first and second DBS were received to remind them to collect their next DBS. If follow-up samples were not submitted, participants received one additional reminder email. Participants placed the DBS card and the accompanying questionnaire in the preaddressed envelopes and either mailed them to the laboratory or dropped them off in assigned locations. Upon receipt, laboratory staff labeled each DBS card with a unique accessioning barcode and checked the questionnaires and cards for any issues. Samples received more than 14 days after collection were marked as unsuitable for testing (n = 2). Participants were notified if their sample was unsuitable and allowed to collect and submit another sample.
Following completion or withdrawal from the study, participants were sent a report of their results which included the individual's nucleocapsid and spike index values, qualitative result, and interpretation for each DBS collected. The report also included the criteria used to determine qualitative results, information to help interpret the data, and a graph showing the mean spike median fluorescent intensity (MFI) index values by days after first dose of vaccine for all study samples submitted up to that point.
Assay procedure. Magplex-C microspheres were coupled as previously described (4). Three bead sets were coupled: a spike S1 subunit antigen from SinoBiological (S) (Beijing, China), and two nucleocapsid antigens, one from SinoBiological (N-SB) and one from Native Antigen (N-NA) (Oxford, UK). DBS cards were punched using Panthera or Wallac punchers (PerkinElmer, Waltham, MA). One 3.2-mm punch was added to each well of a round-bottom, nontreated, polystyrene 96-well plate (Corning, Corning, NY). Punches were then eluted in 250 mL elution buffer (Tris-buffered saline, 1% casein blocker) (Bio-Rad Laboratories, Hercules, CA) for 1 h at room temperature (19°C to 22°C). Beads and eluate (25 mL of each, 1,250 beads/bead set/well) were then transferred to nonbinding 384-well plates (Greiner Bio-One, Monroe, NC) and incubated for 30 min at 37°C and shaken at 300 RPM in the dark. Samples were washed using wash buffer (PBS, 2% BSA, 0.02% Tween, 0.05% azide, pH 7.5) on a BioTek 405 TSUS magnetic microplate washer. After washing, samples were incubated with 50 mL phycoerythrin-tagged goat-anti human IgG at a concentration of 0.4 mg/mL (Thermo Fisher Scientific -Invitrogen-eBiosciences, Waltham, MA or SouthernBiotech, Birmingham, AL). Plates were then incubated for 30 min and washed as previously described. Ninety mL of xMap sheath fluid (Luminex Corp., Austin, TX) were added to each well and plates were shaken at room temperature in the dark at 300 RPM for 1 min to resuspend beads. Assays were performed manually or by an automated Microlab Star liquid handling system (Hamilton Company, Reno, NV) to perform DBS elution, microsphere addition, sample addition, and secondary antibody addition. Plates run using the liquid handlers were incubated on Hamilton Heater Shakers. Samples were analyzed using a FlexMap 3D instrument (Luminex Corp., Austin, TX) and reported as MFI for each bead set.
Cutoff values were established for each bead set based on analysis of 789 known negative newborn and adult DBS samples. The mean MFI 16 standard deviations is considered reactive. Index values were calculated using the sample MFI divided by the reactive cutoff value for each bead set; values of >1.0 indicate a reactive result for that bead set. This assay was validated and approved for clinical testing by New York State Department of Health Clinical Laboratory Evaluation Program.
Self-Collected DBS to Measure Ab to COVID-19 Vaccines Microbiology Spectrum Data analysis. Data were cleaned using Microsoft Excel. Participant's infection status was determined by self-reported positive PCR test, COVID-19 diagnosis, or a nucleocapsid index value >1. Individuals infected prior to starting the study were assigned to the recovered group and individuals not infected at the start of the study were assigned to the naive group. Self-reported DBS collection issues and side effects were categorized using key words and were manually reviewed by a single analyst.
Comparisons of continuous variables were made using Mann-Whitney U tests for comparing two groups and Kruskal-Wallis tests for multiple comparisons. Comparisons of categorical variables were made using Fishers Exact test for comparing two groups or a chi-squared test for trend for comparing multiple groups. Two tailed P-values are reported and P-values of ,0.05 are considered significant. Figures were created using R (v4.0.5) or GraphPad Prism (v9.2.0).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.5 MB.

ACKNOWLEDGMENTS
This work was supported by a Serological Sciences Network grant from the National Cancer Institute (U01CA260508). K. Nemeth and E. Yauney were supported by a cooperative agreement from the U.S. Centers for Disease Control and Prevention (NU50CK000516) and a Serological Sciences Network grant from the National Cancer Institute (U01CA260508). We thank Norma Tavakoli for assisting with the study facilitation and Bryanna Freitas for helping to make DBS collection kits.