Search

In recent years, the importance of describing mortality at the limits of the life span has led to a number of relevant and controversial studies. Whereas considerable efforts have been devoted to collecting data and estimating models on the oldest-old individuals, the testing of statistical confidence about the conclusions of analyses at extreme ages has been largely neglected. How certain can we be in saying that the risk of dying increases, levels out, or, paradoxically, decreases over age 105? Can we recognize particular mortality age patterns at such high ages? In this paper, it is shown that very little can be confidently asserted about mortality at extreme ages. Instead of analyzing actual data, we perform a series of simulation studies mimicking actual scenarios from controlled mechanisms. Our findings are thus robust with respect to factors such as particular observation schemes, heterogeneity, and data quality issues. Given the sample sizes currently available and the levels of mortality experienced in present populations, we show that before age 110, only a Gompertzian increase of mortality may be detected. Afterwards a plateau will be regularly recognized as the most suitable pattern, regardless of the complexity of the true underlying mortality.

Italy was hit harshly by the Covid-19 pandemic, registering more than 35,000 Covid-19 deaths between February and July 2020. During this first wave of the epidemic, the virus spread unequally across the country, with northern regions witnessing more cases and deaths. We investigate demographic and socio-economic factors contributing to the diverse regional impact of the virus during the first wave. Using generalized additive mixed models, we find that Covid-19 mortality at regional level is negatively associated with the degree of intergenerational co-residence, number of intensive care unit beds per capita, and delay in the outbreak of the epidemic. Conversely, we do not find strong associations for several variables highlighted in recent literature, such as population density or the share of the population who are older or have at least one chronic disease. Our results underscore the importance of context-specific analysis for the study of a pandemic.

The Covid-19 pandemic did not affect sub-national regions in a uniform way. Knowledge of the impact of the pandemic on mortality at the local level is therefore an important issue for better assessing its burden. Vital statistics are now available for an increasing number of countries for 2020 and 2021, and allow the calculation of sub-national excess mortality. However, this calculation faces two important methodological challenges: (1) it requires appropriate mortality projection models; (2) small populations implies important uncertainty in the estimates, commonly neglected. We address both issues by adopting a method to forecast mortality at sub-national level and by incorporating uncertainty in the computation of mortality measures. We illustrate our approach to Frenchdépartements(NUTS 3, 95 geographical units) and produce estimates for 2020 and both sexes. Nonetheless, the proposed approach is so flexibility to allow estimation of excess mortality during Covid-19 in most demographic scenarios as well as for past pandemics.

The COVID-19 pandemic is characterized by successive waves that each developed differently over time and through space. We aim to provide an in-depth analysis of the evolution of COVID-19 mortality during 2020 and 2021 in a selection of countries.We focus on five European countries and the United States. Using standardised and age-specific mortality rates, we address variations in COVID-19 mortality within and between countries, as well as demographic characteristics and seasonality patterns.Our results highlight periods of acceleration and deceleration in the pace of COVID-19 mortality, with substantial differences across countries. Periods of stabilization were identified during summer (especially in 2020) among the European countries analysed, but not in the United States. The latter stands out as the study population with the highest COVID-19 mortality at young ages. In general, COVID-19 mortality is highest at old ages, particularly during winter. Compared to women, men have higher COVID-19 mortality rates at most ages and in most seasons.There is seasonality in COVID-19 mortality for both sexes at all ages, characterized by higher rates during winter. In 2021, the highest COVID-19 mortality rates continued to be observed at ages 75+, despite vaccinations having specifically targeted those ages.

Une hausse temporaire du risque de décès au moment de l’entrée dans l’âge adulte a été observée dans de nombreuses populations sans qu’une explication n’émerge clairement jusqu’ici. Plusieurs études récentes permettent cependant de mieux cerner les principales caractéristiques de ce phénomène, qui s’avère historiquement non universel, davantage masculin sans toutefois que les femmes y échappent, et dominé par les morts violentes.

