Family and Friends Produce Greater Social Facilitation of ... · Department of Psychology, Georgia State University, Atlanta GA 30303, USA Received 7 October 1993 DE CASTRO ... anxiety,

Pergamon

0031-9384(94)E0037-5

Physiology & Behavior, Vol. 56, No. 3, pp. 445-455, 1994 Copyright © 1994 Elsevier Science Ltd Printed in the USA. All rights reserved

0031-9384/94 $6.00 + .00

Family and Friends Produce Greater Social Facilitation of Food Intake Than

Other Companions

J O H N M. DE C A S T R O

Department of Psychology, Georgia State University, Atlanta GA 30303, USA

R e c e i v e d 7 O c t o b e r 1993

DE CASTRO, J. M. Family and friends produce greater social facilitation of food intake than other companions. PHYSIOL BEHAV 56(3) 445-455, 1994.--How the presence of other people increases the amount eaten in meals was investigated by studying the impact of different companions on the spontaneous intake of free-living humans. 515 adults were paid to maintain 7-day diaries of everything they ate or drank, the time of occurrence, self-rated hunger, anxiety, and elation, the number of other people present, and their gender and relationship to the subject. Meals eaten with other people were larger and longer in duration compared to meals eaten alone regardless of the relationship of the eating companion to the subject. However, relative to other companions, meals eaten with spouse and family were larger and eaten faster, while meals eaten with friends were larger and of longer duration. This was independent of the time of day with similar effects occurring with morning, noontime, and evening meals. In addition males produced greater social facilitation of intake in females but not in males. These results suggest that the presence of other people at a meal increases intake by extending the time spent at the meal, probably as a result of social interaction, and that family and friends have an even larger effect, probably by producing relaxation and a consequent disinhibition of restraint on intake.

Meal Pattern Eating Drinking Carbohydrate Fat Protein

THE social context is thought to be an important determinant of human and animal behavior (28,29). Field and laboratory studies have clearly demonstrated that the presence of other people can alter many different forms of behavior. This phenomenon is termed social facilitation (38). Recently we have demonstrated social facilitation of eating behavior in humans in their natural environment. We found that 44% more food is ingested when a meal is eaten with other people present than when alone (16).

We 've found social facilitation to be the most salient factor associated with short term food intake in humans. The correlation between the number of people present and the meal size is called the social correlation. It is the strongest correlation we have yet discovered with the size of spontaneously ingested meals (16). In addition, it is a very orderly phenomenon with the amount eaten increasing as a function of the number of other people present in a fashion best described by a power function (14). The presence of other people influences the meal pattern independent of other salient physiological and psychological factors. In particular, when the number of people present is added into a multiple linear regression prediction of the meal size along with the subjective state of hunger and an estimate of the premeal stomach content, the prediction of these latter factors is unchanged. How- ever, the addition of number of people present to the prediction, more than doubles the variance accounted for by the regression. Hence, the social influence appears to operate in addition to, and independent of, other salient influences on the pattern of meal ingestion (16).

Even though this evidence comes from observational data, the phenomenon does not appear to be spurious. Social facilitation of eating is not an artifact of time or place, or snacking, or of the ingestion of alcohol (15). The social correlation, the correlation between the number of people present and the meal size, has been shown to occur for meals eaten either at breakfast, lunch, or din- ner (15) or on weekdays or weekend days (10). In addition, the social correlation is significant for intake classified either as a snack or as a meal, for meals eaten at home, in restaurants, or elsewhere, and for meals ingested along with or without alcohol (15). Furthermore, when subjects were instructed to eat without other people present they reduced their intake by 212 kcal per day over a 5-day period suggesting that the effect of other people on intake is causal in nature (35).

How the presence of other people may operate to influence the amount eaten in a meal is not known. There are, however, a number of potential explanations. It may do so by producing imitation (37). This model suggests that in the presence of someone who eats a relatively large amount of food, an individual will increase intake. Indeed, subjects in the presence of a high intake " m o d e l " tend to eat large amounts while in the presence of a low intake " m o d e l " eat very little (3,32,33) even after 24-h food deprivation (2l). The presence of other people might induce an aroused state that leads to greater consumption (38,39). This explanation postulates that the presence of other people increases the individual's drive level which facilitates the emission of dominant responses. Eating in this enhanced drive state would be

445

446 ,1 ( AS'I'R~;

expected to result in an increase in the rate with which the individual emits the dominant response of ingesting food. Indeed, a pecking companion increases the peck rate of a chick (36).

Social facilitation might operate by inducing an emotional response such as increased anxiety or elation which may produce an increase in intake (22,23). This would predict that there would be an increase in self-reported anxiety or elation with meals eaten socially. Indeed, Harlow (22) observed greater emotionality in rats fed with others than when fed alone. On the other hand, the presence of other people may increase perceived hunger (37). A completely sated chick has been observed to begin eating again when placed in the presence of a hungry chick in the process of eating (1). This would predict higher levels of self-rated hunger prior to a social meal.

