Invisible Women: Exposing Data Bias in a World Designed for Men

by Caroline Criado Perez

  The intransigence of this feminine riddle has not gone unpunished. Women who had often done little more than manifest behaviours that were out of feminine bounds (such as having a libido) were incarcerated for years in asylums. They were given hysterectomies and clitoridectomies. Women were locked up for having even mild post-natal depression: the grandmother of a friend of mine spent her life in an asylum after throwing a scourer at her mother-in-law. At least one US psychiatric textbook, still widely in use during the 1970s, recommended lobotomies for women in abusive relationships.62

  Of course, we’ve moved on from such inhumane treatment of women. We no longer lock women up and cut out parts of their brains. Instead, we give women drugs: women are two and a half times more likely to be on antidepressants than men.63 This is not to condemn antidepressants: they can be life-changing for people with mental health problems. However, it’s still worth asking why women are so much more likely to be on them, because it’s not simply that women are more likely to seek help. A 2017 Swedish study in fact found that it was men who were more likely to report depression.64 So why are more women being treated with antidepressants? Are women simply more ‘feeble-minded’? Does living in a world in which we don’t quite fit affect our mental health? Or are antidepressants the new (and obviously preferable) lobotomy for women dealing with trauma?

  Freud once believed that hysteria might be linked to historic sexual abuse. He later retracted this theory as it would have implicated too many men to be, in his opinion, credible. But recent research suggests that abuse might be linked to certain types of pain women experience65 – and in the wake of the #MeToo global scandal maybe it’s not so incredible after all.

  The full answer to these questions is beyond the scope of this book. But one possible explanation for at least part of the disparity is that women are being prescribed antidepressants when they are not in fact depressed. Women’s physical pain is far more likely to be dismissed as ‘emotional’ or ‘psychosomatic’. The Swedish study which found that men are more likely to report depression also found that women who have not reported depression are twice as likely as men to be prescribed antidepressants. This chimes with studies from the 1980s and 90s which found that while men who reported pain tended to receive pain medication, women were more likely to receive sedatives or antidepressants.66 A 2014 study which required healthcare providers to make treatment recommendations for hypothetical patients with lower back pain similarly found that female patients were significantly more likely to be prescribed antidepressants than men.67

  It seems that Yentl syndrome may be at play again here: it is striking that so many of the stories women tell of undiagnosed and untreated pain turn out to have physical causes that are either exclusively female diseases, or are more common in women than in men. Women are almost twice as likely to have irritable bowel syndrome as men68 and three times more likely to experience migraines69 (a condition about which we know very little despite it being chronic, often deeply debilitating and affecting 37 million Americans70 and one in eight people in the UK71). In fact, many clinical pain conditions are substantially more prevalent in women than men,72 and several studies over the past decades have shown that women are more sensitive to pain than men (which sheds a particularly cruel light on the finding that women are less likely to receive painkillers).

  There is also mounting evidence that men and women may experience pain differently. A woman’s pain sensitivity increases and decreases throughout her menstrual cycle, ‘with skin, subcutaneous tissue, and muscles being affected differently by female hormonal fluctuations’.73 An animal study which found that males and females use different types of immune cells to convey pain signals may provide the beginnings of an answer as to why74 – although only the beginnings: sex differences in pain remain an underresearched area and even what we do know is not widely dispersed. Dr Beverly Collett, who until she retired in 2015 was a consultant at Leicester’s pain management service and chair of the Chronic Pain Policy Coalition, told the Independent that the average GP ‘has no idea that drugs such as paracetamol and morphine work differently in women’.75

  Even if they are treated for their pain, women routinely have to wait longer than men to receive that treatment. A US analysis of 92,000 emergency-room visits between 1997 and 2004 found women had longer waiting times than men,76 and a study of adults who presented to a US urban emergency department between April 2004 and January 2005 found that while men and women presented with similar levels of pain, women were less likely to receive analgesia and women who did receive analgesia waited longer to receive it.77 A US Institute of Medicine publication on chronic pain released in 2011 suggested that not much has changed, reporting that women in pain face ‘delays in correct diagnosis, improper and unproven treatments’, and ‘neglect, dismissal and discrimination’ from the healthcare system.78 In Sweden a woman suffering from a heart attack will wait one hour longer than a man from the onset of pain to arrival at a hospital, will get lower priority when waiting for an ambulance, and will wait twenty minutes longer to be seen at the hospital.79

  The reality that female bodies are simply not afforded the same level of medical attention as male bodies is often brushed aside with the riposte that, on average, women enjoy more years of life than men. But while it is true that female life expectancy remains a few years longer than male life expectancy (although that gap is narrowing as women’s lives have become less prescriptive and occupational safety in male-dominated jobs has become more stringent), there is evidence to suggest that the female mortality advantage isn’t exactly secure.

