Alcohol consumption and mortality. I. Characteristics of drinking groups.

AIMS: This is the first of a set of three papers evaluating drinking status and mortality risk. Analyses of multiple studies describe associations of drinking patterns with characteristics hypothesized to confound the relationships between drinking status and mortality. Characteristics which both significantly differentiate drinking groups and are consistent across studies would suggest that mortality studies not controlling for them may be compromised. DESIGN AND PARTICIPANTS: Associations are evaluated from the raw data of 10 general population studies which contained mortality data. Long-term abstainers are compared to former drinkers, long-term abstainers and former drinkers are compared to light drinkers (by quantity, frequency and volume in separate analyses) and moderate to heavy drinkers are compared to light drinkers. Tetrachoric correlation coefficients assess statistical significance; meta-analysis determines if associations are homogeneous across studies. MEASUREMENTS: Measures of alcohol consumption are quantity, frequency and volume; long-term abstainers are differentiated from former drinkers. Multiple measures of health, social position, social integration and mental health characteristics are evaluated. FINDINGS: Across studies, adult male former drinkers are consistently more likely to be heavier smokers, depressed, unemployed, lower SES and to have used marijuana than long-term abstainers. Adult female former drinkers are consistently more likely to be heavier smokers, in poorer health, not religious, and unmarried than long-term abstainers. Both types of abstainers tend to be of lower SES than light drinkers and report poorer health (not consistent). Female abstainers are more likely to be of normal or overweight than light drinkers. CONCLUSIONS: Characteristics of two groups of abstainers, other than their non-use of alcohol, may confound the associations found between drinking and mortality risk.

Alcohol consumption and mortality. II. Studies of male populations.

AIMS: This is the second of a set of three papers evaluating drinking status and mortality risk. Analysis of eight general population surveys of men evaluated all-cause mortality rates by drinking pattern. DESIGN AND PARTICIPANTS: Raw data from three studies of youth and five studies of adults were evaluated. Logistic regression models controlled for confounding characteristics. Meta-analysis combined study results. MEASUREMENTS: Drinking pattern was alternatively defined by quantity, frequency and volume of drinking. Final models included drinking pattern (as well as abstinence in the youth models and long-term abstainers and former drinkers in adult models), age and other confounding variables. Models also evaluated interactions of age and, respectively, long-term abstinence and former drinking. FINDINGS: No evidence was found for the hypothesis that abstinence is associated with greater mortality risk than light drinking. In the youth samples, abstainers had a lower risk of dying than those drinking less than 15 times per month. One study of the adult samples showed a significant age by former drinker interaction; this did not alter the lack of association of former drinking with mortality risk or the homogeneity of results across studies for this finding. The most consistent finding was the association of heavy drinking with mortality among youth. Among adults, drinking 43 or more drinks per month and drinking 21 or more times per month were associated with increased mortality risk. Quantity per occasion was not significantly associated with mortality risk among adults. CONCLUSIONS: That frequent drinking was related to mortality risk, whereas heavier quantity was unrelated, is inconsistent with the belief that daily consumption of a few glasses of wine has salutary effects. Empirically, however, this pattern tends to be unusual. Findings were homogeneous across studies lending generalizability to results.

Alcohol consumption and mortality. III. Studies of female populations.

AIMS: This is the third of a set of three papers evaluating drinking status and mortality risk. Analysis of three general population surveys of women evaluated all-cause mortality rates by drinking pattern. DESIGN AND PARTICIPANTS: Raw data from three studies of adult women were evaluated. Logistic regression models controlled for confounding characteristics. Meta-analysis combined study results. MEASUREMENTS: Drinking pattern was alternatively defined by quantity, frequency and volume of drinking. Final models included drinking pattern (including long-term abstainers and former drinkers) as well as age and other confounding variables. Models also evaluated interactions of age and, respectively, long-term abstinence and former drinking. FINDINGS: In models in which age was controlled, odds of death for long-term abstainers and former drinkers (defined by volume or quantity) were greater than those for light drinkers; odds of death for moderate and heavy drinkers (defined by quantity) were greater than those for light drinkers. When other psychosocial attributes were controlled, odds of death were similar for abstainers and light drinkers. When other psychosocial attributes were controlled, odds of death for heavy drinkers (defined by volume and quantity) were greater than those for light drinkers. When interactions of age and the two forms of abstinence were introduced, one study showed a significant effect of age and former drinking. CONCLUSIONS: Results were consistent with the hypothesis that characteristics of abstainers other than their non-use of alcohol may account for their higher mortality risk. With the exception of former drinkers compared to light drinkers, when interactions were introduced into models (for measures of quantity and frequency) findings were homogeneous across studies, lending generalizability to results.

