Littérature scientifique sur le sujet « Generalized Age-Period-Cohort »

AbstractGeneralized Additive Models (GAMs) with age, period and cohort as possible covariates are used to predict future mortality improvements for the Irish population. The GAMs considered are the 1-dimensional age + period and age + cohort models and the 2-dimensional age-period and age-cohort models. In each case thin plate regression splines are used as the smoothing functions. The generalized additive models are compared with the P-Spline (Currie et al., 2004) and Lee-Carter (Lee & Carter, 1992) models included in version 1.0 of the Continuous Mortality Investigation (CMI) library of mortality projections. Using the Root Mean Square Error to assess the accuracy of future predictions, the GAMs outperform the P-Spline and Lee-Carter models over intervals of 25 and 35 years in the age range 60 to 90. The GAMs allow intuitively simple models of mortality to be specified whilst also providing the flexibility to model complex relationships between the covariates. The majority of morality improvements derived from the projections of future Irish mortality yield annuity values at ages 60, 65, 70 and 80 in 2007 in the range of annuity values calculated, assuming a 2 to 4 percent annual compound improvement in mortality rates for both males and females.

Summary Many public health databases index disease counts by age groups and calendar periods within geographic regions (e.g., states, districts, or counties). Issues around relative risk estimation in small areas are well-studied; however, estimating trend parameters that vary across geographic regions has received less attention. Additionally, small counts (e.g., $<10$) in most publicly accessible databases are censored, further complicating age–period–cohort (APC) analysis in small areas. Here, we present a novel APC model with left-censoring and spatially varying intercept and trends, estimated with correlations among contiguous geographic regions. Like traditional models, our model captures population-scale trends, but it can also be used to characterize geographic disparities in relative risk and age-adjusted trends over time. To specify the joint distribution of our three spatially varying parameters, we adapt the generalized multivariate conditional autoregressive prior, previously used for multivariate disease mapping. Specified in this manner, region-specific parameters are correlated spatially, and also to one another. Estimation is performed using the No-U-Turn Hamiltonian Monte Carlo sampler in Stan. We conduct a simulation study to assess the performance of the proposed model relative to the standard model, and conclude with an application to US state-level opioid overdose mortality in men and women aged 15–64 years.

To what extent are new generations ‘Thatcherite’? Using British Social Attitudes data for 1985–2012 and applying age-period-cohort analysis and generalized additive models, this article investigates whether Thatcher’s Children hold more right-authoritarian political values compared to other political generations. The study further examines the extent to which the generation that came of age under New Labour – Blair’s Babies – shares these values. The findings for generation effects indicate that the later political generation is even more right-authoritarian, including with respect to attitudes to redistribution, welfare and crime. This view is supported by evidence of cohort effects. These results show that the legacy of Thatcherism for left-right and libertarian-authoritarian values is its long-term shaping of public opinion through political socialization.

Introduction. There is an established fact that Asian breast cancer patients are, on average, younger than their European counterparts. This study aimed to utilize the data from the Cancer Incidence in Five Continents I through XIII (published by the International Agency for Research on Cancer) to examine what contributes to the younger age at onset in the Asian population. Material and Methods. Data (number of breast cancer cases and corresponding population figures) for 29 registries in Europe and 9 registries in Asia for the period of 1953–2002 was accessioned and pooled to form two distinct populations, Asia and Europe. The age specific rates were defined and analyzed cross-sectionally (period wise) and longitudinally (cohort wise). The magnitude and the pattern of age specific rates were analyzed using the age-period-cohort analysis. The constrained generalized linear model with a priority assumption of cohort effect as contributing factor to changing rates was used to analyze the data. Result. During the last 50 years, the rate of breast cancer increased for both populations with an estimated annual percent change of 1.03% (with 95% CI of 1.029, 1.031) for Asia and 1.016% (95% CI of 1.015, 1.017) for Europe. There were stronger cohort effects in the magnitude of rates among the Asian population compared to the European population. The cohort effects, expressed as the rate ratio with cohort born in 1970 as reference, ranged from 0.06 (95% CI 0.05, 0.08) to 0.94 (95% CI 0.93, 0.96) for Asians and 0.35 (95% CI 0.33, 0.36) to 1.03 (95% CI 1.02, 1.04) for Europeans. The estimated longitudinal age specific rates (adjusted for cohort and period effects) showed similar patterns between the two populations. Conclusion. It was concluded that a strong cohort effect contributes to the younger age at onset among Asian breast cancer patients.

