Testing Down Syndrome to Death

From the Winter 2013 issue of the Human Life Review.

At one time or another, all of us have fallen for a lie that we wanted to believe. Have you heard that nine out of ten babies with Down syndrome are aborted?

I first came across this shocking statistic in a New York Times article published in May 2007, just two months after my daughter Magdalena, who has Down syndrome, celebrated her first birthday.

In the years since, I have seen it repeated in countless articles in respected publications, in blog posts and in interviews, in every corner of the ever-expanding media universe. The influential conservative writer George Will referenced the claim last year in a lovely piece about his son John’s 40th birthday. The New York Times columnist Ross Douthat also used it last year in an article about eugenics. The claim shows up all the time in social media. I have used it myself.

But is it true? Winston Churchill said that a lie gets halfway around the world before the truth has a chance to put its pants on, and this 90 percent claim has certainly done some impressive travelling.

As someone who is functionally innumerate, I am rarely inclined to question the veracity of percentages and statistics when I encounter them in print. But if decades of watching disingenuous ideologues arguing on cable television have taught me anything, it’s that statistics are routinely manipulated to support the partisan preferences of those who wield them. There are, as they say, lies, damned lies, and statistics. Those of us who feel called to change hearts and minds on issues of great personal and societal importance are obliged to reach for a higher standard than mere reliance upon statistical claims of dubious provenance. We just might profit from examining and upending our assumptions now and then to see if they are indeed sound. We just might want to do this even if it leads us to question those claims that we hold dearest—the ones that we think we really need to be true.

The claim that 90 percent of children with Down syndrome are aborted has served the pro-life movement well. It is so shocking that it rarely fails to get people’s attention. If true, it surely ranks as one of the most horrifying statistics ever calculated. If true, we should shout it from mountaintops and, if need be, send it halfway around the world.

But only if it’s true.

I put blind faith in the 90 percent statistic because it was useful to me, and it was useful precisely because it was so shocking. If it’s not true, where does that leave those who have come to believe in its symbolic value and rhetorical power? Where does it leave me?

“The 90 percent number is only shocking if you’re the kind of person who thinks babies with Down syndrome should be born,” says Amy Julia Becker, author of the just-released e-book on prenatal testing and Down syndrome, What Every Woman Needs to Know About Prenatal Testing: Insights from a Mom Who Has Been There. I sought her opinion on the 90 percent statistic because a column she wrote last year on the subject got my attention. In it she observed, “The number gets thrown about all the time: 90 percent of babies with a prenatal diagnosis of Down syndrome are aborted. Often it’s reported in shorthand, 90 percent of babies with Down syndrome are aborted. The latter statement is patently untrue. The former is somewhat inaccurate.” Inaccurate? Patently untrue? To someone who has traded on this statistic as often as I have, reading this was bad news indeed. I decided to do my best to get to the bottom of this 90 percent business.

Like it or not, the Pulitzer Prize-winning website Politifact.com has established itself in recent years as the first port-of-call for those seeking to establish the veracity of controversial claims made in the public square. So that’s where I turned.

In 2011, Politifact’s fact checkers gave a “true” rating to Florida Republican state representative Richard Corcoran’s claim, made during a debate over abortion legislation, that “90 percent of babies with Down syndrome are aborted.” However, it only gave a “half true” rating to former Senator Rick Santorum in 2012, when the Republican presidential candidate told CBS’s Bob Schieffer that “90 percent of Down syndrome children in America are aborted.” Why did the latter claim merit only a “half true”? Because, as it turns out, there has never been a comprehensive national study done on the rate of abortion in the United States due to a prenatal diagnosis of Down syndrome. Santorum’s addition of the words “in America” to the 90 percent claim, in Politifact’s eyes, made it somewhat less true.

I began sorting through some of the scientific literature provided to me by Dr. Brian Skotko, co-director of the Down Syndrome Program at Massachusetts General Hospital. A medical geneticist by training, Skotko is beloved by those in what is sometimes called “the Down syndrome community” because, unlike many of the cold-hearted clinicians that we frequently encounter, he is not only what you might call sympatico, he is in fact one of us; Skotko’s sister Kristin has Down syndrome.

