The Trouble With Boys Part 2: Why So Many?

When my colleague Dr Abbey Page was conducting her fieldwork with the Agta, a Philippine hunter-gatherer population, she spent a lot of time watching children play – for two reasons: 1) she was interested in how children in a population with no formal school system interact and learn from each other and 2) there are a lot of Agta children around (Agta mothers, on average, have 7.7 children, one of the highest recorded total fertility rates among hunter-gatherers). Whilst doing so, she started to notice something – there were a lot of boys. To be exact, there were 131 boys aged 5 or under, compared to 87 girls, that’s a sex ratio of 1.51 or 151 boys to every 100 girls! The equivalent figure for the UK is 1.05. This begs the question why?

The Agta live in the Northern Sierra Madre Natural Park of the Philippines, where they make their home in the mountainous rainforest and coastal beaches looking out over the Pacific Ocean. They practice spearfishing, the spoils of which they trade for rice with local farmers, supplemented by the gathering of wild foods and some cultivation. Currently, roughly 1000 Agta live in the area of Palanan where our fieldwork was conducted and my colleagues collected data from 915 of them (500 of whom were males, that’s a sex ratio of 1.21!).

As a first step to figuring out why there are so many more males in a population, it is important to look at how sex ratios may vary across different age groups in the population – for example, the reasons why there are more young boys may differ from why there are more old men. The graph below shows what happens when we breakdown Agta sex ratios by age cohort. This graph shows that the male-bias at the population level (which is 1.21), is being driven by skew towards males at younger ages – while it looks like there is also a dramatic skew towards men in the over 55’s, this is probably just an artefact of the small number of individuals (only 41) surviving to this age.

Cohort sex ratios
This graph shows the percentage distribution of males (bottom, yellow) and females (top, green) across Agta age cohorts. The dashed horizontal line indicates a balanced sex ratio; error bars that do not cross this line indicate a significant difference. The numbers at the bottom of each bar are the sex ratios for each age cohort. Reproduced from Page, Myers, et al. (2019).

So why so many young boys?

Many of you might be thinking the obvious answer is that girls are dying – indeed gendercide (either via selective abortion, infanticide, or neglect) is the primary explanation for why we see male-biased sex ratios in many parts of the world today. Anthropologists have often proposed this is the reason sex ratios of often male-biased in hunter-gatherer populations.

Fieldwork by my colleagues and others lead us to believe this was unlikely to be the case in the Agta for two reasons: 1) infanticide has never been documented in the Agta and 2) the Agta, verbally at least, do not express any preference for sons. However, you can’t always trust what people say, particularly when it comes to sensitive subjects; you have to look at what they do. 35 Agta children were observed for 9 hours each; during this time the form childcare they were receiving and who was providing it were recorded every 20 seconds. Using this highly detailed data on childcare, we found that parents also show no difference in their interactions with children based on their gender.

The graph of sex ratios across age cohorts also suggested the idea that females were more likely to die was wrong, because sex ratios decreased rather than increased as the age of cohorts increased. When we looked at mortality in the under 16’s we found that males are more likely to die across this juvenile period (hazard ratio = 1.557, p = 0.011) and while disease-related mortality was equally likely in both sexes, females were less likely to die of non-disease causes than males.

This leaves us with an intriguing alternative explanation – the Agta give birth to more males…

It is generally assumed that the sex ratio at birth is equal, i.e. 1.00 (well actually it is assumed that the sex ratio at conception is equal and that it is “natural” for a few more females to die in utero, though it is rarely asked why this should be!). However, for simplicity, if the sex ratio at birth is equal and more females are dying, then sex ratios should become progressively male skewed (the yellow, green, and light blue lines in the graph below), and vice versa if more males are dying as they are in the Agta (the dark blue line in the graph below). But if the sex ratio at birth is not equal and instead is male-biased, after which more males die, then the pattern of sex ratios across age cohorts will start high and then decline (the orange line in the graph below). If we look back at the sex ratios across age cohorts in the Agta (above), we see a pattern most similar to this orange line – the sex ratio starts high and then falls.

