Dose-dependent effects of alcohol on appetite and food intake

Physiology & Behavior 81 (2004) 51 – 58

Dose-dependent effects of alcohol on appetite and food intake S.J. Caton, M. Ball, A. Ahern, M.M. Hetherington* Department of Psychology, University of Liverpool, Liverpool, L69 3BX, England, UK Received 18 August 2003; received in revised form 17 December 2003; accepted 22 December 2003

Abstract To examine the potential dose – response effect of alcohol on appetite and food intake, 12 males attended the laboratory on three occasions. On each occasion, they were given a standard breakfast, then lunch 3 h later, and dinner, 4 h after that. Thirty minutes before lunch, Ss received 330 ml of no-alcohol lager (263 kJ: no-alcohol condition), the same amount of lager spiked with 1 unit (1 UA: 8 g ethyl alcohol, 498.2 kJ) or 4 units of alcohol (4 UA: 32 g ethyl alcohol, 1203.8 kJ). Visual analogue scale (VAS) ratings of appetite and mood were recorded before and after preloads and lunch, then hourly across the day. Intake at lunch (excluding energy from the preload) was significantly higher following 4 UA (5786 F 991 kJ) compared to 1 UA (4928 F 1245 kJ). Participants consumed more high-fat salty food items at lunch following 4 UA compared to the other preloads. Hunger was rated higher following 4 UA across the day in comparison to the other preloads, but fullness ratings failed to reflect any difference by condition. Energy intake at dinner was similar in all conditions and total energy intake across the day was significantly higher after 4 UA (14,615 F 1540 kJ) than after 1 UA (13,204 F 2156 kJ). In conclusion, above a certain threshold, alcohol appears to stimulate appetite in part, due to elevated levels of subjective hunger. When this occurs, energy intake is not reduced at subsequent meals. Thus, alcohol may contribute to positive energy balance via its additive effects to total energy intake and by short-term appetite stimulation. D 2004 Elsevier Inc. All rights reserved. Keywords: Alcohol; Appetite; Food intake; Hunger

1. Introduction Alcohol is a macronutrient with an energy density of around 29 kJ (7 kcal) per gram and an estimated average alcohol intake in Western countries constitutes 8– 10% of total daily energy intake [1]. Alcohol is second only to dietary fat in energy density. However, the relationship between alcohol consumption, energy intake and energy balance is complex. The regulation of appetite and energy intake is fundamental to the control of energy balance and the maintenance of body weight. Metabolic fuels, protein, carbohydrate, fat and alcohol exert differential effects on hunger and satiety and this in turn influences how much is consumed, and therefore, influences energy balance. Alcohol is the least satiating macronutrient in the satiety hierarchy [2], yet, it is at the top of the oxidative hierarchy [3]. Alcohol is generally consumed as a beverage and fluids are known to have a weaker effect on energy compensation than

* Corresponding author. Tel.: +44-151-794-1454; fax: +44-138-2344613. E-mail address: [email protected] (M.M. Hetherington). 0031-9384/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2003.12.017

solid foods [4]. Therefore, alcohol may contribute to the development and maintenance of obesity in a number of ways including its high energy density, exertion of weak satiety signals and poor subsequent energy compensation. Epidemiological evidence illustrates the complex relationship between alcohol intake and body weight. Colditz et al [5] examined alcohol intake in relation to diet and obesity in men and women in two large-scale cohort studies, the Nurses Health study and the Health Professionals follow-up study. A positive relationship between alcohol consumption and energy intake was found for both men and women. However, an inverse relationship was noted between alcohol consumption and body mass index (BMI: kg/m2) for females [6 –8]. Women who drink tended to be somewhat lighter than women who abstain [5,9], but this was not true of males [6,8]. Crouse and Grundy [10] added 24% of daily energy intake in the form of alcohol (approximately 90 g/ day) into the habitual diets of a group of males. No weight gain occurred in the lean cohort whereas the obese gained weight, demonstrating differential effects of alcohol on body weight according to body weight status [11]. Laboratory investigations consistently demonstrate that consumers fail to compensate for alcohol ingestion in the

