In this episode of Tiny Show and Tell Us, we read an email from “baby sis” aka Binky aka Sam’s younger sister Caroline who writes in about an alarming pregnancy test that predates today’s at-home tests. Then we talk about a fascinating study that found a dead salmon showed brain activity in an MRI machine. Spoiler: It wasn’t actually alive, scientists just really needed to rethink MRI analysis.
Transcript of this Episode
Sam Jones: Welcome to Tiny Show and Tell Us, the bonus series where you write in with your favorite science news or factoid, we read your email aloud, and then dive deeper. I'm Sam Jones and I'm here, as always, with my co-host Deboki Chakravarti.
Deboki Chakravarti: Hi, Sam. Last time we did this, we talked about insect vibrations, we also talked about longstanding grudges against bad science teachers.
Sam Jones: Important.
Deboki Chakravarti: Yes, very important. And of course, a reminder for all of you listeners that, if you have a story you'd like for us to cover on Tiny Show and Tell Us, all you have to do is write us an email, so you can email us at tinymatters@acs.org or fill out this form linked in the episode description.
Sam Jones: All right, cool. Deboki, I am going to just go first today.
Deboki Chakravarti: Yeah.
Sam Jones: How does that sound?
Deboki Chakravarti: Go for it.
Sam Jones: I'm going to start with something a little bit different, because I'm going to start with an email. There's no science to unpack in this first little thing I'm going to share.
Deboki Chakravarti: Got it.
Sam Jones: But I got an email from listener Cecilia, and I was out when I got this email, so I was on maternity leave. A lot of people don't know that, because we were publishing episodes, but we had prerecorded a lot of materials.
Deboki Chakravarti: I thought you were about to say, "We had a baby."
Sam Jones: We collectively had a baby.
Deboki Chakravarti: Yes, yes. All of you, you also all have a baby now.
Sam Jones: Yeah, so if you could send money, diapers, food.
Deboki Chakravarti: Yes.
Sam Jones: Yeah. Cecilia got in touch with us, I think, a week after I gave birth, so I was not monitoring the Tiny Matters inbox, and then I just emailed her back a few days ago. She forgave me for my tardiness.
Deboki Chakravarti: Very kind.
Sam Jones: And then also said it was fine to share this email, so I'm going to hop into it. And Cecilia writes, "Hello. I've been catching up on the Tiny Matters podcast, binging many of the episodes I missed, and wanted to send an email with some info I thought you might find interesting. I'm the listener who shared the dog's love you proven by MRI on Show and Tell Us number six. This May, I graduated with my bachelor's degree in chemistry, fittingly, right beside my service dog. Having a service dog in chemistry is something I had never heard of before prior to getting one. The college president awarded my service dog with a special patch in recognition of her work, as my service dog was trained to be in the lab with me. Over the years, my service dog has attended many labs and also aided me throughout my undergraduate thesis research in inorganic chemistry, so it felt like a great achievement."
Deboki Chakravarti: Yeah.
Sam Jones: "However, unexpectedly, over 2.6 million others also agreed. On my Instagram account, which is @salad_and_crouton, we went internet viral and ended up being featured on ABC World News, Good Morning America, Fox News, and people.com, along with many other news outlets." Incredible.
Deboki Chakravarti: Wowee.
Sam Jones: I know. And then Cecilia continues. "ACS has been a huge part of my growth throughout the years and has one of the best communities I've seen. Inspired by this community, since 2022, I've been presenting at various events and conferences, advocating for myself and others on the topic of service dogs, invisible disabilities, and navigating STEM careers with a service dog, so I thought it would be interesting to share this news with Tiny Matters, especially as I know the hosts are big dog lovers." I know, Deboki, you have a cat, but you say you like dogs-
Deboki Chakravarti: I love dogs, they're great. Yeah.
Sam Jones: We know I'm pretty obsessed with dogs.
Deboki Chakravarti: And Cecilia's dog sounds brilliant.
