That being said, my rationale has hitherto been based on the assumption that this accommodation effect will even occur. Today, a couple of months after I first issued my hypothesis, I have a study from the Texas Christian University to prove it.
In a series of experiments, Sarah E. Hill, Marjorie L. Prokosch, Amanda Morin and Christopher D. Rodeheffer examined the possibility that the consumption of artificially sweetened drinks and foods might produce unintended physiological, psychological, or behavioral changes that hinder, rather than help, consumers’ weight management goals. In doing so, they tested the impact of non-nutritive sweetener (NNS) consumption (via a diet soft drink) on food-related cognition, consumer choice, and subjective responses to sugar-sweetened snacks. In contrast to previous SuppVersity posts today's approach to the "Sweetener Conundrum" is thus a psychological, not a physiological one. One that is based on the measurement of the impact of nun-nutritive sweeteners on processes, of which we known that they will have an impact on food regulation over time. The purpose is thus to answer the all-important question whether NNS are a hazard to your weight loss / maintenance goals.
So what did the scientists do?
Now that we've successfully set the scene, let's take a look at the experimental approach Hill and her co-workers chose. As I previously hinted at, the study comprised three different experimental conditions:
- Experiment 1 -- The goal of our first experiment was to examine the impact of non-caloric sweeteners on the cognitive accessibility of food items with differing levels of caloric density.
The scientists expected that that – compared to participants who consumed the sugar-sweetened or unsweetened beverage – those who consumed the beverage sweetened with NNS would have shorter response-time latencies to the names of high-calorie, but not low-calorie, food items.
What's the significance? If the hypothesis was correct this would mean that people who consume NNS are more likely to respond to high-calorie food clues. Practically speaking, they would thus be more tempted to give in and buy the a pack of candy from a street seller in New York City, when all they'd actually want to do was to pass by and catch the next taxi.
- Experiment 2 -- The goal of the second experiment was to extend the results of experiment 1 by testing the effects of NNS-sweetened beverages on people’s decision-making in a consumer choice scenario.
In particular, the scientists were interested in whether NNS consumption influences the types of consumer products people choose when offered their choice of either a high-calorie food item (M&M chocolate candies) or one of two low- or zero-calorie, non-food items (a bottle of Ozarka spring water or a pack of Trident sugar-free gum).
As I pointed out in the "what's the significance" paragraph of experiment 1, the scientists expected that eh participants who consumed the NNS-sweetened beverage would choose to take the high-calorie candy at a greater frequency than would those who consumed the sugar-sweetened or unsweetened drinks.
What's the significance? If the hypothesis was correct and the subjects would be more likely to give in to the temptation of high-calorie candies, this would confirm that the abstract and practically irrelevant decrease in response-time latencies translates into a higher likelihood of giving in to sweet temptations.
- Experiment 3 -- The goal of our third and last experiment was to build on the results of experiment 1 and experiment 2 by examining whether consuming NNS disrupts individuals’ subjective responses to subsequently consumed sweetened foods.
As Hill et al. point out, "[m]uch research has found that NNS consumption does not influence the number of calories consumed in a subsequent meal or snack" (see Bellisle & Drewnowski, 2007 for a review). However, because consuming NNS disrupt the natural pairing of sweet taste and energy availability, consuming them may degrade the hedonic response to sucrose-sweetened food, impairing the body’s ability to regulate energy and body weight (Swithers &
Davidson. 2005 & 2008).
Thus, experiment 3 was designed to test the possibility that NNS would disrupt normal sweet perception experimentally by examining the impact of NNS-consumption on participants’ satisfaction with a subsequently-consumed, sugars-weetened snack.
What's the significance? If the subjects would in fact feel less satisfied after consuming a sweet snack, when they'd been consuming non-nutritively sweetened drinks before, this would increase the chance of overeating. In other words, the one piece of chocolate that would have been enough to satisfy the chocolate cravings of someone who abstains from NNS may well trigger a "fatal" chocolate binge in NNS junkies.
41 men). Participants’ ages ranged from 18 to 25 years (M=19.81, SD=3.27) and all received partial course credit in exchange for their participation.
No(!) difference between natural and artificial sweeteners: While you may argue that the different forms of sweeteners could have differential effects on physiological processes, the lowering of the sweetness threshold will depends - if on anything - on the degree of sweetness. Stevia would thus not be better than the (imho) falsely decried aspartame, of which I am getting tired to repeat that it is not toxic in the amounts you will find in foods and artificially sweetened beverages.The design of the study was a 3(drink condition: Sprite vs. Sprite Zero vs. mineral water, between subjects) × 3(word type: high calorie vs. low calorie vs. non-word, within subjects) mixed factorial design. During recruitment, all participants were instructed not to eat or drink anything other than water past midnight prior to their testing session. In order to minimize unwanted influences of the subject's individual eating / drinking habits, anyone who ate or drank anything other than water less than eight hours prior to their session was excluded from all analyses (18 excluded).
