There is a curious finding well known in some academic disciplines: girls who grow up in households without a father tend to start their periods earlier than girls whose fathers do live with them.

At first glance, this seems very surprising – what does family structure have to do with the timing of puberty? Researchers have puzzled over this question for decades, and one conclusion with widespread support is that early puberty is an evolved response to father absence. The argument runs that the lack of a father in the household is an indicator of a stressful environment.

Throughout most of human history a stressful environment would mean that the risk of death is relatively high, and when the risk of dying is high it makes sense for girls to mature early so they can have children and raise them to adulthood before the mother’s (relatively early) death. Mortality rates in human societies are now much lower than in the past, but children may still have evolved physiological mechanisms which speed up their development in stressful environments, as if they needed to get started on reproduction early.

Some people are WEIRD

But there is a big problem with this idea: almost all the data on father absence and puberty comes from WEIRD populations. WEIRD is an acronym coined by anthropologists Joe Henrich and two colleagues for Western, Educated, Industrialised, Rich and Democratic. Henrich and colleagues pointed out that the vast majority of research in psychology (>90%) comes from WEIRD populations (very often university students).

Often such research is generalised far beyond the context in which it’s gathered, and findings are assumed to hold in all other human populations. This is very problematic. The defining characteristic of our species is our behavioural flexibility. Humans have successfully spread across the entire globe because we can adapt to so many different environments. We can’t draw any firm conclusions about evolved responses then, without studying whether such responses exist in many different societies. This is certainly true when it comes to the human family.

What’s the traditional human family?

There is a widespread belief in WEIRD societies that the nuclear family is the ‘traditional’ human family. Most people do live in nuclear families in WEIRD societies, meaning that many children grow up in households with just their Mum, Dad and siblings. Outside WEIRD societies, on the other hand, families take a number of different forms, and children are often raised surrounded by a large network of relatives, including many from the extended family, such as grandparents, cousins, aunts and uncles. Marriage patterns also differ worldwide. Sometimes families involve polygyny (one man married to two or more wives), more rarely polyandry (one women married to several husbands), and societies differ in how socially acceptable divorce is, and for divorced or widowed people to remarry. So children grow up in a range of different family structures, and have variable relationships with their fathers. Some fathers are heavily involved in children’s lives; but in some societies, it may be common for children to grow up without a father constantly present in the household.

Why does this matter?

There has been a lot of research on how children’s early life experiences affects their development and health partly so that policy recommendations can help ensure children have the best start in life. Research on how father absence may be associated with early puberty – which is sometimes linked to poorer health and social outcomes in adolescence and later life – is therefore useful to some WEIRD policy-makers, who have argued that the nuclear family is the best environment to raise children, and enacted policies which encourage marriage and the formation of such families. Such policies risk ignoring the evidence from many populations that children can thrive in a range of different family structures, and may actually benefit from exposure to many different family members. This is particularly important in large and socioeconomically unequal populations such as the US, because WEIRD data are often from relatively wealthy and highly educated sections of Western populations, meaning that results may not generalise to marginalised populations even within the same society. For policies which affect the family, therefore, we need a solid evidence base on which to understand associations between family structure and child outcomes.   

What did we do?

My colleagues, Paula Sheppard and David Coall, and I decided to thoroughly search the literature to find out what the evidence really says about the association between father absence and puberty. We included in our investigation a search for studies which had tested for this association outside WEIRD populations, and a look at whether boys appeared to be affected by father absence. It seems plausible that, if father absence causes a physiological response in girls then it should also do something similar for boys, but most research has focused on the impact of father absence on girls’ puberty. This is probably because there is an easy-to-remember measure of puberty in girls – age at first period (menarche) – whereas for boys, collecting data on puberty involves asking questions about events which may be more difficult to remember or are more sensitive, such as the age at voice-breaking or other physical changes, such as the growth of body hair.

So what did we find?

We found about 80 papers which had data on the association between father absence and the timing of puberty, and counted up the number of papers which found that father absence was associated with early puberty; the number which found associations with later puberty; and the number that found no association. For WEIRD girls, a link between father absence and earlier puberty was found in a majority of studies – about 60% – but not all (see Figure below). But the picture was quite different when non-WEIRD societies, or boys, were studied. Outside WEIRD societies, father absence was sometimes associated with delayed puberty, particularly for boys.

Is there a universal, causal effect of father absence on age at puberty?

So what does this complicated picture mean? There are few studies which have attempted to collect data on the proposed physiological stress mechanisms which underlie associations between father absence and puberty, so that it’s hard to know whether these associations are the result of a direct effect of father absence causing earlier puberty. But some researchers have shown, in WEIRD societies, that earlier puberty is more closely associated with markers of family stress such as arguments within the household than with father absence itself, and that children with close relationships with fathers, even if they don’t live in the household, may not experience early puberty. This supports the idea that stress in childhood may cause early puberty, though it may not be father absence itself that affects children, but other stressful experiences which may be associated with father absence in WEIRD societies – for example, divorce may be relatively stressful in societies where the nuclear family is considered the norm.

