The Obesity Epidemic (part 6): The Effects of Diet

The previous blog in this series discussed some of the genetics surrounding obesity.  More than 20 genes have been discovered that may be involved. 

I’m sure there are more. 

The next issue I’d like to examine is the effects of diet on weight loss.  But interestingly, there is evidence that the type of diet one has is influenced to some degree by the types of genes you have.

First, there are generally four types of diets: low fat, low carbohydrate, low calorie, and very low calorie.

The low-fat diet, as the name implies, reduces the percentage of fat in one’s diet. Concurrently, calories are reduced because energy-dense fat is reduced.

The low carbohydrate diet is higher in protein and fat, since carbohydrates are reduced. These include the Atkins and Protein Power diets; the new Paleo diet would also fit in this category.  

Low calorie diets are what their name suggests—they typically produce a deficit of 500-1,000 calories/day. The DASH diet and Weight Watchers are low calorie diets.

Finally, the very low-calorie diet is a near-starvation diet, providing 200-800 calories/day. 

Well, all of these diets will cause you to lose weight.  An interesting question is, which is better?

Guess what—there is NO scientific evidence that one diet is better than the other.  They all cause weight loss—and perhaps weight loss is observed because they all result in a reduction in calories—a conclusion reached by almost all the studies.  Let me repeat: there is no evidence that having high or low carbohydrates, high or low protein, or high or low fat confers a diet that has any more superiority over another for weight loss.

Such was a 2012 report of 424 men and women who had BMI’s of over 30, tracking fat as well as muscle loss.  Tracking muscle loss makes this study different from most weight studies, who only look at waist circumference and/or total weight.   Further, this study looked at total fat, visceral fat, subcutaneous abdominal fat, as well as hepatic fat, using X-ray analysis.

Simply, visceral fat is “deep fat” that wraps around organs, resulting in a large belly and waist, and is considered the most dangerous of the types of fat.  Subcutaneous abdominal fat is fat found directly under the skin, and is typically measured by a “pinch test.” Hepatic fat is fat leading to “fatty liver disease” wherein fat accumulates in liver cells.

This 2012 paper looked at the effects of four weight loss diets that all had reduced calories (the following percentages are amount of energy provided by the nutrients): 1.) low fat (20% fat), average protein (15% protein), high carbohydrate (65%); 2.) low fat (20% fat), high protein (25%), high carbohydrate (55%); 3.) high fat (40% fat), average protein (25% protein), low carbohydrates (35%).

They found that after 6 months on the diets participants lost an average of 9.3 pounds of total fat and 4.6 pounds of muscle.  Of the fat, 5.0 pounds were abdominal fat, 3.3 pounds were visceral fat.  There was no significant difference for hepatic fat.  Further, as stated above, there was no significant difference between the diets.  Finally, after two years, participants gained back 40% of their weight loss. There was also apparent “cheating” on the diets, as, for example, urinary nitrogen levels were the same between the low and the high protein groups (it is expected that as protein in the diet is increased, then there is more excretion of nitrogen).

In recent years there has been a lot of interest shown in high protein diets.  A 2013 paper reported a meta-analysis of 15 studies involving 1,990 male and female subjects, all obese, and lasting for a year or more. The dietary protocol of all studies was high or low protein, low fat, variable carbohydrates, and 11 studies had energy restriction and 4 did not.

What did they find? No effects of either high or low protein diets on weight, waist circumference, or fat mass. Nor were there effects on total cholesterol, high density lipoprotein (HDL) ,though HDL was near statistical significance, and total triglycerides, or blood pressure.  However, high protein diets DID result in statistically lower fasting insulin levels. And finally, for those of you on high protein diets (like myself sometimes!), there was no difference in renal (kidney) function.

