Uncorking new evidence in the alcohol-hypertension relationship

Klatsky reports a link between elevated blood pressures and increased risk of hospitalizations for coronary heart disease and stroke that is independent of alcohol intake. The study stresses the overall risks of hypertension, but the limited observational data needs to be interpreted with caution.

Using the alcohol intake questionnaires administered between 1978 and 1985 by Kaiser Permanente, data was analyzed on patients with coexistent adequate blood pressure information. As noted in their original article, this yielded an adequate data analysis based on 127,212 subjects. The prevalence of hypertension was notably greater among patients who consumed 3 or more drinks per day, with the heavier drinkers having an odds ratio of 1.7 for having blood pressures greater than 140/90 mm Hg. Even a moderate amount of alcohol intake, defined as 1-2 drinks per day, was associated with an increased risk of hypertension. As noted with other studies, increasing age and body mass index were associated with higher prevalence of hypertension. The risks of coronary heart and cardiovascular disease mortality progressively increased as blood pressure increased to the higher blood pressure strata, defined as 120-129/80-84 mm Hg, 130-139/85-89 mm Hg, or equal to or greater than 140/90 mm Hg. Surprisingly, however, the increase in alcohol intake from life-long abstinence to heavy drinking did not alter the cardiovascular risk within each blood pressure stratum.

These new data seem to be at odds with previous statistics, suggesting an alteration in cardiovascular risks with alcohol intake, often referred to as the J-shape association. Numerous studies, including a previous study by Klatsky et al, have demonstrated that consumption of moderate amounts of alcohol, usually defined as 1-2 drinks per day, significantly lowers the risk of cardiovascular risk compared with abstinence from alcohol.¹ With heavier alcohol intake, this reduction in cardiovascular risk disappears and starts to worsen.

These new data by Klatsky seem to suggest that the observed increase in cardiovascular risk with alcohol intake is secondary to hypertension rather than the effects of alcohol. But we are unable to draw such conclusions from this intriguing but limited observational data. One significant predictor of cardiovascular risk is low-density lipoprotein (LDL) cholesterol, which was unfortunately not evaluated in this study. It is known from the 6-year follow-up in the Copenhagen male study that the inverse association between alcohol intake and cardiovascular risk depends on the concentration of the LDL cholesterol. For example, alcohol intake did not alter the risk of cardiovascular risk in men with low LDL but the risk was significantly lowered in men with high LDL concentrations.² Additionally, several studies have shown that alcohol may quantitatively and/or qualitatively affect LDL. Data remains ambiguous on the effect of alcohol intake and LDL level; some suggest no significant effect while others suggest the possibility of lowering LDL.³ In contrast, many studies support the notion that heavy alcohol intake modifies the chemical composition of LDL, conferring greater oxidation and atherogenic properties in LDL.^3,4

Furthermore, alcohol abstainers may have a disadvantaged risk factor profile, including increased age, higher proportion of diabetes, and lower concentrations of high-density lipoprotein.² Abstainers also often include greater proportion of "sick quitters" who had been drinkers in the past.² In the previous studies by Klatsky et al, they had demonstrated that patients with heavier alcohol intake had greater mortality from cancer, cirrhosis, accidents, and nonmalignant respiratory conditions.¹ It is conceivable that comorbid conditions and the increased risk of death from noncardiovascular disease in the heavier alcohol drinkers altered the risk profile in this study cohort. This study evaluated the primary hospital discharge diagnosis. Consequently, those patients admitted with cancer who subsequently suffered a myocardial infarction in the hospital may not have been captured correctly in the data analysis.

Extensive data analysis on other coexistent cardiovascular disease risk factors will be required for proper interpretation. For example, it is known that derangements seen with metabolic syndrome (ie, obesity, glucose intolerance, dyslipidemia) tend to cluster at 4 times the rate than chance alone; and when these risk factors cluster along with cigarette smoking and left ventricular hypertrophy, the 10-year probability of heart disease can multiply 10-fold from only 4% with isolated systolic hypertension to 40% in subjects with multiple risks.⁵ It is also well known that alcohol enhances adenine nucleotide degradation and worsens hyperuricemia.⁶ In addition, recent data suggest that hyperuricemia is an actual modifier of cardiovascular risk, rather than a mere statistical association.⁷ And to make matters more complex, several medications can alter the serum uric acid levels.⁸ Were there differences in the serum uric acid level from alcohol consumption? Or were there differences in the serum uric acid levels from other medications used? Was there a disproportionate distribution of central obesity and metabolic syndrome amongst the alcohol intake groups that may have elevated the serum uric acid and LDL levels? It would be difficult to draw any conclusions without evaluating these and other interrelated cardiovascular risk factors.

The observational data presented by Klatsky is intriguing in that the J-shape association disappears when the risk is evaluated by blood pressure stratification. The reason for this remains unclear and further analysis that includes other risk factors such as metabolic syndrome and LDL cholesterol is warranted. This study does seem to support the notion that hypertension needs to be aggressively managed and its presence should serve as a flag to explore and treat the associated risks of cardiovascular disease.