Recurrent Miscarriage (Pregnancy Loss)


As soon as a pregnancy becomes recognized, each (prospective) parent generally starts to accept and plan for their new arrival. If the pregnancy is lost, this is often considered a "death within the family" and the couple will go through an intense grieving process. The loss of a pregnancy can be devastating for a couple, regardless of the number of children in the family or the cause for the loss.

Recurrent Pregnancy Loss: grieving process

As soon as a pregnancy becomes recognized, each (prospective) parent generally starts to accept and plan for their new arrival. If the pregnancy is lost, this is often considered a "death within the family" and the couple will go through an intense grieving process. The loss of a pregnancy can be devastating for a couple, regardless of the number of children in the family or the cause for the loss.

Components of the grieving process may be easier to accept and cope with if they are consciously understood. Therefore, I have outlined major issues. Interested couples can either read the original sources or consult a professional psychologist specializing in this area.

The grieving process often includes sequential periods of

  • denial, beginning with the shock of learning that there has been a death
  • anger, often inappropriately directed at anyone the person thinks about or sees
  • bargaining, often involving charitable acts or attempts to reconcile damaged relationships
  • depression, often associated with feelings of helplessness and hopelessness
  • acceptance, enjoying the time spent with family and social groups "more than ever."

There may also be changes in one's self image. The changes that have been described in the context of the loss of a body part may be relevant, including

  • impact, beginning at the point of awareness that there is a problem requiring the loss of a body part (or here essentially a "family part")
  • retreat, where denial of the importance of the loss may occur (a second opinion at this point is often important in allowing movement toward closure)
  • acknowledgment, with acceptance of the need for treatment generating an attempt to place the treatment and loss into an appropriate context
  • reconstruction, a redefinition of self image without the presence of the lost part (or family member)

In this grieving process, if the "redefined self images" of each member of the family can not be accepted by the other members then there is often a long lasting impact possibly resulting in depression. If a couple can "not get over the loss" then professional counseling is often quite powerful and should be recommended.


Recurrent Pregnancy Loss: Incidence Rates

No one really knows how often human pregnancies are lost. Investigation of the "rates of loss" are inherently difficult due to the inaccessibility of information.

  • Many pregnancies are lost "at home" without the aid of medical facilities. These losses may go unreported.
  • In the USA, there is no formal reporting of previable pregnancy losses to a centralized agency (as there is with live birth statistics).

Most of the research on incidence rates of pregnancy loss incorporate their own unique methodologies for identifying these losses. Consequently, the rates reported between studies vary tremendously. This situation has led to the apparent discrepancy often noted between doctors in their discussions on the rates of miscarriage.

The most widely accepted rate of loss for a "single spontaneous abortion" in an unselected population of couples (that is, regardless of characteristics associated with pregnancy loss) is about 15-20% (1 in 6) of "clinically detected" pregnancies (where the woman missed a menses or otherwise knew that she was pregnant).

Many pregnancies are lost prior to clinical detection but the incidence of these very early losses is not clear. A number of studies checked for pregnancy each month with a highly sensitive immunoassay from blood drawn or urine collected in sexually active women not using contraception. This research consistently demonstrates a high rate of "unrecognized pregnancy" in woman who are just "a little late for the menstrual flow." Some studies report a total pregnancy loss rate (nonclinical plus clinical) of more than 50% (1 in 2).

The chance of having a second spontaneous abortion with a history of only one isolated spontaneous abortion is generally considered to remain at 15-20% (for clinically recognized pregnancy). The incidence does not decrease (as if you used up your 1 in 6 and now must have 5 normal pregnancies) or increase significantly.

If there have been 2 spontaneous abortions in a row, then the most reliable information suggests that there is about a 35% chance (1 in 3) that the next pregnancy will be lost. Therefore, the loss rate is approximately doubled.

If there have been 3 spontaneous abortions in a row, then it appears that the couple has a roughly 45-50% chance of a loss with the next pregnancy. There are reports indicating improvement in future pregnancy success for couples with recurrent pregnancy loss after there has been at least one prior live born for the couple (that is, a 40-45% loss rate if no live borns and only a 30% loss rate with a history of a prior live born). Therefore, the couple's prior reproductive history is also important.

The spontaneous abortion rate rises as the woman's age increases, with a gradual increase starting about age 30, more rapid increases after age 35, and much more rapid increases after age 40. The age related increases in spontaneous abortion rates appear to be predominantly due to chromosomal accidents around the time of fertilization, where the egg is given one too many or one too few chromosomes so that the resulting fertilized egg (embryo) has a lethal genetic abnormality. When women over 40 or 45 years age are recipients of donor eggs from younger women they do not have this increased spontaneous abortion rate. This suggests that the cause of this increase in loss rate is related to egg rather than uterine factors.

About 80% (4 in 5) of spontaneous abortions occur in the first trimester of pregnancy (in the initial 13 weeks gestation). In couples without a history of recurrent losses if a fetal heart beat (FH) is seen by ultrasonography at 6 weeks gestational age then there is a reduced loss rate to about 5% (1 in 20). There is a further reduction if an FH is seen at 8 weeks gestation to about 3% (1 in 33). Unfortunately, in couples with recurrent losses the loss rate is still about 4-5 times greater (about 20% or 1 in 5) even after seeing an FH. Of course, seeing the fetal heart beating is reassuring but not as encouraging as if seen in an unselected population.

The high level of uncertainty involving any pregnancy seems to warrant that couples remain "cautiously optimistic" when they recognize a pregnancy. Many couples do not announce that they are "expecting" until seeing the FH or the completion of the first trimester.


Recurrent Pregnancy Loss: Overview of Causes

Many couples blame themselves (often harshly) for their pregnancy losses. In fact, it is rare that either member of the couple has done anything that would result in a pregnancy loss. Additionally, the actual incidence of pregnancy loss in the United States is much higher than typically thought. The result is that many couples might benefit from knowledge of the recognized causes for pregnancy loss.

There is a major difference in both "incidence rates" and "causes" for single spontaneous abortions and recurrent spontaneous abortions. Recurrent abortion is typically defined as three or more consecutive (in a row) pregnancy losses that occur prior to fetal viability (usually 20 weeks gestation or a fetal weight of 500 grams). The reason for this criteria is the reports of a significantly higher chance for further pregnancy loss following the loss of "three in a row."

The two major clinically important categories of causes for spontaneous abortion (miscarriage) are fetal and maternal.

Fetal causes include the genetic composition of the fetus.

  • human live borns have a very low percentage of chromosomal abnormalities (about 0.6% or 1 in 170). This low percentage indicates that almost all chromosomal abnormalities are lethal and aborted early in pregnancy.
  • The only chromosomal abnormalities (other than those involving the X and Y sex chromosomes) that might result in a human live born are trisomy 21 (three of the 21 chromosome, known as "Down's syndrome"), trisomy 18 (three of the 18 chromosome, known as "Edward's syndrome" and all die during infancy) and trisomy 13 (three of the 13 chromosome, known as "Patau syndrome" and all die during infancy).