Social facilitation might operate by producing disinhibition; the presence of a companion relaxing the organism, thereby releasing behavior from inhibition (34). This notion would predict a greater level of calmness with social meals producing a release of restraints on eating. Finally, the presence of other people might quite simply extend the amount of time spent at a meal and thus increase the amount eaten (9). This notion predicts that the rate of intake should be the same regardless of the social conditions but the duration of the meal would be extended when other people were present. Indeed, de Castro (9) observed exactly that, with the duration and not the rate of intake increased when meals were eaten socially. This should produce a higher than normal level of stomach filling and greater satiety at the end of the meal.

These hypotheses yield different predictions regarding the influence of different companions on intake. Because males gen- erally eat more than females, an imitation or increased hunger model would predict that females would tend to eat more in the presence of a male while males would tend to eat less in the presence of a female. The aroused state or the increased emotionality explanations would predict that the more emotionally arousing the companion, the more that would be eaten. Both of these hypotheses, then, would predict that people well known to the subject, with whom they are comfortable, such as family and friends, would produce the least effect on intake. On the other hand, the disinhibition and time extension notions would predict exactly the opposite. The time extension notion suggests that the better the companion is known, the greater the likelihood of conversation and thus the greater the facilitation of intake. The disinhibition notion would likewise suggest that the better known the companion the greater the relaxation and thus the greater the facilitation of intake.

The literature does not provide a clear answer as to what is the effect of different companions on intake. While nonobese subjects tend to increases intake in the presence of others, obese subjects tend to reduce intake (27). However, Edelman et al. (20) found social facilitation of intake in the obese with normal weight companions. In addition, when an obese subject is accompanied by another obese subject they tend to increase intake while normal weight subjects are unaffected by their companions body weight (19). When paired with a desirable male females tend to eat less than when with a relatively undesirable male or a female. Males on the other hand tend to eat less when paired with females regardless of the desirability of the companion (31).

The data needed to evaluate these predictions are available in the data base collected during prior research projects (5-18) that is routinely added to with new data from ongoing research projects. These data were collected by asking adult humans to maintain a 7-day diary of their intake, social conditions, and mood. The various hypotheses of how social influences affect food intake were addressed by reanalysis of these previously collected data. Meals eaten in the presence of specific companions were

identified and compared to meals eaten with dilterent c,+mpan ions and to meals eaten alone.

METHOI)

The data used in this study were available in the data base accumulated from 1985 through 1993 during prior research projects (5 - 18). Details regarding the methods are available in these publications.

Subjects

Data were collected from 276 male and 239 female adult humans who were paid $30 to participate. They also received a detailed nutritional analysis based on their food intake for the 7- day reporting period. The subjects averaged 41.9 years (~r = 13.7), 72.1 kg (a = 9.1) in weight, 25.1 (~r = 4.1) Body Mass Index (BMI) and 1.69 m (a = .10) in height. Informed consent was obtained from all subjects.

Procedure

The subjects were given a small (8 × 18 crn) pocket sized diary and were instructed to record in as detailed a manner as possible every item that they either ate or drank, the time they ate it, the amount they ate, and how the food was prepared, and the nature of the other people eating with them, number of males and females and their relation with the subject (e.g., friend, classmate, date etc.). Self-ratings were obtained of the subjects' degree of hunger at the beginning and again at the end of the meal on a 7-point scale from very sated (1) to very hungry (7), their degree of elation from very depressed (I) to very elated (7), and their degree of anxiety from very calm (1) to very anxious (7). The subjects recorded for a day and were contacted by the experimenter to review the information, correct any problems and answer any questions. They then recorded their intake for seven consecutive days. After this recording period the subjects were contacted by the experimenter and reviewed the diaries, clarify- ing any ambiguities or missing data. After the completed diaries were submitted, two individuals who ate with the subject during the recording period were contacted and asked to verify the subjects reported intake. In some cases, difficulty was encountered in remembering exactly what the subject ate. However, in no case was the subject's diary report contradicted in either the nature or the amount reported.

Data Analysis

The foods reported in the diaries were assigned codes from a computer file of over 3500 food items, by an experienced reg- istered dietitian. The coder was unaware of the experimental hypotheses and did not interact directly with the subjects. Meals were then identified and the compositions of the individual items composing the meal were summed. For a reported intake to be classified as an individual meal it had to contain at least 50 kcal, or more stringently 100 or 200 kcal. It also had to be separated in time from the preceding and following ingestive behaviors by at least 15 min. More stringent definitions of 45 and 90 rain. were also employed. Five different definitions of a meal were used combining these minimum criteria, 15 rain/50 kcal, 45 min/50 kcal, 45 rain/100 kcal, 45 min/200 kcal, and 90 min/50 kcal.

