"Arts, Faculty of"@en . "Psychology, Department of"@en . "DSpace"@en . "UBCV"@en . "Assanand, Sunaina"@en . "2009-02-17T22:42:16Z"@en . "1996"@en . "Master of Arts - MA"@en . "University of British Columbia"@en . "Recent research on hunger and eating has shown (a) that among individuals\r\nwith ad libitum access to food, hunger and eating are not regulated by deviations\r\nin the body's energy resources from energy set-points, and (b) that it is healthier\r\nfor people to consume their daily caloric intake as several small snacks than as\r\nthree large meals. People's beliefs about hunger and eating were assessed in\r\ntwo questionnaire studies. In Study 1, a large sample of university\r\nundergraduates was surveyed; in Study 2, dietetics students, nursing students,\r\nmedical students, dietitians, nurses, and doctors were surveyed. Both studies\r\nrevealed that people's personal theories of hunger and eating were inconsistent\r\nwith recent research findings in ways that are likely to promote over consumption\r\nand ill health. These results suggest that educational programs designed to\r\nmodify the beliefs about hunger and eating of people suffering from problems of\r\nover consumption and of health professionals who treat problems of over\r\nconsumption may increase the effectiveness of current treatment regimens."@en . "https://circle.library.ubc.ca/rest/handle/2429/4735?expand=metadata"@en . "1325714 bytes"@en . "application/pdf"@en . "PERSONAL THEORIES OF HUNGER AND EATING by Sunaina Assanand B.Sc, University of British Columbia A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Department of Psychology We accept this thesis as conforming (PHE UNIVERSITY OF BRITISH COLUMBIA October, 1996 \u00C2\u00A9 Sunaina Assanand, 1996 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of 9 ^ A U ^ The University of British Columbia Vancouver, Canada Date f ^ W W \ \ J 1 L DE-6 (2788) Abstract Recent research on hunger and eating has shown (a) that among individuals with ad libitum a c c e s s to food, hunger and eating are not regulated by deviat ions in the body's energy resources from energy set-points, and (b) that it is healthier for people to consume their daily calor ic intake as severa l smal l snacks than as three large meals. Peop le ' s beliefs about hunger and eating were a s s e s s e d in two quest ionnaire studies. In Study 1, a large sample of university undergraduates was surveyed; in Study 2, dietetics students, nursing students, medical students, dietit ians, nurses, and doctors were surveyed. Both studies revealed that people 's personal theories of hunger and eat ing were inconsistent with recent research f indings in ways that are likely to promote over consumpt ion and ill health. T h e s e results suggest that educat ional programs des igned to modify the bel iefs about hunger and eating of people suffering from problems of over consumpt ion and of health professionals who treat problems of over consumpt ion may increase the ef fect iveness of current treatment regimens. Tab le of Contents Abstract ii Tab le of Contents iii List of Tab les v List of F igures vi Introduction 1 Ev idence Incompatible with Set-Point Theor ies of Hunger and Eat ing 2 Ev idence that the Feed ing Sys tem Does Not Mainta in Energy Rese rves at Homeostat ic Leve ls 2 Ev idence that Hunger and Eat ing are Not Normal ly Tr iggered by Dec l ines in Energy Resou rces 3 Ev idence that Increases in Energy Resou rces Do Not Rel iab ly R e d u c e Hunger and Eat ing 5 Summary 5 Contemporary Theor ies of Hunger and Eat ing 5 Posit ive-Incent ive Theory of Hunger and Eat ing 6 W o o d s ' Theory of Mea l Intolerance: Ev idence that S n a c k s are Healthier than Mea ls 8 Evolut ionary Theory of Hunger and Eat ing 9 Persona l Theor ies of Hunger and Eat ing 10 Study 1 11 Method 11 Part ic ipants 11 Procedure and M e a s u r e s 12 Resu l ts 12 iv 14 Study 2 Method 1 4 Participants 1 4 Procedure and Measures 1 5 Results 1 6 17 Discussion 22 Tables ^ 24 Figures References 2 6 29 Appendix V List of Tables 1. Mean Composite Set-Point Scores for Study 2 22 2. Mean Scores on Items Assessing Beliefs about Snacking and Meal Consumption for Study 2 23 vi List of Figures 1. Distribution of Composite Set-Point Scores in Study 1 24 2. Distribution of Composite Set-Point Scores in Study 2 25 1 Introduction In many industr ial ized countr ies, health problems assoc ia ted with overeat ing have reached ep idemic proportions. In the United States a lone, for example, at least 34 mill ion adults are at risk for developing d iabetes, hypertension, heart d i sease , and certain cancers because they are