Sam Jones: Yes. Brilliant, adorable, all of the things. Then Cecilia closes out by saying, "Thank you for your time in reading this all-over-the-place email." It really was not. Cecilia, don't worry. "Attached below will be the original video link and the ABC video that started it all. Have a wonderful day, Cecilia Hernandez." You can follow Cecilia on Instagram and YouTube @salad_and_crouton. We've linked to the Instagram reel and the ABC video that Cecilia mentions in the show notes, and they're worth the watch. And you know, I was mentioning to Deboki right before we started recording that I saw this. I saw this go viral back many, many, many months ago. Had no idea that Cecilia had then written into us after it went viral, so it's like six months later, I'm getting caught up on everything, but this is one of those things that went viral but is worth revisiting, because it is heartwarming and... Oh, my gosh. Cute dog.
Deboki Chakravarti: And congrats on your degree and everything, too.
Sam Jones: Yeah, congratulations. That's super exciting. I just loved that, I just loved it so much. That is a different way to start a Tiny Show and Tell Us, but I thought it was just so wonderful and it was sent in so long ago, and I was just fully unaware.
Deboki Chakravarti: Yeah.
Sam Jones: Now we are getting into a Tiny Show and Tell Us, we're going to break down the science. This one is from listener Binky. I say Binky, it's my sister. Her name is Caroline. I've talked about her on many episodes, but this is from my sister and this is what she wrote. She wrote in saying, "Hi, Tiny Matters baddies, baby sis here. I've been waiting and waiting and waiting," in all caps, "To have a science fact for you all, but alas, I'm not a science gal, so it took a second. This past weekend, I was out with a gal pal who told me she was pregnant." Yay.
Deboki Chakravarti: Congrats.
Sam Jones: Yeah, congrats. "We then proceeded to talk about pregnancy tests and the accuracy between different brands, types, and how expensive they are. At the end of this conversation, my friend said, 'Well, it's better than killing a bunny,' to which I said, 'WTF?' She then proceeded to tell me how her mom told her that, in our grandmother's generation, they would inject a woman's pee into a bunny and, quote, 'If it died, then she was pregnant.' I did a little digging and, LOL, that's not quite it." And then Binky puts in quotes, "I'll let you science chicks explain. But honestly, it's worse, because they would kill every single bunny, pregnant or not, to find out if you were in fact pregs by killing, dissecting them to see if their ovaries were enlarged. Wow, terrible. Thank goodness for ever-evolving scientific research and practices for both me and my fondness of bunnies, and the bunnies on a whole. Thanks for doing what you do and entertaining me for 100-plus episodes."
You're welcome, Caroline. Let's hop into it, let's get into this one. We're going to start with the now. Home pregnancy tests used today have been around since the late '80s, early '90s to detect the hormone human chorionic gonadotropin, or hCG, in urine around six to 10 days post conception, which is three to four weeks after the start of your last menstrual cycle. It's a lot of ballparking, but generally speaking, that's when it can be detected. It's produced mainly by the placenta during pregnancy, which just goes to show you how early placental cells are making things happen. You want to hear more about the placenta? Go listen to our placenta episode.
Deboki Chakravarti: Yes.
Sam Jones: hCG rises quite fast once a fertilized egg implants in the uterus. Home test strips, they are strips that have antibodies on them that hCG will bind to. Really similar to a COVID test or a flu test. Any of those ones, where you get a nose swab or whatever, it's kind of that, but urine is used instead. If you get that colored line, that's the antibodies binding to hCG and then actually binding to another bit of an antibody strip. The main thing is, if you have hCG and it binds to the antibodies, you're going to get that colored line or you're going to get a test that says "Pregnant" or "Not pregnant," but that's just a digital readout of the chemical reaction that's happening on the strip.
Deboki Chakravarti: Right.
Sam Jones: Just takes even that tiny bit of guesswork out of it for you.
Deboki Chakravarti: Yeah, which especially if you're testing super early and you're like, "I think I'm pregnant, but you're not sure," and then you're looking at that strip and it's a total optical illusion sometimes, where you're like, "If I put it in the light this way, it looks like it's positive, but if it's this way, it doesn't."
Sam Jones: Yeah. Very similar to a COVID test, where you're like, "Is there a faint line? Do I have it?"
Deboki Chakravarti: Yep, yep.
Sam Jones: Yeah. And then, of course, to detect pregnancy earlier, you can have a doctor do a blood test. That is, by far, more accurate. You can detect the concentration of hCG in the blood, which will increase exponentially at the beginning of pregnancy.