Hypothetical births
This graph shows what happens when 1000 hypothetical, simultaneous births are subjected to five different demographic scenarios across 15 population years. Reproduced from Page, Myers, et al. (2019).

Can the Agta really be having more male infants? My colleagues witnessed 16 births while they were conducting fieldwork, of these 12 were males (i.e. a sex ratio of 3.00). This is far too small a sample size to trust, but when we look at the reported reproductive histories of the Agta women censused they appear more likely to have given birth to more boys (graph A below). The sex ratio of births per year also seem more likely to be male-biased (graph B below).

Scatter plots
Graph A shows the sex ratio of children born to individual Agta mothers – each point is a mother and the colour of the point indicates her age (included because older mothers typically have had more children). Graph B shows the sex ratio of yearly birth cohorts between 1990 and 2014 – the number of births per year is indicated by the size and colour of the point. In both graphs the dashed horizontal line indicates a balanced sex ratio. Reproduced from Page, Myers, et al. (2019).

Putting this altogether we suggest that it is plausible that the Agta sex ratio at birth is substantially male biased as a result of a biological mechanism adjusting the sex ratio at conception and/or the sex ratio at birth in response to higher male juvenile mortality. It is beyond the current scope to go into how they might be doing this, but if you’re interested then we discuss in our open access paper.

Why is this important?

We do not claim that this is the explanation for all male sex ratio biases seen in hunter-gatherer populations; however, we hope to have provided sufficient cause for looking more critically at behavioural-based assumptions regarding male-biased sex ratios in foraging populations and, in the absence of evidence, avoid the further invocation of gendercide. A priori assumptions of female infanticide and/or neglect not only unnecessarily stigmatise groups, but also mean that much needed interventions aimed at improving child outcomes in these high mortality populations are potentially misdirected.

This blog is based on data published in Page, A.E., Myers, S., Dyble, M. and Migliano, A.B., 2019. Why so many Agta boys? Explaining ‘extreme’ sex ratios in Philippine foragersEvolutionary Human Sciences1. COPYRIGHT: © The Author(s) 2019. This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (, which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.

The trouble with boys part 1: the story of two results

Postnatal depression is all too common – the standard figure quoted is that it occurs after about 10% of births, but this was recently found to quite dramatically underestimate rates of the condition worldwide. Postnatal depression is associated with a range of detrimental outcomes for both mothers and their children, and the costs of treating are perinatal mental illnesses are immense – for instance, it was recently estimated to cost the UK’s National Health Service £1.2 billion per year.

It is perhaps not surprising then, that when our new study identifying an easily observable (yet previously largely unrecognised) factor increasing the likelihood of postnatal depression by 174% came out, it received some press attention. I say ‘some’ press attention, in fact it was really rather a lot – the study was featured in at least 131 stories in print media, 7 radio broadcasts, and 3 TV news reports within a few days of publication (we even made People Magazine!).

However, the focus of these reports was not the finding that something health professionals don’t currently consider to be a major risk factor for postnatal depression, and yet document as a matter of course, increases the risk of the condition by 174%. Instead, they all focussed on the finding that something else, granted as easily discernible, increased the likelihood of postnatal depression by 73% – why?

Was this seeming discrepancy due to deliberate spin placed on the findings, either by the University or us the researchers, to grab attention? I’d like to think not – both factors are in the title of the paper and the first lines of the press release…

Instead, I think the reason will become more clear when I disclose that the factor we found to  increase the likelihood of postnatal depression by 174% is the experience of complications during childbirth, whereas the factor increasing the likelihood by a relatively meagre 73% is the giving birth to a son.

The tabloids loved it! The tone of much of the reporting is neatly summed up in the headline ‘Mothers of male infants, take note!’ – as if by dramatically warning mothers, they’ll magically be able to prevent themselves becoming depressed, and thereby avoid the danger of damaging society’s precious male heirs. (And yes, I am aware its statements like that which cause the comments fields on such news reports to fill with accusations that I’ve been “blaming men since birth”!)

