The Relative Satiety Value of Candy Bars in American Children

Objective: Sensory specific satiety has been suggested to be a contributor to the regulation of appetite. Based on 13 Australian adults, Holt, in 1995, proposed an index of degree of satiety induced by isocaloric consumption of culturally specific foods. The objective of this study was to determine if this satiety index would also apply to American children using hedonically preferred confectionaries. 

Methods: Twenty-four 12 or 13 year-old seventh graders, self-rated their degree of satiety in response to ingestion of eight different isocaloric (95 +/- 5 Kcal) hedonically positive candies as compared to white bread (Wonder Bread). Confections evaluated were: Tootsie Roll Midgets, Twix, M&M’s, M&M’s Peanut, Smarties, Star Bursts, Hershey’s Kisses, and Gummy Bears. Ratings were performed immediately prior to and after, and at 15 and 30 minutes post ingestion. Using the area under the curve, the satiety index was calculated, and adjusted so that white bread equaled 100%. 

Results: Despite being isoenergetic, the degree of a food’s satiety, as shown by the satiety index, demonstrated substantial variation. Confectionaries were ranked based on their satiety as compared to white bread (100). Of the candies tested, Star Bursts demonstrated the greatest satiety index (135) being 40% more that the candy showing the lowest satiety index, Tootsie Rolls (179). On casual inspection, these candies seem very similar. Tootsie Rolls and Star Bursts both are individually packaged with multiple packages per portion, and are grossly similar in terms of texture, chewiness, and sweetness. However, Star Bursts, unlike Tootsie Rolls, possess different flavors with each portion, and thus, a more diverse chemosensory experience. This suggests that sensory induced satiety may be an important element in production of the satiety index. 

Conclusion: This study validated and extended previous studies on Australian adults and demonstrated that the satiety index could be assessed in American children.

HYPOTHESIS

The goal of this study is to determine if the satiety index, a method of ranking the satiety value of foods, as validated in Australian adults would also apply to American children, using hedonically preferred confections. It is hypothesized that such validation will be found. 

INTRODUCTION  

Obesity is epidemic in our society (Flum, Khan, Dellinger, 2007). This is believed to be partly due to an excess consumption of food (McTigue, Harris, Hempill, et al, 2003). One of the mechanisms for control of such consumption is hunger, and its opposite, satiety, or the sense of fullness (Kruse, 2001). Approaches which influence these may be useful in helping reduce food intake.  

Myriad factors have been postulated to impact satiety of both external and internal origin. External factors in food include bulk and volume, degree of flavoring, visual presentation, and relative fat, protein, and sugar content (Cabanac, 1990).  Internal mechanisms include the psychological, hormonal, physical, and chemosensory spheres. In the psychological realm, anywhere from the number of people one eats with, the duration of the meal, whether the television is playing during the meal, or even the belief that the food is of greater nutritional value, all affect the satiety level (DeCastro, 2000, Rozin, 1990, Lowe, Bocarsley, DelParigi, 2008, Meiselman 2006). Hormones, including gherlin, and leptin similarly can affect satiety (Teff & Kapadia, 2008). Ingestion of food physically expands the stomach, causing stretch receptors to fire, signaling a sense of fullness (Peters & Mela, 2008). This effect of volume of food inducing satiety has been labeled, aliesthesia (Bartoshuk, 1990). The chemosensory system, through smell and taste, regulates feelings of hunger through the process of sensory specific satiety (Rolls, 1993). All of these work together in regulating perception of fullness. Thus, satiety is not solely based on the number of calories consumed, but rather, on more complex internal and external factors.  

Regardless of the underlying physiological mechanisms, assessment of degree of satiety produced through consumption of different foods has been proposed (Holt, Brank-Miller, Petocz, et al., 1995). In this study, isocaloric foods were categorized based on degree of satiety compared to a standard of white bread in 13 adults in Australia. No similar classification has been established in foods preferred and commonly consumed by children. 

Using satiety index of Holt as a model (Holt, Brank-Miller, Petocz, et al., 1995), eight different types of candy were chosen for examination. These were selected based on authors’ hedonic preferences, availability at the local drug store chain, and ease of distribution in isocaloric packaging.  

All candies were segregated for disposition in approximately isocaloric amounts of 95 +/- 5  

Calories as follows:

                                                                                    NUMBER OF

White bread (Wonder Bread, Small), two slices, with a total of 95 calories was also chosen for a baseline value, as per Holt study (Holt, Brank-Miller, Petocz, et al., 1995).

