Twin births aren't exactly common. In humans, they occur in one to three per cent of all births.
Previous studies of the phenomenon have concluded that mothers of twins are more fertile than other women. This is because on average they give birth more often than other mothers. They have been called "supermothers" and are considered more robust and in better health than mothers of single babies.
Now an international research group has found that twin mothers are not actually more fertile than other women.
"On the contrary, when a woman gives birth several times, the chances increase that at least one of these births will be a twin birth. Twin mothers aren't supermothers, but have been given more chances," says Gine Roll Skjærvø, a senior engineer and human behavioural biologist at NTNU's Department of Biology.
[...] "Previous studies are problematic because they can't tell us whether mothers with twins give birth more often because they're especially fertile, or because giving birth more often increases the chance that one of these births will be to twins," said lead author Alexandre Courtiol at the Leibniz Institute for Zoo and Wildlife Research in Germany.
The new results show that women who give birth to twins are not unusually fertile. The previous research mixed up cause and effect.
"If a mother gives birth more often, it's more likely that one of these births will be to twins – just like you're more likely to win if you buy more lottery tickets, or to be in a car accident if you drive a lot," says Ian Rickard, a lead author from Durham University in the UK.
When taking into account this lottery effect, the researchers find that mothers of twins actually give birth less often than others, not more often. This new finding is in stark contrast to previous ones.
Journal Reference:
Rickard, Vullioud, Rousset, et al., Mothers with higher twinning propensity had lower fertility in pre-industrial Europe, Nat Comm, 2022. DOI: 10.1038/s41467-022-30366-9
(Score: 2) by Snotnose on Sunday August 07 2022, @08:28PM (1 child)
They aren't more fertile. Statistics show that when they do get pregnant they have a better than average chance of having twins.
FFS. Just because they have twins doesn't mean they're twice as fertile. Swear to $diety, I really wish basic statistics was a high school graduation requirement.
I just passed a drug test. My dealer has some explaining to do.
(Score: 2) by Immerman on Monday August 08 2022, @03:52PM
>Statistics show that when they do get pregnant they have a better than average chance of having twins.
Oh? And how would you conclude that? Having two sets of twins is extremely rare, and with only a single occurrence you can't conclude anything about a particular woman's odds of having twins. And even two sets could just be a fluke, especially with a large family. At best you might be able to say that some family lines have a higher than average chance of having twins.
(Score: 4, Insightful) by Thexalon on Sunday August 07 2022, @08:35PM (4 children)
Talk to any mother of twins, and tell me if they think they're "lucky" for having to deal with 2 whiny hungry newborns simultaneously. Or in the even more extreme version, I knew one family who had a girl, and then had quadruplet boys, and I have no idea how their mother survived the infant and toddler stage.
The only thing that stops a bad guy with a compiler is a good guy with a compiler.
(Score: 4, Interesting) by choose another one on Sunday August 07 2022, @09:24PM (3 children)
Yeah, this promptly explains their second result: "that mothers of twins actually give birth less often than others, not more often"
Once you have twins, probability that you will want to get pregnant again is almost certainly significantly lowered compared to if you have non-twin birth(s), add in availability of contraception and bingo, probability that you will have further births after twins is reduced.
For further bleeding obvious results, see if the same applies to triplets etc. - we have neighbours who had triplets, they didn't have any other kids, ever, how very odd... not.
(Score: 4, Informative) by Booga1 on Sunday August 07 2022, @10:06PM (2 children)
Agreed. At least the study is from 100,000 pre-industrial births instead of modern society, but similar social effects are present. You can't "flip a coin" 100 times with a woman birth rates. At some point your family is big enough, or you're just tired of having babies.
Although the researcher is comparing giving birth to buying lottery tickets, it seems to me that the journalist is the one making the more absurd conclusions from the research. Though, the researcher themselves are the ones claiming that the prior studies were the ones with flawed methods, the journalist seems to be taking the statements one step further than the researchers were willing to go.
