Daylight Saving Time: Does springing forward cause heart attacks?

[Regina] (0:00 - 0:18)
No, no, no, it is lost to me. Kristin, I will argue this point until I die, because all that matters to me now is I lost that hour and I had to get up too early, and now it's dark in the morning. And can you tell?

I hate daylight saving time.

[Kristin] (0:23 - 0:46)
Welcome to Normal Curves. This is a podcast for anyone who wants to learn about scientific studies and the statistics behind them. It's like a journal club, except we pick topics that are fun, relevant, and sometimes a little spicy.

We evaluate the evidence, and we also give you the tools that you need to evaluate scientific studies on your own. I'm Kristin Sainani. I'm a professor at Stanford University.

[Regina] (0:46 - 0:52)
And I'm Regina Nuzzo. I'm a professor at Gallaudet University and part-time lecturer at Stanford.

[Kristin] (0:52 - 0:57)
We are not medical doctors. We are PhDs. So nothing in this podcast should be construed as medical advice.

[Regina] (0:57 - 1:06)
Also, this podcast is separate from our day jobs at Stanford and Gallaudet University. Kristin, today is the grouchy episode.

[Kristin] (1:06 - 1:07)
Okay. Why?

[Regina] (1:09 - 1:25)
Because everyone in the U.S. listening to this episode on the day it drops is grouchy. We just lost an hour of our lives thanks to social engineering. Thank you, Spring forward.

[Kristin] (1:25 - 1:41)
Oh, yes. Daylight saving time. But, Regina, technically, we're not really losing an hour, right?

Because we get that hour back in the fall. I guess maybe if you die before then. But no, I mean, you're not losing an hour of your life.

You're not living an hour less. So that's not quite right.

[Regina] (1:43 - 2:01)
No, no, no. It is lost to me, Kristin. I will argue this point until I die, because all that matters to me now is I lost that hour and I had to get up too early.

And now it's dark in the morning. And can you tell? I hate daylight saving time.

[Kristin] (2:02 - 2:14)
You know, Regina, I'm kind of agnostic because I'm so busy and never get enough sleep anyway. So I think until my kids are in college, it's not going to make a huge difference to me.

Okay, Regina, this just happened Sunday.

[Regina] (2:15 - 2:41)
Yep. In the U.S., daylight saving time starts the second Sunday in March. And our European listeners still have a couple more weeks to look forward to their stolen hour, because theirs starts the last Sunday in March.

And meanwhile, actually, our listeners in Australia, the Southern Hemisphere, are about to get their extra hour of sleep and are probably feeling pretty smug right about now.

[Kristin] (2:42 - 2:55)
Right. It's reversed for them. But, Regina, I just learned so much.

I've never actually paid attention to the differences around the world in daylight saving time. I do know that daylight saving time is really controversial in the U.S. People hate it.

[Regina] (2:55 - 3:56)
Oh, they do. I think this is one thing that many, many people agree on. Every year, there are calls to abolish it in the U.S. and in Europe. And believe it or not, one of the reasons that people cite for abolishing it, besides just being grouchy, which I think is a valid reason, is the health risks. And I want, Kristin, to quote to you from a BBC article from 2025, just last year, with the headline, How Daylight Saving Time Affects Your Health. And here is the quote.

Increasingly, research suggests that clock changes may adversely affect people's health. The strongest evidence relates to pushing the clocks forward in spring. In 2014, researchers in the U.S. found that the number of people admitted to hospital with heart attacks soared 24 percent the Monday after clocks were put forward.

[Kristin] (3:56 - 4:03)
Wow, 24 percent. That sounds big. Of course, Regina, you know I'm going to want to understand where that number came from.

[Regina] (4:04 - 4:48)
Of course you do, Kristin. This is one of the many, many reasons I love you. But we are going to look at where numbers like these come from because statistics like this really do get cited a lot to argue that we need to scrap the whole thing, just get rid of daylight saving time altogether.

And there are other health risks that have been linked to this whole thing, right? Like car crashes, workplace injuries, mood changes, but heart attacks are really the most studied outcome. So today we're going to focus on that.

And the claim is actually pretty simple. It's this, that spring daylight saving time transition, springing forward one hour, increases heart attacks.

[Kristin] (4:48 - 4:49)
Okay, got it.

[Regina] (4:49 - 5:02)
Mm-hmm. And statistically, what are we going to talk about? Incidence ratios, relative risk versus absolute risk, and harvesting, which, spoiler alert, it's not actually about farming.

[Kristin] (5:03 - 5:16)
Sounds great, Regina. All right, so give me some background. How did daylight saving time get started?

It was something about the kids getting up earlier to farm? No, that's summer vacation. Does it have to do with farming?

Yes?

[Regina] (5:17 - 5:23)
It does not really. It's kind of fascinating. It all started with a bug collector in 1895.

[Kristin] (5:25 - 5:27)
Seriously? A bug collector? Okay, I don't know this story.

[Regina] (5:28 - 6:30)
No, right? George Hudson was his name, a guy who lived in New Zealand, and George Hudson loved bugs. He loved collecting them, studying them, killing them, publishing on them.

But his day job was at the post office. And he realized that, okay, summer days are long, but there wasn't enough evening daylight left over after his post office shift ended for him to go outside and look at the pretty bugs. And he was very sad about this.

And he came up with an idea. He went to a meeting, in fact, of the Royal Society of New Zealand, it's a big deal, 1895. And he proposed that everyone just move their clocks two hours ahead between the spring equinox and the autumnal equinox.

So everyone could have more daylight and go out and look at bugs, I guess. And he was basically laughed out of the meeting. Poor George!

[Kristin] (6:31 - 6:43)
I actually can see the logic, right? It is kind of nice to have more daylight hours in the summer, like you go out in the warm summer evening, throw the ball with nibbles.

[Regina] (6:44 - 7:22)
Yeah, but the thing is, the days get longer no matter what.

But I'm going to try to withhold my ranting as much as possible throughout this episode, I promise you. So we're going to mostly stick to the facts here. Okay, after we've got poor George Hudson laughed out of the Royal Society meeting, then there was a new guy, William Willett, over here in England.

He was a rich homebuilder, and it was about 10 years later, he had the same idea that I think he came up with independently, except his rationale, he wanted more time in the evening to play golf. So I'm thinking, Kristin, this is just about boys and their toys.

[Kristin] (7:23 - 7:26)
Oh, my goodness. Yeah, golf, I don't care at all about, sorry.

[Regina] (7:28 - 8:02)
Well, William Willett, he went all out on this, and he spent a lot of his own money promoting this idea to important people. His suggestion was a little different, though. He said we should shift 80 minutes in the spring and in the fall, but we should do it gradually.

He said, in April, let's move the clock forward by only 20 minutes each weekend. So you end up with 80 minutes over the entire month, and then in November, let's reverse it by falling back 20 minutes each week.

[Kristin] (8:02 - 8:27)
Oh, wow. Well, that's very interesting, because that might solve the potential problem of health risks, because 20 minutes is very incremental compared to an hour. But also, that's going to be very confusing, I think, right?

We can barely keep up with one time shift, let alone four or eight per year, and it's already hard enough to sync up time zones with, you know, people in Europe. Now we have to shift by an hour 20?

