1. health and fitness
2. disease
3. cholesterol

PSYCHOSOCIAL PREDICTORS OF DIETARY FAT REDUCTION:
THE ROLE OF STRESS AND THE TRANSTHEORETICAL MODEL
IN A DIETARY INTERVENTION
A Dissertation
Submitted to the Graduate Faculty of the
Louisiana State University and
Agricultural and Mechanical College
in partial fulfillment of the
requirements for the degree of
Doctor of Philosophy
in
The Department of Psychology
by
Jennifer L. Francis
B.S., James Madison University, 1996
M.A., Louisiana State University, 2000
December 2003
TABLE OF CONTENTS
ABSTRACT..................................................................................................................................... iii
INTRODUCTION .............................................................................................................................1
LITERATURE REVIEW ..................................................................................................................3
Dietary Fat Interventions ...............................................................................................................3
Predictors of Dietary Fat Intake and Reduction.............................................................................5
Transtheoretical Model of Behavior Change.................................................................................8
Stress ............................................................................................................................................14
Rationale for the Study ................................................................................................................20
Research Questions and Hypotheses ...........................................................................................21
METHODS ......................................................................................................................................23
Participants...................................................................................................................................23
Procedure .....................................................................................................................................25
Measures ......................................................................................................................................26
RESULTS ........................................................................................................................................30
Data Analytic Plan .......................................................................................................................30
Research Question #1: Predictors of Baseline Dietary Fat Intake..............................................30
Research Question #2: Predictors of Dietary Change.................................................................34
Research Question #3: Change in TTM Constructs over Six Months........................................40
DISCUSSION ..................................................................................................................................42
Predictors of Dietary Fat Intake...................................................................................................43
Predictors of Dietary Fat Reduction ............................................................................................45
Change in TTM Constructs over Six Months ..............................................................................47
Limitations ...................................................................................................................................49
Conclusion ...................................................................................................................................50
REFERENCES ................................................................................................................................52
VITA ................................................................................................................................................63
ii
ABSTRACT
Dietary fat is related to cardiovascular disease and numerous intensive, controlled clinical
trials have successfully reduced dietary fat in symptomatic populations. However, there has been
less success in large, community-based studies with healthy or mildly at-risk populations. Little is
known about predictors associated with actual change in dietary fat intake and this is an important
omission because dietary interventions are more likely to be successful if they are based on factors
known to influence behavior. The purpose of this study was to investigate the psychosocial
predictors of dietary fat and dietary fat reduction through the framework of the transtheoretical
model (TTM) and minor stress.
Participants were part of a larger study examining reversal of cardiovascular signs in a
healthy population by reducing dietary fat intake to less than 20% of calories from fat. This study
consisted of 179 adults who enrolled and had complete data. Participants were randomly
assigned to the intervention or control groups. The intervention consisted of attending individual
instruction and dietary groups. The following measures were administered at baseline and six
months: Weekly Stress Inventory and the processes of change (i.e., experiential and behavioral
subscales), decisional balance, and self-efficacy questionnaires from the TTM. Dietary fat intake
was measured with four-day food records. Hierarchical regression analyses, with BMI as a
covariate, were conducted to determine psychosocial predictors of dietary fat intake at baseline
and change in dietary fat at six months. The experiential processes variable was the only unique
predictor of dietary fat intake and the relationship was moderated by minor stress. The
experiential processes were also related to dietary fat reduction. Results of a repeated measure
MANOVA revealed that use of behavioral and experiential processes increased for participants in
the intervention group over six months of the intervention. Results revealed a modest relationship
iii
between variables associated with the TTM and dietary fat and dietary fat reduction, with limited
evidence that stress moderates this relationship.
iv
INTRODUCTION
A significant amount of attention has been given to the relationship between diet and
health, particularly in the area of dietary fat and chronic disease. Evidence suggests that dietary
fat is related to cardiovascular disease (CVD), breast cancer, colon cancer, diabetes mellitus, and
obesity (National Cholesterol Education Program [NCEP], 2001). Of these diseases, CVD (e.g.,
coronary heart disease, stroke, and hypertension) is the leading cause of death in the United States,
accounting for 42% of all deaths (Field, Barnoya, & Colditz, 2002). A major causal factor for
CVD is hyperlipidemia (i.e., high cholesterol), which is related to dietary intake of total and
saturated fat (Levine, Keaney, & Vita, 1995; Van Horn & Kavey, 1997).
Reductions in cholesterol have been shown to reduce risk of morbidity and mortality from
cardiovascular disease. Estimates indicate that a 10% reduction in serum cholesterol will reduce
coronary heart disease by 25% within five years and by approximately 50% by age 40 (Law,
Wald, & Thompson, 1994). A recent meta-analysis reported 24% and 14% reductions in non-fatal
and fatal myocardial infarctions due to cholesterol lowering (Roussouw, 1994). Despite the
reduction in CVD attained with cholesterol lowering therapy, the absolute CVD rates are still
above those in a low-risk population (NCEP, 2001). Thus, primary prevention in the form of
guidelines and recommendations for the general population has become a major target for
reducing CVD (NCEP, 2001).
Current recommendations indicate that the general public should eat no more than 30% of
daily calories from dietary fat and 8% to 10% of calories from saturated fat (NCEP, 2001).
However, a significant portion of Americans consume dietary fat at much higher levels. Recent
estimates suggest that Americans are eating 35% of total calories from fat and 17% of calories
from saturated fat (U.S. Department of Health and Human Services [USDHHS]-Public Health
Service [PHS], 1995). From another perspective, only 25% of the population consumes the
1
recommended less than 30% and almost half of the population consumes between 30% and 40%
of dietary fat (Kumanyika et al., 2000).
Although reduction of dietary fat results in significant decreases in morbidity and mortality
in those at risk, as noted above, little is known about how reduction of dietary fat can reverse the
progression of cardiovascular disease in a healthy population. This is important given the large
percentage of adults that are placing themselves at risk for CVD with higher than recommended
levels of dietary fat. An intervention study was conducted at the Pennington Biomedical Research
Center (PBRC) in Louisiana to investigate reversal of cardiovascular signs in a healthy population
by targeting a dietary fat goal of less than 20% of calories from dietary fat. This low level of
dietary fat was needed to detect sufficient change in biomarkers of atherosclerotic disease.
Although large dietary fat reductions have been attainable in carefully controlled dietary studies
using participants symptomatic for chronic disease, significant reductions have not been
demonstrated in a healthy population (see Brunner, 1997, for review).
The goal of this study is to expand on the PBRC intervention study by investigating the
psychosocial predictors associated with dietary fat in a healthy population. A brief summary of
dietary fat interventions and outcomes will be presented below, followed by a general review of
the psychosocial predictors of dietary change, including a model of behavior change entitled the
Transtheoretical Model of Behavior Change. Finally, a review of the role of stress on eating
behavior will be presented with a description of how it may be related to dietary fat consumption
and therefore impact success of dietary interventions.
2
LITERATURE REVIEW
Dietary Fat Interventions
Numerous studies of dietary fat reduction have been conducted over the past 20 years and
overall they have demonstrated modest decreases in dietary fat. A recent meta-analysis examined
randomized controlled trials of dietary interventions that lasted at least three months and targeted
prevention of chronic disease (Brunner et al., 1997). Of the six studies that specifically examined
dietary fat intake over a three to six month period, four included individuals with no health risk or
mild cardiovascular risk (i.e., not symptomatic), and the samples attained a reduction of six
percent of energy from fat. Two studies targeted women at risk for breast cancer and these
participants had a reduction of 40% of energy from fat (Brunner et al., 1997). These results
suggest that significant levels of dietary change are possible; however, it may be groups that are
most at risk for disease that have the largest amounts of dietary fat reduction.
Barnard and colleagues (1995) descriptively examined dietary fat reduction in 30 studies
designed to reduce cardiovascular risk factors. Therefore, results must be interpreted with caution
due to differing sample sizes and lack of control for extraneous variables. Dietary goals for all
studies ranged from 5% to 35% of daily calories from fat and included participants symptomatic
for heart disease, vascular disease, hypertension, obesity, and diabetes. Additionally, there were
nine studies were composed of individuals who were considered “normal” and ten studies with
participants who were considered healthy but high risk. Overall, of the 30 studies examined, 50%
met their stated target dietary fat goals, 27% were within five percentage points above, and 23%
were more than five percentage points above the targeted goal (Barnard et al., 1995). Notably,
there appeared to be little difference between success of studies with longer duration (e.g., one
year) as compared to those with a shorter duration. For example, of the 18 studies that were one
year or longer, final mean fat intake of 29% was attained as compared to 28% of fat intake for the
3
21 shorter studies. Studies that had highly restrictive goals for daily fat intake (e.g., 10%) actually
attained the lowest level of fat intake ranging from 7% to 16% of daily energy from fat. However,
these studies were conducted with populations diagnosed with cardiovascular disease and may
indicate larger levels of fat reduction for those who believe they are at high health risk. Although
a majority of the studies did not report drop-out rates, those that did reported 0% to 29% of
participants were excluded due to dropping out, death, or other factors (Barnard et al., 1995).