Background: The COVID-19 pandemic has caused major shocks in mortality trends in many countries. Yet few studies have evaluated the heterogeneity of the mortality shock at the sub-national level, rigorously accounting for the different sources of uncertainty.Methods: Using death registration data from Belgium, we first assess the change in the heterogeneity of subnational standardized mortality ratios in 2020, when compared to previous years. We then measure the shock of the pandemic using district-level values of life expectancy, comparing the observed and projected districts life expectancy, accounting for all sources of uncertainty (related to the life-table construction at district level and to the projection methods at country and district level). The Bayesian modelling approach makes it easy to combine the different sources of uncertainty in the assessment of the shock. This is of particular interest at a finer geographical scale characterized by high stochastic variation in annual death counts.Results: The heterogeneity in the impact of the pandemic on all-cause mortality across districts is substantial, with some districts barely showing any impact whereas the Bruxelles-Capital and Mons districts experienced a decrease in life expectancy at birth of 2.24 (95% CI:1.33-3.05) and 2.10 (95% CI:0.86-3.30) years, respectively. The year 2020 was associated with an increase in mortality levels ' heterogeneity at a subnational level in comparison to past years measured by both the standardized mortality ratios and the life expectancies at birth. Decisions on uncertainty thresholds have a large bearing on the interpretation of the results.Conclusion: Developing sub-national mortality estimates with their uncertainty is key to understanding why certain areas have been hard hit in comparison to others.

The Lee-Carter (LC) model represents a landmark paper in mortality forecasting. While having been widely accepted and adopted, the model has some limitations that hinder its performance. Some variants of the model have been proposed to deal with these drawbacks individually, none coped with them all at the same time. In this paper, we propose a Three-Component smooth Lee-Carter (3C-sLC) model which overcomes many of the issues simultaneously. It decomposes mortality development into childhood, early-adult and senescent mortality, which are described, individually, by a smooth variant of the LC model. Smoothness is enforced to avoid irregular patterns in projected life tables, and complexity in the forecasting methodology is unaltered with respect to the original LC model. Component-specific schedules are considered in projections, providing additional insights into mortality forecasts. We illustrate the proposed approach to mortality data for ten low-mortality populations. The 3C-sLC captures mortality developments better than a smooth improved version of the LC model, and it displays wider prediction intervals. The proposed approach provides actuaries, demographers, epidemiologists and social scientists in general with a unique and valuable tool to simultaneously smooth, decompose and forecast mortality.

Drawing cohort profiles and cohort forecasts from grids of age-period data is common practice in demography. In this research note, we: (1) estimate the bias in the cohort TFR and life expectancies calculated from such data, and (2) demonstrate that cohort Lee-Carter forecasts drawn from an age-period grid have implausible prediction intervals, especially when drawn from the upper and lower bounds of the time series parameter. These biases are surprisingly large, even when the cohort profiles are created from single-age, single-year period data. The danger is that we overinterpret deviations from expected trends, that in actual fact were induced by our own data manipulation.

L’urgence que représente la pandémie de COVID‑19 a entraîné des différences considérables entre les processus de collecte des données des pays, qui s’efforcent tous de produire des informations en temps réel mais qui restent des statistiques de mortalité imparfaites. Pour remédier à ce problème, nous analysons les décomptes de décès par COVID‑19 provenant de la base de données « La démographie des décès par COVID‑19 » (https://dc-covid.site.ined.fr/fr/) et en examinons les limites. Nous décrivons et illustrons des aspects importants touchant aux données et qui limitent la possibilité de mener des comparaisons internationales. Pour aplanir ces difficultés, nous classons les sources en fonction du caractère exhaustif des données qu’elles fournissent puis nous analysons et comparons les décomptes de décès pour 16 pays. Enfin, nous insistons sur l’importance de bien comprendre les caractéristiques de la collecte des données et formulons des recommandations pour le traitement des statistiques imparfaites.

