Our citizen science journalism project with Covid antibody tests is over, and the results are conclusively inconclusive.
Out of 32 members of the group (including me) who used our in-home antibody tests from Wuhan, China, 6 were positive, 17 were negative, 4 were invalid and 5 were not returned.
Among the people who tested negative were a significant number of people who had previously tested positive for active infection and for antibodies. Among those who were positive were people who were pretty sure they had it, and were months out from an active infection — including me. I was positive in December, 9 months from my original infection, which was very mild, almost unnoticeable.
Background: We launched this project in partnership with the Covid-19 Long-Haulers Discussion Group on Facebook and admin Amanda Finley in December. We had several goals: Figuring out whether there was a reliable antibody test; learning whether people keep antibodies around and for how long; understanding if people who had a negative test or no test earlier in their Covid experience could get a positive antibody test, and thereby prove that they had been infected, and thus gain access to services restricted to “positive-only” patients. We did this with the long-haulers group because there are a number of people who are sure they were sick with Covid, but could not get any positive testing and were therefore in limbo.
Read more about what we did, how we did it and why here.
Surprising and not surprising
It is both surprising and not surprising that the results were inconclusive. The scientists who analyze antibody
tests have long said that reliability is an issue. We reported way back in April that the antibody tests were undependable, and the situation has not gotten much better. After the early part of the pandemic, we believed that the science would improve, and that the reliability of tests for active infection and for presence of antibodies would improve. Sadly, that’s not really the case. Scientists continue to report that antibody tests are less than reliable.
We got what we think is one of the best antibody tests on the market (see why we picked this one here), to understand if people were getting negative results because they had an imperfect test. We can now safely say that’s not the only reason.
Why? Well, some people seem to flush antibodies relatively quickly — or never show them at all.
The story of unreliable antibody tests has yet to be told, but this recent news report of studies, from The New York Times, was informative. “Two officials at the Food and Drug Administration said on Saturday that they had erred by allowing manufacturers to sell Covid-19 antibody tests that had not been proved accurate, flooding the United States with unreliable blood tests early in the pandemic,” The Times reported.
Positive, then negative
A number of people who did have positive tests early on are negative on this and other antibody tests — for reasons we don’t fully understand. I myself had antibodies on this test 9 months out, though I never had a positive Covid test for active infection, and was only very mildly ill for a few days in March. Meanwhile, long-haulers who were very ill and had positive tests were negative on this 3 months later.
We do know that antibody presence does not confer immunity for sure. Also that there are other kinds of immune responses.
Donald J. McNeil Jr., then a New York Times reporter covering Covid, wrote in an email: “Immunologists think primed T cells and B cells are an important part of the immune response, and they aren’t measured by antibody tests (or anything else that can be done easily.)”
A friend who was in the Moderna trial, and did get the vaccine and not the placebo, did this test at my behest. She was negative. Why? McNeil answered: “If she got the vaccine, she absolutely should have antibodies. That’s what all vaccines do — prompt the immune system to make antibodies. But they would only be to the spike protein, not to all the different parts of the virus. So it would depend partly on how good the antibody test was — but mostly on what antibodies it was looking for. For example, some tests look for antibodies to the nucleocapsid protein, not the spike protein. She would not have those.”
Dr. Avindra Nath, an infectious diseases specialist at the National Institutes of Health, said in a video interview that there are other tests to determine if people had previously been infected, though they are uncommon. He said they are called “antigen stimulation tests” or “cell proliferation tests.”
“Antibody levels tend to go down in patients after a few months of having had the original infection,” he said. “What you can do, however, is take people’s cells, their T cells, the lymphocytes, and then treat them with, let’s say, the spike protein. And then if the cells had any memory of having been exposed to it before, they start proliferating. That’s one way to see if they have been exposed. But those tests are not commercially available and have to be done in specialized research labs.”
Marty Makary, a Johns Hopkins professor, wrote in The Washington Post recently: “Having the infection activates both antibodies as well as memory B- and T-cells, which teach your immune system to recognize the same virus in the future to swiftly eradicate it. Even mild infections appear to elicit a persistent and functional immune response. One recent European study found that people who had mild or asymptomatic covid-19 mounted a ‘robust T-cell immunity’ afterward. A separate French study affirmed this, noting that some people who lived with a confirmed covid-infected person developed T-cell immunity even when they did not test positive for covid.”