overweight (see Mil lar & Stephens, 1986). Unfortunately, al though efforts to treat overeat ing have met with some short-term success , long-term results general ly have been poor (Garner & Woo ley , 1991; Goodr ick & Foreyt, 1991; W i l son , 1 9 9 4 ) - i n one study, fewer than 3% of patients beginning a weight-reduct ion program maintained their weight losses for more than 4 years (Kramer, Jeffery, Forster, & Snel l 1989). The studies reported here are, to my knowledge, the first to approach the problems of overeat ing and weight reduction by trying to understand how people think about their own motivations to eat. The studies tested two hypotheses, both of which were der ived from a ser ies of pilot interviews: (a) that most people bel ieve that their eating behavior is regulated by deviat ions in their body's energy resources (e.g., b lood g lucose, body fat) from homeostat ic set points and (b) that most people bel ieve that it is healthier to consume their daily calor ic intake as three large meals than as severa l smal l snacks . Neither of these beliefs is consistent with the empir ical ev idence (see Bol les, 1980; Fr iedman & Strieker, 1976; Toates, 1981; W o o d s , 1991). Below, I review this contradictory ev idence, introduce new theoretical perspect ives on hunger and eating that are more compatible with it, and explain how misconceptions about hunger and eating may contribute to over consumption and ill health among people living in food-replete societies. Evidence Incompatible with Set-Point Theories of Hunger and Eating Set-point theories of hunger and eating make three predictions: (a) that the feeding system maintains the body's energy resources at homeostatic set points, that is, at stable levels that are optimal for health and survival, (b) that hunger and eating are promoted by decreases in the body's energy resources below their set points, and (c) that hunger and eating are inhibited by increases in the body's energy resources to their set points. None of these predictions, however, have been reliably confirmed. Evidence that the Feeding System Does Not Maintain Energy Reserves at Homeostatic Levels Evidence from studies of both humans and laboratory animals indicates that the primary function of the feeding system is riot to maintain the body's energy reserves at levels that are optimal for health and survival. Indeed, humans and other animals living with ad libitum access to palatable foods tend to eat far more than is optimal for their health and survival. This point has been made in two ways: (a) by documenting the high levels of consumption (e.g., 3 F idanza , 1980), obesity (e.g., Mil lar & Stephens, 1986), and health problems assoc ia ted with overeat ing (e.g., Bray, 1992) in populat ions living with ad libitum a c c e s s to palatable foods; and (b) by demonstrat ing the dramat ic improvements in health and longevity that result from reducing consumpt ion levels to well be low ad libitum levels (for reviews see Bucc i , 1992; Masoro , 1988; We ind ruch & Wal ford , 1992). The high heritability index of obesity--.44 in one study (Pr ice, N e s s , & Laskarzewsk i , 1 9 9 0 ) - h a s led some (e.g., Hewitt, Stunkard, Carro l l , S ims , & Turner, 1991) to suggest that the inheritance of high set points may be an important factor in overeat ing and obesity. The high inc idence of overeat ing and obesi ty in some populat ions renders this hypothesis untenable. If the inc idence of obesity were low, then the feeding systems of individuals who are o b e s e could be seen as pathological except ions to systems that have evo lved to maintain most peop le 's energy resources at optimal levels; however, a set-point system that results in obesity, ill health, and premature death in a substant ial portion of the populat ion is unlikely to have evolved. Ev idence that Hunger and Eat ing are Not Normally Tr iggered by Dec l ines in Energy Resou rces Ev idence , again from studies of both humans and laboratory animals, indicates that hunger and eating are not normally caused by d e c r e a s e s in the body's energy reserves. Al though major dec reases in b lood g lucose produced by insul in injections (e.g., Rowland, 1981) and major dec reases in body fat 4 produced by starvation (e.g., Keys , Brozek, Hensche l , M icke lsen , & Taylor, 1950) do stimulate hunger and eating, virtually all bouts of consumpt ion in subjects living with ad libitum a c c e s s to food occur in the a b s e n c e of such deficits (see Weingar ten , 1985). Indeed, b lood g lucose levels are c lose ly regulated by pancreat ic hormones and, as a result, rarely d isp lay major f luctuations