Background: The association of the mobile phone use and risk of brain tumor remains controversial among radiation epidemiologists. Methods: We hypothesized if an association between brain tumor and mobile phone use exists, this association will be manifested as a cohort effect (as a proxy of association between mobile phone use and brain tumor) in the incidence rates of brain tumor during the period of 1990 to 2009. We used age-period-cohort methodology (generalized log-linear model) and compared the distribution of cohort effects in the observed rates of brain tumor from 1990 to 2009 to the cohort effects from rates driven based on epidemiological study results that reported a positive association between brain tumor and mobile phone use in the Nordic population. Three latency period of 1-4 years with odds ratio (OR) of 1.2, latency period of 5-9 years with OR of 1.3, and latency period of more than 10 years with OR of 2.7 were used to estimate expected rates. Results: The distribution of cohort effects between observed and expected rates were more similar among the males compared with females. A shorter latency was more con¬sistent to observed rates. Conclusion: Our study supports a possible a weak association between mobile phone use and brain tumor; further fueling the controversies in association.

Age and gender effects in 10 physical self-concept scales for elite athletes and nonathletes were based on responses from 4 age cohorts (grades 7-10 in high school) who completed the same instrument 4 times during a 2-year period. A multicohort-multioccasion design provides a stronger basis for assessing development differences than a cross-sectional comparison collected on a single occasion or a longitudinal comparison based on responses by a single age cohort collected on multiple occasions. Across all 10 physical self-concepts there were substantial differences due to group (athletes greater than nonathletes), gender (males greater than females), and gender x group interactions (athletes less than nonathletes in gender differences). There were no significant effects of age cohort and only very small effects of occasions. Thus longitudinal and cross-sectional comparisons both showed that mean levels of physical self-concept were stable over this potentially volatile adolescent period and that this stability generalized over gender, age, and athlete groups.

Au cours des deux derniers siècles, l’espérance de vie tout autour du globe a connu un accroissement considérable. Si la tendance sur le long terme est plutôt régulière, l’amélioration de la longévité peut être décomposée sur le court-terme en plusieurs phases, que l’on peut relier le plus souvent aux progrès médicaux et à la diminution de causes de mortalité particulières. L’année 2020 marque un tournant du fait de l’ampleur de la pandémie Covid-19 et de ses conséquences. Ses effets directs et indirects sur l’économie et les systèmes de santé se manifestent également au travers des autres causes majeures de décès. Pour comprendre et anticiper les risques liés à la mortalité, il devient de plus en plus nécessaire pour les acteurs de la réassurance de raisonner et modéliser en termes de causes de décès. Ce type de modélisation pose néanmoins des défis spécifiques, issues de la nature multivariée des modèles, dont la complexité dépasse celle des outils classiques de l’actuaire. Nous proposons dans cette thèse plusieurs axes pour étendre la modélisation de la mortalité à un cadre multivarié. Ces axes sont abordés sous forme d’articles de recherche. La première étude porte sur des aspects techniques des distributions multivariées au sein de modèles linéaires généralisés. Lorsque les variables explicatives sont catégorielles, nous proposons de nouveaux estimateurs pour les distributions multinomiale, multinomiale négative et de Dirichlet sous forme de formules fermées, qui permettent notamment un gain considérable en temps de calcul. Ces estimateurs sont utilisés dans la seconde étude pour proposer une nouvelle méthode d’estimation des paramètres de modèles de mortalité. Cette méthode prolonge le cadre existant pour la mortalité toute cause, et permet de traiter toutes les problématiques de modélisation de mortalité en une seule étape, en particulier par cause de décès. Le troisième axe porte sur les projections de mortalité. Nous étudions des réseaux de neurones spécifiquement adaptés aux séries temporelles. Nous montrons par des exemples concrets auxquels peut faire face l’actuaire que ces modèles sont suffisamment flexibles et robustes, offrant une alternative crédible aux modèles classiques
Over the past two centuries, life expectancy around the globe has increased considerably. While the long-term trend is fairly regular, the improvement in longevity can be broken down into several phases in the short term, which can most often be linked to medical progress and the reduction in specific causes of mortality. The year 2020 marks a turning point due to the scale of the Covid-19 pandemic and its consequences. Its direct and indirect effects on the economy and healthcare systems will also be felt through other major causes of death. To understand and anticipate mortality-related risks, it is becoming increasingly necessary for reinsurance players to reason and model in terms of causes of death. However, this type of modeling poses specific challenges. By its very nature, it involves multivariate models, whose complexity exceeds that of conventional actuary tools. In this thesis, we propose several avenues for extending mortality modeling to a multivariate framework. These are presented in the form of research articles. The first study deals with technical aspects of multivariate distributions within generalized linear models. When the explanatory variables are categorical, we propose new estimators for the multinomial, negative multinomial and Dirichlet distributions in the form of closed formulas, which notably enable considerable savings in computation time. These estimators are used in the second study to propose a new method for estimating the parameters of mortality models. This method extends the existing framework for all-cause mortality, and enables all mortality modeling issues to be addressed in a single step, particularly by cause-of-death. The third axis concerns mortality forecasts. We study neural networks specifically adapted to time series. Based on concrete use cases, we show that these models are sufficiently flexible and robust to offer a credible alternative to conventional models

