Marginal interest posts (was: Saccharin bitterness predicts overactivity)

[Guest guest]

Al: The relationship of these to CRON is marginal and these papers are very technical. It would be appreciated if you would weed out those papers that have wider appeal and only post those; I am guessing that these below have mimimal interest to our group. Let's try to concentrate on what's important (such as studies directly related to better health and longer life).

That said, of course we all appreciate the effort you put into your posts.

on 8/13/2005 7:02 PM, Al Pater at old542000@... wrote:

Hi All,

Saccharin was used in the pdf-available below paper to test the bitterness effect on

behavior dependent on stress.

Are you sensitive to bitterness and what is your reaction to stress?

How do we explain " the puzzling relationship between food consumption and exercise

patterns, more specifically, the relationship between restricted food consumption

and excessive exercise " ?

Int J Eat Disord. 2003 Jul;34(1):71-82.

The bitter truth: sensitivity to saccharin's bitterness predicts overactivity in

highly arousable female dieters.

Craig ML, Hollis KL, Dess NK.

OBJECTIVE: The interaction between taste sensitivity and emotionality in rats

provides a provocative view of hyperactivity. Rats that have been bred selectively

for their reactivity to saccharin exhibit characteristic emotionality. When placed

on restrictive diets, these rats exhibit excessive activity levels, relative to rats

that are not sensitive to saccharin. Because humans who are highly arousable (i.e.,

reactive to environmental stimuli) also exhibit an increase in sensitivity to

saccharin's bitterness, the current study evaluated whether women who are highly

arousable, currently dieting, and sensitive to saccharin's bitterness engage in

excessive exercise. ... RESULTS AND DISCUSSION: As hypothesized, sensitivity to

saccharin's bitterness predicted overactivity in highly arousable female dieters,

which reveals the multidimensionality of activity anorexia. PMID: 12772172

INTRODUCTION

Human studies of physical activity provide several insights into the puzzling

relation-ship

between food consumption and exercise patterns, more specifically, the relationship

between restricted food consumption and excessive exercise. For example, cadets

reported consuming less food on days when activity was high compared with days of

lower activity (Edholm, Fletcher, Widdowson, & McCance, 1955). The number of

calories

consumed by cadets on high-activity days was significantly lower than on days with

less

activity. Similarly, obese children voluntarily reduced the amount of food consumed

during lunch following a prelunch exercise period (Epstein, Masek, & Marshall, 1978)

and workers who engaged in light to medium work ate less than sedentary workers

(Mayer, Roy, & Mitra, 1956). These data suggest an inverse relationship between food

consumption and physical activity in the short term. However, because these studies

necessarily must rely on correlational data, animal models have been useful in

controlled

evaluations of this relationship.

Animal models in which a restricted food schedule results in excessive activity and

paradoxically reduces food intake suggest many commonalties with the relationship

between food restriction and exercise patterns in humans. For example, to explore

the

relationship between food schedule and physical activity in rats, food was made

avail-able

either continually or only for 1 hr per day, and a running wheel was either freely

available or locked (Epling, Pierce, & Stefan, 1981). All rats were housed in the

running

wheel apparatus with an attached side cage. They were permitted free access to

either space,

except for those on the restricted food schedule, which were denied access to the

wheel

during feeding time. When food was restricted and the running wheel was available,

animals showed a dramatic decrease in food consumption compared with control groups.

Among rats that have been bred selectively for differential intake of saccharin

solu-tions,

Dess, Arnal, Chapman, Siebel, VanderWeele, and Green (2000) have shown that

rats that are more sensitive to bitterness in complex tastes (i.e.,

low-saccharin–consuming

rats [LoS]) are significantly more anxious and emotionally reactive, and more likely

to

develop deprivation-induced hyperactivity than rats that do not avoid the substance

(i.e.,

high-saccharin–consuming rats [HiS]). When young adult female LoS and HiS rats are

placed on restrictive diets (i.e., 1 hr per day) and a standard running wheel is

available,

activity increases in both lines. However, the increase in activity is more robust

in LoS

rats than in HiS rats. These findings support the notion that taste and affect, as

well as

restrictive diets, collectively trigger excessive activity.