under ad libitum condit ions (e.g., Smith, G ibbs , Strohmayer, & Stokes, 1972); and injections of insulin that produce the slight dec reases in blood g lucose that do routinely occur in subjects living under ad libitum condi t ions do not elicit consumpt ion (VanderWee le , Haraczk iewics , & V a n Itallie, 1982). A m o n g the observat ions most frequently offered in support of set-point theories of hunger and eating is the tendency for individuals who have lost weight to regain it once they return to their previous eat ing-and-energy-re lated lifestyle. A l though this line of ev idence is seduct ive, it is not sufficient to establ ish a c a s e for energy set-points: No regulat ion at all is required to explain why body weight drifts back to its original level once the factors that inf luence it return to their original states (Bol les, 1980). For example, the temperature of a house with no heating and air condit ioning will drop during a cold spel l , but will drift back to its original level once the outside temperature returns to no rma l - th i s does not imply set-point regulat ion, and neither does the return of body weight to its original level after the termination of a weight- loss program. 5 Ev idence that Increases in Energy Resou rces Do Not Rel iab ly R e d u c e Hunger and Eat ing Ev idence , aga in from studies of both humans and laboratory an imals , indicates that increases in energy resources do not reduce hunger and eat ing. Severa l studies of the effects of calor ic pre loads on hunger and eating have found that pre loads actually stimulate hunger and eating rather than suppress them; for example, premeal infusions of g lucose often increase intake (see Ge i se lman , 1987), and subjects who consume small amounts of food before a meal often exper ience increased feel ings of hunger (see Bol les , 1990). Other studies have found that calor ic pre loads have no effect on consumpt ion (e.g., Folt in & Schuster , 1984). Summary Cons ide red together, the three aforement ioned l ines of ev idence lead unavoidably to the conc lus ion that eating is not regulated by deviat ions in energy resources from energy set points. Rather, they suggest that, under typical ad libitum condit ions, variat ions in the body's energy resources have little impact on hunger and eating. Contemporary Theor ies of Hunger and Eat ing B e c a u s e of their inability to account for bas ic phenomena of hunger and eat ing, set-point theories are gradual ly being supplanted by new theories. Three 6 of these theor ies - the posit ive- incentive theory of hunger and eat ing, W o o d s ' theory of meal intolerance, and the evolut ionary theory of hunger and e a t i n g -together provide a strong theoretical framework within which most research f indings can readily be integrated (see Pine l , 1997). Posit ive-Incent ive Theory of Hunger and Eat ing The central premise of the posit ive-incentive theory of hunger and eat ing is that humans and other animals are not driven to eat by energy deficits but are drawn to eat by the ant icipated p leasure of eat ing- that is, by food's posi t ive-incentive value. Accordingly , the anticipated taste of the avai lab le food is held to be one of the main factors motivating consumpt ion. The major impact of taste on eat ing, an effect that is not accounted for by set-point theories, has been demonstrated in innumerable studies. For example, in one study rats fed a highly palatable diet of chocolate and bread in addit ion to their usual laboratory chow increased their average intake of calor ies by 8 4 % and, after 120 days, increased their body weights by 4 9 % (Rogers & Blundel l , 1 9 8 0 ) - a f inding that is totally at odds with the predict ions of set-point theories. Other ev idence for the major effect of taste on consumpt ion comes from studies of sensory speci f ic sa t ie ty -s tud ies that demonstrate that the satiat ing effect of food consumpt ion is largely speci f ic to the tastes of the foods being eaten. In one study, human subjects were asked to rate the palatabil i ty of eight different foods, and then they ate a meal of one of them. After the meal , they were asked to rate the palatabil ity of the eight foods once again. It was found 7 that their rating of the food they had just eaten had dec l ined substant ial ly more than had their ratings of the other seven foods. Moreover , when the subjects were unexpectedly offered a second meal , they consumed most of it un less it was the same as the first (Rol ls, Rol ls , Rowe, & Sweeney , 1981). Factors other than taste can inf luence the posit ive- incent ive va lue of food. T h e s