Maternal causes include abnormalities in the "environment" in which the embryo and fetus develops. Known maternal causes related to an action of the mother are uncommon, but can include

  • heavy smoking (uncommon for this to result in a loss)
  • alcohol abuse (uncommon for this to result in a loss)
  • irradiation or exposure to chemical toxins
  • medications known to be teratogenic (cause fetal malformation)

Other maternal causes which are not related to any conscious activity of the mother or couple include

  • anatomic abnormalities (typically uterine)
  • hormonal imbalances (typically in progesterone)
  • immunologic system errors (autoimmune and alloimmune)
  • serious or life threatening maternal disease

By far the most common causes for spontaneous pregnancy loss are fetal not maternal. It is difficult for a woman with an undesired pregnancy to consciously create an unfavorable environment for the pregnancy to successfully force a miscarriage.

Often couples blame themselves for "doing something" that must have resulted in the pregnancy loss. Focusing on themselves (often harshly) for doing something wrong is unfortunate since

  • it adds guilt on top of an existing emotionally charged situation, which is counterproductive and may delay or arrest recovery from the event
  • it is misdirected since very few losses are related to conscious maternal actions
  • it often assumes that such losses are rare events when in fact they are common (but not commonly discussed)


Recurrent Pregnancy Loss: clinical evaluation

An evaluation for known causes of recurrent pregnancy loss is usually initiated after 2 or 3 consecutive pregnancy losses. The tremendous emotional impact of each loss may encourage an evaluation sooner than later. A full evaluation includes

  • demonstration of a normally shaped uterine cavity (by either hysterosalpingogram or hysteroscopy)
  • evaluation for a hormonal deficiency in progesterone production (by either endometrial biopsy or bloodwork)
  • analysis of both the maternal and paternal chromosomes (by bloodwork)
  • laboratory testing for immunologic causes of pregnancy loss (by bloodwork)
  • taking a history for maternal disease states, environmental or other toxin exposure

If a full evaluation is completed on couples with either 2 or 3 consecutive losses there will still be about 50% (1 of 2) of couples with "unexplained" recurrent pregnancy loss. That is, roughly half of couples seem to have a reason for recurrent loss that is beyond modern medicine's ability to diagnose this cause. This can be frustrating for both the couple and the physician. In this situation, the couple will at least know that potentially repairable pathology has been ruled out. The couple can then elect to enroll in experimental protocols designed to further our knowledge of recurrent pregnancy loss. In my experience, these experimental treatments often result in reproductive success despite limited knowledge on why they work.

Recurrent Pregnancy Loss: Anatomic Causes

Implantation and development of a human embryo requires both space and an abundant blood supply. Normally both are readily available within the uterine cavity.

In the nonpregnant state, the uterine cavity is primarily a potential space (with a 5-10 milliliter volume) between the anterior and posterior muscular walls of the uterus. Under pressure (such as during operative hysteroscopy) the uterine muscular walls can be distended to about 10-30 milliliters. During pregnancy, the walls of the uterus rapidly adapt to increase the intracavitary volume for the developing fetus. The ability of the uterine walls to adapt is largely due to the muscular composition and ample blood supply. In conditions that compromise either the muscular composition of the uterus or its blood supply there is a tendency for recurrent pregnancy loss.

An estimated 15% of couples (1 in 6) with recurrent pregnancy loss have an anatomic abnormality of the uterus as the primary reason. The four categories of anatomic defects that cause recurrent pregnancy loss are abnormalites in

  • the normal process of uterine fusion during embryonic development (Mullerian duct abnormalities)
  • the development of the uterus and cervix due to maternal diethylstilbestrol (DES) ingestion
  • the size of or circulation to the uterus due to fibroids, endometrial polyps or scar tissue (Ashermann's syndrome)
  • the function of the cervix


(1) Defects in uterine fusion and resorption

The incidence of congenital uterine fusion defects is not well established since these defects are not readily apparent without radiologic imaging (or the equivalent). Therefore, the incidence rates in the literature have a potential selection bias, since they reflect a selected group of women (that is, a population selected to have the testing required for diagnosis). Understanding this, the available reports generally cite incidences of about 1 in 200 to 1 in 600 women. Therefore, these defects are not common but also not rare. About 1 in 4 women with a congenital anomaly of the uterus due to a fusion defect will have significant difficulty with reproduction, possibly including recurrent pregnancy loss.

Mullerian duct abnormalities include

  • septate uterus: a partial lack of resorption of the poorly vascularized tissue within the uterine cavity. The remaining tissue creates a wedge shaped septum in the fundus (top) of the uterine cavity. Implantation of an embryo onto this septum, and within a septate uterus, has a markedly greater chance of spontaneous abortion (miscarriage) compared with a normally developed uterus. There is also an increase in preterm labor and delivery as well as abnormal fetal lie or presentation (such as breech). Fortunately, the repair of the septate uterus is fairly straightforward and usually very effective, requiring an outpatient surgical procedure (hysteroscopy). Reportedly, repair of a septate uterus will improve the miscarriage rate from about 80-90% if untreated to about 10-15% following treatment.
  • unicornuate uterus: failure in development of 1 of the Mullerian ducts most likely due to a failure in the migration (movement) of the duct to the proper location prior to its time for fusion (resulting in its loss). The resulting unicornuate or half uterus has connection to only 1 fallopian tube since the other tube was to be formed from the "lost" Mullerian duct. The caliber (size) of the cavity in the unicornuate uterus is very important in determining the likelihood of reproductive success. Generally, a unicornuate uterus is associated with the highest rate of loss of any of the congenital abnormalities of uterine fusion. Unfortunately there is no accepted benefit for the treatment of these uterine defects. The unicornuate uterus is associated with renal abnormalities (renal agenesis, lack of a kidney, on the side of the missing Mullerian structures) and other reproductive problems (abnormal lie or presentation, intrauterine growth retardation, preterm labor and delivery, incompetent cervix). An incompetent cervix is essentially a "weak" cervix that tends to open without contractions during the course of pregnancy and can result in the delivery of a markedly premature baby or a nonviable fetus (miscarriage).
  • bicornuate uterus: a partial lack of fusion of the Mullerian ducts resulting in a single cervix and two uterine cavities in a "heart shaped" partially unified uterus. Reproductive outcome may be normal so no treatment is indicated unless reproductive problems are identified. Reports suggest an increase in spontaneous abortion (miscarriage), preterm labor and delivery, and abnormal presentation (breech). The patient with recurrent pregnancy loss, a bicornuate uterus, and treatment for all other identified causes for the losses may reasonably consider repair of the uterus. The treatment is surgical repair requiring a laparotomy with unification of the uterine cavities. The laparotomy requires a lengthy (few weeks) postoperative recovery period. Success with this surgery is generally quite good, improving the miscarriage rate from about 90-95% if untreated (and the cause for the recurrent losses) to about 25-30% if treated.
  • didelphic uterus: a complete lack in fusion of the Mullerian ducts with duplication of the uterus and cervix so that the patient has 2 cervices and two uteruses (each smaller than normal). Commonly also associated with a vaginal septum so that there are 2 separate vaginal canals at the top of the vaginal vault. Occasionally, one of the sides will become obstructed and result in pain as blood accumulates in the obstructed hemiuterus. There is an association with abnormal lie or presentation as well as preterm labor and delivery.
  • rare abnormalities: there are an entire host of intermediate or somewhat unique problems associated with abnormal development of the Mullerian structures. Isolated endometrial (lining of the uterus) or cervical agenesis (lack of development) are rare. Communicating and noncommunicating uterine horns that failed to fuse and canulate properly are possible.