  A 2013 paper that examined trends in US mortality rates from 1992-2006 in 3,140 counties reported that even as mortality decreased in most counties, female mortality increased in 42.8% of them.80 And while men’s years of good health have increased in line with their longevity, both women’s longevity and active years have increased at a much lower rate: thirty years of US health data showed that, while women live on average five years longer than men (in Europe it is 3.5 years81), those years are spent in ill health and disability.82

  The result is that US women no longer have more active years than men,83 despite their longer lives, and while women account for 57% of US citizens aged over sixty-five, they make up 68% of those who need daily assistance.84 In 1982 both men and women who lived to eighty-five could expect two and half further years of active healthy life. For women, that figure hasn’t changed, but an eighty-five-year-old man alive now can expect to be active and healthy until he’s eighty-nine. The trend of increasing longevity and good health amongst men can also be found in Belgium85 and Japan.86 A WHO paper into women’s health in the EU reported that in 2013, ‘even in countries with some of the highest overall life expectancy in the Region, women spent almost 12 years of their life in ill health’.87 And, yes, it would be nice to have some sex-disaggregated data on why this is happening.

  A particularly troubling side effect of Yentl syndrome is that when it comes to medical issues that mainly or only affect women, you can forget about including women in trials because here the research is often lacking altogether.

  Premenstrual syndrome (PMS) is a collection of symptoms that can include among other things: mood swings, anxiety, breast tenderness, bloating, acne, headaches, stomach pain and sleep problems. PMS affects 90% of women, but is chronically under-studied: one research round-up found five times as many studies on erectile dysfunction than on PMS.88 And yet while a range of medication exists to treat erectile dysfunction89 there is very little available for women, to the extent that over 40% of women who have PMS don’t respond to treatments currently available. Sufferers are still sometimes treated with hysterectomies; in extreme cases, women have tried to kill themselves.90 But researchers are still being turned down for research grants on the basis that ‘PMS does not actually exist’.91

  Period pain – dysmenorrhea – similarly affects up to 90% of women,92 and according to the American Academy of Family Physicians it affects th
e daily life of around one in five women.93The level of pain women experience on a monthly basis has been described as ‘almost as bad as a heart attack’.94 But despite how common it is and how bad the pain can be, there is precious little that doctors can or will do for you. A rare 2007 grant application for research into primary dysmenorrhea described its causes as ‘poorly understood’ and treatment options as ‘limited’.95 The prescription medications which are available have serious possible side effects and are by no means universally effective.

  When I went to my (male) doctor about period pain that wakes me up at night and leaves me in a moaning foetal position in the daytime, he more or less laughed me out of the room. I haven’t bothered trying again. So imagine my joy when I read about a 2013 study that seemed to have found a cure. The ‘primary outcome’ of a double-blind, randomised, controlled trial of sildenafil citrate, was, ladies, you may want to sit down for this: ‘total pain relief over 4 consecutive hours’, with ‘no observed adverse effects’.96 Imagine.

  Created in 1989, sildenafil citrate is the medical name for Viagra. In the early 1990s, the drug was being tested as a heart-disease medication.97 It turned out not to be great at that, but one thing participants did report was an increase in erections (yes, all the trial participants were men). Total erectile dysfunction affects between 5-15% of men depending on age,98 with about 40% of men experiencing it to some degree – and so naturally the researchers were keen to explore this alternative use for their drug. By 1996, sildenafil citrate had been patented in the US and by March 1998 it was approved by the FDA.99 A happy ending for men, then.

  But what if the trial had included women? The outcome of the 2013 study is suggestive. The trial had to be stopped because the funding ran out, meaning the researchers did not meet their sample size and therefore could not confirm the primary hypothesis. They called for ‘larger studies of longer duration, likely multi-center’ to confirm their findings.

  These studies have not happened. Dr Richard Legro, who led the study, told me he applied twice to the NIH for funding ‘to do a longer and larger study and also to compare sildenafil to the standard of care, a non-steroidal anti-inflammatory agent’. He was rejected both times. In each case, the grant ‘was deemed to be in the lower half of grants submitted’. It wasn’t even reviewed. Legro tells me that the comments he received ‘indicated that the reviewers did not see dysmenorrhea as a priority public health issue’. They also didn’t ‘fully understand clinical trial design of dysmenorrhea trials’. When I ask him if he thinks he will ever get funding, he says, ‘No. Men don’t care or understand dysmenorrhea. Give me an all-female review panel!’

  The failure of pharmaceutical companies to step in here and capitalise on what is surely a gold-plated commercial opportunity may seem baffling, but it’s quite possibly just another data-gap problem. In an email, Legro told me that, for cost reasons, the pharma industry ‘doesn’t usually fund investigator-initiated projects’, particularly of drugs that are available generically. And this may be where the data gap comes in: there simply isn’t much research done on dysmenorrhea,100 which makes it difficult for pharma companies to know exactly how much money could be made on such a drug – and therefore makes it harder for them to decide to fund trials. Especially if the people making the decisions happen not to be women. Legro also suggested that pharma companies may not want to risk doing tests in women in case of negative findings that would endanger the use of sildenafil in men. In short, it seems that pharma companies may in fact not see this as a gold-plated commercial opportunity. And so women carry on being incapacitated by pain on a monthly basis.