e include the time of day in relation to usual mealt imes (e.g., Weingar ten , 1983), the type and quantity of food in the gut (e.g., Ge i se lman , 1987), whether other people are present and eating (e.g., R e d d & De Cast ro , 1991), and whether blood g lucose levels are within their normal range (e.g., Rowland , 1981). Posi t ive- incent ive theory does not deny that major dec l ines in energy resources\u00E2\u0080\u0094although rare under ad libitum cond i t i ons - can increase hunger and eating. Accord ing to posit ive- incentive theory, however, major d e c r e a s e s in the body 's energy resources increase hunger by increasing the posi t ive- incent ive va lue of food. The exper iences of a subject in an experiment on semistarvat ion support this view; when asked how it felt to starve, the subject repl ied: I wait for mealt ime. W h e n it comes I eat slowly and make the food last as long as possib le. The menu never gets monotonous even if it is the same each day or is of poor quality. It is food and all food tastes good. Even dirty crusts of bread in the street look appet iz ing . . . . (Keys, Brozek, Hensche l , M icke lsen , & Taylor, 1950, p. 852) 8 W o o d s ' Theory of Mea l Intolerance: Ev idence that S n a c k s are Healthier than Mea l s W o o d s and his co l leagues (e.g., W o o d s , 1991; W o o d s & Strubbe, 1994) have pointed out that eat ing a large meal , rather than restoring homeostas is , is a major homeostasis-d is turb ing event. Before a meal , the body is in reasonab le homeostat ic ba lance; then, as a meal is consumed, there is a sudden influx of fuels into the b loodstream which disturbs the homeostat ic ba lance of severa l physio logical systems: For example, blood g lucose levels increase, body temperature increases, basa l metabol ic rate increases, and hepat ic temperature increases. In order to minimize these changes , the body undergoes a variety of compensatory changes during the cephal ic stage of digest ion. For instance, just before meal-t ime, insulin is re leased from the pancreas, which extracts g lucose from the blood and reduces the subsequent increase in b lood g lucose result ing from the meal . Indeed, it is likely that it is these premeal preparatory changes , rather than energy deficits, that are the primary stimuli for premeal hunger, and conversely , the homeostat ic d is turbances result ing from meal consumpt ion, rather than energy replenishment, that are the primary stimuli for satiety. G i ven the homeostasis-d isturbing nature of eat ing large meals, W o o d s (1991) conc luded that it is healthiest to consume one 's dai ly allotment of ca lor ies as severa l smal l meals (i.e., snacks) rather than a few large meals . The ev idence supports his view. Regular ly eating severa l smal l snacks rather than a few large meals has been found to result in severa l benef ic ial metabol ic 9 changes , including the lowering of g lucose and lipid levels in the b lood (Jenk ins et a l . , 1989). Converse ly , the regular consumpt ion of large meals has been assoc ia ted with severa l metabol ic abnormal i t ies and an increased risk of card iovascu lar d i sease (Bray, 1972). Evolut ionary Theory of Hunger and Eat ing The evolut ionary theory of hunger and eating (see P ine l , A s s a n a n d , & Lehman, 1996) maintains that the mammal ian feeding system evo lved to anticipate and prevent energy deficits, rather than to merely react to them. Accord ing to this theory, this occurred because the environment in which the mammal ian feeding system evolved was character ized by an inconsistent and unpredictable food supply. For a hunger and eating mechan ism to have been adapt ive under such condit ions, it would have had to promote high levels of consumpt ion when food was avai lable, so that excess energy could be stored in the body as a buffer against unpredictable per iods of food shortage. Accord ing to this view, the mammal ian feeding system has evo lved to promote high levels of consumpt ion of sweet, salty, and fatty foods whenever they are avai lab le because in nature these tastes are invariably rich in calor ies, vitamins, and minerals. From this evolut ionary perspect ive, the preva lence of overeat ing in industr ial ized societ ies can be readily understood. In modern societ ies, the very tastes that humans have evolved to consume abundant ly are almost cont inuously access ib le in a seemingly end less variety. 