(2) DES exposure in utero

Maternal diethylstilbestrol (DES) ingestion during pregnancy was intended to reduce the likelihood for spontaneous abortion (miscarriage). DES was the first orally active commercially available estrogen compound and as it gained popularity it was widely used from the 1940s to 1971. In 1971, the association between DES exposure in utero and the subsequent development of a rare vaginal cancer (in the female offspring) named "clear cell adenocarcinoma" was established. The FDA promptly removed the pregnancy related indications for DES when these findings were revealed. Therefore, females born in 1971-1972 are the youngest to be exposed to this medication. In 1997, these women are 25 years old so reproductive consequences of the medication for this population of women is still important. The abnormalities associated with DES of reproductive importance include:

  • structural abnormalities of the upper vagina, including transverse vaginal ridges or a vaginal hood over the cervix
  • structural abnormalities of the cervix, including cervical collars, a cervical cockscomb appearance, and inadequate development of the cervix
  • abnormal size and shape of the endometrial cavity, generally including a small uterine cavity with possibly a T shape (due to incomplete resorption of the lateral walls of the cavity), constricted regions at the ostia (entrance of the tubes into the cavity) to give an hourglass like image on HSG, and an irregular contour of the cavity

The risk of spontaneous abortion in women who were exposed to DES in utero is far greater than normal, especially when the abnormalities listed above are present.

No currently available treatment for any of these DES changes in terms of spontaneous abortion has been convincingly shown to be effective in the available literature. A common practice is placement of a cerclage into the cervix, since the abnormally developed cervix may have a strong association with pregnancy loss. The literature on the techniques and benefits of this treatment do not allow for a uniform recommendation. Another treatment that has been proposed but of unclear benefit is the hysteroscopic transsection of the lateral walls of the uterus to increase the size of the cavity. This surgery is potentially dangerous since the incisions are close to major uterine blood vessels.

(3) Abnormalities in uterine circulation

The size of the uterine cavity and its circulation is critically important for pregnancy. Uterine circulation can be altered by the presence of fibroids, endometrial polyps or scar tissue (Ashermann's syndrome).

Uterine fibroids, known as leiomyoma uteri, are tumors of the smooth muscle cells in the wall of the uterus. The uterine wall is primarily composed of smooth muscle cells (the myometrium). A uterine fibroid is thought to originate as a mutation within one of these myometrial (smooth muscle) cells that leads to the progressive loss of it own growth regulation. Each fibroid tumor grows from a single progenitor cell (each tumor arises from one single cell) and all the cells within a particular fibroid contain the same abnormal DNA that favors growth. Different fibroid tumor originate from different muscle cells, each with their own genetic (DNA) abnormality so that each tumor may grow at its own rate (some faster and some slower). Fibroid tumors are not malignant (cancer) yet there is an uncommon cancer called "leiomyosarcoma" that is composed of malignant smooth muscle cells. It is not clear whether these cancers develop from benign fibroids or whether they arise independently. Fibroid tumors of the uterus are common. About 75% (3 of 4) of uterine specimens removed during abdominal hysterectomy contain fibroids (many are quite small) and about 15-20% of hysterectomies performed in the USA are for problems involving fibroids.

The role of uterine fibroids in reproduction is usually not clear. If the fibroid is presenting (bulging) into the uterine cavity (submucosal) then it may obstruct one of the fallopian tube entrances or it may present a mechanical or other barrier to implantation. If the fibroid replaces an entire wall of the uterus, then it might interfere with the blood supply to the uterine structures around it or an embryo implanting near it. If the fibroid is predominantly on the outside of the uterus with projection into the pelvis and abdomen then it may outgrow its own blood supply and become degenerate or infected. Degenerating or infected fibroids may result in pain and irritability (contractions) of the uterus that can be associated with complications of pregnancy (preterm labor, severe pain).

Most fibroids do not seem to interfere with fertility and should not be removed unless (a) a reproductive problem is identified and (b) all other treatable causes for the reproductive problem have been evaluated and excluded. An exception is the presence of a large intrauterine filling defect seen on HSG, which should be removed. Another exception is a fibroid compressing the fallopian tubes or creating a tremendous distortion of the uterine cavity.

Endometrial polyps appear to be organized overgrowths of the uterine endometrium, although the precise mechanisms leading to their development are not clearly defined. The endometrial lining of the uterine cavity grows in response to estrogen and is architecturally restructured in response to progesterone. If the response to estrogen is excessive, either in the presence of unopposed estrogen (such as during periods of anovulation) or whenever the bioactivity of the circulating estrogen is increased (such as with increased numbers of estrogen receptors or with decreased metabolism of estrogen) overgrowth of the endometrium may occur. If these overgrowths organize and develop their own blood supply then they become polyps.

The mere presence of polypoid overgrowths in the uterine cavity may (at least theoretically) interfere with implantation and fertility. I have envisioned polyps as acting sort of like IUDs in the cavity, creating a hostile environment for embryo implantation. I remove endometrial polyps in women with reproductive problems and these women (anecdotally) seem to do remarkably well in subsequent fertility efforts. A well designed research project describing fertility outcome after treatment for different types of endometrial polyps would be valuable.

Endometrial polyps are not always benign. I removed one normal appearing endometrial polyp hysteroscopically and this was found to contain an endometrial adenocarcinoma (cancer) on pathology report. Therefore, in the presence of any atypical overgrowth of tissue it is always important to think about the possibility of cancer.

