I was concerned that I had found a small, statistically insignificant way in which Natural Family Planning (NFP) might cause a spontaneous abortion1 I wasn't going to say anything. Who wants to add to the misery in the world? And yet, in searching to see if anyone else had made the argument, I discovered something far more disturbing about NFP.
A while back, I started a discussion about whether "Plan B" acts as an abortificient. The only modes of action demonstrated are suppressed ovulation and increased cervical mucous thickness. Another possible mode of action, that of preventing the egg from implanting, has been hypothesized but never demonstrated. Opponents of "Plan B" who wish to continue arguing that it is an abortificent claim that the experimental data are not accurate enough, and that there is still a small but unacceptable abortion component to how "Plan B" acts.
I tend to argue by analogy, and so the obvious tack was to question whether NFP causes eggs to fertilize that can either no longer implant or implant but do not have sufficient time to trigger changes that prevent sloughing of the uterine lining.2 Thus, the behavioral changes induced by NFP result in spontaneous abortions, making it just as much a form of abortion as "Plan B" is claimed to be.
I was curious: has anyone else ever considered this? So I looked it up. It turns out I'm arguing about insignificant, mostly theoretical effects when something much larger is actually out there.
The article "Timing of conception and the risk of spontaneous abortion among pregnancies occurring during the use of natural family planning" in Am J Obstet Gynecol. 1995 May;172(5):1567-72. caught my eye. I haven't gotten the full article, but the abstract is online at the link. The conclusion is startling:
CONCLUSIONS: Overall, there is no excess risk of spontaneous abortion among the pregnancies conceived during natural family planning use. However, among women with a history of pregnancy loss, there is an increased risk of spontaneous abortion associated with preovulatory or postovulatory delayed conceptions. (Italics added by the author of this post.)
They're not talking about the effect I was hypothesizing. Pregnancy was only counted in their study if it was detectable — meaning the egg successfully implanted and created sufficient hormone changes to be noticed. The quoted spontaneous abortion rate in the article (10.1%) for the 868 pregnancies demonstrates this — it's far too low to include every egg that was fertilized.
What they found was, that for a subset of the population that has had previous spontaneous abortions, NFP about doubles the risk of a spontaneous abortion for non-optimally-timed conceptions.
The study did not rule out an increase in spontaneous abortions in the case of women without a previous history of same. The difference between 9.1% (optimally timed conceptions) and 10.9% (non-optimally timeed conceptions) is not statistically significant. "Statistically insignificant" is not the same as "no difference." This is the same argument used against "Plan B."
Is there some reason that proven occasional spontaneous abortions caused by NFP is acceptable but as-yet unproven abortions with "Plan B" are not? Or is something else going on?
BTW: Physicians for Life references the paper "Timing of conception and the risk of spontaneous abortion among pregnancies occurring during the use of natural family planning" as proof that NFP does not cause spontaneous abortions. Even avoiding the "statistically inconclusive" argument, the paper clearly doesn't say that. Physicians for Life wasn't the only web site promoting NFP to do that, either. Did they not read the abstract, or did they just figure no one else would?
- Note: throughout this article, I will use the term "spontaneous abortion" as opposed to "miscarriage." "Spontaneous abortion" is the preferred medical term, what I'm used to, and what the journal article uses. [back]
- If you want the basic argument, x/28 < x/(28-y), where x is the "window of missed opportunity" where an ill-timed egg can be fertilized but cannot successfully implant in decimal days and y is the days during which intercourse is avoided for NFP. For any positive, non-zero number for y, the first term must be smaller than the second. If this effect is on the order of a hundredth of a percent, I'd be surprised. But it's a non-zero positive number, and that's the point. [back]
1. How does NFP affect the rate of suboptimally-timed conceptions? Are you saying that using NFP’s rules causes couples TTC to conceive at poor times or couples TTA who unintentionally conceive to do so at poor times?
2. Your formula in footnote 2 needs work. There is no such thing as a 28-day cycle; it’s a myth. That’s just an average, and an old one at that. Not only do different women have different cycle lengths, but each woman’s cycle length varies from cycle-to-cycle.
Since when do abstracts prove or disprove points? Sometimes in an effort to express findings in limited space, mistaken impressions are created. Perhaps the full article supports PFL’s claims.
BTW, I linked to this post in a NFP discussion board I frequent. I think registration is required in order to read posts there, but it’s free.
