As little as one alcoholic drink per day may be enough to increase blood pressure over time, with an increase in alcohol consumption leading to greater corresponding increases in blood pressure, according to a recent meta-analysis of international health data.
The dose-response effect indicated by the analysis was found even with low levels of alcohol consumption, in a linear relationship that surprised the researchers, says Marco Vinceti, an adjunct associate professor of epidemiology at BUSPH and a professor of epidemiology and public health at the University of Modena and Reggio Emilia Medical School, Modena, Italy.
“Given the absence of previous dose-response meta-analyses on the relation between alcohol consumption and blood pressure, we were substantially uncertain about the potential results of our endeavor when we started it,” Vinceti says. The study was published in the American Heart Association journal Hypertension and has received substantial media coverage around the world.
Vinceti says the research team expected either no effect or just a slight beneficial effect at low doses of alcohol intake, consistent with what has been suggested in older systematic reviews and meta-analyses of alcohol consumption and incidence of—or mortality from—hypertension, stroke, ischemic heart disease, and overall cardiovascular disease. Given the new results of the new analysis, Vinceti says, “The overall public health message with reference to alcohol intake and blood pressure appears to be ‘the lower the better,’ suggesting either to limit or to avoid consumption.”
The team conducted a meta-analysis of seven longitudinal studies from the U.S., South Korea, and Japan with data published between 1997 and 2021. The 19,548 participants ranged in age from 20 to early 70s when the studies were initiated; studies included 12,701 men and 6,650 women. Four studies included smoking status as a potential confounder, and five included body mass index.
Vinceti’s team selected studies with follow up testing durations ranging from 4 to 12 years, with a median value of 5.3 years. None of the participants had previously been diagnosed with high blood pressure or other cardiovascular diseases, diabetes, liver disease, alcohol use disorder, or binge drinking.
Each study recorded participants’ alcohol intake via a variety of questionnaires about dietary intake, or more detailed interviews about smoking and alcohol consumption. Researchers used the reported alcohol consumption to calculate the amount of grams of alcohol consumed daily at each phase of the studies, and were able to compare blood pressure readings between adults who regularly drank alcohol with those of non-drinkers.
Grams of alcohol are used as a relatively standardized measurement of consumption to account for different types of drinks that contain varying levels of alcohol. In the U.S., 12 ounces of regular beer, 5 ounces of wine or a 1.5-ounce shot of distilled spirits each contain about 14 grams of alcohol.
Systolic blood pressure, the top number in a pressure reading, rose 1.25 millimeters of mercury (mm Hg) in people who consumed an average of 12 grams of alcohol per day, about one drink, rising to 4.9 mm Hg in people consuming an average of 48 grams of alcohol per day. Systolic pressure represents the force of blood against artery walls during the contraction phase of a heartbeat; high systolic pressure is a major risk factor for cardiovascular disease in people over 50, according to the American Heart Association.
Diastolic blood pressure, the bottom number, rose 1.14 mm Hg in people consuming an average of 12 grams of alcohol per day, rising to 3.1 mm Hg in people consuming an average of 48 grams of alcohol per day. Diastolic blood pressure measures the force against artery walls between heartbeats and is not considered as strong a predictor of hypertension risk as systolic.
However, the analysis found that the relationship between consumption and systolic blood pressure increase was lower in women than in men, and the diastolic pressure measured in women plateaued—and even declined slightly—above a threshold of 40 grams per day. Vinceti and the team suggest that the sex-related difference could be possibly due to only two of the studies stratifying their data by sex, which study authors wrote, “substantially decreases the statistical stability and precision of the sex-specific results, particularly in women.”
Vinceti says the study revealed several areas of further research that might be worth exploring, including quantifying whether the effects of a slight blood pressure increase translate into real effects on hypertension, stroke, coronary disease, and overall cardiovascular disease. He and colleagues are currently investigating modifying effects of sex, age and other factors.
“We are doing that by using the statistical approach named ‘one stage’, a rather novel technique that allows us to pool studies with only two exposure categories (i.e. trials with two arms only, or observational—cohort, case-control and cross-sectional—studies based on two exposure categories as well), and to linearly plot their results across the entire range of exposure, in our instance alcohol intake.”
Vinceti and his SPH colleagues Lauren Wise, Elizabeth Hatch, Kenneth Rothman, and Marcia Pescador Jimenez have recently used this approach to assess the relationship between selenium and cadmium intake and the risk of diabetes; greenness and cognitive decline, and benzene exposure and risk of childhood leukemia.
“This is a substantial improvement over previous meta-analytic approaches such as linear ones or those based on forest plots, allowing us to shape the non-linear relations between environmental/dietary exposures and health endpoints, including the identification of possible thresholds,” Vinceti says.