Increased risk of developing PCOS using valproic acid in adult women w/bipolar disorder

From MindLinc Wiki

Context: Polycystic ovarian syndrome (PCOS) is a disorder characterized by irregular menstrual cycles and excess androgens that occurs in approximately 5% of women in the United States . It is highly associated with both obesity and insulin resistance. The source of the menstrual dysfunction is anovulation. The excess androgen state is typically manifested as hirsutism, acne, or male pattern baldness.

Additional Context: There have been conflicting reports in the neurology literature about whether exposure to valproic acid increases the risk of PCOS in women treated for epilepsy. More recently, the psychiatric literature has entered the fray.

Clinical Question: Does use of valproic acid increase the risk of developing PCOS in comparison to other mood stabilizers in adult women with bipolar disorder?

Poster: Joffe H. Cohen LS. Suppes T et al. Polycystic Ovarian Syndrome is Associated with Valproate Use in Bipolar Women. Poster Presentation: American Psychiatric Association Meeting 2004

Poster vs. Paper

Pros

• Poster format is similar to an article abstract
• Allows for delivery of new information in a faster venue than standard publication
• Practically, allows authors to get funded to attend their professional conference

Cons

• Not formally peer reviewed
• Does not elaborate on details of study
• Practically, does not carry the same weight for academic advancement as papers

Methods:

Patient Population:

A. Population: 229 women with bipolar disorder between ages 18-45

B. Inclusion Criteria: bipolar disorder, treatment with one or more mood stabilizers for at least three months

C. Exclusion Criteria: prior diagnosis of PCOS, currently using OCP’s

Setting: 16 sites

Design: Cohort Study

• VPA Use = “Exposed”, Non-VPA use = “Unexposed”
• Retrospective assessment of menstrual-cycle patterns
• Cross-sectional collection of serum hormone assays and skin examinations (dermatologists or psychiatrists?)

Outcome Measure: The development of PCOS after initiation of the mood-stabilizer as defined by oligomenorrhea AND hyperandrogenism.

Results:

• 86 of the women taking VPA, 143 using other mood stabilizers
• 9/86 (10.5%) taking VPA developed PCOS, while 2/143 (1.4%) of VPA non-users developed PCOS (p = 0.002), imparting a relative risk of 7.5

Weaknessness, Weaknessess, Weaknessess . . .

• Retrospective assessment of menstrual-cycle patterns, unclear how far back women are going in recall
• Are potential confounders controlled for between the two groups: age, weight, insulin status?
• Small sample sizes
• Is there potential selection bias in how these women wound up on a given treatment regimen?
• Did not separate out people using VPA alone vs. VPA in combination with another agent in comparison
• VPA alone showed 3/45 (6.7%) people developing PCOS – is this still significant compared to non-VPA users?
• Likewise, 6/41 with VPA + another agent had PCOS => the difference between that and VPA alone not significant (14.6% vs. 6.7%; P = 0.23) but would it be if the sample sizes were larger?
• The rate of PCOS in the general population is estimated at 5% vs. 6.7% in VPA-alone group – is this really significant if extended to large numbers of people?

Bottom Line: In this limited study, there is a suggestion that Valproate exposure increases the risk for developing PCOS, particularly in obese women with elevated insulin resistance. However, significant limitations prohibit drawing definitive conclusions from this data.

Future Study: We all know the answer to this . . . Randomized, controlled, prospective, blinded

Presented by Damon Tweedy, MD, on 4/18/05