The good doctor began by referring me to a 2011 study by James F.X. Egan, a now-deceased obstetrician from the University of Connecticut who compared the expected number of Down syndrome births in various demographic groups in the United States to the actual number of such births. While Egan’s model predicted that 122,519 children with Down syndrome should have been born in the United States between 1989 and 2006, only 65,492 of those children actually were born. Egan graphed the divergence of the expected birth and actual birth trend lines along with the dates when certain prenatal tests were introduced.

Over time, the number of expected births of babies with Down syndrome has risen sharply. This is due, in part, to the well-understood role of maternal age in determining whether a fetus with Down syndrome will be conceived.

As Egan explained, “[t]he child-bearing population in the United States has become progressively older over the past 25 to 30 years,” leading to an overall increase in the absolute number of children born with Down syndrome. Simply put, for a variety of reasons, more and more women are putting off having children until their mid-to-late thirties or even their early forties. This trend has dramatically expanded the population at high risk of conceiving a child with Down syndrome. Despite this expansion, however, the number of actual live births of babies with Down syndrome has declined modestly, opening up a yawning gap in the graph representing the children with Down syndrome who should have been born over the last 25 years but weren’t.

Where are all those kids? Why weren’t they born? Sadly, the questions answer themselves. Thousands upon thousands of babies with Down syndrome were aborted between the late 1980s and 2006. And the abortions increased at an increasing rate as new and more accurate screening instruments became available.

“As new prenatal tests were introduced, more women found out about the diagnosis prenatally and therefore more women had the option of deciding whether or not to continue a pregnancy. And more women chose to take the option of elective termination,” Skotko told me.

I want you to hold that thought.

The Egan paper established an important precedent in my search for the truth regarding the 90 percent stat: It is verifiably true that a monstrous number of abortions are being performed on women who have been told that the fetus they have conceived and are carrying has Down syndrome. Exactly how many abortions have been performed is probably impossible to know, as record-keeping requirements vary from state to state and change over time. As a result, I realized to my dismay that the rate at which these abortions are performed is also probably impossible to know, though several limited studies have been performed in an attempt to find out the answer.

The 90 percent claim was given life in 1999 by a literature review conducted by Caroline Mansfield and her coauthors at the Psychology and Genetics Research Group at King’s College in London. The review considered the results of 20 other studies measuring abortion rates in countries around the world between 1980 and 1998 and concluded that an average of 92 percent of women who received a prenatal diagnosis of Down syndrome chose to have an abortion. Indeed, this is the study that Politifact referenced when giving its “true” rating to Florida State Rep. Richard Corcoran.

I was initially relieved to discover that the 90 percent stat could be traced to a legitimate, non-ideological, and scientific study. However, a close examination of the Mansfield paper revealed certain obvious limitations—obvious even to this untrained author’s eye. Although the authors reviewed fully 20 single-country studies, most of these were conducted using what can only be called extremely small sample sizes. For example, the results of one study reviewed by Mansfield purported to show an 80 percent rate of abortion of fetuses with Down syndrome in Singapore. In reality, however, the researchers examined a mere five cases of women who had received a positive Down syndrome diagnosis. Four of these chose to abort, hence the rate of 80 percent. A 1995 French study recorded that 76 out of 76 women who received the diagnosis aborted their babies. That’s 100 percent, but in a nation of 65 million people, how much can we really extrapolate from the experiences of 76 people?

Mansfield only looked at three studies from the United States. The largest of these, conducted in 1980, followed 2,500 second-trimester amniocenteses, 19 of which resulted in a positive diagnosis of Trisomy 21, the clinical name for the condition known as Down syndrome. Of those 19 women receiving the positive Down syndrome diagnosis, 18 chose to abort their babies, a rate of 95 percent. A second U.S. study, from 1985, saw 42 out of 43 women abort and a third, from 1988, recorded 13 abortions out of 15 positive diagnoses. Each of these studies suffered from what Mansfield called a lack of “sufficiently large sample sizes to enable reliable estimations of termination rates.”

But Mansfield also reviewed a 1998 paper by D. Mutton, et al. analyzing trends in England and Wales—a paper that stands out both for the size of the sample it examined and the scope of years it covered. Between 1989 and 1997, Mutton tallied 4,438 abortions out of a total of 4,824 confirmed prenatal Down syndrome diagnoses for a rate of 92 percent. These numbers dwarf those in any of the other studies looked at by Mansfield and provide a significantly more reliable basis for extrapolating conclusions.