Regular revisions of the classification of diseases and the consequent disruptions of mortality series are well-known issues in long-term cause-of-death analysis. Given basic assumptions and medical knowledge about possible exchanges across causes of death in the revision years, redistribution of counts of causes of death into a new classification can be viewed as a constrained optimization problem. Penalized likelihood within a quadratic programming framework allows estimation of exchanges that vary smoothly over age groups. The approach is illustrated using both German data on malignant neoplasms and French data on heart diseases.

International audience; Non-pharmaceutical interventions have been implemented worldwide to curb the spread of COVID-19. However, the effectiveness of such governmental measures in reducing the mortality burden remains a key question of scientific interest and public debate. In this study, we leverage digital mobility data to assess the effects of reduced human mobility on excess mortality, focusing on regional data in England and Wales between February and August 2020. We estimate a robust association between mobility reductions and lower excess mortality, after adjusting for time trends and regional differences in a mixed-effects regression framework and considering a five-week lag between the two measures. We predict that, in the absence of mobility reductions, the number of excess deaths could have more than doubled in England and Wales during this period, especially in the London area. The study is one of the first attempts to quantify the effects of mobility reductions on excess mortality during the COVID-19 pandemic.

International audience; Empirical research on human mortality and extreme longevity suggests that the risk of death among the oldest-old ceases to increase and levels off at age 110. The universality of this finding remains in dispute because of two main reasons: i) high uncertainty around statistical estimates generated from scarce data, and ii) the lack of country-specific comparisons. In this article, we estimate age patterns of mortality above age 105 using data from the International Database on Longevity, an exceptionally large and recently updated database comprising more than 13,000 validated records of long-lived individuals from eight populations. We show that, in all of them, similar mortality trajectories arise, suggesting that the risk of dying levels off after age 105. As more high-quality data become available, there is more evidence in support of a levelling-off of the risk of dying as a regularity of longevous populations.

Mortality forecasting has crucial implications for insurance and pension policies. A large amount of literature has proposed models to forecast mortality using cross-sectional (period) data instead of longitudinal (cohort) data. As a consequence, decisions are generally based on period life tables and summary measures such as period life expectancy, which reflect hypothetical mortality rather than the mortality actually experienced by a cohort. This study introduces a novel method to forecast cohort mortality and the cohort life expectancy of non-extinct cohorts. The intent is to complete the mortality profile of cohorts born up to 1960. The proposed method is based on the penalized composite link model for ungrouping data. The performance of the method is investigated using cohort mortality data retrieved from the Human Mortality Database for England & Wales, Sweden, and Switzerland for male and female populations.

Mortality forecasting has received increasing interest during recent decades due to the negative financial effects of continuous longevity improvements on public and private institutions’ liabilities. However, little attention has been paid to forecasting mortality from a cohort perspective. In this article, we introduce a novel methodology to forecast adult cohort mortality from age-at-death distributions. We propose a relational model that associates a time-invariant standard to a series of fully and partially observed distributions. Relation is achieved via a transformation of the age-axis. We show that cohort forecasts can improve our understanding of mortality developments by capturing distinct cohort effects, which might be overlooked by a conventional age–period perspective. Moreover, mortality experiences of partially observed cohorts are routinely completed. We illustrate our methodology on adult female mortality for cohorts born between 1835 and 1970 in two high-longevity countries using data from the Human Mortality Database.

Mortality forecasting has recently received growing interest, as accurate projections of future lifespans are needed to ensure the solvency of insurance and pension providers. Several innovative stochastic methodologies have been proposed in most recent decades, the majority of them being based on age-specific mortality rates or on summary measures of the life table. The age-at-death distribution is an informative life-table function that provides readily available information on the mortality pattern of a population, yet it has been mostly overlooked for mortality projections. In this chapter, we propose to analyse and forecast mortality developments over age and time by introducing a novel methodology based on age-at-death distributions. Our approach starts from a nonparametric decomposition of the mortality pattern into three independent components corresponding to Childhood, Early-Adulthood and Senescence, respectively. We then model the evolution of each component-specific death density with a relational model that associates a time-invariant standard to a series of observed distributions by means of a transformation of the age axis. Our approach allows us to capture mortality developments over age and time, and forecasts can be derived from parameters’ extrapolation using standard time series models. We illustrate our methods by estimating and forecasting the mortality pattern of females and males in two high-longevity countries using data of the Human Mortality Database. We compare the forecast accuracy of our model and its projections until 2050 with three other forecasting methodologies.