[Regina] (8:28 - 9:02)
It's all arbitrary, though. It's all arbitrary, because it's not like the sun is following our clock, right? Our clock is the rigid thing.

It's the sun. I will die on this hill. Yes, okay.

So, William Willett actually got people's attention. He was not laughed out of the meeting, like poor George Hudson. And in fact, in England, Winston Churchill got interested in this, yeah.

But do you know what really got people's attention and managed to get the whole thing passed? Can you guess?

[Kristin] (9:02 - 9:09)
Oh, I guess with Churchill, war. We want to be more efficient, save energy during war or something like that, right?

[Regina] (9:10 - 10:14)
There you go. It was actually World War I, and it was that whole need to save energy, and it actually started with Germany. Germany was the first to adopt a one-hour shift in 1916, and that was because they were running out of coal thanks to the British blockade.

And so they wanted to make sure that people were not burning coal for light in the evening. But then it was just three weeks later, so William Willett had been, like, arguing about this for years, but then it took only three weeks after Germany did it for the UK to suddenly adopt it themselves, because now they wanted their military workers to spend all their time in the factories and the shipyards during daylight hours. And then shortly after that, France and most of Europe jumped on board.

Everyone was at war. And then I love this. A couple years later, 1918, the U.S. finally looked around, and they're like, oh, wait, us too. Everyone else is doing this clever wartime efficiency measure thing. We can't fall behind. We want to be one of the cool kids, too.

So they went ahead and passed it.

[Kristin] (10:14 - 10:17)
Oh, I didn't know that's when it started in the U.S. That makes sense, though.

[Regina] (10:18 - 10:33)
But in the U.S. since then, it has been quite a weird love affair, love-hate affair with this daylight-saving time thing. Yeah. Okay.

So adopted in 1918, repealed in 1919 because people hated it.

[Kristin] (10:33 - 10:34)
Oh, it's like prohibition.

[Regina] (10:34 - 11:02)
Actually, it was. Congress tried repealing it because the farmers hated it, because the cows hated it, because cows want to be milked when they want to be milked based on the sunrise. They don't care about this human clock thing.

So Congress tried repealing it, but then President Wilson at the time, he vetoed the repeal twice. Why? Because he loved golfing.

[Kristin] (11:02 - 11:11)
Oh, this is a great story. So it actually got passed that we were going to repeal it, and then the presidential veto, which is hardly ever used.

[Regina] (11:12 - 11:19)
I know. But then Congress overrode both of his vetoes and said, no, that's it. You and your golfing. Just sit down.

[Kristin] (11:19 - 11:23)
So we got rid of it.

So how is it still here?

[Regina] (11:23 - 11:37)
Uh-huh. Okay. Yep.

It's cute. It's a fun story. Adopted again in 1942 during World War II.

This time they called it War Time. I think they were hoping for some better marketing here.

[Kristin] (11:38 - 11:38)
Yes, I like it.

[Regina] (11:38 - 11:39)
Wartime.

[Kristin] (11:39 - 11:40)
It's better than daylight saving.

[Regina] (11:41 - 11:43)
Yes, because you're not saving anything.

[Kristin] (11:43 - 11:45)
Right. You're not saving daylight. I've never understood that term anyway.

[Regina] (11:46 - 12:28)
Marketing. This is what I'm saying. It's a mind op thing.

This is just mass control. Okay. So they repealed it in the U.S. in 1945 at the end of the war, but then there was complete chaos, Kristin, in the 50s and 60s because they let each city decide for itself whether they wanted to do this daylight saving time, whether they wanted to shift clocks to summertime or not. I read somewhere that in the summer, you could take a 35-mile bus ride from a city in West Regina to a city in Ohio, and you would have to change your watch seven times because every city was just doing its own thing.

[Kristin] (12:28 - 12:40)
You could never do that in the Internet age because it would be just total chaos. But I guess back then you just have a watch on your wrist, you know, no biggie.

[Regina] (12:40 - 12:41)
No biggie, but then how do you know the buses? Like is it?

[Kristin] (12:41 - 12:45)
Oh, it might be a biggie for the bus schedule, yes. You're going to miss your bus.

[Regina] (12:45 - 12:50)
Right. Or the broadcasting, the six o'clock news on the radio or the TV. What are you doing?

[Kristin] (12:50 - 12:51)
That's true.

[Regina] (12:51 - 12:51)
Yeah.

[Kristin] (12:51 - 12:51)
Yeah.

[Regina] (12:51 - 13:31)
Okay, so they realized, okay, this American independence thing, letting everyone go their own way, maybe not so efficient after all. 1966, they finally standardized it to the last Sunday in April, which is pretty late. Then in 1987, several years later, they moved it three weeks earlier to the first Sunday in April.

Why did they do this? I read because it was heavily lobbied by industry, golf course industry, barbecue, grill manufacturing industry, theme parks, anyone who wants us to spend more money outside in the summer evenings.

[Kristin] (13:31 - 13:39)
That makes a lot of sense, actually. An extra hour for people to be at Disneyland, yeah, or the golf course. Wow.

That is fascinating, Regina.

[Regina] (13:40 - 14:00)
Isn't it? Okay, 2007, during our recent history, I don't know if you noticed, they moved it three weeks earlier again to the second Sunday in March, which is very, very early. And at the same time, they extended it one week later to the first Sunday in November.

Now, guess why that one got extended later?

[Kristin] (14:00 - 14:05)
Okay, I totally don't remember this, but I'm guessing the one in November for Halloween?

[Regina] (14:06 - 14:10)
It was the candy manufacturing lobby.

[Kristin] (14:11 - 14:18)
You're kidding. Okay, so I have trick-or-treated with my kids many, many years. I don't remember it ever being light outside when we were trick-or-treating. It's always after dark.

[Regina] (14:18 - 14:21)
Well, since 2007, it has been lighter.

[Kristin] (14:22 - 14:40)
Lighter, really? I think we just went trick-or-treating an hour later then because trick-or-treating is during the dark, always. I remember many long walks around the neighborhood in the dark with mom being really bored and like, when can we go home?

And my kids wanting to keep going because, you know, candy.

[Regina] (14:41 - 14:54)
Maybe it makes it more of a difference for younger kids. Yep. Okay, so that is where we are.

First Sunday in November is when it ends. Second Sunday in March. So it's quite long, much longer than what they did during wartime.

[Kristin] (14:55 - 15:03)
Right, and completely arbitrary other than like industry-driven, it sounds like. I had no idea there was so much history to this.

[Regina] (15:03 - 15:14)
I checked on this. I wanted to see whether it corresponded to some inflection point on, you know, when the days start accelerating and getting longer. Nothing, nothing around that.

So it has nothing to do. It's all political.

[Kristin] (15:15 - 15:25)
Wow, that is super fascinating and really kind of stupid, actually. How much time and energy have people wasted on this thing that is pretty useless?

[Regina] (15:25 - 15:27)
And those lobbyists have a lot of power.

[Kristin] (15:27 - 15:37)
I'm sure, yeah. Well, because I keep hearing like, yeah, that they're going to repeal it in Congress. Like you hear that periodically, oh, they've got a bill up to repeal this.