52

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58

short term [4,12 – 17]. For example, Westerterp-Plantenga and Verwegen [2] reported that 24-h energy intake was raised on the days an alcoholic beverage was consumed compared to equienergetic preloads of fat, protein or carbohydrate. Alcohol is both additive to total energy intake, and appears to enhance food intake [14]. Thus, energy intake was significantly higher following 3 units of alcohol added to an alcohol-free lager compared to the alcohol-free lager on its own [16]. Given that fat has the lowest satiating capacity [18], it could be that the most pronounced effects on body weight regulation and energy balance arise when alcohol is consumed alongside a high-fat diet. Tremblay and St.-Pierre [19] examined the hyperphagic effect of a high-fat diet and alcohol preload in comparison to an equienergetic carbohydrate preload. There was an increase in subsequent food intake following the high-fat and alcohol preload in comparison to a carbohydrate preload. Weak satiating signals arising from fat, coupled with the propensity of fat to be more readily stored than oxidised, suggest a potentially synergistic effect when alcohol is consumed with fatty foods. A number of mechanisms may explain the observed effects of alcohol on appetite, including metabolic sensing [20] and psychological processes including expectancy effects [14,15]. Thus, belief about the alcohol content of a beverage appears to moderate its effect on appetite [14,15,21]. Clearly, if consumers typically pair alcohol consumption with food intake and experience increased appetite following alcohol ingestion, then, this contributes to associative conditioning; that is, alcohol-related cues may elicit appetite following repeated pairings. However, it is not clear if this happens even with relatively small doses of alcohol. If alcohol stimulates appetite, does this happen in a dose-related pattern, such that increasing the amount of alcohol consumed directly increases the amount of food eaten? Alternatively, the effect may be dose-sensitive; that is, it may only occur above a certain threshold. The present investigation was designed to examine any dose –response relationship between alcohol and appetite by comparing rated appetite and food intake following preloads containing 0, 1 or 4 units of alcohol. Given the importance of expectancy effects, all conditions were administered using a no-alcohol lager vehicle. In order to examine different psychological processes underlying any dose – response effect, subjective ratings of appetite, mood and palatability were tracked before and after preload drinks and after meals. If the effect of alcohol on appetite follows the typical dose – response pattern, then, rated appetite and energy intake should be higher following 4 units compared to 1 unit of alcohol, and these should be higher following 1 unit compared to baseline. In addition, if the effect of alcohol on appetite is modulated by selectively increasing the pleasantness and desire to eat particular foods or by altering subjective mood states, this should then emerge in a dose –response pattern.

2. Method 2.1. Participants Twelve male participants aged 18 – 35 were recruited from the staff and student population at the University of Liverpool via recruitment posters. Participants were screened prior to recruitment to ensure they were suitable, that is, they were of normal weight (BMI < 26 kg/m2), they were unrestrained eaters with a score of 2.5 or less on the DEBQ-R [22], they consumed no more than 21 units of alcohol per week and were self-reported nonsmokers. Individuals who abstained from any alcohol consumption ( < 2 units/week) were also excluded from the study. The experiment received ethical approval from the Department of Psychology Ethics Committee and participants provided written informed consent. 2.2. Materials Subjective ratings were made on visual analogue scales (VASs) consisting of 100-mm horizontal lines anchored with ‘‘not at all’’ and ‘‘extremely.’’ Positioned above each line was a specific question relating to appetite (hungry, full and thirsty), mood (happy, sad, energetic, tired, clearheaded, lightheaded, relaxed, bloated and anxious) and palatability (pleasantness and the desire to consume foods/drinks). Each question was rated by placing a vertical line on the horizontal line at the point which best represented their current state. Ratings were recorded as the length from the left endpoint of the VAS to the vertical mark made by the participant. Participants were given breakfast on each test day consisting of cereal, milk, toast, bread, margarine, jam and fresh orange juice (Table 1). Participants were told that they must consume all of the food and drink items presented to them. Total energy from this meal was 2604 kJ (see Table 2); this provided 25% of estimated daily energy intake based on a 10,500 kJ/day diet. The no-alcohol preload consisted of 330 ml of chilled noalcohol lager (330 ml/316.8 g, 263 kJ), presented in a pint glass with 96% pure ethyl alcohol wiped around the rim. The 1 and 4 units of alcohol preloads consisted of the same amount of no-alcohol lager with 8 g (340 ml/324.8 g, 498.2 kJ) and 32 g of ethyl alcohol added (370 ml/348 g, 1203.8 kJ), both were presented in a pint glass. The lunch test meal consisted of a variety of lunch items, two low-fat sweet and four low-fat savoury items ( < 10% fat), two high-fat sweet and four high-fat savoury items (>25% fat) and 500 ml of chilled water. The dinner test meal consisted of pasta, tomato pasta sauce and chocolate mini rolls. Each food was weighed before and after to calculate intake in grams. Energy intake was then calculated using WISP (Tinuviel software) or manufacturer’s information. Items rated for pleasantness were aliquot samples of seven of the lunch foods presented in the following order:

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58 Table 1 Food items offered at breakfast, lunch and dinner

Table 3 Summary of procedure

kJ/100 g P Breakfast Cornflakes (Kellogg’s), 30 g 1455.9 Semiskimmed milk, 300 g 194.6 Orange juice (Just Juice), 200 g 180.6 Jam (Hartleys), 20 g 987 Bread (Warburtons medium sliced), 80 g 939.7 Flora 2646

F

CHO

T0

8 0.9 3.3 1.6 .66 0 .3 0 10.1 1.8 0.1 70

82 5 10.89 62.7 44.6 0.1

T1

58.6 54.6 10 8.7 7.5 1.1 49 5.6 43.6 0 3.6 12.1

Lunch Chocolate buttons (Cadbury), 100 g Penguin cake bars (McVitie’s), 71.7 g Fruit salad (Tesco), 135 g Strawberry yoghurt (Muller Lite), 150 g Cheese slices (Kraft), 101.4 g Salami (Tesco), 35 g Plain crisps (Walkers), 50 g Mayonnaise (Hellmans), 50 g Bread rolls (Warburtons), 161 g Tuna in brine (Tesco), 100 g Cottage cheese (Tesco), 150 g Salad cream (Heinz), 50 g Chilled water, 500 ml

2205 2116.8 180.6 222.6 1176 1415.4 2226 2049.6 1066.8 457.8 327.6 567

7.8 5.2 0.4 4.4 13 22.4 6.5 0.8 10.3 25.9 11.9 1.1

29.4 29.5 0 0.1 26 27 34 51.4 4.3 0.6 1.8 8.5

Dinner Pasta (Tesco), 300 g Tomato pasta sauce (Tesco), 700 g Chocolate mini roll (Cadbury), 72 g Chilled water, 500 ml

1468.8 142.8 1822.8

11.5 0.6 5.5

1.9 72.6 0.8 6.1 20.6 55.6

cheese, salami, crisps, tuna, cottage cheese, chocolate, yoghurt and the preload drink. Participants were provided with 500 ml of chilled water. 2.3. Procedure Condition assignment was randomised (Latin Square). Each test day was separated by at least 3 days and no more than 3 weeks (see Table 3 for a summary). Participants were led to believe that the purpose of the investigation was to examine the effect of food and drink on mood. On each occasion, participants attended the laboratory between 08:30 and 09:30 h following an overnight fast for breakfast (T0) and returned around 3 h later for lunch. To ensure participants arrived at the laboratory in a fasted state, they were telephoned the day prior to each test day to remind them not to consume any food or drinks except water for 12 h prior to their visit.

Table 2 Macronutrient composition of test meals Total Total P energy weight (%) (kJ) (g) Breakfast 2604 640 Lunch 10,800 1154 Dinner 6438 1072

F (%)

CHO P (%) (g)

53

F (g)

CHO (g)

14.05 19.62 66.37 21.77 13.51 109.61 16.21 48.87 34.97 104.22 139.63 239.43 11.13 15.34 73.52 42.66 26.13 300.53

T2 T3 T4 T5 T6 T7 T8 T9

(09:00 h) Volunteers arrive at lab in fasted state, standard fixed breakfast (12:00 h) Return to lab appetite and mood, VAS 1, palatability ratings 1 Preload, 30-min rest, VAS 2, palatability ratings 2 (12:50 h) Ad libitum lunch, VAS 3, palatability ratings 3 (13:10 h) (14:10 h) Appetite and mood, VAS 4 (15:10 h) Appetite and mood, VAS 5 (16:10 h) Appetite and mood, VAS 6 (17:10 h) Return to lab appetite and mood, VAS 7, palatability ratings 4, ad libitum dinner Appetite and mood, VAS 8, volunteer free to leave Food diary, end of day appetite and mood, VAS 9