  Male-dominated funding panels may also explain why we have so few drugs available for uterine failure. Every day 830 women around the world die due to complications during pregnancy and childbirth101 (in some African countries more women die annually from childbirth than at the height of the Ebola epidemic102). Over half of these deaths are explained as being a result of problems with contractions, often because the contractions are too weak for the woman to give birth. The only medical treatment available for women whose contractions aren’t strong enough is the hormone oxytocin, which works about 50% of the time. Those women go on to give birth vaginally. Women who don’t respond to oxytocin are given an emergency caesarean. In the UK weak contractions are the reason given for a majority of the 100,000 emergency caesareans carried out each year.

  We currently have no way of knowing which women will respond to oxytocin, which clearly isn’t ideal: all women, including those for whom it will result in a pointless and harrowing delay, have to go through the process. This happened to a friend of mine in 2017. After being in hospital in excruciating pain for two days (on her own for much of it as her partner had been sent home), she was only 4 cm dilated. Eventually she was taken off for a C-section, and she and the baby were fine. But the experience left her traumatised. She had flashbacks for the first few weeks after she gave birth. When she talks about the internal exams and procedures, she describes it as a violent assault. It was, she says, brutal. But what if it didn’t have to be this way? What if they’d known from the beginning that she was going to need a caesarean?

  In 2016 Susan Wray, a professor of cellular and molecular physiology at the University of Liverpool, gave a lecture to the Physiological Society.103 Wray is also the director of the Centre of Better Births in Liverpool Women’s Hospital and she explained that recent research revealed that women with contractions that were too weak to give birth had more acid in their myometrial blood (the blood in the part of the uterus that causes contractions). The higher the levels of acid were, the higher the likelihood a woman would end up needing a caesarean, because oxytocin isn’t, it turns out, that effective on women with an acidic blood pH.

  But Wray didn’t simply want to be able to predict the need for a caesarean. She wanted to be able to avoid it. Together with her fellow researcher Eva Wiberg-Itzel, Wray conducted a randomised control trial on women with weak contractions. Half of them were given the usual oxytocin; half were given bicarbonate of soda, and then given the usual oxytocin an hour later. The change was dramatic: 67% of women given just oxytocin went on to give birth vaginally, but this rose to 84% if they were given bicarbonate of soda an hour before. As Wray pointed out, the bicarb dose wasn’t tailored to body weight, it wasn’t tailored to the amount of acid in the blood, and the women weren’t given repeated doses. So the efficacy could turn out to be even higher.

  This finding could not only be transformative for the tens of thousands of women a year who have what could turn out to be unnecessary surgery (not to mention saving the NHS a substantial amount of money). It could save women’s lives in countries where caesarean sections are risky or not readily available – not that you have to live in a low-income country for a C-section to be risky: you could just be a black woman living in the United States.104

  The US has the highest maternal mortality rate in the developed world, but the problem is particularly acute for African Americans. The World Health Organization has estimated that the death rate of black expectant and new mothers in the US matches that of women in much lower-income countries like Mexico and Uzbekistan. Black women in America have worse health outcomes overall than white women, but when it comes to pregnancy and childcare the comparisons score off the charts: African American women are 243% more likely than white women to die from pregnancy and childbirth-related issues. And it’s not just because African Americans tend to be poorer: a 2016 analysis of births in New York City found that ‘black college-educated mothers who gave birth in local hospitals were more likely to suffer severe complications of pregnancy or childbirth than white women who never graduated from high school’. Even global tennis superstar Serena Williams is not immune: in February 2018 she revealed that she had almost died following an emergency C-section.105 African American women also have higher rates of caesarean section and a 2015 study from Connecticut found that – even when controlling for socio-economic status – black
women were more than twice as likely to have to return to hospital in the month following surgery.106 So Wray’s research could be transformative here.

  But it looks like we aren’t going to see the fruits of her labour any time soon. When Wray discovered that the British Medical Research Council was offering funding for research that would benefit low- and middle-income countries, she decided to apply. And yet, despite all the data about how dangerous weak contractions can be, she was turned down. The research was ‘not a high enough priority’. So currently we have only one treatment for women with weak contractions, and it doesn’t work half the time. Compare this, Wray says, to the around fifty drugs available for heart failure.

  The evidence that women are being let down by the medical establishment is overwhelming. The bodies, symptoms and diseases that affect half the world’s population are being dismissed, disbelieved and ignored. And it’s all a result of the data gap combined with the still prevalent belief, in the face of all the evidence that we do have, that men are the default humans. They are not. They are, to state the obvious, just men. And data collected on them does not, cannot, and should not, apply to women. We need a revolution in the research and the practice of medicine, and we need it yesterday. We need to train doctors to listen to women, and to recognise that their inability to diagnose a woman may not be because she is lying or being hysterical: the problem may be the gender data gaps in their knowledge. It’s time to stop dismissing women, and start saving them.