Think Out Loud

Switching arms between vaccine doses could dramatically improve immune response, OHSU study shows

By Gemma DiCarlo (OPB)
Feb. 16, 2024 5:57 p.m. Updated: Feb. 26, 2024 11:07 p.m.

Broadcast: Tuesday, Feb. 20

A woman in a black shirt gets a shot in her upper arm. All you can see are the patient's shoulder and the provider's hands.

A new study from Oregon Health & Science University shows that switching arms between vaccine doses could lead to as much as a fourfold increase in immune response. (Photo provided.)

OHSU/Christine Torres Hicks


Until recently, doctors didn’t think it made a difference which arm patients got their COVID-19 vaccine in. But a new study from Oregon Health & Science University shows that switching arms between doses could lead to as much as a fourfold increase in immune response. Though further studies are necessary, researchers suspect the improvement has to do with how the body learns to protect itself against viruses, meaning the results could be replicated in other multidose vaccine regimens.

Marcel Curlin is an associate professor of medicine and the medical director for occupational health at OHSU. He joins us to talk about what these findings could mean for future vaccine protocols.

This transcript was created by a computer and edited by a volunteer.

Dave Miller:  From the Gert Boyle Studio at OPB, this is Think Out loud. I’m Dave Miller. Until recently, doctors didn’t think it made much of a difference if people got their doses of COVID-19 vaccines in the same arm or in different arms. But a new study from Oregon Health and Science University found that switching arms between doses led to as much as a four-fold increase in immune response. Marcel Curlin is the senior author of the study. He’s an Associate Professor of Medicine and the Medical Director for Occupational Health at OHSU. He joins us now to talk about the implications of these findings. Welcome back to the show.

Marcel Curlin: Thank you, Dave. Good afternoon.

Miller:  Whenever I’ve gotten vaccines for COVID in recent years, but even before that, they would say, “Which arm do you want?” And I would say, does it matter? And they’d always say, “no.” The only thing I can remember people saying is maybe put it in your dominant arm because then it’ll be less sore. “The soreness will go away faster because you use your muscle more,” is what they would say. How much evidence was there in the past for the efficacy of vaccines based on where the jab is going?

Curlin: There really wasn’t any. I think that basically that advice you got, that we all got and that I got was based on the assumption that it really just didn’t matter.

Miller:  Is that common in medicine, advice based on assumptions as opposed to data?

CurlinI think if we’re really offering a positive intervention of some kind, like taking aspirin because it’ll reduce heart attacks, we wanna have data for that. But in this case, we didn’t have any data either way. So there’s no reason to really suggest a preference.

Miller:  Why do you think it wasn’t studied before?

CurlinThat’s a fascinating question. I really don’t know. Vaccination has been around since the late 1700s in its formal form, but has probably been around for even longer than that in sort of more traditional practices. And as far as I can tell, it’s just never been looked at.

Miller:  You started this study in the early days of the pandemic, in 2020 in an overwhelming… I think both maybe societally and maybe scientifically… a confusing time. What was that like just as a scientist in those early days?

CurlinIt was a really bewildering time. I think it’s worth remembering that we had forest fires. We had COVID bearing down on us. And it felt like a very unstable and unsettling time. And when the vaccines finally were arriving, there was a great deal of excitement, that now we had something that we could use to help protect people. So there were all of these currents going on. I think there was also a sense of urgency among those of us that wanted to study COVID and immune responses. With the vaccines coming, we needed to quickly get some kind of a study going so that we could get baseline values on things before people got vaccinated. So there’s quite a bit of urgency.

And I think it was also a really fun moment. Because in a weird way, it was a time of creativity. We could just come up with a study. And really there was no funding associated with this project. It was just all on the back of a napkin, so this moment of real creativity and, “Can we do something?”

Miller:  Am I right that the subjects for this study were all your colleagues, OHSU employees?

CurlinYeah, that’s right. We were vaccinating about 1000 people a day. We had to vaccinate 20,000 OHSU employees in a pretty short amount of time, some of the earliest people in the state, I imagine, to get vaccinated.

That’s right because they were health care workers. And so people were coming through Occupational Health at OHSU and we had an army of medical students there that really couldn’t do their regular rotations. They didn’t have much to do but they wanted to help with this project. So I taught them how to draw blood and we consented people as they came through. We asked them for their consent if they wanted to join the study. And we enrolled about 2,000 people.

Miller:  So this was something you set out to study from the beginning. This wasn’t an accidental accumulation of data that you could then look back on after the fact and say, “Actually we can just see if they got it in the left or the right or the left and the left.” You did this from the beginning?

