Smell and Obesity

Here is the NY Times, describing the latest weight-loss fad:

Like almost every dieter in America, Wendy Bassett has used all sorts of weight-loss products. Nothing worked, she said, until she tried Sensa: granules she scatters on almost everything she eats, and which are supposed to make dieters less hungry by enhancing the smell and taste of food.

The maker of Sensa claims that its effectiveness is largely related to smell: the heightened scent and flavor of food that has been sprinkled with Sensa stimulate the olfactory bulb -- the organ that transmits smell from the nose to the brain -- to signal the "satiety center" of the hypothalamus. Hormones that suppress appetite are then released.

Unfortunately, I think Sensa's approach is exactly backwards. Delicious smells are what psychologists refer to as a "hot stimulus," a sensory provocation that makes us want to eat. I described this process in my recent New Yorker article on Walter Mischel and the marshmallow task, a simple experiment in which four-year olds attempt to not eat the tempting treat right in front of them. The question, of course, is why only some children are able to resist, while others gobble down the marshmallows (or oreo cookies) in less than thirty seconds:

Mischel's conclusion, based on hundreds of hours of observation, was that the crucial skill was the "strategic allocation of attention." Instead of getting obsessed with the marshmallow--the "hot stimulus"--the patient children distracted themselves by covering their eyes, pretending to play hide-and-seek underneath the desk, or singing songs from "Sesame Street." Their desire wasn't defeated--it was merely forgotten. "If you're thinking about the marshmallow and how delicious it is, then you're going to eat it," Mischel says. "The key is to avoid thinking about it in the first place."

The problem with Sensa is that it makes it harder to not think about food - we can't strategically allocate our attention because those delicious smells keep on reminding us how hungry we are. If anybody loses weight on Sensa, I'm betting that it's placebo effect at work, and not some hypothalamic pathway.

This nasal product, though, seems much more promising.

More like this

Seth Roberts has an interesting theory about this. We know that the body has a set-point for weight and the body eventually bounces back to this set-point. Roberts accidentally found a way to hack the set-point and has a quite convincing theory about the big picture. His diet is called Shangri-la Diet.

It is described here:

I'll try to summarize the idea here. So the idea is that we maintain this flavor-calorie association. After we've eaten a particular taste a few times, our brains automatically learns to associate the taste with the calories it gives. A simple example: an adult who has never tasted Coca-Cola will find the taste awkward, but will learn to like it really quickly, because of the calorie association.

Now, the hypothesis is that the set-point is adjusted by "estimating" how much food is available. When there is not a lot of food available (tastes are not rich), body drops the set-point to live off the body fat already accumulated. In contrast, when there is plenty of calorie-rich food, it is time to store the food and thus body adjusts the set-point higher and generates more hunger.

This is why modern pre-packaged and standardized food leads so easily to obesity: it develops such a strong flavor-calorie association, because the flavor stays constant and the food is calorie-rich. And this is also why most diets do work for a while. When a person starts a diet, he will typically experience new flavors that do not have such tight association with calories. But once the connection is made, the diet no longer works.

But what Roberts accidentally found is that consuming completely flavorless calories lowers the set-point. Consuming flavorless [1] olive oil, for example, in a way simulates this sort of "flavorless non-rich food times" and body lowers the set-point. The set-point is not lowered by that much, because you do eat normal food. But because the set-point is lowered, losing weight becomes pretty much automatic. The body generates just enough hunger so that the set-point is reached. (And this is of course why normal dieting is so hard: it's almost impossible to fight against the set-point.)

All that said, the method doesn't seem to work for everyone. The reason for this is not known. Also, it hasn't been tested with double-blind placebo-controlled studies.

Coming back to the "Sensa" method, I think Roberts's theory says that it shouldn't work. It should probably make it harder to lose weight. It might have this initial effect of introducing new flavors to the people using it, so they wouldn't have such strong flavor-calorie associations at start.

[1] Flavorless means that the oil itself doesn't have flavor, but also that it is consumed as such, not with other food. In fact, it is recommended to not eat anything with flavor at least an hour before and after consuming the oil.