But there’s not a T-cell test that’s easily available. So we were left with a conclusively inconclusive citizen science project. Thank you to our participants!
Responses from our project
Here are some of the responses from the people in our project.
The test looks for both IgM antibodies, the rapid response team, which tends to diminish fairly quickly, and IgG antibodies, which hang around longer. My test in December showed IgG but not IgM, which would be logical since I was 9 months out from being sick.
Leann, a teacher from North Carolina, wrote: “Looks like no antibodies to me. I likely had Covid in early March. There were no tests at my local hospital at the time but a student in one of my classes was hospitalized that week and later diagnosed with Covid. I’m not surprised I don’t have antibodies anymore. My case was pretty mild and it was in March.”
A 44-year-old nursing student and academic tutor from Blue Island, Ill., was positive for IgG antibodies. She wrote: “I was surprised to test positive for antibodies. I was not surprised that it was negative for IgM antibodies as those shouldn’t last very long at all. It has been nearly 9 months since I was sick and still experiencing serious and debilitating Long Covid symptoms as are my children (ages 21, 17, and 14. The 17-year-old is experiencing more severe symptoms then the other two kids). I have mixed feelings on what this means as I am aware that people with antibodies should have some protection but to my knowledge we do not know how much. I am also aware that sometimes antibodies seem to be the type that attack body instead of or along with the virus it is supposed to protect against. I am hopeful it is the protecting kind and enough protection to keep me from getting sick again but will continue to take the same very strict precautions for myself and my family as to not play Russian Roulette with this virus. That being said I still need to get through my nursing classes and clinicals and can only pray that with the proper PPE being used correctly, and with the right type and amount of antibodies, hopefully my family and I can be spared going through this again. Due to contraindications, I am not a candidate for the vaccine which is devastating news for someone like me that nearly died from COVID-19 and is still fighting for my life 9 months after the fact.”
Peg, a massage therapist from Wisconsin: “Test appears to be registering as ‘Invalid,’ Followed directions however kit was in mailbox for an unknown amount of time (1-4 hours) in freezing temps. Was really hoping to get a result other than invalid as my primary care refuses to run an antibody test on me. Tested positive on 10/30 and negative on 12/21.”
Amanda, an archaeologist from Missouri, tested positive. She was ill with Covid two times. She explained: “March 6 (one year ago today) which was three weeks before we had tests in Kansas City, and October 8, tested positive October 12.”
Michelle, an accounting clerk from Yankton, S.D.: “My antibody test was negative. That saddened me more than it should have. I am almost 10 months since symptom onset and have failed every test, covid or medical. I was a double-negative (nasal pap and antibody). Now I am a triple negative (this antibody). There is such divisiveness regarding covid and the way long haulers are treated. That is even worse when you have no proof that you even had covid. There are doctors, employers, friends, and even family that treat you like you’re lying or exaggerating. I can’t imagine why I would make this up.
“I have no doubt that I had covid. My symptoms don’t align with anything else, and I have a large number of symptoms that have continued for almost 10 months. While I understand that this is a new disease and these are newly created tests, our society isn’t prepared to deal with the reality that tests can be wrong. Until accurate long hauler testing is created, I will live under this cloud of doubt.”
Mary from Alabama wrote: “Thanks for the opportunity to participate. I tested negative. This makes sense to me since I tested negative in May about 6 weeks following my positive test for coronavirus. I didn’t trust the physician who administered the test (he called it the China virus and did not know that I was there for a post covid cardiac review until well into our encounter), and so that impacted my trust in those test results. I’m now feeling a little disappointed that I don’t have the antibodies but do feel validated in my choice to continue observing precautionary measures despite having already had the virus.”
Layla, a social worker from Michigan, wrote: “I believe I was positive with symptoms starting 3/20/20. I never had a positive test as there were not tests available at the time.
“I am reading this as negative. I am not surprised as it’s been many months since I first felt sick. However, I am disappointed because I still don’t feel 100% yet. Thank you!”
Shannon, a stay-at-home mom from North Carolina, said, “my results were negative.”
“I’m disappointed. I was very sick in June with what my doc says was probably Covid. My test was negative. I wish I had done more than one, but I was too sick to deal with that.