Meals eaten alone or with five different companion types were identified; friend, family, spouse, work associate, or other (in- cluding roommate, lover, classmate, and others). Meals were also identified that were eaten with a male present or a female present. For comparison purposes, for each of these seven different categories, meals were identified that were eaten with someone other

COMPANION TYPE AND MEAL SIZE 447

O O

v

N

v)

B

I.d 69 +

v

¢-

(3 O

900 t

6 0 0

3 0 0

Intake in Meals With and Without Companion and

C a r b o h y d r a t e

Foe. r t" ,q Protein ix, x. "1 Alcohol

|

Alone

o o° o° ,o o o Oo .o,o

Go

FIG. 1. Mean (+SEM) amounts (kcal) ingested per meal of carbohydrate (solid portion of each bar), fat (cross hatched portion), protein (hatched upward) and alcohol (hatched downward) for meals eaten with a particular companion type (first bar of each set of three), with others but not that companion type present (second bar of each set of three), and alone (third bar). The • above the bar indicates a significant difference (p < .05) between the total meal sizes as assessed with a t test. The • above the first bar signifies the with companion--with other comparison, above the second bar signifies the with other--alone comparison, and above the third bar signifies the with companion--alone comparison. All meals were defined by the minimum 45 min./50 kcal criteria.

than a member of the category (e.g., for the friend category meals were identified that were eaten with someone other than a friend). These meals were then characterized by their total caloric content, carbohydrate, fat, protein, and alcohol content, duration and rate of intake, and the amount of time between meals, the premeal and postmeal intervals. The mean pre and postmeal subjective states of hunger, elation, and anxiety were also recorded. The estimated premeal and postmeal stomach contents were calculated with a computer model in which the reported intake was estimated to empty from the stomach at a rate proportional to the square root of the caloric content of the stomach, s,+j = s, -

5~s-~s~ where s equals the stomach content in little calories and n equals a particular minute of the day (2,6,24-26).

The average of each of these meal characteristics was then calculated for each subject for each companion type. These individual means were then used to calculate overall group means. Because not all subjects ate meals with each companion type, the overall means contain different, but highly overlapping, groups of subjects. Separate averages for the alone meals were calculated for the subject groups appropriate to each companion type mean. Meals eaten alone and with and without a member of each of the categories were compared with a repeated measures ANOVA. Individual comparisons were made with correlated t tests.

Univariate correlations between the number of companions present at meals and the meal size were calculated individually for each subject. Individual correlations were converted to z

scores (4) and averaged for a companion type for all of the subjects who ate meals with that companion type. Individual comparisons were made with correlated t tests. In addition, univariate regressions and multiple linear regression analyses (30) were performed on these data across all meals eaten by all subjects, using subject characteristics, the conditions of meal ingestion, and the companion types as predictors of meal size, duration, or rate.

RESULTS

Although, quantitative differences were apparent between the results obtained for the five different meal definitions, the pattern of results were equivalent and there were no significant qualita- tive differences in the results. Thus, only the minimum 50 kcal, 45 min definition is presented as representative. All statistics presented are significant at the .05 level unless otherwise indicated.

Companion Type

The average meal sizes for meals eaten with and without each of the companion types and meals eaten alone are presented in Fig. 1. There were large significant differences between the mean meal sizes (kcal) for each of the companion types; spouse, F(2, 590) = 145.60, family, F(2, 726) = 99.05, friend, F(2, 646) = 114.47, co-worker, F(2, 538) = 101.88, and others, F(2, 156) = 19.96. For all of these types, meals eaten alone were significantly smaller than meals eaten with the companion type or with others

448 ~ (AS ' IR ( )

v

D L

~J b3 +

v

bJ (/3 +

g

E3

O "I"

t~J

+ v

Meal Duration

Go ~

C,O/~"

Beginning Hour of Meals

| | I

FIG. 2. Mean (+SEM) duration of the meal (top), rate of intake (middle) and beginning of meal time (bottom) for meals eaten with a particular companion type (first bar of each set of three), with others but not that companion type present (second bar of each set of three), and alone (third bar). The • above the bar indicates a significant difference (p < .05) between the meal characteristics as assessed with a t test. The • above the first bar signifies the with companion--with other comparison, above the second bar signifies the with other--alone comparison, and above the third bar signifies the with companion--alone comparison. All meals were defined by the minimum 45 min./ 50 kcal criteria.

but without the companion type; spouse, t(295) = 19.96, 8.81, family, t(363) = 12.26, 12.83, friend, t(323) = 13.33, 13.54, co- worker t(269) = 6.18, 18.13, and others, t(78) = 3.71, 8.95 for the w i t h - - a l o n e and w i t h o u t - - a l o n e comparisons respectively. This basically replicates the previously reported differences between social meals and meals eaten alone (16). In addition, as is apparent in Fig. 1, different categories of companions have differing magnitudes of effects on the sizes of the meals ingested. Meals ingested with spouse, t(295) = 7.26, family, t(363) = 5.14, or friend, t(323) = 5.55, are significantly larger than meals ingested with others but without the companion type, while meals ingested with co-workers, t(269) = 6.87, were significantly

smaller. Exactly similar results were found for the average amounts of carbohydrate, fat, and protein ingested in the meals.