Prior to leaving the laboratory, participants were instructed to consume no food or drinks except water prior to breakfast and in between breakfast and lunch and lunch and dinner. Upon their return to the laboratory, participants were seated in a private cubicle for lunch and asked if they had consumed any food or drink item to check compliance with instructions to abstain. They then completed their first set of VAS and palatability ratings (T1). Participants were given instructions to taste and rate the food samples in the order that they appeared on the ratings sheet. Participants were asked to drink water between each food/drink. The preload was then presented and the participants were given instructions to drink all of the preload within 10 min and to take regular even sips of the drink to standardise drinking rates. They then rested quietly for 30 min and were allowed to read (T2). After this 30-min interval, a further set of appetite and mood VAS was administered together with palatability ratings before lunch was served. They were then invited to eat as much or as little as they liked from the array of foods presented and to ask for more food if this was required. Participants were asked to mix foods on the plate provided. This was done for the convenience of the investigators so that leftover foods could be weighed accurately. Participants rang a bell to indicate when they had finished eating, and lunch trays were then removed. Food plates were weighed from the trays in a separate kitchen. A third set of VAS and palatability ratings (T3) was then administered. After lunch, participants were provided with appetite and mood VAS to be completed hourly until their return to the laboratory for dinner (T4, T5 and T6). Again, they were instructed not to eat or drink anything except water before dinner. Participants returned to the laboratory around 4 h following the end of lunch. Upon arrival for dinner, participants completed a further set of appetite and mood VAS, and a fourth set of palatability ratings of the lunch items. They were then provided with the dinner (T7) and invited to eat as much or as little as they wished. All dinner items were weighed before and after to determine intake. After dinner, participants filled out another set of VAS (T8). Participants were asked to record all foods or drinks

54

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58

consumed during the remainder of the day in a food diary, and to complete a final set of VAS ratings (T9).

way ANOVA. To examine significant differences, pairwise comparisons were used (least significant difference, equivalent to no adjustment for multiple comparisons).

2.4. Data analysis All analyses were conducted on SPSS (v10). Variations in degrees of freedom for different tests are due to some incomplete data as participants occasionally omitted hunger ratings at some time points. Energy intake between conditions was compared in a one-way analysis of variance (ANOVA) with repeated measures (sphericity assumed). Subjective appetite and mood ratings were analysed over time and preload condition, in a 3  9 ANOVA with repeated measures. Assumptions for ANOVA were not met for the analysis of hunger ratings at all time points. These include time points 3, 4 and 8. As a result, Greenhouse – Geisser (degrees of freedom, alpha level) was used. Where it was predicted that significant differences might occur at a particular time point, one-way ANOVA was used to examine the effect across conditions with repeated measures. Individual food item intake was compared across conditions using one-

3. Results 3.1. Energy intake Lunch intake was greatest following the preload containing 4 units of alcohol (4 UA) relative to 1 unit of alcohol (1 UA) and the no-alcohol preload (see Fig. 1). ANOVA revealed a main effect of condition [ F(2,22) = 3.59, P=.045] and pairwise comparisons indicated that participants consumed more energy at lunch following 4 UA in comparison to 1 UA despite consuming a more energy-dense preload ( P < .05). The difference between energy intake at lunch following 4 UA compared to 1 UA represented an enhancement of intake of 17% and relative to no alcohol of 9%. Taking into account the energy from the preloads, ANOVA overall energy intake at lunch again showed a

Fig. 1. Upper panel shows energy intake (kJ) at lunch in each condition. Lower panel shows energy intake (kJ) across the entire day.

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58

significant main effect of condition [ F(2,22) = 14.45, P < .001]. Pairwise comparisons demonstrated energy intake to be significantly higher following 4 UA (6990 F 992 kJ) in comparison to both 1 UA (5426 F 1246 kJ; P < .001) and the no-alcohol preload (5580 F 1256 kJ; P < .01). Energy intake at dinner did not differ significantly, suggesting a lack of compensation later in the day. Energy intake at dinner was 5665 F 1969 kJ following the no-alcohol preload, 5174 F 1608 kJ following the preload containing 1 UA and 5021 F 1333 kJ following the preload containing 4 UA. Total energy intake (breakfast energy intake + lunch energy intake + energy from the preload + energy intake at dinner) was significantly higher following the preload containing 4 UA in comparison to 1 UA [ F(2,22) = 6.13, P < .05; Fig. 1]. Energy intake following the 4-UA preload did not differ significantly when compared to the noalcohol preload. Overall, participants consumed more energy from food on the day when most alcohol was consumed. 3.2. Individual foods Intake of individual food items was examined to establish if more energy was derived from any one particular item across conditions. Participants consumed significantly more energy from crisps following 4 UA (709 F 323 kJ) in comparison to 1 UA (499 F 289 kJ) [ F(2,22) = 5.68, P < .05] and this did not differ from the no-alcohol condition

