Weight Loss Support - Interesting article: Artificial Sweeteners and Obesity




TheWalrus
06-16-2009, 09:09 PM
http://www.medscape.com/viewarticle/704432

"People who use artificial sweeteners are heavier, more likely to have diabetes, and more likely to be insulin-resistant compared with nonusers, according to data presented here during ENDO 2009, the 91st annual meeting of The Endocrine Society.

...

First author Kristofer S. Gravenstein, a postbaccalaureate researcher with the Clinical Research Branch at the National Institute of Aging (NIA), National Institutes of Health (NIH), said the association may reflect the increased use of artificial sweeteners by obese and/or diabetic study participants. 'This is a cross-section study,' Mr. Gravenstein told Medscape Diabetes & Endocrinology, 'so there are limitations we cannot say that artificial sweetener use causes obesity, we can say it is associated with it.' "


mandalinn82
06-16-2009, 09:34 PM
It's interesting, but not new or groundbreaking - variations on this study have been done many times in the past.

The problem with a study like this is the inability to show anything more than correlation, which is NOT causation. Considering what I know about biochemistry, metabolism, and how the body works, I'd actually consider it more likely that people who are overweight are, as a group, more likely to be trying to lose weight, and therefore more likely to be using artificial sweeteners than groups who are normal weight. The study can't tell us ANYTHING about that, it can only say that if you're overweight, you're more likely an artificial sweetener user.

As always, if a sweetener increases your cravings, makes you hungrier, or causes unexplained weight gain or other unpleasant symptoms (aspartame, for example, makes me throw up...ick), avoid it! But don't do it on the basis of a correlational study, which gives no causal data.

Heather
06-16-2009, 11:29 PM
This is why I love Amanda!


kaplods
06-17-2009, 12:20 AM
There was a similar study several years ago which found that people who drank decaf coffee had a lot more cardiovascular disease than those who drank regular coffee. Initially there was some speculation (mostly in the popular media, not among scientists) that regular coffee was perhaps a healthier choice than decaf. Several doctors came out on talk shows to point out that the more likely explanation is that coffee drinkers only switch to decaf AFTER learning that they have health issues, and that healthy coffee drinkers are less likely to drink decaf.

Heather
06-17-2009, 08:20 AM
In my psych courses I spend a lot of time teaching students not only to identify correlation studies, but also causal language, and to come up with alternative explanations, like the one kaplods gave above. That's a great example!

mayness
06-17-2009, 08:42 AM
I'd actually consider it more likely that people who are overweight are, as a group, more likely to be trying to lose weight, and therefore more likely to be using artificial sweeteners than groups who are normal weight. The study can't tell us ANYTHING about that, it can only say that if you're overweight, you're more likely an artificial sweetener user.

This seems like the most simple/logical explanation to me, and I am surprised that they (almost) never mention it in articles about these studies. Well, "surprised" is the wrong word, of course I realize they want to suck people in by suggesting a more exciting/shocking possibility. I guess I'm "disappointed," not surprised.

Rock Chalk Chick
06-17-2009, 09:12 AM
I'm always baffled by how these studies are presented. While I'm not convinced artificial sweeteners are the answer (I lean towards a "less processing, fewer chemicals" approach to food), I also don't understand how they can present such a study as anything more than a relationship. Using artificial sweeteners is more likely to be a symptom/result of excess weight, not a cause.

It's like saying recently married women are more likely to have a giant white dress in their house. Obviously, that means that if you want to get married, you need to go buy a giant white dress.

Or people with babies use more diapers than people without babies. Therefore, buying diapers causes people to have children. (Let's face it, if you're pregnant, not buying diapers isn't going to be the answer!)

While I tend to be of the mindset that we're better off eating real foods, and indulging in the occasional small portion of a real-fat-real-sugar treat, I don't think that artificial sweeteners are directly to blame. And it's frustrating that the media is continually looking for the "magic secret" that fits in a 30-second sound bite, instead of addressing the big picture (an apple will always be better for you than apple pie, made with or without splenda!)

