[Skip to Navigation]
Sign In
Medical News & Perspectives
October 21, 2021

The Flawed Science of Antibody Testing for SARS-CoV-2 Immunity

JAMA. 2021;326(18):1781-1782. doi:10.1001/jama.2021.18919
JAMA Medical News Audio (17:18)
0:00 / 0:00

Early in the COVID-19 pandemic, developers designed SARS-CoV-2 antibody tests to detect whether people had been infected. Some experts thought the blood tests eventually would help to ease lockdowns. One idea was that those with antibodies likely would be immune to reinfection at least temporarily, allowing them to reenter society without putting themselves or others at risk.

iStock.com/Ilze Kalve

But as the pandemic unfolded, the concept of an immunity passport based on having antibodies didn’t pan out. The early consumer tests’ accuracy was unproven, making the results somewhat dubious. More fundamentally, the so-called correlates of protection were unknown. Which specific antibodies guarded against SARS-CoV-2 reinfection? How high did their levels need to be? And how long would they provide a reliable defense?

As the assays’ usefulness for individual patients became less clear and testing for active infections expanded, the public’s clamor for antibody testing waned. But for some, the arrival of COVID-19 vaccines revived their interest in serology. Could a simple blood test reveal whether the vaccine was working or, later, if it was time for a booster shot?

No, says the US Food and Drug Administration (FDA), which discouraged antibody testing as a do-it-yourself immunity check in a communication to the public and clinicians this past spring. A spokesperson for the agency reiterated the position in an email to JAMA in September.

Yet many aren’t heeding the advice. A recent New York Times article described concierge clinicians who regularly test their clients for SARS-CoV-2 antibodies. And Florida-based Epitome Risk Solutions promises that the direct-to-consumer SARS-CoV-2 serology assay it sells for $170 “correctly identifies the number of neutralizing antibodies you have with 100% specificity” so that in “24-48 hours you’ll know if your immune system is still protecting you from COVID” after vaccination or infection.

Elitza Theel, PhD, director of the Mayo Clinic’s Infectious Diseases Serology Laboratory, confirmed the worrisome trend. “I am aware of individuals getting tested just to see if they’ve ‘reacted’ to the vaccine or if they have ‘immunity,’ despite our best efforts to educate,” she wrote in an email.

Serology Oversimplified

The SARS-CoV-2 serology tests that eventually received FDA Emergency Use Authorization (EUA) have demonstrated high sensitivity and specificity, but that accuracy is for detecting antibodies. Their ability to predict protection against the virus based on those antibodies hasn’t been proven. Plus, the FDA cautioned that some tests detect antibodies the immune system likely produces only after natural infection with the virus. Depending on the assay, people who weren’t previously infected could test negative for antibodies despite having vaccine-induced immunity.

Therefore, the agency in its May 19 communication stated that “results from currently authorized SARS-CoV-2 antibody tests should not be used to evaluate a person’s level of immunity or protection from COVID-19 at any time, and especially after the person received a COVID-19 vaccination.”

The problem isn’t simply that the tests weren’t designed to assess immunity, experts told JAMA. It’s also that the protective antibodies and their thresholds still haven’t been fully worked out.

In an email, Nicole Doria-Rose, PhD, a staff scientist and chief of the Humoral Immunology Core at the National Institutes of Health’s Vaccine Research Center, noted that antibodies that bind to the SARS-CoV-2 spike protein—particularly neutralizing antibodies—“do correlate with protection.” For example, higher titers were associated with increased protection in 2 phase 3 studies: one not yet peer-reviewed of the mRNA-1273 (Moderna) vaccine and another of the ChAdOx1 nCoV-19 (Oxford/AstraZeneca) vaccine. Still, she wrote, “it’s not a simple relationship—there is no clear titer at which you can say whether a particular person is protected.” What’s more, different neutralizing antibody thresholds, or even different immune responses altogether, may correlate with protection against asymptomatic, symptomatic, or severe disease.

Theel pointed to a recent study of breakthrough infections among 1497 fully vaccinated health care workers in Israel. The 39 workers who became infected after receiving their second dose of the BNT162b2 (Pfizer-BioNTech) vaccine had lower neutralizing antibody levels than their uninfected colleagues. Although antibody levels were associated with protection, the researchers were unable to pinpoint a threshold. None of the cases were severe, but antibodies obviously weren’t foolproof against the virus.