Study subjects were recruited from the seventh grade science classes of Ms. Duncan at Daniel Wright Junior High School, Lincolnshire, Illinois. Selection of criteria of volunteers was based on absence of allergy to nuts, no restriction with regard to eating of candy, and signed parental and child consent. Ten and 14 subjects were selected from each class. All students were age 12 or 13. Forty-two percent (10) were boys and 58% (14) were girls. None experienced negative hedonics toward the foods tested. No one ate anything for one hour prior to testing, or anything other than test material during the testing session. Each class was provided a different candy bar each day in a randomized order to avoid any influence of effect of order on presentation. Subjects, on a visual analog scale, rated their degree of hunger from extremely hungry to full. (See Illustration 1)

ILLUSTRATION 1

This was completed immediately prior to, immediately after, and 15 and 30 minutes after eating, as signaled by the principal investigator. Subjects were required to complete consumption of candy bar within two minutes of initiation. Forms were filled out independently and without sharing of information. Data was collected, collated, and statistically analyzed similar to that of the methods of Holt (Holt, Brank-Miller, Petocz, et al., 1995). In that study, prolonged degree of satiety and amount consumed in snacks and at the next meal correlated with self-rating of satiety at 15 and 30 minutes. Degree of satiety was calculated by determining the area under the curve, the AUC (see Illustration 2).

ILLUSTRATION 2

Example: Area Under Curve 

Integer values from one to seven were provided for each self-assessment hunger/satiety level, one being extremely hungry, and seven being full.

The satiety index was calculated by determining the 15 and 30 minute post-prandial satiety level, adjusting for the initial level of hunger (i.e., if initially full, even consumption of more food can not have an even greater effect on fullness, whereas if initially extremely hungry, a smaller amount of food may have a greater effect).  

This was accomplished through the following calculation:

Average AUC – (2 x average pre-prandial satiety level)   x 100

2 x subject size

This data was normalized (by multiplying by 1.15) to set white bread satiety index value equal to Holt’s standard 100% (Holt, Brank-Miller, Petocz, et al., 1995). 

With white bread as a reference of 100%, 100kcal of different candies demonstrated substantial differences in their satiating capacities, despite isoenergetic servings (See Table 2). 

Data arranged in order from smallest to largest, with all values round to two decimal places.

Starbursts and Gummy Bears were approximately a third more satiating than white bread. Star Bursts were 40% more satiating than Tootsie Rolls. 

This study extends Holt’s satiety index in regards to candy bars (Holt, Brank-Miller, Petocz, et al., 1995). In that study, she found the average satiety index score for snacks and confectionary was 100; with Mars Bars being 70 and Jelly Beans 118. Those results were very close to what our findings were with similar candies: Twix 95.8, and Star Bursts, 135. Furthermore, this was confirmed in a subject size twice that of Holt’s (Holt, Brank-Miller, Petocz, et al., 1995). Our findings further extend the concept of satiety index from Australian adults to American children and refines it in reference to specific American confectionaries. 

Limitations of findings:

Findings have to be considered with reference to subjects’ selected and food assessed. While subject size was double that of the published reference (Holt, Brank-Miller, Petocz, et al., 1995), a larger sample size would have been preferred. Subjects were all from suburban Chicago and thus were not geographically or demographically diverse. Different results may have been found in other geographic regions in the US. Likewise, different age groupings of children may present with alternative satiety index scores. Like so many other physiological parameters which change throughout childhood and adolescence, the satiety value of foods consumed by 5 year olds or 17 year olds may not be identical when consumed by 12 and 13 year olds.  

The food chosen to evaluate may have influenced the results. Non-confectionary foods were not evaluated, and this may have provided additional important satiety index values. The total calorie provided in our study was approximately one half of that of Holt (Holt, Brank-Miller, Petocz, et al., 1995). We don’t believe this had any substantial influence since the satiety index we found closely resembled the confectionary satiety index of Holt’s. (See Illustration 2A and Illustration 2B) Part of the rationale for choosing the candies in the study was the author’s hedonic preference. It is unknown to what degree hedonics influence satiety index. This deserves further exploration. 

Experimental error:

There are several sources of possible experimental error. Experiment design mandated a period of fasting prior to initiation of study. While not at lunch, it is possible students secretively may have eaten during these food-free class periods. If this happened, initial hunger may have been less, and a ceiling effect may have occurred and thus, true candy-induced satiety may have exceeded what was actually recorded. The Hawthorne Effect may have also skewed data. The Hawthorne Effect is whereby the mere act of observation and recording effects results (Rosenthal, 1996). Forcing subjects to record their level of satiety, may in and of itself, have influenced their levels of satiety. While this may have occurred, a similar effect would have been anticipated to be seen in the earlier published study of adults (Holt, Brank-Miller, Petocz, et al., 1995). 

