A Probabilistic Framework for Estimating the Modal Age at Death

arXiv:2411.09800 · stat.AP · Submitted 2024-11-14 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Statistics. The day in statistics: methods, theory and the inference under them.

Tom: I'm Tom.

Jane: And I'm Jane, here with Lu, Meng and our in-house large language model, Lalam.

Tom: We're looking at the paper: "A Probabilistic Framework for Estimating the Modal Age at Death".

Jane: Here's the gist of the paper. The modal age at death is an increasingly used measure for understanding longevity and mortality patterns This study addresses this gap by introducing a probabilistic framework for estimating the probability distribution of the modal age at death.

Tom: Context and Motivation Life expectancy at birth has long been the standard summary measure of mortality Life expectancy reflects mortality conditions observed during a specific calendar year rather than the actual lifespan of any population.

Jane: The modal age at death offers a more robust alternative by identifying the age at which deaths are most concentrated.

Tom: This mode is largely unaffected by infant or early-adult mortality and is driven mainly by changes in survival at older ages.

Title and authors: Tom: So we're looking at this paper called "A Probabilistic Framework for Estimating the Modal Age at Death," and it seems like they're tackling a problem where most current ways of figuring out the modal age at death just give you one single number, but that number doesn't really tell you how much uncertainty there is, or what the range of plausible ages actually looks like.

Jane: That’s a big deal because it means we can start talking about risk and uncertainty in longevity data instead of just looking at a static point estimate. The authors are introducing a whole new way to look at this, which they base on treating the mode as something that has its own probability distribution rather than just an outcome of some other model.

Lu: Exactly. They build this framework using a multinomial model for age-specific death counts and then use a Gaussian approximation when the total number of deaths, which they call n, gets pretty large. That lets them capture the variability inherent in the data while still keeping things computationally manageable by using that normal distribution approximation for big n values.

Meng: Okay, so they're taking these age-specific death counts and treating them like outcomes of a multinomial experiment conditional on the total number of deaths, which they call DN = n ∼ Multinomial(n, p) according to page two of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death <ref:2411.09800#pg3,the total number of deaths, which>. But what does that actually change for us when we're trying to deploy a model?

Lalam: It means they are not just modeling the counts themselves, but describing how those counts depend on each other because the total number of deaths n is fixed and there’s this dependency where an increase in one age category forces a decrease in others, which is an inherent property of that multinomial distribution according to page three of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Tom: Right, so the key mechanism here is using that fixed total n constraint to show how deaths across different ages are linked probabilistically, and then using that structure to estimate the probability that any specific age interval is actually the mode of those counts.

Jane: And they do this by calculating P(M = x), which is the probability that age interval x has the highest death count compared to all other intervals, which they write as P(M = x) = Z Rx fYx(yx) dyx, referring to page two of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death <ref:2411.09800#pg3>.

Lu: That’s where the probabilistic formulation really shines because it directly quantifies uncertainty around the modal age at death, which is what they call a key property of this framework according to page two of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death <ref:2411.09800#pg3>. They are moving away from just model-based propagation to estimating the mode as a stochastic quantity in its own right.

Title and authors: Meng: But how do we know if that Gaussian approximation they use for large values of n is actually reliable when we're dealing with real-world data, especially when the population sizes aren't always huge? What are the limitations they admit?

Lalam: The paper itself flags that the Gaussian approximation assumes a large total number of deaths n and smooth mortality distributions, and they also mention that data quality issues can introduce biases in those age specific probabilities p, which affects how reliable their estimated modal age distribution is.

Tom: So it’s not a perfect model for everything because of those assumptions about size and smoothness, but the whole point is that this probabilistic formulation captures variability better than just a single point estimate would.

Jane: Right, so what does this mean for us who are trying to use these models? It means we can now quantify the uncertainty around the modal age at death directly, which is something that traditional methods often overlook when they only give you one number.

Lu: And it gives us a way to see the full probability distribution of the modal age at death, which provides a better picture of the dynamics of longevity and its improvements across populations.

Meng: For me, if I were deploying this in production, knowing that we have this distribution instead of just one number helps me build better guardrails around what we predict when mortality data gets noisy or suddenly changes. It addresses that variability often overlooked by traditional point estimation methods.

Lalam: And they also say that as the total number of deaths n increases, the probability distribution of the modal age at death converges to a point mass at the age interval with the highest underlying death probability px star, which means for large n, P(M = x star) converges to one according to page seven of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Tom: So they’re telling us that if we collect a really big amount of data, we can be very confident that the real modal age is right around that specific age interval x star, which is where the underlying death probability px star is highest.

Jane: But for finite n, they stress that random fluctuations in the data introduce variability, which means there's still uncertainty about the exact mode even if n gets large according to page seven of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Lu: And that’s actually really useful because it allows us to quantify this uncertainty directly and aligns with the discrete nature of mortality data, which is something they emphasize on page two of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death <ref:2411.09800#pg3>. It moves beyond just a model-based propagation.

Title and authors: Meng: I wonder how this maps to my work on running models at scale, since we’re dealing with massive datasets—does calculating this full probability distribution add too much overhead compared to just getting a mean vector?

Lalam: The paper suggests that for large values of n, the vector of independent Poisson counts when conditioned on its sum is equivalent to a single multinomial distribution according to page two of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death <ref:2411.09800#pg3,for large values of n, the>. So they are using this structure to simplify things computationally, even with those complex distributions.

Tom: It sounds like they’ve built a way to handle that complexity by using the multinomial framework and then simplifying it with the Gaussian approximation when n is big enough, which is a smart approach for modeling large datasets.

Jane: And this whole probabilistic approach gives us a clearer picture of the likelihood that each age interval contains the mode, which helps us understand where those longevity improvements are actually happening across different countries.

Lu: They even show empirical applications from mortality data in six different countries covering the period from one thousand nine hundred sixty to two thousand twenty and they found an overall upward trend in the modal age at death across those countries according to page eight of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Meng: That trend is interesting, but I need to know if that upward trend holds up when we look at smaller cohorts or different demographic shifts—does this framework handle those sudden changes well, or does it just smooth them over?

Lalam: The methodology was applied to mortality data from six countries covering the period from one thousand nine hundred sixty to two thousand twenty and the results show an overall upward trend in the modal age at death from one thousand nine hundred sixty to two thousand twenty across those six countries according to page eight of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Tom: So they’ve shown that this framework can capture these long-term trends, and they also look at gender differences, where females consistently exhibit a higher modal age at death than males across those six countries.