“Individuals can have neutralizing antibodies and still get infected,” Theel said in an interview. “We know that higher antibody levels, specifically higher neutralizing antibody levels, are better. But we don’t know how high is high enough.”

Meanwhile, the laboratory tests haven’t been standardized. “That’s a problem when people say, ‘Okay, I want to go see if I should get a booster or not,’” Theel said. Some SARS-CoV-2 serology assays simply give a positive or negative result, without antibody values. Those that are quantitative use varying methods, detect different antibody classes, and report values using different units of measurement.

According to Theel, if and when correlates and thresholds of protection are determined, the tests will need to be standardized and calibrated, as has been done with antibody tests for other vaccine-preventable diseases, including tetanus, diphtheria, and measles. So far, only one commercially available SARS-CoV-2 antibody test, from Ortho-Clinical Diagnostics, has been calibrated to the World Health Organization’s reference standard, she said.

What’s more, all antibodies bind but only some neutralize, and almost none of the authorized clinical tests distinguish between them. Although some studies have shown a correlation between levels of binding and neutralizing antibodies, they’re still an imperfect match.

Doria-Rose noted that measuring neutralizing antibodies requires a complex test that is run in only a few laboratories and hasn’t been scaled up for diagnostic use. The test sold through Epitome Risk Solutions’ subsidiary, FourthWall Testing, appeared to be an at-home version of the FDA-authorized GenScript cPass SARS-CoV-2 Neutralization Antibody Detection Kit, a surrogate assay that indirectly measures neutralizing antibodies. According to an FDA spokesperson, the cPass test does not have authorization for sample collection at home. And despite Epitome’s marketing hype, GenScript states that customers “should not interpret the results of this test as an indication or degree of immunity or protection from reinfection.”

Managing Expectations

For Paul Offit, MD, director of the Vaccine Education Center at Children’s Hospital of Philadelphia, there’s another strong argument against using blood tests to assess whether individual patients are protected: circulating antibodies don’t give a complete picture of SARS-CoV-2 immunity.

“The immunological component that is associated with protection against severe disease is immunological memory B cells,” Offit said in an interview. “These cells aren't making antibodies, but they remember that they have seen this SARS-CoV-2 spike protein before.”

Theel noted that circulating antibodies against the virus peak 2 or 3 months after natural infection or vaccination and then begin to decrease. But the immune system’s ability to mount a defense lasts longer. “Just because we’re seeing a decrease in antibody levels doesn’t necessarily mean that your immunity is gone,” she said. “We know from a lot of studies now that your memory T cells and memory B cells persist for at least 6 to 8 months and continue to evolve and mature, and none of that information is relayed by an antibody test.”

Offit explained that reinfection with the virus activates memory B cells to differentiate into antibody-secreting cells. Because this process can take 3 to 5 days, it doesn’t stop SARS-CoV-2 infections from occurring, but it does help tamp down severe COVID-19. And that’s a success story, in his view: “The goal of the vaccine is to protect you against serious illness.”

So far, it has. Although data suggest that mRNA vaccine–induced protection against infection and symptomatic illness is decreasing in the US, the shots continue to prevent serious disease, hospitalization, and death.

The data seem to contradict a widely reported rise in hospitalizations among vaccinated individuals in Israel. But according to Offit, Israel’s COVID-19 hospitalization trends can’t be applied directly to the US. “They do have a different definition to some extent for what they consider severe,” he said. “It’s less rigid than ours.”

Offit served on the FDA vaccine advisory committee that recommended against a booster dose for most adults who received the BNT162b2 vaccine. For now, a booster is only authorized for people with the greatest risk of severe illness: adults aged 65 years or older and adults who have certain underlying medical conditions, live or work in high-risk settings, or live in long-term care facilities. (People who are moderately to severely immunocompromised and received the BNT162b2 or mRNA-1273 vaccine should get an additional dose, according to the CDC.)

