|
Article Excerpt
EPIDEMIOLOGICAL STUDIES SHOW a reduced risk of coronary heart disease (CHD) and all-cause mortality for individuals who consume alcohol in moderate amounts compared with individuals who consume very little alcohol and abstainers (Camargo et al., 1997; Corrao et al., 2000; Keil et al., 1997; Shaper et al., 1988). However, these results have been challenged as many studies on the CHD consequences of alcohol consumption suffer from epidemiological weaknesses, such as heterogeneity of the investigated exposed group or residual confounding. In a previous analysis of the KORA/MAGiC-Control study (KMC), we observed an association of various social, demographic, and behavioral factors as well as health conditions, with light alcohol consumption and abstinence (Konig et al., 2005). The fact that CHD risk factors are likely to confound studies that are designed to investigate the direct health effects of alcohol has also been shown impressively by data from a large population-based U.S. telephone survey. Of the 30 investigated cardiovascular disease-associated factors, 90% were significantly more prevalent among nondrinkers (assessed during the last 30 days) than among moderate drinkers (Naimi et al., 2005).
Nondrinkers, assessed during a short period, are a heterogeneous group comprising occasional drinkers, nonlifetime abstainers (NLAs), and lifetime abstainers (LAs) (Konig et al., 2005; Wannamethee and Shaper, 1997). Cryer et al. (2001) analyzed data from a British health and lifestyle survey. Using a cluster analysis of sociodemographic and health status variables, both groups of 1,120 male and 2,683 female teetotalers were each divided into an optimum number of two alcohol-abstainer clusters. For each gender, one cluster reflected an LA group, containing, on average, younger and healthier individuals, who were also more likely to be working, whereas the other cluster reflected a "sick quitter" group, containing, on average, older participants with poorer health status, more disabled persons, and a greater proportion of exsmokers (Cryer et al., 2001).
For epidemiological studies, it is crucial to differentiate and to appropriately define subgroups with low alcohol consumption. Assuming that the group of alcohol abstainers contains a number of former alcohol consumers who gave up drinking because of ill health or because alcohol interacts with prescription drugs they are taking (the sick quitter hypothesis), one could expect a higher rate of morbidity or mortality within the class of abstainers (Cryer et al., 2001; Gmel et al., 2003; Shaper et al., 1988; Wannamethee and Shaper, 1988). Several epidemiological studies have yielded different risk estimates for various diseases and mortality (Doll et al., 2005; Wannamethee and Shaper, 1997). A large meta-analysis detected statistically significant higher mortality risk estimates for exconsumers compared with LAs in both men and women (Gmel et al., 2003). In a prospective study of the Augsburg population (median follow-up time of 10 years), we collected data on alcohol consumption twice during a span of 3 years, allowing differentiation of study participants in alcohol consumers, quitters, and nonconsumers. The hazard rate ratios for CHD and overall mortality were higher for nondrinkers compared with quitters and current drinkers (Wellmann et al., 2004). Although this may not provide the full explanation of a higher CHD risk of abstainers, heterogeneity of the exposure group, inclusion of sick quitters, and residual confounding may severely affect study results (Klatsky et al., 1990; Konig et al., 2005; Naimi et al., 2005).
This report describes a population-representative sample of persons with very low alcohol consumption in southern Germany with regard to the presence of alcohol abstinence, starting point of abstinence, and duration of abstinence as well as reasons for becoming abstinent. Moreover, the reliability of self-reported alcohol abstinence was investigated. The goals of this study...
|
