Over the course of the last several issues we have examined three nutritional tools that were created for the purpose of predicting how the consumption of various foods may affect ones energy levels throughout a typical day. Specifically, the ever-popular Glycemic Index (GI), the more practical Glycemic Load (GL) and perhaps the most useful of all, albeit the most obscure of all, Insulin Index (II) have all been previously discussed.
In summary, the GI, used to measure the rate at which blood glucose rises in response to 50 grams of a food item, clearly shows that the customary dichotomy of “simple” vs. “complex” carbohydrates is grossly inadequate. The GL multiplies the GI score by the number of grams of total carbohydrates in a practical serving size, thereby taking into account the actual volume of a given food item.
And finally, the II, although research is relatively scant, reveals the fallacies associated with the GI and assesses the rate at which insulin is directly affected by the given food. This month we will sink ravenous teeth into yet another potential resource for stabilizing unwanted energy fluctuations and maintaining desirable bodyweights. This resource is known as the Satiety Index (SI).
The Satiety Index
The Satiety Index is a means of measuring the degree of fullness that a given food item provides following its consumption. The idea was originally conceptualized by Dr. Susan B. Holt, a nutritionist working at the University of Sydney in Australia. In short, the methodology employed by Dr. Holt and her team of associates involved feeding students a 240 calorie (1000 kjoules) portion of a given food. Following intake of the given food, students were inquired about their feelings of hunger every 15 minutes for a period of 2 hours.
These feelings were rated along a Lickert scale of 1-10, with 1 being “not hungry” and 10 being “very hungry”. At the end of the two hour interval, the students were permitted to eat ad libitum at a buffet while the team of researchers covertly observed and recorded the quantities of food each one consumed. The purpose was to see if the participants’ subjective ratings of hunger paralleled their actual eating behaviors following the two hour “fasting” period.
Indeed, the correlation was positive and significant; higher ratings of hunger were associated with higher consumptions of food at the buffet table. In regards to the food items tested, there were some noteworthy findings in the received scores.
In total, 38 food items were tested (the same items that were tested in the II studies), which had been grouped into 6 broad categories: Bakery Products, Snacks and Confectionary, Breakfast Cereals and Milk, Carbohydrate-Rich Foods, Protein-Rich Foods, and Fruits. (Please refer to the chart provided at the bottom of this article.) As was the case in the GI and the II discussed previously, white bread was used as a basis of comparison and assigned a value of 100.
Foods that received scores higher than 100 were determined to be proportionately more satisfying than white bread, whereas foods that received scores lower than 100 were determined to be less satisfying. So, for example, cheese and eggs, both receiving a value of 150, would be considered 1.5 times as satisfying as white bread. Ice cream, on the other hand, with a score of 96, would be somewhat less satisfying.
The study showed some provocative and head-turning results, especially when taking into account the GI’s and II’s of these same foods. Of paramount importance, it must be emphasized that despite the caveats concerning the flagrantly flawed assumptions on which it stands (the topic of the last issue), the GI nevertheless remains robust in predicting energy sways and insulin responses to the 38 foods tested for the II!!
In Other Words
Although the underlying assumptions of the GI were shown to be violated, and many foregone conclusions proved to be false, the GI and the II are nevertheless statistically very highly and positively correlated. Perhaps this fact alone suffices to allow the GI to save its battered face and retain integrity in the “trainer’s toolbox”. So is GI back in the house?? Well, that I guess is contingent upon the Trainer’s Perspective…but I digress here.
Returning now to the subject at hand, no significant correlation was shown to exist between the GI and the SI or between the II and the SI! Hence, the degree of satiety could not be reliably predicted by either the glucose response or the insulin response of the consumed food. This is very strange indeed; for it seems only natural that a food that induces an increase in blood sugar should also induce feelings of satiety. And yet the statistical analysis did not support this anticipated response. There is yet another conundrum with which to contend.
Despite the non-significant relationships existing between GI and II with the SI, carbohydrates were nevertheless found to be more satisfying than either protein or fat-laden food items. This flies in the face of other data that have shown protein and fat to be more satisfying than carbohydrates, due to the longer length of time required for their digestion. Holt reasons that fat is seen by the body as an emergency fuel, and it is therefore stored in cells instead of being used for immediate energy.
Consequently, the brain does not cut hunger signals, and the appetite rages on. She argues further that carbohydrates are the opposite, since they raise blood glucose levels. This conclusion, however, seems a non-sequitur, since the GI and SI were not found to correlate! So we find ourselves immersed in a hodgepodge that swallows up and corrodes our preconceived notions of how things are supposed to work.
And the sizzling flames of query only serve to thicken the stew as Holt goes on to note that, although most high-SI food items were satisfying for the full two-hours prior to the time of unrestrained eating, fruit and other high-carbohydrate items such as rice were not.