Instead of being helpful to new or expectant mothers, it is arguably very harmful to engender fear over something they cannot control – the mechanism underlying both these findings is thought to be an inflammatory immune response, triggered in response to gestating a male foetus or injury during labour. The paper was firmly aimed at health professionals, with a view to increasing their knowledge of the factors which may leave a woman at elevated risk of becoming depressed, so that the limited resources which accompany mental health provision can be more efficiently targeted. Rather than articulating this, much of the media reporting simply fed into a wider, highly problematic narrative surrounding motherhood which emphasises the risks of maternal behaviour towards infants.

How as researchers do we prevent our findings causing unintended harm? We could have not put out a press release, and left academics to find our results for themselves. But this radically reduces the chances of their being noticed by the wider public health community – as it is, our paper came to the notice of the Royal College of Midwives (so mission accomplished!).

A more productive solution seems to be to actively engage with the media, wherever possible, and otherwise to convey the purpose and nuance of research findings to a popular audience ourselves, wresting control of the message of our work. Hence this blog…

Myers, S. and Johns, S.E. (2019). Male infants and birth complications are associated with increased incidence of postnatal depression. Social Science and Medicine 220: 56-64

Two is enough! Postnatal depression stops women having more children…

Women who experience postnatal depression have fewer children than women who don’t, and the worse the depression, the less likely they are to carry on having children. This may seem unsurprising, but until now no one had ever checked. How many children women have is important, not just to the women themselves but also in terms of population structure; the fewer children born, the slower population growth and, past a certain point, the quicker population ageing. Postnatal depression is very common, yet prevention and treatment are poorly funded – the prospect of slowing population ageing may just provide a much needed incentive to address this.

It is often reported that postnatal depression, i.e. depression experienced in the first year after giving birth and usually within 6 weeks, has a prevalence of 13%. However, it is also estimated that more than 50% of cases go undiagnosed in the UK, and in the US as many as 63% of women experience symptoms of depression after giving birth. What is for sure is that not enough is done to prevent it and not enough is known about the consequences of postnatal depression for those women who do experience it. Given so many women undergo dramatic reductions in their emotional wellbeing after childbirth, it should be surprising (although perhaps it isn’t given the dearth of scientific research into female experience) that researchers had never asked what this does to the likelihood a woman will go on to have another child. In fact, without such information having been documented, a prominent evolutionary hypothesis was developed suggesting that postnatal depression is an adaptive mechanism evolved to enable women to save or gain additional resources, thus benefiting females.

Very little in the way of existing data was available to address this knowledge gap, so as part of my PhD under the supervision of Dr Sarah E. Johns and Dr Oskar Burger at the University of Kent I gathered the reproductive histories of around 300 (what evolutionary anthropologists would call) WEIRD postmenopausal women – WEIRD standing for Western, Educated, Industrialised, Rich, and Democratic. I found that women who did not have postnatal depression at their first birth had an average of 2.30 children, while women who did have postnatal depression had 1.96 – this might not look like a big difference, but it is statistically significant and when one considers women must have 2.07 children or more for the population not to age, it starts to look like a lot.

The impact postnatal depression has on female reproductive behaviour seems to be cumulative, either influencing a woman’s physical or emotional capacity to have increasing numbers of children. As the severity of postnatal depression increased in the period after a woman’s first birth, the odds of her having a third birth (rather than a second) decreased (Figure 1). Having two bouts of postnatal depression also made a striking difference to a woman’s childbearing, such women were 76% less likely to have a third child compared to those who had no postnatal depression – this is particularly important in terms of impact at a population level, because women who have postnatal depression at their first birth have a 50% chance of having it at their second as well.

Odds of a third birth based on EPDS b1 with line
Figure 1. The graph shows the odds of having a third birth depending on a woman’s score on the Edinburgh Postnatal Depression Scale (EPDS) at her first birth. As scores on the EPDS increase, depressive symptoms get worse. The dashed vertical line indicates a common cut-off for postnatal depression – scores to the right indicate cases of postnatal depression.