CurlinYeah, this was a prospective study. We knew before the vaccine was available that we wanted to study immune responses. And so we prepared a protocol, submitted it to our IRB, Institutional Review Board, for ethical approval. They said, “Yes, go ahead.” And then as the vaccine became available and we invited people to join we took blood before their first vaccine. And then when they came back for that second vaccination, it occurred to us that we should look at arm preference.

So we randomized the entire list of participants, to either the same or opposite. And when they came back for that second vaccine, we said, “Would you like to participate in this arm randomization?” If they said yes, we gave them what their preassigned category was and had them pick that arm. Then we had them come back for a few more visits after that. And we studied immune responses.

Miller:  What did you find in terms of immune response? And maybe let me take a step back. What are you charting in terms of immune response? Then we can talk about the differences you found?

CurlinIn a nutshell, we’re looking at antibody responses. Antibodies are these proteins that circulate in our blood. They’re specific for a certain foreign invader or pathogen. And when they encounter that they glom onto it and basically neutralize it. That’s how antibodies work. So we’re looking at people, taking their blood, and measuring the level of those antibodies.

Miller:  What did you find?

Curlin: If you switch arms and get your first dose in the right arm and then come back and get your second dose in the left arm, you have about a one-and-a-half to two times higher antibody response over time, compared to people that got it in the same arm. And we’re talking about antibodies specifically directed at the type of virus that the vaccine was made from.

Miller:  The first version, the original COVID?

CurlinAnd what was really surprising about this result was that that effect seemed to increase over time. So we studied out to 1.1 years after the randomization. And we found that the difference got larger and larger.

Miller:  Is that because there was a decrease in immune response among people who had it in the same arm and the other was constant? Or was it just an increase in the two-arm cohort?

CurlinYeah, it was a relative effect. So in everyone, antibodies peak after vaccination and then wane again. And our immune system is designed to be that way. So in both groups, at 1.1 years after vaccination, the antibody response is on the down slope. It’s waning. But if you look at the relative amount in each group, it’s much higher in the [arm.

And then the second fascinating thing about it was that if we looked at antibody responses to a new future virus like Omicron, which at that time had never shown up before - the Omicron would only come in February of 2022…


Miller:  You say new. We also called it a variant. It’s similar, but it had some different qualities?

CurlinYeah, that’s right. It’s a variant of COVID. It’s still in the COVID family but it looks a little different because of mutations. And so if we sort of stopped the clock and said, “We know this virus is coming in a year. Let’s see how your immune response is now,” matched up against that virus. The difference was fourfold instead of twofold.

Miller:  Over the summer, just not that long ago, researchers in Germany made headlines for a study that seemed to come to the opposite conclusion that your study arrived at. They reported that people who got successive doses in the same arm actually had a better immune response than the alternators did. What were the differences in your two studies that help you understand these very different results?

CurlinYeah, that’s right. And that was a good study. It was well done. The only difference was they only measured two weeks after the vaccination.

Miller:  So two weeks after the second dose?

Curlin[Right] so at that point, people that got the vaccine on the same side twice, you’re getting a recall response at two weeks, which is pretty efficient. But beyond two weeks, if you got the vaccine in the opposite arm, now you’re getting two centers of the immune system that are both working. And so beyond two weeks, the advantage becomes more clear that the opposite arm is more advantageous.

Miller:  This reminds me of a finding we’ve heard about in the past about bilingual learners. That early on when a kid is learning two languages at once, they may not be as good at reading or comprehension as people who are learning only one language. But eventually they can sometimes surpass their unilingual peers?

CurlinYeah, that’s a really good analogy. And in those kids that learn two languages, they might even be more adept at learning a third language.

Miller:  Do you think that there’s maybe a limit to taking metaphors from other life and applying it to science? But let’s go to the mechanism of what’s happening because maybe that’s more important. Then we can get back to metaphors if we want to. How do you explain this finding?

CurlinIt gets to how the immune system works. The immune system is quite complicated and there are lots of parts to it. But one major determinant is the ability to make antibodies against the specific pathogen. Those antibodies are made by B cells. So, in some way, we’ve got to get to a point where we have B cells in our blood that are specifically able to make the antibody we need. Essentially, by vaccinating in two arms, the system of immune education happens in the lymph node.

Let’s say I get a cut in my right hand. And I get a little infection. You may sometimes feel lymph nodes swelling up in your right armpit because those bacteria and or parts of those bacteria get moved to that lymph node. And immune education happens. And then our immune response in the future emerges from that learned environment. Now, if I’m gonna stimulate the other arm next, the same thing happens on the other side. So, essentially I have two different systems that are responding. And therefore the breadth and the magnitude of the response is bigger.

Miller:  Even though if there’s an invader and it gets say in your blood, it could be picked up anywhere, right? Then the way you described it, it makes sense. But I’m evolutionarily, I guess, still a little bit confused by the localization of immune response?