I guess it never happens because of the lobbyists.

[Regina] (15:38 - 15:42)
I think so. Every year. Every year.

I think the bill is up and it just doesn't get passed.

[Kristin] (15:42 - 15:42)
I know.

[Regina] (15:42 - 15:44)
Maybe this is why it hasn't passed.

[Kristin] (15:45 - 15:52)
Yeah, like I said, I don't pay too close of attention to it. I have no strong feelings one way or the other. But now I will.

I will pay attention just out of pure curiosity.

[Regina] (15:52 - 15:54)
Join me in my anger and indignation.

[Kristin] (15:55 - 16:05)
Okay, so that was some fascinating history. I feel like we could just do an episode of the history of this, Regina, but there was some science we were going to talk about. How do we get to the heart attacks?

[Regina] (16:06 - 16:59)
Right, right. That was the important thing because, you know, people didn't like it. The cows didn't like it or whatever.

But that's just, you know, people being grouchy. It's not until we started talking about the health effects that we actually had some ammunition for why we should get rid of this. The proposed mechanism behind these health effects is what?

It's basically that you're losing an hour of sleep and also you may be disrupting your circadian rhythm when you spring forward, kind of like an hour of jet lag. And this is stressful to the heart. That's the idea.

So the first big splash for this was 2008, right after, in fact, the U.S. changed their daily saving time transition. But this one was in Sweden, two Swedish researchers, and it was a publication in the New England Journal of Medicine.

[Kristin] (17:00 - 17:08)
Oh, wow. New England Journal of Medicine, that is the holy grail of medical journals. So this must be a serious study.

[Regina] (17:08 - 17:31)
It is. And actually, the origin story is kind of funny because the first author is an MD, PhD, cardiac epidemiologist. But he started looking into this because he was grouchy after a spring time change.

He said in an interview with the Associated Press, I was on the bus, quite sleepy, and I thought of this.

[Kristin] (17:32 - 17:43)
I have to say, Regina, yeah, this is how research questions come to life sometimes, right? Somebody in the shower go, aha, I had something that affects me personally. Yes.

[Regina] (17:44 - 17:49)
So this was not, Kristin, a full research article. It was a letter.

[Kristin] (17:49 - 18:08)
Oh, OK. We should explain the difference, Regina. So a letter in academia is a really short version of a research paper, so you don't have a lot of room for details.

But still, I'm not going to discount this research just because it's a letter. You know, getting into the New England Journal, whether letter or big article, is still a big deal.

[Regina] (18:08 - 18:32)
Oh, absolutely. We'll see in a moment why that makes a bit of a difference. So a little bit more about the background, though.

So this was a registry study, because Sweden has these great national health registries, and the research team looked at all recorded heart attacks, myocardial infarctions, from 1987 to 2006.

[Kristin] (18:32 - 18:40)
Wow. This is why these Scandinavian health registry data are so great, because they're so comprehensive. You can look at everybody with a heart attack.

[Regina] (18:41 - 19:07)
Right, right, exactly. The design was pretty simple here. So for each daylight-saving time transition, so the spring and the autumn, what they did is count the number of heart attacks on each day in the week right after the time change.

But then, of course, how do you know if that's a lot, right, because that's a lot of heart attacks. You need a baseline. You have to compare it to something.

[Kristin] (19:07 - 19:08)
You need to compare it to something, yeah.

[Regina] (19:08 - 19:12)
Right, right. What did we say my tattoo was going to be compared to what?

[Kristin] (19:12 - 19:13)
Compared to what, yes.

[Regina] (19:14 - 19:26)
Compared to what. So the baseline, the comparison, they decided was going to be the average number of heart attacks on that same weekday, two weeks before and two weeks after the time change.

[Kristin] (19:27 - 19:40)
Right, because presumably the two Mondays before and the two Mondays after the transition aren't affected by daylight-saving times, but that gives us something to anchor to, like what is just the background rate of heart attacks on, you know, a Monday.

[Regina] (19:40 - 19:45)
Exactly. Perfect. So just to illustrate this, how about if I just make up some numbers to give an example?

[Kristin] (19:45 - 19:46)
Yeah, make up some easy numbers, yes.

[Regina] (19:47 - 20:27)
Okay. Let's say we're going to look at the Monday right after the time change. So like today, the day the episode is dropping.

Okay, let's say there's 100 heart attacks today after the spring shift on this Monday, and let's say two Mondays earlier there were 70 heart attacks. And let's fast forward in time two Mondays from today and there are 90 heart attacks. So our baseline is going to be the average of that 70 and 90, which is going to be 80.

And then what we do is we just compare that 100 from this Monday after the time shift to the 80, that baseline.

[Kristin] (20:27 - 20:40)
Okay, so Regina, you're talking about the number of heart attacks per day, per a specific day. So that actually involves time, and technically this is a rate. We call it an incidence rate.

So what they're doing here is they are comparing incidence rates.

[Regina] (20:40 - 21:21)
Exactly. Exactly. Incidence rate is the formal term that we use in epidemiology.

I might just say, you know, number of heart attacks, but it is that. And then we want to compare these two incidence rates, these two counts. And to do that, the researchers just divided those two numbers, and that is what is called the incidence ratio.

So in my hypothetical example, then, it would be 100 heart attacks on the Monday after the time shift divided by the baseline of 80 heart attacks two weeks before and two weeks after. So 100 divided by 80, that ratio, that incidence ratio is 1.25. That's the number we're interested in.

[Kristin] (21:22 - 21:48)
And let's talk about how to interpret incidence ratios. For an incidence ratio, a value of one would mean that there was no increase or decrease in heart attacks, right? If you divide two things and they're equal, you will get a value of exactly one.

Value greater than one means an increase in heart attacks, like we have here, or a value less than one would actually mean a reduction in the rate of heart attacks. So 1.25 would indicate a 25% relative increase in the rate of heart attacks.

[Regina] (21:48 - 22:05)
Exactly. So the researchers did this not just for Monday, but for each of the seven days individually. And they did it separately for the spring time shift and the autumn time shift.

And they also calculated an aggregated effect over the entire week.

[Kristin] (22:06 - 22:13)
Oh, okay. So, all right. Now, that was the fake numbers that we made up.

So now I want to know what were the real numbers that they found in the New England Journal of Medicine paper?

[Regina] (22:13 - 22:19)
I think you're going to find this very interesting, but I suggest a short break first.

[Kristin] (22:41 - 22:56)
Welcome back to Normal Curves. Today we are looking at the claim that daylight saving is bad for your health, specifically that it increases the risk of heart attacks. And Regina, you were about to tell us what this New England Journal of Medicine letter on this topic found.

[Regina] (22:56 - 23:18)
Actually, Kristin, we did not get a lot of specific numbers, as I am going to talk about. Now, they did say in the text that the aggregate of the entire week after the spring shift was significantly more heart attacks over baseline. And they said that that incident ratio was 1.05. Okay.

[Kristin] (23:18 - 23:30)
So, there was an increase. If you averaged across the entire week, this is not looking at individual days, 1.05, that's a 5% increase in the rate of heart attacks over the entire week.

