A press release last week indicated that diet soda drinkers had a 61% higher risk of stroke than people who didn’t drink soda (regular or diet) at all. Is that really likely?
The Northern Manhattan Study (NOMAS) was launched in 1993 to study variations in risk of stroke between ethnic groups and any risk factors that might contribute to those variations. Many papers have been written based on the ongoing NOMAS. The study referred to in the press release linking diet soda and stroke has not yet been published.
Participants were given a questionnaire at the start of the study. The questionnaire asked a lot of food and lifestyle questions, but the important ones for this study concern soda intake.
2564 people over the age of forty were divided into seven groups according to their soda intake.
- those who drink less than one soda a month, whether diet or regular
- those who drink diet soda six times or less in a week
- those who drink diet soda seven or more times a week
- those who drink regular soda six times or less a week
- those who drink regular soda seven or more times a week
- those who drink both diet and regular soda six times or less a week
- those who drink both diet and regular soda seven or more times a week
The press release contrasts the group of people who drink diet soda seven or more times a week with the group of people who drink less than one soda a month. The people who drink diet soda had a 61% higher risk of stroke than the people who don’t drink soda.
It’s a surprising number. I have no reason to doubt it, but it’s just a correlation at this time. In order to be really convincing, we need both replication of these results from an independent study and a mechanism for diet soda to cause so many more strokes.
Another bit from the press release that I found interesting was the number of vascular events in the study population. A vascular event is described elsewhere in NOMAS as a stroke or myocardial infarction, fatal or non-fatal. Over the course of 9.3 years, the 2564 study participants experienced 559 vascular events. Researchers accounted for known risk factors, such as age, sex, race or ethnicity, smoking status, exercise, alcohol consumption, and caloric intake.
In the general U.S. population of over-40 individuals, 11.4% will have vascular events over the course of 9.3 years. Extrapolating to a population of 2564, that means we would expect 295 vascular events over that time period. 559 is nearly double the expected rate. It seems everyone in the NOMAS group has a surprising rate of vascular events.
Sources: Press Release for NOMAS
Image by Joelk75, used with Creative Commons license.