Deboki Chakravarti: Yeah, and so that's also really important for tracking. If you have a healthy viable pregnancy going on early on, they'll use the blood test for that.
Sam Jones: Right. Typically, I think it's an exponential growth curve. You're seeing double-
Deboki Chakravarti: Yeah double every day, I think.
Sam Jones: Yeah, so you'll go in and then they want you to come back a few days later to double check. What happened before the late '80s, '90s, when we had this available? Well, in the '70s, the FDA approved what the Association for Diagnostics and Laboratory Medicine refers to as, quote, "Many chemistry sets that required users to mix urine with solutions in test tubes and wait two hours for a result."
Deboki Chakravarti: Two hours.
Sam Jones: Yes, yes.
Deboki Chakravarti: Even waiting two minutes is agonizing.
Sam Jones: Yeah, two hours, two hours. These were annoying to use, which by the description is clear that it would be annoying, and apparently, they had a false negative rate of around 20%, which is quite high.
Deboki Chakravarti: Yeah.
Sam Jones: I found an incredible article in the conversation about the history of pregnancy tests written by Helen King, who is a PhD historian of medicine and specializes in the history of obstetrics and gynecology. And she also wrote a book that came out, actually, in 2025, called Immaculate Forms: Uncovering the History of Women's Bodies. That sounds really interesting. Before we get to the bunnies specifically, let's talk urine as it relates to pregnancy. Three ancient Egyptian papyri, I guess the plural of papyrus, which I hadn't ever known before.
Deboki Chakravarti: Yeah, yeah, yeah. Is it papyri, papyri? Papyri.
Sam Jones: Papyri, papyri show that urine was being used as long as 4,500 years ago to determine if a woman was pregnant, and I'm putting "Determine" in air quotes. In these texts, a woman is described who either wants to know if she'll conceive or if she's pregnant, really different things, by urinating on wheat and barley over the course of several days, and then they say, if the barley sprouts first, it's a boy. If the wheat grows first, it's a girl, but if no seed sprout, she isn't pregnant, which is like, "Okay." Yeah, it sounds a little wonky, but thousands of years later, in the 1930s, people were reevaluating these seed tests. They found that, 70% of the time, a pregnant person's urine did make the seeds grow. Again, it's more than 50/50, but still, it's 70%, so this is not foolproof.
It's not a great error rate, but also the whole... Also, the “if barley sprouts, it's a boy, wheat sprouts, it's a girl,” of course, that has no merit whatsoever. But the idea that there was something detectable in a pregnant person's urine was legit. There's something going on, and so around the same time, here's where you get into the bunnies, Caroline, sorry, but also, you asked for this explainer, so that's on you. Rabbits were being injected with a pregnant person's urine and then killed to see how their ovaries changed. The researchers doing this would look for larger ovaries and the corpus luteum, which is the endocrine gland that's made up of a yellow mass of cells and lipids that forms in the ovary after ovulation. But of course, to do this check, the rabbits had to be killed.
Later, live toads were used instead of rabbits and injected with urine, and if a person was pregnant, then the female toads would release eggs. The good news was that these toads weren't typically killed, but the bad news is, I found out, that their mass export may have spread a deadly fungal disease to wild amphibian populations across the globe.
Deboki Chakravarti: Excellent.
Sam Jones: Yeah. Yeah, because when I first read about that transition to toads, I was thinking, "Great." Is it great for them to live in a lab where they're injected with urine? No, but you don't need quite as many of them, you're not cycling through them the way that you would if you had to kill them every single time, but then I found out about the whole deadly fungal disease thing and...
Deboki Chakravarti: Yeah…
Sam Jones: Yeah, I don't know. I'm like, "What's better? I'm not sure." Yeah, none of these tests were great or ethical, but thankfully, antibody work, particularly in the 1960s, led to what we now have available today, which is a highly-accurate at-home test that does not use animals. Binky, I hope that you are feeling informed and-
Deboki Chakravarti: And also grateful to have not had to be pregnant at a time of rabbit tests.
