Baby’s DNA in Mom’s Blood: Noninvasive Prenatal Testing

Until recently, the moment of birth was a surprise. We anxiously awaited the obstetrician’s announcement: “It’s a boy!” or “It’s a girl!” Then we checked to see if any crucial parts were missing and we counted the fingers and toes. We had to wait for a baby to be born before we could know its sex and whether it was normal. Today, thanks to prenatal testing, we can know the sex of a fetus, diagnose a number of genetic abnormalities and malformations, and we can even operate on the fetus in utero to correct certain problems before birth. I had amniocentesis for my two pregnancies because of the higher risk of Down syndrome at my age (37 and 39). It was reassuring to know the baby didn’t have Down syndrome, and it was fun for my husband to point to my burgeoning belly and introduce it to people as “our daughter Kristin.”

Amniocentesis is invasive, carries risks, and can’t be done until the 15th to 20th week of pregnancy. Now there is a safe, noninvasive, accurate blood test that can be done as early as the 9th week. It analyzes cell-free fetal DNA (cfDNA) circulating in the mother’s blood. It sounds ideal, but there are some caveats. It’s not yet appropriate to recommend it to all pregnant women. An editorial in The New England Journal of Medicine expressed concern that pressures are promoting diffusion of cfDNA testing beyond the boundaries of available evidence.

What tests are available?

  1. Standard screening tests on maternal blood for serum markers associated with chromosomal abnormalities (like the trisomy of Down syndrome) and neural tube defects (like spina bifida). Indicates elevated risk but doesn’t make a diagnosis. Further invasive testing required.
  2. Ultrasound.
    1. Nuchal translucency, measured at end of first trimester
    2. Nuchal fold, measured at end of second trimester
    3. Can also visualize gross anatomical abnormalities
  3. Combining serum marker tests with ultrasound nuchal measurements can decrease the number of false positives to 4-5% with a true positive rate of up to 90%. Not diagnostic: indicates a need for further invasive testing.
  4. Chorionic villus sampling (CVS) samples placental tissue.
    1. Can be done at 8-10 weeks gestation
    2. 2 approaches: through the vagina or the abdomen
    3. Risks: miscarriage (0.5-1%), infection, amniotic fluid leakage, limb reduction defects
    4. False positives can occur with mosaicism, where the placental cells are abnormal but the fetus is normal, and false negatives can occur if the sample is contaminated with maternal cells
    5. If the fetus is Rh positive and the mother is Rh negative, the procedure can introduce fetal cells into the maternal circulation and cause Rh sensitization
  5. Amniocentesis.
    1. Needle inserted into amniotic fluid
    2. Done later than CVS, between 15th and 20th week of pregnancy
    3. Risks similar to those of CVS, but miscarriage rate is lower (estimates vary from 1 in 200 to 1 in 500)
    4. Can also predict fetal lung maturity
  6. Percutaneous umbilical cord blood sampling
    1. Needle inserted through abdomen into baby’s umbilical cord under ultrasound guidance
    2. Can detect chromosomal abnormalities, infections, blood disorders and some metabolic disorders
    3. Can detect fetal anemia and treat with blood transfusion
    4. Done after 17 weeks pregnancy
    5. Risks: similar risks to amniocentesis but with a higher rate of miscarriage (1-2%)
  7. cfDNA tests
    1. Noninvasive: require only drawing blood from the mother
    2. Can be done as early as 9 weeks
    3. High sensitivity and specificity

What can prenatal testing diagnose?

All the invasive tests and the new cfDNA tests collect fetal DNA and analyze it. They all require a 1-2 week wait for results. They can detect chromosomal abnormalities include Down syndrome, other trisomies, sex chromosome abnormalities like Turner syndrome, and fragile X syndrome. Specific tests can be done for cystic fibrosis, Tay-Sachs, sickle cell disease, muscular dystrophy, and other inheritable conditions when indicated by family history.

What if results are abnormal?