Jane: Yes, but what’s interesting is that some fluctuations are visible in the distribution shapes—temporary plateaus or slight declines in the mode occurring during specific periods—which shows that it’s not just one steady line going up.

Lu: The figure also highlights variability in the distribution shapes; broader distributions, particularly in Denmark and the Netherlands, point to greater statistical uncertainty, potentially tied to smaller population sizes, while narrower distributions suggest more stable estimates in countries like Japan according to page seven of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Title and authors: Meng: So it’s a trade-off between accuracy and confidence based on the underlying data structure; that makes sense when you’re trying to deploy something reliably.

Lalam: The framework can also be extended to leverage the Human Cause-of-Death series recently incorporated into the Human Mortality Database, allowing them to model age specific death counts disaggregated by cause, which would offer a clearer view of how different causes shape the distribution of deaths around the mode according to page seven of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Tom: That sounds like a way to add another layer of detail; not just age, but what’s causing those deaths, which gives us more granular insights into longevity dynamics.

Jane: Overall, this framework’s robustness in reflecting the uncertainty surrounding the modal age offers a valuable tool for demographic analysis because it shows you the full probability distribution instead of just a single estimate.

Lu: Yeah, by showcasing the full probability distribution of the modal age at death, it provides a better picture of the dynamics of longevity and its improvements across populations according to page seven of THIS PAPER — A Probabilistic Framework for Estimating the Modal Age at Death.

Meng: I think having that full picture is crucial for risk stratification tools because it allows us to quantify variability directly and aligns with the discrete nature of mortality data.

Lalam: The probabilistic framework itself provides a probabilistic perspective that explicitly quantifies variability, reveals competing age intervals, and improves robustness to data fluctuations, which is what’s called the gains in this approach.

Tom: So it’s about moving past just point estimation and getting a more robust measure of how certain we are about the modal age at death by using this framework.

Jane: So to wrap up on "A Probabilistic Framework for Estimating the Modal Age at Death," it’s a sophisticated way to quantify uncertainty around longevity trends, showing us not just where the mode is, but how likely different ages are actually that mode.

Lu: And future work could look at methods that account for overdispersion or clustering, like the Bell distribution or Bayesian approaches, which might give us even more detail on those fluctuations.

Meng: If we can get better at handling those small population scenarios where the Gaussian approximation fails, I think we can make this deployment much more reliable for real-world use.

Lalam: This framework can also be extended to leverage the Human Cause-of-Death series recently incorporated into the Human Mortality Database, allowing them to model age specific death counts disaggregated by cause, which would offer a clearer view of how different causes shape the distribution of deaths around the mode.

Tom: That’s a solid way forward for more detailed demographic research. We'll keep an eye on those extensions as they come out.

The paper's summary: Tom: So, basically, this paper is taking the modal age at death—which is just where most people die—and instead of giving you one single guess for that age, it gives you a whole probability distribution of ages where that mode might be located.

Jane: That’s right. It moves away from just saying "the mode is seventy-five" to saying "there's a fifty percent chance the mode is between seventy-two and seventy-eight." It quantifies the uncertainty, which is a huge deal when you're trying to model real-world mortality data.

Lu: The paper sets up this framework using a Poisson-based method, similar to what Brillinger did before, but they treat those age specific death counts as something that follows a multinomial distribution conditional on the total number of deaths.

Meng: So it's not just looking at the individual age groups separately; it’s acknowledging that they are all part of one big pool where the total sum matters, and that connection is what defines the probability structure.

Tom: Exactly. And then they use a Gaussian approximation when the total number of deaths, which we call n, gets pretty big. That lets them turn that complex multinomial structure into something manageable using a multivariate normal distribution with a mean vector and a covariance matrix based on those age specific probabilities p.

Jane: It’s like when you have thousands of data points; instead of tracking every single one individually, you can use the average and the spread to get a really good picture of where most things are clustered.

Lu: The main goal they’re trying to solve is computing P(M equals x), which is the probability that a specific age interval x ends up being the age with the most deaths in that group, compared to all other intervals.

Tom: That probability calculation involves this joint multivariate normal density function, fYx(yx), which lets them calculate exactly how likely each age interval is to be the mode of death counts.

Jane: And they show something pretty interesting about convergence; as n gets really big, that whole distribution of probabilities for the mode starts to settle down into a single point mass at the age interval x star where the underlying death probability px is highest.

Meng: So it means that in theory, if you collect enough data, you can be very sure that the true modal age will be right around that specific age interval x star where deaths are most concentrated.

Tom: But they also stress that for any finite number of deaths n, there’s still this random fluctuation from the data itself introducing variability and uncertainty about where exactly the mode is located.

Jane: That randomness is what makes this framework so useful for risk assessment; it lets us see not just a single best guess, but the range of plausible modal ages.

Lu: This whole probabilistic approach gives you a better handle on that variability than traditional point estimation methods, which often overlook these kinds of fluctuations in the data.

Tom: It’s about moving from just getting a median lifespan to understanding the full likelihood of different ages being the most common one, which helps us build more robust systems.

Jane: And they’re not just looking at general mortality; they applied this to actual data from six countries spanning one thousand nine hundred sixty to two thousand twenty and they found an overall upward trend in that modal age across all those nations.

Meng: That upward trend is clear, but the paper also points out that there are gender differences; females consistently have a higher modal age at death than males in their empirical application.

Tom: And they noted some fluctuations too, temporary plateaus or slight dips during certain periods, which shows it’s not just a steady climb.

Jane: It’s really useful because by looking at the full probability distribution, we get a much clearer picture of how longevity is changing and where those changes are happening across different populations.

Lu: They also mentioned that this framework can be extended to use the Human Cause-of-Death series from the Human Mortality Database, allowing them to model how different causes shape that distribution around the mode.

Tom: That means we could look at not just when people die, but what’s causing those deaths and how that affects where the mode lands.

Jane: Overall, this framework is a solid tool for demographic analysis because it shows you the uncertainty around those longevity trends instead of just one number.

Meng: The main caveat they give is that their Gaussian approximation assumes a large total number of deaths n and smooth mortality distributions, so in really small populations or during sudden shocks, those assumptions might not hold up well.

Tom: So the method’s strength is in its probabilistic structure, but we have to be careful when it comes to applying it where the population isn't massive or the data is really noisy.