10 Persona l Theor ies of Hunger and Eat ing I ant ic ipated on the bas is of pilot research that, in contrast to the empir ical ev idence, most people would bel ieve (a) that their exper iences of hunger result from dec reases in their body's energy resources below levels that are optimal for health and (b) that it is healthier for them to consume their dai ly calor ic intake as three meals than as severa l snacks . The belief that hunger results from energy deficits may have adverse implicat ions for health. Implicit in this belief is the idea that the main tenance of good health requires that one \"listen to one 's body\" and eat whenever one is hungry- that hunger is proportional and appropriate to need. G i ven the recent ev idence that, under normal condit ions, hunger does not reflect a need state but rather reflects either the ant icipated p leasure of eating or the preparatory changes in the body assoc ia ted with the expectat ion of a meal , eat ing in response to each sensat ion of hunger is clearly not a requisite for good health. Indeed, in modern soc ie t i es -soc ie t i es in which a virtually end less variety of high posit ive- incent ive value foods are readily ava i l ab le -ea t i ng in response to each sensat ion of hunger is likely to result in overeat ing and, thus, to contribute to ill health. The belief that it is healthier to consume one 's daily ca lor ic intake as three meals than as severa l snacks may also have adverse implicat ions for health. Th is belief is likely to lead people to regularly engage in three extended bouts of eat ing per d a y - a s opposed to severa l smal ler bou ts - thus expos ing 11 them to the potentially health-threatening homeostat ic perturbations assoc ia ted with eat ing large meals. The present research was concerned with document ing the preva lence of these potentially heal th-compromising bel iefs about hunger and eat ing among both lay people and health professionals. Al though an understanding of the bel iefs of lay people is clearly important, an understanding of the bel iefs of health profess ionals is, perhaps, even more important g iven the inf luence that they are likely to have both on the people whom they treat and on the genera l public. Accord ingly , two studies were c o n d u c t e d - o n e examin ing the bel iefs of university undergraduates, and one examining the bel iefs of students in health professional programs and health professionals. Study 1 In Study 1, the personal theories of hunger and eating of a large sample of university undergraduates were a s s e s s e d . Method Part ic ipants. The participants were 206 undergraduates enrol led in a health psycho logy c lass at the University of British Co lumbia . The sample was 6 9 % female, with an average age of 21.3 years (SD = 3.7). 12 Procedure and measures . The students were asked to complete a brief quest ionnaire about their beliefs about hunger and eating. They completed the quest ionnaire during c lass time, and they received course credit in exchange for part icipation. The quest ionnaire compr ised eight items that measured respondents ' set-point bel iefs. T h e s e items are listed as 1 through 8 in the Appendix . They formed a rel iable sca le , with a Cronbach 's a lpha of .85. In addit ion, pr inc ipal -components factor analys is of the items indicated that they formed a unid imensional sca le . The analys is produced one distinct factor that accounted for 4 9 % of the total var iance, and all item loadings were greater than .55. Two items in the quest ionnaire measured respondents ' bel iefs regarding the heal th iness of consuming their daily calor ic intake as three meals and as severa l snacks . T h e s e items are listed as 9 and 10 in the Appendix . Two addit ional items in the quest ionnaire asked respondents to provide their height and weight. T h e s e measures were obtained for the computat ion of respondents ' body mass indices. Body mass index (BMI), def ined as weight (kg) div ided by height (m) squared (kg/m2), is highly correlated with weight (r = .86) but relatively independent of height (r = - .03; Romieu , Willett, & Stampfer, 1988). Resu l ts The eight items measur ing set-point beliefs were averaged to produce a composi te set-point score for each respondent. Compl icat ing the ana lys is was the fact that the items differed from one another in terms of which va lues on the 13 sca le represented the lack of set-point beliefs. On items 1 through 4 sco res of 1 indicated the absence of set-point beliefs, whereas on items 5 through 8 sco res \ of 1, 2, or 3 indicated the absence of set-point bel iefs (assuming that items 7 and 8 are reverse scored). Accordingly , a respondent who did not endorse any of the set-point beliefs a s s e s s e d by the items could have obtained a maximum total score of 16 (the sum of the highest scores represent ing the a b s e n c e of set-point bel iefs on the items) and, thus, a maximum composi te set-point score of 2.0 (the average of the highest scores represent ing the a b s e n c e of set-point bel iefs on the items). For the sample, the distribution of composi te set-point scores was approximately normal, with a mean of 3.52 and a standard deviat ion of .54, indicating that most respondents held set-point bel iefs. Indeed, none of the respondents had composi te set-point scores below 2.00, 17% had scores between 2.01 and 3.00, 5 7 . 