“I had pneumonia in Nov and Dec – my first time. I didn’t think it was Covid (started with sinus infection symptoms) and was tested four times, all negative.
“I wish I could get confirmation that my sickness in June was Covid but hopefully this will all be behind me soon and it won’t matter. (I have C. diff now from my antibiotic use in Nov-Dec).
“It feels like I’ve been sick since June bc of lingering symptoms from June to Sept, then jumping to the November illness.
“Thanks for the opportunity to test my antibodies. I really appreciate it. It was all easy to do though the printout image of supplies was hard to read and may be challenging for others who don’t know how to use the equipment already.”
Jean, from Florida, tested negative. “I guess I’m not much help, but this far out from Jan it doesn’t surprise me. Thank you for allowing me to participate.”
Pat from central Indiana wrote: “Negative. Surprised I was negative. I just expected antibodies, I guess. I was sick back in January and February, so no covid test available.”
Chantelle from Austin wrote: “I have attached the result of the home antibody test. This shows that I am negative for antibodies. This is good to know but scary because it lets me know I am at risk for reinfection. While I know antibodies fade, I was still surprised by this result. However, I am glad to know, as I can be extra vigilant and discuss with my dr, especially as it may relate to me being vaccinated and handling returning to work as a nurse, while still dealing with longhaulers symptoms.
“I tested positive for covid on August 8, 2020, I was tested at the occ health dept. for my job. I have had 2 negative tests since then and prior to this test have never tested for antibodies. I had what they would classify as a mild to moderate case as I wasn’t hospitalized and despite me having asthma I did not have any lung involvement so they say. I did have periodic cough, chest tightness and pressure, some sob, in addition to the constant fatigue, body aches, tinnitus, tingling, blackouts, etc. I am still having symptoms that are severe to me. Along with the previously mentioned symptoms, minus the cough, I still have increased hr when standing and walking, along with feeling faint, balance issues and seizure like episodes, to name a few.”
Deborah, a retired teacher from Tampa, Fla., wrote that she “didn’t test positive in August or November but caught the virus while on a cruise in January 2020.” On the test, she added: “Unfortunately my husband used too much buffer and I didn’t get results so it’s invalid. I am a long-hauler so have ongoing symptoms that change almost weekly. Just finished treatment for my 3rd bout of pneumonia since Jan.2020.”
Sue from Wisconsin wrote: “Hello, I did my test yesterday. Results were invalid. I guess I am not sure what invalid means, user error? It was easy to use and I liked being able to do it at my home. Thank you for the opportunity to participate and help!”
Amanda from Texas wrote “no antibodies.” She tested positive for active infection in July.
Stephanie, a photographer from Savannah, had an invalid result. “Nothing happened even though I followed the instructions exactly as it said.” She tested positive for active infection in late July at the health department in Savannah and never tested positive for antibodies. She is still experiencing long-haul symptoms.
Heather, a stay-at-home mom and teacher from Michigan, wrote “Here are my results, negative. I wasn’t surprised, but I’d hoped my body would show some. Thank you for the opportunity to check for the antibodies. “
She fell ill in September and tested “inconclusive” for active infection. She has never had an antibody test before.
What you can do
In this part of our coverage, we try to marshal our best knowledge on what you should do.
- If you want to take a test, be clear about why you want it.
- Even if you do find a test that says you have antibodies, there’s no evidence that this provides short- or long-term immunity from Covid. So … get vaccinated, and continue to observe masking and distancing practices. Even after you’re vaccinated, these practices can save lives.
- We have learned through our reporting that many antibody tests are not covered by insurance. We surmise that this is because they are a) unreliable and b) not particularly predictive of any health status, so providers and insurers and the government are increasingly reluctant to pay for them. So you will probably be paying out of pocket. We have heard of some wildly inflated prices for antibody tests, so if you want one, you might want to — after asking yourself why you want it — shop around for both a measure of reliability and some comparison pricing.
- What’s the best antibody test? When we started out doing this, we asked that question several times. We wanted an easily available, affordable test that had some bona fides from scientists who know more than we do. We wound up with this one, which we bought from a manufacturer in Wuhan, China, because it was used in a study that I was in at Mount Sinai Health Systems in New York. But for the reasons mentioned here (durability of antibodies, significance of antibodies) we are not certain that antibody testing is relevant or useful for most people.