These results suggest that dietary change is possible for studies of longer duration and more
restrictive dietary goals. Again, however, a majority of the studies were intensive, highly
controlled clinical trials involving participants who were symptomatic for chronic diseases.
The significant dietary fat reductions of the intensive, controlled trials may not generalize
to less restrictive studies (Kumanyika et al., 2000). For example, large work-site interventions
have not been shown to have large decreases in dietary fat (e.g., reductions of .9% to .36%;
Sorensen et al., 1996; Tilley et al., 1999). The Multiple Risk Factor Intervention Trial (MRFIT)
that included 12,000 men at risk of CVD reported dietary fat reductions of 4% that were
maintained after six years, but the overall dietary fat intake remained above the recommended
30% of calories from dietary fat (Gorder, Bartsch, Tillotson, Grandits, & Stamler, 1997).
In summary, dietary interventions, even those with goals of 20% or less of daily calories
from fat, have demonstrated successful adherence. However, many of the studies were highly
controlled and included individuals who were at risk or symptomatic for chronic disease. The
significant reductions that were demonstrated in these studies do not appear to generalize to larger,
less intense, community-based interventions (Kristal, Hedderson, Patterson, & Neuhauser, 2001;
Kumanyika et al., 2000). Therefore, it may be important to identify characteristics that are
associated with dietary change in order to facilitate generalization of successful health behavior
4
change from intense controlled trials to less intensive dietary interventions and potentially to
community-based samples.
Predictors of Dietary Fat Intake and Reduction
Although identification of the factors related to dietary fat consumption and the process of
health behavior change appears crucial to the design of dietary interventions, only a few studies
have examined predictors or characteristics associated with success in dietary interventions (e.g.,
Glanz, Patterson et al., 1998; Gorder et al., 1997). This is surprising given the known difficulty of
adherence to lifestyle change (e.g., diet, exercise) in general and the lower rates of adherence in
less controlled dietary interventions with asymptomatic individuals (Brownell & Cohen, 1995;
Carmody, Matarazzo, & Istvart, 1987; Chrisler, 1997; McCann, Retzlaff, Dowdy, Walden, &
Knopp, 1990). Interventions are most likely to be successful if they are based on factors that are
known to influence food choice and are related to health behavior change theory (Glanz, Lewis, &
Rimer, 1997). Identification of these factors is more difficult than it may seem since food
selection is a complex behavior that is a function of physiological, psychosocial, and cognitive
influences (Lewis, Sims, & Shannon, 1989). There is a significant amount of literature regarding
food selection processes (e.g., Birch, 1999; Eertmans, Baeyens, & VandenBergh, 2001; Nestle et
al., 1998). However, it is only recently that researchers have begun examining these factors in
relation to dietary fat consumption. The sociodemographic and psychosocial variables associated
with dietary fat intake are discussed below.
Sociodemographic Predictors
Sociodemographic predictors such as gender, age, education, socioeconomic status, and
weight characteristics have been associated with dietary fat intake. Cross-sectional studies have
indicated women tend to have lower dietary fat intake (Glanz et al., 1994; Roos,
Lahelma,Virtanen, Prattala, & Pietinen, 1998; Steptoe & Wardle, 1999; Steptoe, Wijetunge,
5
Doherty, & Wardle, 1996), whereas longitudinal studies have had mixed results (Kristal, Glanz,
Tilley, & Li, 2001; Steptoe, Doherty, Kerry, Rink, & Hilton, 2000). Furthermore, some have
suggested that women have healthier diets but are not more likely to respond to dietary
interventions (Anderson et al., 1992; Ehnholm et al., 1982; Steptoe et al., 2000). Older individuals
have been shown to have diets lower in fat than younger individuals (Glanz et al., 1994; Gorder et
al., 1997; Van Horn, Dolecek, Grandits, & Skweres, 1997) and are more likely to achieve dietary
fat reduction (Kristal et al., 2000; Winkleby, Flora, & Kraemer, 1994). Higher education level has
been shown to be related to diet (Glanz et al., 1994) and to dietary change (Kristal et al., 2000;
Urban, White, Anderson, Curry, & Kristal, 1992). Researchers have reported that individuals with
low socioeconomic status have higher fat intake (Roos et al., 1998; Shimakawa et al., 1994), but
data from a large U.S. survey indicated that there were no differences between those above and
below poverty level (Federation of American Societies for Experimental Biology [FASEB], Life
Sciences Research Office [LSRO], 1995).
Although people with a higher body mass index (BMI; an index of body weight to height)
tend to have diets higher in fat (Drewnowski, 2002), these individuals tend to report being farther
along in readiness to change behavior, suggesting they may be amenable to dietary change (Glanz
et al., 1994). However, other research suggests that lower BMI was associated with increased
adherence in the MRFIT trial for men (Van Horn et al., 1997).
Psychosocial Predictors
Recent research has examined psychosocial determinants of dietary fat intake such as
social support, knowledge, stress, and beliefs. Social support and encouragement from others,
particularly family and spouses, has been shown to be associated with success in efforts to achieve
dietary fat reduction (Bovbjerg et al., 1995; Feunekes, de Graaf, Meyboom, & van Staveren, 1998;
Kelsey et al., 1996; Steptoe et al., 2000). Knowledge, including use of nutrition labels, has been
6
shown to be a necessary component of dietary interventions and some studies have indicated that
it is associated with reduction of dietary fat (Kristal et al., 2001; Neuhouser, Kristal, & Patterson,
1999; Patterson, Kristal, & White, 1996). However, it is generally acknowledged that knowledge
is not sufficient for dietary change (Brownell & Cohen, 1995). Personal beliefs, attitudes, and
motivations regarding changing dietary behavior have been shown to be associated with dietary fat
intake and reduction (Glanz, Kristal, Tilley, & Hirst, 1998; Glanz et al., 1994; Kristal et al., 2000).
Finally, there is preliminary evidence that stress affects dietary fat intake and predicts
success in dietary fat reduction interventions, including the intervention in this study. Research
with dietary compliance in diabetic patients indicates that situations that could be classified as
minor stressors, such as time pressure, planning, and competing priorities, tended to be reported as
barriers to dietary adherence (Schlundt, Rea, Kline, & Pichert, 1994). In the MRFIT trial for men,
a baseline predictor of successful dietary fat reduction was no occurrence of major stressful life
events over the past year (Gorder et al., 1997). Furthermore, there is an area of research
suggesting that stress may affect individual’s responsiveness to high-fat, sugary (i.e., palatable)
foods, indicating that stress may impact success in dietary fat reduction interventions (Laitinen,
Ek, & Sovio, 2002; Oliver & Wardle, 1999). A full discussion of the role of stress in eating
behavior will be detailed further.
In summary, sociodemographic and psychosocial predictors have been shown to be related
to dietary fat intake. It will be crucial for future dietary interventions to include psychosocial
predictors of change in order to identify targets that facilitate dietary fat reduction (Kristal et al.,
2001). Interventions are most likely to be successful if they are based on factors that are known to
influence behavior and are related to health behavior change theory (Glanz et al., 1997). The
Transtheoretical Model of Behavior Change, as detailed below, was used in this study to examine
factors associated with dietary fat intake.
7
Transtheoretical Model of Behavior Change
The Transtheoretical Model of Behavior Change, also known as the Stages of Change Model,
is a widely used model of behavior change that provides a theoretical framework to explain and
predict health behavior change (Prochaska & DiClemente, 1983). Initial research on this model
developed from investigations into how people change in therapy (Prochaska, 1984). Prochaska
compared different strategies of behavior change compiled from various theoretical orientations
and found 10 processes that appeared to be associated with change. This integration from
differing orientations led to the term transtheoretical (Prochaska, Redding, & Evers, 1997). Initial
research within this model was conducted in the area of smoking cessation (DiClemente &
Prochaska, 1982). However, its use has expanded to a wide variety of health behaviors including:
exercise adoption, substance abuse, weight control, mammography utilization, condom use, and
dietary fat (Curry, Kristal, & Bowen, 1992; DiClemente & Hughes, 1990; Galavotti et al., 1995;
Marcus, Rossi, Selby, Niaura, & Abrams, 1992; Prochaska, Norcross, Fowler, Follick, & Abrams,
1992; Rakowski, Fulton, & Feldman, 1993).