HOWEVER, all may not be so simple—which seems to be a general conclusion in dietary studies. A 2013 meta-analysis of “all-cause mortality” which looked at 17 studies covering 272,216 people, the largest such analysis to date, concluded that low carbohydrate diets had a statistically significant increase in mortality, and there was no effect one way or the other on cardiovascular disease. The authors conclude:

“Given the facts that low-carbohydrate diets are likely unsafe and that calorie restriction has been demonstrated to be effective in weight loss regardless of nutritional composition…it would be prudent not to recommend low-carbohydrate diets for the time being.”

Now, to turn to the point I introduced at the beginning of this blog: the types of genes you have may influence the effectiveness of weight loss diets as well as where on your body you lose weight.

As mentioned in the Obesity-5 blog, some FTO genes have been associated with obesity.  Actually, I glossed over the point that there are different kinds of FTO genes—some are associated with obesity and some are not.  Of relevance here is that one of these genes is known by scientists as rs1558902.   Let’s call it gene “A” for this discussion.  Further, let’s call the normal FTO gene (not causing obesity), “T.”

In a study published in 2012, 742 obese adults were classified as having the “A” gene or “T” gene.  They next randomly divided these individuals into four diets: 20% fat, 15% protein, 65% carbohydrate; 20%, 25%, 55%; 40%, 15%, 45%; and 40%, 25%, 35%.    Interestingly, the found that HIGH PROTEIN diets showed the most impact on weight loss parameters amongst those that that the “A” gene.   So in other words, if you only had the “T” gene you did not respond to a high protein diet. 

This constitutes proof that your response to a particular diet depends on the types of genes you have—or to say it differently, people that do not respond to high vs. low protein diets may do so for innate reasons.

I can only assume that given the large number of genes that influence obesity, many other such interactions would be found—leading to the conclusion that not one diet fits all people. 

But, the question is NOT just which diet causes weight loss, but HOW sustainable is the diet? That is, is the diet one that folks will stay on for years? The rest of their life?  How about other life-style changes that are incorporated in to some diets, such as exercise—is it sustainable?  A rather shocking observation in all of the diet studies that I looked at is that the drop out rates were discouraging. The 2012 study above had 20% drop out rates; the 2013 meta-analysis of 15 studies had dropout rates ranging from 8% to 55%, with most around 30%.  Plus, there is evidence of dietary cheating in some studies—which must raise questions about some conclusions.

In summary, there is no scientific consensus as to which diet is best for weight loss.  The ONLY consensus (so far) is that calorie reduction has the greatest impact, and even that may depend on they types of genes you have. 

And now I’m done with Obesity!! On to other subjects. 

Useful References:

http://www.cmaj.ca/content/174/1/56.full

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3041737/

http://www.nejm.org/doi/full/10.1056/NEJMoa0804748#t=articleBackground

http://ajcn.nutrition.org/content/95/3/614.long

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3636027/         

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2646098/

The Obesity Epidemic (part 5): The Effect of Genes

In the previous blog of this six-part series I concluded that increased sugar consumption in general, and sugar sweetened sodas specifically, seem to account for a large amount of our obesity.  And obesity is correlated with a suite (no pun intended) of conditions known as “metabolic syndrome,” of which diabetes is not only associated but may be, in some cases, reversible with weight loss. Same can’t be said for cardiovascular disease, it appears.

So, it seems the major reason for obesity in the world and the U.S. specifically is excess calories, of which sugar would be the major contributor to the epidemic.

Although the above reasons may be the major cause of obesity, there are factors other than consuming too much sweetened soft drinks—such as particular genes, unique gut flora, or perhaps even viruses.

It appears that genes play a major role in obesity.  Now, this does not mean that the increase in obesity in the last 50 years in the US or worldwide is due to an increase in obesity genes.  In order for that to have happened there would have to have been a selection for obesity—that is, obese people having more children.  And even if that were to have been the case, of which there is no evidence, there would not have been enough time for this to have had much of a measureable effect. 