Offit said that as antibodies wane, susceptibility to asymptomatic or milder infections increases. In a recent US study, the BNT162b2 vaccine was 93% effective against hospitalizations with the Delta (B.1.617.2) variant after 6 months, but its effectiveness against Delta variant infections overall fell from 93% in the first month after full vaccination to 53% after 4 months. The authors attributed the reduction mainly to waning immunity rather than the variant’s escape from vaccine protection.

Offit acknowledged that any infection can be disruptive but pointed out the impracticality of using periodic boosters to keep neutralizing antibodies high. He said that’s akin to “rolling the stone up the hill only to have it come back again.”

And like Theel, he knows that some infections will still break through. “You can have a robust or high level of virus neutralizing antibodies in your circulation,” he said. “That’s not going to prevent the virus from attaching to your nose and beginning to reproduce itself. That’s an asymptomatic infection.”

Managing expectations for COVID-19 vaccines may be in order. “I think there was maybe an initial misunderstanding amongst the public that this would just halt all infections, and that’s just not the case,” Theel said. Ultimately, her advice for avoiding a SARS-CoV-2 infection in the first place comes down to a now-familiar combination of measures: “getting vaccinated, frequently washing your hands, wearing a mask, and avoiding high congregate in-door settings, particularly in areas that have high case rates.”

Back to top
Article Information

Conflict of Interest Disclosures: Dr Theel reported serving as an advisory board member or consultant for Roche Diagnostics, Euroimmun US, Serimmune Inc, and Ortho-Clinical Diagnostics. Drs Doria-Rose and Offit reported no disclosures.

Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    9 Comments for this article
    Thomas Pliura, M.D., J.D. | Private Practice (self-employed)

    This article quotes Elitza Theel, PhD, director of the Mayo Clinic’s Infectious Diseases Serology Laboratory, who expresses her opinion that handwashing, among other public health measures, decreases one's chances of becoming infected with Covid-19.

    I am all in favor of people washing their hands but to my knowledge there has not been a published study to suggest that washing one's hands frequently prevents transmission of Covid.

    Thomas Pliura, M.D., J.D.

    Antibody Tests and Booster Vaccine Doses
    Tien-Lan Chang, M.D. | Private Pediatrician over 65 years of age
    This article points to the limitation of the antibody test as an indication of immunity. It depends on how and why people do the test. The academics used the test and learned that immunity wanes over time, and it wanes more in older individuals than younger population. This resulted in the FDA recommendation for booster doses for people older than 65 and people who are immunocompromised. CDC broadened this recommendation for all healthcare workers because of the greater exposure risk. For someone like me who was vaccinated 10 months ago and felt doubt about the need for the booster, a negative antibody test result has convinced me to get the booster.
    Efficacy Antibody Serology and Immunocompetence
    David Loiterman, MD, FACS | Past President and Emeritus Board Chair CMS
    Thank you for publication of this timely piece.

    What is lacking at this point in the pandemic trajectory is meaningful analysis of the incidence of asymptomatic seropositivity.

    The technology exists. It is not clear why the will to obtain this information is not stronger.
    More standardized triage and ventilation management protocols would go a long way in reducing current morbidity and mortality.
    Antibody Testing in COVID-19
    Subhradip Karmakar, PhD | All India Institite of Medical Sciences, New Delhi, India
    Excellent piece by Abbasi. Antibody serology is indeed confusing and sometime could be erroneous. There is a possibility that different populations located in different countries and geographical locations might have variant specific antibody repertoire and titres that further makes its hard to compare to know if there is adequate protection. Also, different ethnic populations may have variable immune response. Last but not least, T cell response is a very strong determinant of immune response against SARS-CoV2 virus. T cells could be more resistant than antibodies to threats posed by emerging variants. T cell responses target potentially hundreds of different viral epitopes (1). Someone with a low antibody titres but with an adequate number of T cells or memory T cells will be better off avoiding severe COVID disease as compared to someone with just antibodies alone. So making a case of immune protection based on just circulating antibodies is not only misleading but also flawed as the author rightly mentioned.


    1. Tarke, A. et al. Comprehensive analysis of T cell immunodominance and immunoprevalence of SARS-CoV-2 epitopes in COVID-19 cases.  Cell Rep Med. 2021 Feb 16;2(2):100204. doi: 10.1016/j.xcrm.2021.100204.Epub 2021 Jan 26.