If a condition like Down syndrome is detected, about 90% of parents choose to terminate the pregnancy. For some conditions like spina bifida, fetal surgery may be an option. Parents who continue the pregnancy can use the information to plan for the care of a child with special needs, anticipate lifestyle changes, and identify support groups and resources.

Some people reject any testing because of religious or moral objections to pregnancy termination, because they are comfortable with the idea of raising the child no matter what handicaps it has, or because they reject any testing that poses any risk to the fetus.

Currently available cfDNA tests in the US are offered by 4 companies (Verinata, Sequenom, Ariosa, and Natera) at a cost of $795 to $2762. They all screen for trisomy 21 (Down syndrome), trisomy 18 (Edwards syndrome), and trisomy 13 (Patau syndrome). Some additionally offer sex determination and screening for sex-chromosome abnormalities like XO, XXX, XXY and XYY. They do not test for anything else. They are being aggressively marketed though channels like Facebook, and consumers may not understand the trade-offs and the limitations.


Patients considering cfDNA testing should be aware that:

  • They do not provide all the information available through other tests
  • They are generally not covered by insurance
  • They are said to be accurate, but the reported sensitivities and specificities of these tests are based on studies with a high proportion of abnormals (with a prevalence as high as 1 in 8 for Down syndrome). They will be less accurate when applied to the whole population of pregnant women.
  • If the test is positive, what’s the likelihood that the individual really has the condition? In other words, what is the test’s positive predictive value (PPV)? We don’t know, because the companies have not reported that information. Sensitivity and specificity depend on the prevalence of the condition in the population being tested; we want to know the PPV for an individual in the specific population we are testing. In a low risk population a positive result will have a lower PPV than in a high risk population.

ACOG recommendations

The American Congress of Obstetricians and Gynecologists and other professional organizations recommend cfDNA testing only for high-risk pregnancies, and specify that abnormal results should be confirmed by invasive testing before any action is taken.


Who would have imagined that fetal DNA was circulating in the mother’s blood? Or that it could be separated from her blood to allow analysis of the fetus’s genetic makeup? This is an exciting scientific development. These new cfDNA blood tests have great promise, especially if they can be expanded to include broader screening of DNA for more abnormalities. Meanwhile, the currently available tests are only able to detect a few specific abnormalities and they shouldn’t be used as a screening test for all pregnancies. They are useful in high-risk pregnancies but abnormal results must still be confirmed by invasive testing.

Posted in: Diagnostic tests & procedures, Obstetrics & gynecology

Leave a Comment (25) ↓

25 thoughts on “Baby’s DNA in Mom’s Blood: Noninvasive Prenatal Testing

  1. Alia says:

    What a coincidence, just yesterday I was browsing my practice’s website (looking for information on whether they already have flu vaccine available) and found an ad for this test. Not that I’m going to need it any time in the nearest future, but still, thanks for your article, dr Hall.

  2. cphickie says:

    The ability to do newborn genetic disease testing in utero would be very medically useful (provided the technique can be expanded to screen for more abnormalities at a lower costs). What concerns me (and I could be completely wrong about this concern) is whether this before-birth information on the baby’s genetic makeup could be used by insurance providers to deny coverage for the baby once it is born– based on the argument that until the baby is born, the baby is not legally a person.As an example, if the in utero genetic screening shows the baby to have cystic fibrosis, could an insurance company, before the baby is born, tell the parents that their baby will not receive health insurance coverage and be able to do so because legally this might not be considered a pre-existing medical condition because the unborn child does not legally exist as its own self until after birth?

    1. Jann Bellamy says:


      ” . . . could an insurance company, before the baby is born, tell the parents that their baby will not receive health insurance coverage and be able to do so because legally this might not be considered a pre-existing medical condition because the unborn child does not legally exist as its own self until after birth?”

