Periodic Reporting for period 1 - Born Once - Die Once (Towards a transdisciplinary demographic theory of birth and death trajectories)
Reporting period: 2022-09-01 to 2025-02-28
Birth and death are the cornerstones of existence and essential for population renewal. Complicated processes determine why and when individuals are born and die. Yet, despite this, scientists have identified major patterns. Especially regularities in how risk of death changes over age, and how lifespan improves over time are well established. Birth patterns are however harder to capture theoretically, and formal models are less developed due to the multitude of underlying dimensions that lead to the birth of a child. Stronger regularities and models of birth would improve forecasting of fertility and reduce the large uncertainty in population projections. From a larger perspective, population science lacks a theoretical framework with shared measures to study birth and death jointly. This is important because, although seemingly separated at opposite ends of the life cycle, the basic proposition of this project is that birth and death are connected via development and aging. It hypothesizes that constraints on growth and renewal and on dysregulation and decay share similar underlying principles. I will identify shared principles by means of finding shared regularities over age and time across population, such as the trend in postponing of birth and death. In this way, it envisions to work towards a transdisciplinary, demographic theory across systems and scales, that works for any kind of individuals, i.e. molecules, cells, or ecosystems. To capture aging across scales of observation, part of this project will consider groups of individuals as individuals themselves, who are born and die. E.g. a marriage of two individuals can be considered a “social individual”, that is born when signing the marriage certificate, and dies with signing the divorce certificate. By modeling birth and death of individual groups and linking them to that of their constituent individuals, new ways of modeling aging across scales are sought.
Based on a novel formal demographic method called born once—die once approach, our team is systematically investigating whether eight specifically known regularities of death patterns also can be found in birth patterns. Specifically, the born once die once approach allows calculating the “hazard of birth”, in analogy to the “hazard of death”. Thereby, we can capture and compare the full birth and death patterns with the same measure, which is an unexplored area of research.
First, we investigated whether birth patterns, just as death patterns, rise exponentially over adult ages. As our key first finding, we could verify that not just death, but also birth follows an exponential hazard. We consider this a substantial advancement because based on current knowledge, it was unexpected that birth can be modeled with the same mathematical model as death, which turns out to be the simple and mathematically convenient exponential function. Interestingly also from the perspective of aging research, we find that for both birth and death, the exponential pattern is observed over the age-range that is associated with respectively reproductive and actuarial senescence.
We consider this a major theoretical leap forward in the field of fertility analyses, not just because it increases the analytical power of modeling fertility by offering an analytically simple and universal parametric model. With the same, shared and simple model of the hazard of birth and death, our finding now enables developing novel formal demographic models and theory of fertility by importing and mirroring existing models and theory of mortality patterns.
Second, we investigated the trend in the mean age at birth in the front runner countries who experience, on average, the latest age at childbirth. Strikingly, we discovered a linear trend in the maximum mean age at birth that has been persisting over four decades. We argue that, together, the linear trends in maximum mean age at birth and at death reveal a shared constraint of aging. This constraint shapes and limits the potential postponement of birth and death to higher and higher ages.
Surprisingly, we also found that the slope of the linear trend in maximum mean age at birth falls right within the same ballpark as the rate of reproductive aging. This rate is reflected in the slope of the birth hazard function, discovered in our first project. More surprisingly, this value is also similar to the slope of the death hazard function, identified by previous research, and commonly referred to as the rate of actuarial aging. We are now digging deeper into this unexpected finding that birth and death seem to share the same pace of aging using a shared model of birth and death, that is, a model that applies to both processes equally and that can separate the senescent (aging) component of birth and death from its behavioral and environmental components.
Thirdly, we also investigated the postponement of birth by percentile measures, that is by interquartile ranges. Just as mortality researchers have discovered an "advancing front of old-age human survival", here we find that also birth patterns exhibit an advancing frontier of age-specific fertility contributions, visible in constant distances between birth survival percentiles.
In parallel to the empirical projects, we also developed an integrated analytical framework of fertility, which quantifies fertility patterns in a manner similar to mortality. It also translates between the novel and the traditional fertility frameworks. For the first time, this opens the door to study birth and death patterns together, as done in this project, and thereby connects two up to now almost disparate fields of demographic research.
