Iron Helps Infertility: Study

ByABC News
October 30, 2006, 6:24 PM

Oct.31, 2006— -- Iron deficiency may be linked to infertility, suggests a new study in the Obstetrics & Gynecology journal by researchers from Harvard University. The study focused on more than 18,500 female nurses who were trying to get pregnant.

The women answered detailed questionnaires about their health, including diet and other aspects of their lifestyle. Doctors found that women who reported taking iron supplements were 40 percent less likely to have infertility problems due to ovarian failure.

Given these findings, the obvious assumption is that taking iron supplements would help prevent infertility or perhaps even improve fertility in women who are currently struggling to get pregnant.

But don't run out and purchase those iron pills just yet.

The new research is based on the long-running Nurses' Health Study, a large study of female nurses in which they respond to a survey every two years whose questions touch on breast cancer, pregnancy, the number of Tylenol they take, how much soda they drink, and more.

The survey allows scientists to examine the relationship between different health and lifestyle factors, such as whether drinking coffee is linked to diabetes (it isn't, says the study) or whether iron pills help ward off infertility.

The only problem is that the Nurses' Health Study is an observational study. No one told these women how much coffee to drink or whether to take iron pills; they made the decisions on their own and scientists observed the outcome.

If women who take iron supplements also happen to have other traits that are linked to good health and fertility, one could mistakenly conclude that it's the iron supplements and not these other traits that affect fertility. Indeed, the scientists have found in this study that women with low iron or who did not take iron supplements also tend to weigh more, smoke more and exercise less -- all variables that can negatively affect a woman's ability to get pregnant.

Researchers try to take these variables into account in their statistical analysis, but it is difficult to account for all relevant behavioral variables. Because of this, two variables that appear to be related in an observational study can turn out not to be related in the way the research indicated.