So, the presumption is that the (recent) world population has always had the same genetic potential to become obese—certainly over the last couple hundred years, anyway.  But some people may have more of a tendency towards obesity than others. Certainly in our own anecdotal experience some people seem to eat and remain thin, while others have a difficult time keeping the pounds off.

Now, for us here the interesting question, or at least the question that interests me, is: HOW MUCH do genes influence obesity?

For geneticists, getting a handle on this would almost have to start with a concept known as “heritability”—or in other words, how much do genes influence the trait in question.

It is intuitively obvious that many traits are influenced by both the genes and the environment.  Other traits would not be influenced by the environment. If we look at eye color, we’d say that the environment does not affect it very much; geneticists would say it is highly “heritable”—it really doesn’t matter what your environment is like.  Your eye color is going to depend on what kind of genes you have, not by what kind of environment you grow up in.  Another is your blood type.   For traits like this we’d be safe to say that heritability (h) is “1.0” (h ranges from 0 to 1.0).  However, for other human traits, like alcoholism or schizophrenia or some reading disabilities, it is easy to see that there would be an environmental influence.  For example, an estimate of “h” for personality is 60%, meaning that 60% of the total variation for personality is  due to genetic causes, and 40% is due to the environment.

We have to be careful of our interpretation of this 60% value.  What it means is that when you look at a particular group of people (those that were examined for their personality type), and measure all the types of variation that exists for a particular personality trait (of which there would be a lot!), 60% of the variation is due to genetics, and 40% is due to environmental influences.  The “h” value DOES NOT apply to individuals.  It doesn’t mean that YOUR personality is 60% determined by genes and 40% by the environment.

Now, heritability estimates have been done for years over many many characteristics, in everything from fruit flies to chickens to corn to humans. 

So, how do we estimate heritability?

It turns out that for humans, there is a wonderful genetic “tool” known as “twins.” As we all know, identical twins are genetically identical (for all intents and purposes for our discussion (critics please don’t get all excited about epigenetics—a whole other discussion), while fraternal twins are no more genetically identical than another sibling.  So, you can see that if you looked at identical twins, which are genetically identical, and measured some trait, such as personality, ANY variation in their personalities would be due to their environment.  Conversely, if you looked at fraternal twins, which only share ½ of their genes—like “normal” brothers and sisters—their differences in personality would be due to both genetics and the environment. 

So in this way I think you can kind of see that by examining both types of twins you can make an estimation of “h”.  So, what do they find when they look at “h” for obesity?

Since 1997 many twin studies have found that variation for BMI (and recall from my Obesity-1 blog how inaccurate that is) has a strong genetic basis, with “h” ranging from 0.55 to 0.85 (remember, where 1.0 represents 100% control of the trait by genes).

Now, most studies have looked at adult twins.  A problem with this is that the older a person is, the more influence the environment may have.  However, by looking at children, in theory, you’d get a better estimate of “h” because, well, your results would not be influenced by years of environmental “entanglement.” Or in other words, adults have lived longer in an “obesogenic environment.” (I’ve just been waiting for an excuse to use the word “obesogenic”!).

A 2008 study in the United Kingdom looked at 5,092 twin pairs aged 8 to 11 years old, with an average age of 9.9 years.   BMI’s were determined, as well as waist circumference (a better measure of obesity).  What did they find? “h” was 0.77 for BMI and 0.76 for waist circumference.

So the conclusion here is obvious: as with other studies, genetics has a majority influence on obesity.

But I just need to make a quote from this 2008 paper, as their words are much better than mine:

“The fact that siblings’ experience of being served similar food, being given the same options for television viewing and active outdoor play, seeing the same behaviors modeled by parents, and going to the same school does not make siblings more similar is a challenge for etiologic models that highlight the home environment as the root cause of obesity. This finding will, however, come as no surprise to parents, who are well aware that their children come in different shapes and sizes despite having a similar upbringing.…. Results from the present study highlight the fact that excessive weight gain in a child is unlikely to be the fault of the parents and is more likely to be due to the child's genetic susceptibility to the obesogenic features of the modern environment.” (Emphasis mine).