The average meal durations for meals eaten with and without each of the companion types and meals eaten alone are presented in the top portion of Fig. 2. There were highly significant differences between the mean meal durations (min) for each of the companion types; spouse, F(2, 590) = 10.27, family, F(2, 726) = 14.97, friend, F(2, 646) = 80.00, co-worker, F(2, 538) = 10.61, and others, F(2, 156) = 8.25. As was the case for meal size, for all of these types, meals eaten alone were of significantly shorter duration than meals eaten with the companion type or with others but without the companion type; spouse, t(295) =


3.57, 3.92, family, t(363) = 3.46, 5.96, friend, t(323) = 9.74, 3.74, co-worker t(269) = 2.82, 4.64, and others, t(78) = 3.38, 3.91 for the wi th - -a lone and wi thout - -a lone comparisons respectively. This, again, replicates previous results (9). However, in contrast to the differential effects of the companion types on meal size, there were no significant differences between the meal durations of social meals eaten with or without a particular companion type except for friend, t(323) = 8.66.

The average rate of intake for meals eaten with and without each of the companion types and meals eaten alone are presented in the middle portion of Fig. 2. There were significant differences between the mean rates of intake (kcal/min) for each of the companion types; spouse, F(2, 590) = 9.04, family, F(2, 726) = 7.17, friend, F(2, 646) = 13.67, and co-worker, F(2, 538) = 5.42, except for others. The larger meal sizes in the spouse and family categories were due to higher rates of intake than occurred with social meals eaten without the companion type, t(295) = 4.17, t(326) = 4.26, respectively, while the smaller meal sizes eaten with co-workers resulted from slower rates of intake t(269) = 3.17. More was eaten with friends due to the marked extension of meal duration as the intake rate was actually less, t(323) = 6.30.

As is often the case with observational evidence, there is a third factor, time of day, that covaries with companion type. The average beginning hour of the day for meals eaten with and without each of the companion types and meals eaten alone are presented in the bottom portion of Fig. 2. There were significant differences between the mean beginning of meal time (hour of the day) for each of the companion types; spouse, F(2, 590) = 167.92, family, F(2, 726) = 100.01, friend, F(2, 646) = 88.99, co-worker, F(2, 538) = 181.52, and others. F(2, 156) = 15.21. In general, the pattern is equivalent to that seen with the meal sizes.

Morning, Noontime, and Evening Meals and Companion Type

Because meals tend to get larger throughout the day (6) and also tend to occur with different types of companions, the observed differences in meal sizes with the companion types could be an artifact produced by this covariation. To investigate this possible artifact meals were separated into morning (0600- 1000), noontime (1100-1300), and evening (1700-2100) categories and meals eaten with the companion types during each of these periods were compared and are presented in Fig. 3. In- spection of Fig. 3 reveals a very similar pattern of results regardless of time of day. Even when the results are nonsignificant they are in the same direction as those described over the entire day.

There were significant differences between the mean meal sizes (kcal) for each of the companion types, except for others, for morning, noontime, and evening meals; spouse, F(2, 220) = 10.02, F(2, 300) = 22.03, F(2, 200) = 11.14, family, F(2, 176) = 4.26, F(2, 332) = 26.01, F(2, 272) = 18.09, friend, F(2, 136) = 5.35, F(2, 262) = 18.59, F(2, 216) = 12.64, and co-worker, F(2, 144) = 3.91, F(2, 296) = 17.95, F(2, 58) = ns, respectively. For all of these types, except others, meals eaten alone were significantly smaller than meals eaten with the companion type or with others but without the companion type, for morning, noontime, and evening meals; spouse, t(110) = 4.52, ns, t(151) = 7.23, 4.51, t(100) = 6.43, 2.47, family, t(88) = 3.57, ns, t(166) = 6.80, 6.52, t(136) = 5.92, 3.66, friend, t(68) = 3.05, 2.23, t(131) = 5.43, 5.50, t(108) = 4.20, 2.60, co-worker, t(72) = ns, 3.46, t(148) = 4.74, 6.12, t(29) = ns, ns, respectively. Addition- ally, different categories of companions had differing magnitudes of effects on the sizes of the meals regardless of whether the meals were ingested in the morning, noontime, or evening. Meals

ingested with spouse, t(110) = 2.44, t(150) = 3.02, t(100) = ns, family, t(88) = ns, t(166) = 2.14, t(136) = 2.31, or friend, t(68) = 2.12, t(131) = 2.00, t(108) = 3.98, respectively, are significantly larger than meals ingested with others but without the companion type, while meals ingested with co-workers, t(72) = 2.50, t(148) = 2.28, t(29) = 2.38, respectively, were significantly smaller.

Subjective Self-Ratings

Figure 4 presents the mean subjective self-ratings of pre and postmeal hunger (top), elation (middle), and anxiety (bottom) at the time of meals eaten with the different companion types. There were highly significant differences between the mean self-ratings for each of the companion types except others both pre and postmeal; spouse, F(2, 590) = 14.62, 47.31, 18.84, 19.33, family, F(2, 726) = 22.33, 47.12, 20.92, 3.04, friend, F(2, 646) = 15.99, 35.03, 61.84, ns, co-worker, F(2, 538) = 8.56, 26.68, 22.46, 13.85, and others, F(2, 156) = ns, ns, 7.83, 3.06, for pre and postmeal hunger, elation, and anxiety, respectively.