All testing sessions were conducted between 8:00 and 11:00a.m. After arriving in the laboratory, participants were given a participant ID number that was linked to their testing condition. Upon being seated, participants were given an unmarked, red plastic Solo cup that contained one of three 12 ounce (355 ml) beverages: (1) sugar-sweetened (Sprite), (2) non-calorically sweetened (Sprite Zero), or (3) unsweetened (Kroger brand lemon-lime sparkling mineral water) -- The nutritive value of the test drinks is depicted in Table 1.
"I see sweet high calorie feeds, wherever I look!"
Participants (all condition-blind) were given five minutes to consume their drink while watching images from the Hubble Telescope on their computer screens. After this time had elapsed, participants were asked to complete a lexical decision task (described below). The study closed with participants being asked to respond to a series of questions about themselves (e.g., sociodemographic questions, height, weight) and about their compliance with the experimental procedure. After the experiment was complete, a hypothesis-blind research assistant used the ID numbers to match participants’ computerized data with the drink condition to which they were assigned.
"After participants finished their beverages they completed a lexical decision task to measure the cognitive accessibility of high and low-calorie foods. During this task, participants were pre sented with 28 letter strings. These letter strings, presented in random order, consisted of seven high-calorie food words (e.g., burger, cookie, pizza), 7 low-calorie food words (e.g., celery, radish, carrot), and 14 non-words (e.g., ebusun, ganeap, tigne). Each letter string flashed on the screen for 250 ms and participants had to indicate whether each letter string was a word or non-word by pressing the “z” or “m” key, respectively. The response latencies (i.e., how long it took for participants to indicate if a string of letters was a word or non-word) served as our dependent variable, with lower response latencies indicating greater cognitive accessibility. To familiarize participants with this task, before completing the experimental trials, participants completed 30 practice trials consisting of 15 neutral words and 15 non-words."In a pre-test the scientists had made sure that all participants were aware of he high energy content and potential fattening effect of the items.
|Table 2: Descriptive statistics for total number and the reaction time (in milliseconds) of words categorized correctly during a lexical decision task. Longer reaction time latencies indicate lower cognitive accessibility (Hill. 2014)|
"Orthogonal planned comparisons revealed that participants who consumed the NNS sweetened beverage responded more quickly to the names of high calorie food items compared to those who consumed either the sugar-sweetened or unsweetened drink (p=.001, CI:−84.89,−20.11)."What's also interesting is that there were no differences were found between the sugar-sweetened drink and those who consumed the mineral water (p=.28). The effect is thus NNS and not "sweet"-specific!
"Sweets, I see you!" Question: Will you also consume thee?
In the follow up experiment, which had the exact same baseline design and was conducted with the exact same test drinks, the participants were instructed to take the lid off of an opaque (6”×10”) green IKEA box that was located on the far right of their partitioned computer space. Inside the box were three consumer products: (a) a 12 ounce bottle of Ozarka natural spring water, (b) a pack of Trident gum (18 sticks, original flavor), and (c) an 8 ounce bag of plain chocolate M&M candies.
"Participants were prompted by the computer to pick each product up one at a time (order was randomized via Qualtrics) and evaluate the product’s logo, packaging, and their familiarity with the product. At the end of the experiment, participants were told that they could choose one of the products to take with them when leaving. Participants were then thanked, debriefed, and dismissed." (Hill. 2014)After participants exited the lab, the research assistant inspected each of the boxes and made note of which product was chosen by each participant using their ID number.
|Table 3: Percent of participants in each drink condition who chose a specific food option to take home. Three participants took none of the items (Sprite:n=1, Sprite Zero:n=2 | Hill. 2014)|
A similar result was observed in the last and final experiment. After having realized that there is no sex-difference between male and female participants, the researchers conducted this last experiment, in the course of which their now exclusively female subjects were - once again! - randomly assigned to consume (1) sugar-sweetened (Sprite), (2) non-calorically sweetened (Sprite Zero), or (3) unsweetened (Kroger brand lemon-lime sparkling mineral water).
|Table 4: Descriptive statistics for participants’ food ratings and calorie consumption by drink condition (Hill. 2014).|
Artificially sweetened beverages do not trigger overeating in healthy females
Since the participants were told to eat at least one cookie, but that they could eat as many as they liked, the scientists were able to determine that the type of drink type did not influence how good/bad participants thought that the cookies tasted (p=.51). It did yet also reveal (and that comes as a positive surprise) that the different beverages did not influence how much of the product participants consumed, either (p=.65).
What the NNS consumption did do, though, was leaving the participants significantly less satisfied with what they had eaten compared to those who consumed either the sugar sweetened or unsweetened drink. Whether that would have made them grab another Oreo, piece of chocolate or whatever only minutes after the test, however cannot be determined on the basis of this study.
- Hill, Sarah E., et al. "The effect of non-caloric sweeteners on cognition, choice, and post-consumption satisfaction." Appetite (2014) | Accepted Manuscript.
- Wright, Katharine Mary. "Nonnutritive Sweetener and Weight Management: A Potential Paradox in Modern Dieting." UNF Theses and Dissertations (2014).