With so much inaccurate information about diet and health out there (looking at you, social media and blogs), the last thing you would believe is misleading -- and the one we have trust in most -- are research studies.
People rely on unbiased research to find out important statistics about all facets of nutrition, from sugar intake and food supplements, to genetically modified food and cereals.
However, recent research from the Charles Perkins Centre at the University of Sydney suggests there is bias in industry-funded research studies, but the full extent of which is still unknown.
"We're starting a whole program of work in this area. The first thing we did was a review of all the studies that have looked at the association of industry sponsorship with the outcomes of nutrition studies," Lisa Bero, professor, chair of Medicines Use and Health Outcomes and head of the bias node at the Charles Perkins Centre at the University of Sydney, told The Huffington Post Australia.
Bero and her team reviewed 775 reports in the medical literature, narrowing down to 12 relevant reports, to determine whether nutrition studies funded by the food industry were "associated with outcomes favourable to the sponsor".
There's lots of ways to spin the wording so that the results don't agree with the conclusion.
"It was a little surprising because most of the studies only looked at the conclusion of the research. By that I mean the author's interpretation. So, if it were industry sponsored, they were more likely to have a conclusion that favoured the industry sponsor," Bero said.
"What we found is that only three of the studies looked at the actual results or data. That's something we're really interested in doing in the future."
Bero's investigation has confirmed that researchers know little about the influence of corporate sponsors on nutrition studies. This latest paper comes off the back of Bero's previous study which found nutrition studies funded by artificial sweetener companies are more likely to lead to favourable results.
"It's really how the authors interpret the results. For example, a study might say that it has a statistically significant effect, and then the authors conclude that it's showing a 'trend' or 'effect'," Bero said.
"Often the studies, if they're measuring a harm, they may discount some of the data that shows harm and so they can conclude overall they didn't demonstrate a harm."
As Bero highlights, this bias can be subtle.
"There's people who measure what's called 'spin' in conclusions and there are ways to look at spin. What these studies were really looking at was whether the results agreed with the conclusions," Bero said.
"There's lots of ways to spin the wording so that the results don't agree with the conclusion."
As for the industries which were reviewed, the main associations were sugar, food supplements and GM foods.
"We looked at reviews of reviews, and they were on different topics. Some of them were related to sugar sweetened beverages, some of them were related to food supplements, some were on genetically modified foods," Bero said.
If you have a study that's unfavourable or parts of it are unfavourable, it's hard to tell if it's gotten published, or if all of it has gotten published.
The wheat and dairy industries will be part of the bias node's next investigations.
"We're actually looking into that now. Unfortunately I can't tell you what the result of that is, but there's a big need [for research] because if you're doing a clinical practice guideline or a systematic review, it's really these results that matter. We want to look at the effect on the actual conclusion," Bero said.
And what's the first (or only) part most people read when looking at research papers? The easily readable conclusion.
"The conclusions are important, definitely from the perspective of the public because a lot of time that summary of the article is the take-home message or conclusion that people get," Bero.
So, what happens if more industry sponsored nutrition studies are proven to be biased, and this becomes known?
"If you look at other areas where the effects of industry sponsorship have been shown, like in the pharmaceutical research area and the tobacco research area, people have actually applied more consistent quality criteria to make sure the research is of a certain quality," Bero said.
There's also another part of food industry bias in nutrition research that we don't even think about.
"You'd also want to try to make sure that all the data is being published. In the nutrition area they don't have things like clinical trial registries like they do for drug studies, for example," Bero said.
"So if you have a study that's unfavourable or parts of it are unfavourable, it's hard to tell if it's gotten published, or if all of it has gotten published. That's a huge bias in the pharmaceutical and tobacco studies."
As you can imagine, being in the field of studying corporate bias with large businesses' integrity at stake has its pitfalls, the most recent for Bero being that she's being 'monitored' by Coca-Cola.
"That was just an email that came to light, and frankly I have no clue as to how it did. It was through DC Leaks," Bero said.
"The email basically described how a consultant who was working for Cola-Cola came to the Charles Perkins Centre and attended some of the seminars. The email had a little bio of me that said they were 'monitoring' my research in the future and that's how the email ended."
We have a lot of work ahead of us.
However, this was not a surprise for Bero, who has dealt with corporate intimidation before.
"I've been studying corporate influences on research for over 20 years and so when I saw that email I thought, this is just like the tobacco industry. I wasn't very surprised. But it is creepy, that's what Stephen Simpson said," Bero explained.
"A lot of junior faculty members I know related to the tobacco industry have felt threatened because the risk is always that they can try to suppress and discredit your work if they don't like it. I do think if you're more early in your career it's more threatening.
"The tobacco industry tried to get me fired once when I was very early in my career. It didn't work, though. I have a thick skin."
From here on, Bero and the Charles Perkins Centre bias node will continue to research whether industry sponsored nutrition studies may be more likely to have favourable conclusions for the industry, rather than non-industry sponsored studies.
"It's very early on for us and I wish we'd found more data, frankly, because there's so much written about bias in this area and then to find it hadn't been studied nearly enough as some of the other industry areas," Bero said.
"We have a lot of work ahead of us."
Click below to follow HuffPost Australia on Snapchat!