Jane: And that’s what we’ll be looking at next—how researchers are trying to fix those limitations with methods like the Bell distribution or Bayesian approaches.

The paper's improvements: Tom: So, we're looking at where this paper suggests they can take things next; they’re not just stopping at the Gaussian approximation when n gets large, right?

Jane: They suggest exploring methods that account for overdispersion or clustering, like using a Bell distribution or maybe Bayesian approaches to handle those tricky data patterns.

Lu: That makes sense because the Gaussian assumption is pretty strict about how smooth the mortality distribution has to be; if the data has more bumps and dips than a smooth curve, that approximation breaks down.

Meng: If we want this framework to be deployable for real traffic, we need something that handles those sudden shocks or weird clusters in deaths, not just a smooth average.

Tom: Exactly. And they also have an idea to extend the framework into a continuous age setting using Gaussian processes, which would give us a more precise representation of mortality patterns when you treat age as something fluid rather than just discrete intervals.

Jane: That’s interesting because it means we could analyze the whole lifespan, not just those fixed age groups, capturing finer nuances in how people age.

Lu: They also mention leveraging that new Human Cause-of-Death series from the Human Mortality Database; modeling death counts disaggregated by cause would give us a clearer view of how different causes shape the distribution around that mode.

Meng: So instead of just one modal age for everyone, we could get a modal age broken down by why people are dying, which is way more useful for targeted interventions.

Tom: It sounds like they’re pushing toward models that can handle that complexity and provide more detailed insights into the underlying processes.

Jane: Overall, the improvement is moving toward methods that are more flexible with data structure and cause-specific details.

Lu: The authors are setting up a path to incorporate those kinds of richer data sources to make the modal age estimate even more powerful for prediction.

Tom: This whole direction shows they’re thinking about how to make this tool more versatile for different real-world scenarios, not just textbook cases.

Jane: So what we need to watch next is whether these extended methods actually hold up when they are applied to the kind of messy, real-world data we actually get from health databases.

Conclusion: Jane: So we’re wrapping up on "A Probabilistic Framework for Estimating the Modal Age at Death," which basically shows how to use a probabilistic model to estimate not just one modal age, but a whole distribution of possible modal ages.

Tom: Right. It moves us past just getting a single number and lets us quantify the uncertainty directly, which is what we need when we’re building models for risk assessment.

Lu: The implication is that AI systems could predict not just a median lifespan, but the entire range of plausible modal ages, which improves accuracy in predicting longevity trends over time.

Meng: For deployment at scale, having this distribution instead of just a point estimate helps us build better guardrails around our predictions when mortality data gets noisy or suddenly changes.

Jane: And they showed that this approach is robust against those data fluctuations because it explicitly quantifies the variability often overlooked by traditional methods.

Tom: They also demonstrated that as the total number of deaths, which we call n, gets very large, the probability distribution for the modal age settles down into a point mass at that age interval with the highest underlying death probability.

Lalam: That means for huge datasets, you can be really sure that your prediction is centered right around that specific age interval x star where deaths are most concentrated.

Lu: But they also made it clear that for finite n, there's still this random fluctuation from the data itself introducing uncertainty about the exact mode location.

Meng: So for us deploying models, we have to be careful about those small population scenarios or sudden shocks where their Gaussian approximation might not hold up as reliably.

Jane: That’s a key caveat; it tells us exactly where this framework stops working perfectly and where we need to look for improvements.

Tom: We're looking forward to seeing how researchers use these limitations to guide the next generation of models, especially with extensions into continuous age settings.

Lu: Exactly, because extending it into a continuous setting using Gaussian processes could give us a much more nuanced understanding of mortality dynamics across the entire lifespan.

Jane: That sounds like the right direction for deeper research, exploring how those cause-specific details or structural uncertainties play into this probabilistic framework.

Tom: We’ll keep an eye on those future papers to see if they manage to make that continuous extension work reliably in practice.

Interdisciplinary Center on Population Dynamics University of Southern Denmark

stat.AP

Submitted: 2024-11-14

Updated: 2026-11-02

Comments: 26 pages, 9 figures, 2 tables; appendices included. Code and replication materials: https://github.com/scpatricio/modal_age_posterior

Code: https://github.com/scpatricio/mode_estimation

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 77/100

The gist: The modal age at death is an increasingly used measure for understanding longevity and mortality patterns This study addresses this gap by introducing a probabilistic framework for estimating the…

Key concepts

Modal Age at Death (M)
The age interval where the highest number of deaths occurs within a population. It is more stable than life expectancy because it focuses on survival patterns at older ages rather than early mortality.
Multinomial Distribution
A statistical model used here to describe the vector of death counts across different age intervals. It treats the observed number of deaths as outcomes from an experiment where each age interval has a specific probability of a death occurring.
Gaussian Approximation
When the total number of deaths is large, complex multinomial data can be approximated by a multivariate normal distribution. This mathematical tool allows researchers to calculate probabilities related to the distribution of the modal age at death.
Convergence Property
As the total number of deaths increases, the probability distribution of where those deaths are most concentrated converges toward a single point—the age interval with the highest underlying death probability. This means long-term trends become more predictable.

Terminology

Summary

The modal age at death is an increasingly used measure for understanding longevity and mortality patterns This study addresses this gap by introducing a probabilistic framework for estimating the probability distribution of the modal age at death. p. 0

Where the paper says it

However, existing estimation methods focus on point estimates, overlooking the inherent variability and uncertainty in mortality data. This study addresses this gap by introducing a probabilistic framework for estimating the probability distribution of the modal age at death.

Context and Motivation

Life expectancy at birth has long been the standard summary measure of mortality Life expectancy reflects mortality conditions observed during a specific calendar year rather than the actual lifespan of any population. The modal age at death offers a more robust alternative by identifying the age at which deaths are most concentrated. p. 0

Where the paper says it

The modal age at death offers a more robust alternative. It identifies the age at which deaths are most concentrated, summarizing the prevailing location of adult mortality (Bergeron-Boucher et al.

This mode is largely unaffected by infant or early-adult mortality and is driven mainly by changes in survival at older ages. p. 0
Where the paper says it

, 2015; Canudas-Romo, 2008). Unlike life expectancy, the mode is largely unaffected by infant or early-adult mortality and is driven mainly by changes in survival at older ages.