The fact that father absence tends not to be very often linked to earlier puberty outside of WEIRD populations could have several explanations. One is that father absence doesn’t cause stress to children in all societies. Another is that an important role fathers often play is to provide food for children, so that where fathers are absent, children may be relatively poorly nourished. This may explain why in some societies father absence is associated with delayed puberty: growth and development takes energy, and poorly nourished children may experience relatively late puberty. If this explanation is true, though, it does raise questions about the evolved response to stress which WEIRD girls are proposed to experience in stressful family environments: throughout most of human history, food supply was relatively restricted, meaning that there may have been little opportunity for accelerated development to evolve in response to early stress.

The bottom line is…

We don’t yet have enough information to really know whether associations between father absence and either early or late puberty reflect truly causal relationships. But this literature review does show that any relationships between father absence and the timing of puberty are likely to depend on the role of fathers within the family, as well as other factors which influence puberty but vary between populations, such as food availability.

The most important take-home message is that it is impossible to draw conclusions which apply universally to all individuals in all populations without having good data from outside the very narrow slice of humanity which is Western, Educated, Industrialised, Rich and Democratic.

Read the full article here: https://royalsocietypublishing.org/doi/full/10.1098/rstb.2018.0124

For more information on the authors, please see the following:

Paula Sheppard: https://sites.google.com/site/paulajsheppard/ 

David Coall: https://www.ecu.edu.au/schools/medical-and-health-sciences/our-staff/profiles/senior-lecturers/dr-david-coall

Rebecca Sear: https://www.lshtm.ac.uk/aboutus/people/sear.rebecca

Published On: August 21, 2019

Rebecca Sear

Rebecca Sear

Rebecca Sear is a Reader at the London School of Hygiene and Tropical Medicine (LSHTM), teaching demography and researching human reproductive behaviour from an evolutionary perspective. She is trained in zoology, biological anthropology, and statistics, and subsequently worked first in a social science institution (London School of Economics) and then in an institution of global and public health (LSHTM). Having been exposed to a variety of disciplines, she is particularly interested in how the natural, social and medical sciences can be integrated as we try to understand our own species, and aims to conduct research somewhere inbetween these disciplines. She is particularly interested in taking a comparative perspective to understanding human reproductive behaviour, and exploring why such behaviour varies between, as well as within, populations.

Read her academic bio here: https://www.lshtm.ac.uk/aboutus/people/sear.rebecca


  • It makes a great deal of sense to consider a) multigenerational family history and b) the family system around the mother/absent father/child in terms of contact, stability or level of stress. Good to see this careful review of the research.

  • EmmaB says:

    It seems to me that association observed in WEIRD societies but not much in other societies, can have a much simpler explanation : father absence is associated with lower income, with is itself associated with higher BMI and earlier puberty.

    • Rebecca Sear says:

      Yes, it’s important to distinguish between different hypotheses for why father absence may be associated with puberty, and a loss of income associated with father absence may be one factor which helps explains any association. But some studies have controlled for family socioeconomic status when investigating the link between father absence and puberty and still find an association even after including this control. Associations between father absence and puberty are likely to be quite complex and may have several different explanations.

      But it’s also important to identify the exact causal pathway through which father absence may influence the timing of puberty. For example, one of the reasons that low income may be linked to puberty is a stress response, so that the causal arrow may flow from father absence through low income to stress responses which lead to earlier puberty.

      The causal pathway described above is also entirely compatible with the evolutionary hypothesis that the reason why father absence and low income cause a stress response which leads to earlier puberty is that such a response was adaptive (i.e. lead to more surviving children) in previous environments, when father absence/low income meant a higher risk of early death. This evolutionary explanation (an ‘ultimate’ explanation in evolutionary jargon) is a different kind of explanation to the ‘proximate’ explanations for why father absence and puberty are linked in the paragraph above.

      • EmmaB says:

        I do not think that it is really possible to statistically control for family socioeconomic status in the studies that you are describing. These studies, by definition, compare families with one parents to families with two parents. Even if the income and/or profession of all parents are known, it is very difficult to correctly estimate the economic cost of having only one parent and thus it is not possible to control correctly for socio economic status.

        The causal pathway described above (no father –> lower income –> higher BMI –> earlier period) is entirely INcompatible with the evolutionary hypothesis that the reason why father absence and low income cause a stress response which leads to earlier puberty is that such a response was adaptive because until extremely recently, the relationship between “ressources” and BMI was in the opposite direction (lower ressources/income –> lower BMI).

        Selected plasticity (if father absent then earlier period) is certainly not impossible but I think that an (unselected!) confounding effect is much more likely.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.