The study published by researchers from Stanford University found that there was no difference in how much weight subjects lost whether they followed a low carb or low fat diet. The researchers also found that there were no significant diet-insulin or diet-genotype interactions.
Conflicting results on influence of genetic factors
This finding conflicted with previous research, which had suggested that individuals may respond better to a particular type of diet depending on their genetic make-up. In other words, those with a low-fat-responsive genotype might lose more weight when following a low-fat, rather than a low-carb, regime.
“In this 12-month weight loss diet study, there was no significant difference in weight change between a HLF diet vs a HLC diet, and neither genotype pattern nor baseline insulin secretion was associated with the dietary effects on weight loss,” wrote first author Professor Christopher Gardner. “In the context of these 2 common weight loss diet approaches, neither of the 2 hypothesized predisposing factors was helpful in identifying which diet was better for whom.”
In an earlier study, researchers from the University of Copenhagen had postulated the fasting glucose and fasting insulin response markers would play a role in what diet approaches worked best. In their study, published last year, they found that prediabetic prediabetic individuals were extremely susceptible to weight gain when consuming a high-glycemic load diet, but saw substantial weight loss when consuming a diet with a low glycemic load or a diet high in fiber and whole grains — even without restricting calories.
Likewise, diabetic participants lost more on a high-fat and low-carbohydrate diet than on a low-fat and high-carbohydrate diet. This is the opposite of individuals with normal glycemic levels, who lost more on a low-fat and high-carbohydrate diet.
For entrepreneurs developing genetic testing platforms to inform nutrition choices, the confounding results are to some degree to be expected. The picture of how genetics affect nutrition is a highly complicated one, one that is bound up with lifestyle factors. But developers of these platforms said there’s no doubt that the opportunity is huge and will only grow as more evidence comes in.
Plugging confounding results into overall system
Mehdi Maghsoodnia, CEO of Vitagene, said the conflicting study results are to be expected as a more full picture of the issue emerges. Vitagene is a personalized nutrition firm that feeds the genetic data of its subscribers into an algorithm to determine the best supplementation regime and diet plans for them. Maghsoodnia, who has a background as a computer scientist, said he has approached the personalized nutrition as basically a data problem.
“I entered this space not knowing much about health care at all,” Maghsoodnia told FoodNavigator-USA. “What we take for granted in other industries in terms of data analysis and systems design has been wholly missing in health care.”
Maghsoodnia said that questions like weight loss and optimum eating have been dealt with piecemeal when it comes to nutrition research. While researchers do build on one another’s work, it is still very early days. And with the lack of an overarching direction, the rise of confounding results should come as no surprise.
“We learned a long time ago that if you want to expect repetitive results from a system you have to take a systems design approach. For instance, to get a package from New York to San Francisco in a predictive fashion we have designed a very elaborate system around package transportation,” Maghsoodnia said.
“In the delivery of nutrition and health care, though, it seems as if every day people are making decisions randomly based on personal knowledge. It’s not based on a system but on the clinician’s or dietitian’s personal determination. If we equate that to package delivery, it would be like if every package got there by an individual route and at a different cost,” he said.
Building a big data set
Maghsoodnia said the goal of his company’s platform is to plug individual consumers’ data into an overall model to continually refine and build the power of the proprietary algorithm. That way, a picture will emerge based on thousands or hundreds of thousands of data points, rather than trying to glean answers from a few dozen subjects in a research study, or in the case of a meta analysis, several hundred.
“These studies are really too early to make these kinds of definitive claims about which diets might work and why. We are really trying to bring that discipline of systems analysis to this field,” he said.
“Genetics is a signal but it is rarely an overwhelming signal in your eating behavior. Our algorithm has shown so far that genetics clearly has an impact. It accounts for about 25% to 30% of the overall impact of nutrition choices,” Maghsoodnia said.
Multiple factors in play
Gil Blander, PhD, founder and chief science officer of Segterra, which offers the Inside Tracker testing and analytics kit service which now incorporates DNA info, agreed that it is far to early to say that anyone has found ‘the answer’ for what might predict the performance of diets or of individual nutrients when talking about an individual consumer.
“There are many genetics companies that are recommending interventions such as supplementation based on genetic data only, which is wrong. For example, having a predisposition to low Vitamin D might suggest that you have high risk, but doesn’t prove you have low vitamin D. To prove it you need to test your blood levels of vitamin D. [The Stanford] paper shows that out knowledge and understanding of genetic data is still limited and we need to be carful in giving recommendations based on it,” he said.
Blander said the translation of how a person’s genetic makeup affects how they metabolize certain foods is too complex to be explained by looking at a couple of genetic factors. It’s like describing the tail of the elephant—it’s a start, but there is still much more to understand.
“What the authors were looking for was a complex trait—a response to a particular diet—and these are best explained by many genes that have a very small effect. Even the most robust genetic scores of 50+ genetic markers only explain up to 5% or so of the variability in a particular trait,” Blander said.
“In general, lifestyle-based inputs and blood data far outweigh any genetic inputs from the users. These factors could include the amount of exercise, overall dietary habits such as the ratio of processed to unprocessed foods, stress or inflammatory stimuli, etc. We adjust our recommendations proportionately to the published effects of these factors. Thus, the overall results of this study generally align with the hierarchy of our recommendations,” he concluded.