Hedonically positive popular isoenergic confections have a heirarical influence on long-term satiety as determined by the satiety index. The reason for this is unknown, but may have to do with chemosensory influence (Rolls, 2005, Rolls, Rolls, Rowe, et al., 1981). For instance, Star Bursts have a 40% greater satiety value than Tootsie Rolls. Yet, on casual inspection, these candies seem very similar. Tootsie Rolls and Star Bursts both are individually packaged with multiple packages per portion, and are grossly similar in terms of texture, chewiness, and sweetness. However, Star Bursts, unlike Tootsie Rolls, possess different flavors with each portion, and thus, a more diverse chemosensory experience. This suggests that sensory induced satiety may be an important element in production of the satiety index. This can be investigated further by testing other foods with similar calorie and varying sensory characteristics. Moreover, expanding the procedure to a greater variety of foods may allow for the development of a “children’s satiety index” which ultimately, may have utility as part of a child’s weight control program. In light of the current state of childhood obesity, such further exploration is warranted. 

Bartoshuk, L.M. (1990) Distinctions between taste and smell relevant to the role of experience. In: Capaldi, E.D., Powley, T.L. (Eds.). Taste, experience, and feeding. Washington, D.C.: American Psychological Association, pp. 62-72.  

Cabanac, M. (1990) Taste: the maximization of multidimensional pleasure. In: Capaldi, E.D., Powley, T.L. (Eds.). Taste, experience, and feeding. Washington, D.C.: American Psychological Association, pp. 28-42.

DeCastro, J.M. (2000) Eating behavior: lessons from the real world of humans. Ingestive Behavior and Obesity, 16, pp. 800-813. 

Flum, D.R., Khan, T.V., Dellinger, E.P. (2007) Toward the rational and equitable use of bariatric surgery. JAMA,  298, pp. 1442-1444. 

Holt, S.H.A., Brand-Miller, J.C., Petocz, P., et al. (1995) A satiety index of common foods. European Journal of Clinical Nutrition, 49, 675-690.  

Kruse, H.P. (2001) Can satiety be measured? Nahrung/Food, 45, pp. 298-301.  

Lowe, M.R., Bocarsley, M.E., DelParigi, A. (2008) Human eating motivation in times of plenty: biological, environmental, and psychosocial influences. In: Harris, R.B.S., Mattes, R.D. (Eds.). Appetite and food intake: behavioral and psychological considerations. Boca Raton: CRC Press, pp. 95-112.  

McTigue, K.M., Harris, R., Hemphill, B., et al. (2003) Screening and interventions for obesity in adults: summary of the evidence for the US Preventative Services Task Force. Ann Intern Med, 139, pp. 933-949.  

Meiselman, H.L. (2006) The role of context in food choice, food acceptance and food consumption. In: Shepherd, R., Raats M. (Eds.). The psychology of food choice. Cambridge: CABI, pp. 179-199. 

Peters, H.P.F., Mela, D.J. (2008) The role of the gastrointestinal tract in satiation, satiety, and food intake: evidence from research in humans. In: Harris, R.B.S., Mattes, R.D. (Eds.). Appetite and food intake: behavioral and psychological considerations. Boca Raton: CRC Press, pp. 187-211.  

Rolls, B.J., Rolls, E.T., Rowe, E.T., et al. (1981) Sensory specific satiety in man. Physiol Behav, 27, pp. 137-142 

Rolls, E.T. (1993) The neural control of feeding in primates. In: Booth, D.A. (Ed.). Neurophysiology of ingestion. Tarrytown, NY: Pergamon Press, Inc., pp. 137-169.  

Rolls, E.T. (2005) Taste. olfactory, and food texture processing in the brain, and the control of food intake. Physiol Behav, 85, pp. 45-56. 

Rosenthal, R. (1996) Experimenter effects in behavioral research. New York: Appleton-Centruy-Crofts, p. 464.  

Rozin, P. (1990) The importance of social factors in understanding the acquisition of food habits. In: Capaldi, E.D., Powley, T.L. (Eds.). Taste, experience, and feeding. Washington, D.C.: American Psychological Association, pp. 255-269.  

Teff, K.L., Kapadia C. (2008) Role of postabsorptive endocrine factors on human feeding and regulation of body adiposity. In: Harris, R.B.S., Mattes, R.D. (Eds.). Appetite and food intake: behavioral and psychological considerations. Boca Raton: CRC Press, pp. 235-252.