By Jarno Virtanen (not verified) on 19 Jun 2009 #permalink

What Mischel calls "strategic allocation of attention" is the psychological defense mechanism called suppression. The ability to use suppression at an early age to inhibit impulsiveness towards a smaller sooner reward (SSR) and wait for a larger later reward (LLR) is a powerful predictor of success later in life. Howard Rachlin (2000) in his book "The Science of Self-Control" presents an elegant example of people waiting for a bus on a cold night in New York and how their delay-of-gratification is controlled by the steepness of their hyperbolic temporal discounting (HTD) curves. Ability to suppress SSR for LLR indicates well developed top-down control by the Prefrontal Cortex to flatten HTD curves.

Mischel says. "The key is to avoid thinking about it in the first place."

This is not as easy as it sounds. I have literally stayed awake all night thinking about buying and consuming food the next day. I've had therapists suggest distraction -- play a game, read a book -- to counteract the cravings for food. No go.

Even now when I'm not doing too bad on wellbutrin, I still think about eating. At least now I am able to channel and think about small portions at appropriate times instead of large portions whenever I am alone.

For some of us, it is just very hard to silence that little voice in our heads that says eat a lot when no one is looking or when you are home alone.

Is it possible that two different tasks are being described here? In the marshmallow example, children are delaying instant gratification so they can get more later. In the Sensa example, Wendy is not so much delaying gratification as she is moderating her food intake. Maybe a heightened sense of smell allows her to be more aware of her food, and therefore control her intake because she is paying more attention to it. This goal seems different to me than trying to maximize a reward or payout.

Sensa seems to be the exact opposite of an appetite suppressant. I've never tried Sensa of course, but I (like almost everyone) have had head colds that suppressed my sense of smell (& taste). I found that not being able to smell and taste my food is one of the most powerful appetite suppressants I've ever known.

I described this process in my recent New Yorker article on Walter Mischel and the marshmallow task, a simple experiment in which four-year olds attempt to not eat the tempting treat right in front of them.

AHAHAHAHAHAHAH! Try that task on me with Jameson!

That's a really interesting theory, Jarno. I have a personal anecdote about the topic.

About six months ago, mostly due to boredom with food preparation, I started taking swigs of olive oil instead of snacking on something else (usually one before bed, and maybe another one earlier in my day). It was maybe 400 calories a day. Over the course of about 3 months I lost 15 pounds. It was odd, since I wasn't trying to lose weight, and I wasn't overweight to start with. Normally my weight is very stable. Calculating it after the fact, I went from a BMI of 24.7 to 22.5, and have stayed there.

At the time I just thought there might be some special weight-loss property of olive oil. But I also noticed I didn't seem to eat as much as I used to, and I didn't feel any hungrier. There might be something to Seth Roberts' theory. Thanks for mentioning it.

By Ben Seeley (not verified) on 19 Jun 2009 #permalink

One of the things that has been observed in some obese and bulimics is an almost unconscious eating. Oh, I'll have a couple of bites, then a few more, then a few more, until it's oh, well, I should just finish it off. The motions and the taste become almost unconscious. The attention goes elsewhere and the spoon keeps going to the mouth. I know that I have been surprised when I actually paid attention to saying to myself "I can't believe I ate the whole thing".

What the sensa might do is raise the consciousness of the act of eating so that there is a more of a consciousness of satiety.

By natural cynic (not verified) on 19 Jun 2009 #permalink

Ben Seeley... I agree with you about the olive oil, the taste of olive oil lingers in a satisfying way; there are also some lip balms that act the same way.

By Elizabeth (not verified) on 20 Jun 2009 #permalink

RE: Natural Cynic, #8

This. I can tell you from experience that when I finally give in to the constant thoughts and buy a bag of Doritos, I hide the bag and wait until I'm alone before I eat it. I only actually taste the first 4 or 5 chips. After that, my mouth almost becomes numb or overloaded from the chips[I normally eat *very* bland food]. But that doesn't stop me from eating the whole bag.

It really is a kind of automatic eating. I'm reading and eating and 15 minutes later[no joke] the bag is empty. There's almost a drive to finish the bag. I have a hard time stopping. In fact, the only way I can stop before I eat the whole bag is to crumple it all up and dump the few remaining chips in the trash. I'm just glad I'm not one of the people who would actually dig the crumbled chips out of the trash.