Prochaska, DiClemente, & Norcross (1992) proposed that individuals progress through
five stages of change when attempting to cease high-risk behaviors or adopt health-promoting
behaviors. The initial stage of the model is Precontemplation in which the individual has no
intention to change behavior in the next six months and hasn’t even considered changing.
Contemplation occurs when the individual is thinking about changing a behavior in the next six
months but hasn’t done anything to initiate the process. Preparation is the stage in which there is
some action towards change but it doesn’t meet the criteria required for the behavior (e.g., using
skim milk instead of whole milk but still consuming >30% dietary fat). The Action stage occurs
when the actual behavior change takes place. It requires the most effort and time and has occurred
in the past six months (Prochaska, DiClemente, & Norcross, 1992). People in this stage must
8
attain the criterion determined by professionals in the area to reduce disease. For example, in the
dietary fat literature, 30% of calories from dietary fat is the general consensus. Maintenance is the
stage at which the individual has successfully made a behavior change for at least six months. The
authors propose that this model is not linear, but more of a spiral pattern suggesting that people
relapse in their attempts to change behavior and fall back to earlier stages of change (Prochaska,
DiClemente, & Norcross, 1992).
In addition to the stages of change, the transtheoretical model is composed of three other
constructs that have been associated with movement through the stages of change and with health
behavior change: self-efficacy, decisional balance, and processes of change (Prochaska,
DiClemente, & Norcross, 1992). These constructs will be discussed in detail below. Only one
study has applied stages of change and all of the transtheoretical constructs to dietary fat intake
(Ounpuu, Woolcott, & Greene, 2000). However, the researchers utilized a cross-sectional design
examining dietary fat intake across stages of change in a sample of women.
One potential reason for the limited research investigating the TTM and dietary fat intake
is the difficulty in assessing stage of change for dietary fat. There has been discussion regarding
the classification scheme for this measure (Brug et al., 1997; Greene & Rossi, 1998; Ni Mhurchu,
Margetts, & Speller, 1997). A major problem is that individuals are unable to estimate their actual
amount of dietary fat intake, leading to inflated and inaccurate representation in the action and
maintenance stages of change (Kristal et al., 1999; Povey, Conner, Sparks, James, Shepherd,
1999). Therefore it has been suggested that nutrient intake cut-offs be used based on food
frequency records to determine actual stage of change (Greene & Rossi, 1998). However, this
method becomes tautological when one uses the dependent variable to define the independent
variable (Kristal et al., 1999). Given these considerations, the stage of change construct will not
9
be used in this study. Rather the psychosocial constructs associated with the model will be used as
predictors of actual dietary fat intake.
Each construct of the transtheoretical model will be discussed below, first in relation to
general patterns identified across several health behaviors, followed by a summary of the available
data related to dietary fat reduction.
Decisional Balance
The decisional balance construct of the transtheoretical model was based on the decisionmaking model of Janis and Mann (1977). Velicer and colleagues (1985) tested this decisionmaking model across the stages of change for smoking cessation and produced a two-factor model
that they identified as the pros (i.e., advantages or benefits) and cons (i.e., disadvantages, barriers)
of behavior change. For example, a pro for reducing dietary fat is “People close to me disapprove
of my eating a diet which is too high in fat” and a con is “I am more content with myself when I
am eating the high fat foods I enjoy” (Rossi, Rossi, Prochaska, & Velicer, 1993). Early research
demonstrated that the pros and cons vary according to stages of change and the general pattern is
robust across 12 health behaviors, including dietary fat intake (Prochaska et al., 1994). Results
demonstrated that in all 12 behaviors, individuals who were not planning on changing a behavior
endorsed more negative than positive views of changing. Conversely, for those who were actively
engaging in behavior change, the pros outweighed the cons in 11 of the 12 behaviors.
Furthermore, it was determined that 7 of the 12 behaviors demonstrated a crossover of the pros
and cons during the contemplation stage and the remaining 5 behaviors, including dietary fat, had
a crossover in the action stage. Although this pattern was demonstrated cross-sectionally, it
suggests that as individuals begin to think about behavior change the benefits of changing become
more important than the disadvantages. Overall, the decisional balance construct has been
demonstrated to be an important component of the transtheoretical model.
10
Dietary Fat and Decisional Balance
Decisional balance and dietary fat intake have only been investigated in relation to
validation of the decisional balance scale. Three validation studies (Rossi, 1993; Rossi, Rossi et
al., 1993; Rossi et al., 1994b) demonstrated that the pros and cons were found to vary by stages of
change and in the same pattern as the other health behaviors in the seminal study noted above
(Prochaska et al., 1994). In other words, the cons were more important in the precontemplation
stage and the pros were more important in the action stage. Again, this suggests that as
individuals are reducing dietary fat, the pros of changing behavior become more important and
may have implications for dietary interventions. For example, if the pros of dietary fat reduction
are found to predict actual dietary change then this indicates that changing attitudes would be a
target for an intervention. Clearly, further research is necessary to examine the relationship
between decisional balance and dietary fat consumption.
Processes of Change
Processes of change in the transtheoretical model are covert and overt activities that people
use to progress through the stages of change. They consist of two groups of processes:
experiential and behavioral. Experiential processes focus on thoughts, feelings, and experiences
and include: consciousness raising, dramatic relief, environmental reevaluation, self-reevaluation,
social liberation and self-liberation. The behavioral processes focus on behaviors and
reinforcement and include: reinforcement management, helping relationships,
counterconditioning, stimulus control, and interpersonal systems control.
The processes of change are the least researched of all the transtheoretical model
constructs (Greene et al., 1999). Research has shown that processes of change vary across stages
for many different health behaviors, including dietary fat intake. Experiential processes are
11
typically used in the early stages of change and behavioral strategies are used in the later action
and maintenance stages (Rossi, 1992).
Dietary Fat and Processes of Change
There is a paucity of research investigating the role of the processes of change in dietary
fat reduction. This may be an important omission as the processes are explicit attitudes and
behavioral strategies that could provide information about the predictors of actual dietary change.
Understanding what processes are used by those who successfully reduce dietary fat consumption
may provide areas to target in future dietary fat reduction studies (Ni Mhurchu et al., 1997).
Self-Efficacy
Self-efficacy is conceptualized as a person’s confidence in or perceived ability to perform
a task (Bandura, 1977). It is a core construct in Bandura’s Social Cognitive Theory (1977) and
has been adapted for use in the transtheoretical model (DiClemente, Prochaska, & Gilbertini,
1985). A major tenet of self-efficacy theory is that the belief in one’s ability to perform a behavior
is predictive of the actual ability to engage in that behavior, how long they persist in doing it, and
how much distress is experienced when challenged (Bandura, 1997). Self-efficacy is influenced
by several factors including: mastery experiences, vicarious experiences, social persuasion, and
somatic and emotional states, particularly stress (Bandura, 1997).
A comprehensive review of self-efficacy and health behavior interventions suggests that
the effectiveness of interventions is significantly mediated by self-efficacy (Bandura, 1992). Selfefficacy has been shown to be associated with a variety of health behaviors including smoking
cessation (DiClemente et al., 1985), weight loss (Stotland & Zuroff, 1991), adoption and
maintenance of exercise (Marcus, Selby, Niaura, & Rossi, 1992), and adherence to medical
regimens (Kavanagh, Gooley, & Wilson, 1993; McCaul, Glasgow, & Schafer, 1987).
12
Within the transtheoretical model, a linear progression of increasing self-efficacy across the
stages of change from precontemplation to maintenance has been demonstrated in the following
health behaviors: condom use (Galavotti et al., 1995), weight control (Prochaska, Norcross et al.,
1992), exercise (Marcus et al., 1992), and smoking (DiClemente et al., 1985). Although relatively
few longitudinal studies have examined self-efficacy in relation to actual health behavior change,
it has been shown to predict exercise behavior and smoking cessation (McAuley & Courneya,
1993; Prochaska, Velicer, Guadagnoli, & Rossi, 1991).
Dietary Change and Self-Efficacy
A recent review of self-efficacy and eating behaviors indicated that few studies have
examined the association between self-efficacy and dietary fat reduction (AbuSabha &
Achterberg, 1997). Cross-sectionally, higher self-efficacy has been shown to be related to lower
fat intake (Steptoe et al., 2000; Van Duyn et al., 2001) and to progression of stages of change
(Brug et al., 1997; Sporny & Contento, 1995). Even fewer studies have investigated the role of
self-efficacy over time. McCann and colleagues (1995) demonstrated increases in self-efficacy
that were associated with dietary fat reductions for situations characterized by negative affect after
four weeks of dietary instruction. These associations remained at the three month follow-up but
not at 12 months. However, this study suffered from a small sample size (N=25). A more recent
study reported higher self-efficacy was associated with lower dietary fat consumption in
individuals entering a primary care behavioral counseling intervention but not at four months post
intervention (Steptoe et al., 2000). Interestingly, the change in self-efficacy was associated with
change in dietary fat. Based on this information, the authors hypothesized that self-efficacy is
directly involved in the processes used during behavior change but is not predictive of who will
benefit from a dietary intervention (Steptoe et al., 2000).