3 % had scores between 3.01 and 4.00, and 2 5 . 7 % had scores between 4.01 and 5.00 (see Figure 1). A statistically signif icant correlat ion between composi te set-point scores and BMI was found (r = .18, p< 05), indicating that respondents with higher BMIs held stronger set-point bel iefs than respondents with lower BMIs. R e s p o n s e s to items 9 and 10 were compared to a s s e s s whether the respondents be l ieved that it is healthier to consume their dai ly ca lor ic intake as three meals or as severa l snacks . Sco res on item 9 (eating three meals per day, M = 4.25, S D = .63) were significantly higher than scores on item 10 (snacking severa l t imes per day, M = 3.03, S D = 1.23; t[205] = 11.63, p<.001), suggest ing 14 that respondents bel ieved that it is healthier to eat three meals per day than severa l snacks per day. Of the respondents, 66.1 % rated eating three meals per day as more healthy than snack ing throughout the day, 15 .5% rated eat ing three meals per day as as healthy as snack ing throughout the day, and 18.4% rated eating three meals per day as less healthy than snack ing throughout the day. Moreover , relatively few respondents bel ieved that snack ing is a healthy pract ice. Of the respondents, only 3 7 . 3 % rated snack ing throughout the day as healthy. Dif ference scores generated by subtracting item 9 from item 10 for each respondent were not correlated signif icantly with either composi te scores or BMIs. Study 2 In Study 2, the personal theories of hunger and eating of samp les of students in health professional programs and samp les of health profess ionals were a s s e s s e d . Method Part ic ipants. Six samples were recruited: 44 students in their final year of study in dietet ics (93% female; average age = 23.0 years, S D = 2.8), 35 students in their final year of study in nursing (94% female; average age = 22.9 years , S D = 2.59), 25 students in their final year of study in medic ine (52% female; average age 25.6 = years, S D = 3.9), 32 dietitians working in local hospi ta ls (97% female; 15 average age 33.8 = years, S D = 8.26), 31 nurses working in local hospi ta ls (97% female; average age = 36.2 years, S D = 6.9), and 30 phys ic ians working in local hospi ta ls (37% female; average age = 41.5 years, S D = 7.8). Procedure and measures . The respondents completed the quest ionnaire descr ibed in Study 1. The dietetics, students and nursing students completed the quest ionnaire during c lass time, and the dietit ians, nurses, and phys ic ians completed the quest ionnaire during staff meetings. B e c a u s e a c c e s s to medical students during their c l asses could not be obtained, quest ionnai res were left in their mai lboxes along with a brief introductory letter containing contact te lephone numbers; they had 2 weeks to complete the quest ionnaire and were asked to return them to a dropbox p laced near their mai lboxes. The eight items measur ing respondents ' set-point bel iefs formed a rel iable sca le in each of the six samples. A lphas for the samp les ranged from .86 to .90. In addit ion, pr inc ipal-components factor ana lys is of the items in the combined-sample (generated by co l laps ing the six samples) produced only one distinct factor. The factor accounted for 5 4 % of the total var iance, with all item loadings greater than .59. The six samples were combined for factor ana lys is because the means for each item did not did not differ signif icantly among the samples , and because the combined-sample s ize a l lowed for a more rel iable factor ana lys is . 