Ashermann's syndrome is the occlusion or obliteration of the uterine cavity due to damage to the lining of the cavity (endometrium). This is not common but is important to recognize it if indeed present. When the endometrium is destroyed beyond a certain depth (believed to be the basalis level which is the level that promotes subsequent growth) in the context of hypoestrogenism (a low circulating estrogen concentration) then permanent scar tissue can easily form within the cavity. Clinical situations that increase the chance of Ashermann's Syndrome include

  • overzealous dilatation and curettage (especially for a missed abortion, postpartum bleeding, or septic abortion),
  • intrauterine surgery to remove fibroid tumors, uterine structural defects (septum, bicornuate uterus, large polyps), or at cesarean section
  • infections related to IUD use (or the placement of any foreign object within the uterine cavity),
  • some uncommon infections of the uterus (such as intrauterine tuberculosis or Schistosomiasis)
  • radium insertion into the uterus for the treatment of gynecologic cancers

The finding for Ashermann's Syndrome on hysterosalpingogram (HSG) exam is intrauterine filling defects.These are irregular areas within the normally triangular shaped cavity where the distending media is excluded due to the presence of the adhesions. Thin adhesions may be primarily composed of fibroconnective tissue with little blood supply. The thicker the adhesions, the greater the likelihood that they are vascular and possibly also partially muscular. Vascular and muscular adhesions are much more difficult to repair and seemingly pose a greater problem for fertility.

Repair of intrauterine adhesions is most easily and safely performed by hysteroscopy. Operating scissors can be used through some hysteroscopes but tend to be a bit flimsy for any but very thin filmy adhesions. A type of operating hysteroscope called a resectoscope allows the surgeon to apply electrical current through a monopolar cutting instrument attached as the operating element of the hysteroscope and lysis (cutting) of the adhesions can then be performed. In more complex cases of adhesions, repeated procedures may be required to accomplish complete lysis of the adhesions. After each hysteroscopic repair in which cautery is used or extensive lysis of adhesions is accomplished, the patient is typically placed on higher dose estrogen replacement (say, Premarin 1.25 or 2.5 mg by mouth each day for 30-60 days, with a Provera withdrawal flow brought on at the end of this time) to promote the regrowth of endometrium (lining) over the repaired sites. Occasionally, a stent (such as an IUD or pediatric foley balloon) is also placed within the cavity to keep the sides of the uterus apart during the repair period.

For mild to moderate adhesions, you might expect a 60-80% chance of successful pregnancy after repair. For more extensive adhesions the chance of a successful pregnancy is lower. If a pregnancy does occur after repair of Ashermann's Syndrome there is a greater chance of preterm labor and delivery (delivery of a premature baby), placenta accreta (where the placenta invades the uterine wall into the muscular component of the wall and becomes difficult to impossible to remove) and postpartum hemorrhage (heavy bleeding after the delivery of a baby).

(4) Abnormal cervical function

The cervix is very important in the development of a pregnancy. The nonpregnant cervix is normally composed of a dense collagenous fibroconnective tissue with small amounts of smooth muscle to give it a tough texture. In pregnancy, the increased water content and vascularity in the cervix leads to a softening and a blue coloration. Throughout pregnancy the cervix and lower uterine segment change but maintain a "functionally intact" internal os.

If the internal os of the cervix dilates or effaces during pregnancy this can be an ominous sign. In the beginning of pregnancy, cervical dilatation with some bleeding is known as an "inevitable abortion." In later pregnancy, cervical dilatation or effacement associated with lower abdominal cramps or pressure is a sign of labor (which is preterm if it occurs prior to 37 weeks gestation).

If there is painless dilatation or effacement of the cervix, usually occurring between the mid second trimester (about 20 weeks gestation) to the early third trimester (about 27-30 weeks), this is usually the result of an incompetent cervix. Pregnancy losses at progressively earlier gestational ages often reflect an incompetent cervix that gives way earlier with each subsequent pregnancy. The fetal membranes (chorionic and amniotic membranes) can sometimes be found bulging from the open cervix and can indeed hourglass through the cervix to fill the entire vaginal vault (which can be difficult to distinguish from a fully dilated cervix).

The causes of cervical incompetence can be congenital or acquired and include

  • congenital abnormality in the composition of the cervix, with a relative deficiency in the tougher collagenous fibroconnective material or relative increase in concentration of the less tough smooth muscle
  • congenital hypoplasia (underdevelopment) of the cervix, such as with in utero exposure to DES
  • trauma to the cervix, such as with mechanical dilators for dilatation and curettage (D+C), cervical conization or extensive biopsy, and precipitous labors or cervical lacerations during labor and delivery.

Establishing the diagnosis of cervical incompetence with certainty can be difficult. Generally, a suggestive history of late painless pregnancy losses with the history of a plausible cause is all that is used to diagnose the condition. Additional testing sometimes suggested to confirm the diagnosis (none of which have been widely accepted) includes

  • passing an 8 mm dilator into the nonpregnant uterus through the internal os (an office procedure)
  • a hysterosalpingogram (HSG) on the nonpregnant uterus to look for funneling of the lower uterine segment and an open internal os
  • ultrasonography of especially the pregnant uterus looking for shortening or dilatation of the cervix and bulging of the fetal membranes

Treatment of an incompetent cervix is surgical. The cerclage is an attempt to strengthen the cervix, with the two most commonly used modern techniques having been developed in the 1950s by Drs. Shirodkar and McDonald. These techniques involve the surgical placement of a suture or Mersilene band around the cervix to hold it closed. In appropriately selected women, the improvement of pregnancy outcome with a cerclage is seemingly impressive. Generally, 80-90% of women with cervical incompetence as their cause for recurrent pregnancy loss will deliver a viable live born following cerclage placement.



Recurrent Pregnancy Loss: Hormonal Causes

Hormonal causes for recurrent pregnancy loss are generally considered "luteal phase defects." Luteal phase defects are most often thought to result from inadequate progesterone effect on the uterine endometrial lining. The existing literature on luteal phase defects is inconsistent and many physicians question the significance or even the existence of these defects. More often than not, infertility specialists in the USA accept that these luteal phase defects indeed exist and are capable of playing a significant role in a small group of couples (thought to be less than 5% but some supporters claim up to 40%) with recurrent pregnancy loss.

To successfully implant into the uterus the embryo must be available during a window of time limited to a few days per cycle, referred to as the "window of uterine receptivity." If this window of uterine receptivity is not properly timed with respect to ovulation then either infertility or pregnancy loss may occur. The primary regulation for this window of receptivity appears to be hormonal (progesterone). Molecular events (currently poorly understood) change in response to hormonal shifts and most likely allow for and guide implantation of the developing embryo (fertilized egg). Research is active in this area. Descriptions of cell adhesion molecules that allow the embryo to adhere to the uterine lining (the "molecular glue" that sticks the developing embryo to the endometrium) and their hormonal (or other) regulation is just one exciting area of investigation.

Progesterone appears to have a critical role in implantation and the development of a normal pregnancy. Limited exposure to progesterone may result in infertility (severe) or recurrent pregnancy loss (milder). Characteristically, decreased progesterone results in a shortened (less than 11 day) luteal phase (period between ovulation and the onset of the next menses) or a persistently abnormal endometrial biopsy (greater than 2 days out of phase). When these changes are more severe, the impact on reproduction can be greater.

