Researchers try to pin down a statistical association between diet drinks and cardiovascular disease:
Some studies have suggested that consumption of diet soft drinks may be associated with Type 2 diabetes and development of the condition known as metabolic syndrome — high blood pressure, abdominal obesity and other risk factors. Now a 10-year epidemiological study has found a link between diet soft drinks and cardiovascular disease.
The analysis, published online in The Journal of General Internal Medicine, included 2,564 adults over 40 living in Manhattan. Researchers found that diet and regular soft drink consumption were both associated with a number of risk factors for cardiovascular disease.
Even after controlling for many of those risks, including diabetes, the researchers found that daily consumption of diet soda was still independently associated with an increased risk for stroke, heart attack and death. The reasons for the association are unclear, the authors said, and the results must be interpreted with caution.
“The message for diet soft drink drinkers is not to be alarmed,” said the lead author, Hannah Gardener, an epidemiologist at the University of Miami. “What we’ve found is an association, and it might be due to chance or other unmeasured variables.”
Weird luck is always a possibility but my guess is that "unmeasured variables" is the answer - people who favor diet drinks also tend to possess some characteristic or engage in some behavior that leads to heart disease. However, that suggests that the researchers have an incomplete list of the behaviors and characteristics for which they need to control (or are able to observe), which may leave one wondering about the accuracy and relevance of many other studies.
SURELY I CAN ILLUSTRATE THIS WITH AN INAPT METAPHOR: Suppose NASCAR fans show a reduced life expectancy after controlling for obvious things like age, weight, race, and physical activity. Is the solution to stop watching NASCAR? Well, maybe further digging would establish that NASCAR fans are over-represented among groups that pick up speeding tickets and die in car crashes. That might leave the social scientists wondering about the direction of causality - do folks who like to drive too fast also like to tune in to NASCAR or is NASCAR prompting people to drive too fast?
Or, to swing back to diet drinks, maybe people who drink diet soda are more inclined to lie to themselves and the researchers and under-report their food consumption. Who knows?