16 Resul ts A s in Study 1, the eight items measur ing set-point bel iefs were averaged to produce a composi te set-point score for each respondent. The mean composi te set-point score for each sample is l isted in Tab le 1, a long with its s tandard deviat ion. The distribution of composi te set-point scores for e a c h sample is shown in Figure 2. The composi te set-point scores of the samp les did not differ signif icantly, nor did they differ signif icantly from the composi te set-point scores obtained from the undergraduate sample in Study 1, indicating that the students and health professionals in the present study were as likely to maintain set-point beliefs as the undergraduate students in Study 1. The correlat ions between composi te set-point scores and BMIs were not signif icant for any of the samples ; however, the variabil i t ies in BMI in the samp les were relatively smal l and may have been insufficient to constitute reasonab le tests of this trend. M e a n scores and standard deviat ions on items 9 and 10 for each of the samp les are listed in Tab le 2. Sco res on these items did not differ signif icantly in any of the samples , suggest ing that the respondents be l ieved that snack ing throughout the day is as healthy as eating three meals per day. A c r o s s the six samples , 2 5 . 9 % of the respondents rated eating three meals per day as more healthy than snack ing throughout the day, 56 .4% rated eating three meals per day as as healthy as snack ing throughout the day, and 17 .8% rated eat ing three meals per day as less healthy than snack ing throughout the day. Dif ference 17 scores generated by subtracting item 9 from item 10 for each respondent were not correlated signif icantly with either composi te set-point scores or BMIs in any of the samples . D iscuss ion The present research was the first to examine how people interpret their own hunger and eating behavior. It was predicated on the premise that most people hold bel iefs about hunger and eating that are likely to promote excess ive consumpt ion and ill health. O n e hypothesis was that most people would bel ieve that their hunger and eating behavior are regulated by deviat ions in their body 's energy resources from levels that are optimal for their health. The results of both studies were consistent with this hypo thes is -a l l of the undergraduate students in Study 1, and almost all of the students in health professional programs and health profess ionals in Study 2 had composi te set-point scores of more than 2.0, indicating at least some commitment to the belief that their feel ings of hunger are produced by dec reases in energy resources below levels that are optimal for health. In Study 1, there was a posit ive correlat ion between set-point thinking and BMI, indicating that respondents with higher BMIs had stronger set-point bel iefs than respondents with lower BMIs. Th is f inding lends support to the argument that set-point bel iefs contribute to excess ive consumpt ion. Never the less, I did 18 not expect to see ev idence of such a correlat ion for two reasons . First, factors in addit ion to level of consumpt ion (e.g., level of energy expenditure and metabol ic rate) greatly inf luence BMI (see K e e s e y & Powley, 1986). S e c o n d , g iven that most students were expected to ev ince set-point beliefs, I did not ant icipate that there would be enough variation in that measure for a signif icant correlat ion to emerge. Another hypothesis was that most people would bel ieve that it is healthier to consume their daily calor ic intake as three meals than as severa l snacks . The results of Study 1 were consistent with this hypo thes i s - the majority of the undergraduate students bel ieved that the consumpt ion of three meals per day is healthier than snack ing throughout the day. However, the results from Study 2 revealed that the majority of the students in health professional programs and health profess ionals bel ieved that the consumpt ion of three meals per day is as healthy as snack ing throughout the day. Al though this belief is more consistent with the ev idence than was anticipated, it is important to recogn ize that it does not d iscourage meal consumpt ion. Indeed, the only belief that is likely to d iscourage the consumpt ion of meals is the belief that eat ing three meals per day is less healthy than snack ing throughout the day- -and relatively few of the students in health professional programs and health profess ionals held this belief. Interestingly, al though most of the students in health professional programs and health professionals surveyed in Study 2 did not differ from the 19 undergraduate students surveyed in Study 1 in their bel iefs about energy set points, the majority of the students in health professional programs and health profess ionals held beliefs about snack ing and meal consumpt ion that were intermediate between common beliefs and recent research f indings: They bel ieved that snack ing and meal consumpt ion are equal ly healthy. Th is may be expla ined by the greater exposure that students in health profess ional programs and health