      I doubt it, because the baby’s insurance coverage depends on his status as an “insured” under the policy, not his existence as a person under the law. Insurance policies can differ in their coverage, but I’ve never heard of one that doesn’t cover a baby “as is” when he is born, regardless of when he developed the condition or when it was discovered. As Dr. Hall pointed out, we are already able to determine that a baby has Down’s syndrome prior to birth and insurers have not denied coverage on that basis.

      As a practical matter, an insurance company would face a PR disaster if it tried to deny coverage on this basis and if there were some gap in the law that allowed this to happen Congress would swiftly move to amend the Affordable Care Act to fill the gap.

    2. Chris Hickie says:

      What I’ve been trying to find (but haven’t) are any cases/reports of parents being denied coverage on their newborn based on something that was found through amnio…because that would be, I think, a precedent for this happening with the non-amnio DNA analysis. I’ve heard that they have shown the ability to isolate enough fetal DNA from maternal blood to do the newborn screening tests that you currently get in the first 2 weeks after your baby’s birth (typically done by a heelstick for about 5-10 drops of blood). I guess if there aren’t any precedents for this from amnio, it shouldn’t be a big deal for this newer method.

  3. windriven says:

    “[C]ould an insurance company, before the baby is born, tell the parents that their baby will not receive health insurance coverage and be able to do so because legally this might not be considered a pre-existing medical condition?”

    One of the advantages of ACA is that it generally bars insurers from refusing coverage for pre-existing conditions. Of course it remains to be seen how that will play out in practice.

    But taking that notion a step farther, could a country with a fully socialized* health care system mandate testing of this nature and in the event of serious congenital problems pressure the parents to abort? Should it? It seems to me that a powerful argument could be mounted favoring that approach for conditions that preclude any meaningful quality of life, anencephaly for instance. I could mount an equally powerful argument that such actions would be the first step toward eugenics**.

    *This is not a closeted attack on socialized medical systems, several of which offer arguably the best care on the planet. It is simply a thought experiment about control over choices for which someone else pays the bill..

    **And I think that anyone who thinks about the issue carefully will accept that some form of eugenics will shape – in a few cultures already shapes – humanity’s future though we find the idea generally abhorrent today.

    1. gewsin says:

      Good for you, for bringing up the obvious ethical challenges we will face. If people refuse to face such questions, they will be decided by someone else when they can no longer be ignored. A republic is not designed to address such issues easily. A group such as the National Science Foundation or NIH could be charged with responsibility to educate what such topics are and develop curricula for HS students, hold townhall meetings on such topics, and even have programs to create public ed and adult ed teachers in how to elicit thoughtful responses from citizenry. Much like Oak Ridge Nat’l Labs produced teachers/lecturers on energy science for school assemblies across the country. (The best assembly presentations of my young LIFE!)

      1. windriven says:

        Addressing these issues requires people to come face-to-face with some very uncomfortable choices. Much easier to let their surrogates (churches, political parties, etc.) do their thinking for them. Then all they have to do is wear the t-shirt.

        The current ‘debate’ over ACA is a case in point. Those on the left embrace it as nirvana with whipped cream and a cherry – precluding any serious effort to fix some of its more glaring defects. Those on the right decry it as the end of civilization as we know it complete with death panels and the specter of Barack Obama snooping through the details of your sex life – precluding using ACA as a reasonable starting point for meaningful improvements to our badly broken health care delivery system.

  4. Sullivanthepoop says:

    With my last pregnancy I actually got a false result on one of these tests for Trisomy 13. I didn’t have an amnio because both the doctor and the genetic expert said that said the likelihood she actually had trisomy 13 was less than risk of miscarriage from an amnio and something like trisomy 13 would not be missed in an ultrasound. Anyway, my daughter didn’t have any problems, definitely not trisomy 13.

  5. anthro49 says:


    *If (most) everyone contributes to the system through taxes, then the “someone else” paying the bill is also (most) everyone. Isn’t that the whole idea of socialism?

