Deaths Associated with High-Volume Drinking of Alcohol among Adults in Canada in 2002: A Need for Primary Care Intervention?
Patra, Jayadeep, Taylor, Benjamin, Rehm, Jürgen, Contemporary Drug Problems
Alcohol is causally related to many different causes of death, including different cancers, cardiovascular diseases, mental health conditions, digestive diseases and injury categories (English et al. 1995; Gutjahr et al. 2001; Rehm et al. 2003a; Ridolfo & Stevenson 2001; Single et al. 2000). The World Health Organization Comparative Quantification of Health Risks estimated that death caused by alcohol accounted for 3.2% of total global mortality in 2000 (WHO 2002), increasing from the previous estimate of 1.5% for 1990 (Murray & Lopez 1996). Approximately half of these deaths were due to acute causes, primarily from unintentional injury categories such as motor vehicle accidents, poisonings and falls. The remaining portion was due to chronic conditions, with cancer and cardiovascular diseases contributing the majority of deaths in this category (Rehm et al. 2004; WHO 2002).
Average daily volume of alcohol consumption is particularly important for studies of mortality (English et al. 1995; Gutjahr et al. 2001; Ridolfo & Stevenson 2001; Single et al. 2000). Previous work by English et al. (1995) has found high-volume drinking (20+ grams per day for women, 40+ grams per day for men) to be causally related to over 40 acute and chronic causes of death. Significant also is that the beneficial effects of alcohol, such as the protective effect of moderate drinking reported for ischemic heart disease, ischemic stroke, gallstones and diabetes, are much less pronounced in high-volume drinkers.
Despite identification of this higher-risk group, however, the alcohol literature to date has mostly reported the effects of total consumption rather than parsing out subpopulations of drinkers and analyzing mortality attributable to different average consumption scenarios separately. In Canada, a significant portion of the population makes up this group of drinkers, although the impact of their daily consumption on mortality has not been reported. Rehm and colleagues (Rehm et al. 2006a), using data from the Canadian Addiction Survey, reported that 12.8% of men and 8.2% of women fell into this group, with a higher prevalence reported in younger age groups.
This study aims to fill a gap in the literature by quantifying the impact of high-volume drinking on mortality in Canada for the year 2002. As well, we discuss the meaning and interpretation of these results with an emphasis on prevention and early detection methods. It is a follow-up study to an analysis of the mortality attributable to moderate drinking in Canada for the same time period (Rehm et al. 2007a)
The aim of the present study was to estimate the proportion of deaths attributable to high-volume drinking in Canada for the year 2002. Several elements necessary for this estimate are described below: methods for identification of diseases and injuries in part or wholly caused by alcohol, determination of risk relations, estimations of prevalence of exposure in Canada, the method used to combine risk relations and exposure to calculate the attributable fractions due to high-volume drinking and the source of the mortality data, to which these fractions were applied.
Identification of diseases and metaanalyses
To identify the chronic and nonchronic disease conditions attributable to alcohol, this analysis was guided by the following criteria in judgments of causality: consistency, strength of association, temporality, dose-response, experimental evidence and underlying biological pathway (Hill 1965; Rehm et al. 2003a; Rothman & Greenland 1998). Once identified, the conditions were translated into corresponding codes from the tenth revision of the International Classification of Diseases (ICD-10). Finally, a comprehensive search strategy of current meta-analyses was performed for each disease category and its risk relationship with alcohol consumption.
The most comprehensive meta-analyses for different conditions were identified from the literature (Rehm et al. …