Probabilistic Approach for Modal Age Estimation

The approach builds on the Poisson-based framework for analysing death counts introduced by Brillinger (1986). A key property of this framework is that the vector of independent Poisson counts, when conditioned on its sum, is equivalent to a single multinomial distribution. p. 1

Where the paper says it

Our approach builds on the Poisson-based framework for analysing death counts introduced by Brillinger (1986). A key property of this framework is that the vector of independent Poisson counts, when conditioned on its sum, is equivalent to a single multinomial distribution, with deaths distributed across age categories according to their underlying probabilities.

The authors treat age-specific death counts as outcomes of a multinomial experiment conditional on the total number of deaths. p. 1
Where the paper says it

A key property of this framework is that the vector of independent Poisson counts, when conditioned on its sum, is equivalent to a single multinomial distribution, with deaths distributed across age categories according to their underlying probabilities. In this setting, we treat age-specific death counts as outcomes of a multinomial experiment conditional on the total number of deaths.

Multinomial Model for Mortality Data

The vector of death counts is defined as D = (D1, D2,..., DK)⊤, and the total number of deaths across all age intervals is given by n = D1 + D2 + · · · + DK. p. 1

Where the paper says it

The vector of death counts is defined as D = (D1, D2, . , DK) ⊤, and the total number of deaths across all age intervals is given by n = D1 + D2 + · · · + DK.

The vector D follows a multinomial distribution: DN = n ∼ Multinomial(n, p), where px is the probability that a death occurs in age interval x. p. 1
Where the paper says it

Assume that each death is an independent trial resulting in one of the K age categories with probabilities p = (p1, p2, . In this case, the vector D follows a multinomial distribution: D|N = n ∼ Multinomial(n, p), where px is the probability that a death occurs in age interval x.

The probabilities p = (p1, p2,. p. 1
Where the paper says it

Our approach builds on the Poisson-based framework for analysing death counts introduced by Brillinger (1986). A key property of this framework is that the vector of independent Poisson counts, when conditioned on its sum, is equivalent to a single multinomial distribution, with deaths distributed across age categories according to their underlying probabilities.

.., pK)⊤ are derived using a life table approach. p. 1
Where the paper says it

The probabilities p = (p1, p2, . , pK) ⊤ are derived using a life table approach.

Gaussian Approximation for Large Populations

For large values of n, the multinomial distribution can be approximated by a multivariate normal distribution. Under these conditions, D can be expressed as: D ≈ N (µ, Σ), where µ = np is the mean vector. p. 2

Where the paper says it

In the context of the multinomial distribution, when n is large and each expected count npx is sufficiently large, the distribution of the death counts D can be wellapproximated by a multivariate normal distribution. Under these conditions, D can be expressed as: D ≈ N (µ, Σ), where µ = np is the mean vector.

The covariance matrix is given by Σ = n Diag(p) − pp⊤. p. 2
Where the paper says it

This represents the expected number of deaths in each age interval. The covariance matrix is given by Σ = n Diag(p) − pp⊤ , which captures the variances and covariances between age categories.

Probabilistic Estimation of the Modal Age

The goal is to compute P(M = x) = P (Dx = max[D1, D2,..., DK]), which represents the probability that Dx is the largest component of the death count vector D. p. 2

Where the paper says it

We aim to compute P(M = x) = P (Dx = max{D1, D2, . , DK}), which represents the probability that Dx is the largest component of the death count vector D.

This probability can be expressed as: P(M = x) = Z Rx fYx(yx) dyx, where fYx(yx) is the joint multivariate normal density function of Yx. p. 3
Where the paper says it

where fYx (yx) is the joint multivariate normal density function of Yx, and the integration region Rx is defined as: Rx = {yx ∈ R K−1 | Yx,j ≥ 0 for all j ̸= x}. Evaluating this integral requires integrating over a (K−1)-dimensional space, with boundaries determined by the inequalities Yx,j ≥ 0.

Convergence of the Modal Distribution as n Increases

As the total number of deaths n increases, the probability distribution of the modal age at death converges to a point mass at the age interval with the highest underlying death probability px⋆. This result shows that as n becomes large, P(M = x⋆) converges to 1. p. 4

Where the paper says it

Therefore, limn→∞ P(M = x ⋆ ) = 1. This result shows that as the total number of deaths n becomes large, the modal age at death converges in probability to the age interval x ⋆ with the highest death probability px⋆ .

Empirical Application

The methodology was applied to mortality data from six countries covering the period from 1960 to 2020. The results show an overall upward trend in the modal age at death from 1960 to 2020 across the six countries. p. 5

Where the paper says it

This dataset provides a robust foundation for evaluating the applicability of our method to real-world scenarios, as it includes varying population sizes, cultural contexts, and mortality patterns across these nations. The results, shown in Figure 1, show an overall upward trend in the modal age at death from 1960 to 2020 across the six countries.

Concluding Remarks and Future Directions

The framework’s robustness in reflecting the uncertainty surrounding the modal age offers a valuable tool for demographic analysis. Future research could explore methods that account for overdispersion or clustering, such as the Bell distribution or Bayesian approaches. p. 6

Where the paper says it

Dependencies in mortality due to shared environmental or social factors, such as clustering in epidemics, also challenge the model’s assumption of independent deaths (L´eger and Mazzuco, 2021). To address these limitations, future research could explore methods that account for overdispersion or clustering, such as the Bell distribution, negative binomial models, or Bayesian approaches (Gelman et al.

This framework can also be extended to leverage the Human Cause-of-Death (HCD) series recently incorporated into the Human Mortality Database. p. 6
Where the paper says it

Gaussian processes and methods from extreme value theory may offer promising directions for analyzing the distribution of the mode in continuous age contexts (Ludkovski et al. This framework can also be extended to leverage the Human Cause-of-Death (HCD) series recently incorporated into the Human Mortality Database.

p. 6
Where the paper says it

Gaussian processes and methods from extreme value theory may offer promising directions for analyzing the distribution of the mode in continuous age contexts (Ludkovski et al. This framework can also be extended to leverage the Human Cause-of-Death (HCD) series recently incorporated into the Human Mortality Database.

Acknowledgments

We gratefully acknowledge the financial support from the AXA Research Fund through the funding for the “AXA Chair in Longevity Research”.

References

Adler, R. J. and Taylor, J. p. 4

Where the paper says it

Instead, we can analyze the probability density function of the age at death and study the behavior of the process around its maximum (e. , Adler and Taylor, 2009).

E. (2009). Random fields and geometry. p. 6
Where the paper says it

and Taylor, J. Random fields and geometry.