    **I hesitate to say this, but I think eugenics has got somewhat (emphasize the somewhat) of an unfair reputation. The concept isn’t the problem so much as the ignorance and prejudices of the times in which it was practiced. I think you are correct that some form of what might be termed “eugenics” will eventually be part of most cultural systems. We have been doing so in a limited way since the advent of the testing Dr. Hall describes here.


    Interestingly, Dr. Hall, I did NOT have amnio at 37. Given that the preliminary blood test was completely normal, the OB said it was a gray area for 35 – 40 age group. I decided not to do it as I had a history of miscarriage (not that that was related to amnio-induced miscarriage, but I’m not sure I thought it through like that in those days). I also never wanted to know the sex ahead of time–talk about UN”natural” :-)

    1. windriven says:


      “*If (most) everyone contributes to the system through taxes, then the “someone else” paying the bill is also (most) everyone.”

      Exactly. Health care is (in the economic sense) a limited resource. Is it ethically justifiable to spend large sums of money on the medical treatment of individuals with no hope of meaningful quality of human life? Remember that as a limited resource, the long term care and support of an anencephalic reduces the available resources for some other patient(s).

      But then one might take the position that society might choose to provide only limited palliative care to those appearing with one of the degenerative motor neurone diseases. There are no cures and it is arguable that life as a lump of inert flesh has little intrinsic meaning. But then what of Stephen Hawking?

  6. Bruce says:

    Testing alone does not convey wisdom. If these types of screenings become commonplace, it is essential that those receiving screenings (and those administering them) understand the conditions that are revealed. As the sibling of a middle-aged person with Down Syndrome, and the parent of a now-deceased child who had cerebral palsy, I would say that my sister with Down Syndrome has had a very good quality of life, and she contributed to the quality of life of our family (although not in the way that my parents had initially envisioned). It was much more difficult to deal with a child who had severe cerebral palsy, but it is not a condition that can be generally revealed in screening tests (and typically is not present until just before, during, or shortly after birth). There is a wide-spread assumption that Down Syndrome is a horrible diagnosis. While it does come with some challenges, it is not the terrible problem that many seem to assume. Many conditions are much more challenging and many of those (e.g. severe autism) cannot be currently diagnosed. So a positive screening result for Down Syndrome may not be the bad news it is assumed to be, and a clear screening may not be deserving of the relief that many prospective parents feel.

    1. fishchick says:

      I agree. I am currently pregnant with my second, and being of “advanced maternal age” this time around, it’s been trisomy, trisomy, trisomy. My first child has autism, so just by the numbers, I have a higher chance of having another child with autism than I do/did a trisomy… but there’s no test for that, so nobody’s even brought it up.

  7. angorarabbit says:

    By coincidence, today’s NY Science Times has an article about how chimeric many people actually are, and that a blood sample may reveal DNA of their own mother (moving from mother to fetus) or one’s past children (as per above). I first heard about this two yrs ago, in a talk from the head of Baylor’s genetic screening program. Using the new sequencing technologies that require vanishingly small DNA quantities, they find that parents of children with genetic birth defects were sometimes highly chimeric and the extent depended on when developmentally the mutation emerged vs. the segregation of the germ cells. Layering onto this was the realization of maternal-fetal cell transfer across the placenta, the most egregious example being (IIRC) golden tamarins where males can actually posses a surprising percentage of maternal somatic cells.

    All this long-winded comment leads to my observation that, as cfDNA testing expands, it will be interesting to see how much cross-cell populations will be present, even in people who have never been pregnant.

    Also, I can’t help but wonder aloud if some of these cross-contaminating cells, as they persist, may contribute to immune disorders and perhaps chronic fatigue syndrome in both men and women. (Because the cells can cross both directions). Sheer speculation, of course, but if I were starting over again it’s a question I’d consider investigating.

