Monday, February 22, 2016

Welch's Study Finds Grape Juice Makes You Smarter #NotTheOnion

Oh, and a better driver.


The findings were part of a study published ahead of print in the American Journal of Clinical Nutrition. The funded at least in part by Welch's study, Concord grape juice, cognitive function, and driving performance: 1 12-wk, placebo-controlled, randomized crossover trial in mothers of preteen children, started with 25 participants randomly assigned to consume 1.5 daily cups of Concord grape juice (containing 233 calories and more than a quarter cup of sugar (13.5 tsp)), or an energy and sugar matched non-concord grape juice "placebo" for 12 weeks followed by a cross-over of treatment arms.

19 participants described as "working mothers" completed the battery of cognitive tests, and 11 sufficiently completed driving simulations. Outcome measures for cognitive function included the number of correctly identified sequences, false positive responses, reaction times, a grooved pegboard assessment of manual dexterity (fill the holes with the pegs), and "Tower of Hanoi" testing. Outcome measures for driving performance included a 25 minute virtual driving scenario.

Now up front I should disclose I'm neither a statistician nor an expert in cognitive function testing so please consider that with my assessment. Ultimately, within the battery of tests (and there were plenty), a few in fact were found to show a statistically significant improvement in favour of the grape juice - most with p values of 0.05. Of course, do enough tests, and chance alone will dictate one will turn out to have a p value of 0.05. Putting that aside, I also can't help but question the significance, even if true, of for instance completing a psychomotor skill test in 60.4s vs. 63.2s (P < 0.05), or as it would pertain to driving, a "car following" accuracy of 0.96 vs 0.97 (P = 0.05).

Sounds like the authors may have wondered the same given their use of the word "subtle" in describing the improvements and their proviso when considering the battery of tests that were run,
"it is important to acknowledge that there were no effects on most cognitive outcomes".
Not surprisingly, that proviso did not stop Welch's from putting out a press release that opened with,
"New research by the University of Leeds in the UK suggests that drinking Concord grape juice daily can benefit certain aspects of memory and everyday tasks in people with stressful lifestyles – specifically working mothers".
So there you have it - the 13.5 teaspoons a day of sugar way to just barely, sort of, in a very small number of very specific tests, as revealed by an incredibly small sample size, make yourself smarter and a better driver.

[And to consider too - did these, acknowledged even by the authors to be meagre, findings really warrant publication in the American Journal of Clinical Nutrition, a journal with an impact factor (nearly 7) higher than roughly 95% of all other scientific journals, and one that was selected by the Special Libraries Association as one of the top 100 most influential journals in Biology and Medicine over the last 100 years? And/or does this just speak to impact factors' lack of utility?]