Eating with their spouse was associated with greater premeal hunger, t(295) = 3.15, 5.95, lower postmeal hunger, t(295) = 4.31, 9.67, and lower anxiety, t(295) = 5.57, 4.78, than eating with others or alone, respectively, and greater elation in comparison to the alone condition, t(295) = 5.16. Eating with family was associated with greater premeal hunger, t(363) = 5.23, 6.41, and lower postmeal hunger, t(363) = 3.40, 8.65, than eating with others or alone, and lower anxiety, t(363) = 2.55, and greater elation, t(363) = 5.50, in comparison to the alone condition.

Eating with friends was associated with greater elation, t(323) = 5.94, 9.99, than eating with others or alone, and lower premeal hunger, t(323) = 4.31, than with others, and lower postmeal hunger, t(323) = 7.59, than when alone. Eating with coworkers was associated with lower anxiety and postmeal hunger, t(269) = 5.08, 2.87, than eating with others respectively, and greater premeal hunger, t(269) = 3.46, lower postmeal hunger, t(269) = 3.76, and greater anxiety and elation, t(269) = 2.57, 4.90 than eating alone. Eating with others was associated with greater elation, t(78) = 2.88 than when alone and greater anxiety, t(78) = 2.03 than when with others.

Companion Gender

The average meal sizes for meals eaten with opposite sex and same sex companions and meals eaten alone are presented in Fig. 5. There were highly significant differences between the mean meal sizes (kcal) for all subjects, F(2, 656) = 58.06, for males, F(2, 332) = 26.35, and for females, F(2, 324) = 33.55. For all comparisons, meals eaten alone were significantly smaller than meals eaten with opposite or same sex companions; all subjects t(328) = 11.07, 8.22, males, t(166) = 7.32, 6.03, and females, t(162) = 5.73, 8.12. Males and females responded differently to the presence of same and opposite sex companions resulting in significant Gender x Companion Gender interactions both when the alone condition was included, F(2, 654) = 3.28, and when it was not, F(1,335) = 4.16. Whereas meal sizes were equivalent for males eating with either males or females, they were significantly larger for females eating with males than with females, t(162) = 2.92. Hence the presence of males appears to produce a greater social facilitation of intake in females but not in males.

Regression Analyses

The univariate correlations between the number of companions present at meals and the meal size are presented in Fig. 6 separately for male and female subjects. As is apparent, males

45/) I~t ( 'ASTR()

v

N ~

W u3 +

v

~3

0J

Ld U I +

v

900- M o r n i n g M e e l 8

m

600--

3O0--

" - - - T . . . . . . . o "y'~ o~.~'r~o £9o o o ~'d t~o C~O

9oc~ N o o n t i m e M e a l s

C,o/"

f Go

FIG. 3. Mean (+SEM) amounts (kcal) ingested per meal of carbohydrate (solid portion of each bar), fat (cross hatched portion), protein (hatched upward) and alcohol (hatched downward) for meals eaten in the morning (0600 to 1000, top), noontime (1100 to 1500, middle), and evening (1700 to 2100, bottom), for meals eaten with a particular companion type (first bar of each set of three), with others but not that companion type present (second bar of each set of three), and alone (third bar). The • above the bar indicates a significant difference (p < .05) between the total meal sizes as assessed with a t test. The • above the first bar signifies the with compan- i o n - w i t h other comparison, above the second bar signifies the with other--alone comparison, and above the third bar signifies the with companion--alone comparison. All meals were defined by the minimum 45 min./50 kcal criteria.

and females did not differ in their response to a companion eating with them. As has been observed in the past (16), there were clear and significant correlations between the total number of other people present at a meal and the size of the meal. However, there were differences in the magnitude of the responses with different gender companions. There was a stronger response to an opposite sex than a same sex companion, F (1 ,495) = 41.80, for both males and females. In addition, there were marked differences in the magnitudes of the responses with the five companion types, F(4, 1980) = 73.74. Whereas the correlations for

spouse, family, and friends were significant, positive, and did not differ, the correlations for co-workers and others were tiny or nonsignificant and significantly differed (co-worker, t(497) = 10.05, 11.83, 12.64; other, t(497) = 11.91, 11.74, 8.70) from those for spouse, family, and friend, respectively.

In order to remove the variance due to a number of covariates in the analyses, multiple linear regressions were performed across all of the 9277 meals reported by all subjects employing 16 independent variable predictors of meal size. The ~ coefficients from this regression are presented in Figure 7. Overall, the re-


E

n

LU 01 +

0 Q.

3~

I

~J 03 +

Pre - & P o s t m e a l Sa ted - Hungry 8 e l f - r a t i n g

° ' ~ . , ~ o o ~

c_Y

~o~ o ~ ~o~ ~o~°'~'-~ °~° ~,~ o ~ ~° .~o o ~ o ~ o ~ ° Go ~

O~ r

I

g~

kd O3 +

Go I

FIG. 4. Mean (+SEM) subjective self-ratings of hunger (top) with the premeal ratings represented by the bar above 0 and the postmeal ratings by the bars below 0, elation (middle), and anxiety (bottom) for meals eaten with a particular companion type (first bar of each set of three), with others but not that companion type present (second bar of each set of three), and alone (third bar). The • above the bar indicates a significant difference (p < .05) between the self-ratings as assessed with a t test. The [] above the first bar signifies the with companion--with other comparison, above the second bar signifies the with other--alone comparison, and above the third bar signifies the with companion--alone comparison. All meals were defined by the minimum 45 min./ 50 kcal criteria.