F.D.,
The equation stands. We’re talking back-of-the-envelope estimation. For the purposes of evaluating the effect, one could sum over the various possible cycles multiplied by their probability. It wouldn’t change a thing, but I’d have to figure out how to type a sigma sign in HTML and then pair up the fractions. It doesn’t change the result.
I suppose one could argue x/z < x/z-y for all possible positive z and y
Anyway, that’s just a theoretical estimation that I put at around .01%. Far more significant is the result of the research.
As far as the abstract: The conclusion is pretty clear. The numbers quoted back it up. If the actual paper contradicts the conclusion that there is an identifiable subgroup for whom NFP causes spontaneous abortions, then how the heck did those statements wind up in the conclusions? There’s no way a journal editor would allow that sort of mistake to happen. They let all sorts of mistakes happen, but not that sort.
You want to make a bet on the paper? You could easily go to Scaife Medical Library and dig that one up. I’ll pay you for the photocopying — can you scan docs in? You might want to check online to see if Scaife has it — but I’d bet they do.
I’m just astounded. NFP doesn’t cause pregnancy. Nor does it cause infertility. It doesn’t cause lack of pregnancy. NFP is ONLY a variety of methods by which a woman may observe physical symptoms that indicate the likelihood of being fertile on a given day.
Now sex, THAT causes pregnancies. And a pregnancy has to be caused in order for a miscarriage to be caused… so don’t these researchers REALLY mean that SEX causes an increased incidence of spontaneous abortions in women who have had previous spontaneous abortions?
Pam,
Thanks for commenting.
I’m not saying NFP causes pregnancy. But when pregnancy occurs, because of the timing, it’s more likely to result in miscarriages in a susceptible subgroup — at a rate far in excess of anything possible for “Plan B” — when it fails.
NFP does fail on occasion.
lol — yes, although i’d be more likely to blame user error over method failure.
that said, somehow i get the impression that it’s still not nfp “causing” greater miscarriage incidence, EXCEPT that couples using nfp ttc are more knowledgeable re. when they are likely to conceive when trying to, than the general population. that really means that nfp works. the problem lies in what ever causes some women to miscarry repeatedly, and i think it’s just silly to lay the blame for that on nfp.
There are a couple other issues to deal with.
1. How many of the couples were TTA?
2. How many early miscarriages happened to couples TTA?
3. How can we compare early miscarriage rates for NFP couples versus non-NFP couples given that most non-NFP couples are unaware that they’ve miscarried?
Anyhow, this is a siple problem to fix. For those miscarrying couples TTA, the rules of NFP should be adjusted to be stricter around the fertile period. For those m.c. TTC, the rules need to adjusted so that intercourse is timed to be closer to ovulation.
Um, Rob, I’m not sure what you’re getting at with this post.
The abstract reads:
“…among 171 women who had experienced a spontaneous abortion in a prior pregnancy, the rate of spontaneous abortion in the index pregnancy was significantly higher with non-optimally timed conceptions (22.6%) as compared with optimally timed conceptions (7.3%).”
Translation:
Women who historically have trouble with pregnancy, are more likely to have trouble when they conceive outside the optimal window.
Seems to make sense to me. How this is a ‘disturbing discovery’ about NFP I have no idea. Am I missing something?
He’s implying that the rules of NFP lead to poorly timed conceptions. Refer to my previous comment for how to remedy the situation. One of the great things about charting is that you *can* remedy the situation by tweaking the rules.
Dear FD,
I may be misreading the abstract, but it seems to me the researchers have used NFP’s own definitions to distinguish between optimal and non-optimal conceptions.
They’ve simply concluded that conceptions during non-optimal times (as defined by NFP) resulted in higher rates of spontaneous abortions for women who have had previous pregnancy troubles.
What remedy are you suggesting?
Non-optimality is defined as resulting from either sex more than a day before ovulation or sex after ovulation. Optimality is sex on the day of or before ovulation. I think the gist is that old sperm and old ova are somehow damaged.
If that’s the case, controlling time of conception is definitely doable. See comment #8 for remedies.
Got it–thanks. (Sorry about missing the remedies reference–saw it after: got lost in the TTA and TTC talk).
So I guess I’m still confused about what the ‘disturbing discovery’ is. The researchers set out to determine whether the timing of conception made a difference on the likelihood of a spontaneous abortion. They found that it did for women who had previous spontaneous abortions. This phenomenon exists whether or not a couple practices NFP, so how is NFP the ‘proven cause of occasional spontaneous abortions’?