On top of all that, several studies have shown that mothers of twins are more likely to have twins again if they get pregnant again. They've even identified genetic markers for it: https://www.livescience.com/54597-fraternal-twin-genes-found.html [livescience.com]
(Score: 2, Informative) by aafcac on Monday August 08 2022, @03:19PM
Yes, and in industrial societies, women tend to want 3 or fewer children. There are some that want more than that, but childbirth, even in developed countries, is still rather risky. Not to mention that the process is rather unpleasant and you've then got a bunch of inconveniences afterwards.
(Score: 3, Interesting) by Immerman on Monday August 08 2022, @04:09PM
>At some point your family is big enough, or you're just tired of having babies.
Absolutely. But in pre-industrial times that didn't necessarily make much difference. Neither herbal spermicides, nor sheep-gut condoms, rhythm methods, nor pulling out provide particularly effective birth control. Even latex condoms aren't incredibly effective. It took the invention of the pill to really give people reproductive choice.
Complete abstinence is obviously effective - but not an option many married couples are willing to make.
And given the historical power imbalance lets be honest - not a choice many married men would be willing to accept. Heck, even today there's many places where women have to seek out undetectable birth control in secret to keep their family size down to something manageable, because their men see more children as proof of virility, even as supporting them drives the family to destitution.
(Score: 4, Touché) by Opportunist on Sunday August 07 2022, @09:47PM
After having to deal with two hungry poop-machines at the same time, you NEVER EVER forget the pill again. Ever. And you use a condom on top of that. And abstain from vaginal sex. Or just become lesbian. Or do anything else to make abso-fucking-lutely sure that this will NEVER EVER happen again.
(Score: 1) by Sjolfr on Sunday August 07 2022, @10:45PM (9 children)
Each pregnancy has the exact same chances of being twins, 1-3%, when you adjust for individual genetics. Being pregnant more does not increase your chances.
I would suspect that there is a genetic tie to having twins when you control for fertility drugs and such. Not "fertility" related genes, but genes that split the egg or cause 2 eggs to deploy.
----------------------------------------------------------------------------------------------------------------------
Why am I replying to online articles and conversations ... some day I will learn but today is not that day.
(Score: 3, Informative) by Immerman on Monday August 08 2022, @04:39PM (8 children)
>Being pregnant more does not increase your chances.
Not your chances of the next pregnancy being twins - but it *dramatically* increases the odds that you will have had twins.
E.g. with a 2% chance to have twins:
1 pregnancy: = 98% chance of no twins = 2% chance of twins
2 pregnancies = (98%)^2 chance of no twins = 4% chance twins
10 pregnancies = (98%)^10 chance of no twins = 18% chance of twins
There is definitely a genetic component that makes the women in some families more prone to twins, you even mentioned it yourself:
(Score: 1) by Sjolfr on Wednesday August 10 2022, @04:40PM (7 children)
It seems to me that the driver is more likely to be the genetic predisposition of having twins over time and not simply "the odds go up with more pregnancies".
Woman #1 with no predisposition
Pregnancy 1: = 98% chance of no twins = 2% chance of twins
Pregnancy 2: = 98% chance of no twins = 2% chance of twins
Pregnancy 3+: = 98% chance of no twins = 2% chance of twins
Woman #2 with predisposition
1 pregnancy: = 98% chance of no twins = 2% chance of twins
2 pregnancies = (98%)^2 chance of no twins = 4% chance twins
10 pregnancies = (98%)^10 chance of no twins = 18% chance of twins
(Score: 2) by Immerman on Wednesday August 10 2022, @07:12PM (6 children)
I think you're misunderstanding what I'm saying. I probably should have been more explicit with my math.
I'm absolutely assuming the probability of having twins remains the same with each pregnancy, but the odds that one of those births will have been twins *must* go up.
Consider flipping a fair coin three times - the ultimate predisposition-free outcome. Obviously, flipping three coins gives you a better chance of getting tails on at least one of them than only flipping one, right? To get the same odds regardless of the number of coins you throw, all but the first would have to be double-headed coins that *couldn't* be tails.
