Defining Infertility, Part 1

I was recently asked by WebMD to comment on a newly published study that highlights a possible link between epilepsy, its treatment, and infertility, but a finding in the study has broader implications for all patients trying to conceive.  First, a little about the study, and afterwards I will show how it applies to everyone trying to build a family.

Epilepsy is a medical disorder characterized by seizures, which are paroxysmal episodes of disorganized neurological activity.  The manifestations of seizures can be quite varied: loss of consciousness, confusion and muscle spasms are some examples of how seizures can appear.  In this recent study, the authors reported that patients seeking treatment for seizure disorders had a high rate of infertility when compared to the general fertility rate of the community in which the study was performed.  Although there were methodological problems with that study that weaken the observed link between epilepsy and infertility, the authors observed that of the patients that entered into their study, the majority who became pregnant did so in the first or second year of follow-up and that after three years very few additional patients became pregnant.  Not coincidentally, this observation is the same for everyone and it is the basis for how infertility is defined.

General “fecundability,” or the probability that conception will occur in any given menstrual cycle, is approximately 20%.  Another way of saying that is that, on average, the probability of conception is about 1 in 5 in any given month.  This is not a large number, and most patients don’t realize how “uncommon” it is to conceive, at least on a per attempt basis.  In fact, this skewed observation is a good example of the concept of “sampling bias” in statistics: because most people only hear when their friends are pregnant, it sounds like everyone is getting pregnant except them.  What most women don’t hear is when their friends don’t get pregnant (except maybe for a few whose friends are really, really forthcoming about their personal lives.)  If we heard about all of the months of trying that didn’t result in a pregnancy, we’d hear, “I didn’t get pregnant this month,” a lot more than we would hear, “I’m pregnant!”  By hearing only the times that a pregnancy has occurred, and by not collecting information on all of the times it didn’t, we develop a biased perspective on how easy it is to get pregnant.  The frequent reminder that one’s friends are “always” getting pregnant when a woman who is trying has not yet conceived can be extremely disheartening on a personal level in spite of feelings of joy for a friend’s happiness.  Sometimes, only more time is needed for conception to occur, but after a trial period, if conception has not occurred, medical evaluation may be needed.

The estimate of a 20% chance per month is a statistical calculation that does not apply to any particular individual.  For example, the chances of success are age-dependent, so younger patients will likely have better chances than older patients.  The reason for this is based in biology and will be the subject of a future column.  This statistic describes events that occur to the entire population, not any specific person in that population.  (See: Infertility care: I am not a statistic for why it is important to understand this concept.)

If we use pure theoretical math rather than real life, we can get an approximate idea of what will happen over the course of a year.  Starting with a group of 100, after one month, approximately 20 will conceive and 80 will not.  (20% of 100 = 20.  100 – 20 = 80.)  In the second month, approximately 16 patients will conceive and 64 patients will remain.  (20% of 80 = 16.  80 – 16 = 64.)  In the third month, approximately 13 patients will conceive and 51 patients will remain.  Obviously, we can’t have a part of a person, so I am rounding the numbers to make it easier to follow.  (20% of 64 = 12.8.  12.8 is approximately 13.  64 – 13 = 51.)  In the fourth month, 11 more patients will be successful.  (20% of 51 is 10.2, which is approximately 11.  51 – 11 = 40.  In the fifth month 8 will conceive and 32 will remain (20% of 40 = 8, 40 – 8 = 32).  The sixth month: 7 will conceive, 25 will remain (20% of 32 is 6.4, or approximately 7.  32 – 7 = 25.)  And so on.  At the end of one year, approximately 93 patients in this example will have conceived, and 7 will remain.  If this calculation is continued ad nauseum, eventually less that one person would remain.  It would be a perfect world if this theoretical example were true to life, but clearly it is not, and this is yet another good example of why statistics are useful to describe a group, but don’t tell you about the group’s individual members.

If we actually follow a large group of individuals who are trying to conceive, we find that the “more fertile” patients conceive rather quickly, leaving “less fertile” patients behind.  Eventually, the curve would flatten as fewer women become pregnant as time goes on, and it is from this effect that we define the infertile population.  Once the curve flattens out, very few additional patients will get pregnant, and rather than ignore their plight, medical evaluation and intervention is warranted.

So the authors of the epilepsy study observed in their population a function that is also true of the general population: after a trial period, the chances of conception for the patients who have not conceived are lower individually than the average chance of the entire population that started the study.  In general, we define infertility as one year of unprotected intercourse without conception.  Because success with treatment for infertility is also dependent on age, we recommend that patients older than 35 seek evaluation after 6 months of trying.

Editor's Note: This post was picked up from RMA of New York's blog.The blog posts do not intend to diagnose, treat, or cure any condition and are not a substitute for consultation with a physician. The postings are presented for educational purposes only.