13
Although the studies in this area have generally shown an inverse relationship between selfefficacy and dietary fat consumption, there is little consistency in the measures used for measuring
self-efficacy and many use only one or two item questions (Brug et al., 1997; Glanz et al., 1994;
Steptoe et al., 2000). Measurement of self-efficacy with one item does not provide an accurate
picture of perceived abilities and has been shown to be less predictive of health behavior
(Bandura, 1997; Lee & Bobko, 1994).
One factor that has been neglected in both the transtheoretical model and dietary fat
reduction literature is the role of stress. This is surprising given the fact that stress is known to
affect eating behavior (see Greeno & Wing, 1994 for review), impede success in dietary
interventions (Gorder et al., 1997), and interact with self-efficacy (Bandura, 1997). Many dietary
interventions contain relapse prevention components apparently presuming stressful situations
affect the ability to adhere to dietary change (Brownell & Cohen, 1995). However, there is
usually no actual measure of the impact of stress on behavior change. Prior to delineating the
links between stress and health behavior change, a brief review of important constructs in the area
of stress will be presented.
Stress
Early conceptualizations of stress stemmed from the works of Cannon (1914) and Selye (1956)
who focused on the body’s predictable patterns of physiological response to external demands in
the environment. Selye defined stress as a “nonspecific result of any demand upon the body, be
the effect mental or somatic” (1982, p. 7). The stress response first identified by Selye has been
shown to affect a variety of areas including: physiological functioning, performance, sleep habits,
aggressive behavior, and engagement in high-risk health behaviors (Dougall & Baum, 2001).
These early works formed the foundation of stress theory and initially defined stress as a response
(Dougall & Baum, 2001).
14
Subsequent psychological theories of stress have typically focused on the role of stress as a
stimulus or “stressor.” Holmes and Rahe (1967) were the first to attempt to operationalize
environmental demands as life events and measure their impact on disorders and disease. Their
initial research posited that life events, positive and negative, were considered stressful if they
created change in the environment. This seminal work led to a long line of research investigating
life events and health status. Later theories, such as the transactional theory of stress, emphasized
that individual characteristics such as cognitions and personality mediate one’s response to stress
and has generated research in the area of coping (Lazarus & Folkman, 1984).
Major and Minor Stressors
With the advent of the Holmes and Rahe study (1967), major life event research grew
significantly (Monroe & McQuaid, 1994). Examples of major life events include the death of a
family member or loss of a job. Instruments such as the Social Readjustment Rating Scale
(Holmes & Rahe, 1967) and the Life Experiences Survey (Sarason, Johnson, & Siegel, 1978) are
typically used to measure life events. Although early research suggested that life events had a
significant impact on physical health and disease, these associations have only been modest
(Rabkin & Struening, 1976). This has led to strong criticisms of major life event research, one of
which is the difficulty in establishing a temporal relationship between major life events and illness
or disease (Monroe & McQuaid, 1994). As a result, investigation into minor stressors was
initiated to address the methodological problems associated with major life event research.
Minor stressors are events that occur on a day-to-day basis that would be considered daily
hassles. Examples would include having to wait in line, having car trouble, or having an argument
with a spouse, and are typically measured using instruments such as the Daily Stress Inventory
(Brantley, Catz, & Boudreaux, 1997), Weekly Stress Inventory (Brantley, Jones, Boudreaux, &
Catz, 1997), and the Hassles Scale (DeLongis, Coyne, Dakof, Folkman, & Lazarus, 1982). Minor
15
stressors have been shown to have a larger impact on well-being than major life events (Garrett,
Brantley, Jones, & McKnight, 1991; Holahan & Holahan, 1987; Monroe, 1983; Weinberger,
Hiner, & Tierny, 1987). Additionally, they have been shown to be better predictors of health
status than major life events (Brantley & Jones, 1993; DeLongis et al., 1982; Kanner, Coyne,
Schaefer, & Lazarus, 1981).
Stress and Health Behavior
Stress is believed to affect health directly through psychophysiological processes (e.g.,
autonomic and neuroendocrine) or indirectly through modification of health behaviors (Steptoe,
1991). Given space limitations and that the aim of this project was to assess change in health
behavior, the relationship between physiological responses and stress will not be discussed. The
remaining literature review will concentrate on the effects of stressors on health behaviors. Minor
stressors can impact health behavior by affecting an individual’s ability to engage in health
promoting behavior or by increasing engagement in negative health behaviors (Delongis et al.,
1982; Wiebe & McCallum, 1986). For example, stressors have been associated with increases in
smoking (Cohen & Lichtenstein, 1990), alcohol consumption (Cooper, Russell, Skinner, Frone, &
Mudor, 1992), and eating (Greeno & Wing, 1994). Individuals may engage in these types of
negative health practices to reduce tension or regulate affective responses to stress (Carmody,
1989).
Less information is known about the effects of stress on positive health behaviors, such as
exercise, diet, and medical regimen adherence (Griffin, Friend, Eitel, & Lobel, 1993; Stetson,
Rahn, Dubbert, Wilner, & Mercury, 1997). Stress may impede one’s ability to schedule or engage
in healthy activities or it may lead to a belief that one is unable to carry out healthful behaviors
(Griffin et al., 1993). For example, minor stressors were associated with decreased exercise and
lower self-efficacy for meeting exercise goals in a sample of women (Stetson et al., 1997).
16
Additionally, poor adherence to medication and dietary regimens in patients diagnosed with
diabetes and HIV have been related to minor stressors (Peyrot, McMurry, & Kruger, 1999;
Proctor, Tesfa, & Tompkins, 1999).
In summary, stress appears to impact health behaviors by either affecting the ability to
engage in health promoting behavior or increasing negative health behavior. Although there is
little understanding of the mechanisms related to stress and health behavior change, several
models have been developed to investigate stress and eating behavior specifically. A summary of
the research on stress and eating will be presented below, followed by an examination of the role
of stress on dietary fat intake.
Stress and Eating Behavior
It is generally accepted that stress affects eating behavior in humans and in animals (Greeno &
Wing, 1994). The dominant theory in human literature is the individual differences model that
posits that learning history, attitudes, or biology determines the effects of stress on eating (Greeno
& Wing, 1994). The overall conclusion resulting from this model is that women and restrained
eaters tend to eat increased amounts of food under stress (Grunberg & Straub, 1992; Polivy &
Herman, 2002). There are mixed results about obese individuals (Greeno & Wing, 1994). A
limitation of these studies is that many were conducted in the laboratory setting with acute
stressors (e.g., mood induction, cognitive tasks) that may not be generalizable to stressors that
would occur on a daily basis. Additionally, all of the studies conducted in the restrained eating
population were composed of college-aged women (Greeno & Wing, 1994).
In summary, restrained eaters and women are more likely to eat increased amount of food
under stress. Little is known about predictors of stress-related eating in individuals that are not
obese or chronically dieting. Furthermore, there is little investigation into particular nutrients
(e.g., sugars, fats) that would be affected by stress. This is surprising, given the assumption that
17
individuals eat more fatty foods when under stress (Greeno & Wing, 1994). Studies examining
the specific impact of stress on preference for high fat foods in “normal” healthy populations will
be discussed below.
Stress and Dietary Fat Intake
Despite the common perception that individuals eat more high fat foods under stressful
circumstances, there is a paucity of research regarding types or quality of food selected during
these times. A shift in preference towards more palatable food under stress may serve to obtain
energy quickly when demands are high and eating has taken a low priority (Greeno & Wing,
1994). This was demonstrated in one study where women ate two times more sweet foods in a
high stress condition than in a low stress condition (Grunberg & Straub, 1992). In addition,
Michaud et al. (1990) reported that fat consumption increased in high school students, particularly
girls, on stressful days. Two survey studies also reported increased intake of higher fat foods
when individuals were stressed. Laitinen, Ek, & Sovio (2002) surveyed 31-year-old men and
women in a large population based study and reported that men and women who were labeled
“stress driven eaters” ate high fat foods (e.g., sausages, pizza, chocolate) more frequently than
those who were not stress driven eaters. However, because the authors did not assess intake of
non-fatty foods it cannot be determined whether the stress driven eaters had an overall increase in
food consumption. Oliver & Wardle (1999) administered a survey to young adults and reported
that all participants reported an increase in snacking behavior when stressed, particularly for
sweets, chocolate, cake, biscuits, and “savory” snacks. The authors point out that these foods
share properties of being highly palatable, easy to prepare, and of high caloric density.
