How Does Childbirth Affect Ageing?

Getting your Trinity Audio player ready...

Childbearing vs Longevity: The Trade-Off

Species that produce more offspring throughout their lives typically also have shorter natural lifespans. This makes sense from an evolutionary standpoint – there is a limited amount of energy obtainable from the environment. Many species spend most of this energy on producing many offspring as quickly as possible at the cost of lifespan. This rapid turnover of generations allows a species to multiply quickly. The main disadvantage of this strategy is that such species are vulnerable to sudden changes in their environments that last longer than their natural lifespans, such as a prolonged famine.

Some species instead put most of their energy into maintenance and repair of their cells, tissues and organs, allowing them to live longer, grow larger, and produce fewer offspring that are more likely to survive to reproductive age. Such species are better able to survive environmental changes, but must dedicate significant amounts of energy to resisting infectious disease and predation. It’s no use being able to live for 100 years if you get eaten by a tiger when you’re 15, before you have a chance to reproduce. This also partly explains why species do not just evolve to have longer and longer lifespans, even though it might seem like an evolutionary advantage to live longer and therefore have more time to reproduce. Sooner or later something is going to kill you, and at that point any energy spent on living longer is essentially wasted resources that could have gone into reproduction.

This trend doesn’t just occur across different species, but also within some species. An individual deer that produces more fawns in a season is more likely to have a shorter lifespan compared to a deer that produces fewer offspring. Producing and raising more offspring places greater physiological stress on the mother and requires more environmental resources, which means there’s less energy available for lifespan extension. The same relationship also exists for early reproduction. Animals that reproduce earlier in life tend to have shorter lifespans compared to animals from the same species that reproduce later, partly because pregnancy takes a greater physiological toll at a younger age, but also possibly because that diversion of energy from survival towards reproduction begins at an earlier stage of life.

Does This Apply To Humans?

Where do humans fit in to this pattern? Humans are, of course, members of the longer-lived, fewer offspring club. However, we are quite unusual in the fact that women only remain fertile for a little over half of their average life expectancy. It is common for fertility to decline significantly with age in animals, but species in which females become completely unable to reproduce long before the end of their natural lifespans are rare.

Several ideas have been put forward as to why humans live so long beyond reproductive age. The most popular and well-known of these ideas is the ‘grandmother hypothesis’. Compared to most other species, newborn humans remain completely helpless for many years after birth. Older humans past reproductive age may still aid the reproductive success of their offspring by assisting with childcare, gathering resources, providing knowledge and taking care of children if their parents die before said children can survive by themselves – things that are only possible because humans live in communities.

This brings us to another difference between humans and most animals – our environment is extremely complex due to the existence of socioeconomic factors. This makes the nature of the biological relationships between reproduction and longevity a lot harder to discern because wealthier, better educated individuals are more likely to wait longer before having children, and to have fewer of them. This finally brings us to this article, recently published in Nature, which attempts to plot the associations between reproductive history, lifespan, and estimates of biological ageing. While previous studies have looked at the relationship between number of births and ageing trajectories, this study attempted to combine data on the number of births and the age at which those births occurred, while using data from twins to help control for confounding factors.

What Did The Researchers Do?

The study data included 10,783 women from a Finnish twin cohort study. Using this data, researchers modelled 6 distinct ‘reproductive trajectories’ shown in the graph below. It plots the probability of giving birth according to age, in three-year age ‘bins’. For example, class 3 represents women who have at least one live birth between the ages of 27 and 30, with the probability of giving birth before or after that age range falling off sharply. Class 6, on the other hand, represents women who have more children spread out over a much longer period of their lives.

Researchers then compared survival over time between these reproductive classes. We already mentioned how confounding factors like education level affect both health and chance of pregnancy, so researchers had to use statistical methods in order to attempt to adjust for this, alongside other confounding factors like smoking and alcohol use. Since women in the twin cohort study were not all born in the same year, researchers also had to control for changing quality of life over time. While the study did not directly compare twin pairs to each other, they did use the fact that twins are genetically similar and usually grow up in similar environments in order to help with their statistical adjustments.