Background: Women earn less than men at most career stages. Explanations for a gender gap in wages include gender differences in the allocation of household and domestic work. At the family level, a marital age difference is an important shared characteristic that might play a role in determining a woman’s career trajectory, and, therefore, her income. Since women tend to marry older men, we investigate whether women whose husbands are older have lower incomes than women whose husbands are the same age or younger. Objective: This study investigates whether the age gap between a woman and her partner was associated with her income in Denmark in 2010. Methods: We use data on Danish female twin pairs in 2010. Our design includes samples within twin pairs (n = 4,716) and pooled twin samples (n = 13,354) to account for differences in early household environments and uses unconditional quantile regression to model the association between the age gap and the woman's income. Results: We find a statistically significant association between the marital age gap and the woman's income. The form of this association appears to be complex and varies across the income and age gap distribution. However, the magnitude of the estimated effects is small in economic terms. Conclusions: These results suggest that the marital age gap is unlikely to be an important determinant of a woman’s income, at least in Denmark. Contribution: To our knowledge, this is the first study that explores the association between marital age differences and a woman's earnings using a twin design and high-quality register data. Background: Women earn less than men at most career stages. Explanations for a gender gap in wages include gender differences in the allocation of household and domestic work. At the family level, a marital age difference is an important shared characteristic that might play a role in determining a woman’s career trajectory, and, therefore, her income. Since women tend to marry older men, we investigate whether women whose husbands are older have lower incomes than women whose husbands are the same age or younger.Objective: This study investigates whether the age gap between a woman and her partner was associated with her income in Denmark in 2010.Methods: We use data on Danish female twin pairs in 2010. Our design includes samples within twin pairs (n = 4,716) and pooled twin samples (n = 13,354) to account for differences in early household environments and uses unconditional quantile regression to model the association between the age gap and the woman's income.Results: We find a statistically significant association between the marital age gap and the woman's income. The form of this association appears to be complex and varies across the income and age gap distribution. However, the magnitude of the estimated effects is small in economic terms.Conclusions: These results suggest that the marital age gap is unlikely to be an important determinant of a woman’s income, at least in Denmark.Contribution: To our knowledge, this is the first study that explores the association between marital age differences and a woman's earnings using a twin design and high-quality register data.

Age-at-death distributions provide an informative description of the mortality pattern of a population but have generally been neglected for modelling and forecasting mortality. In this paper, we use the distribution of deaths to model and forecast adult mortality. Specifically, we introduce a relational model that relates a fixed ‘standard’ to a series of observed distributions by a transformation of the age axis. The proposed Segmented Transformation Age-at-death Distributions (STAD) model is parsimonious and efficient: using only three parameters, it captures and disentangles mortality developments in terms of shifting and compression dynamics. Additionally, mortality forecasts can be derived from parameter extrapolation using time-series models. We illustrate our method and compare it with the Lee–Carter model and variants for females in four high-longevity countries. We show that the STAD fits the observed mortality pattern very well, and that its forecasts are more accurate and optimistic than the Lee–Carter variants.