With the new theoretical framework in place, we now can translate between different perspectives to measure fertility and can predict how strongly outcomes should diverge depending on alternative measures. This is currently helping us to double check the empirical results from our projects above to ensure that findings and conclusion we draw are not critically dependent on the specific measures applied or data that newly become available.
Taken together, our current findings support the main hypothesis underlying this ERC that birth and death are connect as expressed by mirrored regularities in population level patterns. These results encourage further pursuit of the principles of aging by defining, measuring and modeling aging as reflected in age-patterns of birth and death together.
First, we investigated whether birth patterns, just as death patterns, rise exponentially over adult ages. As our key first finding, we could verify that not just death, but also birth follows an exponential hazard. We consider this a substantial advancement because based on current knowledge, it was unexpected that birth can be modeled with the same mathematical model as death, which turns out to be the simple and mathematically convenient exponential function. Interestingly also from the perspective of aging research, we find that for both birth and death, the exponential pattern is observed over the age-range that is associated with respectively reproductive and actuarial senescence.
We consider this a major theoretical leap forward in the field of fertility analyses, not just because it increases the analytical power of modeling fertility by offering an analytically simple and universal parametric model. With the same, shared and simple model of the hazard of birth and death, our finding now enables developing novel formal demographic models and theory of fertility by importing and mirroring existing models and theory of mortality patterns.
Second, we investigated the trend in the mean age at birth in the front runner countries who experience, on average, the latest age at childbirth. Strikingly, we discovered a linear trend in the maximum mean age at birth that has been persisting over four decades. We argue that, together, the linear trends in maximum mean age at birth and at death reveal a shared constraint of aging. This constraint shapes and limits the potential postponement of birth and death to higher and higher ages.
Surprisingly, we also found that the slope of the linear trend in maximum mean age at birth falls right within the same ballpark as the rate of reproductive aging. This rate is reflected in the slope of the birth hazard function, discovered in our first project. More surprisingly, this value is also similar to the slope of the death hazard function, identified by previous research, and commonly referred to as the rate of actuarial aging. We are now digging deeper into this unexpected finding that birth and death seem to share the same pace of aging using a shared model of birth and death, that is, a model that applies to both processes equally and that can separate the senescent (aging) component of birth and death from its behavioral and environmental components.
Thirdly, we also investigated the postponement of birth by percentile measures, that is by interquartile ranges. Just as mortality researchers have discovered an "advancing front of old-age human survival", here we find that also birth patterns exhibit an advancing frontier of age-specific fertility contributions, visible in constant distances between birth survival percentiles.
In parallel to the empirical projects, we also developed an integrated analytical framework of fertility, which quantifies fertility patterns in a manner similar to mortality. It also translates between the novel and the traditional fertility frameworks. For the first time, this opens the door to study birth and death patterns together, as done in this project, and thereby connects two up to now almost disparate fields of demographic research.
With the new theoretical framework in place, we now can translate between different perspectives to measure fertility and can predict how strongly outcomes should diverge depending on alternative measures. This is currently helping us to double check the empirical results from our projects above to ensure that findings and conclusion we draw are not critically dependent on the specific measures applied or data that newly become available.
Taken together, our current findings support the main hypothesis underlying this ERC that birth and death are connect as expressed by mirrored regularities in population level patterns. These results encourage further pursuit of the principles of aging by defining, measuring and modeling aging as reflected in age-patterns of birth and death together.
The dynamics of birth and death are an integral part of any population science. By discovering and modeling shared macro level regularities of birth and death, this project seeks to provide the foundation for better forecasting and projection models of population dynamics, applicable to human and non-human populations. With a shared formal framework to model and measure birth and death patterns, this project aims to pioneer a new, transdisciplinary field of research that would synergistically investigate the basic principles of formation, renewal, and dissolution. For example, we expect that empirical results above can have a major impact on aging research and inspire new empirical research questions and direction. Overall enabling fundamental research, the shared regularities, constraints and models of birth and death investigated in this project could have manifold impact and implications in a range of different disciplines and applications.