It's my understanding that there was a retrospective study done that suggested an association, but as I know you're aware that is hardly a definitive answer of anything. Generally I consider retrospective studies as basis for deciding what warrants further study rather than drawing conclusions from them, but sadly I am in the minority, especially when it comes to the internet.
ETA: Kafka, here's a comment on a subscription site (can't link, sorry) in response to the same question:
"The client is likely referring to the JAVMA papers published in 2000 by Glickman et al. They compared breed related and non-dietary factors of dogs that developed GDV, and associated risk factors of GDV from an epidemiological standpoint. Although not clearly stated, it has been gleaned by many readers that pre-moistening dry food that was preserved with citric acid was a risk factor for GDV. These, plus similar papers subsequently published, look at the ingredient lists of diets from affected and unaffected dogs and evaluate commercial pet-food labels as to the order of ingredients and guaranteed analysis to try to draw conclusions regarding risk factors for GDV. In order to do this, huge assumptions had to be made, and these studies do not take into account variables such as manufacturing, ingredient quality, diet changes and formula changes from entry to the end of the study, or individual unreported feeding practices.
These studies underscore the limited information that can be interpreted from product labels. Unfortunately, we now have more myth regarding the impact of nutrition on GDV that will take years to sort through, as apposed to valuable scientific data. We simply cannot make the sweeping statements often heard (no citrate, no preservatives, no corn, no oils, meat only proteins). What is so intriguing is that these are tangible statements people can grasp onto as “the cause” of GDV. After reviewing numerous papers, what information we do have regarding risk factors is that feeding large, single meals per day, rapid ingestion, increased thoracic chest depth to width ratios, and having a 1st degree relative are risk factors, while feeding smaller, multiple meals daily is not a risk factor. The best recommendation to date is to feed multiple small meals, yet provide adequate energy for these large and giant dogs. And do not breed dogs that have had a GDV or a 1st degree relative with GDV. We do not have any evidence that a particular type of food, dry, canned, or raw, has any impact or inciting ability with regards to GDV. We do have evidence that our feeding methods may influence the incidence."