Studies with naturalistic designs have demonstrated mixed results with regard to stress and
fat intake. Some investigators have reported higher fat intake during stressful time periods
(Michaud et al., 1990; Wardle, Steptoe, Oliver, & Lipsey, 2000; Weidner, Kohlmann, Dotzauer, &
18
Burns, 1996). Other studies have found increased fat intake in relation to objective stress (e.g.,
hours worked) rather than perceived stress (McCann, Warnick, & Knopp, 1990). Finally, others
have found no change in fat intake between high and low stress times (Bellisle et al., 1990;
Pollard, Steptoe, Canaan, Davies, & Wardle, 1995; Stone & Brownell, 1994). Clearly, there is no
clear pattern regarding the relationship between stress and increased dietary fat intake. This may
be due to differing measures of stress that are not based on validated instruments, poor
measurement of food intake, and problems inherent to cross-sectional and laboratory based
designs.
In summary, self-reported stress has been shown to increase food consumption and
possibly dietary fat in some studies. Although none of the studies addressed potential mechanisms
that would explain increases in high fat foods, it has been suggested that stress may affect
performance of health behaviors by impacting frequency and patterns of eating (Griffin et al.,
1993). Time pressures may make speed and convenience more important than actual nutritional
quality of food (Oliver & Wardle, 1999).
These factors may have a bearing on success in dietary interventions, particularly those
with dietary fat reduction as a goal. There is little information on the effects of stress on success
in dietary interventions. The MRFIT trial found that individuals who reported no occurrence of
major stressful life events over the past year had better dietary compliance compared to those with
one or more stressful life events (Gorder et al., 1997). Furthermore, preliminary data from this
study indicated that stressful minor life events and global perception of stress was associated with
lower self-efficacy to resist dietary fat (Martin et al., 2001). Given the known effects of minor
stressors on health behaviors and the hypothesized increase of high fat foods under stress for
certain populations, investigation into the impact of stress on dietary fat reduction in an
intervention would be beneficial.
19
Rationale for the Study
The relationship between dietary fat intake and cardiovascular disease has been firmly
established (NCEP, 2001). Numerous intense, controlled clinical trials have successfully reduced
dietary fat in symptomatic populations (see Barnard et al., 1995; Brunner, 1997, for reviews).
However, there has been less success in large, less controlled studies with healthy or mildly at-risk
populations (e.g., Sorensen et al., 1996). Although various predictors associated with dietary fat
consumption have been investigated cross-sectionally, little is known about predictors associated
with actual change in dietary intake. This is an important omission because dietary interventions
are more likely to be successful if they are based on factors that are known to influence behavior
and are related to health behavior change theory (Glanz et al., 1997).
Although the transtheoretical model is widely used to explain health behavior change,
there has been very limited work in the area of dietary fat and this research is still in the
descriptive and explanatory phases (Burkholder & Evers, 2002). Three studies have
longitudinally investigated the stages of change in relation to dietary fat and they did not include
an examination of decisional balance, processes of change, or self-efficacy (Glanz, Patterson, et
al., 1998; Greene & Rossi, 1998; Kristal et al., 1999). It has been suggested that it is important to
use all dimensions of the transtheoretical model rather than just the stage dimension alone in order
to “capture the richness, complexity, and multivariate nature of behavior change” (Greene et al.,
1999, p. 674).
Stress has a demonstrated effect on eating behavior with some evidence that it increases
consumption of dietary fat (e.g., Greeno & Wing, 1994; Oliver & Wardle, 1999). This has
implications for the success of dietary fat interventions. Therefore, it is important to understand
the association of stress to dietary fat intake and possibly with psychosocial predictors of dietary
change.
20
This study investigated the psychosocial predictors of dietary fat and dietary fat reduction
through the framework of the transtheoretical model and its relationship with minor stress. In
addition, constructs of the transtheoretical model were measured at baseline and six months to
determine if change in the variables corresponded to changes in dietary fat.
Research Questions and Hypotheses
Question #1a: What psychosocial variables associated with the transtheoretical model predict
dietary fat intake of healthy participants entering a dietary intervention (See Table 1)? Is minor
stress a predictor of dietary fat intake?
Hypothesis: Higher self-efficacy will be the strongest predictor of lower dietary fat intake.
It is hypothesized that higher pros of changing to a low fat diet and more frequent use of
the behavioral and experiential processes will be associated with lower dietary fat intake.
Finally, higher numbers of minor stressors are expected to be associated with greater
dietary fat intake.
Question #1b: For participants entering a dietary intervention, does minor stress moderate the
relationship between the transtheoretical model variables and baseline dietary fat intake?
Hypothesis: It is hypothesized that stress will moderate the relationship between selfefficacy and dietary fat such that at higher levels of stress, lower self-efficacy will be
associated with higher dietary fat. It is hypothesized that at high levels of stress, lower use
of behavioral and experiential processes will be associated with higher dietary fat. Finally,
it is hypothesized that at high levels of stress, more cons of reducing dietary fat will be
associated with higher dietary fat intake.
Question #2a: What are the baseline psychosocial constructs associated with the transtheoretical
model that predict change in dietary fat intake from baseline to 6 months for participants enrolled
in a dietary intervention? Is minor stress a predictor of dietary fat reduction?
21
Hypothesis: It is hypothesized that higher self-efficacy will be the most significant
predictor of dietary fat reduction. In addition, it is hypothesized that higher pros of
changing to a low fat diet and more frequent use of the behavioral and experiential
processes, will predict dietary fat reduction. Finally, it is hypothesized that higher levels of
minor stress will predict lower dietary fat reduction.
Question #2b: Does minor stress moderate the relationship between the transtheoretical model
variables and dietary fat reduction?
Hypothesis: It is hypothesized that minor stress will moderate the relationship between
self-efficacy and dietary fat such that at higher levels of stress, lower self-efficacy will be
associated with less dietary fat reduction. It is hypothesized that at high levels of stress,
lower use of behavioral and experiential processes will be associated with less dietary fat
reduction. Finally, it is hypothesized that at high levels of stress, the more cons of
changing dietary fat will be associated with lower dietary fat reduction.
Question #3: Are there significant changes in the constructs associated with the TTM between the
intervention and control groups over the course of the intervention?
Hypothesis: It is hypothesized that participants in the dietary intervention will increase
self-efficacy, increase the pros of changing to a low fat diet, increase the use of the
behavioral processes, and decrease the use of the experiential processes.
Table 1
Predictor, moderator, and outcome variables for research questions #1 and #2
Predictors
Self-efficacy
Decisional Balance
(pros-cons)
Experiential Processes
Behavioral Processes
Moderator
Weekly Stress Inventory
22
Outcomes
Baseline percentage of dietary fat intake
Change in dietary fat intake (baseline – six
months)
METHODS
Participants
Over a two-year period, participants were recruited to enroll in a study investigating the
effects of dietary fat reduction on markers of atherosclerotic disease and on atherosclerotic disease
progression. Participants were healthy men and women between the ages of 45 and 70.
Exclusionary criteria included: presence of coronary artery disease or use of lipid lowering
medication; diabetes mellitus; uncontrolled hypertension or use of hypertensive medication; renal,
hepatic, endocrine, gastrointestinal or other systemic diseases; body mass index greater than 35;
history of alcohol or drug abuse; history of eating disorder; presence of psychotic disorder; or use
of antipsychotic or mood stabilizing medications.
Three hundred fifty two participants were enrolled in the study and 278 completed the sixmonth follow-up indicating high retention rates for the overall sample. Although this full sample
was intended to be the sample used for this study, there was a clerical error and administrative
decision that limited the sample size for these analyses. At baseline, 153 participants did not
receive the decisional balance questionnaire due to an error in the copying of data packets and 20
participants did not receive the processes of change questionnaire. Additionally, at the six-month
time interval, it was necessary to shorten the questionnaire battery due to subject complaints about
the length of the battery. Therefore, the decisional balance questionnaire was removed at the sixmonth interval since there was no baseline information for many participants on this measure. The
total sample used for this study was composed of the 179 participants who had complete data,
including the decisional balance questionnaire, at baseline. To achieve power of .80 for both
regression analyses, the recommended sample size is 130 participants using the rule, N > 50 + 8m
(m = number of predictors; Green, 1991).
23
At the six-month follow-up, 162 (90%) participants remained in the study, indicating high
retention rates similar to the original larger sample. There were no significant differences on age,
race, education level, baseline BMI, or baseline percent of dietary fat between those with or
without the decisional balance questionnaire. Additionally, there were no significant differences
on age, race, education level, baseline BMI, baseline percentage of dietary fat, or on any of the
baseline TTM measures between those who dropped out at six months and the remaining sample.
