In the Journals

Alcohol-mortality associations underestimated in cohort studies

Researchers report that the connections between alcohol and mortality established in cohort studies may underestimate negative health consequences of moderate drinking.

The study, published in the Journal of Studies on Alcohol and Drugs, showed that many cohort studies exploring this connection enroll people aged 50 years and older, eliminating those who have died prematurely due to alcohol consumption before age 50.

“Findings about alcohol-mortality relationships are rarely contextualized with respect to the age of study participants, and in fact information about the mean age or mean weighted age (in the case of meta-analyses) of study participants is rarely reported,” Timothy S. Naimi, MD, MPH, of the clinical addiction research and education unit at Boston Medical Center, and colleagues wrote. “This makes it difficult to assess the potential magnitude of selection bias from premature mortality.”

Using 2006 to 2010 data from the CDC’s Alcohol-Related Disease Impact software app, researchers compared the distribution of mortality-related benefits and harms from alcohol use among adults aged 20 years and older with the distribution among those aged 50 years and older.

They evaluated the distribution of alcohol-attributable deaths and the years of potential life lost that was attributable to or prevented by alcohol for 54 medical conditions in the U.S. by 15-year age groupings (aged 20 to 34 years, 35 to 49 years, 50 to 64 years, and older than 65 years). In addition, they estimated the proportion of net deaths and years of potential life lost for each age group, overall and by cause of death.

From 2006 to 2010 in the United States, adults aged 20 to 49 years experienced 35.8% of the deaths and 58.4% of the years of potential life lost attributable to alcohol, while this age group amassed only 4.5% of alcohol-attributable deaths and 14.2% of years of potential life lost gained, according to the results. In contrast, Naimi and colleagues found that the group aged 65 years and older accounted for about 35% of alcohol-attributable deaths caused by alcohol and about 80% of alcohol-attributable deaths prevented by alcohol.

Across all medical condition categories, 46.3% of the total net deaths and 64.7% of the net years of potential life lost occurred among adults aged 20 to 49 years. When the researchers combined the net deaths occurring among those aged 20 to 49 years and those occurring after age 50 years, they found an 86.3% relative increase in net deaths.

“Premature mortality is a type of selection bias, because deceased persons cannot be enrolled in cohort studies,” Naimi and colleagues wrote. “It is tempting to say that this issue could be addressed by carefully contextualizing findings from cohort studies on the basis of age. However, findings from studies of established drinkers might not be the same as effects among older persons who might begin to drink, or purposefully continue to drink, in order to prolong life.” – by Savannah Demko

Disclosure: The authors report no relevant financial disclosures.

Researchers report that the connections between alcohol and mortality established in cohort studies may underestimate negative health consequences of moderate drinking.

The study, published in the Journal of Studies on Alcohol and Drugs, showed that many cohort studies exploring this connection enroll people aged 50 years and older, eliminating those who have died prematurely due to alcohol consumption before age 50.

“Findings about alcohol-mortality relationships are rarely contextualized with respect to the age of study participants, and in fact information about the mean age or mean weighted age (in the case of meta-analyses) of study participants is rarely reported,” Timothy S. Naimi, MD, MPH, of the clinical addiction research and education unit at Boston Medical Center, and colleagues wrote. “This makes it difficult to assess the potential magnitude of selection bias from premature mortality.”

Using 2006 to 2010 data from the CDC’s Alcohol-Related Disease Impact software app, researchers compared the distribution of mortality-related benefits and harms from alcohol use among adults aged 20 years and older with the distribution among those aged 50 years and older.

They evaluated the distribution of alcohol-attributable deaths and the years of potential life lost that was attributable to or prevented by alcohol for 54 medical conditions in the U.S. by 15-year age groupings (aged 20 to 34 years, 35 to 49 years, 50 to 64 years, and older than 65 years). In addition, they estimated the proportion of net deaths and years of potential life lost for each age group, overall and by cause of death.

From 2006 to 2010 in the United States, adults aged 20 to 49 years experienced 35.8% of the deaths and 58.4% of the years of potential life lost attributable to alcohol, while this age group amassed only 4.5% of alcohol-attributable deaths and 14.2% of years of potential life lost gained, according to the results. In contrast, Naimi and colleagues found that the group aged 65 years and older accounted for about 35% of alcohol-attributable deaths caused by alcohol and about 80% of alcohol-attributable deaths prevented by alcohol.

Across all medical condition categories, 46.3% of the total net deaths and 64.7% of the net years of potential life lost occurred among adults aged 20 to 49 years. When the researchers combined the net deaths occurring among those aged 20 to 49 years and those occurring after age 50 years, they found an 86.3% relative increase in net deaths.

“Premature mortality is a type of selection bias, because deceased persons cannot be enrolled in cohort studies,” Naimi and colleagues wrote. “It is tempting to say that this issue could be addressed by carefully contextualizing findings from cohort studies on the basis of age. However, findings from studies of established drinkers might not be the same as effects among older persons who might begin to drink, or purposefully continue to drink, in order to prolong life.” – by Savannah Demko

Disclosure: The authors report no relevant financial disclosures.