Sam Jones: Rabbit tests, toad tests, or having to, essentially, use a mini chemistry set and then wait for two hours.
Deboki Chakravarti: Or having to go pee out in a field.
Sam Jones: Yeah. Yeah. Yeah, go get that barley.
Deboki Chakravarti: Of all the ones we mentioned, that's the one that I'm like, "Yeah, you know what? That one, I can handle."
Sam Jones: Yeah, me too.
Deboki Chakravarti: I don't like the idea of having to wait to see if the seeds are going to grow. By that point, back then, you wouldn't like...
Sam Jones: It sounds like they had to go back multiple times, so you just keep going back and peeing in the same spot day after day.
Deboki Chakravarti: It almost feels like, by the time you get the results from that test, you'll start to feel that you're pregnant, you'll start to potentially have started to show and have more signs, but it was also a very different time. You couldn't ultrasound.
Sam Jones: 4,500 years ago, not a lot of options. I was reading also that, and I believe it was Ancient Greece, there was something where it was like, "As soon as you've conceived, you can feel your womb close," and it was like, "Yeah, what?" There have been so many things, also then, related to blood or urine or, "If you bleed on this type of cloth and then wash it this type of day."
Deboki Chakravarti: Right.
Sam Jones: There's so many things, but I thought that it was really interesting that there has been this urine thread for thousands of years and that, although peeing on barley or wheat is not an accurate way to see if you're pregnant, it was just interesting that people were already thinking, "There must be something in urine specifically that is an indicator of pregnancy."
Deboki Chakravarti: Right.
Sam Jones: And I thought that was cool.
Deboki Chakravarti: For sure. Well, Sam, I have a story for you from listener Madison. She sent us a link to this study and wrote, "Ever since I heard about this study, I could not get it out of my head. It's an excellent story about false positives and accounting for natural variations in data. It is also an excellent mental image, a dead fish lighting up an MRI machine. I would love to hear about how this is being implemented today and the lasting impact it might've had. Love the podcast."
Sam Jones: I remember this.
Deboki Chakravarti: Yeah.
Sam Jones: Also, when I saw that someone wrote in with this, I remember also seeing this and thinking, "This is fascinating," and I'm so glad we're talking about it.
Deboki Chakravarti: Yeah. For those of you who are like me, who actually didn't know about this, Madison's link is to the story from improbable.com, and it's actually about a recent study looking about the reliability of functional MRI to find biomarkers of disease, but it's also about this older study, which is where the dead fish lighting up an MRI machine comes up, so we're going to get into both of these. Some background information, functional MRI looks at blood flow, which scientists often use as a proxy for brain activity. The idea is that, if you have neurons in a part of your brain that are being more active, then there will be more blood flow going on. Something you might do in an experiment is have your subject do some task and use the MRI to look at their brain and see where there's activity going on.
There's this more recent study that the news article is about, which gets into how reliable FMRI is for that, and I'll get into that in a bit, but we're going to go first to the dead fish lighting up the MRI machine, which the article talks about too, and I want to get into, because I didn't really know that much about this. I didn't really do much brain stuff, so this was newer for me. Yeah, so in 2012, the Ig Nobel Prize for neuroscience was given to a group of researchers who bought some dead Atlantic salmon from the store and put it into an FMRI machine. They showed the salmon pictures of people doing different social things, they asked the salmon to figure out what emotion the person in the picture was possibly experiencing, and again, when I say all of this, I emphasize that the salmon was dead, so you really should not be expecting any brain activity, right?
Sam Jones: Right.
Deboki Chakravarti: But the FMRI data showed that the dead salmon had brain activity, it was like thinking about the pictures. This was all just a big, "What? They should not be showing this."
Sam Jones: Yeah. I will say also just a side note on the Ig Nobel Prize, is that it's a prize that I think on their website says it honors achievements that first make people laugh and then make them think.
Deboki Chakravarti: Yeah.
Sam Jones: And it's been around for going on 35 years. It's been around for a while and it's a take on the Nobel Prize. It's like the funny, fun one, but what's great about it is that you're actually learning a ton from these silly... The research that has a silly premise, but is legitimate research. And we do have an episode on the Ig Nobel Prize that came out I think the end of our first year or maybe second year. I'm forgetting now. But anyways, I just wanted to interject in case a listener hadn't heard of the Ig Nobel Prize.