[Regina] (23:30 - 23:58)
And looking at that fall back, the time shift in the fall when you aggregate over the entire week, the ratio was 0.985, so slightly fewer heart attacks, but not statistically significant. But as I said, Kristin, the rest of the numbers were not in the text, and there was no table. And I wanted to know the numbers for the individual days, so I had to do some sleuthing.

[Kristin] (23:58 - 24:07)
Oh, that's very annoying that they didn't just give you those. But Regina, this was just a letter, and sometimes you don't have a lot of space in the letter, so maybe I can forgive them this one.

[Regina] (24:08 - 24:24)
That is very generous of you, Kristin, because honestly, I was kind of cursing them. And here's why. Instead of giving numbers in the text or in a table, guess what they did instead.

I am going to rant now.

[Kristin] (24:24 - 24:29)
Oh, I can guess this, Regina. Did they display the numbers in bar graphs?

[Regina] (24:30 - 24:49)
Kristin, you know me so well. That is exactly what they did. They used all of their precious space in a letter to show bar charts with stupid little error bars.

No table of numbers, no incidence ratio numbers, just bars.

[Kristin] (24:50 - 25:01)
Okay, I am taking back my forgiveness then, Regina, because if they had enough space for a useless bar graph, then they certainly had enough space for adding a few numbers in the text. Right?

[Regina] (25:01 - 25:06)
I mean, we're just like a simple, beautiful little compact table of numbers.

[Kristin] (25:06 - 25:26)
We like tables. Oh, yeah. I mean, I see so many papers that I review where there's like 18 different bar graphs, and I'm like, what the heck, just put all of those numbers in a table.

That would be so much easier for me, the reader. I don't need your stupid bar graphs. They're not easy to read.

And then I got to weed through, you know, where's figure 17 with the numbers I'm looking for? Put them all in one table, please.

[Regina] (25:26 - 25:44)
Put them in a table. Okay, Kristin, clearly you and I have talked about this before in the podcast. We have ranted because in addition to bar charts needlessly just taking up a lot of space, they are supposed to be used for categorical data, which we do not have here.

We have incidence ratios.

[Kristin] (25:45 - 26:05)
Yes, we talked about this back in the pheromones episode. Bar charts are supposed to be for categorical data. Incidence ratios are a number.

So this is not the right way to display the data. And Regina, I'm going to cite one of our own normal curves, methodologic morals. Bar charts are not for numerical data.

Repeat after me. And that was from our pheromones episode.

[Regina] (26:06 - 26:12)
Are you sure I didn't include some more like expletives and swears when I gave that methodologic moral?

[Kristin] (26:12 - 26:24)
I edited them out. Don't worry. So, Regina, did you have to go dumpster bar chart diving and extract the numbers from the chart in order to get those numbers for us to talk about today?

[Regina] (26:25 - 26:29)
I love that. Dumpster bar chart diving.

[Kristin] (26:29 - 26:31)
That's what it reminds me of somehow.

[Regina] (26:34 - 27:00)
I know that's what it feels like, like you just have to dive into the mess. And that is exactly what I did. Dumpster bar chart diving, because I was really annoyed and I wanted the actual numbers.

I was ready to dive in. Oh, but Kristin, that was when I remembered our great new boyfriend, whose name is graph2table App.

[Kristin] (27:01 - 28:12)
Yes. This is the bar that now all potential boyfriends have to aspire to, because we love Graph2table. And we talked about this tool back in our marathon fueling episode.

This is an AI driven tool that extracts numbers from the figures like bar charts. And the great thing is you can upload like 18 different bar charts from your paper and it will automatically extract every single mean and error bar and it will extract the labels from the axes and it creates these downloadable data sets with the data ready to go. And so I have been using WebPlotDigitizer to do this for years, which is a great program except it is so much work and Graph2table is saving me so much time.

When I discovered this tool, I actually reached out to the creator of Graph2table and they are now an affiliate partner of Normal Curves, our very first affiliate partner, which I think is just a very fitting partner. So our listeners get a discount. So you can find Graph2table if you just Google it and you can use our discount code, which is for 20% off, normalcurves20.

That's all lowercase.

[Regina] (28:12 - 28:21)
I have known you for 30 years and

[Kristin]
I'm about as excited as I get about anything nerdy, yes.

[Regina]
Pure love. This is it.

This is pure love.

[Kristin] (28:22 - 28:35)
But you don't know how much time I have spent on WebPlotDigitizer clicking on axes and dots. You have no idea how much time I have spent on that. And I love things that improve efficiency.

So, so I assume you ran the bar graphs from this paper?

[Regina] (28:35 - 29:46)
I did. And I am right there with you now because I was annoyed with the researchers, right? I fed it this annoying bar chart.

It thought for just a minute and then it spit out all the things I wanted, the incidence ratio numbers, confidence interval endpoints, and it was a beautiful CSV spreadsheet. OK, so Monday and Wednesday, right after the spring time shift, the incidence ratios were about 1.06. So that means 6% more heart attacks. And that was statistically significant, statistically discernible.

Now, Tuesday was even higher. That was about 1.1, so 10% more heart attacks. And the other four days of the week, they had slightly more heart attacks, but they weren't statistically significant.

And like I said, when you look over the whole week, that was 5% more heart attacks. That was significant. And in the autumn, I was able to dive into the numbers a little bit more.

So Monday's incidence ratio was 0.95. So about a 5% drop in heart attacks, statistically significant, but again, the entire week, no real difference.

[Kristin] (29:46 - 30:31)
OK, so interesting, Regina, the pattern is lining up with losing an hour being bad for you and gaining an hour maybe at the beginning being good for you. But those effects are pretty darn tiny, Regina. Those Scandinavian data sets are so huge, so many heart attacks, that even very tiny effects are going to come out to be statistically significant.

Normally, we'd look at relative risks like this and say the effect is basically negligible. But daylight saving time affects so many people and heart attacks are not rare. So even a small increase in risk could translate into real additional cases.

So, Regina, the question we need to ask is, are these effects big enough to care about clinically?

[Regina] (30:31 - 30:54)
I am so glad you asked that question, because, Kristin, here is something they did do well. On top of those annoying little bars in the bar chart, they gave us the actual counts. So they gave the count of the number of heart attacks on that Monday after the time shift and then the average of the heart attacks on those two baseline weeks.

[Kristin] (30:54 - 31:02)
So technically, you could have hand calculated the incidence ratios and confidence intervals, although that would have taken you some larger amount of time than graph2table.

[Regina] (31:03 - 31:10)
It is true. I could have done that, but I am lazy and my boyfriend was right there, graph2table, willing to help out. So I let him help out.

[Kristin] (31:11 - 31:12)
As all boyfriends should be.

[Regina] (31:13 - 31:38)
As they should be, just waiting by my side, ready to do some math for me. But more importantly, Kristin, those raw counts let us look at the absolute differences, not just the relative differences. And this is really important, as you and I talk about when you are looking at a paper with a relative effect, because relative risk is not the same as absolute risk if they give you different stories.

[Kristin] (31:38 - 31:56)
This is a really important point and why we talk a lot in this podcast about relative versus absolute risk. So imagine I'm a policymaker in Sweden. It doesn't help me to just hear a 5% increase.