gression was significant, F(15, 9261) = 149.54, and produced a multiple correlation of .453. Significant/~ coefficients were produced for all of the subject characteristic variables of gender, age, weight, height, and Body Mass Index, and for the conditions of the meal of hunger, prior interval duration, and prior stomach content, but not the beginning of meal time. There were significant /~ coefficients for the companion types of spouse, family, and friend, but not co-worker or other, and there was a significant influence of the number of opposite sex individuals present but not same sex individuals. Hence, even taking into consideration a large number of other variables, the presence of certain com-

panion types has a significant positive relationship with the amount eaten in meals.

DISCUSSION

The present study again demonstrates social facilitation of intake as in prior studies (9,10,14-16,35). Meals were larger and longer in duration regardless of the nature of the eating companion. However, different companion types had differing magnitudes of effects on the subjects' intake. In comparison to meals eaten with other companions, meals eaten with spouse and family

p ~

0 (D

i v

N

b5

©

bJ 09 +

C u ©

7 5 0 -

5 0 0

2 5 0

Intake and All Subjects

Companion M a l e s

Gender

i C a r b o h y d r a t e

i i I f F a t

I / / ~ P r o t e i n

A l c o h o l

II

F e m a l e s

0

f S FIG. 5. Mean (+SEM) amounts (kcal) ingested per meal of carbohydrate (solid portion of each bar), fat (cross hatched portion), protein (hatched upward) and alcohol (hatched downward) for meals eaten with a companion of a particular gender (first bar of each set of three), the other gender (second bar of each set of three), or alone (third bar). The • above the bar indicates a significant difference (p < .05) between the total meal sizes as assessed with a t test. The • above the first bar signifies the comparison with the second bar, above the second bar signifies the comparison with the third bar, and above the third bar signifies the comparison with the first bar. All meals were defined by the minimum 45 minJ50 kcal criteria.

c o

k_ o

o

LO O0 +

v

C o

0 . 4

0.3

0.2

0.1

0 .0

- 0 . 1

M e a l S i z e C o r r e l a t i o n s

# P e o p l e S a m e O p p o s i t e S e x ~ e x

|

| • i

w i t h N u m b e r of C o m p a n i o n s P r e s e n t

spouse Family Friend Co-worker Other

• | i II |

i II i II

|

FIG. 6. Mean +_ SEM z transformed correlation coefficients between the amounts ingested in the meals and the number of other people present, in total, of the same sex, opposite sex, or in the spouse, family, friend, co-worker, or other companion category. The mean for male subjects is represented by the first bar of each set of 2 while for female subjects is represented by the second. All meals were defined by the minimum 45 min,/50 kcal criteria. • indicates that the mean is significantly (p < .05) different from zero as assessed with a t test.


0 . 2 5 - -

0 . 2 0 - -

o 0.15 - (D

.~ o . to -

o,o5 -

0 . 0 0

ne

- ~ - - 0 . 0 5 -

- 0 . 1 0 - -

- o ~ 5 -

o "o

o - 0 . 2 0 - -

i

- 0 . 2 5

Meal Size Multiple Regression

lOli Beta Coefficients

!! ! !! ~£!"~

~ "~ "~ ~ - ~ ® g ~ ~ ~ ¢ ~ ~ o o

~ "C o & o ~ 0

o ~

FIG. 7. Mean __. SEM/~ coefficients from the multiple linear regressions predicting the amount ingested in the meals on the basis of 16 independent variables. All meals from all subjects were included in the regression. All meals were defined by the minimum 45 rain./50 kcal criteria. • indicates that the fl coefficient is significantly (p < .05) different from zero as assessed with a t test.

were larger and eaten at a higher rate, while meals eaten with friends were larger and of longer duration. This effect of companion type was shown to be independent of the time of day with similar effects occurring with morning, noontime, and evening meals. Male companions at meals were found to produce a greater social facilitation of intake for females but not for males. In addition, social correlations, in both univariate and multivariate analyses, were found to be significant only for spouse, family, and friends but not for other companion types.

The results of the present study do not support an imitation explanation of social facilitation (37). This model suggests that an individual would alter their intake in the direction of a companion's intake, eating more with a high intake companion and less with a low intake companion. Hence, women should eat more when in the presence of relatively high intake males while men should eat less when in the presence of relatively low intake females. In the present study, females ate more with males than with females as predicted by the model. But contrary to the model ' s prediction, males ate the same amount regardless of whether they ate with males or females. This was not due to males being less responsive to social influences than females, as males ate 36% more in social conditions than when alone while females ate 40% more and the two genders had social correlations of equivalent magnitude.

