Methodological Approaches to Conducting Pooled Cross-Sectional Time Series Analysis: The Example of the Association between All-Cause Mortality and per capita Alcohol Consumption for Men in 15 European States

Abstract

Aim: To compare different statistical models in order to estimate the association of alcohol consumption and total mortality when time series data stem from different regions. Data and Methods: Data on per capita consumption in 15 European countries were combined with standardized mortality rates covering different periods between 1950 and 1995. An indicator of region-specific drinking patterns was measured without reference to a concrete time point, thus generating a hierarchical data structure. Two groups of models were compared: pooled cross-sectional time series models with different error structures and hierarchical linear models (random coefficient models). Results: If historical time is not controlled for in cross-sectional models, this might result in estimating a negative association between alcohol consumption and total mortality. Hierarchical linear models or cross-sectional models controlling for historical time, however, resulted in the expected positive association. Only hierarchical linear models were able to adequately estimate the moderating effect of drinking patterns on the association between alcohol consumption and total mortality. Conclusion: For pooled cross-sectional time series data, control for the potential impact of historical time is of utmost importance. Hierarchical linear models constitute a superior alternative to analyze such complex data sets, especially as time-independent characteristics of regions can be implemented in the model.

Tea Consumption, Alcohol Drinking and Physical Activity Associations with Breast Cancer Risk among Chinese Females: a Systematic Review and Meta-analysis

“…For example, 10398G allele in the mitochondrial genome was reported to influence the alcohol metabolism, which may also modify the association between alcohol drinking and breast cancer (Pezzotti et al, 2009). Additionally, consistent J-shaped curve of alcohol drinking on the risk of diseases was found in many cardiovascular diseases (White et al, 1999;Gmel et al, 2001). Whether the J-shaped curve also existed in the incidence of breast cancer, it really deserved further studies.…”

Steps towards Constructing a Global Comparative Risk Analysis for Alcohol Consumption: Determining Indicators and Empirical Weights for Patterns of Drinking, Deciding about Theoretical Minimum, and Dealing with Different Consequences

“…From a substantive point of view, patterns of drinking were found to influence the overall mortality attributable to alcohol to a considerable degree, especially in young people in Europe [10]. This is not surprising, as young people often die from acute causes of death, especially accidents.…”

“…Calendar year was used to control for omitted variable bias and the time structure [10]; per capita gross national product was included to control for poverty as a potential confounder. The data were taken from the following sources.…”

“…Random intercept and random coefficient models were analyzed to determine the influence of patterns of drinking on consumption. Greater detail on the equations used can be found in Rehm and Gmel [9]; a detailed comparison of alternative methods and their results is given in Gmel et al [10]. Separate analyses were conducted for males and females.…”

“…One way of integrating average volume and patterns into the risk relationship that also takes into account the availability of data across most countries is to model the relationship between volume, patterns and mortality with hierarchical models [9,10]. In these models, patterns of drinking are seen as moderating the influence of volume on disease.…”