    1. Young CC Prof says:

      You know, I believe some folks were looking into the role of chimeras (mostly child cells in mothers) in causing autoimmune diseases. It is known that a lot of autoimmune diseases are most commonly diagnosed in women of later reproductive years or early middle age…

      1. angorarabbit says:

        Yeah, I should look into it some more. I think it was a SciAm article a few years ago. The demographics are right, but then correlation is not causation. :)

  8. Kathy says:

    @angora, I’m also wondering how long foetal DNA persists in the mother. Could it be possible that such a test would pick up DNA from a previous child?

    1. That is exactly what I am wondering now, faced with having this test done on me next week!

  9. Allan says:

    Less “serious” matter than blood testing and you might already know about the babyCOmpact CO monitor by Bedfont but as I find it particularly easy to use and pretty reassuring for a regular prenatal measuring of the amount of carbon monoxide…thought I’d share!

    1. WilliamLawrenceUtridge says:

      Is there any reason to expect this comment to be naught but spam? Is there evidence that infant CO poisoning is any more of a threat than adult CO poisoning, or that special monitors are needed above and beyond what a house should anyway?

  10. Rosemary says:

    “If the test is positive, what’s the likelihood that the individual really has the condition? In other words, what is the test’s positive predictive value (PPV)? We don’t know, because the companies have not reported that information. Sensitivity and specificity depend on the prevalence of the condition in the population being tested; we want to know the PPV for an individual in the specific population we are testing. In a low risk population a positive result will have a lower PPV than in a high risk population.”

    Right. But even though I’m low-risk, I had the cell-free DNA test because, given the best information available, its PPV would MUCH higher for me than any other test I would be routinely offered.

    Obviously I understand that we don’t have the exact numbers, but is there really any doubt that it’d be more accurate than the quad screen? Which no one would bat an eye at me getting? (It’s routinely offered to low-risk mothers despite its crummy, crummy PPV.)

    Anyway. After talking to an OB who specializes in prenatal screening, I declined all fetal testing (AFP/quad screen, first trimester ultrasound, etc) except for the cell-free fetal DNA test and the 20-week A/S. So far, so good . . .

  11. HReece says:

    I read where you said only 4 companies do these prenatal testings and listed all they tested for and noticed that non invasive paternity testing wasn’t included in that. I had a prenatal paternity test done through DDC, who actually sends the samples for testing to be done at Natera in California. I’ve been concerned that the test results were wrong. Should I trust the results? I’m due in less than 50 days and am nervous the results may be wrong.

  12. patp says:

    Im 45 years old I had a IVF with a egg donor(19 years old,my daugher) we are pregnant my first U.s with my OBG was perfect at 10 weeks. at 12 weeks I ask for th NT ultrasound at the perinatal center .the actual NT was not done because they said could not get a clear picture but they said the shape of the baby head looks pointed. i have another u.s 2 weeks after at 14 weeks with my regular OBG he said the head shape looks normal to him he wants me to keep the follow up with the perinatal center now at 14 weeks i was told someting doesnt look normal at the baby brain top of the head like to much fluid which the doctor said is normal for the 15 weeks but taking in consideration the shape of the head and he cant see the nose he wants me to do a blood work he is thinking trisomy 13. we pay extra $ to have the blood sent to a lab in CA that supposs to be give more accurate result that LAB CORP after 10 agony days the result are back negative for down sindrome,t18 and t13. now my question is what is next?? can i related on the blood work or be hunted on what the doctor think he sees on the Us at 12 and 14 weeks? I have 2 diferente opiniosn form my OBG who has 28 eyar in the field and a docotr at the perinatal cneter who My IVF doctor reconazie as beeing a good one, now the blood work is negative

    1. Harriet Hall says:

      I can’t advise you about what to do. I explained the reliability of the various tests in the article; perhaps it would help to re-read it. You can discuss it with your doctors: perhaps you can get them to discuss it between themselves and resolve the conflict, or get a 3rd opinion.

  13. patp says:

    I need to mention that the bloodwork that was done call noninvasive prenatal or NITT. How accurate can this overrode the U.s?

Comments are closed.