Wannabeskinny
06-17-2009, 10:03 AM
I'm no scientist but I have explored this field extensively... well by that I mean that I was a regular soda drinker for years and years. I was overweight, probably 185 when I decided it would be better for me to start using equal in my coffee and switched to diet soft drinks as well. I thought this would cut major calories out of my day, possibly up to 500-600! What actually happened though is that I shot up to 215lbs in 2 years. For some reason in made me hungrier and I definitely craved a lot of sweets eventhough I had never had a sweet tooth before. Obviously I can't explain this scientifically but I avoid all artificial sweeteners now. I finally felt the cobwebs clearing away after I stopped artificial sweeteners, even my mind feels better.

kaplods
06-17-2009, 11:27 AM
.... I also don't understand how they can present such a study as anything more than a relationship. Using artificial sweeteners is more likely to be a symptom/result of excess weight, not a cause...

It's like saying recently married women are more likely to have a giant white dress in their house. Obviously, that means that if you want to get married, you need to go buy a giant white dress.

Or people with babies use more diapers than people without babies. Therefore, buying diapers causes people to have children. (Let's face it, if you're pregnant, not buying diapers isn't going to be the answer!)


Usually it's not the researchers that present the study this way, it's people in the common media, with no more education on the subject that the average person (which means slim to none). Even with the training and education (an undergrad degree in behavioral psych and a graduate degree in developmental psych and statistic classes at both levels), I sometimes find myself thinking down these paths myself. Because correlation does not mean causation, however correlational data does bring up questions about causation. That is, the appropriate next step is to consider the possible explanations for the correlation and then do more studies to test them.

Tobacco research for example is mostly correlational, done many ways to give a support to a theory of causation, giving strong evidence that people who are prone to cancer aren't more likely to seek out cigarrettes.

I think it's the fact that causation would generally show a correlation that gets people stumped. If a causal factor would be reasonable, our minds tend to go there. I think because it's an instinctive factor, a biological hardwiring to look for correlation. Operant and classical conditioning ARE correlational in nature. If bad (or good) things happen under a specific circumstance, the natural response, it's a natural tendency to try to avoid (or attempt to recreate or anticipate) the circumstances in the future.

Even pigeons (very tiny brains) will exhibit "superstitious" behaviors during intermittent reinforcement. Their food may be on a timer, but if they happen to be turning around or flapping their wings when the food comes down the shute, it's very likely that they will increase wing flapping and/or turning.

So it's natural for our minds to "go there," especially if the cause hypothesis makes any sense. With diapers and white dresses, it's so obvious you'd have to be pretty stupid or superstitious to fall for it - or would you? When it comes to many goals, but especially one that is emotionally charged such as getting married and starting a family (or losing weight, or making money), otherwise intelligent people are often willing to resort to superstitions, knowing there is no rational basis in fact, but "it's worth a chance."

How many people waste their time, money and effort on "too-good-to-be-true" products, programs and schemes on the very, very slim chance that it might bring them the answer to their prayers.

Consumers have to be educated and aware, and vigilant against superstition and the tendency to assume causal relationships.

Heather
06-17-2009, 11:41 AM
I agree that we're probably hard-wired to seek causal explanations. It's the best way to try to understand a dangerous world. If we think X causes Y and we want to avoid Y as an outcome, we'll avoid X. And if we want Y as an outcome, well, why not try X? It's our best shot sometimes about having control over things we otherwise can't control.

And then the psychology kicks in, the cognitive biases. For example, we use confirmation bias to further see causality that isn't there. That is, we remember the the times when X was followed by Y and don't remember the times when X was not followed by Y. This further bolsters our feeling that we are right about X causing Y, and that we have control over our world... even though we don't.

I think we see this in weight loss a lot. Yesterday I did X and the scale went down, X has just been reinforced as a behavior, so I'll do that again. But maybe one time I do X and the scale doesn't go down. Doesn't matter, if X has been reinforced enough times, I'll "ignore" this one time (perhaps not consciously) and continue to do X.

And all along, the real cause for the scale going down might be something else... maybe behaviors from 2 days ago that I've failed to consider! :lol:

TheWalrus
06-17-2009, 02:18 PM
I thought this article and the scientists did a good job -- as evidenced in the quotation I provided -- of sticking to the facts, showing correlation and not causation. Can you all tell me which language they used that made you think otherwise? (Disclaimer: This is not a challenge, just a question -- please feel free to ignore it.)