    Antibody Testing
    Shalom Reich, MBA | None
    While getting antibody tests may say nothing about immunity status, it is the closest thing available to the public. If someone will approve a neutralizing antibody detection test for COVID, I would be happy to use that instead of an antibody test. In the interim, the antibody test is the closest proxy for such a test.

    Saying one shouldn't take an antibody test is reminiscent of the instruction NOT to mask at the beginning of the pandemic.
    Humoral Antibody Levels as a Correlate of Protection Against SARS-CoV-2
    Stephen Strum, MD, FACP | Private Practice of Hematology, Oncology & Internal Medicine
    I read with interest the article entitled "The Flawed Science of Antibody Testing for SARS-CoV-2 immunity by Abbasi. In the in-depth reading I have done as an MD, and also as an elder at age 79, as well as having a compromised immune system due to light chain amyloidosis (AL), I find this article, itself, to be flawed. What has typified this pandemic can be characterized by a few quotes:

    ▶︎ "What we have learned about what man learns from history is that man learns nothing from history." —Churchill
    ▶︎ "Common sense is not so
    ▶︎ "If you put two doctors in a burning room, they won't agree whether to stay or to leave."

    There have been close to a hundred medical articles, many peer-reviewed, on the association of humoral antibody levels (HALs) with both natural infection and with vaccination against SARS-CoV-2, as well as articles on the effect of vaccination in both SARS-CoV-2 naive patients and those with past COVID-19 infection.

    Yes, there are many lab tests out there, and some are only qualitative and others are semi-quantitative. Furthermore, some are tests to assess IgG Ab (antibody) responses involving the S1 protein at the receptor-binding domain (RBD) and others assess nucleocapsid (NC) Ab responses. In many publications, the RBD-pan-Ig-quant ( Roche Elecsys Anti-SARS-CoV-2 S) has been used and has been found to (a) correlate with titers relating to natural infection and to vaccination (b) wane over time with major drops at ≥ 4 months post- full vaccination, and (c) correlate with virus-neutralizing antibodies (nAbs). Two of the national labs in the USA (LabCorp and Quest) both use the same Roche test for a semi-quantitative assessment of RBD IgG against SARS-CoV-2.

    ▶︎ LabCorp test code 164090: SARS-CoV-2 Semi-Quantitative Total Antibody, Spike using Roche Elecsys.
    ▶︎ Quest Test Code 39820 SARS-CoV-2 Total Antibody, spike, semi-quantitative using Roche Elecsys.

    What has been missing in almost all publications is the simple inquiry of Ab level vs. breakthrough infection. This is near-impossible to find in any study. Perhaps the only study that relates this information is in a preprint case-cohort study based on RCT data by Gilbert et al. that reports that Day 57 Spike IgG, RBD IgG, cID50, and cID80 neutralization levels were each inversely correlated with risk of COVID-19: hazard ratios 0.66 (95% CI 0.50, 0.88; p=0.005); 0.57 (0.40, 0.82; p=0.002); 0.42 (0.27, 0.65; p<0.001); 0.35 (0.20, 0.61; p<0.001) per 10-fold increase in marker level, respectively, multiplicity adjusted P-values 0.003-0.010 (1). The authors concluded: "Binding and neutralizing antibodies correlated with COVID-19 risk and vaccine efficacy and likely have utility in predicting mRNA-1273 vaccine efficacy against COVID-19."

    The Abassi commentary focuses on how little physicians can agree on anything. Because MDs and other scientists cannot agree on anything early on, or because we don't seem to learn from the experience of others ("not invented here" syndrome), or that we fail to use common sense, is not a reason not to delve deeper into an issue and conclusively come to a consensus. This is a pandemic of not only the un-vaccinated, but also the functionally unvaccinated. But underlying all of this, it is a pandemic of misinformation and miscommunication-- and that does not have to be so.