After making these adjustments, they were able to compare hazard ratios across different classes. In this context, the hazard ratio (HR) represents the probability that, at any point during the followup period, a person from that class will have died, with this probability being relative to class 3. For example, an HR of 2 would essentially mean that survival for women from that class is dropping off twice as quickly as class 3.

As you can see, there is a trend towards lower survival at both ends of the spectrum. After adjustment, women who had not given birth (nulliparous) were on average 41% more likely to die than those in class 3, a statistically significant difference. The next highest hazard ratios are in class 6 (more births across a longer period) and class 1 (more births earlier in life) although these only reached statistical significance before controlling for confounding factors.

Next, researchers made the same analysis but for epigenetic age instead of survival. Epigenetic age is an estimate of how ‘biologically old’ someone is, and is obtained by measuring DNA methylation – the addition of molecular ‘tags’ called methyl groups to the DNA molecule. These tags alter how the genetic code is read, and they change in a predictable way throughout life.

Mirroring the survival findings, researchers found that women at the extreme ends of the spectrum had the oldest epigenetic age estimates compared to their actual chronological age after confounding factors were accounted for. Classes 3 and 5 had the lowest estimates, with class 5 representing women who had the fewest total births, and tended to give birth later in life. Epigenetic age in these classes was around 1.5 years younger than chronological age on average. It should be emphasised that having a lower epigenetic age doesn’t necessarily mean that someone is actually biologically younger, as these tools are still ultimately estimators of true biological ageing. However, the specific epigenetic algorithm used here (GrimAge) is know for its ability to predict age-related disease risk and mortality, so it is reasonable to propose that people with lower estimates may have a reduced risk of disease and death.

The researchers also repeated these analyses using a few alternative modelling approaches and obtained similar results suggesting that those in classes 1, 6 and nulliparous women were most at risk, though some differences that were not previously statistically significant proved to be significant in these analyses.

What Are The Implications?

What can we conclude form these findings? Well, they mostly line up with existing ideas about how reproduction influences biological ageing. Having more offspring is resource intensive and shifts the balance between reproduction and biological housekeeping away from the latter, while having children at a younger age causes this shift to occur earlier in life. This results in a ‘U shaped’ relationship between reproductive trajectories and longevity. However, this doesn’t explain why women with no children at all had the lowest survival rates and highest epigenetic ages.

We should be careful about reading too much into this finding because of the potential influence of confounding factors. There is good evidence that pregnancy protects against certain age related diseases. For example, changes in breast tissue and in hormonal balance during pregnancy appear to protect against multiple cancers. Lacking these protective effects could have contributed to the apparent reduced survival among nulliparous women.

However, woman often don’t have children due to external reasons beyond their control such as health problems. This study did try to control for such factors, but it is impossible to perfectly control for all confounders. Nulliparous women were defined as those who had no recorded live births, which would suggest that women who had miscarriages or terminations were included in this group. Multiple factors that were uncontrolled for (such as certain infectious diseases, exposure to environmental toxins or extreme pyschological stress) can increase the risk of miscarriage, and miscarriage is associated with poorer subsequent mental health.

It could be argued that the extent of the risk increase is unlikely to be fully explained by the above confounding factors. However, even if there is a causal relationship between lack of births and mortality, we have to consider that not all of the benefits of having children are biological. For example, children may provide emotional, financial and other forms of support to their parents in old age. This means that even if having children does extend lifespan, it isn’t clear how much of this is due to direct biological effects of childbearing.

Are These Findings Actionable?

According to this data, the ‘optimal’ age for childbirth appears to be one’s late 20s, with the ‘optimal’ number of children being 2. However, the researchers caution against making plans based on this data. Outside of the extremes, number and timing of births did not have a particularly large impact on survival or rates of ageing. The fact that there are associations at the population level doesn’t necessarily mean that individual choices will have the same outcome. Furthermore, the population studied here may not be representative of all populations – these were women from Finland, over half of whom were born before 1950.