AAN: Smoking May Cancel Stroke Benefit of Drinking
Reported April 13, 2010
TORONTO — Moderate drinking may protect against stroke, but not if the drinker is also a smoker.
An analysis of stroke incidence in a long-running cohort study showed that both nondrinkers and heavier drinkers were at higher risk than moderate drinkers, according to Yangmei Li, a PhD candidate at Cambridge University.
That effect — known as a U-shaped curve — has been seen in some studies but not in others, Li told reporters at the annual meeting here of the American Academy of Neurology.
The reason, she said, may be the impact of smoking.
She and colleagues looked at data from the European Prospective Investigation into Cancer (Norfolk) study, which also included information on cardiovascular outcomes.
Volunteers ages 39 through 79 began taking part in the study between 1993 and 1997. At baseline, they were free of heart attack, stroke, and cancer and were followed up for incident stroke until 2008.
All told, the researchers looked at the relationship between alcohol and stroke in 10,171 men and 12,353 women.
Over the follow-up period — 12 years on average — there were 876 strokes, Li said in a press conference before her late-breaking poster presentation.
Drinking behavior was quantified in units of alcohol, where one unit was roughly equivalent to a glass of wine, she said.
Cox regression analyses showed that, compared with nondrinkers:
* Those who drank between one and three units of alcohol a week had a hazard ratio for stroke of 0.72.
* Weekly drinking of 3.1 through seven units yielded a hazard ratio of 0.56.
* The hazard ratio for 7.1 through 14 units was 0.55.
* Weekly drinking of 14.1 through 21 units yielded a hazard ratio of 0.71.
* The hazard ratio for 21.1 through 28 units was 0.82.
* And weekly drinking of more than 28 units yielded a hazard ratio of 0.96.
The risk was statistically significant for the first four categories, Li said.
The pattern was much the same for a multivariate analysis adjusting for — among other things — smoking, she said.
But smoking had a statistically significant interaction with alcohol, she said, when the researchers compared those who never drank or smoked with other possible combinations of substance use.
Because of small numbers, many of the comparisons failed to reach statistical significance, she said. The exception was people who combined drinking and smoking, whose risk of stroke was 75% greater than nondrinking nonsmokers.
“We haven’t found any previous studies reporting this interaction term,” Li said, adding it may be the reason not all studies have reported the same U-shaped curve.
“Different results from previous studies may be caused by different proportions of smokers in the cohorts,” she said.
Several studies have “documented that moderate alcohol consumption is associated with a lower risk of stroke, according to Ralph Sacco, MD, of the University of Miami’s Miller School of Medicine, who was not part of the study. “Our own work from the Northern Manhattan Study has shown the J-shaped relationship between stroke and alcohol use.”
But, he added, the data from that study “did not suggest an interaction between smoking and alcohol.”
On the other hand, Sacco said, smoking is a “very potent risk factor for stroke and may outweigh the protective effects of alcohol.” And it’s possible that smokers may on average drink more than nonsmokers, he added.
“While moderate alcohol use may be protective for stroke, the adverse health effects of smoking may eliminate any benefits,” he said. “We urge smokers to quit and others to never start.”
Source : MedPage Today