(ProQuest: ... denotes formulae omitted.)1.IntroductionThe spectacular longevity gains achieved by low mortality countries in the last two centuries are the direct consequence of ongoing progress in fighting a wide variety of diseases. During the first half of the 1900s, advances in immunization, sanitation, and nutrition led to the near-eradication of major infectious and parasitic diseases, as well as maternal, perinatal, and nutritional disorders (Horiuchi 1999, Omran 1971, Preston 1976). Throughout this period, life expectancy at birth increased steadily and very rapidly, reaching a level of about 65-70 years old for both sexes combined by the mid20th century (Human Mortality Database). As fewer individuals succumbed to these diseases in the early years of life and increasingly survived to older ages, the risk of developing chronic degenerative diseases became more likely. By the early 1950s these age-associated conditions replaced infectious diseases as the most common cause of death in industrialized societies. After the fall of mortality at young and middle ages to historically low levels, gains in life expectancy at birth slowed down for about ten years, leading some demographers to argue that mortality improvements were no longer possible.However, the late 1960s ushered in a new era of progress characterized by the decline of chronic degenerative diseases, particularly cardiovascular diseases. Kannisto (2001) referred to an "era of delayed aging," as mortality at older ages dropped faster than at younger ages and deaths above age 80 started to decline considerably for the first time (Kannisto et al. 1994). Thereafter, advances in the extension of human life were mainly fuelled by improvements in old-age survival (Mesle and Vallin 2006, Wilmoth et al. 2000). Consequently, the upturn in life expectancy at birth resumed but the pace of increase was slower during this new era than in preceding years.Recent evidence suggests that life expectancy at birth underestimates progress made in longevity extension when such progress is principally driven by improvements in the survival of the elderly (Canudas-Romo 2010, Kannisto 2001). Therefore, life expectancy at some middle or early old ages, such as 50 or 65, takes center stage in analyses monitoring old-age survival. One of the main disadvantages of these lifespan measures, however, is their dependence on an arbitrary selection of the age limit, and hence on an arbitrary definition of 'old' (Kannisto 2001). The conditionality upon survival to an age threshold constitutes another important disadvantage, as it reflects the average length of life still to be lived by the most robust individuals.To make up for these shortcomings, Kannisto (2001), building on earlier work by Quetelet (1835) and Lexis (1877, 1878), proposed the late modal age at death (M) as a promising lifespan indicator in an era where the extension of human life is primarily due to the reduction of old-age mortality. Under a given mortality regime, this measure represents the most common (i.e., most frequent) or 'typical' length of life among adults. In the early years of its introduction to contemporary demography the attractiveness of M was its freedom from any arbitrary age limit. More recently, studies have also demonstrated that the late mode is solely determined by old-age survival and that the pattern of trends and differentials in M can differ greatly from those in life expectancy, at birth, or at some early old age (Canudas-Romo 2010, Cheung et al. 2009, Horiuchi et al. 2013, Office of National Statistics 2012). Therefore, a growing number of researchers have been monitoring longevity gains in low mortality countries through the lens of the late modal age at death (Brown et al. 2012, Canudas-Romo 2008, Cheung and Robine 2007, Cheung et al. 2005, 2009, Cheung, Robine, and Caselli 2008, Kannisto 2007, Ouellette and Bourbeau 2011, Ouellette, Bourbeau, and Camarda 2012, Robine 2001, Thatcher et al. …

Representing the conditional mean in Poisson regression directly as a sum of smooth components can provide a realistic model of the data generating process. Here, we present an approach that allows such an additive decomposition of the expected values of counts. The model can be formulated as a penalized composite link model and can, therefore, be estimated by a modified iteratively weighted least-squares algorithm. Further shape constraints on the smooth additive components can be enforced by additional penalties, and the model is extended to two dimensions. We present two applications that motivate the model and demonstrate the versatility of the approach.

We propose a method to decompose the young adult mortality hump by cause of death. This method is based on a flexible shape decomposition of mortality rates that separates cause-of-death contributions to the hump from senescent mortality. We apply the method to U.S. males and females from 1959 to 2015. Results show divergence between time trends of hump and observed deaths, both for all-cause and cause-specific mortality. The study of the hump shape reveals age, period, and cohort effects, suggesting that it is formed by a complex combination of different forces of biological and socioeconomic nature. Male and female humps share some traits in all-cause shape and trend, but they also differ by their overall magnitude and cause-specific contributions. Notably, among males, the contributions of traffic and other accidents were progressively replaced by those of suicides, homicides, and poisonings; among females, traffic accidents remained the major contributor to the hump.