I don't know what that means. I need to know how does that translate into heart attacks? How many extra heart attacks are we talking about?

[Regina] (31:56 - 32:52)
Exactly. So I went through this this whole calculation and I'm going to walk through it a little now just so other people can see how to walk through it in case they want to do it themselves. OK, so let's just take Tuesday in that spring transition, because that was the biggest jump, actually.

That was the 10% increase in heart attacks. And the actual counts were this, that Tuesday after the time change, there were sixteen hundred and forty four heart attacks versus in the comparison weeks, fourteen hundred and ninety four. And I will subtract that out for you so you don't have to do it in your head.

That's one hundred and fifty extra heart attacks right after the time change. But then you have to take into account that this was over the entire study period. That was 15 years included in here.

So one hundred and fifty divided by 15. That's about 10 extra heart attacks for each spring time change.

[Kristin] (32:52 - 33:02)
And that's over all of Sweden. So that is not a lot. But you said that was just for Tuesday, Regina.

And there was also increases in Monday and Wednesday. So what happens when you add in Monday and Wednesday's extras?

[Regina] (33:02 - 33:46)
Right. So Monday, Tuesday, Wednesday, all together, that works out to about 20 extra heart attacks per spring forward change. But if you're like me and you have no idea how many people are in Sweden, I really didn't.

I did a back of the envelope calculation and I figured out that there are about four point eight million adults, 35 or older in Sweden, on average during that study time period. So 20 extra heart attacks divided by about five million people in Sweden is about four extra heart attacks per million adults per spring transition.

[Kristin] (33:47 - 33:59)
So that's not nothing, but also it's not exactly a public health catastrophe. That is a pretty small amount. And Regina, were these just any heart attacks or were these heart attacks that resulted in death?

[Regina] (34:00 - 34:09)
Oh, good question. I am glad that you asked that these numbers were for all heart attacks that were listed as a primary diagnosis, fatal or not.

[Kristin] (34:09 - 34:27)
OK, so this makes it even less of a public health catastrophe. Right. But, you know, Regina, I just realized something.

So you said that there were fewer heart attacks in the fall, at least on Mondays. So maybe you have a few extra heart attacks in the spring, but then you avoid a few in the fall. So maybe it just cancels out.

[Regina] (34:27 - 34:58)
Now, Kristin, you are actually bringing up a very profound and interesting question here. And if you don't mind, I'm just going to broaden it out a bit further because I love it. It's a fair question.

You're saying, hey, we saw these extra heart attacks in the few days after the March time change. But the question is, would these people have had a heart attack anyway? And this concept, it actually has a name.

It's called harvesting. Harvesting.

[Kristin] (34:59 - 35:06)
Harvesting sounds kind of morbid, right? Like sounds like they're harvesting souls like the devil. Is that what we're talking about?

[Regina] (35:08 - 36:15)
That is actually kind of the idea. The harvesting effect. Sometimes they call it mortality displacement, but I kind of like harvesting effect better.

It's much more evocative. And it's basically when something causes deaths or, you know, in this case, health problems, heart attacks, causes them to happen slightly earlier than they otherwise would have. But it doesn't actually increase the number of events overall.

And I don't know, think of it like harvesting like a crop of apples on a tree. Right. Some of them are ripe and a strong wind just comes through and knocks down the ripe fruit today instead of tomorrow.

So it's not like the wind is creating new fruit. You don't get new apples. The wind just accelerated when those ripe apples fell and are harvested by the devil.

So the total apple harvest is the same. It's just redistributed in time. That's the idea.

[Kristin] (36:15 - 36:51)
Right. So you're saying like maybe somebody gets a heart attack on the Monday after daylight saving, but they would have gotten that heart attack two weeks later anyway. And so we have just slightly blown the heart attack off the tree a little bit early.

But it's not a new we haven't created a new heart attack. Yep, exactly. But Regina, I mean, mortality displacement.

I mean, technically, we're all getting harvested eventually. Like so I guess we're talking days or months of the timescale is important here. Harvesting is referring to a small timescale, not to like all of us over all time because we're all getting harvested.

[Regina] (36:52 - 36:56)
Well, I am. That's a super cheerful thought.

[Kristin] (36:58 - 37:07)
But it's true, right? I mean, like mortality, displacement, somebody who dies of a heart attack tomorrow that could have lived another 10 years like that's harvesting, too, in some sense.

[Regina] (37:08 - 37:13)
Then everything, everything is harvesting. Wow. We just got like deeply existential right now. And we went from boyfriends now to it's just about death.

[Kristin] (37:14 - 37:31)
Well my point being that when we're talking about harvesting in a methodological sense, we're talking about in a very short time frame, a month or a day, half a year, not in our lifetimes.

[Regina] (37:31 - 37:37)
Oh, absolutely. No, this is an excellent point. I'm just, you know, slightly a little bit more depressed than I was 30 seconds ago.

[Kristin] (37:38 - 37:42)
That's OK. Graph2table. Just think of graph2table.

It'll cheer you up.

[Regina] (37:44 - 38:15)
Oh, my boyfriend. OK, we're going to come back to harvesting later. I hope we come back to it because it is actually it is very interesting.

But back to right now, this New England Journal of Medicine letter, it got enormous attention when it was published in 2008. It has been cited in medical society position letters that argue against daylight saving time. It shows up in state health board reviews of the whole thing.

People are still citing it prominently. It's influential.

[Kristin] (38:16 - 38:23)
Well, you know, anything that affects a lot of people, which this affects everyone, it does make sense that this would get a lot of attention.

[Regina] (38:24 - 38:26)
It affects more people than even sex, maybe.

[Kristin] (38:26 - 38:58)
Yes, everyone has to get up in the morning, but not everyone is having sex. Yeah, you're right. And actually, that just gave me a great idea for a paper, Regina, because daylight savings, it probably affects the amount of sex people are having.

Like maybe you have less sex when you lose an hour of sleep and you have more sex in the fall when you gain an hour of sleep. That could be an outcome. Imagine the headlines, you could combine sex and daylight saving time.

We could have that'll be a New England Journal of Medicine paper that gets so much attention.

[Regina] (38:58 - 39:05)
We should do it. We can absolutely do this.

[Kristin]
I'm surprised nobody has done that one already.

[Regina]
How do you spend that extra hour? I mean, come on, really.

How are you going to spend it come November?

[Kristin] (39:06 - 39:10)
Yeah, I don't think anybody studied this, but somebody should. Dissertation people out there.

[Regina] (39:10 - 39:16)
And I would also like to say, thank goodness we finally got sex into the episode because I was worried it was not going to happen.

[Kristin] (39:16 - 39:28)
We did it. Checkbox. OK, back to daylight saving time.

OK, so this was 2008. Has anybody done any follow up studies on this yet?

[Regina] (39:29 - 39:36)
Absolutely. Oh, yes. Lots of people.

And I think after 2008, a lot of daylight saving time haters started studying this.

[Kristin] (39:36 - 39:51)
Oh, so we may have some biases here. Not conflicts, not monetary conflicts of interest, but I want my hour of sleep back conflicts of interest. And that's a pretty powerful conflict of interest.