This shared mechanism has spurred human investigations. Dess and Edelheit (1998)

and Dess (1991; 1993) hypothesized that individual differences in emotionality, as

well as

stress, should augment saccharin’s bitter taste in humans. To test this hypothesis,

they

gave college students a set of unsolvable anagrams to work on while a startling

noise was

produced, which acted as a mild stressor. Following this task, participants rated

the

bitterness and sweetness of four saccharin solutions. A week later, the participants

completed a temperament questionnaire that measured pleasure, dominance, and

arousability, a dimension of emotionality. As predicted, high emotionality (i.e.,

high

arousability), as well as stress, enhanced saccharin’s aversive taste.

Although the previous study (Dess & Edelheit, 1998) evaluated the complex

relation-ship

among taste, emotionality, and stress, it did not consider eating or exercise

patterns.

Consequently, the current study extended Dess and Edelheit’s (1998) study by

evaluating

whether emotionality, restrained eating, and taste predicted obligatory exercise,

which is

a measure of physical activity. Participants completed a self-administered

questionnaire

packet that assessed trait arousability, eating patterns, and exercise patterns. On

another

occasion, they completed a body contour drawings handout, and their weight and

height

were measured. They also rated the bitterness and sweetness of five saccharin

solutions

following a stressful event. We predicted that women who are highly arousable,

cur-rently

dieting, and sensitive to saccharin’s bitterness would engage in excessive exercise.

We were also curious whether stress would modulate sensitivity to saccharin’s taste,

as it

had in a smaller, mixed-sex sample (Dess & Edelheit, 1998).

.... RESULTS

....Dieting, Taste, and Affect as Predictors of Exercise

The new and interesting questions addressed by this study concern the utility of

individual differences in taste and temperament as predictors of obligatory

exercise.

The first set of analyses focused on individual differences as predictors.

Sample and population means and standard deviations for the five questionnaires

used in this current study are shown in Table 2. The means and standard deviations

for the Trait Pleasure-displeasure, Trait Dominance-submissiveness, and Trait

Arous-ability

scales were similar to the population means for each scale (Mehrabian, 1994a,

1994b, 1994c, respectively ). Population means for the Restrained Eating Inventory

and

the Obligatory Exercise Questionnaire are not available.

Although Mehrabian (1996) asserted that the three dimensions of the

Pleasure-Arousal-

Dominance (PAD) temperament model are nearly orthogonal, bivariate correla-tions

among the three dimensions in the sample revealed significant associations.

For example, Arousability was associated negatively with Pleasure-displeasure and

Dominance-submissiveness was associated positively with Pleasure-displeasure,

rs(235)

=.207 and -0.153, ps <.05, respectively.

Dess et al. (2000) found that young adult female LoS rats, which are more

emotionally

reactive than young adult female HiS rats, develop a more robust case of

deprivation-

induced hyperactivity when access to food is limited. In the current study, the

interaction

of dieting, affect, and taste was expected to predict exercise patterns. Tests of

this

prediction began with an evaluation of pairwise correlations between the total score

for

obligatory exercise and other measures, including restrained eating, trait

arousability,

pleasure, and dominance, as well as ratings of saccharin’s bitterness and sweetness

at the

highest saccharin concentration. Obligatory exercise was correlated significantly

and

positively with restrained eating, arousability, bitterness, and sweetness,

rs(209)=.427,

..136, .135, and .138, respectively, ps </=.05.

The strong relationship between restrained eating and obligatory exercise could be

primary, with the other predictors of obligatory exercise deriving from it, as would

be the

case if restricted eating accounted for the taste ratings. This idea was tested in

two ways.

First, the pairwise correlations between restrained eating and the other predictors

were

examined. These tests yielded quite a different pattern from obligatory exercise.