CurlinIn normal infections, there’s no localization. There’s a portal of entry, but there’s no localization because an infection is disseminated. When we vaccinate, we’re essentially giving you something that’s inert. Maybe it’s a bit of a protein from that virus or that spectrum. So there’s no systemic infection. That protein gets taken up by professional antigen-presenting cells. They travel through these lymphatic pathways which are basically like little highways that kind of lead centrally into the body. And they stop at a regional lymph node. So it’s a local process actually, in this case with vaccination.

Miller:  And so we’re imagining how this would actually benefit people. So let’s say that you had the booster and the initial COVID vaccine in just one arm. Then you go to a party. You breathe in some COVID virus particles and they go into your system. And then they eventually hit both sides, but only say your right arm has been trained to make these antibodies?

CurlinWell, the immune response is general. So regardless of which arms you got, once you’re immunized in some way, by vaccination or previous exposure, you have an immune response. And it doesn’t matter. The arm part is now no longer particularly relevant. It’s how big that immune response is and how broad it is. How able is it to cope with slightly different incoming threats? So if you got your vaccination in the opposite arm and you go to that party, you have a slightly higher level of antibodies.

Miller:  Because you have, in a sense, two factories that are making it?

CurlinThat’s right. And you have a broader repertoire of antibodies. If you bought shoes at Nike and you bought shoes at Reebok, you might have two different kinds of shoes. Whereas if you bought them just at Nike, you only have one kind. So you go to the party. And now a slightly different variant comes along. You get exposed. You breathe it in but you have great antibodies and they’re broad enough to neutralize. So you’re better.

Miller:  This study is only about the SARS COVID-2 virus. It’s only for one particular virus and one set of vaccines for it. How generalisable do you think this initial data is for other vaccines?

CurlinYeah, it’s a very good question and we don’t know. I think that for inactivated vaccines that do not establish a systemic infection - of course some vaccines that we use, like measles vaccine or something, is live, attenuated, and creates a modest infection in the body. So it’s a different category of vaccine. But the COVID vaccines are inert. They don’t cause any infection. And so I think that this might be general for all inert vaccines. That if we put them in one location and then in another, we get a slightly better response than otherwise.

Miller:  On your own, separate from giving medical advice to people on the basis of one study that has yet to be replicated, will you already make behavioral changes when you get vaccines going forward?

CurlinThis is not medical advice for anyone else but for me personally, if I get a new vaccine that I have never had before and there are multiple doses and it’s an inert vaccine, I will switch arms.

Miller:  Medical data record programs that we all see…half the time doctors, when we encounter them, now are typing and maybe looking at us, maybe looking at the screen. There’s a lot of data that’s being kept. Is there a field for which side of our body we got vaccines on? I mean is that even a record that has a metadata field right now?

CurlinYeah, it depends on which medical system you’re looking at. For example, in occupational health here at OHSU where we record vaccines that were given for the protection of our workers. We do have a field that says specifically which arm. But it really depends on medical systems. In some places, the information is not captured and in others, it is. It’s probably captured a lot in different research studies where they’re quite meticulous about details like that.

Miller:  But if we’re talking about pediatric vaccines or just any other vaccines that people might be getting, it seems like if we’re going to follow your own personal model going forward, it might be up to us to remember this?

Curlin: I know. That’s difficult and that’s actually true. My wife joked that because our last name, Curlin, has an R and an L, we remember to do right first and left second.

Miller:  I have an L and an R so I could do that as well. What do you see as the broadest implications of your study?

CurlinThere are two things that I would point out. One is that if it really is true that there’s a benefit, it could change practice for some people. And for certain vaccines, I think those people who are particularly vulnerable to infection afterwards would benefit the most. So if we take COVID as the example, for most of us, it might not matter that much if we develop a pretty good immune response. But if we’re in that unfortunate category of people that have a predisposing factor for severe COVID or low antibody responses, we might significantly benefit from that.

So it might change practice. I think it might also help with vaccine development. Because there are some vaccine candidates that have never been successful or illnesses that represent a very difficult challenge to develop a vaccine for. For example HIV. And in those cases, anything we can do that boosts the immune system a little bit would potentially push us over the threshold to success.

Miller:  Marcel Curlin, thanks very much.

CurlinThank you.

Miller:  Marcel Curlin is a researcher and infectious disease doctor at OHSU. He is a senior author of a new study that found that switching arms for COVID vaccine doses led to a significant increase in immune response.

Contact “Think Out Loud®”

If you’d like to comment on any of the topics in this show or suggest a topic of your own, please get in touch with us on Facebook, send an email to, or you can leave a voicemail for us at 503-293-1983. The call-in phone number during the noon hour is 888-665-5865.