Luteal phase defects can be categorized into classes that guide treatment. The ovary's corpus luteum cyst that develops following ovulation produces progesterone. Initially, the stimulus for progesterone production is pituitary LH, which supports the progesterone production (by the corpus luteum's granulosa cells) for about 11-14 days. As LH support declines in the presence of a pregnancy progesterone production by the corpus luteum is normally rescued by placental hCG (which is functionally similar to pituitary LH) until about 7-10 weeks gestation. After 8-10 weeks gestation, the primary source of circulating progesterone changes from the ovarian corpus luteum cyst to the uterine placenta.

Three potential causes (classes) of luteal phase defects are

  • inadequate luteal phase production of progesterone. This may be the result of ovulation from a small ovarian follicle (and thus corpus luteum cyst) which would be lined by fewer hormone producing granulosa cells. There also might be inadequate circulating pituitary LH. This would effect implantation and very early development. Increasing progesterone during this time could be accomplished by preimplantation direct supplementation (progesterone medication) or promoting increased follicular (and consequently corpus luteal) cyst development (typically with clomiphene citrate).
  • inadequate progesterone production after luteal rescue by the placental hCG. Circulating hCG concentration must increase dramatically until about 10 weeks to adequately support the corpus luteum. Inadequate hCG production might be related to a small (or abnormally growing) placenta due to either a uterus with an inadequate blood supply or an abnormal embryo (fetus). This would effect the embryo from about 4 to 10 weeks gestation. Increasing progesterone during this time could be accomplished by direct supplementation (progesterone medication). Generally, progesterone supplementation is not powerful enough to prevent a miscarriage if there is an abnormal embryo.
  • inadequate placental production of progesterone. This may be due to either a small placental mass or a biosynthetic abnormality in placental progesterone production. This would effect pregnancies primarily after 10 weeks gestation. Increasing progesterone after 8 weeks gestation could be accomplished by direct supplementation (progesterone medication). This appears to be the least common cause for pregnancy loss since most losses occur prior to this time.

Infertility specialists occasionally draw upon a small handful of reports describing women with documented low progesterone concentrations in pregnancies resulting in normal outcomes. These reports force one to question the absolute necessity of progesterone, and suggest the possible importance of other nonhormonal regulators of the "window of uterine receptivity." These reports include:

  • women with the congenital abnormality known as "abetalipoproteinemia" have cells that are unable to take up and use VLDL-cholesterol. VLDL-cholesterol is a primary source for cellular cholesterol. Since cholesterol is required for the synthesis of progesterone these women have very low circulating progesterone concentrations. There are reports of women with abetalipoproteinemia who have documented low progesterone concentrations throughout pregnancy and have carried their pregnancy to term
  • fetuses with a rare deficiency in one of the enzymes required for progesterone production, such as "3-beta hydroxysteroid dehydrogenase" or the "cholesterol side chain cleavage complex," may be delivered at term despite the inability of these fetuses (and presumably also their placentas) to produce adequate progesterone. Prenatal diagnosis of these conditions has never been early enough to actually document low progesterone throughout pregnancy (at least from the time of placental takeover of progesterone production)
  • an In Vitro Fertilization patient with a diagnosis of unexplained infertility discontinued her prescribed progesterone when she noted vaginal bleeding at 4-5 weeks gestation (and assumed that she was not pregnant). Bloodwork documented a progesterone concentration of less than 2.0 ng/ml at 5-6 weeks gestation, she did not return to progesterone supplementation and she delivered a normal fetus at term. It is generally accepted that a progesterone concentration of less than 7 ng/ml at the time of hCG rescue (the usual nadir in progesterone concentration which occurs at about 4 weeks gestation) is ominous and predicts spontaneous abortion.
  • a mouse with a knockout mutation of the gene encoding the progesterone receptor has recently been described. These mice appear to be incapable of ovulation. Future experiments with these mice may reveal important aspects of the role of progesterone in reproduction, including implantation.

The endometrial biopsy is the "gold standard" diagnostic test for luteal phase defects. It only detect defects that are due to inadequate luteal phase progesterone production. The other hCG or placental progesterone defects are not determined. Pregnancies lost due to early hormonal defects occur throughout the first trimester suggesting abnormal embryogenesis (embryo development) rather than an immediate uterine rejection.

The reliability of the endometrial biopsy has been questioned. Research on the biopsy includes

  • introduction of histologic criteria for dating endometrium in 1951, with a stated mean error in dating of 1.8 days
  • more than 60 luteal biopsy specimens were examined by the same experienced pathologist at two different times to compare the assigned dates and only 24% (about 1 in 4) were read as the same day and in 10% of specimens there was a greater than 2 day discrepancy in dating (which would change the diagnosis of LPD)
  • more than 60 luteal phase biopsy specimens were read by 5 different pathologists and in about one third of the specimens there was a difference in dating of 2 days or more.
  • when multiple methods of ovulation detection were compared head to head in the assessment of greater than 25 biopsy specimens it was demonstrated that the percentage of out of phase biopsies is related to the method of ovulation detection used. The percent of out of phase biopsies was 4% with ultrasonography, 15% with ovulation predictor kits, 23% with basal body temperatures and 30% with next menstrual period.
  • a total of 39 luteal phase biopsies were performed in repeated months in 5 regularly menstruating fertile women with normal circulating thyroid hormone and prolactin concentrations. Using the usual criterion of a greater than 2 day discrepancy to define an out of phase biopsy it was determined that 31% of random single biopsies and 6-7% of (two) consecutive biopsies were out of phase.

These studies emphasize the importance of attention to detail in timing and performing endometrial biopsies. In particular, an experienced pathologist or infertility specialist trained to assign dates to these tissues should be sought, there should be use of multiple ovulation detection techniques (I typically use a combination of the next menstrual period, the basal body temperatures and the ovulation predictor kits), and ideally an abnormal result should be repeated prior to diagnosing a LPD (most fertility specialists define the LPD as two consecutive out of phase biopsies) since there is a high background rate of abnormality in single random biopsies.

If an inadequate progesterone effect is documented or believed to exist during the luteal phase of the menstrual cycle then either supplemental progesterone (either as oral micronized progesterone, vaginal suppositories in a gel, or by injection in an oil base), supplemental hCG (as injections every few days following ovulation to enhance the ovary's own progesterone production) or clomiphene citrate in the follicular phase (to increase the final follicular size, number of granulosa cells and luteal progesterone production) are treatment options. Supplemental progesterone medication is usually administered until about 10 weeks gestation (after the placenta takes over progesterone production).

Reports of treatment success with progesterone supplementation for LPD often lack appropriate controls (such as a similar group of women who did not receive progesterone treatment). However, existing reports taken together support the use of progesterone supplementation in documented cases of LPD.