Determinants of life-course variation in the frequency of alcohol consumption: meta-analysis of studies from the collaborative alcohol-related longitudinal project.

OBJECTIVE: This report evaluates the relative contribution of predictors of change in the frequency of alcohol consumption among drinkers, based on the quantitative synthesis of data from 27 longitudinal studies of the general population. The analysis has two objectives: (1) to evaluate the impact of selected demographic characteristics on the magnitude and trajectory of change in drinking across multiple samples, and (2) to assess the influence of methodological characteristics on the consistency of results across studies. METHOD: Raw data from studies including two serial measures of the frequency of alcohol consumption are analyzed. Fixed, random and mixed effects models for meta-analysis are used to pool measures across observations and model the influence of predictors on variability between results. RESULTS: Gender-based variation in the patterning of change is present across all observations, but concentrated in early periods of the life course. Age displays significant predictive effects across all observations, but statistically uniform results are obtained for subjects aged 30 and over. The national origin of study predicts larger amounts of variation than do other demographic predictors in the models. Significant effects are observed for several methodological characteristics of studies. Variation among effect estimates is associated with differences between samples in the interval between first and final measurements, the date of first measurement (a proxy for the historical context of the sample), the percentage retention of subjects between measurements and the time frame of the original alcohol measure. CONCLUSIONS: Based on the synthesis of data from multiple longitudinal samples, this study (1) characterizes normative developmental patterning in the frequency of alcohol consumption and demonstrates the varying effects of demographic factors across the life-course: (2) indicates the key influence of cultural and historical context on the establishment of drinking patterns; and (3) confirms the impact of methodological differences on variation in the results of studies.

Aggregate-level predictors of the prevalence of selected drinking patterns in multiple studies: a research synthesis from the Collaborative Alcohol-Related Longitudinal Project.

This paper examines the prevalence of two "at-risk" alcohol drinking patterns (infrequent heavy drinking and frequent heavy drinking) within age/gender groups in multiple general population studies. When heterogeneity in findings across studies is found, we test the hypotheses that suicide, divorce, unemployment rates, and the per capita consumption of alcohol in each country are associated with the prevalence of these drinking patterns. These analyses should inform the literature on the relationships between societal factors and the prevalence of persons in different societies and periods in history that drink at these levels.

Gender differences for the risk of alcohol-related problems in multiple national contexts.

The primary research question asked is: After holding alcohol consumption constant, will men and women be at equal risk for a variety of alcohol-related problems? Since women are actually at a higher blood alcohol content at the same consumption levels, a physiological argument would suggest that women are at equal or greater risk for alcohol problems than men. However, variation in societal norms surrounding gender roles and/or societal-level stress may mediate the experience of men and women, regardless of the differences in physiology. Ten cross-sectional general population studies are used. Analyses control for individual-level variables (age, quantity, and frequency of drinking) and societal-level variables (proportion of women in the work force and female suicide rate) that might confound these relationships; cross-study homogeneity is examined.

Relationships of measures of alcohol consumption with alcohol-related problems in multiple studies: a research synthesis from the collaborative alcohol-related longitudinal project.

