A Swedish nationwide register study of 2.93 million children born 1982–2010 found that the heritability of childhood-onset type 1 diabetes remained high (~0.83) and stable over 30 years, while changes in measured environmental factors explained only a small fraction of the rise in incidence.
0:00Welcome to Base by Base, the papercast that brings genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. You know, when you look back at the history of human health over the last century.
0:12There is this um, this incredibly compelling global medical mystery, sitting right in plain sight. Yeah, it really is hiding in plain sight. Right, because since the 1950s, the rates of childhood onset type one diabetes have just been skyrocketing.
0:28And, well, when you see a massive population wide disease spike happening in just a few decades, human instinct immediately kicks in to blame the modern environment. Oh, absolutely. We wanna point our fingers at everything.
0:40Exactly. Our highly processed diets or microplastics changing lifestyles, urban pollution. Because if you think about evolutionary timelines, human genetics simply do not mutate fast enough to cause a massive sudden surge in autoimmune disease over a single generation.
0:55No, I mean, it takes millennia to rewrite our code, not 20 years. We all want to find the smoking gun in the world around us. It's just the most intuitive leap to make. If the genetic landscape of humanity is, by comparison, like an ancient Sloven glacier, then a sudden explosion in sick children must mean some new toxic element in the modern environment is doing all the damage.
1:18So we operate on the assumption that the environment is the sole culprit. But as we'll unpack in the deep dive, that assumption might actually be blinding us to the real story. How could this change our fundamental understanding of autoimmune diseases?
1:31What really happens when a modern environment collides with ancient DNA? Well, answering those questions requires us to move past those broad assumptions, and look at human health data on a scale that is, honestly, almost difficult to wrap your head around, just to untangle what is actually nature and what is definitively nurture.
1:48And that data actually exists, which is amazing. Today we celebrate the work of UxioA, Sophia Carlson, and their research team at the Karolinska Institute in Sweden, who have advanced our understanding of the genetic and environmental drivers of type one diabetes.
2:02They've done some incredible work. They really have. And we're basing this deep dive on their article, stable heritability of type one diabetes in a Swedish nationwide cohort study, which was published in Nature Communications on June 17, 2025.
2:18Right. And to really appreciate the magnitude of what this team accomplished, We should probably set the stage with the clinical reality of the problem first. Yeah, let's do that. So type one diabetes, or T1D, is a severe autoimmune condition.
2:33And just be clear for everyone listening, it operates very differently from type 2 diabetes, which is often tied to, you know, metabolic syndrome later in life. Right, they're fundamentally different mechanisms.
2:43Exactly. And type one, an individual's own immune system basically goes rogue. It mistakenly attacks the insulin producing beta cells in the pancreas, and once those cells are gone, the body just can't regulate blood sugar.
2:54And the incidence of this childhood onset, T1D has been climbing by, what, about 3 to 4%, every single year globally? Yeah, 3 to 4%, every single year. Wow. I mean, if you compound a 3 to 4% increase year over year, decade after decade.
3:09You are looking at a terrifying public health crisis. massive. And it makes complete sense why the scientific community has historically thrown its weight behind environmental explanation. Sure. I mean, if you were a researcher in the 1990s watching these numbers climb, you'd naturally look at everything else that was changing in society at the time.
3:27Precisely. The timeline of the T1D spike perfectly aligns with the rise in childhood obesity. It aligns with shifts in maternal smoking habits, rising maternal agent delivery, and even changing norms around breastfeeding, and introducing cow's milk to infants.
3:45So the overarching logic was pretty solid. Like our environment changed dramatically over the late 20th century. So one of those lifestyle or dietary shifts must be the trigger for this immune system malfunction.
3:56That was the standard thinking, yes. But let me push back on that premise for a second, because there is an obvious paradox here for you to address. If human DNA doesn't radically change over a single generation.
4:07How could genetics possibly be responsible for a sudden modern spike in diabetes? And that paradox is exactly the engine driving this entire study? To solve it, the researchers had to measure something called heritability.
4:22And uh, this is where we have to discard the colloquial way we usually use that word. Oh, like when someone says, I got my mom's blue eyes. Exactly. In everyday conversation, that's what heritable means.
4:33But in population genetics, caretability is a strict mathematical ratio. Okay, so how do they define it mathematically? It's the proportion of a disease's variance across an entire population that is explained by genetic factors, as opposed to environmental factors?
4:48Okay, so imagine you have a giant pie chart labeled like all the reasons kids in this country get type one diabetes. Heritability isn't about the specific genes you have. It's simply measuring the size of the genetic slice of that pie versus the environmental slice.
