Abstracts

A prospective study of postmenopausal women found those with any prior exposure to anticonvulsants (AEDs) had a 2.3 fold increase in foot fractures (Seeley, et al., 1996). Several retrospective studies show an association between AED use and osteopenia. Hahn (1993) summarized the risk factors for AED bone loss: polytherapy, long term therapy, low vitamin D, immobility, older age, and institutionalized patients.
Women with untreated hypothalamic amenorrhea develop progressive bone loss (Biller, et al., 1991). Women with polycystic ovary syndrome however do not show bone loss as bone mass is positively correlated with androgen levels (Kasperk, et al, 1989). Reproductive endocrine disorders are common in women with epilepsy (Herzog, et al., 1986).
We wished to survey risk factors for fractures in women taking AEDs and specifically address amenorrhea as a potential risk.
An anonymous survey was offered to every person in the Neurology Clinic at Henry Ford Hosptial for the month of December 2002. The data from the survey was entered into a database by author SD and checked by author LS. Surveys missing critical information were excluded from final analysis. We identified those taking AEDs with a history of seizures or epilepsy, those with AED exposure for other reasons, AEDs used, supplement use, fracture history, sex, age, race, and history of amenorrhea for more than 6 months unrelated to pregnancy, menopause or surgical menopause. Characterization of nonresponders was not performed. Statistics used were chi square, student t test, mean and standard deviation.
Two hundred and twenty patients completed the questionairre from a total of 941 outpatients seen in December 2002. 34 surveys were excluded because critical information was missing. 63% of surveys were completed by women. 51% of women reported exposure to AEDs. 41 women took AEDs for seizures and 27 for other reasons. An average of 2 AEDs had been used by each woman. 53% of women responders were Caucasian and 36% African American. Fracture rates were similar between men (43%) and women (39%). Fracture rates for women with amenorrhea and AED exposure (70%) and for caucasian women exposed to AEDs (54%) were significantly higher than other risks (p[le]0.025 and p[le]0.01 respectively). There was no difference in fracture rates between women with and without amenorrhea and no AED exposure (p[le]0.10). Age (p=0.8), supplement use (p[le]1.0), exercise (p[le]1.0), and number of AEDs used (p=1.0) did not impact fracture risk.
These findings suggest women with AED exposure and a history of amenorrhea or of Caucasian descent are at greater risk for fractures. This relationship may not be causally related and there are clearly limitations to this survey study without medical record review. We can make no comment on the etiology of amenorrhea as this was not part of the survey, nor can we comment on issues such as duration of treatment. Regardless, these possible relationships deserve further investigation with a prospective study.