Introdução: Os linfomas abrangem um grupo heterogêneo de neoplasias originadas no sistema linfático, diferentes quanto à sua histologia, prognóstico e epidemiologia, embora possa haver grande número de aspectos clínicos comuns. De acordo com sua morfologia, dividem-se dois grupos: os linfomas Hodgkin (LH) e os linfomas não-Hodgkin (LNH). Objetivo: Avaliar a tendência temporal da incidência de linfomas no período de 1997 a 2012, identificando a influência de fatores como o sexo, a idade, o período e a coorte e nascimento. Metodologia: Trata-se de um estudo ecológico. Foram obtidas, do Registro de Câncer de Base Populacional de São Paulo (RCBP-SP), informações sobre todos os casos novos de linfomas no Município de São Paulo, diagnosticados no período de 1997 a 2012. Informações sobre o número de habitantes do Município foram obtidas online através do site do Departamento de Informática do SUS. Para avaliar a tendência da incidência de linfomas no decorrer do período, segundo sexo e faixa etária, foram ajustados modelos lineares generalizados (MLG). Para avaliar a influência da idade, do período de diagnóstico e das coortes de nascimento na tendência da incidência dos linfomas, foi utilizado o modelo idade-período-coorte (IPC). Resultados e conclusões: Dos 18.037 casos analisados, 20,5% eram do tipo LH e 79,5% do tipo LNH. Entre os casos de LH, 52,6% eram homens e 70,0% tinham entre 20 e 39 anos. A taxa de incidência de LH padronizada por idade, por 100 mil habitantes, variou de 5,0 em 1997 para 4,0 em 2012, entretanto, não foi detectada nenhuma tendência significativa na incidência ao longo do tempo (p>0,05). O risco de desenvolver LH foi maior no sexo masculino do que no feminino apenas na faixa etária de 0 a 14 anos (p<0,001), nas demais, o risco foi semelhante para ambos os sexos (p>0,05). O risco de desenvolver LH segundo a faixa etária apresentou um padrão etário bimodal. No sexo feminino, os maiores riscos ocorreram nas faixas de 20 a 39 e de 65 anos e mais e, no masculino, nas faixas de 15 a 19 e de 65 anos e mais. O modelo IPC apontou uma influência da coorte de nascimento na incidência de LH em mulheres: para aquelas nascidas antes de 1960, quanto mais antiga a coorte, maior o risco de LH; para as nascidas após 1960, o risco relativo permaneceu estável. Entre os casos de LNH, 51,6% eram homens e 77,4% tinham mais de 40 anos. A taxa de incidência de LNH padronizada por idade, por 100 mil habitantes, passou de 22,6 em 1997 para 17,0 em 2012. Foi detectada uma tendência de decréscimo na incidência de LNH de 1,7% ao ano em todas as faixas etárias, em ambos os sexos (p<0,001). O risco de desenvolver LNH aumentou continuamente com o avançar da idade, em ambos os sexos. O modelo IPC não detectou efeitos de período. Para os homens, as coortes mais velhas apresentam maior risco e, as mais jovens, menor risco. Para as mulheres nascidas antes de 1960 o comportamento foi semelhante ao dos homens, para as nascidas após 1960, o risco relativo permaneceu estável.
Introduction: Lymphomas comprise a heterogeneous group of neoplasias originating in the lymphatic system, different in their histology, prognosis and epidemiology, although there may be a large number of common clinical aspects. According to their morphology, two groups are divided: Hodgkin\'s lymphomas (HL) and non-Hodgkin\'s lymphomas (NHL). Objective: To evaluate the temporal trend of the incidence of lymphomas in the period from 1997 to 2012, identifying the influence of factors such as sex, age, period and cohort and birth. Methodology: This is an ecological study. Information about all new cases of lymphomas in the city of São Paulo, diagnosed in the period from 1997 to 2012, were obtained from the São Paulo Population Based Cancer Registry. Information on the number of inhabitants of the Municipality were obtained online through the website of the SUS Department of Informatics. To evaluate the trend of lymphoma incidence during the period, according to sex and age group, generalized linear models (GLM) were adjusted. The age-period-cohort (APC) model was used to evaluate the influence of age, diagnosis period and birth cohorts on the trend of lymphoma incidence. Results and conclusions: Of the 18,037 cases analyzed, 20.5% were HL type and 79.5% NHL type. Among the cases of HL, 52.6% were men and 70.0% were between 20 and 39 years old. The age-standardized incidence rate of HL per 100,000 population ranged from 5.0 in 1997 to 4.0 in 2012, however, no significant trend in incidence over time was detected (p> 0.05). The risk of developing HL was greater in males than in females only in the age group 0 to 14 years (p <0.001), in the others, the risk was similar for both sexes (p> 0.05). The risk of developing HL according to the age group presented a bimodal age pattern. In females, the highest risks occurred in the 20-39 and 65-year-olds, and in the male, in the 15-19 and 65 years and older ranges. The APC model pointed to an influence of the birth cohort on the incidence of HL in women: for those born before 1960, the older the cohort, the higher the risk of HL; for those born after 1960, the relative risk remained stable. Among the cases of NHL, 51.6% were men and 77.4% were over 40 years old. The age-standardized incidence rate of NHL per 100,000 population increased from 22.6 in 1997 to 17.0 in 2012. A trend of a decrease in the incidence of NHL of 1.7% per year was observed in all ranges in both sexes (p <0.001). The risk of developing NHL increased steadily with advancing age in both sexes. The APC model did not detect period effects. For men, older cohorts are at higher risk, and younger cohorts are at lower risk. For women born before 1960 the behavior was similar to that of men, for those born after 1960, the relative risk remained stable.