5:02That is a highly effective way to visualize it, yeah? Now, apply that pie chart to our historical assumption. Okay. If the environment is the massive driving force behind this sudden doubling of T1D cases.
5:15Right. If modern diets or new pollutants are suddenly making 1000s of new kids sick. Exactly. Then the mathematical heritability should be actively shrinking over time. The environmental variance is expanding so rapidly that it should mathematically crowd out the genetic variants.
5:32Because the environmental piece of the pie should be growing larger and larger, making the genetic piece relatively smaller. Yes, exactly. Because if a brand new highly processed diet is suddenly the main reason kids are getting sick, well, genetics matter a whole lot less than they used to.
5:46Yeah. And surprisingly, until this Carolenski Institute team took on the challenge, no one had actually investigated whether the heritability of T1D had genuinely changed over these critical decades of rising incidences.
5:59I mean, I imagine the reason no one had done it is because proving that mathematically across a whole population sounds nearly impossible. Like, you can't exactly put 1000000s of kids in a controlled laboratory for 20 years to see who gets sick and why.
6:13No, you can't put them in a lab, but you can study them if you live in a country that keeps the most meticulous public health records on Earth. Oh, right. The Swedish registers. Exactly. The researchers leverage Swedish nationwide registers, and the sheer scale of this data is just breathtaking.
6:29They tracked nearly 3000000 children. Wait, 3 million? Yes. Specifically, 2,928,704 children, born in Sweden between 1982 and 2010. That is wild. That is essentially the entire fan base of ABA neatly organized into a single epidemiological spreadsheet.
6:49It practically is an entire generation of the country. Yeah. And they follow these individual's medical histories all the way up to age 19. Wow. But having 1000000s of people isn't enough to separate the genetic slice of the pie from the environmental slice.
7:01For that, they utilize the Swedish multi-generation register to employ a sibling design. Okay, a sibling design. Yeah, so out of those nearly 3000000 kids, they analyzed over 1500000 full sibling pairs.
7:15Oh, using siblings is such a clever filter. It's like looking at 2 C's from the exact same plant grown in slightly different soils to figure out exactly what makes them sprout. The sibling filter is incredibly powerful because, you know, full siblings share, on average, 50% of their segregating genes.
7:34But just as importantly, they share a massive amount of their early background environment. They have the same parents, they live in the same household, they experience the same social economic status, and they usually eat the same family meals.
7:46So by comparing siblings against the broader population, the researchers can mathematically filter out all of that shared background noise. Exactly. They can isolate what is genetically driven versus what is driven by unique environmental experiences.
8:00Okay, so you have this monumental data set of over a 1000000 sibling pairs. What kind of mathematical framework? Do you even feel that into to actually extract the answers? Well, they use an epidemiological tool called an AE liability threshold model.
8:15And um, without grounding in the statistical formula. Gee, let's not drown in formulas. Right. A simply stands for additive genetics. That is, the combined cumulative effect of all the individual risk genes a person carries.
8:28And E, since we're non-shared environment. non-shared environment. So we are explicitly ignoring the things siblings have in common and only looking at things that happen to one sibling, but not the other.
8:40Precisely. For example, maybe one sibling contracted a severe anerovirus infection in their 6th month of life, and the other sibling didn't. Oh. Or one was born prematurely or suffered a unique childhood trauma.
8:52The researchers knew from previous data that shared environmental factors, like having the same household income level or living the same zip code, actually contribute virtually 0 to the variants of T1D.
9:03Wow, virtually zero. That surprising. Yeah, it's the unique individual interactions with the environment that actually matter. So the AE model specifically splits the disease causes into the additive genetics and those unique non-shared environmental factors.
9:18Okay, that sorts out the broad categories of nature versus nurture. But I know they also wanted to test the specific villains we talked about earlier, like obesity and maternal smoking. How do you measure the exact impact of a single factor on a massive population trend?
9:33For that, they used causal mediation models. And this is where the methodology gets really, really elegant. How does a causal mediation model work in practice? Think of it like a traffic jam. If you know heavy rain is linked to you being late for work.
9:47A causal mediation model looks at the exact steps in between. Okay, tracking the dominoes. Right. Did the rain cause a specific accident on the bridge, which caused the lane closure, which caused the traffic, which actually made you late.
9:59It isolates the exact domino effect. Got it. In this study, the researchers didn't just want to know if childhood obesity was linked to T1D. They wanted the model to calculate precisely what percentage of the actual disease spite over time could be blamed entirely on the rising rates of obesity or the changing age of mothers.