The baseline sample for this study was composed of 117 female (65%) and 62 male (35%)
participants, with a mean age of 55.37 (SD = 5.25). A majority of the sample was Caucasian and
married with some college education. The mean dietary fat intake at baseline was slightly above
recommended values (range 9% - 48%) and average BMI indicated the sample was overweight
(See Table 2). Following randomization, 65.4% of participants were enrolled in the dietary
intervention and 34.6% of the participants in the control group.
Table 2
Baseline Demographics
Variable
Age
% DF
BMI
Gender
Female
Male
Ethnicity
Caucasian
AfricanAmerican
Other
Marital
Status
Single
Married
Total Sample
N=179
M
SD %
55.37 5.25
31.18 7.62
27.26 4.16
Control
n=62
M
SD
54.98 5.39
30.63 5.90
26.55 3.80
%
Intervention
n=117
M
SD
%
55.57
5.19
31.48
8.39
27.64
4.30
65.4
34.6
69.4
30.6
63.2
36.8
92.2
7.30
91.9
6.5
92.3
7.7
.6
1.6
3.9
71.5
4.8
72.6
24
--
3.4
70.9
(Table continued)
Divorced
21.2
Widowed
3.4
Education
< 12
.6
years
12 years
15.1
12-16 yrs
30.7
16 yrs
25.7
>16 yrs
25.7
Unknown
1.1
Note. M = mean; SD = standard deviation.
22.6
--
20.5
5.1
--
.9
11.3
32.3
30.6
25.8
--
17.1
29.9
24.8
25.6
1.7
Procedure
Participants were recruited through radio and newspaper advertisements for a two-year study
examining the effects of a low fat diet on heart disease progression. Once participants were
enrolled, they completed a four-week baseline period in which pre-intervention data was obtained.
They were instructed not to modify their normal eating habits or food selections during this time.
Following the baseline period, participants were randomized to either a no-intervention control
group or to a dietary-intervention group. Those placed in the no-intervention control group
completed questionnaires and food records on the same schedule as individuals in the intervention
group. All measures were collected at baseline and at 6, 12, 18, and 24 months. Psychosocial
constructs and food records administered at baseline and six months were used for this study.
Dietary Intervention
Participants in the dietary intervention arm received intensive group and individual dietary
counseling from trained dietitians. Goals for the study were to decrease total dietary fat to less
than or equal to 20% of daily caloric intake and saturated fat to 4% to 6% of daily caloric intake.
For the time frame of this study, group dietary instruction involved 10 dietary sessions over six
months. For the first four months of the study, group interventions were conducted every other
week and then decreased to once a month for months five and six of the study. Groups contained
25
12-15 participants and lasted approximately one hour. The group interventions were didactic and
were composed of education and information on behavioral strategies for decreasing dietary fat.
Topics included: reading nutrition labels, types of dietary fat, how to choose low fat foods, how to
modify recipes, how to follow a low-fat diet in social situations and restaurants, and shopping and
food preparation to meet low-fat dietary goals.
Individual dietary interventions occurred once during months one, three, and six were
conducted by the same dietitian running the group intervention. During these individual sessions
the dietitian provided feedback to the participant about dietary intake based on the food records.
These sessions were designed to target areas of dietary fat improvement and generate strategies to
improve compliance in these areas.
Measures
The demographic information and the decisional balance questionnaire were administered at
baseline. The remaining measures were completed at both baseline and at the six-month interval.
Demographic Information
The following demographic information was obtained by self-report during the initial
screening phone call that determined eligibility for the study: age, gender, marital status, race, and
educational level.
Weight
Participants were weighed two consecutive times without shoes or heavy clothing. The
average of these two weights was used for the final weight.
Body Mass Index (BMI)
BMI was calculated as kg/m2.
26
Decisional Balance Questionnaire for Dietary Fat Reduction (Rossi, Rossi, et al., 1993)
The decisional balance questionnaire is an eight-item inventory designed to measure the
pros and cons of changing to a low fat diet. There are four items measuring the pros (e.g., People
close to me disapprove of my eating a diet which is too high in fat) and four items measuring the
cons (e.g., I am more content with myself when I am eating the high fat foods I enjoy). Items are
rated on a five-point Likert scale from 1 (not important at all) to 5 (extremely important). Higher
scores indicate higher importance of pros or cons. Internal consistency (alpha) coefficients
indicate good reliability for pros (.82 - .86) and cons (.83 - .84; Rossi et al., 1994b; Rossi, Rossi et
al., 1993). The pros and cons were found to vary across stages of change indicating that the
decisional balance questionnaire is externally valid. In this study, the decisional balance score was
calculated by subtracting the cons from the pros resulting in a score range of –16 to 16. Therefore,
a positive score indicates endorsement of more pros and a negative score indicates endorsement of
a larger number of cons.
Processes of Change for Dietary Fat Reduction (Rossi & Rossi, 1993)
The processes of change instrument is a 33-item measure assessing the frequency of the
use of behavioral and experiential strategies for decreasing dietary fat. This instrument uses a
five-point Likert scale ranging from 1 (repeatedly) to 5 (never), with higher scores indicating less
use of the processes of change. It is composed of two subscales: the behavioral processes (e.g.,
not bringing high fat foods into the home or avoiding others who are eating high fat foods) and the
experiential processes (e.g., seeking information about lowering dietary fat or caring about the
consequences of eating high fat foods). The behavioral processes include: helping relationships,
reinforcement management, interpersonal systems control, stimulus control, and
counterconditioning. The experiential processes include: consciousness raising, self-liberation,
social liberation, self-reevaluation, and environmental reevaluation. Internal consistency for the
27
eleven scales was adequate and ranged from .73 (counterconditioning and stimulus control) to .90
(dramatic relief; Rossi & Rossi, 1993; Rossi et al., 1994a). Additionally, the processes were found
to vary across stage of change suggesting that the measure is externally valid.
Self-Efficacy Questionnaire for Dietary Fat Reduction (Rossi et al., 1994c)
The self-efficacy questionnaire is a 12-item inventory designed to measure confidence in
resisting high fat foods in three different situations: Positive/Social Situations, Negative/Affective
Situations, and Difficult Situations. The positive/social situations subscale measures ability to
resist high fat foods when there is a positive feeling and a social situation (e.g., while eating out at
a restaurant with close friends). Negative/affective situations involve a bad feeling and a negative
situation (e.g., had argument with someone close to me), and a difficult situation is when there is
trouble in obtaining or preparing lower fat foods (e.g., when only high fat foods are available).
Items are rated on a five-point scale from 1 (extremely confident) to 5 (not at all confident) with
higher scores indicating lower self-efficacy. Internal consistency is adequate with reliability
coefficients for the subscales ranging from .73 (difficult situations) to .95 (negative/affective
situations) and .88 for the total scale (Rossi et al., 1994c). Self-efficacy was found to increase as
stage of change progressed in the developmental sample (Rossi et al., 1994c).
Weekly Stress Inventory (WSI; Brantley, Jones, Boudreaux, & Catz, 1997)
The WSI is an 87 item self-report measure assessing the number and appraised
stressfulness (on a 0 to 7 Likert scale) of minor stressors that occurred over the past week. Two
scores are obtained from the WSI, the event score and the impact score. The event score is the
total number of events that occurred during the past week and the impact score is the sum of the
ratings of the events that occurred over the past week. However, only the event score was used for
this study. Internal consistency has been shown to be high with alphas ranging from .92 - .96 for
the WSI event score. Test-retest reliability of the WSI was conducted on 170 students in which
28
they were given the WSI on two occasions separated by 1-3 hours. Reliability coefficients were
adequate for both scales (WSI-event, r = .83). The WSI was compared to the Hassles Scale
(Kanner et al., 1981) and found to have adequate concurrent validity (r = .61 - .69).
Dietary Intake
Dietary intake was measured using a 4-day food record that was designed for use at the
Pennington Biomedical Research Center and has been included in several published studies (e.g.,
Lovejoy, Champagne, Smith, deJonge, & Xie, 2001). Participants were instructed on how to
complete food records and were provided with sample records. They were instructed to record
everything they ate or drank during the time period. The four days required for recording dietary
intake included one weekend day and three weekdays within a span of a 7-day period. For this
study, food records that were collected at baseline and 6 months were used. Following a review of
the records by the dietitian, the records were analyzed using the PBRC Food Diary Program
(Pennington Biomedical Research Foundation). As part of the extensive dietary output from the
food records, percentage of energy from fat was used for this study.