55

(578 F 363 kJ). Energy derived from all other food items did not vary significantly between conditions. 3.3. Rated appetite and mood Subjective appetite and mood ratings were plotted across the day from baseline (T1) to the end of the day (T9). For rated hunger, significant main effects of time [ F(2,22) = 28.29, P < .001], condition [ F(8,88) = 11.27, P < .001] and a significant Time  Condition interaction [ F(16,176) = 4.25, P < .001] were found. A oneway ANOVA was carried out on ratings of hunger at T1; there were no significant differences between ratings indicating that all participants started each condition in a similar state of hunger. Significant differences were found between conditions at T3 [ F(1.26,13.92) = 16.02, P < .001 Greenhouse – Geisser], immediately following lunch, and again at T4 [ F(1.19,13.14) = 7.12, P < .05 Greenhouse – Geisser], T5 [ F(2,22) = 4.18, P < .05] and T6 [ F(2,22) = 4.49, P < .05]. These differences clearly relate to a higher level of hunger reported by participants immediately following the 4-UA preload relative to the other preloads which was sustained well after lunch (see Fig. 2, upper panel). Significant differences in rated hunger were also found at T8 [ F(1.18,12.971) = 10.5, P < .01 Greenhouse – Geisser] and T9 [ F(2,22) = 5.92, P < .01] ratings taken immediately after dinner and at the end of the day. This suggests that the

Fig. 2. Upper panel shows rated hunger across the test days in each condition. Lower panel shows rated fullness across the test days in each condition.

56

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58

elevation in rated hunger associated with drinking 4 UA was still present at least 6 h after consuming the preload. In contrast to rated hunger, fullness ratings only revealed a significant main effect of time [ F(8,72) = 27.65, P < .001; see Fig. 2, lower panel]. It is striking to note, however, that fullness was rated as similar following lunch despite a significantly higher energy intake after 4 UA. Ratings of feelings, such as clearheadedness, revealed a significant effect of time and a Time  Condition interaction, [ F(16,112) = 1.90, P < .01]. Individual one-way ANOVAs demonstrated a nonsignificant difference at baseline; however, participants reported feeling significantly less clearheaded 30 min following 4 UA at T3 in comparison to ratings following 1 UA and the no-alcohol preload [ F(2,20) = 11.33, P < .01]. This was also found at T5 [ F(2,18) = 4.12, P < .05; n = 10], implying that the reduction in feeling clearheaded following 4 UA lasted for at least 1 h. Similarly, ratings of lightheadedness were influenced by condition and time [interaction effect: F(16,176) = 3.11, P < .001]. The 4-UA preload increased ratings of lightheadedness compared to the other preloads 30 min following the preload and for the following 3 h. Feelings of tiredness also differed by condition [ F(2,22) = 4.11, P < .05] and time [ F(8,88) = 3.69, P < . 01]. One-way repeated-measures ANOVA and pairwise comparisons reveal that participants were significantly more tired following 4 UA at T4 [ F(2,22) = 5.04, P < .05] and T5 [ F(2,22) = 3.46, P=.049]. However, given that these ratings were completed after lunch, it is not clear if fatigue increased as a function of the preload itself or the combination of the preload and a larger lunch. 3.4. Pleasantness ratings Analysis of the pleasantness of the preloads at baseline revealed a significant main effect of preload [ F(2,22) = 7.257, P < .01]. Pairwise comparisons revealed that participants found the no-alcohol preload to be significantly more pleasant than the two alcohol-containing preloads ( P < .05). There were no significant differences between ratings of desire to consume the preloads at baseline (Table 4). Rated pleasantness of the preload declined by 4 F 2.54 mm following ingestion and this decline was similar across all conditions [main effect of time: F(3,33) = 3.76, P < .05]. In contrast, pleasantness ratings of the other foods that were tasted but not eaten increased by 1.9 F 1.63 mm following