10:18That makes total sense. Let's dig into the results then. because this is where the mystery really takes a turn. First, let's quantify the spike they observed in the Swedish registers. Yeah, it was significant.
10:29They found that from 1982 to the year 2000, the incidence of T1D almost doubled. It went from a cumulative incidence of .50% in the 1982 birth cohort to .93% in the 2000 birth cohort. And then the curve sort of plateaued between 2001 and 2010.
10:48Still, a near doubling of a severe autoimmune disease in under 2 decades is a massive public health earthquake. It really is. So think back to our pie chart. With an environmental explosion of that magnitude.
10:59The environmental slice of the pie should have surged, and the heritability slice should have shrunk drastically. Right, because the environment is doing all the extra damage now. But here is the major reveal of the paper.
11:08It didn't. Despite the massive jump in sick children, the heritability of T1D stayed incredibly stable and incredibly high. Wait, really? That blows my mind. The total number of kids developing the disease doubled, but the genetic weight of the disease remain entirely unchanged.
11:25It was essentially locked in place. The researchers calculated the overall heritability across the entire timeline at .83 or 83%. 83%. Yeah. When they broke it down year by year, it was .80 for children born in 1982, it was .83 for children born in 2000 and it was .83 for children born in 2010, it barely fluctuated.
11:46That is wild. And the non-shared environmental contribution stayed perfectly locked at around 17% for the entire 30-year observation period. Okay, let me see if I am tracking this. The mathematical importance of genetics stated 83% while the disease rate doubled.
12:00That feels like a total contradiction. How's that mathematically possible if the genetic code of the Swedish population didn't suddenly mutate in 1995. It feels completely counterintuitive, right? And the researchers anticipated that skepticism, so they ran a computer simulation to provide a mathematical proof.
12:16Oh, brilliant. What did they do? They essentially asked the model. Let's pretend the historical assumption is correct. If this massive increase of 40 new cases per 10,000 children was completely driven by changing environmental risk factors pushing healthy kids into illness.
12:31What would the heritability metric look like? Okay, and what did the simulation reveal? The simulation proved that if the modern environment was acting alone to cause the spike, the heritability should have plummeted from around 80% down to 59%.
12:44Oh, wow. It should have dropped off a cliff. Exactly. The environmental slice of the pie should have crowded out the genetics, but the actual observed data showed it remained rock solid at 83%. So the simulation is basically the mathematical smoking gun.
12:58Yes. It proves definitively that the observed rise in T1D cannot be solely explained by the environment independently attacking children. Okay, but if the environment isn't solely responsible, What about all those specific environmental dominoes we've been pointing at for decades?
13:15You know, the rising childhood obesity rates, the shifting demographics of motherhood. Well, this is what we might call the great environmental letdown of the study. Yeah. They track the known, measurable factors.
13:27For instance, maternal smoking during pregnancy, which, strangely enough, is inversely associated with T1D, meaning it somehow lowers risk that decreased over this time period. Okay, and the others. internal age at delivery increased.
13:42Childhood obesity increased significantly. But when they ran that causal mediation analysis, our traffic jam model, to see exactly how much of the doubling was caused by these massive societal shifts, the result was shockingly small.
13:55How small are we talking half a quarter? Combined, the changing prevalence of all these known measured factors explain less than 10% of the T1D incidents increase? Less than 10%. To be exact, they accounted for only 5 to 8% of the rise.
14:09That is staggering. We have witnessed a doubling wave of an autoimmune disease, and the primary environmental factors the medical community has been confidently blaming for 20 years account for, at best, 8% of the wave.
14:23It completely upends the traditional narrative. We spent decades looking at the wrong suspects or, at the very least, assigning them way too much blame. So we are left with an incredible puzzle then. If genetics didn't radically mutate, because 20 years isn't enough time for evolution, and the measured environmental factors like diet and demographics barely move the needle, what actually happened?
14:44How do we have twice as many sick kids with the exact same 83% genetic heritability? To solve this puzzle, the researchers point to a profound biological concept. The diabetogenic environment theory. We have to fundamentally reevaluate how we think about the load required to make someone sick.
15:00What do you mean by the load? So to trigger T1D, a child needs a combined load of genetic susceptibility and environmental pressure. What the liability threshold model suggests is that the total combined load required to trigger the disease has decreased over time.
15:14Ah I see. The modern environment has become more diabetogenic, meaning it is far more reactive. It is easier for the environment to pull the trigger on genetics that previously would have been harmless.