Claims of progesterone teratogenicity (cause of fetal malformations) are unproven. A "Collaborative Perinatal Project" report in 1977 suggested an association between fetal cardiac defects and first trimester exposure to female hormones or birth control pills. The data in this study was reevaluated (published in 1984), revealing that the timing of the hormonal drug exposure was inconsistent with the cardiac effects suggested and that if the examined pregnancies which involved Down's syndrome were removed from the data then there was no increased risk of cardiac anomalies.

In a report (published in 1985) of women who were given progesterone supplementation for prevention of spontaneous abortion (upon presentation with a threatened abortion) over 2,700 infants were examined and did not have an increased number of anomalies compared to the general population.

Overall, there seems to be no known significant increased risk of fetal anomalies in taking natural progesterone supplementation during pregnancy. However, the couple taking the medication should be aware of this potential for risk and the data that addresses this risk.



Recurrent Pregnancy Loss: Chromosomal Causes

At least 50% of the clinically recognized pregnancies that are lost in the first trimester have a major chromosomal abnormality when the products of conception are examined. In pregnancies lost later in gestation there is also a high rate of chromosomal abnormality, roughly 30% in the second trimester and 5% in the third trimester.

The "largest single class" of chromosomal abnormality found in spontaneous abortions are the autosomal trisomies (roughly 50% of abnormal specimens), some of which reveal a maternal age effect. Monosomy X is the "single most common" chromosomal abnormality found in spontaneous abortions (roughly 25% of abnormal specimens), which usually occurs due to a loss of the paternal sex chromosome (Y chromosome) and is not more common with advancing maternal age.

Certain chromosomal abnormalities are universally accepted by infertility specialists as a cause for recurrent pregnancy loss. Fortunately, these major chromosomal abnormalities are uncommon. They may occur within either the maternal or paternal chromosomes. The overall incidence of chromosomal abnormality as the cause of recurrent pregnancy loss is low (less than 5% of couples with recurrent losses). The abnormalities associated with recurrent pregnancy loss include

  • translocations, which are either "Robertsonian" (the result of centric fusion following the loss of the short arms of two participating chromosomes and possibly one of the centromeres) or "reciprocal" (interchanges of genetic material among different chromosomes following the breaking off of pieces of chromosomes such that the total chromosome complement may still contain all of the genetic material)
  • inversions, which are either "paracentric" (rearrangements in the genes on a chromosome following the breaking off of a piece of chromosomal material and its reattachment in a different order, not including the centromere) or "pericentric" (rearrangements within a chromosome that do include the centromere)
  • possibly recurrent aneuploidy (numeric chromosomal abnormality), where given couples appear to show a predisposition toward particular types of chromosomal abnormalities (the most common being recurrent trisomies) and couples with a loss showing an abnormal chromosomal complement (in the products of conception) for an initial pregnancy loss have a much greater chance of another abnormal chromosome analysis (in the products of conception) if they suffer another loss (up to 80%).


Robertsonian translocations

  • involve only the acrocentric chromosomes (consisting of the "D" chromosomes 13, 14 and 15 and the "G" chromosomes 21 and 22) where the lost short arms of the involved chromosomes appear to only contain genes for ribosomal RNA (which does not result in malformations so that the carriers appear totally healthy),
  • may involve the same chromosomes, such as 13/13 or 21/21, where there is no chance for a normal offspring. A 21/21 translocation (seemingly the worst case) results in the offspring only having either a lethal monosomy (50%) or Down's syndrome (50%),
  • may result from a 21/D or 21/22 translocation where there are theoretically 6 possible combinations with three of these being lethal and of the other three there is an equal 33% chance of having a Down's syndrome, a balanced translocation (normal appearing as with the parent carrying the translocation) or a normal set of chromosomes. In reality, for an unknown reason, experience has shown that there is only about a 10-15% chance of Down's syndrome if the mother is the carrier and 2-5% chance of Down's syndrome if the father is the carrier, while the balanced translocation rate is about 50% in liveborns.


Reciprocal translocations

  • are rearrangements between chromosomes that do not necessarily entail loss of genetic material in the "parent" (they appear normal)
  • may be passed on to the offspring such that small pieces of genetic material are either added to or subtracted from the total chromosome complement (an unbalanced translocation) of the offspring. This can be either lethal or very damaging (usually resulting in severe mental retardation and major anomalies)
  • in reality, experience has shown that there is only a 10-15% risk of an unbalanced translocation in offspring independent of whether the mother or father is the carrier


Paracentric inversions

  • are very difficult to detect on chromosome analysis
  • are less common than pericentric inversions
  • are rare causes for recurrent pregnancy loss
  • may result in early pregnancy loss


Pericentric inversions

  • are usually easy to identify on chromosome analysis
  • are detected in about 1 in 1000 couples with recurrent pregnancy loss
  • are more likely to result in abnormal crossover of genetic material during meiosis to yield unbalanced offspring
  • females with a pericentric inversion have about an 8% chance, while males with a pericentric inversion have about a 5% chance, of an abnormal liveborn (depending on the specific inversion detected)
  • pericentric inversions can develop as the cells taken for chromosome analysis are grown in culture so that the lab should be contacted whenever an inversion is reported to identify the labs experience with similar inversions (to determine the likelihood of culture artifact)

Whenever a couple has suffered three (or if desired by the couple, two) spontaneous abortions, an investigation of the chromosomes for both of the female and male is indicated. This investigation is normal 95% of the time. If an abnormality is identified, prompt consultation with an experienced human geneticist is indicated since the theories and actual experiences in this field are complex and constantly changing. Preconceptional counseling and planning with respect to prenatal diagnosis is also important. Unfortunately, there is no available treatment to "fix abnormal chromosomes" at this time.


Recurrent Pregnancy Loss: Immunologic Causes

Immunologic causes of recurrent pregnancy loss are poorly understood. The theories proposed by authorities in this field appear to be constantly evolving and most of the theories that have been proposed to date have been proven to be either incorrect or largely incomplete.

Two major categories of immunologic causes of recurrent pregnancy loss are

  • autoimmune, in which the woman's immune system attacks her own organs and tissues, and
  • alloimmune, in which the immune system attacks tissues considered foreign.

The immune system is designed to protect oneself against infectious organisms and their toxins. The system identifies, immobilizes and eliminates "invaders." The two major mechanisms of surveillance are

  • nonadaptive immunity, in which cells respond nonspecifically to foreign molecules or material via either phagocytosis and lysis (by macrophages), lysozyme secretion (by lachrymal cells) or cell lysis (by natural killer cells). This type of response does not adapt and so its efficiency is not improved with further exposure
  • adaptive immunity, in which action against specific foreign molecules (antigens) is enhanced by reexposure. This is mediated by lymphocytes which produce highly specific antibodies that bind to the foreign molecules to further elicit (amplify) an immune response.