Now, I may quibble a bit with these authors in terms of not blaming the parents—I mean, who else is responsible for a child’s “obesogenic” environment?  Seems to me that if it is observed that a particular child has a predilection for gaining weight the parent can intercede and restrict that child’s caloric consumption.  Now, the “fairness” of this in child rearing is another matter, but certainly a parent could restrict calories, at least up to the point they have control over the child’s ability to sneak off and buy soft drinks, for example.  But, being a parent of six I recognize this as easier said than done!

We digress here! The point is that a CULTURE that supports unrestricted caloric consumption, which ours does, will lead to obesity—and some will gain faster than others.  But we will ALL probably get fatter.

Well, if genes control our tendency to get fat, the next question for a geneticist is: “Just WHAT genes are those?”

Around 20 genes have been implicated so far in controlling obesity.  Two of them have received recent press: the FTO gene, the KSR2 gene, and the IRX3 gene.

The FTO gene was the first to be discovered in 2007, and it is associated with increased food intake, not energy expenditure.  A recent 2013 study reports that the FTO gene may do this through the hormone “ghrelin” (a hormone produced by gut cells to stimulate hunger)—people with the gene found high calorie food (particularly fat) more desirable and were hungrier than those that don’t. 

Further, a 2013 study showed that MRI brain scans showed that people with the FTO gene showed brain activity in areas associated with motivation to eat remained high before and after a meal.

However, it is also a good example of how the genetic control of obesity is probably controlled by hundreds, if not thousands, of genes, and FTO’s effects are “minor,” since it only accounts for around 6.6 pounds of excess weight compared to a normal person without this gene.  Interestingly, it is found in 45% of Europeans, 52% in West Africans, and 14% in Chinese/Japanese.

In March of 2014 it was reported that the IRX3 gene may be the main controlling of the FTO gene.  This further illustrates the complexity behind the genetics of obesity.

The KSR2 gene has been found to impair fatty acid and glucose degradation.  Presumably, neither fat nor glucose is broken down and therefore they are converted to stored fat instead.  This is the first time evidence has been shown that obesity is due to “slow metabolism.”  However, only 2% of obese individuals have this gene, so it certainly does not affect a majority of the obese population. 

And so, it is clear that the future of such studies will uncover many more genes, perhaps even those that have more major effects.  And perhaps now that specific genes are known, new drugs can be developed that target these specific sites.

It is easy to speculate that humans have had selection for the ability to get “fat” during times of plenty, in order to make it through lean times.   It doesn’t take much of an imagination to picture a tribe people huddled in a cave, completely cut off by an extended, severe snowfall, running out of food.  Certainly under those conditions the fattest, and those with the slowest metabolism, would have the greatest probability of surviving and living to pass on their “fat” genes.  Us skinny folk would not be so lucky under those conditions. Conversely, during times of plenty those that got fat the quickest would be those that would have the ability to survive the lean times.

Today we reflect the genetic selections that have taken place over the last 50,000 years of Homo sapiens time on this earth—we are just cave men dressed in suits with a genetics designed for survival under conditions that don’t exist any more.

In our obesogenic environment our bodies just yell “Hooray!” and get on with getting fat—because, well, that’s what they’ve always done.

Looks like this obesity series will extend to at least one more blog—I can hardly leave the topic without looking at diets, which I’ve not been looking forward to: I can hear comments now.  Atkins Diet is the best! No, Paleo is for me! No, no, it’s the South Beach diet! And so on. 

Useful References

http://ajcn.nutrition.org/content/87/2/398.full

http://news.sciencemag.org/2013/07/obesity-gene-linked-hunger-hormone

http://www.jci.org/articles/view/44403

http://www.nature.com/nature/journal/v507/n7492/full/nature13138.html