15:26Oh, so think of these genetic risks like a flammable forest. In the 1970s, the environment was like a weak, damp match. You might toss it into a forest of moderate risk genes, and it might spark, but it wouldn't catch fire.
15:39The child stays healthy. But by the year 2000, the modern environment became a blowtorch. The forest itself hasn't become more flammable. The DNA is exactly the same, but the fire starter has become so intense that it's burning down force that used to be safe.
15:55That metaphor captures the gene environment interaction perfectly. For the researchers give a specific genetic example in the paper, actually. We know there are certain human leucocyte antigen or HLA genotypes that carry a severe high risk for T1D, but there are many other HLA genotypes that only carry a moderate risk.
16:14Historically, a child with moderate risk genes would likely live their entire life without ever developing T1D because the older environment simply wasn't stressful or reactive enough to push them over the disease threshold.
16:26But now because the environment is acting like a blowtorch, those moderate risk kids are crossing the threshold into full-blown disease. Yes, exactly. The modern environment is unmasking existing genetic risks that previously lay completely dormant.
16:41The genetic slice of the pie is still 83% of the reason the variants exists, but the environmental push required to activate those dormant genes has gotten incredibly light. This raises a massive question for me.
16:53If the environmental factors we've been blaming, like obesity and maternal smoking only account for a fraction of the rise, are we completely missing the real environmental culprits? We absolutely are.
17:03And the researchers are highly transparent about this reality. The study acknowledges important limitations in their data. Like what? Well, for instance, while they had amazing nationwide registries, they lacked individual level clinical data for childhood obesity.
17:17They had to rely on nationwide summary statistics to estimate that 2 to 4% contribution. Got it. It's possible that with highly granular individual BMI data. That number could shift slightly, though likely not enough to explain the doubling.
17:33What about viral infections? I know I've read studies linking certain severe viruses to the onset of type one diabetes. You're likely thinking of endrovirus infections, which have been heavily researched as potential triggers for T1D.
17:45The limitation here is that the Swedish registers, as vast as they are, couldn't isolate specific viral streams like entrovirus. Ah, they just had broader data. Right. They only contain broad diagnostic codes for general infections.
17:59So a specific emerging viral strain that has become more prevalent in the last 30 years, could absolutely be part of this unmeasured, intense environmental pressure. If the big obvious societal targets like obesity and maternal smoking were a letdown, where should researchers be hunting for the real trigger?
18:15The data points us to a very narrow, highly specific window. The unshared environment, particularly in the very 1st year of life. Really? Just the 1st year. Yeah, because the shared sibling environment, the neighborhood, the parents' income, the general household diet, didn't matter much to the variants.
18:33So researchers need to hunt for microscopic events that happen to one infant, but not their sibling. So we are talking about highly individualized encounters with the modern world. Exactly. We need to investigate specific viral exposures that one child caught, but the other missed.
18:50We need to look at the exact timing of when dietary factors like cow's milk, gluten or certain proteins are introduced to a specific infant's immune system. Oh, that makes a lot of sense. Or we should be examining shifts in the infant gut microbiome, following early antibiotic use.
19:05These early microscopic unshared immune interactions are likely where the missing environmental blowtorches reside. It shits the entire paradigm of how we view autoimmune diseases. It moves the spotlight away from broad societal lifestyle choices and forces us to look at microscopic early life immune interactions.
19:22It's a huge shift in perspective. It really is a humbling reminder that in epidemiology, the most obvious answer is rarely the complete one. It truly is. It forces us to stop viewing nature and nurture as competing forces, and instead view them as a highly volatile chemical reaction.
19:40As we distill this massive deep dive into something you can take with you, how would you summarize the core insight of this monumental Swedish study? I would synthesize it like this. The incidence of childhood onset type one diabetes doubled in Sweden over 2 decades, yet its mathematical heritability remained exceptionally stable at 83%.
20:01This proves definitively that known environmental shifts alone cannot explain the rise. Instead, intense modern environmental triggers are likely unmasking existing moderate genetic risks that previously lay completely dormant.
20:16It makes you look at the unseen world around us entirely differently. What does this mean for how we define and hunt for environmental triggers in a world where our ancient genetics are constantly colliding with rapid modern changes?
20:26That's the $1000000 question. This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a five-star rating.
20:42If you'd like to support our work, use the donation link in the description. Now stay with us for an original track created especially for this episode and inspired by the article you've just heard about.
20:52Thanks for listening and join us next time as we explore more science base by base.
