Conditional Poisson Regression with Random Effects for the Analysis of Multi-site Time Series Studies.

The analysis of time series studies linking daily counts of a health indicator with environmental variables (e.g., mortality or hospital admissions with air pollution concentrations or temperature; or motor vehicle crashes with temperature) is usually conducted with Poisson regression models controlling for long-term and seasonal trends using temporal strata. When the study includes multiple zones, analysts usually apply a two-stage approach: first, each zone is analyzed separately, and the resulting zone-specific estimates are then combined using meta-analysis. This approach allows zone-specific control for trends. A one-stage approach uses spatio-temporal strata and could be seen as a particular case of the case-time series framework recently proposed. However, the number of strata can escalate very rapidly in a long time series with many zones. A computationally efficient alternative is to fit a conditional Poisson regression model, avoiding the estimation of the nuisance strata. To allow for zone-specific effects, we propose a conditional Poisson regression model with a random slope, although available frequentist software does not implement this model. Here, we implement our approach in the Bayesian paradigm, which also facilitates the inclusion of spatial patterns in the effect of interest. We also provide a possible extension to deal with overdispersed data. We first introduce the equations of the framework and then illustrate their application to data from a previously published study on the effects of temperature on the risk of motor vehicle crashes. We provide R code and a semi-synthetic dataset to reproduce all analyses presented.

Alcohol as a trigger of migraine attacks in people with migraine. Results from a large prospective cohort study in English-speaking countries.

OBJECTIVE: To assess whether alcohol intake is associated with the onset of migraine attacks up to 2 days after consumption in individuals with episodic migraine (EM). BACKGROUND: Although alcohol has long been suspected to be a common migraine trigger, studies have been inconclusive in proving this association. METHODS: This was an observational prospective cohort study among individuals with migraine who registered to use a digital health platform for headache. Eligible individuals were aged ≥18 years with EM who consumed alcohol and had tracked their headache symptoms and alcohol intake for ≥90 days. People who did not drink any alcohol were excluded. The association of alcohol intake ("Yes/No") and of the number of alcoholic beverages in the 2 days preceding a migraine attack was assessed accounting for the presence of migraine on day-2 and its interaction with alcohol intake on day-2, and further adjusted for sex, age, and average weekly alcohol intake. RESULTS: Data on 487 individuals reporting 5913 migraine attacks and a total of 40,165 diary days were included in the analysis. Presence of migraine on day-2 and its interaction with alcohol intake on day-2 were not significant and removed from the model. At the population level, alcohol intake on day-2 was associated with a lower probability of migraine attack (OR [95% CI] = 0.75 [0.68, 0.82]; event rate 1006/4679, 21.5%), while the effect of alcohol intake on day-1 was not significant (OR [95% CI] = 1.01 [0.91, 1.11]; event rate 1163/4679, 24.9%) after adjusting for sex, age, and average weekly alcohol intake. Similar results were obtained with the number of beverages as exposure. CONCLUSIONS: In this English-speaking cohort of individuals with EM who identified themselves as mostly low-dose alcohol consumers, there was no significant effect on the probability of a migraine attack in the 24 h following consumption, and a slightly lower likelihood of a migraine attack from 24 to 48 h following use.

Bayesian model selection for the study of Hardy-Weinberg proportions and homogeneity of gender allele frequencies.

Standard statistical tests for Hardy-Weinberg equilibrium assume the equality of allele frequencies in the sexes, whereas tests for the equality of allele frequencies in the sexes assume Hardy-Weinberg equilibrium. This produces a circularity in the testing of genetic variants, which has recently been resolved with new frequentist likelihood and exact procedures. In this paper, we tackle the same problem by posing it as a Bayesian model comparison problem. We formulate an exhaustive set of ten alternative scenarios for biallelic genetic variants. Using Dirichlet and Beta priors for genotype and allele frequencies, we derive marginal likelihoods for all scenarios, and select the most likely scenario using the posterior probabilities that each of these scenarios is the one in place. Different from the usual frequentist testing approach, the Bayesian approach allows one to compare any number of models, and not just two at a time, and the models compared do not have to be nested. We illustrate our Bayesian approach with genetic data from the 1,000 genomes project and through a simulation study.

An integrated analysis of individual and aggregated health data using estimating equations.

Analyses of individual disease-exposure data within a population are useful when exposure of interest varies sufficiently within the population. When the within-population variance of exposure is limited, however, power of the individual-data analysis is reduced. In such situations, aggregated-data analyses of disease data across populations, with a sample of individual exposure data from each population, can be powerful in estimating the exposure effect if between population variation of exposure is large. In this paper, we consider a new analytical framework that is a combination of the individual- and aggregated-data analyses, based on an estimating equation approach. The proposed analysis utilizes strengths from individual data and aggregated data in the estimation of the exposure effect of interest, depending on which of the exposure variations (within- versus between-population) dominates. Simulation studies under various different scenarios were performed to show the strengths of the proposed approach in the estimation of the exposure effects of interest.

[Mortality of gastric cancer in Catalonia, Spain: geographical distribution and time trends from 1986 to 2000]. / Mortalidad por cáncer de estómago en Cataluña: distribución geográfica y evolución temporal entre 1986 y 2000.