The results of the present study also do not support either an aroused state (38,39) or increased emotionality (22,23) explanation. These models predict that there would be an increase in the rate of intake in social conditions. In fact, the only significant differences between the rate of intake when alone as compared

to eating with a companion were for friends and co-workers, where the rates were actually lower. Further, these models would predict that the more emotionally arousing the companion, the more that would be eaten. Hence, eating with people well known to the subject, with whom they are comfortable, such as family and friends, should produce the least effect on intake. In fact, they had the greatest impact on eating. In addition, the increased emotionality explanation would predict that when social facilitation occurs it should be accompanied by increased elation and/ or anxiety, and when it doesn't there should be no change in elation or anxiety. In fact, the self-ratings obtained with meals eaten with spouse, family, or friends, where social facilitation occurred, indicated less or equivalent anxiety. In addition, meals eaten with co-workers and others, where no social facilitation was apparent, were eaten with higher levels of elation and anxiety.

The results of the present study also do not support an increased hunger explanation (37). This model predicts that self- rated hunger would be higher at the beginning of socially facil- itated meals but should be about the same at the end. In fact, in all cases of social facilitation the people ate to a significantly higher level of satiety at the end of the meal than when alone. In addition, when meals were eaten with friends, where marked social facilitation occurred, there were equivalent levels of premeal hunger as when alone.

The results tend to support the disinhibition (34) and time extension models (9) of social facilitation. The disinhibition notion predicts that the better known the companion the greater the relaxation and thus the greater the facilitation of intake. Similarly,

454 ~ (AS'I 'R~)

time extension suggests that the better known the companion, the greater the likelihood of conversation and thus the greater the facilitation of intake. The facts that there were longer meal durations for all companion types and that social facilitation was greatest with spouse, family, and friends clearly supports these hypotheses.

Although highly speculative, it is possible that both of these processes occur and account for social facilitation of intake. Time is extended, regardless of companion type resulting in increased intake with all companions. However, the multiple regression suggests that even when the number of people present is consid- ered in the regression certain companion types continue to be associated with higher intake. These people, spouse, family, and friends, are the ones with which the subject should feel the most comfortable. According to the disinhibition model, these people tend to have a relaxing effect on the subject and thereby increase intake by releasing inhibitions. The fact that where anxiety levels are the highest, meals with co-workers and other companions,

social facilitation is the smallest tends to support ~uch an inter pretation.

The present results reinforce the notion that social facilitation of food intake is a powerful and ubiquitous phenomenon. They demonstrate that it occurs regardless of the nature of the companions, primarily by extending the amount of time spent eating a meal. However, these results further demonstrate that family and friends exert an effect beyond that of other companions, probably as a result of a release of inhibitory restraints on intake.

ACKNOWLEDGEMENTS

The author acknowledge the substantial contributions of Dixie K. El- more, Sandor Goldstein, Sara Orozco, Margaret Pedersen, and Marie Redd without whose assistance the work could not have been performed. Sup- ported in part by Grant R01-DK39881 from the National Institute of Diabetes and Digestive and Kidney Diseases, from a grant from the National Live Stock and Meat Board, and from a Biological Research Support Grant.

REFERENCES

1. Bayer, E. Beitraege zur Zweikomponententheorie des Hungers. Z. Tierpsychol. 118:283-349; 1929.

2. Booth, D. A, A simulation model of psychobiosocial theory of human food-intake control. Int. J. Vitam. Nutr. Res. 58:119-134; 1988.

3. Conger, J. C.; Conger, A. J.; Costanzo, P. R.; Wright, K. L.; Matter, J. A. Effects of social cues on the eating behavior of obese and normal subjects. J. Pers. 48(2):258-271; 1980.

4. de Castro, J. M. Meal pattern correlations: facts and artifacts. Physiol. Behav. 15:13-15; 1975.

5. de Castro, J. M. Macronutrient relationships with meal patterns and mood in the spontaneous feeding behavior of humans. Physiol. Be- hav. 39:561-569; 1987.

6. de Castro, J. M. Circadian rhythms of the spontaneous meal patterns, macronutrient intake, and mood of humans. Physiol. Behav. 40:437- 446; 1987.

7. de Castro, J. M. A microregulatory analysis of spontaneous fluid intake by humans: Evidence that the amount of liquid ingested and its timing is mainly governed by feeding. Physiol. Behav. 43:705- 714; 1988.

8. de Castro, J. M. Physiological, environmental, and subjective determinants of food intake in humans: A meal pattern analysis. Physiol. Behav. 44:651-659; 1988.

9. de Castro, J. M. Social facilitation of duration and size but not rate of the spontaneous meal intake of humans. Physiol. Behav. 47(6):1129-1135; 1990.

10. de Castro, J. M. Social facilitation of the spontaneous meal size of humans occurs on both weekdays and weekends. Physiol. Behav. 49(6):1289-1291; 1991.

11. de Castro, J. M. Seasonal rhythms of human nutrient intake and meal patterns. Physiol. Behav. 50(I ):243-248; 1991.

12. de Castro, J. M. Weekly rhythms of spontaneous nutrient intake and meal pattern of humans. Physiol. Behav. 50(4):729-738; 1991.

13. de Castro, J. M. The effects of the spontaneous ingestion of particular foods or beverages on the meal pattern and overall nutrient intake of humans. Physiol. Behav. 53(6): I 133-1144; 1993.