Rob,
BV has made an excellent point. The study provides no data to suggest that women with prior miscarriage who use NFP miscarry more often than those who do not. If I understand your post correctly, you are trying to argue that use of the rules of NFP result in a higher rate of miscarriages for that group than would result from random timing of intercourse. The study doesn’t say that, though.
FD:
Couples who use NFP have two options: They can either have sex at their regular rate and just accept the missing days, or they can make up for lost time.
If sex would normally be once a day or more often, the odds are that either choice would not increase the risk of having a child via miscarriage. Sperm can live inside the female body for a surprising time (even more surprising for lizards and birds, but that’s another post that no one but me cares about).
If sex would normally be once a month (or less), again, neither choice would make a difference. I’m not even sure you’d need NFP for that case.
But, if someone would have a base rate of, say 3x a week, there’s no sex during the fertile period. (Note for non-Catholics: Catholics aren’t permitted to “take care of things” via other methods permissible to most other Christians — even Pat Robertson (again, another post).)
Now, either the couple makes up for lost time or they don’t. If they make up for lost time, the odds of having sex druing the time when an egg could accidentally be fertilized and then miscarry go up. If they don’t make up for lost time, there’s still the high probability that they will have sex either right before or right after the “lost time,” again increasing the risk of an accidental conception and micarriage.
You’re making no sense. Either you follow the rules of NFP or you don’t. If you’re TTA, don’t have sex on days when you might be fertile. If you’re ambivalent, have sex whenever. If you’re TTC, either have sex whenever or time it optimally if you need to (such as in the case of low sperm count; cf Taking Charge of Your Fertility).
The vast majority of “failures” for NFP are the result of couples throwing caution to the wind and being “risky” while TTA. If a couple knows they are prone to miscarriages, they ought to be more careful about when they have sex. If they’re TTA, they should allow for a longer potential fertile period, even if it means some false positives and fewer days for sex. If they’re TTC, they should do the same, but then have intercourse on the day before or the day of ovulation.
The “making up for lost time” stuff is utter nonsense.
First, failure to follow the rules exactly is as much a failure mode for NFP as it is for condoms. If failing to use a condom properly can be used in the condom failure rates, than so can failure to use NFP be counted in it’s failure rate.
I’d bet that those kind of failures are far more common than “throwing caution to the wind.” Given that the average IQ is 100, I’d be surprised if even a half of the population could follow NFP correctly.
NFP will also fail occasionally, even if it is followed exactly. Determining the fertile period is not exact. Surprises can happen, such as ovulation occurring much later than would have been indicated according to the charts.
If an absolutely positively exact method of determining fertility were available, would it even be permissible for Catholics?
The “making up for lost time” is not utter nonsense. If a woman’s chart says that she will have ended her fertile period on Tuesday and she has sex on Tuesday, if the chart has failed to predict the fertile period accurately, then you are at greater risk for a late conception than if the couple waits until Friday.
And, as far as making up for lost time…I’d say it’s not nonsense. I’d say it’s fun. Granted, our lost time was because of conflicting work schedules and forced overtime, but the point remains.
Rob, I’m still confused about your point with the post. In comment 15 you elaborate:
Interesting hypothesis, but not one tested or supported by this study.
Again, interesting hypothesis, but still not one tested or supported by this study.
The study merely concludes that conceptions taking place during non-optimal time frames are more likely to lead to miscarriages if the female has a prior history of miscarriage. This fact has nothing to do with NFP, this study makes no claims to compare rates of miscarriages of NFP to non-NFP couples, nor does it examine whether the dates of conception for NFP couples fall outside the fertile time window identified by the method.
In comment 17 you go on to talk about how not following NFP as designed should be counted as a “failure”, that NFP itself can “fail”, that “making up for lost time” can be legitimate, and questioning whether a perfect fertility system (if it existed) would be allowed for Catholics. While I may have things to say, these all seem to be tangents.
Back to the point: How exactly does this study make a ‘disturbing discovery’ about NFP?
[...] For the record, I came up with my analysis independently. I quoted some research that actually shows this happens. I pointed out that this related directly to the argument about “Plan B.” Funky Dung, of Ales Rarus, correctly suggested that this will require a modification to NFP. [...]
[...] Now we must ask if Bovens third assumption, about HF and fringe conceptions having different embryo survival rates, is right. According to this article ("Timing of conception and the risk of spontaneous abortion among pregnancies occurring during the use of natural family planning", AM J OBSTET GYNECOL 1995; 172:1567-72.) in the American Journal of Obstetric Gynecology, it is not. (Fedora Tip: UnSpace) [...]