Andreev, E. p. 0

Where the paper says it

, Booth and Tickle, 2008). In all cases, these approaches yield a single point estimate of the mode, and any assessment of uncertainty must be obtained indirectly—through resampling procedures such as bootstrapping or, when the model’s structure permits, by applying the delta method to estimated parameters (Andreev and Shkolnikov, 2010; Keyfitz and Caswell, 2005).

M. and Shkolnikov, V. p. 0
Where the paper says it

, Booth and Tickle, 2008). In all cases, these approaches yield a single point estimate of the mode, and any assessment of uncertainty must be obtained indirectly—through resampling procedures such as bootstrapping or, when the model’s structure permits, by applying the delta method to estimated parameters (Andreev and Shkolnikov, 2010; Keyfitz and Caswell, 2005).

M. (2010). Spreadsheet for calculation of confidence limits for any life table or healthylife table quantity. p. 6
Where the paper says it

and Shkolnikov, V. Spreadsheet for calculation of confidence limits for any life table or healthylife table quantity.

Bergeron-Boucher, M. p. 0

Where the paper says it

The modal age at death offers a more robust alternative. It identifies the age at which deaths are most concentrated, summarizing the prevailing location of adult mortality (Bergeron-Boucher et al.

-P., Ebeling, M. p. 6
Where the paper says it

Bergeron-Boucher, M. , Ebeling, M.

, and Canudas-Romo, V. p. 0
Where the paper says it

It identifies the age at which deaths are most concentrated, summarizing the prevailing location of adult mortality (Bergeron-Boucher et al. , 2015; Canudas-Romo, 2008).

p. 0
Where the paper says it

It identifies the age at which deaths are most concentrated, summarizing the prevailing location of adult mortality (Bergeron-Boucher et al. , 2015; Canudas-Romo, 2008).

(2015). Decomposing changes in life expectancy: Compression versus shifting mortality. p. 6
Where the paper says it

, and Canudas-Romo, V. Decomposing changes in life expectancy: Compression versus shifting mortality.

Bishop, Y. M., Fienberg, S. p. 6

Where the paper says it

Demographic Research, 33:391–424. , Fienberg, S.

E., and Holland, P. p. 6
Where the paper says it

, Fienberg, S. , and Holland, P.

W. (2007). Discrete multivariate analysis: Theory and practice. p. 6
Where the paper says it

, and Holland, P. Discrete multivariate analysis: Theory and practice.

Bongaarts, J. p. 0

Where the paper says it

Yet, despite its apparent simplicity, its interpretation is not straightforward. In a period life table, life expectancy reflects mortality conditions observed during a specific calendar year rather than the actual lifespan of any population (Bongaarts, 2006).

(2006). How long will we live? p. 6
Where the paper says it

Bongaarts, J. How long will we live?

Population and development review. p. 6
Where the paper says it

How long will we live? Population and development review, pages 605–628.

Booth, H. and Tickle, L. p. 0

Where the paper says it

Across both parametric and nonparametric frameworks, most methods focus on modeling the mortality curve, from which the modal age at death is subsequently derived (see, e. , Booth and Tickle, 2008).

(2008). Mortality modelling and forecasting: A review of methods. p. 6
Where the paper says it

and Tickle, L. Mortality modelling and forecasting: A review of methods.

Brillinger, D. p. 1

Where the paper says it

This probabilistic formulation quantifies uncertainty directly and aligns with the discrete nature of mortality data. Our approach builds on the Poisson-based framework for analysing death counts introduced by Brillinger (1986).

R. (1986). A biometrics invited paper with discussion: the natural variability of vital rates and associated statistics. p. 7
Where the paper says it

Brillinger, D. A biometrics invited paper with discussion: the natural variability of vital rates and associated statistics.

Canudas-Romo, V. (2008). The modal age at death and the shifting mortality hypothesis. p. 7

Where the paper says it

Canudas-Romo, V. The modal age at death and the shifting mortality hypothesis.

Castellares, F. p. 6

Where the paper says it

To address these limitations, future research could explore methods that account for overdispersion or clustering, such as the Bell distribution, negative binomial models, or Bayesian approaches (Gelman et al. , 1995; Castellares et al.

, Ferrari, S. p. 7
Where the paper says it

Castellares, F. , Ferrari, S.

L., and Lemonte, A. p. 7
Where the paper says it

, Ferrari, S. , and Lemonte, A.

J. (2018). On the bell distribution and its associated regression model for count data. p. 7
Where the paper says it

, and Lemonte, A. On the bell distribution and its associated regression model for count data.

Daley, D. J., Vere-Jones, D. p. 7

Where the paper says it

Applied Mathematical Modelling, 56:172–185. , Vere-Jones, D.

, et al. (2003). An introduction to the theory of point processes: volume I: elementary theory and methods. p. 7
Where the paper says it

, Vere-Jones, D. An introduction to the theory of point processes: volume I: elementary theory and methods.

Feller, W. (1991). An introduction to probability theory and its applications, Volume 2, volume 81. p. 7

Where the paper says it

An introduction to the theory of point processes: volume I: elementary theory and methods. An introduction to probability theory and its applications, Volume 2, volume 81.

Fries, J. F. (2002). Aging, natural death, and the compression of morbidity. p. 7

Where the paper says it

John Wiley & Sons. Aging, natural death, and the compression of morbidity.

Gelman, A., Carlin, J. p. 7

Where the paper says it

Bulletin of the World Health Organization, 80(3):245–250. , Carlin, J.

B., Stern, H. p. 7
Where the paper says it

, Carlin, J. , Stern, H.

S., and Rubin, D. p. 7
Where the paper says it

, Stern, H. , and Rubin, D.

B. (1995). Bayesian data analysis. p. 7
Where the paper says it

, and Rubin, D. Bayesian data analysis.

Genz, A. and Bretz, F. p. 3

Where the paper says it

As K increases, the dimensionality of the integration space grows, making exact computation increasingly challenging. Methods for calculating multivariate normal probabilities over such regions include numerical integration, orthant probabilities, and approximation techniques (Genz and Bretz, 2009).

(2009). Computation of multivariate normal and t probabilities, volume 195. p. 7
Where the paper says it

and Bretz, F. Computation of multivariate normal and t probabilities, volume 195.