14. de Castro, J. M.; Brewer, E. M. The amount eaten in meals by humans is a power function of the number of people present. Physiol. Behav. 51(1):121-125; 1992.

15. de Castro, J. M.; Brewer, E. M.; Elmore, D. K.; Orozco, S. Social facilitation of the spontaneous meal size of humans is independent of time, place, alcohol, or snacks. Appetite 15:89-101; 1990.

16. de Castro, J. M.; de Castro, E. S. Spontaneous meal patterns in humans: influence of the presence of other people. Am. J. Clin. Nutr. 50:237-247; 1989.

17. de Castro, J. M.; Elmore, D. K. Subjective hunger relationships with meal patterns in the spontaneous feeding behavior of humans: Evi- dence for a causal connection. Physiol. Behav. 43:159-165; 1988.

18. de Castro, J. M.; McCormick, J.; Pedersen, M.; Kreitzman, S. N. Spontaneous human meal patterns are related to preprandial factors regardless of natural environmental constraints. Physiol. Behav. 38:25-29; 1986.

19. De Luca, R. V.; Spigelman, M. N. Effects of models on tood intake of obese and nonobese female college students, Can. J. Behav. Sci. 11(2):124-129; 1979.

20. Edelman, B.; EngeU, D.; Bronstein, P.; Hirsch, E. Environmental effects on the intake of overweight and normal weight men. Appetite 7:71-83; 1986.

21. Goldman, S. J.; Herman, C. P.; Polivy, J. Is the effect of a social model on eating attenuated by hunger? Appetite 17:129-140; 1991.

22. Harlow, H. F. Social facilitation of feeding in the albino rat. J. Genet. Psychol. 43:211-221 ; 1932.

23. Harlow, H. F.; Yudin, H. C. Social behavior of primates 1. Social facilitation of feeding in the monkey and its relation to attitudes of ascendance and submission. J. Comp. Physiol. Psychol. 16:171-- 185; 1933.

24. Hopkins, A. The pattern of gastric emptying: a new view of old results. J. Physiol. (Lond) 182:144-150; 1966.

25. Hunt, J. N.; Knox, M. T. Regulation of gastric emptying. In: Code, C. F.; Heidel, W.; eds. Handbook of physiology: Alimentary canal, vol. 4: Motility. Washington D.C.: American Physiological Society. 1917- •935; 1968.

26. Hunt, J. N.; Stubbs, D. F. The volume and content of meals as determinants of gastric emptying. J. Physiol. (Lond) 245:209-225: 1975.

27. Krantz, D. S. A naturalistic study of social influences on meal size among moderately obese and nonobese subjects. Psychosom. Med. 41(1):19-27; 1979.

28. Latane, B. The psychology of social impact. Am. Psychol. 36:343- 355; 1981.

29. Latane, B.; Harkins, S. Cross-modality matches suggest anticipated stage freight a multiplicative power function of audience size and status. Perception and Psychophysics 20(6):482-488; 1976.

30. Lunneborg, C. E.; Abbott, R. D. Elementary multivariate analysis for the behavioral sciences: Applications of basic structures. N.Y.: Elsevier Science Publishing; 1983.

31. Mori, D.; Chaiken, S.; Pliner, P. "Eating lightly" and the self-pre- sentation of femininity. J. Pers. Soc. Psych. 53:693-702; 1987.

32. Nisbett, R. E.; Storms, M. D. Cognitive and social determinants of food intake. In: London, H.; Nisbett, R. E., eds. Thought and feeling: Cognitive alteration of feeling states. Chicago: Aldine Publishing Co.; 1974.

33. Polivy, J.; Herman, C. P.; Younger, J. C.; Jaeger, J. Effects of a model on eating behavior: The induction of a restrained eating style. J. Pers. 47(1):100-117; 1979.

34. Rajecki, D. W.; Kidd, R. F.; Wilder, D. A.; Jaeger, J. Social factors in the facilitation of feeding in chickens: Effects of imitation,


arousal, or disinhibition? J. Pets. Soc. Psychol. 32(3):510-518; 1975.

35. Redd, E. M.; de Castro, J. M. Social facilitation of eating: Effects of instructions to eat alone or with others. Physiol. Behav. 52(4):749-754; 1992.

36. Tolman, C. W. Social facilitation of feeding behaviour in the do- mestic chick. Anim. Behav. 12:245-251; 1964.

37. Tolman, C. W. The role of the companion in social facilitation of animal behaviour. In: Simmel, E. C.; Hoppe, R. A.; Milton, G. A., eds. Social facilitation and imitative behaviour. Boston: Allyn and Bacon; 1968:33-54;.

38. Zajonc, R. B. Social facilitation. Science 149:269-274; 1965. 39. Zajonc, R. B. Compresence. In: Paulus, P. B., ed. Psychology of

group influence. Hilisdale, N.J.: Erblaum Press; 1980:35-60.

Family and Friends Produce Greater Social Facilitation of ... · Department of Psychology, Georgia State University, Atlanta GA 30303, USA Received 7 October 1993 DE CASTRO ... anxiety,

Documents