Like, I don't think that you could study this objectively, Regina. Maybe your bias would creep in, you know.

[Regina] (39:52 - 40:10)
I am doing my best to be unbiased here. I want to put that out. And I put my human biases right up front so you know that.

And to be fair, I mean, maybe there was some weirdo daylight saving time lovers out there who wanted to disprove the 2008 results.

[Kristin] (40:10 - 40:15)
The like Disneyland and golf lobbyists. So maybe if you're getting paid from them, you have the opposite bias.

[Regina] (40:15 - 40:47)
Could be. I wonder if that would be enough to counter my anti daylight saving time bias as if I were funded by Disneyland. OK, so after 2008, a lot of studies out there, different countries, different datasets, different designs.

I am going to skip ahead to 2024 because that is when a German research group published a systematic review and meta analysis. And Kristin, gold star, they pre-registered the review on Open Science Framework.

[Kristin] (40:47 - 40:53)
Yay, pre-registration. I feel like we need normal curves gold stars to hand out, right, along with smooches.

[Regina] (40:53 - 41:09)
I love that idea. Oh, let's work on that. So this Goldstar team, they identified 12 studies that had unique data that could be included in their meta analysis, but only 11 strangely made it to the final analysis.

[Kristin] (41:09 - 41:11)
Oh, interesting. Do tell.

[Regina] (41:11 - 41:43)
Yeah, that's because the authors of the meta analysis, you know, they have to go through, they need to look at the numbers in each of the papers. That's how they put it together, synthesis for the meta analysis. And it's basically a little like the statistical sleuthing that we do on this podcast.

It can be. And they found major calculation errors in one of the papers. And it appears that they even tried to work with the author of that paper to get the correct numbers and could not do that.

So they ended up just having to ditch that paper, exclude it entirely.

[Kristin] (41:44 - 42:19)
Wow, Regina, this is interesting because it's telling me that we have a lot of episodes in our future, because if you look at the denominator here, there's 12 studies and one in 12 had blatantly wrong calculations. And if the literature is anything like that, which you and I probably believe it may be, there is literally a universe of papers for us to sleuth out in this podcast. Potentially infinite, really.

Well, yeah, infinite relative to the time we have to go through papers before we're harvested. I don't know why I'm so morbid today.

[Regina] (42:20 - 42:25)
Wow. Yeah. Bring it right back.

I'm going to keep bringing in the sex. You're going to keep bringing in the death.

[Kristin] (42:25 - 42:27)
I know this is terrible. I don't know.

[Regina] (42:28 - 42:47)
Okay. So while we're still alive talking about these studies, we had 11 studies in 10 different countries, six in Europe and four outside of Europe. And now I bring that up.

It's important because, remember, the countries don't all switch to daylight saving time at the same time.

[Kristin] (42:48 - 42:54)
So you could be capturing slightly different effects if it matters when you're doing this switch. Right. Right.

[Regina] (42:54 - 43:36)
So, and the studies were done over different time periods, different sources of data, different statistical methods. So it's not like it's one clean experiment that's replicated over and over. Right.

Mm-hmm. Okay. Let me tell you what the meta-analysis researchers did.

They combined everything. So they were looking at that change in heart attacks for the full week after the time change, not individual days. And they reported everything essentially as that same ratio we were talking about before.

And for spring, that combined ratio was 1.04, which was statistically discernible, statistically significant. And autumn, non-significant. Basically null.

Basically 1.00.

[Kristin] (43:36 - 44:09)
Oh, interesting.

So they are seeing that increase in the spring, but they're not seeing any decrease in the fall. And 4% increase in heart attacks over that entire week, that's actually pretty similar to what they found in the New England Journal paper. That was a 5% increase.

So even though there is some variability in the underlying studies, this overall pooled effect is very consistent with the Swedish study. And bottom line, again, looks like a small effect, very small effect, that potentially translates to some actual heart attacks, but certainly not a public health catastrophe. Maybe a few extra events per million again?

[Regina] (44:09 - 44:28)
That's exactly what it was. Yeah, I did a back-of-the-envelope calculation. This time I was curious if that rate applies to the U.S., and it might not. But just if it does, it would work out to an extra 600 heart attacks in that week after the spring time change in the entire U.S. Right.

[Kristin] (44:28 - 44:51)
So that's a lot more than the 20 in Sweden because the U.S. is so much bigger. But 600 is actually not a big number compared to the just total number of heart attacks that happen in all of the U.S. And again, we're talking heart attacks, not death. So not all of these are fatal heart attacks.

So again, Regina, it's not really clear to me that this is a clinically meaningful increase. Do we have any other studies that we're going to look at?

[Regina] (44:51 - 45:01)
We do. We, in fact, have a 2024 study from the U.S. only. That is very interesting.

But how do you feel about a short break right now?

[Kristin] (45:01 - 45:37)
Sounds good, Regina.

Welcome back to Normal Curves. Today, we are talking about whether daylight saving time increases the risk of heart attacks.

And we were about to talk about a 2024 study done in the U.S.

[Regina] (45:37 - 46:07)
Right. So while that meta-analysis was being done over in Germany, there was another group of researchers at Duke University here in the U.S. looking at this whole topic and thinking, hey, heart disease treatment and heart attack prevention have changed a lot since the olden days, right? Like 1987 is when that Swedish study started. And that is definitely not the same medical landscape as United States in 2022. And they said, hey, let's do a giant modern study here in the U.S.

[Kristin] (46:07 - 46:36)
I think this is really important, actually, Regina, because the U.S. has actually done really well on treatment and prevention of heart disease. And that might affect things. If we have good treatments, then maybe people who are right on the edge of a heart attack, we've pulled it back a little bit so they're not susceptible to something that's a small stressor like an hour of sleep loss.

So this seems like it was important to look at this more modern landscape.

[Regina] (46:36 - 46:56)
Absolutely. So they used a formal research registry called the National Cardiovascular Data Registry and has data from over 1100 hospitals in the U.S. And it's high quality data that is collected specifically for research purposes, not specifically for this question, but research in general.

[Kristin] (46:56 - 47:16)
And this is a super important point, because we have talked about using big databases for research on this podcast before, but sometimes we're using things like electronic health records, which are not collected for research. They're collected for billing and clinical care. And those data can be messy.

So this is really good that these data were specifically collected with research in mind.

[Regina] (47:16 - 48:34)
I think that's an excellent point to keep in mind, because sometimes we just see, OK, they have a lot of data, but we forget to ask, OK, what's the quality of the data behind it? And in this case, it was good, high quality data. There was standardized data collection and there were regular data audits that keep quality high.

So that was good. So the research team looked at about 169,000 heart attack patients over 10 years in the U.S. from 2013 to 2022, so more modern. And they looked at those where the heart attacks occurred right around the daylight savings time clock changes.

And basically, they did a similar kind of thing to what the Swedish study did. They looked at the heart attacks the week of the time change, compared it to baseline the week before the time change, and also separately to the week after the time change. And they looked at both daily measures and weekly measures.

Yeah. Oh, and Kristin, you will like this because that was the kind of the bone, the skeleton, but they also did a lot of fancy models. Including one where they accounted for dependent data, meaning they accounted for the fact that patients are clustered within hospitals.