MortHumpis an R package designed to provide ready-to-use methods to analyse the young adult mortality hump. It contains functions to•format all-cause and cause-of-death data from the Human Mortality Database (HMD) and the Human Cause-of-Death Database (HCD) respectively,•identify and group causes of death that are likely to contribute to the young adult mortality hump,•estimate parametric and non-parametric models that isolate the young adult mortality hump from the rest of the force of mortality, decomposing when needed by cause of death,•measure the young adult mortality hump by computing summary statistics about its magnitude, location and spread, optionally by cause of death.This technical paper is meant as a user guide for the MortHump package and provides examples on how to use its functions

Every time the classification of causes of death is changed, time series of deaths by cause are disrupted in more or less profound ways. When changes involve only the merging of several items or splitting a single item into several new categories, the problems caused by these ruptures are not too difficult to solve. A more or less severe imbroglio occurs, however, each time a new item results from recombining portions of different split items. Sometimes, but very rarely, some countries proceed to a bridge coding during the year of transition, which can help reconstruct coherent time series. This article first summarizes the general principles of the method developed for France by Meslé and Vallin to reconstruct complete series for France from 1925 to 1999 in the detailed list of the 9th WHO International Classification of Diseases (ICD), doing so by successively bridging a posteriori the five versions of the ICD that were in use during that period. Second, it reports on several methodological improvements that have been developed with the aim to reconstruct and analyze mortality trends by cause in sixteen industrialized countries. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s10680-017-9453-1) contains supplementary material, which is available to authorized users.

Women earn less than men at most career stages, and they also tend to partner with older men. This study investigates whether being the younger partner in a marriage reduces a woman’s incentive to pursue an independent career. We hypothesize that the income gender gap might be partially explained by the age differences between spouses. Using both a within-twin (n = 4716) and pooled-twin (n = 13354) design to more readily account for differences in early household environments, we investigated for Denmark whether the age gap between a female twin and her partner has any influence on her income. The hypothesis could not be confirmed, as the age gap between partners did not appear to be associated with women’s earnings. The finding that women’s wages were generally unaffected by partnering with an older man could be a result of heterogeneous groups of women entering men-older partnerships. Future research should explore this question further by using the number of promotions to assess the career success of women, and should extend this work to countries with different social welfare systems and less egalitarian gender norms.

We propose a method to decompose the young adult mortality hump by cause of death. This method is based on a flexible shape-decomposition of mortality rates that separates cause-of-death contributions to the hump from senescent mortality. We apply the method to US males and females from 1959 to 2010. Results show divergences between time trends of hump and observed deaths, both for all-cause and cause-specific mortality. The study of the hump shape reveals age, period and cohort effcts, suggesting that it is formed by a complex combination of different forces of biological and socioeconomic nature. Male and female humps share some traits in all-cause shape and trend, but also differ by their overall magnitude and cause-specific contributions. Notably, among males the contributions of traffic and other accidents were progressively replaced by those of suicides, homicides and poisonings, whereas among females traffic accidents remained the major contributor to the hump.

Summary Digit preference is the habit of reporting certain end digits more often than others. If such a misreporting pattern is a concern, then measures to reduce digit preference can be taken and monitoring changes in digit preference becomes important. We propose a two-dimensional penalized composite link model to estimate the true distributions unaffected by misreporting, the digit preference pattern and a trend in the preference pattern simultaneously. A transfer pattern is superimposed on a series of smooth latent distributions and is modulated along a second dimension. Smoothness of the latent distributions is enforced by a roughness penalty. Ridge regression with an L1-penalty is used to extract the misreporting pattern, and an additional weighted least squares regression estimates the modulating trend vector. Smoothing parameters are selected by the Akaike information criterion. We present a simulation study and apply the model to data on birth weight and on self-reported weight of adults.

Ageing results from the accumulation of multifactorial damage over time. However, the temporal distribution of this damage remains unknown. In seasonal species, transitions between seasons are critical periods of massive physiological remodelling. We hypothesised that these recurrent peaks of physiological remodelling are costly in terms of survival. We tested whether captive small primates exposed to an experimentally increased frequency of seasonal transitions die sooner than individuals living under natural seasonality. The results show that experiencing one additional season per year increases the mortality hazard by a factor of 3 to 4, whereas the expected number of seasons lived is only slightly impacted by the seasonal rhythm. These results demonstrate that physiological transitions between periods of high and low metabolic activity represent a major mortality risk for seasonal organisms, which has been ignored until now.