More recently, Jaime Natoli, et al. reviewed seven population-based studies conducted in California, Hawaii, and Maine between 1995 and 2011. (The term “population-based” means simply that the researchers attempted to study as representative a cohort as possible.) According to Skotko, a population-based study attempting to determine rates of abortion due to Down syndrome must account for the reality that a mere 2 percent of all pregnant women undergo either amniocentesis or CVS (chorionic villus sampling), the only definitive diagnostic tests for Down syndrome currently available.

“It’s not a question of ‘What do all women do?’ because 98 percent of expectant mothers do not get any sort of definitive invasive testing,” says Skotko. “Only a portion of the 2 percent who opt for testing will be told that their fetus has Down syndrome. The question really is: What percentage of those will go on to terminate their pregnancies?” The answer, according to the studies reviewed by Natoli, is between 61 and 93 percent, with the average at 74 percent.

So here we have our first real, concrete statistical divergence from the oft-cited 90 percent statistic. The question that immediately leaps to mind is: Should we simply revise downward the 90 percent rhetoric to say that 74 percent of babies with Down syndrome are aborted? Not so fast.

Natoli also looked at nine “hospital based” studies conducted in six states (CA, CT, MA, MI, NY, SC) and the District of Columbia between 1995 and 2011. As the authors note, hospital-based studies can show great variation depending on various demographic factors that make comparisons difficult. For instance, the authors write, “The number of pregnancies with a prenatal diagnosis of Down syndrome ranged from ten pregnancies over a 2-year period at Georgetown University Hospital (2002–04) to 449 pregnancies over a 20-year period at the University of California at San Francisco (1983–2003).” The rate of abortion due to Down syndrome observed by these studies ranged between 60 and 90 percent, with the average at 85 percent.

I began to suspect that my initial reservations about tangling with numbers and figures were not unfounded. A decidedly non-Churchillian warning came to mind: Mess with the bull—get the horns. The further I dug into these studies, the messier the picture got. I was looking for one number—90 percent—and ended up with at least three that I had to take seriously—74, 85, and 92 percent. Which was the real statistic? Which was the answer I was looking for? My time spent with the scientific literature on the subject seemed to confirm nothing more dramatic than the less-than-useful claim that the rate of abortion due to a prenatal Down syndrome diagnosis, depending on where you live, is somewhere between 61 and 100 percent.

While my faith in the 90 percent claim was quashed, what replaced it was an even more troubling insight about the future of Down syndrome, prenatal testing, and abortion. As someone who has lived the bulk of his life in the progressive bastion of the Northeastern United States, with only a brief detour to the progressive bastion of Southern California, experience tells me that abortion has become a socially acceptable response to a prenatal diagnosis of Down syndrome and other genetic disorders. It’s not just that there’s no stigma attached to the notion. In certain circles, having an abortion because you’ve been told your child will have Down syndrome is not a matter that should prompt even the most routine ethical reflection. It’s just what you do.

So Amy Julia Becker is probably right. The claim that 90 percent of babies with Down syndrome are aborted is shocking only if you think babies with Down syndrome should be born. If you don’t accept that premise, no percentage or statistic, no matter how gaudy, is likely to shock you. If you believe that abortion is always justified so long as it is the free choice of the mother, why would you pause to raise an eyebrow at the number of babies with Down syndrome who are aborted? I can’t think of a single reason.

So who exactly are we trying to shock when we repeat the claim that 90 percent of babies with Down syndrome are aborted? Who exactly have I been trying to shock? Those who already believe, as I do, that all life is precious? Am I trying to shock myself? Why? What good will that do? My heart and mind have already been changed. Becker has concluded that the people we should be talking to—the people we should be trying to convince to join us in the quest to reduce the number of abortions due to a prenatal diagnosis of Down syndrome—are precisely that large cohort of women who, for one reason or another, have given birth to a child with Down syndrome without having received a prenatal diagnosis. As Becker told me:

For people who have categorically decided that they want prenatal testing for the purposes of selective abortion, I certainly want to have a conversation in which I argue for the value of children with disabilities, but I recognize that it’s a hard argument to make given our different presuppositions. And for women who wouldn’t abort under any circumstances, I also want to offer encouragement, support, and hope. But it’s that large middle group—the women who don’t know what they should do or would do—that I most want to address. And it is those women who need to know that there is not one prevailing attitude or choice when it comes to prenatal diagnosis and Down syndrome in this country.