Restrained eating was not correlated with taste ratings: bitterness, r (210)=.021;

sweetness, r (210)=.067, ps >.05. It was, however, correlated with all three

temperament

measures, with high restraint being associated with an anxiety pattern (Mehrabian,

1987),

that is, predicted by high arousability, r (210)=.296; low pleasure, r (210)=.145;

and

low dominance, r (210)=.147. These results indicate that the relationship of

obligatory

exercise to taste and affect is not simply a byproduct of restricted eating. In

particular,

unlike restricted eating, obligatory exercise is predicted by taste ratings and,

among the

temperament measures, by only arousability.

Next, hierarchical regression was used in an attempt to better account for

obligatory

exercise. Two four-step regressions were performed. In each regression, stress

condition

and trait pleasure were entered at the first step as control variables. At Step 2,

restrained

eating, a taste rating (either bitterness or sweetness), and arousability were

entered.

At Step 3, the two-way interactions between those variables were entered. Finally,

at

Step 4, the three-way interaction (Restrained eating Taste Arousability) was

entered. Results for the hierarchical regression using bitterness ratings, including

beta

values and standard errors, incremental R2s, and significance tests, are shown in

Table 3.

Table 3. Hierarchical regression analysis using restrained eating, arousability, and

bitterness rating to predict obligatory exercise (N=212)

..............................

-------B SE t

..............................

Step 1

Stress condition 0.36 1.33 0.27

Trait pleasure 0.03 0.03 1.01

Incremental R2=0.005, F(2, 208)=0.54

Step 2

Restraint 0.23 0.04 6.72*

Arousability 0.01 0.02 0.52

Bitterness rating 0.07 0.04 2.18*

Incremental R2=0.214, F(5, 205)=18.70*

Step 3

Restraint × Arousability –0.001 0.001 –1.01

Restraint × Bitterness –0.001 0.002 –0.52

Arousability × Bitterness –0.0005 0.001 –0.48

Incremental R2=0.006, F(8, 202)=0.55

Step 4

Restraint × Arousability Bitter –0.0001 0.00001 –2.02*

Incremental R2=0.015, F(9, 201)=4.09*

...........................

*p <.05.

The control variables were not significant predictors. At Step 2, both restrained

eating

and bitterness were significant positive predictors of obligatory exercise. None of

the

two-way interactions at Step 3 were significant, but the three-way interaction at

Step 4

was significant.

The three-way interaction was interpreted by repeating the regression separately

for participants above versus below the median bitterness rating (Table 4). For both

participants with relatively low bitterness ratings and relatively high bitterness

ratings, restrained eating was a significant predictor of obligatory exercise.

However,

the interaction between restrained eating and arousability was significant only

among

participants with relatively high bitterness ratings. The interaction in this

subgroup

was interpreted following the procedure recommended by Cohen and Cohen (1983).

Specifically, the interaction regression equation was solved for a hypothetical low

arousability score and a hypothetical high arousability score. This analysis

revealed a

positive association between restrained eating and obligatory exercise at both low

and

high arousability levels, but with the slope decreasing and the y-intercept

increasing

at higher arousability levels. This pattern of an increasing but flattening function

may

indicate saturation, or a ‘‘ceiling effect,’’ at high levels of bitterness,

arousability, and

restrained eating.

Table 4. Hierarchical regression analysis using restrained eating, arousability, and

bitterness rating to predict obligatory exercise among participants with low and

high bitterness ratings (N=212)

.........................

----B SE t

.........................

I. Low-bitterness participants (N=105)

Step 1

Stress condition 1.72 1.96 0.88

Trait pleasure 0.06 0.05 1.35

Incremental R2=0.023, F(2, 102)=1.19

Step 2

Restrained eating 0.24 0.05 4.77*

Arousability 0.04 0.03 1.24

Incremental R2=0.228, F(4, 100)=15.23*

Step 3

Restraint × Arousability 0.001 0.001 0.78

Incremental R2=0.005, F(5, 99)=8.38

II. High-bitterness participants (N=106)

Step 1

Stress condition –0.36 1.84 –0.19

Trait pleasure –0.01 0.05 –0.29

Incremental R2=0.001, F(2, 103)=0.06

Step 2

Restrained eating –0.21 0.05 –4.35*

Arousability 0.003 0.02 0.14

Incremental R2=0.17, F(4, 101)=10.08*

Step 3

Restraint × Arousability –0.003 0.001 –2.29*

Incremental R2=0.04, F(5, 100)=5.26*

Regression equation for interaction=0.345r + 0.203a – 0.003ra + 22.22

Solution for arousability score 1 SD below mean = 0.357r + 21.49

Solution for arousability score 1 SD above mean = 0.120r + 36.52

..............................