I did find these points interesting:

* "Artificial sweeteners activate sweet taste receptors in enteroendocrine cells, leading to the release of incretin, which is known to contribute to glucose absorption. Recent epidemiologic studies...showed an association between diet soda consumption and the development of obesity and metabolic syndrome." (I thought "development" in particular was interesting)

* " 'In our study, we were actually able to isolate what type of sweetener was used at a certain point in time, as we used food diaries, and not food questionnaires,' Mr. Gravenstein pointed out.

'When we first did this analysis, we found that people ate more fat before 1983, which is the year [of] a big increase in artificial sweetener consumption in the American population — it was actually when aspartame was approved and diet Coke was introduced,' he explained."

* "the data show that artificial sweetener users 'were not different than the prediabetics, ie, they had the same prevalence of prediabetes,' he said, adding that 'in our population, people who used artificial sweeteners were twice as likely to have diabetes, 8.8% compared to 4.4% for controls.' "

* And this, from a pediatric endrocinologist: " 'I diet screen all my patients, and they are not drinking enough milk. Usually, they replace the milk with something else, sweetened tea, Gatorade, etc, not just water.' "

Please note this statement:

" 'Also, it could be that artificial sweeteners are causing diabetes, or it could be that there is a higher use of them because a lot of physicians actually recommend people to use artificial sweeteners to prevent diabetes....' Mr. Gravenstein said. The researchers are planning to address this question with a prospective analysis." (emphasis mine)

I just found it interesting to read how endocrinologists are thinking about artificial sweeteners today -- they are a very new substance that we are ingesting, and I'm curious as to what effect they have on the human body. We as a species have accepted them wholeheartedly into our diets, and now that we've been ingesting them for 26 years, it's intriguing to me to see what the effects might be...please note, in my opinion, neither the scientists nor the writers are stating that artificial sweeteners concretely cause obesity or diabetes. But I find their interpretation of the results of their examination of extremely detailed longitudinal data very interesting and look forward to following the course of their investigations.

Jacqui_D
06-17-2009, 02:54 PM
I happen to believe that there is a connection. I was just saying on another thread that I can't drink Diet Dr. Pepper because it triggers my appetite for high carb/simple carb/sweets like crazy! The craving is so strong that I will resort to eating bad foods eventually. I really think I'd be able to drink regular Dr. Pepper, only I don't like regular Dr. Pepper as well because I think it is almost syrupy sweet. That's probably why it doesn't make me want more sweets! I think aspartame just causes my brain to expect sweets/carbs and it's not satisfied until it gets them! It's just my personal opinion.

mandalinn82
06-17-2009, 03:00 PM
I think its a subtle sort of implication in the way the conclusions are phrased.

For example:

"People who are obese are more likely to use artificial sweeteners than people who are not obese"

suggests a different conclusion than

"People who use artificial sweeteners are more likely to be obese than people who do not use artificial sweeteners"

It's not that both conclusions don't say the same exact thing, and you're right, the authors never out and out say this is a causation type of thing. But just phrasing differently implies causality. The only neutral way to express a conclusion like this would be to say something like "Artificial Sweetener Use and Overweight are correlated. When one is present, the other is more likely to be present also". Which isn't a good headline!

I tend to jump on the causal vs. correlational difference, in any case, not just with your article (which again, other than subtleties that are hard to avoid without sounding ridiculously academic, didn't make any spurious conclusions).

Heather
06-17-2009, 04:05 PM
I can't access the article without registering and I don't want to do that right now, but let's start with some general info.

First, to me, causal language happens when someone describes the relationship between 2 variables in such a way that one is shown to, well, be the cause of the other. Usually this is some sort of active language. For example, the headline: "Sexual lyrics prompt teens to have sex" is causal, I would argue, because the use of the word "prompt" suggest that the lyrics are causing sex to happen. Other "active" verbs that often mean causation: X improves Y, X increases Y, X enhances Y , X affects Y, X causes Y,etc.

Correlational language, however, merely describes the relationship between 2 variables, and shows that as one increases, so does the other (or as one increases, the other decreases, in the case of a negative correlation. So, a headline like: "Sexual lyrics associated with teen sex" is more correlational. It's as if you could see the scatterplot, or graph of the data. In this case, you could also reverse the variables and the meaning is the same: Teen sex associated with sexual lyrics. Other correlational language includes phrases like: tend to, more/less likely... as in "people who do X tend to Y" or "people who have X are more likely to have Y."