Horiuchi, S. p. 0

Where the paper says it

Moreover, it is highly sensitive to changes in mortality at younger ages and aggregates the influence of mortality across all ages. As survival improves at older ages, this averaging can mask where mortality improvements actually occur and how they reshape the distribution of deaths (Fries, 2002; Wilmoth and Horiuchi, 1999).

, Ouellette, N. p. 0
Where the paper says it

Nonparametric approaches have been developed to address these limitations. One class of methods uses smoothing techniques—typically penalized splines—to estimate mortality rates and then determine the modal age from the resulting smooth surface (Ouellette and Bourbeau, 2011).

p. 0
Where the paper says it

Nonparametric approaches have been developed to address these limitations. One class of methods uses smoothing techniques—typically penalized splines—to estimate mortality rates and then determine the modal age from the resulting smooth surface (Ouellette and Bourbeau, 2011).

, Cheung, S. p. 7
Where the paper says it

, Ouellette, N. , Cheung, S.

L. K., and Robine, J. p. 7
Where the paper says it

, Cheung, S. , and Robine, J.

-M. (2013). Modal age at death: lifespan indicator in the era of longevity extension. p. 7
Where the paper says it

, and Robine, J. Modal age at death: lifespan indicator in the era of longevity extension.

Human Mortality Database (2024). p. 5

Where the paper says it

Advances in Gaussian process analysis and extreme value theory could further enhance this extension. 3 Empirical Application To validate our methodology, we applied it to mortality data from six countries: Denmark, France, Italy, Japan, the Netherlands, and the United States (US), covering the period from 1960 to 2020, obtained from the Human Mortality Database (2024).

Human mortality database, hmd. p. 7
Where the paper says it

Human Mortality Database (2024). Human mortality database, hmd.

Kannisto, V. p. 0

Where the paper says it

Accurate estimation of the mode is crucial for studying aging patterns, lifespan variation, and the dynamics of population aging. Yet, obtaining precise estimates is not straightforward, as most existing methods either estimate the mode from fitted models or rely on specific mathematical properties of the mode itself (Kannisto, 2001; Horiuchi et al.

(2001). Mode and dispersion of the length of life. p. 7
Where the paper says it

Kannisto, V. Mode and dispersion of the length of life.

Keyfitz, N. p. 0

Where the paper says it

, Booth and Tickle, 2008). In all cases, these approaches yield a single point estimate of the mode, and any assessment of uncertainty must be obtained indirectly—through resampling procedures such as bootstrapping or, when the model’s structure permits, by applying the delta method to estimated parameters (Andreev and Shkolnikov, 2010; Keyfitz and Caswell, 2005).

and Caswell, H. p. 0
Where the paper says it

, Booth and Tickle, 2008). In all cases, these approaches yield a single point estimate of the mode, and any assessment of uncertainty must be obtained indirectly—through resampling procedures such as bootstrapping or, when the model’s structure permits, by applying the delta method to estimated parameters (Andreev and Shkolnikov, 2010; Keyfitz and Caswell, 2005).

(2005). Applied mathematical demography. p. 7
Where the paper says it

and Caswell, H. Applied mathematical demography.

L´eger, A. p. 6

Where the paper says it

Data quality issues, such as age misreporting or incomplete records, can introduce biases in age-specific probabilities p, affecting the reliability of the estimated modal age distribution (Preston et al. Dependencies in mortality due to shared environmental or social factors, such as clustering in epidemics, also challenge the model’s assumption of independent deaths (L´eger and Mazzuco, 2021).

-E. and Mazzuco, S. p. 6
Where the paper says it

Data quality issues, such as age misreporting or incomplete records, can introduce biases in age-specific probabilities p, affecting the reliability of the estimated modal age distribution (Preston et al. Dependencies in mortality due to shared environmental or social factors, such as clustering in epidemics, also challenge the model’s assumption of independent deaths (L´eger and Mazzuco, 2021).

(2021). What can we learn from the functional clustering of mortality data? p. 7
Where the paper says it

and Mazzuco, S. What can we learn from the functional clustering of mortality data?

an application to the human mortality database. p. 5
Where the paper says it

Advances in Gaussian process analysis and extreme value theory could further enhance this extension. 3 Empirical Application To validate our methodology, we applied it to mortality data from six countries: Denmark, France, Italy, Japan, the Netherlands, and the United States (US), covering the period from 1960 to 2020, obtained from the Human Mortality Database (2024).

Ludkovski, M. p. 6

Where the paper says it

Additionally, extending the framework to a continuous setting could provide a more nuanced understanding of mortality dynamics. Gaussian processes and methods from extreme value theory may offer promising directions for analyzing the distribution of the mode in continuous age contexts (Ludkovski et al.

, Risk, J., and Zail, H. p. 7
Where the paper says it

Ludkovski, M. , and Zail, H.

(2018). Gaussian process models for mortality rates and improvement factors. p. 7
Where the paper says it

, and Zail, H. Gaussian process models for mortality rates and improvement factors.

Mi, X., Miwa, T., and Hothorn, T. p. 7

Where the paper says it

ASTIN Bulletin: The Journal of the IAA, 48(3):1307–1347. , and Hothorn, T.

(2009). mvtnorm: New numerical algorithm for multivariate normal probabilities. p. 7
Where the paper says it

, and Hothorn, T. mvtnorm: New numerical algorithm for multivariate normal probabilities.

Missov, T. I., Lenart, A. p. 7

Where the paper says it

1, 1(1):37–39. , Lenart, A.

, Nemeth, L. p. 7
Where the paper says it

, Lenart, A. , Nemeth, L.

, Canudas-Romo, V. p. 0
Where the paper says it

It identifies the age at which deaths are most concentrated, summarizing the prevailing location of adult mortality (Bergeron-Boucher et al. , 2015; Canudas-Romo, 2008).

, and Vaupel, J. p. 7
Where the paper says it

, Canudas-Romo, V. , and Vaupel, J.

W. (2015). The gompertz force of mortality in terms of the modal age at death. p. 7
Where the paper says it

, and Vaupel, J. The gompertz force of mortality in terms of the modal age at death.

Muirhead, R. p. 2

Where the paper says it

This result follows from the multivariate central limit theorem (e. , Muirhead, 2009).

J. (2009). Aspects of multivariate statistical theory. p. 7
Where the paper says it

Muirhead, R. Aspects of multivariate statistical theory.