In many applications data can be interpreted as indirect observations of a latent distribution. A typical example is the phenomenon known as digit preference, i.e. the tendency to round outcomes to pleasing digits. The composite link model (CLM) is a useful framework to uncover such latent distributions. Moreover, when applied to data showing digit preferences, this approach allows estimation of the proportions of counts that were transferred to neighbouring digits. As the estimating equations generally are singular or severely ill-conditioned, we impose smoothness assumptions on the latent distribution and penalize the likelihood function. To estimate the misreported proportions, we use a weighted least-squares regression with an added L1 penalty. The optimal smoothing parameters are found by minimizing the Akaike’s information Criterion (AIC). The approach is verified by a simulation study and several applications are presented.

We investigate whether heterogeneity can explain the differences in mortality between the United States and a more homogeneous country, i.e. Japan. The background of the analysis is the growing gap between life-expectancy in the United States and the world record leader since the 1980s. This development is mirrored in the decelerated improvements made against mortality during the last 25 years - especially for women. Our hypothesis is that it is possible to have survival improvements at a steady pace across all population subgroups. The proportion of the population subgroups which perform relatively well decreased over time whereas poor performing subgroups became more important. The consequence of this compositional change since the 1980s would be the observed levelling off in survival improvements on the population level. Since official mortality estimates in the US are of poor quality, we estimate mortality rates via the extinct cohort method using Medicare enrollment data from the Social Security Administration due to their higher quality. If our hypothesis was true, we should see little differences in the development of mortality rates between the best performing subgroups in the United States and Japan. We compared mortality rates between Japanese women and white women in the US from states who (a) either had the highest life expectancy or (b) the highest median household income or (c) the largest gross state product per head. Regardless of the choice of indicator, mortality rates levelled-off in the United States for those best performance subgroups as for the whole population. The preliminary conclusion is therefore that compositional changes (heterogeneity) in the US population can not explain the widening gap between US life expectancy and best practice life expectancy.

ObjectiveTo measure the burden of the COVID-19 pandemic in 2020 at the subnational level by estimating excess mortality, defined as the increase in all-cause mortality relative to an expected baseline mortality level.DesignStatistical and demographic analyses of regional all-cause mortality data.SettingThe vital statistics systems of 21 European countries.ParticipantsThe entire population of 561 spatial units in 21 European countries.Main Outcome MeasuresLosses of life expectancy at ages 0 and 60 for males and females.ResultsWe found evidence of a loss in life expectancy in 391 regions, while only three regions exhibit notable gains in life expectancy in 2020. For 12 regions, losses of life expectancy amounted to more than 2 years, and three regions showed losses greater than 3 years. We highlight geographic clusters of high mortality in Northern Italy, Spain and Poland, while clusters of low mortality were found in Western France, Germany/Denmark and Norway/Sweden.ConclusionsRegional differences of loss of life expectancy are impressive, ranging from a loss of more than 4 years to a gain of 8 months. These findings provide a strong rationale for regional analysis, as national estimates hide significant regional disparities.‘What is already known’Reported numbers of covid-19 deaths are subject to changes within and across countries due to inaccuracy, and incompleteness.Excess mortality measured by loss in life expectancy is widely considered a relevant indicator for assessing the total mortality impact of the COVID-19 pandemic.Whereas national estimates has been largely provided, few scattered regional studies for specific countries have been conducted.‘What this study adds’This study provides the first coherent analysis of excess mortality at regional level covering a large number of European countries.It allows to properly map COVID-19 pandemic using official mortality data routinely collected by vital registration systems, which are less sensitive to misclassification.How this study might affect research, practice or policy’This study provides a strong rationale for regional analysis, as national estimates hide significant regional disparities