The immune system is constantly operational (turned on) since it must synthesize an enormous catalog of different antibodies and cell surface receptors to deal with the wealth of foreign material that it is presented with.

An important feature of the immune system is its ability to distinguish foreign (unwanted) material from its own (desired) self. If this ability to distinguish non-self from self fails, then the system produces an immune response against itself (or its own tissues). This is called "autoimmune" disease.

Autoimmune disease or dysfunction may play a role in up to 10% of recurrent pregnancy loss. Phospholipids are molecular building blocks that help to make up a large portion of the walls around the cells of the body, including placental cells. Anti-phospholipid syndrome (APS) is the autoimmune dysfunction that is classically associated with recurrent pregnancy loss.

APS is associated with pregnancy loss in any trimester, placental thrombosis (blood clots), and small placentae. The interruption of the circulation to the fetus via these blood clots is a possible reason for the fetal losses.

Identifying the mechanism behind the fetal losses would allow specific treatment to be developed. Clotting mechanisms are difficult to understand without a background in this area (this paragraph is included for completeness sake). Thrombosis may be caused by a relative deficiency in prostacyclin production within the cells that line the blood vessels (endothelial cells) since prostacyclin is a potent vasodilator and inhibitor of platelet aggregation. Thrombosis may also be caused by a relative insufficiency of the active form of the endogenous anticoagulant protein C, which normally degrades certain clotting factors to limit thrombosis, since phospholipids are required to activate protein C. At this time, the mechanism of thrombosis and fetal loss with APS is largely unknown.

Establishing the diagnosis of APS is important since most of the treatment options involve considerable expense and some added risk. Antiphospholipid antibodies are a large varied group of immunoglobulins directed against several different negatively charged cell surface phospholipids. Many of these phospholipids have been identified, with the best known being cardiolipin. Tests for APS can be divided into coagulation based tests and tests that detect the presence of the antibodies directly.

A group of phospholipid dependent coagulation tests are available (such as the kaolin clotting time, the plasma clotting time, dilute Russell viper venom time, and activated partial thromboplastin time) and serve as popular screening tests for antiphospholipid antibodies. Each of these coagulation tests relies on the activation of a "prothrombin activator complex" to allow for clot formation. Antiphospholipid antibodies block this activation to delay clot formation, such that in the presence of these antibodies there will be a prolongation of the time required for clotting and an abnormal result for these coagulation tests.

Other causes for an abnormal coagulation test do exist and should be ruled out if an abnormal result is found. The ways to exclude the other causes for abnormality include

  • diluting the test plasma with a control plasma containing normal concentrations of the necessary clotting factors and then rerunning the assay to determine if the test results normalize
  • reversing the anticoagulant effect of the antibodies by adding a surplus of phospholipid (such as contained in the cell wall fragments resulting from freeze thawing platelets)

These manipulations are not always run if the test ordered is abnormal (especially if a simple aPTT is ordered) but most labs are equipped to run these additional tests if requested by the physician.

There are a several available sensitive and specific assays for anti-cardiolipin antibodies, one of which should be obtained when there is a history of recurrent pregnancy loss. The classic assay for anti-cardiolipin antibodies is the Loizou ELISA, which has generally been modified over the years. At this time, the physician ordering any of the anti-cardiolipin antibody tests should become familiar with the particular assay used and its reference ranges since this information is necessary to interpret the results. When the units of measurement are GPL (IgG phospholipid units) and MPL (IgM phospholipid units) the results are not necessarily the same as when the units of measurement are IU (international units). When GPL and MPL units are used the cutoffs of normal are usually about 30 and 11, respectively. It is often easiest if the lab reports the sample results in terms of "multiples of the median" with interpretation being negative if less than 2, low positive if 2-3, positive if greater than 3. Results that are negative or low positive are generally considered clinically irrelevant and do not require treatment.

There are commercial assays for some of the other phospholipids such as phosphatidyl-serine, phosphatidyl-inositol, phosphatidyl-ethanolamine, phosphatidly-choline and phosphatidyl-glycerol. Rather than testing for each phospholipid individually, the more cost efficient test is one that detects a panel (usually all) of these phospholipids (such as an "antiphospholipid antibody package"). If the panel is positive then more specific detection of specific phospholipids can be considered. Clinically, it is not necessary to test for each of these specific antibodies since the treatment is the same for any of them. Specific testing is most appropriate in a research setting.

To summarize, all couples with recurrent pregnancy loss should be screened for APS. The tests that I routinely order include

  • one of the coagulation tests (aPTT) that relies on the activation of the "prothrombin activator complex" and which will be appropriately diluted with normal plasma when abnormal,
  • the anti-cardiolipin antibody test (positive in 2-3% of the general population, 7-45% of women with recurrent pregnancy loss- depending on what level is considered abnormal), and
  • the lupus anticoagulant test (positive in 1-2% of the general population, 10% of women with recurrent pregnancy loss)
  • I do not routinely order specific anti-phospholipid antibody tests since my management is not altered by the results. Some research centers may order these tests to determine experimental treatment protocols.

APS is classically defined as a triad of recurrent pregnancy loss, thrombosis and autoimmune thrombocytopenia (decreased platelet concentration). For those couples with recurrent pregnancy loss, the positive finding (on 2 separate occasions) of either an appropriately performed coagulation based test or a direct antibody test is generally all that is required to propose treatment.

Without treatment, couples with APS have a poor chance of carrying a fetus to term. The worst prognosis appears to occur when there is a prior fetal loss and high anti-cardiolipin antibodies. Treatment options for APS include

  • low dose aspirin (81 mg per day) starting prior to pregnancy. Rationale for this treatment is based on the theory that a relative decrease in prostacyclin is the cause for thrombosis. Aspirin at these low doses has the effect of increasing the prostacyclin to thromboxane (its natural competitor) ratio to enhance the effect of prostacyclin.
  • prednisone (30-60 mg per day) to suppress the immune system. This corticosteroid can have several potentially serious complications. When given during pregnancy for this indication, prednisone has been associated with preterm premature rupture of the fetal membranes, preterm delivery and pregnancy induced hypertension. This medication should only be given by physicians experienced in its use for this indication and typically in a research setting.
  • heparin (15,000 units per day in the first trimester after fetal viability is seen on ultrasound, 20,000 units per day starting in the second trimester). Typically the aPTT test is used to monitor heparin dosing but these test results are abnormal in APS so cannot be used. Use of heparin is based on the theory that decreased levels of activated protein C may be responsible for the thrombosis seen, and acts as an anti-coagulant. There appears to be small risk for serious morbidity with the use of heparin, however, until there is much more experience with this medication I believe it is prudent to administer this treatment in a research facility with considerable experience in its use.
  • Immunoglobulin (Ig) therapy, with intravenous injections of Igs, has been used for several decades in the treatment of immunodeficiency conditions and more recently in the treatment of autoimmune disorders. The mechanism of action is not known, the dosing is not standardized for recurrent pregnancy loss, and this treatment is very expensive. On the other hand, early experience with this treatment has been very encouraging. I believe that it is important that this treatment should be administered in a research facility until it is better understood.