29
RESULTS
Data Analytic Plan
Prior to testing the study hypotheses, descriptive statistics were determined for all of the study
variables at baseline and six months collapsed across group (see Table 2). The baseline and six
month scores on the variables associated with the TTM, dietary fat intake, and the WSI were
examined for assumptions of univariate and multivariate analyses. The decisional balance
difference score was positively skewed. Therefore, a logarithmic transformation was applied after
first adding a constant of 16 to avoid taking the log of a negative or zero number. The baseline
and six-month WSI event score was also positively skewed and a square root transformation was
applied. For ease of interpretation, the original means of these two variables are reported in Table
3. No multivariate outliers were revealed using the mahalanobis distance test (Tabachnick &
Fidell, 2001).
Research Question #1: Predictors of Baseline Dietary Fat Intake
Prior to testing the research question, correlations between baseline percentage of dietary
fat, the constructs associated with the TTM, minor stress, and demographic variables were
conducted and are presented in Table 4. Education, marital status, and gender were not
significantly associated with dietary fat intake at baseline and were not added to the subsequent
regression analyses as covariates. Experiential processes, self-efficacy, and BMI were
significantly correlated with baseline percentage of dietary fat. Higher dietary fat was associated
with larger BMI, less use of experiential processes (as indicated by a higher number), and lower
self-efficacy (as indicated by a higher number). Minor stress was not significantly associated with
dietary fat or any of the variables associated with the TTM.
To address research question #1, (i.e., What are the psychosocial variables associated with
the transtheoretical model that predict dietary fat intake of healthy participants entering a dietary
30
intervention? Is minor stress a predictor of dietary fat intake and does it moderate the relationship
between the transtheoretical model variables and baseline dietary fat intake?), a hierarchical
regression analysis was conducted. Prior to conducting the regression, the predictor variables
were centered to prevent the negative impact of multicollinearity (Aiken and West, 1991). The
centered variables were computed by subtracting the overall mean of the predictor from each
individual score to create variables with means of zero. These centered predictors were then
multiplied to create the interaction terms. Significant interactions were examined using simple
slope analyses and plots. Post-hoc probing was conducted using t-tests of the significant
interactions to determine which slope was significantly different from zero. Plots were created by
examining the moderator at one standard deviation above and below the mean (Aiken & West,
1991; Holmbeck, 2002).
Table 3
Mean scores of psychosocial constructs for total sample at baseline and
six months (n = 179)
Variable
Baseline
Mean
SD
Six Monthse
Mean
SD
% Dietary Fat
31.18
7.62
23.81
9.48
Behavioral processesa
45.36
9.82
42.85
9.69
Experiential processesa
44.94
11.90
43.21
11.51
Self-efficacyb
29.66
8.49
28.67
7.55
Decisional Balancecd
(pros-cons)
2.76
4.75
--
--
Weekly Stressc
26.73
13.34
25.93
13.95
Note. aHigher scores indicate less use of processes.
b
Higher scores indicate less self-efficacy.
c
Original values are presented for ease of interpretation.
d
Higher scores indicate greater pros.
e
Decisional balance was not collected at six months.
31
Table 4
Bivariate correlations between demographic variables, psychosocial constructs and baseline
dietary fat intake (n=179)
Variable
1. % Dietary
Fat
1
-
2. Gender
3. Education
4. Marital
Status
5. BMI
2
-.09
-
3
.03
-.28*
-
4
-.05
6
.14
7
.25**
8
.17*
9
-.10
10
.01
-.18
-.15*
-.27**
.05
-.05
-.08
-.09
.03
.16*
.01
-.06
-.02
-.08
-
.01
-.09
-.10
-.11
.12
-.08
-.12
.06
.09
.08
.09
-
.60**
.07
-.28**
-.01
.17*
-.30**
.09
-.24**
.13
.21*
5
.20**
-
6. BP
7. EP
-
8. SE
-
9. DB
-
10. WSI
-.09
-
Note. BP = behavioral processes; EP = experiential processes; SE = self-efficacy; DB = decisional
balance. * p < .05. ** p < .01.
BMI was entered into step 1 of the regression to control for the association between BMI
and dietary fat intake. To examine the contribution of each psychosocial variable, scores on
behavioral processes, experiential processes, self-efficacy, log transformed decisional balance
difference, and the square root transformation of the WSI were entered into Step 2. Finally, the
interactions between WSI and each of the constructs associated with the TTM were entered into
Step 3 to address whether stress moderates the relationship between the elements of the TTM and
dietary fat intake.
The results of this analysis indicated that the model was significant (See Table 5). BMI accounted
for 4% of the variance in dietary fat, F(1,177) = 7.56, p < .01. After controlling for BMI, the
second step as a whole was significant, F(6,172) = 3.66, p < .01, accounting for an additional
7.2% of the variance. The interactions as a whole were significant, F(10,168) = 3.03, p < .01 but
32
did not significantly improve the variance accounted for dietary fat intake above BMI and the
elements of the TTM. Examination of the variables within the second step revealed that
experiential processes was the only significant predictor (β = .20, p < .05), demonstrating that
lower use of experiential processes was related to higher dietary fat at baseline. Additionally,
there was a significant interaction between experiential processes and minor stress, (β = .20, p <
.03). This interaction is illustrated in Figure 1. There were no significant interactions between
minor stress and the constructs associated with the TTM.
Follow-up simple slope analysis on the interaction revealed that the interaction was
significant at higher levels of WSI, t (174) = 4.00, p < .001, but not at lower levels of stress, t
(174) = .73, p > .05. These results indicate that at higher levels of minor stress, more use of
experiential processes was associated with a lower percentage of dietary fat intake. In contrast,
among individuals who reported lower levels of minor stress there was no significant association
between experiential processes and dietary fat.
Table 5
Summary of hierarchical regression analysis examining the psychosocial predictors of dietary fat
consumption at baseline
Variable
Step One
.37*
.20*
Step Two
Step Three
B
β
B
β
.35*
.19*
.29*
.16*
BP
.02
.02
.01
.01
EP
.13*
.20*
.13*
.20*
SE
.11
.12
.12
.13
DB
-1.79
-.03
-2.14
-.03
BMI
B
β
(Table continued)
33
WSI
-.24
-.04
-.26
-.05
BP x WSI
.01
.01
EP x WSI
.09*
.20*
SE x WSI
-.01
-.03
DB x WSI
5.16
.09
Note. R2 = .04 for Step 1**; ∆ R2 = .07* for Step 2; ∆ R2 = .04 for Step 3. BP=behavioral
processes; EP=experiential processes; SE = self-efficacy; DB = decisional balance. *p < .05; **p
< .01.
35
33.98
31.79
30.6
30
27.86
High Stress
Low Stress
25
20
High EP
Low EP
Experiential Processes
Figure 1: Interaction effect of minor stress and experiential processes in the prediction of dietary
fat intake at baseline
Research Question #2: Predictors of Dietary Change
Participants in the intervention group significantly reduced their dietary fat intake from
31.24% to 19.44% by the six month follow-up, t (105) = 13.77, p < .001, indicating that the
intervention was successful. In order to identify baseline predictors of dietary change, a dietary fat
change score was computed for those in the intervention group: baseline percentage of dietary fat
minus six-month percentage of dietary fat. Larger numbers were indicative of greater decreases in
dietary fat intake.
34
Bivariate correlations between percent change in dietary fat, baseline variables associated with
the TTM, minor stress, and BMI are presented in Table 6. Greater change in dietary fat was
significantly associated with higher baseline BMI and, contrary to expectation, less use of
experiential processes. The relationship between dietary fat change, minor stress, and the other
TTM variables failed to reach significance.
To test research question #2, (i.e., What are the baseline psychosocial constructs associated
with the transtheoretical model that predict change in dietary fat intake from baseline to six
months for participants enrolled in a dietary intervention? Is minor stress a predictor of dietary fat
reduction?), a hierarchical regression analysis was conducted. Just as in the first analysis, the
predictor variables were centered and then multiplied to create the interaction terms (Aiken &
West, 1991; Holmbeck, 2002). BMI was entered into Step 1, to control for the significant
association between BMI and dietary fat intake. Behavioral processes, experiential processes,
self-efficacy, log transformed decisional balance difference score, and the square root
transformation of the WSI were entered into Step 2. Finally, to examine research question #2b
(i.e., Does stress moderate the relationship between the variables associated with the TTM and
dietary fat reduction?), the interactions between minor stress and each of the TTM constructs were
entered into Step 3.
The results of this analysis indicated that the model was significant (See Table 7). BMI
accounted for 5.8% of the variance in dietary fat change, F(1,104) = 6.41, p < .02. After
controlling for BMI, the second step as a whole was significant, F(6,99) = 2.24, p < .05 but did not
account for a significant increase in variance over BMI. The interactions in step three were not
significant, F(10,95) = 1.67, ns, and they did not add significant variance to the model. None of
the individual baseline variables of the TTM or the interactions with minor stressor predicted
percent change in dietary fat intake.
