Here’s the scapegoat unhappy spouses have been waiting for: According to a paper out last week by Michael R. Gillings, premenstrual syndrome (PMS) can ruin a marriage. Gillings claims that PMSing women—in infertile couples in particular—may use feelings of “animosity” as well as risk-seeking and competitive behaviors to leave their husbands and find someone new. I’m sorry to say, but the evidence in favor of this hypothesis is thinner than Always Infinity menstrual pads.
The first problem with Gillings’ paper is that it does not define PMS symptoms, and in fact, regularly confuses PMS and premenstrual dysphoric disorder (PMDD), by citing papers on PMDD while making claims about PMS. PMDD is a serious disorder, which affects 5 to 8 percent of women, compared to PMS affecting around 80 percent. PMS, broadly speaking—that is, any negative change in any physical, behavioral, or psychological factors during the premenstrual phase—is frequent for people who have ovulatory menstrual cycles. You tend to need ovulation to get a rise in progesterone in the second half of your cycle, and you need the rise in progesterone to get the decline in progesterone that leads to a number of these symptoms. The severity doesn’t seem to increase with progesterone in within-population comparisons, but cultural and population-specific differences in PMS symptomology and incidence suggest ties to cultural context and environment. Many papers have demonstrated that there is a full spectrum of PMS symptoms from minor, to distressing, to completely impeding normal function. And there’s a hefty literature suggesting at least some of this variation is related to one’s sensitivity to hormones, influenced by both genetic and environmental factors.
The next curious thing about Gillings’ paper is its long section devoted to convincing the reader that PMS has a genetic origin. As my fellow University of Illinois professor Charles Roseman said over email, “Find me something that varies that is not heritable.” That is, the fact that PMS is heritable and variable tells us nothing about whether women with PMS have more children than those who don’t, and this is the true test for adaptation. This crucial point—the third and most crucial condition for natural selection—is absent from the paper.
The paper hits another major bump when it makes the claim that PMS causes marital distress, as the papers it cites to support this notion say nothing of the kind. The first, by Patricia C. Coughlin in 1990, demonstrates an association between PMS, married state, and career: those women who were married and had a job (not by choice) had more severe symptoms. The main finding of the paper suggests that economic variables that force women into jobs they don’t want cause emotional distress, and in fact marital distress only influences PMS symptoms once the author controlled for all the other major life stressors. Most important to Gillings, Coughlin’s paper suggests the effect is the reverse of what he claims, in that life stressors, marital distress, and career issues worsen PMS—not that PMS worsens these life variables.
Two of the other main papers cited by Gillings to support an association between PMS and poor marital outcomes—by Halbriech in 2004 and Graze and colleagues in 1990—aren’t even on these topics. Halbreich’s paper is theoretical, meaning it isn’t reporting on new data. The bulk of the paper is taken up with important discussions on distinctions between PMS and PMDD. The paper by Graze and colleagues does mention that, in their study of women with PMDD (again, not PMS), family and work responsibilities felt to them like greater challenges. But this only supports Coughlin’s similarly miscited work that demonstrates relationships between premenstrual symptoms and issues both inside and outside the home.
Let me continue to direct my PMS-fueled “animosity” at this paper, as there are major, problematic assumptions that went into even thinking this hypothesis was worthy of exploration. The author assumes that pair bonds are not only always heterosexual, but always motivated by childbearing decisions. We don’t see this in the animal kingdom, and it’s a safe bet these weren’t the only considerations ancestral humans made either. Further, as Gillings himself points out, ancestral humans cycled far less frequently, having maybe 50-100 reproductive cycles per lifetime to our 400 or more.
There’s also a problem with making these suppositions given the very limited dataset we have on PMS and PMDD: the majority of the literature is on Western, white women. Part of this is related to a biased literature and historical context that doesn’t pay much attention to the health of women of color from any country; another part, though, is that it seems like many forager and agricultural societies don’t report experiencing many PMS or PMDD symptoms. As I’ve written in the past, these data together suggest that PMS is bioculturally-bound, rather than being a syndrome that is motivated only by biology or culture. Hormone concentrations are very strongly tied to developmental conditions, such that Westerners, who sit around and eat a lot more than the human norm, have much higher hormones than those in more agricultural or forager contexts.
As the occasional sufferer of PMS symptoms and one half of an infertile partnership of 14 years—my husband is infertile from cancer treatment as a child—I’m glad to know that I need not worry that any occasional blueness might drive us apart.