This cohort seems ripe for conversion to the pro-life cause. These mothers may have been pro-choice before their children were born. They may still think of themselves as pro-choice. But odds are that they are now more sympathetic to the argument that a child should not be killed in utero simply because that child has Down syndrome than they were before they became the parents of beautiful, charming, and unambiguously human children.

Becker, also the parent of a daughter with Down syndrome, surmises that about half of all children conceived with Down syndrome actually make it to a live birth. This includes babies born to mothers who know their baby will have Down syndrome as well as babies born to mothers who, for one reason or another, don’t know. But just as with the 90 percent claim, this 50 percent claim may or may not be true. As noted earlier—it’s almost impossible to get accurate and useful information on the number and rate of abortions in this country. It’s even harder to get information on precisely why a woman had an abortion.

“This is an area where we just don’t have a lot of accurate data,” concedes Skotko. “The best we can do is to look at the studies that have been done well and say, ‘This is what we know. This is what we don’t know.’ And in the realm of what we don’t know, everyone can make their own educated guesses about how to fill in the blanks.”

Which is not as easy as or as satisfying as it seems. My motivation for launching this investigation in the first place was to try to get beyond the habit of making educated guesses—often informed by little more than my own biases—and get closer to something approaching the truth. For the sake of clarity, let me restate my bias: The claim that 90 percent of babies with Down syndrome are aborted sounds about right to me. But I have reluctantly accepted that it is unsupportable. The vast majority of women in this country who give birth to a baby with Down syndrome have not received a definitive prenatal diagnosis of their baby’s condition, either because they opted out of the screening process or because the screening process failed to recommend invasive testing. Recall that only 2 percent of all pregnant women undergo amniocentesis or CVS in the first place, and these are the only tests currently available which can give a definitive prenatal diagnosis of Down syndrome. The tests just aren’t administered widely enough to come anywhere close to identifying nine out of ten babies with Down syndrome.

Let me rephrase that: They aren’t administered widely enough yet. Remember when I asked you to hold that thought about all the “missing” kids with Down syndrome? This is why: Four noninvasive and highly accurate prenatal tests for Down syndrome that can be given as early as the tenth week of pregnancy have recently come on the market. These tests are only available on a limited basis now, but they will soon take their place in the standard battery of routine blood tests and ordinary lab work that a woman in the early stages of pregnancy undergoes as a simple matter of course. When they do, they will render this entire discussion moot.

In the current environment, where only 2 percent of pregnant women are given a definitive diagnostic test, it may be accurate to say that 50 percent of babies conceived with Down syndrome are born. But the current environment will shortly be transformed beyond recognition. Soon, any woman who wants to know, will know. Even women who don’t want to know will probably end up knowing because of the entirely human tendency to go along with doctor-recommended testing without any real consideration of the possible consequences. And the available evidence, such as it is, points to only one conclusion: An astonishing percentage—probably well north of 75 percent—of women who find out that their baby has Down syndrome abort that baby. The more women who know, the more will abort.

The only percentage that should matter at all in the context of Down syndrome and the abortion debate is the one that provides an answer to the following question: How many women who know they are going to have a baby with Down syndrome choose abortion? For we can reasonably expect that when doctors begin offering pregnant women highly accurate and noninvasive tests on a large scale there will no longer be any surprises. There will no longer be a large cohort of women who found out only after delivery that their babies had Down syndrome. There will no longer be an identifiable group of pro-choice parents who unexpectedly found themselves falling in love with a child conceived with an extra copy of the 21st chromosome and so are ripe for conversion to the pro-life cause. These new tests will eliminate 100 percent of such families.

And that’s a statistic you can put some serious faith in.



  1. Mary Sullivan says:

    Hello Matthew, I am the mother of a beautiful little girl with Down Syndrome. She is the love of our lives…. our 9th child. Thank you for your article, it was very moving. I am also a registered nurse and a graduate student at Arizona State University. In 2 days I will be doing my last presentation on this very subject and this will complete my program. I am always uplifted by people who write about the injustices being done to Down Syndrome fetuses. God Bless You!!! Mary Sullivan, Chandler, Az

  2. Great post Matthew. Here is an article I wrote about the international situation, it also covers the US figures breifly. The situation is indeed very serious now.

%d bloggers like this: