Does The Pill Impact Fertility? (Episode 630)

Full Transcript: Episode 630

Lisa Hendrickson-Jack:

This is the Fertility Friday Podcast, episode number 630.

In today’s episode, I am tackling an important topic. This is a topic that I have tackled in my writing previously in both The Fifth Vital Sign and a touch in Real Food for Fertility, but it is the topic of the phenomenon that we can refer to as delayed return of fertility post-contraception.

In previous episodes and in my previous writings, I have shared studies. They’re quite hard to find. The few and far between studies that actually look at what happens when women discontinue using hormonal contraceptives of various types.

And when I say rare, what I mean by that is that it’s not often that you actually find a study that will look at what happens when women come off contraceptives and actually look at it kind of month over month.

Most studies will look at what happens in a 12-month period. Most studies will take a group of women, they’ll come off of birth control, they’ll exclude anybody that had pre-existing conditions that could affect their menstrual cycle, and they will look at the 12-month mark and say x percentage of women conceived within 12 months — therefore, the birth control pill is wonderful and there’s no delay in the return of fertility. That’s generally how studies were designed, or the majority of studies that I’ve looked at. And that makes it quite challenging to be able to accurately analyze what is really going on when women come off of birth control.

Today I’m actually going to be sharing a little bit of information from two different studies. You could think of this as a bit of a critique. I do these types of episodes because I think it’s important when we’re looking at data for us to really be critical of the data and for us to read beyond the abstract, look at the study design, and really kind of identify what they actually studied, what the researchers are claiming — like what their conclusions are — and if those conclusions match the data, and also how the study was designed, like if it’s even reasonable that we could make those same conclusions.

There are two recent studies. One came out in 2020, one came out in 2023. And what they are looking at and reporting on is the discontinuation of birth control and how it potentially impacts fertility.

The first is actually looking at fertility after coming off of hormonal contraception among women of Northeast Ethiopia. On its face, it seems pretty reasonable. Essentially what the study claims — what their main finding is — is that about 89% of women, 88.6% if we want to be specific, returned to fertility within 12 months of coming off of hormonal contraception.

Now, what’s interesting about this study is they actually did things quite differently to what I’m used to seeing in these discontinuation studies. The researchers recruited 402 women who were already pregnant. That was their sample. And then they did essentially a retrospective study asking them to recall how long it took them to conceive after coming off of birth control.

Every single woman in the study was literally pregnant before the study even started. Logically, anybody who had come off contraception and didn’t get pregnant is literally not even part of the study. This is a really interesting example of how study design shapes results.

They did find some data that is interesting. They’re looking at these women who were already pregnant, they’re asking them when they came off of birth control, and they are able to kind of see how long it took this particular sample. They did find that the women who had used the shot — Depo-Provera — took longer to conceive overall in comparison. They were able to look at contraceptive pill users versus implant versus all these different devices.

But ultimately, going back to the issue with the study design — do you think that it’s fair for the researchers to make this claim that about 89% of women conceive within a year of trying, when the study is designed in such a way to only include women who successfully conceived? By definition, anybody who did have trouble, or anybody who did take longer to conceive, they certainly are not part of the study at all.

I do find that the true and honest objective look at this question — does hormonal contraception have an impact on fertility — I do find that researchers design these studies in a way that would prevent an objective, thorough look at the data.

My initial critique, as I’ve mentioned already, is that the studies I’ve looked at before — the studies I’ve reported on in The Fifth Vital Sign and Real Food for Fertility — those studies tend to exclude women that had long or irregular cycles, or a history of cycle disruption, or suspected fertility issues, or anything like that. They are making sure that the study population doesn’t have any of these other issues because they don’t want pre-existing issues to skew the results. They want a view of birth control by itself, which makes sense when you’re designing a study — to try to exclude extraneous factors like that.

But it would be useful to conduct a study of women who have pre-existing issues, because a significant percentage of women who use birth control are put on it for non-contraceptive reasons. There are plenty of women who are put on the pill because of period pain, or because of acne, or because of irregular cycles. Doing a study with that particular group — where you take the women who you normally exclude and then you do a discontinuation study where you actually see what happens when they come off the pill — would be pretty useful.

In The Fifth Vital Sign I do share the results of one such study where they actually measured that, and they did show a significantly longer time to the return of menstruation. That is logical. But given that so many women are in that situation where they’re using hormonal birth control for other reasons, it would make sense to study it. It seems to me that they typically and intentionally limit what they are studying, because they certainly want to continue to be able to say that there’s not really an issue here — that the majority of women just get pregnant within a year.

Now I want to talk about study number two. This study was published in the BMJ in 2023 — relatively recent. On its face it looks a lot better than the previous one. They are looking at a pretty big sample size — almost 18,000 women across different cohorts in Denmark and North America. These women were not enrolled when they were actually pregnant, so that’s certainly a step up from the previous study.

They recruited women who had started trying to conceive and then followed them forward, asking them every other month if they had conceived. This study actually has promise because it’s not just looking at a 12-month number — it’s actually checking in on them every other month to see what’s happening. They are looking at women who have come off of birth control, who are trying to conceive, and then kind of following them along.

At least in this study, they weren’t already pregnant. We haven’t excluded everybody who wasn’t pregnant by a certain time. One of the potential issues, then, with this study is that anybody who reported that they were trying to conceive for more than six months at the point the study started was automatically excluded.

They’re able to take these women who are trying to conceive — either they just started or they’ve been trying for under six months — and then they start there and make the assessment of the impact of hormonal birth control on fertility based on that.

They were looking at the fecundability rate, which is the per-cycle chance of conception. When we look at the results, they’re kind of saying that there wasn’t really much of an issue within three months or so — everything was looking great in terms of fertility, these women were getting pregnant. Consistent with pretty much every study that’s ever been done on contraceptives, Depo users certainly have the largest reduction in fecundability, showing an 87% reduction in their per-cycle chance of conception at cycle three. They are the ones most likely to be still trying to conceive at the six-cycle mark.

But by designing the study in this way and excluding everybody who had been trying to conceive for six months or more, they excluded about one-fifth to one-quarter of the sample. They excluded 20% to 25% of the potential sample just by making that exclusion.

Logically, if you have this cohort of women who have come off of birth control recently, and there’s a percentage of these women who’ve been trying to conceive for at least six months, and we’re excluding them from the study — but then still making the claim that we’re providing meaningful answers for how quickly fertility resumes after coming off the pill — but we’re literally excluding anybody who took a little bit longer to get pregnant, then it does certainly seem like we’re not looking to find objectively what is going on. We can’t really present an accurate picture if we’re excluding anybody that took longer to conceive.

The researchers’ stated reason for excluding that population is that it reduces recall bias by having women meet that cutoff — essentially that six-month cutoff. But I would say there are obvious issues. The results in this study are quite different, I would say, to the results in the study that I’ve quoted previously and shared on the podcast — the one study that I found that actually did look month by month: month one post-contraceptives, month two, month three, month six, month nine, all the way until 12 months — showing this clear difference between women who had been using non-hormonal methods like condoms versus the pill versus the shot. I have not seen a study since that has been as thorough.

I would imagine that the reason is because if you have a larger data set and you actually show that there is this delay in the return of fertility, it’s just bad press. It’s bad press for the birth control industry.

Another interesting finding in the BMJ study was that they didn’t really find a difference between what they termed long-term users versus short-term users. Again, this is something where I’ve seen the opposite result in other studies. But again, if you exclude the women who took the longest to conceive, then we just don’t have that data. We don’t know what percentage of women who took longer than six months to try to conceive before the start of the study had used hormonal contraceptives long-term, compared to the women who had been trying for less than six months. There’s just data that we don’t have.

For them to make that assertion conclusively — that whether or not she was using contraceptives long-term didn’t make a difference in the results — how do we know that? Because you excluded a pretty significant chunk of your potential data set.

If you were following a doctor who was saying, “See, the pill’s fine — this BMJ paper shows that there’s really no effect,” without this context, without someone pointing out why they excluded the women who had been trying to conceive longer than six months, you wouldn’t even know. You would just think, oh, well, this is a newer study — newer studies must be more current, more accurate, they must be better.

Now, a couple of other critiques that apply to both of these papers. These studies are literally just looking at pregnancy rates in different ways. One is looking at how many women conceived and how long did it take, and the other is looking at fecundability — the per-cycle chance of conception based on the assessment of the data. But ultimately they’re looking at pregnancy outcomes.

There are different categories of studies. When I’ve looked at discontinuation studies, or time-to-pregnancy studies, that’s what they’re looking at. But there is a completely other class of studies that don’t necessarily look at pregnancy, but look at cycle characteristics. One critique of these studies could be that while they’re telling us about fertility and the potential delay in fertility post-pill in their interesting ways based on their study design, they are not telling us what is happening with the cycle.

We don’t actually know how long the average cycle was, how long it took the cycles to normalize, if they had a normal mucus pattern, how long the luteal phase was, or if ovulation was delayed. We just don’t have that information. That could also be a critique. But again, I feel like there are different classes of studies. What I’ve seen is that when you’re looking for that type of data, you’re typically looking at a different kind of study — a discontinuation study that’s specifically looking at cycle characteristics. Whereas if you’re looking for the time-to-pregnancy data, that’s typically a separate kind of study.

With that said, if you could design a really robust study, it would be really interesting to be able to measure both of those things simultaneously with the same population. It would give you the opportunity to see if there is a correlation between the cycle characteristics and the duration of time that it takes a woman to conceive.

After looking at these two studies, it’s kind of interesting for me because initially when I saw these newer discontinuation studies, I was excited. I’m always looking for research to talk about and to update the data that we already have. I’m always looking for discontinuation studies because, like I said before, it’s so hard to find studies that are actually looking at this issue in the type of depth that I would love to see.

I would love to design a study. Any researchers out there who have the ability to do this, please do it. See if you can find a way to actually study this properly — where instead of excluding all these different groups, you include them, but you organize the data statistically and run the analysis. You can run the analysis separately on the women who didn’t have the pre-existing conditions. I would just love to see a study that looks at all the data. It would probably be a bit complex to include all of that, but wouldn’t it be helpful to actually get conclusive answers instead of these studies that just dance around the issue?

I was initially excited when I saw these studies because I thought, great — new studies looking at discontinuation, looking at the impact of hormonal contraceptives on fertility. But I was actually pretty surprised to see these huge holes in the data and to see how flagrantly they are — in my opinion — manipulating the data to create the illusion of a favorable result for hormonal contraception.

I know that’s a very strong statement. But the first study, only including women who are already pregnant — if that’s what you’re doing, then you can’t just boldly claim that almost 90% of women got pregnant, because you don’t even know. You only sampled women that were already pregnant.

And then the other study — by excluding all of the women who took longer to conceive — isn’t that the whole thing you’re measuring? If it really has no effect, then why are we excluding that part of the sample — the women who took a little bit longer to conceive? If you’re purposely making choices like that when you’re coming up with a data set, it does seem like you are looking to get a certain result when you publish the study.

I’ll be really curious as to what you think about this. Feel free to hit me up on Instagram at Fertility Friday. If you’re on my newsletter, I will be writing about this very soon — feel free to hit reply. I would love to hear what you think about it. Is it just me looking at this stuff and thinking, what’s going on? Is there something I’m missing?

If you found this episode to be interesting, if you can think of anybody who would appreciate hearing it, feel free to share. You will find the links to this episode in your favorite podcast player. You can also head over to fertilityfriday.com and this episode will be there in our podcast section.

I hope that you have an amazing week, weekend, whenever you’re tuning into the show. And of course, as always, until next time, be well and happy charting.