*p <.05.

The four-step hierarchical regression was repeated using sweetness ratings in place

of

bitterness ratings. Restrained eating was a significant predictor of obligatory

exercise at

Step 2, incremental R2=.204, F(5, 205)=17.62, p <.05. However, sweetness rating,

neither by itself nor interacting with other variables, accounted for a significant

amount

of variance in obligatory exercise.

These analyses reveal the multidimensionality of activity anorexia. Other factors

(i.e.,

arousability and sensitivity to saccharin’s bitterness) interact with restrained

eating to

predict excessive exercise. By itself, arousability does not predict obligatory

exercise but

moderates the relationship among obligatory exercise, restrained eating, and

bitterness.

Although the amount of additional variance accounted for is modest, it suggests that

investigation of factors other than restrained eating would further our

understanding of

activity anorexia. These results serve as an excellent springboard for future

studies of the

complex relationship between dispositional variables rooted in genetics and early

experi-ence

and eating disorders.

Effects of Stress, Saccharin Concentration, and Arousability on Taste Perception

Dess and Edelheit (1998) reported that stress increases sensitivity to saccharin’s

bitter-ness,

especially among highly arousable people. The current design included a replica-tion

of the experimental manipulations (saccharin concentration, stress) used by Dess

and Edelheit (1998). Figure 1 shows the mean bitterness ratings for the five

saccharin

solutions in the stress and no-stress conditions.

To determine if stress increased sensitivity to saccharin’s bitterness, a 2 × 5

mixed

design ANOVA was performed. The main effect for solution concentration was

significant,

F(4, 848)=10.267, p <.05; bitterness rating increased with solution concentration.

Like-wise,

the Concentration Condition interaction was significant, F(4, 848)=2.827, p <.05;

the concentration of the saccharin solution and the stress condition in which the

solutions

were tasted jointly affected bitterness ratings. However, contrary to Dess and

Edelheit

(1998), the mean bitterness rating for the strongest solution, Solution 5, in the

stress con-dition

was lower than the mean bitterness rating for the same solution in the no-stress

condition. Finally, the main effect for condition was not significant, F(1,

212)=0.000, p >.05;

stress did not modulate sensitivity to saccharin’s bitterness.

Effects of Arousability and Stress on Taste Perceptions

Figure 2 shows the mean bitterness ratings at the highest saccharin solution

concentra-tion

(Solution 5) for high and low arousability women in the stress and no-stress

condi-tions.

Although the mean bitterness ratings for high and low arousability women in the

stress and no-stress conditions are slightly different, the ANOVA revealed that the

difference was not statistically significant.

Dess and Edelheit (1998) found that arousability and stress modulated sensitivity to

saccharin’s bitterness. To determine if stress increased bitterness ratings among

highly

arousable individuals in the current study, a 2 × 2 independent groups ANOVA was

performed. Pleasure-displeasure was used as a covariate in this analysis because it

was

associated negatively with arousability. The main effect for condition was

significant,

F (1, 209)=4.560, p <.05; stress modulated sensitivity to saccharin’s bitterness.

However,

contrary to Dess and Edelheit (1998), the mean bitterness rating for Solution 5 in

the

stress condition (M=16.62, SE=1.69) was lower than the mean bitterness rating for

the

same solution in the no-stress condition (M=21.67, SE=1.66). The main effect for

arousability was not significant, F (1, 209)=.436, p >.05; arousability did not

modulate

sensitivity to saccharin’s bitterness. In addition, the Arousability Condition

interaction

was not significant, F (1, 210)=.753, p >.05; the interaction of arousability and

stress did

not modulate sensitivity to saccharin’s bitterness. Although bitterness ratings were

expected to differ between high and low arousability women following a stressful

event (Dess & Edelheit, 1998), they did not in the current study.

