But the risk for individual women remains extremely low, particularly in younger women.
Danish researchers wrote in the New England Journal of Medicine that the findings suggest a higher risk of stroke in particular for women using vaginal rings, and possibly hormonal skin patches -- though the second finding was based on a smaller group of women and could have been due to chance.
Dr. James Simon, a women's health researcher at George Washington University in Washington, D.C. told Reuters Health other factors -- such as the belief that a patch or a ring might be safer for women thought to be at risk -- may explain the higher rate of stroke in that group.
Simon, who wasn't involved in the new research, said the findings probably shouldn't change how doctors prescribe birth control. The risks seen in the study, he said, pale in comparison to the risks of stroke, heart attack or death faced by women who get pregnant.
"None of the hormonal contraceptives studied… were associated with an excess risk of stroke that was unacceptable, considering their contraceptive and noncontraceptive benefits," Dr. Diana Petitti of Arizona State University in Tucson wrote in an editorial accompanying the study.
Previous attempts to assess the risk of stroke or heart attack due to hormonal contraceptives have produced conflicting results.
The researchers on the new study, led by Dr. Ojvind Lidegaard from Copenhagen University Hospital, crunched the records of 1.7 million Danish women -- essentially the entire female population age 15 to 49 -- to assess the potential dangers. The women, all without a history of heart disease or cancer, were followed for 15 years beginning in 1995.
Each individual woman's risk was small. One in every 4,700 women had a stroke each year and one in every 9,900 suffered a heart attack.
Women taking contraceptive pills with a combination of estrogen and progestin tended to have a higher risk of stroke and heart attack than those not using hormonal contraception.
For some hormone combinations that difference could have been due to chance. But women using estrogen with norethindrone or desogesterel at certain doses, for example, had double the risk of both complications compared to non-users.
Still, Simon said the new research "shows very little difference between the different pills for the same dose of estrogen, which will make women's choices larger."