    1. Gilbert, P.B.; Montefiori, D.C.; McDermott, A.; et al. Immune Correlates Analysis of the mRNA-1273 COVID-19 Vaccine Efficacy Trial. medRxiv 2021, doi:10.1101/2021.08.09.21261290

    Booster vaccination policies in general not dependent on antibody testing
    Jesse Baker, None | Private individual
    I agree that citizens shouldn’t waste time or money on antibody tests in efforts to determine personal Covid risk absent a doctor’s specific advice. But the question of national booster policy ought to be kept separate from that of correlates of protection as these are unrelated except insofar as antibody surveillance testing at the population level might be used by public health officials to assess need for a booster. The Israeli Ministry of Health’s booster program, far from looking at antibodies, was instead motivated by increasing incidence of confirmed Covid and incidence of severe disease as defined by pulse oximetry reading <94% or respirations >30/min. While American population characteristics and severe case ascertainment indeed differ from those abroad, my understanding is that our FDA considered data from Israel and the UK when deciding on boosters this September (1).

    Issues swirling around Covid in the USA have sadly been contested from the outset, with more politics entering discussions here than anywhere else in the world. Two FDA employees who have since resigned, Philip Krause and Marion Gruber, took a parting shot at the Biden Administration on their way out the door by opposing the universal eight-month booster dose the president had announced—I assume without consulting them—in August, stating that it lacked sufficient supporting vaccine efficacy data, that it would compete with the imperatives of equitable global distribution and of reaching the unvaccinated in the US, and that it might undermine public confidence in the vaccines. Although a Lancet opinion article in which they appear as coauthors mentions neutralizing antibody titers in passing, the argument that article makes against routine boosting in the USA relies mostly on the comparative efficacy findings it charts (2).


    1. Bar-On, Goldberg, et. al., “Protection of BNT162b2 Vaccine Booster…,” NEJM, Sept. 15, 2021. doi:10.1056/NEJMoa2114255

    2. Krause, Fleming, et. al., “Considerations in boosting…,” Lancet, Sept. 13, 2021. doi:10.1016/S0140-6736(21)02046-8
    A Voice of Reason
    Lawrence Devoe, M.D. | Professor Emeritus, Department of OBGYN, Medical College of Georgia at Augusta University
    Thank you Ms. Abbasi for this timely summary of the confusing and often misguided public discussion of evaluating individual or group immune status following SARS-CoV2 vaccination. Any medical student who managed to stay awake during his or her basic immunology lectures would understand the difference between antibody types and their transient titers and the population of memory T and B cells that provide a durable defense network against viral infection.

    As a medical journal editor I am "rewarded" with 30 to 40 COVID-related emails every single day, many of them from similar sources. What concerns
    me about this vast and growing sea of information is that a number of the studies brought to my attention are either unpublished or have not undergone peer review. The latter process is essential to maintain the integrity and presumable reliability of such articles. On the other hand, I have also read hastily reviewed and published manuscripts to enhance journals' impact factors that have been later disproven by more careful and robust investigations.

    Many so-called experts in this field have added to the ongoing confusion rather than clarified it by making unfounded predictions of disease course or the ability of the available vaccines to eventually put COVID-19 in the background if not to make it completely go away. What all of us in the medical profession need to accept is that COVID-19 is a new disease, there is much still to be learned over time and that "haste makes waste" while "patience is a virtue" when it comes to addressing this global pandemic.

    Antibody Testing to Assess SARS-CoV-2 Immunity in Immunocompromised People
    Douglas Duncan, MD | Ferrell-Duncan Clinic Inc., Springfield, Missouri
    There are some things I believe all are likely to agree on:

    1. The higher circulating or neutralizing antibody titers are the lower rates of severe COVID-19 disease.
    2. Having antibodies to COVID-19 after vaccination in persons who have not had an infection likely indicates a response to vaccination.
    3. Higher antibody levels following a booster indicates a response to the booster.
    4. It is good to know if immunocompromised patients respond to a vaccination or booster; a marked increase in antibody levels following a booster likely indicates a good response to a booster.

    If there is agreement
    on these 4 principles, antibody testing patients who are immunocompromised may be indicated to see if their immune system was capable of mounting an immune response as a result of a vaccination or booster.

    If there is not agreement on these principles then it is less clear we should work to assess the ability of an immunocompromised person to mount an immune response to vaccination, and we are left with guessing how to manage these patients.

    A question: If an immunocompromised person has the same level of antibodies as a non-immunocompromised person, is that good? If they don’t is that bad? Granted, an immunocompromised person should take extra precautions no matter what their antibody levels are, but there is peace of mind that following the mitigating principles will help protect them and they don’t have to become hermits.