DISCUSSION

The current study reveals the multidimensionality of activity anorexia. Arousability

and sensitivity to saccharin’s bitterness interact with eating patterns to predict

excessive

exercise. Although eating patterns play the biggest role in predicting activity

anorexia,

emotionality and sensitivity to saccharin’s bitterness also have an important role

in the

activity anorexia paradigm. This finding, in conjunction with the work by Dess and

Edelheit (1998), not only reveals the complexity of this paradigm, but also provides

evidence for a revised theory of activity anorexia. Specifically, restricted eating

patterns,

in conjunction with sensitivity to saccharin’s bitterness and high levels of

arousability

among women, reinforce the hyperactivity associated with activity anorexia.

In addition to theoretical implications, these findings have clinical implications

as well.

For example, the current study provides a preliminary predictive model of activity

anorexia, which, through additional research, may develop into a useful method for

evaluating the disorder. However, additional psychometric development of this

potential

diagnostic tool is crucial due to ambiguity regarding the clinical distinction

between

women with activity anorexia and women with anorexia nervosa. Activity anorexia is a

physiologically mediated disorder that produces self-starvation, whereas anorexia

nervosa is a cognitive disorder that also produces self-starvation (Pierce & Epling,

1994). In addition, excessive exercise is characteristic of many patients with

clinically

diagnosed eating disorders (Bruch, 1965; Gull, 1874; Yates, 1991).

Little empirical evidence is available regarding not only clinical distinction, but

also

prevalence of excessive exercise among women with eating disorders (, 1997).

Aside from the current findings, little is known about the relationship among eating

patterns, exercise patterns, and taste in women with clinical and nonclinical

chronicity.

However, to a greater extent than previously imagined, it may be possible to

integrate

seemingly disparate reports on the relationship between taste and syndromes with

affective and psychomotor components, such as depression (Amsterdam, Settle, Doty,

Abelman, & Winokur, 1987; Whittemore, 1986) and anxiety (Mascie-, McManus,

MacLarnon, & Lanigan, 1983). The resulting increase in understanding and awareness

of the

complex nature of activity anorexia will aid in mainstreaming the disorder and,

subse-quently,

enable clinicians to better detect, diagnose, and treat patients with the disease.

Dess and Edelheit (1998) found that stress increased the sensitivity to saccharin’s

bitterness. In the current study, which is only the second study to compare

explicitly

baseline sensory assessments with poststress sensory assessments, stress did not

increase

sensitivity to saccharin’s bitterness. Many of the women commented that they were

not

skillful at solving anagrams, suggesting that they were anxious about performing the

cognitive task. Although the women in the no-stress group were not exposed to the

stress

manipulation (i.e., they did not have to solve three unsolvable anagrams in the

presence

of two startling noises), the mere fact that they were asked to perform cognitively

may

have elicited a stress response. Whether women were assigned to the stress or

no-stress

group, stress induction may have occurred before the start of the task. This may

explain

the similarity in bitterness ratings between the stress and no stress groups, as

well as

between high and low arousability women.

Although researchers have begun to unravel the relationship between affect and

perceptions of saccharin’s bitterness, as well as the relationship among stress,

tem-perament,

and perceptions of saccharin’s bitterness, the current study is the first to

evaluate the complex relationship among stress, temperament, saccharin’s bitterness,

eating patterns, and exercise patterns in humans. Certainly, this relationship is

complex,

multidimensional, and variable across cases. Because this area of research straddles

many psychological subfields, including social, clinical, and behavioral, increased

com-munication

between these subfields may foster new insights into the etiology, activation,

maintenance, and treatment of activity anorexia in all individuals.

Al Pater, PhD; email: old542000@...

____________________________________________________

Start your day with - make it your home page

http://www./r/hs