Ouellette, N. and Bourbeau, R. p. 0

Where the paper says it

Nonparametric approaches have been developed to address these limitations. One class of methods uses smoothing techniques—typically penalized splines—to estimate mortality rates and then determine the modal age from the resulting smooth surface (Ouellette and Bourbeau, 2011).

(2011). Changes in the age-atdeath distribution in four low mortality countries: A nonparametric approach. p. 7
Where the paper says it

and Bourbeau, R. Changes in the age-atdeath distribution in four low mortality countries: A nonparametric approach.

Preston, S. p. 0

Where the paper says it

Keywords: Modal age at death, probabilistic framework, multinomial distribution, demographic analysis, mortality patterns 1 Context and Motivation Life expectancy at birth has long been the standard summary measure of mortality. It condenses the entire life table into a single number that is convenient for comparison across time and populations (Preston et al.

p. 0
Where the paper says it

Keywords: Modal age at death, probabilistic framework, multinomial distribution, demographic analysis, mortality patterns 1 Context and Motivation Life expectancy at birth has long been the standard summary measure of mortality. It condenses the entire life table into a single number that is convenient for comparison across time and populations (Preston et al.

, Heuveline, P. p. 7
Where the paper says it

Preston, S. , Heuveline, P.

, and Guillot, M. p. 7
Where the paper says it

, Heuveline, P. , and Guillot, M.

(2001). Demography: Measuring and modeling population processes. p. 7
Where the paper says it

, and Guillot, M. Demography: Measuring and modeling population processes.

Preston, S. H., Elo, I. T., and Preston, S. p. 0

Where the paper says it

Keywords: Modal age at death, probabilistic framework, multinomial distribution, demographic analysis, mortality patterns 1 Context and Motivation Life expectancy at birth has long been the standard summary measure of mortality. It condenses the entire life table into a single number that is convenient for comparison across time and populations (Preston et al.

H. (1999). Effects of age misreporting on mortality estimates at older ages. p. 7
Where the paper says it

, and Preston, S. Effects of age misreporting on mortality estimates at older ages.

Seeger, M. (2004). Gaussian processes for machine learning. p. 7

Where the paper says it

Population studies, 53(2):165–177. Gaussian processes for machine learning.

Vazquez-Castillo, P. p. 0

Where the paper says it

Patricio silviocabralpatricio@msn. com Paola Vazquez-Castillo pavaz@sdu.

, Bergeron-Boucher, M. p. 0
Where the paper says it

The modal age at death offers a more robust alternative. It identifies the age at which deaths are most concentrated, summarizing the prevailing location of adult mortality (Bergeron-Boucher et al.

-P., and Missov, T. p. 0
Where the paper says it

A variety of methods have been developed to estimate the mode. Parametric models, such as the Gompertz, Weibull, and logistic models, estimate the mode analytically from mortality curves (Kannisto, 2001; Missov et al.

I. (2024). Longevity a la mode: A discretized derivative tests method for accurate estimation of the adult modal age at death. p. 7
Where the paper says it

, and Missov, T. Longevity `a la mode: A discretized derivative tests method for accurate estimation of the adult modal age at death.

Wilmoth, J. p. 0

Where the paper says it

Moreover, it is highly sensitive to changes in mortality at younger ages and aggregates the influence of mortality across all ages. As survival improves at older ages, this averaging can mask where mortality improvements actually occur and how they reshape the distribution of deaths (Fries, 2002; Wilmoth and Horiuchi, 1999).

R. and Horiuchi, S. p. 0
Where the paper says it

Moreover, it is highly sensitive to changes in mortality at younger ages and aggregates the influence of mortality across all ages. As survival improves at older ages, this averaging can mask where mortality improvements actually occur and how they reshape the distribution of deaths (Fries, 2002; Wilmoth and Horiuchi, 1999).

(1999). Rectangularization revisited: Variability of age at death within human populations. p. 7
Where the paper says it

and Horiuchi, S. Rectangularization revisited: Variability of age at death within human populations.

--- Page 6 ---

The results, shown in Figure 1, show an overall upward trend in the modal age at death from 1960 to 2020 across the six countries. This trend reflects consistent improvements in longevity over time. p. 5

Where the paper says it

The results, shown in Figure 1, show an overall upward trend in the modal age at death from 1960 to 2020 across the six countries. This trend reflects consistent improvements in longevity over time.

Additionally, we observe gender-specific differences, with females consistently exhibiting a higher modal age at death than males. p. 5
Where the paper says it

This trend reflects consistent improvements in longevity over time. Additionally, we observe gender-specific differences, with females consistently exhibiting a higher modal age at death than males.

Some fluctuations are visible in the distribution, with temporary plateaus or slight declines in the mode occurring during specific periods. p. 5
Where the paper says it

This finding highlights the persistent gender gap in longevity. Some fluctuations are visible in the distribution, with temporary plateaus or slight declines in the mode occurring during specific periods.

--- Page 7 ---

The figure also highlights variability in the distribution shapes. Broader distributions, particularly in Denmark and the Netherlands, point to greater statistical uncertainty, potentially tied to smaller population sizes. p. 5

Where the paper says it

The figure also highlights variability in the distribution shapes. Broader distributions, particularly in Denmark and the Netherlands, point to greater statistical uncertainty, potentially tied to smaller population sizes.

Meanwhile, narrower distributions in countries like Japan suggest more stable and consistent modal age estimates. p. 5
Where the paper says it

Broader distributions, particularly in Denmark and the Netherlands, point to greater statistical uncertainty, potentially tied to smaller population sizes. Meanwhile, narrower distributions in countries like Japan suggest more stable and consistent modal age estimates, highlighting the importance of considering variability when interpreting the distribution of the mode.

--- Page 8 ---

Overall, the application of our methodology to mortality data across diverse national contexts highlights its utility in capturing trends and disparities in the modal age at death. By showcasing the full probability distribution of the modal age at death, it provides better picture of the dynamics of longevity and its improvements across populations. p. 5

Where the paper says it

Overall, the application of this methodology to mortality data across diverse national contexts highlights its utility in capturing trends and disparities in the modal age at death. By showcasing the full probability distribution of the modal age at death, it provides better picture of the dynamics of longevity and its improvements across populations.

--- Page 7 ---

This result shows that as n increases, P(M = x⋆) converges to 1. This result implies that the real modal age falls within the interval x⋆ as n becomes large. p. 4

Where the paper says it

, limn→∞ P(M = x ⋆ ) = 1, where M is the modal age at death. This result implies that the real modal age falls within the interval x ⋆ as n becomes large.