[Kristin] (48:34 - 48:51)
Oh, this is great. So they get a normal curves gold star too, because we've talked on this podcast before about correlated observations. Turns out that patients within the same hospital, they are correlated.

They're more similar to each other than to patients in other hospitals. And it's really good to take that into account in your modeling. Excellent.

[Regina] (48:52 - 49:06)
Absolutely. Gold star. So they did other good things.

They did sensitivity analyses, Kristin. They used Hawaii and Arizona as negative controls. Why those?

Because those states are rebels and they do not change the clocks.

[Kristin] (49:07 - 49:34)
Oh, that is excellent. We love negative controls. Because say, for example, you found a 1.05 incidence ratio in your study for all the other states, but then you went to Hawaii and Arizona and you saw that they also had a 1.05 incidence ratio for the time period right after daylight saving, but they didn't have daylight saving. Then you would know that that 1.05 was likely due to other confounders, other reasons and not the daylight saving itself. So it's great to have that control in there.

[Regina] (49:35 - 49:49)
Exactly. Great, great explanation there. They also expanded the comparison window out from just plus or minus one weeks to plus or minus three weeks just to see, you know, is there something particular about the one week or does this effect generalize out?

[Kristin] (49:49 - 49:55)
Right. Really want to make sure it's robust and not just a fluke or noise. All right.

So that sounds great. What did they find?

[Regina] (49:55 - 50:26)
Yeah. So in the spring time change week compared to one week prior, the incidence ratio was 1.01, so one percent more heart attacks, tiny and not statistically discernible, not significant. And looking at that week of the spring time change versus one week later, so that other control group, the incidence ratio was 1.03, so three percent more heart attacks, also not statistically significant.

[Kristin] (50:27 - 50:50)
OK, so they didn't average the before and after baseline comparator groups. They looked at them separately, but they're getting 1.01, 1.03, not significantly different. So this is now even closer to 1.0, closer to null, although I have to say it's not wildly different than the Swedish study because 1.05 versus 1.03, they're not that far away. In all cases, really small, maybe nothing there.

[Regina] (50:50 - 51:01)
Right, right. Now, I just want to mention autumn, too. Those incidence ratios for the autumn time change were hovering right around one, not statistically significant.

[Kristin] (51:01 - 51:02)
So they didn't find anything protective there.

[Regina] (51:03 - 51:26)
No, they did not. And Kristin, that was for all 10 years in the study. But they also looked at the ratios year by year just in case the effect was changing over time.

They wanted to be able to catch that. And they found no consistent trend except for one weird blip, and that was March 2020.

[Kristin] (51:27 - 51:38)
Yeah, this is actually super interesting, Regina. So you sent me that graph, and there is a very noticeable blip on that graph in March of 2020. And we can all guess what was happening in March of 2020.

[Regina] (51:40 - 52:24)
I tried to block it out of my memory. But yes, there were 20% more heart attacks for that spring time change week compared to the week after. Like you said, it was very noticeable.

Everything was around 1% or 3%. All of a sudden, you see this big blip of 20%. So I went back to look at the history timeline of the pandemic.

So Sunday, March 8th, was the day that we changed the clocks to spring ahead. And then on Sunday, March 15th, which was exactly one week later, the states started to shut down. They started the shutdowns of schools and bars and restaurants and everything.

And the streets were just deserted and eerie. That was then.

[Kristin] (52:25 - 53:05)
Right. So actually, that blip totally makes sense if you think about it carefully. We are comparing the week of the spring time change to the week after.

And it turns out that that week after was when all the lockdowns happened. So guess what? There was probably an artificially low number of heart attacks that week because people didn't leave their houses to go into the hospital or to the doctor and get diagnosed with a heart attack.

So the Daylight Savings Week only looks higher because the Comparison Week had an unusually low number of recorded heart attacks. So this is the kind of like noise or blip in your data that can sometimes make effects appear when they don't actually exist. It's really just a coincidence.

[Regina] (53:06 - 53:24)
Exactly. It's just a very bad coincidence here. So the researchers tried excluding the pandemic-era data just to see what happened.

And remember that 3% increase in heart attacks I mentioned? That went down to just a 1% increase in heart attacks. Not significant.

[Kristin] (53:24 - 53:31)
Wow. So if you take out pandemic-era data, basically there's nothing there. What did they find with the negative controls?

[Regina] (53:32 - 53:38)
So also nothing, no effect in Hawaii, no effect in Arizona, which we would expect because there was no time change then.

[Kristin] (53:38 - 53:45)
Right. It would be really weird if we found nothing in the rest of the U.S., but we found a time-change blip in the states that didn't have the time change yet.

[Regina] (53:45 - 54:32)
Right, right. There was also nothing when you go three weeks out. So this is all reassuring.

I just want to like point out how you and I, Kristin, as statisticians, look at this. These kind of negative controls, they're really good protection if it turns out that your main results show something significant. And of course, we didn't need them here because there was nothing significant in our main results.

Right, but as statisticians, you and I look at this and we say, that's okay, it's good that they did it and that they reported they did it because it's showing us that they know good statistical best practices and planning. And it means that we trust these researchers just a bit more. But overall, yeah, I would say these are a lot of null results.

[Kristin] (54:33 - 55:23)
Yeah. Yeah, this is very interesting because the Swedish study and that meta-analysis, those both had a very small but statistically significant effect, whereas this massive modern U.S. study says basically nothing. It could be that, you know, the modern era part matters, that since, again, we're treating heart disease better, maybe this just doesn't make as much of a difference in the modern era.

But Regina, I want to go back to something you said at the beginning of the episode, because you cited, I think it was a 2025 BBC story, where it said that there was a 24% increase in heart attacks after daylight saving, if I'm remembering correctly. So where did that 24% statistic come from? Because it's definitely not lining up with any of the studies that we've been talking about today.

[Regina] (55:23 - 56:37)
Kristin, you have an excellent memory, and I'm very glad that you brought that up.

[Kristin]
Only for numbers.

[Regina]
Yes, so let me show you behind the scenes and what did not make it into that BBC article.

First of all, that was just one study in 2014, and it involved just two hospital centers in Michigan in the U.S. Oh, so this is a much smaller study, probably a lot fewer heart attacks, and maybe that 24% was just some kind of fluke noise in the data just due to chance? Absolutely, and the authors of that study actually admit that. They admit that their analysis was exploratory because there was a high chance of false positives, they were looking at a lot of things.

And Kristin, even more interesting, that 24% was Monday only. If you look at Tuesday through Friday in that same study, the heart attacks were actually slightly lower than baseline, about 2% to 3% lower each day. And when you average across the entire week, you're looking at something closer to a 3% increase overall, not 24%.

And, by the way, that weekly effect, not statistically significant.

[Kristin] (56:37 - 57:12)
Wait a minute, so that's very much in line with all the other studies that we're talking about, 1.04, 1.05. So that BBC reporter really, really cherry-picked that result because a more fair number, even from that single paper, would be a 3% increase. And they ignored all the other data on this topic that we've been talking about because the BBC story was 2025. They cited a paper way back in 2014, when there's been all of these other studies that don't line up with that.