Gerontologists are often skeptical of age as a presumed cause of the aging process. Although age is an indispensable marker of life experiences, it is a rather crude indicator of the many factors that actually shape the aging experience, including senescence. To address the multiple meanings associated with age, some gerontologists have advanced concepts such as biological age or functional age. These are useful concepts, isolating one domain or facet of aging, but even these concepts must be applied with a skepticism for age effects. Gullible gerontology ensues when well-meaning persons accept age as an explanatory variable and disregard or minimize other factors or processes that are associated with age differences. Gerontologists prioritize longitudinal research designs and urge caution when one attempts to generalize from age differences. Concepts such as terminal drop and the age-period-cohort are used to illustrate this axiom.

In the current literature, there are several papers which have considered the modelling and forecasting of population mortality using the Lee-Carter framework. According to Booth (2006), the Lee-Carter-based approach is widely considered because it produces fairly realistic life expectancy forecasts, which are used as reference values for other modelling methods. In recent years, there have been several extensions of the standard LC method, retaining some of its basic properties, but adding additional statistical features too. In 2006, Renshaw and Haberman developed a special adaptation of the LC method. They transformed the basic LC model into a more general framework in order to analyse the relationship between age and time and their joint impact on the mortality rates. This transformation gave birth to the so-called age-period-cohort (APC) log-bilinear generalized linear models (GLM) with Poisson error structures. In this paper, we take into consideration a family of generalised log-linear models of the LC type structure with Poisson errors that includes the basic LC model too. In this framework, we implement a specialised iterative regression methodology based on Poisson likelihood maximization process. In particular, we make use of the approach proposed and illustrated in Renshaw and Haberman (2006), which generalises the basic LC modelling framework to develop a tailored iterative process for updating the parameter estimates. In order to assess the goodness of fit of the regression, we provide a range of residual analyses with corresponding target fitted values. Diagnostic plots are provided to show the results.