Rather, these items were satisfying for very short periods of time, probably due to the combination of high overall volume and the fast gastric emptying (since fruit is mostly water, fiber, and sugar). This leads one to believe that all three macronutrients have definite roles in the promotion of long-term satiety, but that the precise mechanisms by which each one operates are convoluted and elusive. So, is there anything that can be derived from this study that indicates or provides a small taste of the foods that really satisfy?
In a nutshell, water, fiber, and protein contents were all shown to correlate positively with feelings of satiety. Fat, on the other hand, was shown to have a significant negative correlation. So it appears as though sheer volume has a greater bearing upon satiety than does calorie density. Moreover, it appears that in the case of fat and refined high-caloric foods, appetite really does come while eating!
Let us now briefly examine several of the specific food items that have served to account for some of the equivocal statistical analyses emerging from Dr. Holt’s investigations. First of all, it is very surprising to note that potatoes, which score very high on the GI and the II, are by far the most satisfying of the foods tested. In fact, they are more than 3 times as satisfying as white bread (SI = 323). French fries, on the other hand, are much less satisfying (SI = 116).
The food with the lowest marker of satiability is the croissant. The SI of 47 indicates that a croissant is less than half as satisfying as the same amount of white bread! Of the groups tested, Fruits proved to be most satisfying, with an average SI of 170. However, it should be stressed that only four fruits were tested (oranges, apples, grapes, bananas) and that there was great variability in the scores among these items.
Within this group, the SI’s correlated very well with the GI’s—in a negative direction. In other words, those items having the lower GI’s (apples and oranges) were much more satisfying than those with the higher GI’s (grapes and especially bananas). This is in stark contrast to the “potato phenomenon” described above, where the high-GI potato elicits a high SI response. (This is an example of a positive correlation.) The SI’s of the high-protein sources are also somewhat enigmatic.
Note that fish is substantially more satisfying than beef, eggs, or cheese. Assuming that the protein-to-fat ratios in these sources are similar, one might speculate that the particular type of fat found in the sources may be of relevance. It could be that many of the benefits ascribed to the Omega-3 fatty acids found in fish spring from the very fact that these essential fatty acids assist in producing feelings of satiety.
The fact that yogurt (SI = 88) does not sufficiently satisfy is somewhat disconcerting, given its high water and protein content. It is another prime example of the apparent chaotic relationship between the GI (upon which it registers low), II (it registers high), and SI (it registers low). Other observations that deem highlighting are that cornflakes, white pasta, and jelly beans all have equivalent SI values.
However, Dr. Holt mentions that jellybeans (a very high-GI food) in fact were not as satisfying, although the students tended to munch afterwards as if they had been. She conjectures that the jelly beans may have induced a feeling of nausea that prevented the participants from fully indulging their appetites. In the Breakfast Cereal and Milk group, it was discovered that porridge (oatmeal) is more than twice as filling as Muesli.
Looking on at the Carbohydrate-Rich Food group, one observes that whole grain breads are 50% more filling than white bread. and Confectionary category, popcorn ranks twice that of either a candy bar or peanuts. And finally, very surprisingly, among the Protein-Rich Food items, lentils do not fair as well as baked beans; they in fact show a rather modest SI value.
Whew! Alright Physio-Feasters, let us now briefly recap some of the highlights that Dr. Holt heaped upon our plates. It is time to step back from the table, pat our rounded bellies, and digest the sustenance upon which we have just savagely overindulged!
- The Satiety Index (SI) is a means of measuring the degree of satiety experienced following the consumption of 240 calories of a given food item.
- The Glycemic Index (GI) is positively correlated with the Insulin Index (II).
- The GI is not correlated with the SI.
- The II is not correlated with the SI.
- The most satisfying of the 6 groups on average was the Fruits group.
- The least satisfying group was the Bakery Products group.
- Much variability exists amongst the SI’s of the food items within any given group.
- Short-term satiety appears to be most influenced by the overall volume of the food item.
- Water, fiber, and protein are all positively associated with the SI.
- Fat is negatively associated with the SI.
- Carbohydrates, when consumed without fat and / or protein, are filling, but only for a short time.
- Potatoes are more than three times as filling as white bread.
- Croissants are less than half as filling as plain white bread.
Okay! Deep breath! Hopefully by now it is strikingly clear that the mechanisms that govern appetite and satiety are nearly incomprehensively complex. However, several research projects directed at unearthing these mechanisms have been undertaken over the years, and some valuable insights have emerged from the sludge. So even though this discussion has been a mere appetizer before the feast, this science scavenger can’t possibly eat another bite!! Until next time—CHOW!!
For questions or general feedback on this article please feel free to contact me at email: firstname.lastname@example.org.
The Satiety Index
Each of the following foods is rated by how much food people ate after consuming them to satisfy their hunger.
All are compared to white bread, ranked as "100"