There are a variety of factors well known for their influence on female childbearing. Particularly salient in contemporary WEIRD populations, where childbearing is often ‘delayed’, is a woman’s age; this is because fertility rapidly declines from the late-20’s until it reaches effectively zero in the mid-late 40’s (technological assistance aside). I found postnatal depression has a comparable effect – in terms of the reduction in the chances of having a third birth after the second, every increase of a point on the Edinburgh Postnatal Depression Scale was similar to increasing in age by a year. Having two bouts of postnatal depression was equivalent to experiencing major birth complications.

Why might this be? It is possible that women are simply put off having children by postnatal depression, and this is undoubtedly the case for some women. However, the fact that symptom severity at the first birth decreased the odds of women having a third birth but not a second, suggests there are long term physiological, as well as emotional, effects. Postnatal depression is known to be linked to changes in reproductive hormones and women with a lifetime history of depression are at risk of earlier menopause. Not only are women more likely to have postnatal depression if they’ve already had it, they are more likely to have depression at other points in the future too; thus, women who experience postnatal depression may be physically less able to have more children.

What good does knowing postnatal depression is linked to lower levels of childbearing do? For starters, it sheds light on a previously undocumented consequence of female experience, which is important in its own right. Second, it contributes to ongoing theoretical debates regarding the evolutionary origins of postnatal depression, and depression more generally (and for those interested you can find the discussion generated within the field by these results here and here). Finally, as hinted at the top, on a practical level these results should interest governments struggling to deal with the massive financial implications of population ageing. Population ageing occurs when individuals, on average, do not have enough children to replace themselves in the population, leading to population structures with more old people than there are young people to support them. This presents major issues for health and social security systems, amongst other things, leading governments to seek ways to raise fertility levels. Typically this involves improving practical things like parental leave – while this is obviously to be encouraged, the emotional lives of parents go unconsidered. At the same time, investment in mental health is widely sub-optimum. It was recently calculated that an extra £400 per birth in the UK would elevate perinatal mental health services to the level needed to effectively deal with the mental health needs of all mothers, spending that seems unlikely in the current political climate purely for the sake of mothers and their children (if only this were enough!). However, recognition that £400 per birth could in fact lead to more births may tip the scales, and without wishing to sound pronatalist, anything that encourages spending on maternal mental health can only be a good thing.

Open access to the award winning original research article on which this blog was based can be found below:

Myers, S., Burger, O., and Johns, S.E. (2016). Postnatal depression and reproductive success in modern, low-fertility contexts. Evolution, Medicine, & Public Health 2016(1): 71-84.

First blog

Knowledge matters! The research academics do is important. They are constantly providing insights that can make a difference to people’s lives. But far too often those insights don’t make it out of academia, being buried in obscure research journals, written in incomprehensible language, or misrepresented by the media. The aim of Matters of Reproduction is to provide a platform for researchers interested in sex, reproduction, parenting, and family dynamics to translate their findings, to tell people why their work really matters!

I have often found my self in conversations with academics bemoaning the fact that their research could make a difference to peoples lives if only more people knew about it! Because of the research that I do, these conversations have typically been with evolutionary anthropologists who passionately believe that their work should be informing public discourse, changing public health policies, and altering the ways we view the causes and consequences of inequality. That research more often than not doesn’t seem to reach its full impact potential (to borrow jargon from funding bodies) is not, from what I can gather, due to a lack of agreement from people outside of academia. I came up with the idea for this site after reflecting on the fact that when I talk about the research my colleagues and I do with non-academic friends, and other poor souls who have the misfortune to ask, they seem genuinely interested and on board with the idea that it is important and empowering. And so here is my own humble stab at getting the word out there…

The aim is to feature short blog posts highlighting new research on all things reproduction related. The posts will describe the findings and explain what they mean in comprehensible terms, but more importantly will give researchers the chance to state why their works matters, why we should care, and what difference their findings could make to the world. And of course, for those who want to read more, there will be a link to the original research publication. I shall obviously be featuring my own research, but I’d really like it if other researchers would contribute too, because who better to ensure that the real message isn’t lost!

So if you’re interested watch this space (and in the meantime follow me on Twitter!)…