With getting a tail being analogous to having twins:
With the first flip you have a 50% chance of heads = 50% chance of tails
With the second flip there's a 50% chance that you got tails the first time PLUS 50% of the remaining chance (=25%) that you got tails this time = 75% that you've gotten a tail
With the third flip there's the 75% chance you already got tails, PLUS a 50% of the remaining 25% chance (=12.5%) that you got tails this time = 87.5% that you've gotten a tail
Mathematically, the formula is
(odds that thing will happen within N tries) = 1-(odds that the thing won't happen in one try)^N
So after 3 coin flips with a 50% chance of NOT getting tails each time, there's a 1-(0.50)^3 = 87.5% chance that you got a tail at least once
And after 5 pregnancies with the same 98% chance of NOT having twins each time, there's a 1-(0.98)^5 = 9.6% chance that you've had twins at least once.
(Score: 1) by Sjolfr on Thursday August 11 2022, @09:40AM (5 children)
Flipping 3 coins gives you more chances at the same odds. The odds of you flipping heads does not change from flip to flip. Pulling on a slot machine handle 100 times does not increase your chances of winning. The odds are the same each handle pull. Past outcomes do not impact future outcomes ... which is exactly the definition of the gambler's fallacy.
Gambler's Fallacy "is the incorrect belief that, if a particular event occurs more frequently than normal during the past, it is less likely to happen in the future (or vice versa), when it has otherwise been established that the probability of such events does not depend on what has happened in the past." --wikipedia.
If you flip a coin 2000 times it does not impact the next flip. The 2001st flip will be 50/50 chances of head/tails.
(Score: 2) by Immerman on Thursday August 11 2022, @03:54PM (4 children)
Exactly. And I never suggested otherwise. After this one, re-read my previous posts more carefully if you still think that I have.
At the risk of making an appeal to authority, I have a degree in mathematics and am VERY certain I'm correct about this, so please give me the benefit of the doubt and assume that I know what I'm talking about as you consider what I'm saying.
You are looking at the odds of an individual event - which in this case never changes, regardless of the outcome of previous events.
I'm taking that assertion as my baseline (it's what allows me to use the formula I gave - the math gets a lot more complicated if events can influence each other), and then using that to compute the odds of a particular sequence of events occurring - specifically, the sequence in which I never get tails.
Again, I am explicitly avoiding the Gambler's Fallacy by choosing a formula for computing compound probabilities that REQUIRES that individual event outcomes can't influence each other.
When I say the probability of twins after 10 pregnancies is 18% I am NOT saying there's an 18% chance that her tenth pregnancy will be twins. I'm saying that with a 2% chance per pregnancy, the combined probability that her 1st pregnancy is twins OR her 2nd pregnancy is twins, OR her 3rd, ... OR her 10th all adds up to 18% (including all combinations where two or more pregnancies are twins). Which is to say that if you look at all women who have had 10 pregnancies, you should expect ~18% of them to have had one or more twins.
If that's not what you see (within the bounds of expected probabilistic noise) then there are only two likely possibilities: either the probability per pregnancy was NOT 2%, or the individual events actually are influencing each other's outcomes.
Clearly something about my explanation isn't making sense to you... so let me try explaining with one of the simplest possible method of computing compound probabilities - listing every possible sequence of events and computing the probabilities from there: (I'm going back to coins since this is only simple when all possible outcomes of a single event are equally likely)
If I flip a fair coin 2 times there are four possible outcomes, all with the same probability of occurring:
HH (heads both time)
HT (heads, then tails)
TH (tails, then heads)
TT (tails both times)
Notice that there is only one possible sequence where I don't get any tails, but three different ways I can get at least one tail. So the odds of never getting tails is 1/4=25%, and the odds of getting at least one tail is 3/4=75%
If I add a third flip I double the number of possible sequences:
HHH, HHT, HTH, HTT, THH, THT, TTH, TTT
Notice that despite doubling the number of equal-probability sequences, there is still only one possible outcome where I never get tails. So my odds are 1/8=12.5% of never getting tails, and 7/8=87.5% of getting at least one tail
The formula I used simply lets you compute the probability that if you make N attempts, X (e.g. getting tails) will happen at least once. Assuming that none of the attempts influence each other in any way, and even if it's not a 50/50 split between possible outcomes of each event.
(Score: 1) by Sjolfr on Thursday August 11 2022, @09:51PM (3 children)
I'm not disputing your math, I'm disputing the outcome. The Law of Large Numbers corrects your increasing percentages over more time with more iterations. The probability of heads v tails returns to 50/50 (or close enough). So, over time, the probability of twins in pregnancy remains ~2%.
