Why do women experience menopause?

At Research the Headlines, the news stories we cover span the whole range of quality, from excellent reporting about research on exercise and academic performance to news stories with completely made-up research. A colleague recently alerted me to an article in the Guardian, Menopause: nature’s way of saying older women aren’t sexually attractive?, that exemplifies a more nuanced problem with media reporting of research: misrepresenting the experts.

What does the research say?

Only humans and a few species of whale experience menopause, the end of female fertility that occurs significantly before the end of female lifespan. This is an evolutionary puzzle because we don’t yet understand why this pattern was favoured by natural selection—why is it that modern humans are descended from women who experienced menopause, rather than women who stayed fertile until the end of their lives?

One potential solution to this puzzle is the Grandmother Hypothesis, first suggested by George Williams in his seminal 1957 paper on senescence, which suggests that, after a certain age, women could make a greater contribution to their reproductive success by investing in existing children and grandchildren, rather than by investing in new children. This hypothesis led to a huge amount of research, both supporting and refuting this idea.

What did the media say?

The article in the Guardian gets the science right, but makes some serious missteps in reporting the views of the experts. I spoke with Dr Rebecca Sear, who was extensively quoted in the article. Dr Sear is an evolutionary demographer at the London School of Hygiene and Tropical Medicine and a world-leading expert on how kin influence fertility. She alerted me to several problems with the way her views were represented in the piece.

Dr Sear’s views are often presented in quotation marks, implying that they are direct quotes of things she said. However, the interview took place face-to-face and was not recorded.  Snippets from the conversation have been used throughout the piece to tell a story, but these quotes are misleading, taken out of context, or fabricated. For example, the article ends with:

Among the explanations out there, Sear, like many researchers in the field, does at least find the grandmother hypothesis plausible. “I do also like hypotheses that give women some agency in evolution,” she says. “It is encouraging that, hypothetically, there might be a use for me after menopause.”

Given that Dr Sear doesn’t have children and is, therefore, unlikely to do much grandmothering herself, this last quote seemed implausible to me. Indeed, Dr Sear explained that she was very reluctant to discuss with the journalist how she ‘felt’ about the research and preferred to speak only about the science. However, Dr Sear told me, “she pushed very hard on this, saying that she knew scientists really didn’t like answering such questions, and gave her own personal views – which were essentially the final quote she has attributed to me.”

Other statements were portrayed as coming directly from Dr Sear, but were actually paraphrased versions of leading questions that she simply agreed with. For example, Dr Sear wrote to me:

During our conversation, she [the journalist] raised the issue of whether research on menopause was coloured by gender bias, and observed that researchers on the grandmother hypothesis were female while those who had come up with the male mate choice argument were men. I agreed with this latter statement (because it is true) yet the article gives the misleading impression that these ideas all originated form me.

A final example of misrepresentation of an expert’s opinion comes from another misleading quotation:

However, with evolutionary research of this kind, which looks at behaviour and culture, theories are so difficult to prove that almost anything can be true, says Sear. “The problem with the evolution of a trait is you can never know for certain. Basically, you can make a model tell you anything.”

Dr Sear explained that this quote is close to something she said, but was taken out of context and twisted to completely change the meaning. Her original point was that empirical analyses of extant human populations, the kind of work that Dr Sear does, can’t tell us much about the evolution of traits. She said that mathematical models are needed to test hypotheses about the evolution of traits and said — later on in the conversation — that such models are dependent on their assumptions. However, the term “assumptions“, simply means the parameters of the model, which are informed by theory and empirical data, not the colloquial definition of assumption, meaning something that is accepted without proof.

What can we do?

So how can we address the problem of mischaracterisation of expert views? First, journalists can run finished articles past their sources before publication to ensure the accuracy of quotes. While the journalist in this case did promise to do that, she never did send a draft to Dr Sear. Journalists are usually under intense time pressures, so experts can help by responding to such requests as quickly as possible. Second, journalists should only use quotation marks when providing direct quotations for which there is documentation (e.g., recordings) or clear assent from the source. Experts can help to prevent misunderstandings by learning to provide clear soundbites that reflect their views and only discussing issues with which they are happy to go on record. This and other advice for scientists are available at the Science Media Centre’s Publications for Scientists.