So I would argue that the examples Amanda gave above are both correlational, in terms of language (whether we insert causation when we read it is something else, but the language is correlational).

Another way you see implied causality is when people note a correlation, and then note how to fix it, by changing one of the variables. For example, if sexual lyrics are associated with teen sex, then there is a relationship, but lots of things could cause it. But if the reporter or researcher suggests that the way to reduce sex is to get kids to stop listening to sexual lyrics, then they are absolutely implying causality. [BTW, I picked this example, which I use in my classes, because it is an actual example of correlational research being described in a causal way].

So, in the case of artificial sweeteners, there IS a relationship, but we don't know if it's a causal one or not. But suggesting that people can lose weight by reducing intake of artificial sweeteners does suggest causality.

Heather
06-17-2009, 04:21 PM
As for the article itself, in the original post we have this excerpt:

"People who use artificial sweeteners are heavier, more likely to have diabetes, and more likely to be insulin-resistant compared with nonusers, according to data presented here during ENDO 2009, the 91st annual meeting of The Endocrine Society.

-- I would argue the section in bold is, appropriately correlational.

...

First author Kristofer S. Gravenstein, a postbaccalaureate researcher with the Clinical Research Branch at the National Institute of Aging (NIA), National Institutes of Health (NIH), said the association may reflect the increased use of artificial sweeteners by obese and/or diabetic study participants.

-- Here the researcher is suggesting a possible causal relationship. They may have other reasons than just this study for thinking there's a causal relationship. Experimental animal research, for example.

'This is a cross-section study,' Mr. Gravenstein told Medscape Diabetes & Endocrinology, 'so there are limitations — we cannot say that artificial sweetener use causes obesity, we can say it is associated with it.' "

-- And here he's back to carefully stating, again, that this is just a correlational relationship. (I'm not putting down correlational research, it's just that it can't definitively answer causal questions).


And here are the other parts excerpted:
* "Artificial sweeteners activate sweet taste receptors in enteroendocrine cells, leading to the release of incretin, which is known to contribute to glucose absorption. Recent epidemiologic studies...showed an association between diet soda consumption and the development of obesity and metabolic syndrome." (I thought "development" in particular was interesting)

-- This is interesting, because he's saying sweeteners cause outcomes: activate receptors, release incretin. But then can't say it causes obesity -- just an "association". But this is the kind of research that makes them think there may, in fact, be a causal connection. It's just very difficult to demonstrate directly.


* " 'In our study, we were actually able to isolate what type of sweetener was used at a certain point in time, as we used food diaries, and not food questionnaires,' Mr. Gravenstein pointed out.

-- That's certainly more reliable, but as it's not experimental research, can't answer a causal question.


'When we first did this analysis, we found that people ate more fat before 1983, which is the year [of] a big increase in artificial sweetener consumption in the American population — it was actually when aspartame was approved and diet Coke was introduced,' he explained."

-- Now nowhere here does he state causality directly, but he does seem to be using it to lead up to a causal argument. But lots of other factors have also happened in that time.


* "the data show that artificial sweetener users 'were not different than the prediabetics, ie, they had the same prevalence of prediabetes,' he said, adding that 'in our population, people who used artificial sweeteners were twice as likely to have diabetes, 8.8% compared to 4.4% for controls.' "

-- Just correlational language


" 'Also, it could be that artificial sweeteners are causing diabetes, or it could be that there is a higher use of them because a lot of physicians actually recommend people to use artificial sweeteners to prevent diabetes....' Mr. Gravenstein said. The researchers are planning to address this question with a prospective analysis."

-- So, here an alternate explanation is introduced and more research suggested!

So, from these bits, not a lot of inappropriate causality at all.

Out of curiosity, what was the headline?

TheWalrus
06-17-2009, 04:34 PM
"Use of Artificial Sweeteners Linked to 2-Fold Increase in Diabetes"

Heather
06-17-2009, 04:42 PM
"Linked to" A tricky one. In the past I've called it correlational. But I guess it depends on whether "linked" in this case is an active verb or not... I think it isn't necessarily so...