Two measures of alcohol consumption were used to predict groups of alcohol problems in 21 general population studies from 11 countries to determine (a) if quantity of drinking per occasion or frequency of drinking per month constituted significant "risk" for alcohol problems, having controlled for each as well as individual-level and aggregate-level variables which might confound these relationships and (b) if these associations were homogeneous across studies. A two-tiered analysis assessed these relationships within each study by modeling age, sex, quantity per occasion and frequency per month as predictors of alcohol problems. Meta-analysis combined test statistics to determine if they were homogeneous across studies. The meta-analysis was repeated, blocking for per capita consumption of alcohol (a trait of nations thought to measure drinking norms) and the female rate of suicide (a trait of nations thought to measure societal-level stress). When only individual-level variables were controlled (age and sex), both quantity and frequency were risk factors for each drinking problem. However, except in the case of the association of quantity with alcohol treatment, the magnitude of these risks were heterogeneous across studies. When blocking for the societal-level traits, each had more relevance for some, but not all, of the relationships between consumption and problems. Particularly striking was the well-documented finding that per capita consumption of alcohol significantly distinguished the relationships of frequency of drinking and health problems (while the female suicide rate did not) and the previously undocumented finding that the female suicide rate significantly distinguished the relationships of both quantity and frequency with treatment (while the per capita consumption of alcohol did not). These findings suggest that the impact of norms and the impact of societal stress in groups have different but significant consequences for the relationships of consumption to problems.

Societal-level predictors of groups' drinking patterns: a research synthesis from the Collaborative Alcohol-Related Longitudinal Project.

OBJECTIVES: Period effects of per capita consumption of alcohol and divorce rates are assessed for change in quantity and frequency among age/sex groups in multiple longitudinal studies. METHODS: Twenty-five studies of quantity and 29 studies of frequency are used. Studies are from 15 nations and cover periods of 1 to 21 years. Models predict the standardized mean difference for quantity and frequency based on period effects and group-level and methodological variables. RESULTS: When both the period effects of per capita consumption and the divorce rate are considered, the divorce rate significantly predicts change in quantity and frequency. An increase in the divorce rate is associated with a stronger decrease in frequency among younger people; men are more likely than women to decrease their frequency of drinking when divorce rates rise. CONCLUSIONS: Multiple societal-level factors should be considered critical in influencing the drinking patterns of groups. These results suggest that an increase in the divorce rate is associated with more "dry" social contexts, characterized possibly by drinking patterns of a more "volitive" nature (i.e., heavier quantity per occasion and less frequent drinking).

A cross-study contextual analysis of effects from individual-level drinking and group-level drinking factors: a meta-analysis of multiple longitudinal studies from the collaborative alcohol-related longitudinal project.

In contextual (cross-level) analysis within multiple longitudinal general population studies, individual-level drinking behaviors (quantity per typical occasion, frequency of drinking per month and total volume of drinks per month) at final measurement are assessed by three models that simultaneously enter individual- and group-level measures. Two age groups (15-20 and 21-30) are independently assessed. In each model, the Time 1 individual-level drinking behavior and one of three group-level factors are entered. The group-level factors are (1) the percentage of abstainers at Time 1 for each age/sex cohort, (2) the Time 1 group mean for the drinking measure for the age/sex cohort and (3) the mean difference of the age/sex cohort's change in the drinking measure over time. All variables in the model are controlled by variations to exposure in per capita consumption of alcohol during the age/sex cohort's formative years and at Time 2. Meta-analysis assesses the homogeneity of the findings across studies. Models were proposed with the rationale that (1) understanding of individual drinking behavior can be advanced if individual-level data and group-level data are considered in the same models, and (2) integration of these two levels of analyses are, to date, rare. The rationale for using meta-analysis is that findings from the models can be assessed across social contexts with respect to their generalizability. The mean difference model, controlling for individual drinking at Time 1, is the most influential of the group-level models for the younger age group: the degree to which the group changes its drinking pattern is positively related to individual-level drinking behavior at final measurement, over and above the individual's drinking behavior at Time 1, for individual-level frequency of drinking among males (homogeneous among drinkers only). Younger females show more significant relationships for the mean difference females show more significant relationships for the mean difference model. Findings are significant for all relationships examined for the mean difference of the drinking of the group and the individual drinking among the older males and females. Measures of individual-level drinking for all measures at Time 1, controlling for the group-level effects, are significantly related to individual-level drinking at final measurement. The results are homogeneous for quantity (drinkers only) and volume among the young. Findings indicate that characterizations of the drinking for both the individual and the group to which the individual belongs predict measures of drinking practices on the individual level over time.