Deboki Chakravarti: Yeah, and that's a good point, because that's actually the story of this paper. This was a paper that was originally rejected by journals, but those scientists presented it as a poster at a conference, and researchers started forwarding it to each other and it ended up leaving its mark on the field. Because basically, one of the things that this paper was showing is, depending on how you filter your data, you can get results like this, so it raises a lot of questions about how to best analyze this data, because one of the challenges is like, yes, you can create a strong filter on your data that's going to say like, "Actually, this dead salmon doesn't have any brain activity," but then you run the risk of actually losing data that's also real and that does reflect actual brain activity.
Sam Jones: Right.
Deboki Chakravarti: This paper has had this longstanding effect on the field, including this more recent work that this article was about from 2020. We're fast-forwarding from 2012 to 2020 and scientists are still trying to figure out the reliability of FMRI, and there are these researchers who are working on a long-term study of 1,300 Duke undergrads looking for biomarkers that might explain why people process thoughts and emotions differently. And one of the ways that they're doing that is looking at brain scans, but they're also wondering how reliable these measures will be. And one of the central questions in this case is, "Just how consistent can FMRI results be in an individual?" If you measure them on one day on a task, are they always going to have that same activity on a task if you measure them a week from now, two weeks from now? And that's really important, because if you're using their brain activity as a biomarker, you need that biomarker to actually be consistent.
It needs to be something that you would see reliably. As part of a different study, some of the researchers I think from this group, but also just part of this team of researchers decided to see the reliability by looking at a number of different sources. One of their sources was actually 56 papers that used FMRI data, they found that the correlation between multiple scans on a person were not that great. They also looked at something called The Human Connectome Project, specifically looking at tests of seven brain function measures in 45 people and then retests of those from four months later, and the correlation between those test results was pretty weak. Also looked at data from a study in New Zealand, where people were put through a task-based FMRI twice to space two to three months apart. Again, very poor correlation. It's just showing that this individual measurement is really not that reliable.
But this also doesn't mean that FMRI data is necessarily bad, it just means we have to be really thoughtful about how we use the data. For example, we might not be able to use it to say specific things about a specific person's brain, but we might be able to average it out and understand population level responses in general. We might be able to say things about the structure of the brain looking across a group of people, and there also might be ways to improve on FMRI. Maybe instead of collecting data, I think more on the order of minutes, maybe we should be collecting that data for longer on the order of hours or using the data to see more about how parts of the brain are connected, rather than just looking at what parts are active. I think what it points to is that this field is still figuring out how to use FMRI data and what's the best way to use this information and to think about it going forward.
Sam Jones: That's really cool. I'm so glad you got into that, because I knew about this study, but I actually just knew that it existed and didn't think about the implications, or realize that the field took it seriously enough to then be reevaluating how we use MRI data or how we analyze and then interpret that data, and that's a really important part of science.
Deboki Chakravarti: Yeah. Yeah, I think it's really good that the scientists working in the field aren't just taking it for granted that they can rely on their data points this way.
Sam Jones: Yeah, absolutely.
Deboki Chakravarti: Thank you to Cecilia, Madison, and to Caroline, Sam's sister, for submitting to Tiny Show and Tell Us, a bonus series from Tiny Matters, created by the American Chemical Society and produced by Multitude.
Sam Jones: You can send us an email to be featured in a future Tiny Show and Tell Us episode at tinymatters@acs.org, or you can fill out this form that's linked in the episode description, and we'll see you next time.
References:
- Check out Cecilia’s viral reel and Follow salad_and_crouton on Instagram
- Rollins College Awards Service Dog During Commencement
- A brief history of pregnancy tests – from toads and rabbits to rosewater
- NHS: Doing a pregnancy test
- How a pregnancy test caused a catastrophe for frogs
- Egyptian Papyrus Reveals This Old Wives’ Tale Is Very Old Indeed
- fMRI Brain Research: The Dead Salmon Has Lots of Company
- Studies of brain activity aren’t as useful as scientists thought
- Scanning Dead Salmon in fMRI Machine Highlights Risk of Red Herrings