Table 4 Pleasantness and desire to consume preloads at baseline

Baseline 1 unit 4 units

Mean pleasantness rating (mm)

Mean desire to consume rating (mm)

59.7 ( F 13.1) 49.1 ( F 16.5) 36.7 ( F 24.4)

46.0 ( F 23.1) 38.4 ( F 17.5) 35.7 ( F 20.1)

the preload. Change in pleasantness of the taste of the preload was greater than the change in pleasantness of the uneaten foods [interaction between food and time: F(7,77) = 2.87, P < .05]; therefore, sensory-specific satiety was observed.

4. Discussion Administration of a relatively high dose of alcohol (4 units) increased energy intake at lunch by 17% and was associated with higher hunger levels. Compensation for the additional energy in the high-dose alcohol condition did not occur at either lunch or dinner. No dose –response pattern for alcohol was found, because a single unit of alcohol failed to stimulate appetite relative to the no-alcohol preload. However, the findings of the present study suggest that the effect of alcohol on appetite and food intake is dosesensitive, with a set threshold for the stimulation of appetite by alcohol. More importantly, it appears that energy derived from a higher dose of alcohol is additive to total daily energy intake. Results from the present study confirm and extend earlier findings on the stimulatory effect of alcohol on appetite and food intake [14,16]. It appears that a low dose of alcohol of around 1 unit fails to have any impact on hunger or subsequent food intake. Similarly, Yeomans and Phillips [15] reported that low doses of alcohol did not increase food intake compared to the no-alcohol condition (water). It is possible that the pharmacological action of 1 unit of alcohol is simply too weak to have an effect on appetite. Thus, alcohol has a threshold below which it fails to promote food intake. The higher dose of alcohol promoted intake of salty potato crisps relative to the low-dose and no-alcohol conditions. Mattes and DiMeglio [23] observed increased perception of saltiness with beer, but only at certain NaCl concentrations. The current findings may reflect changes in taste perception of salty foods following a high dose of alcohol. An alternative explanation to changes in taste perception underlying the increased selection of this food after 4 UA may simply be accounted for by familiarity. Salty foods are typically eaten with alcoholic beverages; thus, increased intake may be a result of prior learning coupled with the pharmacological effect of 4 UA. Participants reported feeling more hungry after lunch and for up to 3 h after consuming the higher dose of alcohol despite consuming more energy. Immediately before dinner, rated hunger was similar in all conditions and intake at dinner was similar; however, by the end of the day, ratings of hunger were higher following 4 UA. This implies that elevated hunger produced by the high dose of alcohol lasted for many hours beyond ingestion and indeed its metabolism. Given that it takes around 1 h to fully metabolise a single unit of alcohol, hunger is elevated beyond the expected time taken to fully metabolise 4 UA.

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58

Ratings of fullness followed a typical profile throughout the day, with fullness increasing immediately after a meal and then gradually decreasing until the next meal. It is significant, however, that intake of the higher dose of alcohol and larger lunch and a similar size dinner failed to produce significantly greater fullness ratings compared to the other conditions. This could be attributed to a delayed satiety effect produced by alcohol. Westerterp-Plantenga and Verwegen [2] demonstrated extended meal duration and delayed satiety development following alcohol ingestion compared to other preloads. It appears that alcohol has the capacity to disrupt postingestive satiety mechanisms and suppression of satiety may contribute to the additive effects of energy from alcohol to the habitual diet. The preloads containing alcohol were rated as less pleasant than the vehicle (no-alcohol lager); thus, lunch presented an opportunity to consume something more pleasant. Hill et al. [24] reported that rated hunger was lower following a lesser preferred meal; thus, this may explain in part the lack of immediate increase in hunger 30 min following the preload containing 4 UA. Ingestion of 4 UA produced significant changes in mood; thus, participants reported feeling less clearheaded and more lightheaded. These subjective sensations may be associated with feeling hungry and may have contributed to increased food intake [2]. Again, prior learning, for example, associating lightheadedness with hypoglycaemia, may promote food intake after intake of alcohol. In the present study, the failure of a dose-dependent effect could be attributable to expectancy effects. The control drink was a nonalcoholic lager which was chosen on the basis that it shares many of the sensory properties of lager (taste, smell, appearance and carbonation). The cues associated with this beverage may elicit the expectancy that alcohol will be ingested, even when no alcohol is present because cues that predict the presence of a drug become associated with the effects of that drug. It remains possible that a drink, such as water with no association to alcohol, could have produced a more pronounced dose– response effect. This has been shown by Yeomans et al. [14] who used a vehicle which is typically ingested in the absence of any alcoholrelated cues (carbonated apple juice). In this study, a low dose of alcohol administered in the beverage was additive to total energy intake in comparison to the no-alcohol preload. However, in the present study, because alcohol-containing preloads were rated as being less pleasant as the noalcohol preload, participants may have been aware that they were consuming alcohol, especially the higher alcohol preload. This mitigates against expectancy as an explanation for the lack of dose – response effect in the present context. Overall, energy intake was higher at the end of the day by around 11% following 4 UA, suggesting that energy received in the preload was added to the diet and confirming previous investigations [2,12,15,19]. If this pattern is maintained over several days of alcohol intake, this could contribute to positive energy balance.