--- Page 7 ---

The interval with the highest px serves as a probabilistic approximation. For finite n, however, random fluctuations in the data introduce variability, resulting in uncertainty about the modal age. p. 4

Where the paper says it

Unlike approaches that rely on the derivative of an estimated density function, our method avoids the challenges associated with identifying local maxima or dealing with saddle points. For finite n, however, random fluctuations in the data introduce variability, resulting in uncertainty about the modal age.

--- Page 7 ---

Estimating the probability distribution of the mode allows us to quantify this uncertainty and provides a clearer picture of the likelihood that each age interval contains the mode. The application of our methodology to mortality data across diverse national contexts highlights its utility in capturing trends and disparities in the modal age at death. p. 5

Where the paper says it

Meanwhile, narrower distributions in countries like Japan suggest more stable and consistent modal age estimates, highlighting the importance of considering variability when interpreting the distribution of the mode. Overall, the application of this methodology to mortality data across diverse national contexts highlights its utility in capturing trends and disparities in the modal age at death.

--- Page 7 ---

By showcasing the full probability distribution of the modal age at death, it provides better picture of the dynamics of longevity and its improvements across populations.

--- Page 7 ---

This framework can also be extended to leverage the Human Cause-of-Death (HCD) series recently incorporated into the Human Mortality Database. By modeling age-specific death counts disaggregated by cause, the framework can estimate the probability distribution of the modal age at death for each specific cause. p. 6

Where the paper says it

This framework can also be extended to leverage the Human Cause-of-Death (HCD) series recently incorporated into the Human Mortality Database. By modeling age-specific death counts disaggregated by cause, the framework can estimate the probability distribution of the modal age at death for each specific cause.

--- Page 7 ---

This extension would offer a clearer view of how different causes shape the distribution of deaths around the mode—both in terms of central tendency and variability.

Where the gains come from

The approach provides a probabilistic perspective that explicitly quantifies variability, reveals competing age intervals, and improves robustness to data fluctuations. This method aligns with the discrete nature of mortality data and addresses the variability often overlooked by traditional point-estimation methods. p. 6

Where the paper says it

By modeling age-specific death counts as outcomes of a multinomial experiment and incorporating the dependency among counts due to the fixed total number of deaths, our method provides an empirical estimate of the probability distribution of the modal age. This approach aligns with the discrete nature of mortality data and addresses the variability often overlooked by traditional point-estimation methods.

Limitations

The Gaussian approximation assumes a large total number of deaths (n) and smooth mortality distributions. In small populations or during mortality shocks, these assumptions may not hold, potentially leading to inaccuracies in the estimated probabilities. p. 6

Where the paper says it

In small populations or during mortality shocks (e. , pandemics or sudden socio-economic disruptions), these assumptions may not hold, potentially leading to inaccuracies in the estimated probabilities.

Data quality issues can introduce biases in age-specific probabilities p, affecting the reliability of the estimated modal age distribution. p. 6
Where the paper says it

, pandemics or sudden socio-economic disruptions), these assumptions may not hold, potentially leading to inaccuracies in the estimated probabilities. Data quality issues, such as age misreporting or incomplete records, can introduce biases in age-specific probabilities p, affecting the reliability of the estimated modal age distribution (Preston et al.

Future Directions

To address these limitations, future research could explore methods that account for overdispersion or clustering, such as the Bell distribution or Bayesian approaches. Extending the framework to a continuous setting could provide a more nuanced understanding of mortality dynamics. p. 6

Where the paper says it

Integrating Monte Carlo methods or hierarchical models could enhance the model’s robustness to mortality shocks and small population sizes. Additionally, extending the framework to a continuous setting could provide a more nuanced understanding of mortality dynamics.

This framework can also be extended to leverage the Human Cause-of-Death (HCD) series recently incorporated into the Human Mortality Database.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve uncertainty quantification in mortality modeling by shifting from single point estimates to estimating a probability distribution of modal age at death, which quantifies uncertainty directly and aligns with the discrete nature of mortality data. This allows AI systems to predict not just a single median lifespan but also the range of plausible modal ages, improving risk assessment accuracy.

  2. Enhance demographic forecasting by using the convergence property that as the total number of deaths n increases, the probability distribution of the modal age at death converges to a point mass at the age interval with the highest underlying death probability pxstar. This enables AI models to robustly project long-term mortality trends based on observed data, providing more reliable future population aging predictions.

  3. Develop more accurate risk stratification tools by utilizing gender-specific modal age estimates derived from the empirical application, which show[s] that females consistently exhibiting a higher modal age at death than males. AI systems can then use this to create tailored health and longevity risk profiles for different genders with greater demographic fidelity.

  4. Increase robustness against data fluctuations by employing a framework that addresses the variability often overlooked by traditional point-estimation methods. This makes AI models less susceptible to noise or sudden shocks in mortality data, leading to more stable and reliable predictions in real-world scenarios like pandemic response planning.

  5. Extend predictive modeling to continuous age settings by modeling death counts as a Gaussian process, which provides a more precise representation of mortality patterns when age is treated as a continuous variable. This allows AI systems to analyze mortality dynamics across the entire lifespan rather than just discrete intervals, capturing finer nuances in aging processes.

Abstract

The modal age at death is used to track longevity, but existing methods report it as a single number. We estimate its posterior distribution. Death counts are modeled as Poisson given the death rates and exposures, and posterior draws of the rates, passed through the life table, give the probability that each age is the mode and a credible interval for an interpolated modal age. With independent gamma priors, the posterior is computed exactly and concentrates on the true mode as the population grows. Because single-year rates are noisy, we also use a second-order random-walk prior on the log rates, which shares information across neighboring ages and estimates the degree of smoothing from the data; this posterior is computed with a Laplace approximation, which we check against posterior sampling. In simulations, the random-walk posterior mean was 1.4 to 10.6 times more accurate than the posterior mean under independent priors. When the distribution of deaths was unimodal, its errors were close to those of P-spline and Gompertz estimates, which were slightly more accurate; with a secondary peak, P-splines were the most accurate. The random-walk credible intervals were conservative, covering the true mode in 98% to 100% of replications, whereas bootstrap intervals for the other estimators either covered it less often or were much wider. For women in Denmark, France, Japan and Sweden since 1960, the random-walk estimates hardly changed with the starting age of the estimation, and the posterior gives the probability that the modal age declined in 2020.

Related papers