In your math, that combines probabilities over successive iterations, it takes 2000 iterations to come to a 100% chance of having twins. It seems to me that those small set probability percentages describe the thinking when someone engages in the gambler's fallacy. That's also why casinos rely on the Law of Large Numbers to maintain their profitability and mitigate perceived short term probability variations.
(Score: 2) by Immerman on Thursday August 11 2022, @11:43PM (2 children)
What do you mean by "gets to 100%"? It *never* gets to 100%, that next event could *always* be yet another single birth. 2000 births still has a better than 1 in 10^16 chance that there would be no twins. Certainly close to a 100% chance of twins, but not quite there. And you could say the same about 1000 births, or 100. Only the definition of "close" changes.
I'm not sure what you're getting at with the bit about small set probabilities. Set probabilities are the antithesis of the gamblers fallacy.
Just top make sure we're using the term the same way:
Proper probability calculation says that flipping 10 coins has a 1 in 1024 chance of getting all heads.
The gambler's fallacy is recognizing that you just flipped 9 heads, and thinking that means there's only a 1 in 1024 chance that the next one will be heads. Which is obviously false, because in that case you want the probability of flipping 10 heads GIVEN THAT you have already flipped 9 heads. And since the odds of reaching your current situation were already 1 in 512 against, that comes out to (1/1024)/(1/512) = 50%.
However, that doesn't change the fact that if you have a million people flip 10 coins, roughly 1000 of them will get 10 heads. The gambler's fallacy simply doesn't apply to population-level analysis that way. Not unless you're doing something to pre-bias your population.
(Score: 1) by Sjolfr on Friday August 12 2022, @12:29AM (1 child)
I get that you're arguing on the side of probability math. Yes, that's accurate. Here's your formula: 1-(0.50)^3 = 87.5% We can all check it and see it.
Use larger numbers instead of 3: 1−(0.50)^100 = 0.9999999999999999999999999999992111390947789881945882714347172137703267935648909769952297210693359375
1−(0.50)^334 is the threshold where my 100 decimal place calculator rounds up to 1 (100%). I think that is a reasonable place to round up for practical purposes. I actually think that this kind of math just shows that reality is far more complicated than this kind of probability math.
You can say it as many times as you like and go over your math as many times as you like. More detail doesn't change the practical facts of the matter. In the end the percentages don't change in a statistically meaningful way, hence the Law of Large Numbers.
(Score: 2) by Immerman on Friday August 12 2022, @03:03AM
Of course percentages don't change in a meaningful way. Not on a per-event basis. At least that's the default assumption.
We're just trying to answer the seemingly simple question:
What are the odds that a woman with N pregnancies will have had twins?
The answer to that is obviously ~2% for one pregnancy
And it obviously goes up with each additional pregnancy, simply because each additional "try" adds another chance for it to happen
The law of large numbers just means that the probabilities we compute should map directly to the number of examples in the population.
E.g.
for every 100 women with one pregnancy, 2 of them should have twins
for every 100 women with 2 pregnancies, 4 of them should have twins
for every 100 women with 10 pregnancies, 18 of them should have twins
etc.
That is what we expect if every pregnancy has exactly the same odds of being twins, regardless of how many pregnancies a woman has had.
Your example above with 2% each time is ridiculous. If women with 1 pregnancy have a 2% chance of having had twins, AND women with 2 pregnancies ALSO have a 2% chance of having had twins, then that means there was NO chance the second pregnancy could have been twins, because the first pregnancy already supplied the full 2%.
(Score: 2) by Rupert Pupnick on Monday August 08 2022, @12:26AM
Never read about a probability estimate that varies by a factor of three before.
(Score: 0) by Anonymous Coward on Tuesday August 09 2022, @04:14AM
There is a liquor store around here that has had quite a few lottery winnings in the past. Because of that, during potentially large lottery winnings, the line gets really long and can (or at least used to) go for blocks. This creates a self fulfilling prophecy, people have ended up winning the lottery from that location again, and again, and again. But it's not because that place is somehow luckier than any other place, it's because people keep buying tickets from there thinking that it's lucky and this increases the chances of a winning occurring at that location.