57

It is likely that physiological and metabolic mechanisms underlie the lack of short-term compensatory reductions in subsequent food intake. Alcohol is a toxin and so must be eradicated as fast as possible from the body. This process may occur at the expense of other nutrient stores, especially triacyglyceride. The oxidation of alcohol takes the place of lipids in the fuel mixture, resulting in satiety signals originating from dietary fat going undetected [3]. Suppression of hepatic fatty acid oxidation by substances such as Etoxomir has been demonstrated to increase energy intake [25], the effect of alcohol on appetite may operate through a similar mechanism. Alternatively, energy from alcohol may go undetected by the body as it is not recognised or metabolised in the same way as other dietary fuels. Suter et al. [26] reported that alcohol is associated with a transitory decrease in lipid oxidation. There may be a synergistic effect of alcohol on lipid metabolism and this is particularly evident when alcohol is taken in conjunction with a high-fat meal [27]. Metabolic and psychological accounts contribute to our understanding of the consequences of alcohol ingestion [28]. It seems that a combination of pre- and postabsorptive metabolic signals arising from the gut and the liver, together with the sensory experience of ingestion, produces the effects of alcohol on intake. This is a clear example of synergy between direct (stimulation of gut receptors) and indirect (metabolic status) controls of food intake [29] operating together to enhance appetite following alcohol intake. To conclude, the current investigation supports previous evidence that high doses of alcohol have a modest stimulatory effect on food intake; however, a dose – response effect was not observed. One unit of alcohol was below threshold because the pharmacological action was too weak to elicit effects on food intake and appetite. In contrast, 4 units of alcohol increased rated hunger and appeared to produce similar fullness ratings despite intake of a large test meal. Taken together, this suggests that the effect of alcohol on food intake operates by increasing appetite and delaying the development of satiety. Finally, energy derived from alcohol was not compensated in the short term; therefore, alcohol energy was additional to energy intake. Future investigations should elaborate both the threshold and longer-term effects of alcohol on food intake given the potential implications of alcohol intake for positive energy balance.

References [1] MAFF. National food survey 1993. London: HMSO; 1994. [2] Westerterp-Plantenga MS, Verwegen CR. The appetizing effect of an aperitif in overweight and normal-weight humans. Am J Clin Nutr 1999;69(2):205 – 12. [3] Schutz Y. Role of substrate utilization and thermogenesis on bodyweight control with particular reference to alcohol. Proc Nutr Soc 2000;59(4):511 – 7. [4] Mattes RD. Dietary compensation by humans for supplemental energy

58

[5]

[6]

[7]

[8]

[9] [10]

[11]

[12]

[13]

[14] [15]

[16]