Flavorless means that the oil itself doesn't have flavor ...

Sheer nonsense. Olive oil definitely has flavor in and of itself.

By llewelllly (not verified) on 20 Jun 2009 #permalink

"Sheer nonsense. Olive oil definitely has flavor in and of itself."

I didn't word it carefully enough. I meant that for the oil to have an effect like this, it must not have any flavor. Typical olive oil does have a flavor, but some olive oil products like extra-light olive oils do have none or very little flavor.

By Jarno Virtanen (not verified) on 20 Jun 2009 #permalink

Huh, lip balms... Thanks for another interesting idea, Elizabeth :).

By Ben Seeley (not verified) on 22 Jun 2009 #permalink

Regarding the flavor of oils, the olive oil I use has some flavor (same goes for my coconut oil), but the flavor to calories ratio is still quite low.

By Ben Seeley (not verified) on 22 Jun 2009 #permalink

interesting theory. But how do you reconcile that with what happens in the absence of olfactory input altogether- anosmia.

According to this paper (

"These authors report that ~14% of anosmic patients experienced a body weight gain exceeding 10% while ~6.5% experienced a loss of at least that amount."

It's obviously not too terribly cut and dry.

smell- the final sensory frontier.

By Jordan Donohue… (not verified) on 27 Jun 2009 #permalink

Thanks for the info, Jordan.

By Ben Seeley (not verified) on 29 Jun 2009 #permalink

ET Rolls has written a few papers about the integration of vision, smell, taste, and texture. If I recall correctly, vision, smell taste and texture combine to form a multimodal sensory experience of food, and this multimodal sensory representation is linked to the hypothalamus, where leptin signaling occurs, and to the amygdala. Thus, turning up the smell of food increases the intensity of the multimodal sensory representation that drives leptin signaling. Here is a link to a paper by ET Rolls on smell and satiety: .

I came across sensa recently, while I was doing research to file a patent on treating obesity by altering vision that is in a similar vein to both Sensa and mirror therapy for phantom limb pain. By one of various means, it's possible to create illusions while eating such that the perceived ratio between hand size and food size is altered such that the food appears bigger. Thus, a person can drink 8oz of soda, but the multimodal sensory representation the brain makes will be based on a visual illusion that 24 ozs are being drunk.

Anyone with a webcam and a shotglass can experience the dramatic difference paying attention to a magnified drink image and unmagnified hand has on how much liquid it feels like you are drinking.

In packaging and marketing food, the ratio between hand size and food size is severely distorted by cartoon characters drawn around the food, and by cutting between close ups of food and longer shots of people on TV ads. These changes have occurred over the last 100 or so years. In all cases prior to that, since hands and eyes evolved, the ratio between hand size and food size in learning about eating and eating was consistent. Now, in two or three generations, the relationship between hand size and food size differs dramatically across learning and eating.

The changes in hand size/food size between learning and eating appear similar to the ebbinghaus illusion. We are learning about food from visuals that create a magnified image of food, but eating food in a minimized condition. This is similar to the recent study that showed that arthritic pain was reduced when the hand was exercised in a minimized condition. We know that drawings of hands can be confused with our actual hands in the phantom hand illusion. We know that the ebbinghaus illusion fools both conscious and unconscious vision. We know that exposure to TV (where illusions are common) correlates with obesity. So I think it's both reasonably likely and currently untested that recreating the hand size/food size ratio in marketing by using illusions while eating will influence leptin signaling (and/or) satiety in response to a portion of food.

Now that I've filed a provisional patent for medical treatments based on this hypothesis, I am hoping to do studies on this. I think Sensa is very much on the right track. However, I think sight is a better target than smell for treating obesity, because hacking vision while eating recreates distortions in the images we learn from. It may very well be the case that the obesity epidemic is soluble to an intervention similar to reading glasses which balance sensory experience with earlier learning.

By Brian King (not verified) on 29 Jun 2009 #permalink

Wow, that is absolutely fascinating, Brian. Thank you for the information and perspective. Good luck with your research!

By Ben Seeley (not verified) on 19 Jul 2009 #permalink