Trevor Butterworth, Forbes.com Contributor
According to the Daily Mail, the New York Times’ “rival” as the most-read English language newspaper in the world, “research” has revealed that “drinking just a single can of diet fizzy drink every day can increase the risk of a heart attack or stroke.”
It should be noted that “research” has also revealed the risk of the Daily Mail misreporting a study’s findings, especially when there’s an opportunity to write an alarming headline. As Dorothy Bishop, a Professor of Neurodevelopmental Psychology at Oxford University, noted in giving the paper her “Orwellian Award for Journalistic Misrepresentation” the Mail sets the standards for inaccurate reporting of academic research.
But, in this case, appearances would seem to confirm journalistic accuracy: researchers did, indeed, find an “increased risk” for vascular events (heart attack and stroke) among high consumers of diet soft drinks. The question is, however, what did this mean? Was the risk is real or the product of something other than diet drink consumption?
This proved tortuous for publications covering the study: on one hand they had an amazing, counter-intuitive finding that made for a great, attention-grabbing headline – who knew that diet soda could give you a heart attack? On the other, there was a critical absence of the kind of health and behavior information in the study that might actually buttress this suggestion of cause and effect – and the most its authors were willing to commit to was that their research “suggests a potential association,” which is about as adamantine as wet tissue.
The answer was simple, embrace contradiction. Bait and switch.
Men’s Health, for instance, declared “Another Strike Against Diet Soda,” saying “if you’ve been looking for a reason to nix your soda-drinking habits, the latest research may offer you the perfect inspiration.” But midway through the news story, the magazine suddenly and weirdly switched track. “So what does this mean for you?” it asked. “Frankly, very little,” it answered. For good measure, it went on to highlight in yellow why the study is so weak just to drive home how schizophrenic its own article was. Similarly, the The Daily Mail, having started on top of the Mountain of Imminent Doom, slid inexorably downwards to the fog of suggestive association by the end of their story.
The problem with the study, Diet Soft Drink Consumption is Associated with an Increased Risk of Vascular Events in the Northern Manhattan Study (Gardner et al., Journal of General Internal Medicine) is that the numbers are low enough to leave us uncertain as to whether diet drinks were the cause or simply correlated with other factors that led to these events.
There were 163 people who consumed one or more diet drinks per day at the start of the study, and there were 51 vascular events in the following 9.8 years. The mean age at baseline was 68.6 years of age. Light drinkers of diet soda, defined as between one and six drinks a week, were not at an increased risk.
One of the key limitations in the study is that the researchers only accounted for baseline dietary and health data. Over the follow-up period of 9.8 years they did not account for how a person’s diet might have changed, how their weight changed, or how their consumption of diet soda might have changed. All of these absences are highly problematic.
As it is reasonable to assume that people are more likely to consume diet drinks due to concerns about weight and health, the question is whether this correlated with lifestyles that by themselves were risk factors. For instance, if the daily diet soda drinkers saw their weight yo-yo (perhaps because they were drinking diet soda to lose weight), then the correlation with vascular events might lie with these changes in weight.
The second issue with having just baseline measurements for this group is whether, as diet drink consumers, they are more likely to make significant changes to their diet and behavior over the course of a decade. These sorts of factors become very important in teasing causality from correlation in small numbers over time.
At the same time, there may also be cultural factors that contribute to drinking diet soda which are also related to vascular health but which are not caught by controlling for “demographics.” This is important given the significant demographic differences among diet and regular soft drink consumers in the study. As the researchers note, “frequent diet soft drink consumption was uniquely associated with white race, former smoking, hypertension, elevated blood sugar, lower HDL, elevated triglycerides, increased waist circumference, BMI, peripheral vascular disease, previous cardiac disease, and the metabolic syndrome.”
Another weakness, in terms of the study being able to determine whether diet soft drinks have a causal role in vascular events, is that the researchers didn’t have data on the range of ingredients. Diet soft drinks are not just one undifferentiated product with different labels. Given that the association between diet drinks and with vascular events is unexpected and unsupported by other research, a study that controlled for specific ingredients, such as the type of sweetener and caffeine content, is needed to assess whether there are grounds for a plausible mode of action.
In sum, the small numbers in this study combined with the kind of data collected (limited baseline data, no details of drink ingredient differences) and the narrow demographics do not support the contention that drinking diet soft drinks is going to put you at a risk for a heart attack.
But why let that get in the way of a good headline?
The opinions expressed are solely those of the author and do not necessarily reflect the views of Comcast.