S.J. Caton et al. / Physiology & Behavior 81 (2004) 51–58 provided as ethanol or carbohydrate in fluids. Physiol Behav 1996; 59(1):179 – 87. Colditz GA, Giovannucci E, Rimm EB, Stampfer MJ, Rosner B, Speizer FE, et al. Alcohol intake in relation to diet and obesity in women and men. Am J Clin Nutr 1991;54(1):49 – 55. Gruchow HW, Sobocinski KA, Barboriak JJ, Scheller JG. Alcohol consumption, nutrient intake and relative body weight among US adults. Am J Clin Nutr 1985;42(2):289 – 95. Fisher M, Gordon T. The relation of drinking and smoking habits to diet: the Lipid Research Clinics Prevalence Study. Am J Clin Nutr 1985;41(3):623 – 30. Jones BR, Barrett-Connor E, Criqui MH, Holdbrook MJ. A community study of calorie and nutrient intake in drinkers and nondrinkers of alcohol. Am J Clin Nutr 1982;35(1):135 – 9. Prentice AM. Alcohol and obesity. Int J Obes Relat Metab Disord 1995;19(Suppl. 5):44 – 50. Crouse JR, Grundy SM. Effects of alcohol on plasma lipoproteins and cholesterol and triglyceride metabolism in man. J Lipid Res 1984; 25(5):486 – 96. Clevidence BA, Taylor PR, Campbell WS, Judd JT. Lean and heavy women may not use energy from alcohol with equal efficiency. J Nutr 1995;125(10):2536 – 40. Poppitt SD, Eckhardt JW, McGonagle J, Murgatroyd PR, Prentice AM. Short-term effects of alcohol consumption on appetite and energy intake. Physiol Behav 1996;60(4):1063 – 70. Poppitt SD, McCormack D, Buffenstein R. Short-term effects of macronutrient preloads on appetite and energy intake in lean women. Physiol Behav 1998;64(3):279 – 85. Yeomans MR, Hails NJ, Nesic JS. Alcohol and the appetizer effect. Behav Pharmacol 1999;10(2):151 – 61. Yeomans MR, Phillips MF. Failure to reduce short-term appetite following alcohol is independent of beliefs about the presence of alcohol. Nutr Neurosci 2002;5(2):131 – 9. Hetherington MM, Cameron F, Wallis DJ, Pirie LM. Stimulation of appetite by alcohol. Physiol Behav 2001;74(3):283 – 9.

[17] Buemann B, Toubro S, Astrup A. The effect of wine or beer versus a carbonated soft drink, served at a meal, on ad libitum energy intake. Int J Obes Relat Metab Disord 2002;26(10):1367 – 72. [18] Blundell JE, MacDiarmid JI. Fat as a risk factor for overconsumption: satiation, satiety, and patterns of eating. J. Am. Diet. Assoc. 1997;97(7 Suppl.):S63 – 9. [19] Tremblay A, St-Pierre S. The hyperphagic effect of a high-fat diet and alcohol intake persists after control for energy density. Am J Clin Nutr 1996;63(4):479 – 82. [20] Lieber C. Alcohol, liver, and nutrition. J Am Coll Nutr 1991;10(6): 602 – 32. [21] Herman CP, Polivy J. Effects of alcohol on eating behavior: influence of mood and perceived intoxication. J Abnorm Psychol 1976;85(6): 601 – 6. [22] Van Strien T, Frijters JER, Bergers GPA, Defares PB. The Dutch eating behavior questionnaire (Debq) for assessment of restrained, emotional and external eating behavior. Int J Eat Disord 1986; 5(2):295 – 315. [23] Mattes RD, DiMeglio D. Ethanol perception and ingestion. Physiol Behav 2001;72(1 – 2):17 – 29. [24] Hill AJ, Magson LD, Blundell JE. Hunger and palatability: tracking ratings of subjective experience before, during and after the consumption of preferred and less preferred food. Appetite 1984;5(4):361 – 71. [25] Kahler A, Zimmermann M, Langhans W. Suppression of hepatic fatty acid oxidation and food intake in men. Nutrition 1999;15(11 – 12): 819 – 28. [26] Suter PM, Schutz Y, Jequier E. The effect of ethanol on fat storage in healthy subjects. N Engl J Med 1992;326(15):983 – 7. [27] Tremblay A, Wouters E, Wenker M, St-Pierre S, Bouchard C, Despres JP. Alcohol and a high-fat diet: a combination favoring overfeeding. Am J Clin Nutr 1995;62(3):639 – 44. [28] Yeomans MR, Caton S, Hetherington MM. Alcohol and food intake. Curr Opin Clin Nutr Metab Care 2003 (Nov.);6(6):639 – 44. [29] Smith GP. The direct and indirect controls of meal size. Neurosci Biobehav Rev 1996;20(1):41 – 6.