Alloimmune dysfunction resulting in recurrent pregnancy loss has also been proposed. Allogeneic antigens are molecular structures that occur in different members of the same species and have the ability to elicit an immune response. Normally, a person will reject dissimilar (non-self) tissues or structures from the body using the immune system. In pregnancy, the placenta and growing embryo are not entirely "self" but rather are a result of both the maternal and paternal genetic heritages (referred to as a semi-allograft). The placenta (and pregnancy) has a "privileged" relationship with the pregnant woman that allows for it to escape rejection. The mechanism for this privilege is not known.

There have been several interesting and complex theories attempting to describe how the normal pregnancy achieves its privileged status in the maternal uterus. Thus far, none of these theories has been generally accepted and proven. Some of the theories are based on

  • increased sharing of HLA types (genes encoding antigens that distinguish and mark tissue as "self") within the maternal and paternal chromosomes. With increased sharing the placenta may not trigger the production of special blocking antibodies which confer privilege
  • decreased numbers of blocking factors that normally allow the placenta to be retained as a privileged site, either due to increased HLA sharing or other factors
  • decreased numbers of natural suppresser cells in the uterus, which may control the activity of the natural killer cells and allow for placental survival within the uterus

The diagnosis of alloimmune recurrent pregnancy loss is one of exclusion. That is, when all other tests have been performed and the findings have come back normal then some of those with "unexplained" losses are thought to fall into this category.

Several physicians refuse to treat alloimmune recurrent pregnancy loss since there are no direct diagnostic tests, treatment options are expensive and their benefits are largely unproven, and treatment options potentially involve risk. I think that it is prudent to limit treatment to a research facility with expertise in these therapies. Having said this, I can honestly recall from my own experience several couples with no abnormal findings in their testing who decided to undergo this "experimental" treatment and surprisingly went on to deliver at term. It does appear that this treatment can be beneficial in some subsets of patients, its just not clear how to predict which patients will benefit. Also, you must consider that there is reportedly up to a 60-70% chance of carrying a pregnancy to term even after 3 spontaneous abortions without treatment.

The two main treatment options include

  • unified leukocyte (white blood cell, WBC) immunization with paternal or donor blood cells, using 200-300 million mononuclear cells from the isolated buffy coat of blood, once the woman is pregnant and prior to 6 weeks gestation on one occasion only
  • immunoglobulin (Ig) therapy, with intravenous injections of Igs. The mechanism of action is not known, the dosing is not standardized for recurrent pregnancy loss, and this treatment is very expensive.

With treatment, viable pregnancy rates of 70-80% have been reported in uncontrolled studies. In my experience, better candidates for this treatment are couples who have no other treatment options available and are willing to commit themselves to the time, energy (especially emotional) and money required to pursue experimental techniques.


Recurrent Pregnancy Loss: Miscellaneous Causes

The unusual causes for recurrent pregnancy loss that are not included in other sections include:

  • debilitating maternal disease states
  • substance abuse with either heavy smoking, alcohol or drug abuse
  • irradiation or exposure to chemical toxins
  • medications known to be teratogenic

Any life threatening maternal disease can compromise reproductive performance either through an ovulatory dysfunction or immunologic disorder. Additionally, women with insulin dependent Diabetes Mellitus who are in poor control have a greater spontaneous abortion rate while those in good control most likely do not have an increased rate. The glycosylated hemoglobin level is a reasonably good assessment of longer term control, and several reports agree that spontaneous abortion rates increase as the glycosylated hemoglobin becomes increasingly abnormal (especially when greater than 3-4 standard deviations over the mean).

Substance abuse is associated with spontaneous abortion. Cigarette smoking is associated with an increase in chromosomally normal spontaneous losses, implying a direct effect on the fetus. Alcohol abuse has been associated with spontaneous abortion if in high quantities, but results within this literature on alcohol are occasionally conflicting (generally excessive consumption is drinking at least 2-3 times per week). Illicit drug abuse affects ovulation and can result in an ovulatory dysfunction. Little is known about the early effects of these drugs on pregnancy and their association to spontaneous abortion.

Industrial or environmental toxins associated with recurrent pregnancy loss include arsenic, benzene, ethylene oxide, formaldehyde, and lead. There has been a concern especially among health care professionals regarding anesthetic gases and miscarriage, with mixed findings in the literature making it prudent to avoid routine intense exposure if possible. Irradiation during diagnostic studies with a total exposure of less than 10 rads is thought to confer only a small increase in risk of spontaneous abortion.

Medications taken during pregnancy should be reviewed with an obstetrician. The current understanding of the effect of drugs on pregnancy include

  • from the time of conception (about day 14 of a 28 day cycle) until about 2 weeks later (31 days gestation in a 28 day cycle) harmful effects of drugs typically have an "all or none" effect on pregnancy. This means that if there is a damaging effect then the pregnancy is generally lost while if the pregnancy goes on then there are typically no anomalies.
  • the classic teratogenic period is from 31 days gestation (in a 28 day cycle) to 71 days gestation (about 10 weeks gestation), during which time damage can occur to specific organs as they form and fetal anomalies may result.
  • after the classic teratogenic period there is still significant development of the brain, which is possibly affected by medications


The FDA (Food and Drug Administration) uses 5 categories of labeling for drugs in pregnancy, including

  • "A": controlled studies in women fail to demonstrate a risk to the fetus in the first trimester, and the possibility of fetal harm appears remote
  • "B": animal studies do not indicate a risk to the fetus and there are no controlled human studies, or animal studies do show an adverse effect on the fetus but well controlled studies in pregnant women have failed to demonstrate a risk to the fetus
  • "C": studies have shown the drug to have animal teratogenic or embryocidal effects, but no controlled studies are available in women or no studies are available in either animals or women
  • "D": positive evidence of human fetal risk exists, but benefits in certain situations (eg., life threatening situations or serious diseases for which safer drugs cannot be used or are ineffective) may make use of the drug acceptable despite the risks
  • "X": studies in animals or humans have demonstrated fetal abnormalities, or evidence demonstrates fetal risk based on human experience, or both and the risk clearly outweighs any possible benefits


Visit Dr. Daiter's website:

Related Videos
Fertility treatment challenges for Muslim women during fasting holidays | Image Credit:
The importance of maternal vaccination | Image Credit:
Matthew Zerden, MD
Marci Bowers, MD | Image Credit:
Angela Dempsey
Haywood Brown, MD | Image credit: © USF Health
Beth Garner, MD, MPH
Related Content
© 2024 MJH Life Sciences

All rights reserved.