BACKGROUND AND OBJECTIVE: The aims of this study are to describe the time trends and the changes in the spatial distribution of stomach cancer mortality by gender, in Catalonia, Spain, in the period 1986-2000. MATERIAL AND METHOD: The mortality data comes from the Mortality Register for Catalonia at the Health Department and the population data from the Institute of Statistics for Catalonia. To analyze time trends, a Poisson regression model was adjusted for each gender. To analyze the geographical distribution, a Bayesian hierarchical model was used. RESULTS: During the period 1986-2000 the number of deaths from stomach cancer was 8,627 for males and 5,831 for females. During this period the estimated decrease in mortality was 3.13% for males and 3.91% for females. The spatial analysis showed the lowest mortality risk areas along the coast while the mortality risk increased toward the zones in the interior. This geographical pattern is very similar for both sexes but in the lasts years of the period it has been fading. CONCLUSIONS: The time trends and the geographical pattern of stomach cancer mortality in Catalonia is similar for both sexes and it is consistent with the trends observed in other developed countries. This suggests a relationship with improved food habits and a better accessibility to health care in the areas of higher risk.

Mortalidad por cáncer de estómago en Cataluña: distribución geográfica y evolución temporal entre 1986 y 2000 / Mortality of gastric cancer in Catalonia, Spain: geographical distribution and time trends from 1986 to 2000

Fundamento y objetivo: Este estudio analiza la evolución temporal y los cambios en la distribución espacial, por comarcas, de la mortalidad por cáncer de estómago en Cataluña, para cada sexo, durante el período 1986-2000. Material y método: Las defunciones por cáncer de estómago de los años 1986-2000 proceden del Registro de Mortalidad de Cataluña del Departament de Salut y la población del Institut d'Estadística de Cataluña. Para el análisis de la evolución temporal se ajustó un modelo de regresión de Poisson para cada sexo. Para el análisis espacial en cada comarca se ha utilizado un modelo jerárquico bayesiano. Resultados: En los años 1986-2000 se produjeron 8.627 defunciones por cáncer de estómago en varones y 5.831 en mujeres. Durante este período se ha estimado un descenso de la mortalidad del 3,13% en los varones y del 3,91% en las mujeres. El análisis espacial muestra que las zonas con menor riesgo de mortalidad son las del litoral y este riesgo aumenta de forma gradual hacia las zonas del interior. Este patrón geográfico es muy similar para los 2 sexos y se ha ido difuminando con el paso del tiempo. Conclusiones: La evolución temporal y la distribución geográfica de la mortalidad por cáncer de estómago en Cataluña son similares en ambos sexos y coherentes con las tendencias observadas en otros países desarrollados. Este patrón indica una mejora en los hábitos alimentarios y una mejor accesibilidad a los servicios sanitarios de las zonas con mayor riesgo

Background and objective: The aims of this study are to describe the time trends and the changes in the spatial distribution of stomach cancer mortality by gender, in Catalonia, Spain, in the period 1986-2000. Material and method: The mortality data comes from the Mortality Register for Catalonia at the Health Department and the population data from the Institute of Statistics for Catalonia. To analyze time trends, a Poisson regression model was adjusted for each gender. To analyze the geographical distribution, a Bayesian hierarchical model was used. Results: During the period 1986-2000 the number of deaths from stomach cancer was 8,627 for males and 5,831 for females. During this period the estimated decrease in mortality was 3.13% for males and 3.91% for females. The spatial analysis showed the lowest mortality risk areas along the coast while the mortality risk increased toward the zones in the interior. This geographical pattern is very similar for both sexes but in the lasts years of the period it has been fading. Conclusions: The time trends and the geographical pattern of stomach cancer mortality in Catalonia is similar for both sexes and it is consistent with the trends observed in other developed countries. This suggests a relationship with improved food habits and a better accessibility to health care in the areas of higher risk

Análisis de la evolución temporal de la mortalidad mediante modelos lineales generalizados / Analysis of mortality time trend using generalized linear models

En este trabajo se muestra cómo los modelos lineales generalizados permiten describir eficientemente diferentes patrones de evolución temporal de datos de mortalidad y, a su vez, llevar a cabo una fácil interpretación. Como aplicación práctica se analiza la evolución de la mortalidad por cáncer de mama en las mujeres de Cataluña entre 1986 y 2000. De los resultados destaca que la mortalidad por cáncer de mama experimenta un aumento y un posterior descenso para todos los grupos de edad. El año en que se inicia el descenso es más reciente en los grupos de edad mayor

In this work it is shown how generalized linear models allow one to describe different patterns of temporary evolution of mortality data, while at the same time allow for an easy interpretation. As a practical application, the evolution of the female breast cancer mortality in Catalonia from 1986 to 2000 is analyzed. Remarkably, the mortality from breast cancer first increases and then decreases for all age groups. Moreover, the year in which the cancer rate starts decreasing is more recent in the older age groups

[Analysis of mortality time trend using generalized linear models]. / Análisis de la evolución temporal de la mortalidad mediante modelos lineales generalizados.

In this work it is shown how generalized linear models allow one to describe different patterns of temporary evolution of mortality data, while at the same time allow for an easy interpretation. As a practical application, the evolution of the female breast cancer mortality in Catalonia from 1986 to 2000 is analyzed. Remarkably, the mortality from breast cancer first increases and then decreases for all age groups. Moreover, the year in which the cancer rate starts decreasing is more recent in the older age groups.