profess ionals are likely to have had to recent research f indings on hunger and eating and by the relative e a s e with which the idea that snack ing is healthy is likely to be accommodated into people 's bel iefs about hunger and eating. Th is idea could be added to most people 's bel iefs about hunger and eating without present ing a major chal lenge to how they think about hunger and eating. In contrast, the idea that hunger is not normally regulated by dec reases in energy resources below set points would require a more thorough revis ion of most people 's thinking. The f indings of the presents studies have three unsettl ing implicat ions. First, they suggest that misconcept ions about hunger and eating that may have . adverse implicat ions for health are common among both lay people and health professionals. S e c o n d , they suggest that the training that students in health professional programs receive does not adequately modify their misconcept ions about hunger and eating. A n d third, they suggest that health profess ionals are d isseminat ing misinformation about hunger and eating to the genera l publ ic that may be encourag ing the very problems that they are trying to treat\u00E2\u0080\u0094this may be 20 one reason why current efforts to prevent and treat eat ing-related problems are largely unsuccess fu l . Recent ev idence suggests that the problem of overeat ing is far more pervas ive than may be commonly assumed (see Soha l & We ind ruch , 1996). A d libitum consumpt ion in the absence of obesity has been found to increase the r isks of numerous d i seases (e.g., d iabetes, hypertension, heart d i sease , cancer) , to accelerate aging-related physio logical and behaviora l changes (e.g., dec reases in immunological response, dec reases in musc le mass , d e c r e a s e s in protein production), and to dec rease life e x p e c t a n c y - b e c a u s e it p roduces high levels of oxygen molecu les that cause damage to cel ls (e.g., base alterations in DNA, s ingle-strand breaks in DNA, modif ication of proteins). Be that as it may, the pilot interviews that I conducted suggest that many people bel ieve that all of the adverse effects of overeat ing on health are mediated by obes i ty - tha t people who eat a lot but who are not obese because they have high levels of energy expenditure are immune to the negat ive health consequences of overeat ing. Th is may be another misconcept ion that contributes to excess ive intake, and thus it warrants systemat ic study. Future research needs to evaluate the ef fect iveness of educat ional programs des igned to treat over consumpt ion by modifying peop le 's bel iefs about hunger and eating. G iven the prevalence of overeat ing in many parts of the world, its dire consequences for health, and the role that bel iefs about hunger and eating may play in promoting over consumpt ion, the potential benefits of such educat ional programs are immense. In much the same way that many people who have understood the adverse effects on health of eat ing too much fat have been able to reduce their fat consumpt ion, people who deve lop an understanding for why they feel hungry at regular meal t imes or when high-incent ive-value foods are avai lable may be better able to reduce their intake. 22 Tab le 1 M e a n Compos i te Set-Point Sco res for Study 2 Compos i te scores Samp le n X S D Dietet ics students 44 3.54 .75 Nurs ing students 35 3.30 .70 Med ica l students 25 3.42 .78 Dietit ians 32 3.71 .65 Nurses 31 3.50 .69 Phys ic ians 30 3.60 .56 Note. Compos i te set-point scores ranged from 1 to 5. A higher score represents higher endorsement of set-point beliefs. The highest composi te score that was obtainable by a respondent who did not endorse any set-point bel iefs was 2.0. 23 Tab le 2 M e a n Sco res on Items A s s e s s i n g Bel iefs about Snack ing and Mea l Consumpt ion for Study 2 Snack ing throughout the day Three meals per day Samp le X S D X S D Dietetics students 4.18 .90 4.32 .47 Nurs ing students 3.86 .91 4.11 .72 Med ica l students 4.08 .70 4.24 .88 Dietit ians 4.63 .55 4.72 .63 Nurses 4.19 .79 4.35 .71 Phys ic ians 4.37 .61 4.50 .51 Note. Sco res ranged from 1 (very unhealthy) to 5 (very healthy). Distribution of Composite Set-Point Scores in Study 1 35 30 25 20 15 10 0 ' / / ' i 1 \u00E2\u0080\u00A2 <$> _ \u00C2\u00ABp -<$> <$> <\u00C2\u00A7> <$ V Qy 0 / o>- O j - fc- fc-^ & C> "Thesis/Dissertation"@en . "1996-11"@en . "10.14288/1.0087215"@en . "eng"@en . "Psychology"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "Personal theories of hunger and eating"@en . "Text"@en . "http://hdl.handle.net/2429/4735"@en .