And that 24%, I still think, Regina, this could just be a fluke. We happen to get a little bit higher on Monday in this smaller data set.

[Regina] (57:12 - 58:25)
Absolutely, Kristin. This was the most striking finding. So, hmm, that's the one that gets pulled and gets quoted, right?

And you're right. It could be a fluke. But the authors actually talked about this in the discussion in their paper, and they offer a very interesting explanation that I think is somewhat plausible, and that is the harvesting.

They don't say harvesting in particular. They don't call it harvesting the souls. But the idea is that some of the patients were already extremely vulnerable.

That's what they said, and kind of on the brink of a heart attack. And the author said that the time change could potentially just accelerate heart attacks that were likely to occur in patients anyway. You know, you take away an hour of sleep and a little stress, that pushes them over the edge.

They might have had the heart attack a bit later anyway. That's the harvesting idea. And they said it looks like daylight saving time is just shifting the timing.

Temporal fluctuations is what they called it. Not increasing the total number of heart attacks in the population. So, they were actually arguing that daylight saving time is not actually increasing heart attacks overall.

[Kristin] (58:26 - 58:39)
Oh, interesting. So, again, the BBC reporter ignored the entire context of the study as well. But that's an interesting thought.

So, maybe the people who are really vulnerable, their heart attack happens on a Monday, but it would have happened on Thursday anyway.

[Regina] (58:40 - 58:44)
Yep. The BBC reporter said heart attacks soared by 24%.

[Kristin] (58:44 - 58:50)
Well, that's not a good verb choice there. That's very misleading. It's more like heart attacks were reshuffled.

[Regina] (58:53 - 58:55)
That's not quite as exciting though, is it?

[Kristin] (58:55 - 59:16)
So, Regina, it sounds like we've covered the landscape here. So, I think we are ready now to wrap it up. We always wrap it up by rating the strength of the evidence for the claim using our 1 to 5 highly scientific smooch rating scale where one smooch means little to no evidence for the claim and five smooches means really strong evidence for the claim.

And so, Regina, can you remind us what the claim was today?

[Regina] (59:16 - 59:24)
Yep. It's that the spring daylight saving time transition, that spring ahead one hour, increases heart attacks.

[Kristin] (59:24 - 59:27)
All right. What do you think today, Regina? Kiss it or diss it.

[Regina] (59:28 - 1:00:28)
Oh, you know, I am so torn on this one because I really wanted this to be ironclad, as I have admitted the entire episode. I am tempted to give it just one smooch because I feel like the whole preponderance of evidence is really saying that there's probably nothing, even if there is something, it's a very small but not very meaningful effect policy-wise and it might just be this reshuffling, you know, this harvesting. And it just doesn't make any sense to me that one hour of time change is enough to increase heart attacks because people fly from Chicago to New York all the time and they are not dropping dead, you know, in LaGuardia.

So, I think the whole thing. Okay, so I want it to be one smooch, but I'm going to go ahead and give it two smooches just because I'm human and I want it to be true.

[Kristin] (1:00:30 - 1:01:46)
I love the honesty, Regina. I'm actually going to go and I have no stake in this one, really. I don't care either way.

But I'm going to give it 1.5 smooches, not the absolute bottom, because there might be a tiny real effect. And I'm going to relate it to something familiar to me, which is research on running. There's good evidence that marathons can trigger a small number of heart attacks during the race, especially in men.

The risk is extremely tiny, something on the order of five heart attacks per million male runners. So, obviously, no doctor would say don't run a marathon. But I find it biologically plausible that the stress of the race might push someone who's already vulnerable over the edge.

So, I could imagine something similar going on with daylight savings, losing an hour of sleep, disrupting your circadian rhythm, might be a small stressor that nudges just a few vulnerable people into having an event a little earlier than they might have otherwise. The effect, though, if it exists, it's extremely tiny. And the Duke study suggests that maybe it doesn't even exist in the modern era.

So, overall, definitely not a meaningful public health problem. But I'm not going to rule out a very, very small effect. So, 1.5 smooches for me.

[Regina] (1:01:46 - 1:01:49)
Oh, I like your very open-minded attitude. That's good.

[Kristin] (1:01:50 - 1:01:54)
Yes. Since I don't care either way, that's the unbiased take.

[Regina] (1:01:55 - 1:02:14)
I want us to go to the Roman method of timekeeping. I looked this up. Apparently, you just said everything relative to when the sun was rising.

So, if I'm talking about a Monday morning class, instead of a Monday morning class being at 8 a.m., I say the Monday morning class is three hours after sunrise.

[Kristin] (1:02:16 - 1:02:23)
Well, that makes a lot of biological sense, but logistically, that's going to be way too hard to keep track of. Sorry, Regina.

[Regina] (1:02:23 - 1:02:26)
Oh, logistic. Who needs it?

[Kristin] (1:02:28 - 1:02:33)
Our biological clock certainly might be happier, but there's no way I'm going to be able to keep track of all of that.

[Regina] (1:02:34 - 1:02:36)
Okay, practical you.

[Kristin] (1:02:36 - 1:02:54)
And I'm sorry, Regina, but I don't think we're going to be able to send this episode to Congress and have them use it as evidence to pass the legislation to get rid of daylight saving. So, I think your ultimate goal for this episode was not realized, but good on you for being objective.

[Regina] (1:02:54 - 1:03:00)
See, look at that. Look at that. It is possible, despite my deep, intense hatred for daylight saving time.

[Kristin] (1:03:01 - 1:03:03)
All right, Regina, how about methodologic morals? What do you have for us?

[Regina] (1:03:03 - 1:03:22)
I think I'm going to go with the harvesting, because we have not talked about this yet, and I am kind of fascinated by harvesting the soul. So, how about this? A bump in time isn't always a bump in total.

The wind can shake a few apples loose early, but it does not grow more apples.

[Kristin] (1:03:23 - 1:03:42)
Oh, I love that. The harvesting effect. Very good.

I'm going to pick on the BBC reporter. I really find that cherry-picking quite annoying, actually. So, mine is, if you already know the story you want to tell, you can always find a number to tell it, but that's cherry-picking.

[Regina] (1:03:43 - 1:03:59)
Oh, absolutely. Absolutely. And cherries and apples.

[Kristin]
Yeah. Cherry-picking, apple-picking. It seems we have a fruit theme going in our methodologic worlds.

[Regina]
I'd like some pie, some cherry apple cobbler, maybe.

[Kristin] (1:04:01 - 1:04:07)
Or just that the raw fruit is healthier. That could be a moral, too, but not methodologic. That's a health moral.

[Regina] (1:04:07 - 1:04:10)
No, no. It needs to be with brown sugar and with ice cream on top.

[Kristin] (1:04:12 - 1:04:22)
That sounds delicious, Regina. Thank you so much for keeping an open mind. And I learned a lot, actually, on this episode.

I had no idea about the history of daylight saving. So, thank you.

[Regina] (1:04:23 - 1:04:33)
You are welcome. Join me. Join me in the hatred.

I welcome you.

Thanks, Kristin. And thanks, everyone, for listening.

Have a great Daylight Saving Day.