35
Table 6
Bivariate correlations between change in dietary fat, BMI, and TTM constructs (n=106)
Variable
1
2
3
4
5
6
7
1. ∆DF
-
.24*
.12
.25*
.14
-.01
-.04
-.09
.08
.12
.19
.16
-
.63**
.16
-.14
-.06
.34**
-.29*
.01
-.23*
.05
2. BMI
-
3. BP
4. EP
-
5. SE
-
6. DB
-
7. WSI
-.03
-
Note. aHigher numbers indicate larger decrease in dietary fat. BP=behavioral processes;
EP=experiential processes; SE = self-efficacy; DB = decisional balance. *p < .05. ** p < .01.
Table 7
Hierarchical regression analysis examining psychosocial predictors of dietary fat consumption (n
= 106).
Variable
BMI
Step One
B
β
.50*
.24*
Step Two
Step Three
B
β
B
β
.45*
.22*
.38
.18
BP
-.03
-.04
-.01
-.01
EP
.18
.24
.16
.21
SE
.06
.05
.07
.07
DB
2.38
.03
2.19
.03
WSI
-.54
-.08
-.41
.06
(Table continued)
36
BP x WSI
.07
.10
EP x WSI
.01
.02
SE x WSI
.01
.02
DB x WSI
-8.28
-.12
Note. R2 = .05 for Step 1*; ∆ R2 = .06 for Step 2; ∆ R2 = .03 for Step 3. BP = behavioral
processes; EP = experiential processes; SE = self-efficacy; DB = decisional balance. *p < .05.
Follow-up Analysis
The sample for the previous regression analysis included the 106 individuals with complete
data who remained in the intervention group at the six-month follow-up interval. The power rule
used in this study (see Participants Section) indicated that 130 subjects are required for adequate
power (i.e., .80), indicating that the sample size for the previous regression analysis was
inadequate. Therefore, after conducting the analysis as originally proposed, an additional analysis
was conducted to determine whether the results were likely the result of a lack of power.
To clarify the results of research question #2, the follow-up analysis was conducted with the
entire intervention sample by eliminating the decisional balance questionnaire from the regression.
This is a slight modification of the research question, but allows for sufficient power to investigate
the remaining elements of the TTM and minor stress. The results of this analysis are presented
below.
As reported in the method section, 352 participants were enrolled in the study and 236
(67%) were randomized to the dietary intervention. At the six-month interval, 187 were remaining
in the intervention. These participants significantly reduced their dietary fat intake from 31% to
20% by the six month follow-up, t (186) = 18.56, p < .001, indicating that the intervention was
successful. In order to identify baseline predictors of dietary change, a dietary fat change score
was computed for those in the intervention group: baseline percentage of dietary fat minus six-
37
month percentage of dietary fat. Larger numbers were indicative of greater decreases in dietary fat
intake.
Bivariate correlations between percent change in dietary fat, baseline elements of the TTM,
minor stress, and BMI are presented in Table 8. Greater change in dietary fat was significantly
associated with higher baseline BMI and, contrary to expectation, less use of experiential
processes. The relationship between dietary fat change, minor stress, and the other elements of the
TTM failed to reach significance.
To test research question #2, (i.e., What are the baseline psychosocial constructs associated
with the transtheoretical model that predict change in dietary fat intake from baseline to six
months for participants enrolled in a dietary intervention? Is minor stress a predictor of dietary fat
reduction?), a hierarchical regression analysis was conducted without the decisional balance
questionnaire. Just as in the previous analyses, the predictor variables were centered and then
multiplied to create the interaction terms (Aiken & West, 1991; Holmbeck, 2002). BMI was
entered into Step 1, to control for the significant association between BMI and dietary fat intake.
Behavioral processes, experiential processes, self-efficacy, and the square root transformation of
the WSI were entered into Step 2. Finally, to examine whether stress moderates the relationship
between constructs associated with the TTM and dietary fat reduction (e.g., research question
#2b), the interactions between minor stress and each of the remaining TTM constructs were
entered into Step 3.
The results of this analysis indicated that the model was significant (See Table 9). BMI
accounted for 2.3% of the variance in dietary fat change, F(1,185) = 4.27, p < .05. After
controlling for BMI, the second step as a whole was significant, F(5,181) = 2.97, p < .05 and
accounted for an additional 5.3% of the variance in dietary fat change. The interactions in step
three were significant, F(8,178) = 2.03, p < .05, but they did not add significant variance to the
38
model. Examination of the variables within the second step revealed that the experiential
processes variable was the only unique predictor (β = .26, p < .01) of change in dietary fat. None
of the interactions were unique predictors of dietary fat change. Although it appears there was a
lack of power to detect significant results in the original analysis, these results further demonstrate
only modest relations between the TTM variables, stress, and dietary fat change. Interestingly, the
results suggest that participants in the intervention who used less of the cognitive processes (e.g.,
thought less about committing to change) at baseline tended to have greater reductions in dietary
fat intake over the course of six months.
Table 8
Bivariate correlations between change in dietary fat, BMI, and variables associated
with the TTMa,b
Variable
1
2
3
4
5
6
1. ∆DFc
-
.15*
.08
.22**
.10
.03
-.13
.01
.06
.17*
-
.61*
.28**
-.01
.26**
.05
2. BMI
3. BP
-
4. EP
-
5. SE
-
6. WSI
.13
-
Note. BP = behavioral processes; EP = experiential processes; SE = self-efficacy;
DB = decisional balance. aN=187. bAnalysis does not include decisional balance;
c
Larger numbers indicate a larger decrease in dietary fat. *p < .05. ** p < .01.
39
Table 9
Hierarchical regression analysis examining psychosocial predictors of the change in dietary fat
consumption
Variable
BMI
Step One
Step Two
Step Three
B
β
B
β
B
β
.33*
.15*
.30*
.14*
.29
.13
-.05
-.07
BP
-.07
-.08
EP
.19**
.26**
.18**
.24**
SE
.05
.05
.06
.06
-.13
-.02
-.11
-.02
BP x WSI
.07
.10
EP x WSI
-.03
-.05
SE x WSI
.01
.02
WSI
Note. N=187. R2 = .02 for Step 1*; ∆ R2 = .05 for Step 2*; ∆ R2 = .01 for Step 3. BP =
behavioral processes; EP = experiential processes; SE = self-efficacy; DB = decisional balance.
*p < .05. **p < .01.
Research Question #3: Change in TTM Constructs over Six Months
To address hypothesis #3 that elements of the TTM would change over time for the
intervention group as compared to the control group, a 2 (Group: Intervention, Control) by 2
(Time: Baseline-Six Months) MANOVA with repeated measures on time for the TTM variables
was conducted for the following dependent variables: self-efficacy, behavioral processes, and
experiential processes (Table 10). The decisional balance questionnaire was not included in this
analysis since it was not administered at the six-month interval. The sample was the same as that
used to test the original research question #2. Post-hoc ANOVA’s were conducted to examine
significant effects.
40
The analysis revealed no significant main effect for time (Wilks’ λ = .96, F(3,158) = 2.31, ns)
or group (Wilks’ λ = .96, F(3,158) = 2.17, ns). The group by time interaction was significant,
Wilks’ λ = .90, F(3,158) = 5.89, p < .001.
Three ANOVA’s were conducted on the within subject factors as follow-up tests to the
significant interaction. Using the Bonferroni method, each ANOVA was tested at the p < .02
level. Use of behavioral processes significantly increased in the intervention group, F(1,160) =
10.58, p < .001. The experiential processes also significantly increased in the intervention group,
F(1,160) = 10.24, p < .01. There was no significant difference between the two groups on selfefficacy at six months, F(1,160) = 5.17, p < .05.
Table 10
Baseline and six-month means and standard deviations for intervention and control participants on
the TTM constructs
Variable
Control group (n=57)
Baseline
6 month
Mean (SD)
Mean (SD)
Intervention (n=105)
Baseline
6 month
Mean (SD)
Mean (SD)
Self-efficacya
29.75 (8.19)
31.07 (8.02)
29.16 (8.22)
27.52 (7.35)
Experiential
Processesb
45.58 (11.47)
46.98 (13.7)
44.05 (11.77)
41.13 (9.51)c
Behavioral
45.75 (9.94)
46.25 (10.36)
45.76 (9.88)
41.06 (8.80)c
b
Processes
Note. aHigher scores indicate less self-efficacy; bHigher scores indicate less use of processes.
c
significantly different from baseline, p < .01.
41
DISCUSSION
This study examined responses to psychosocial questionnaires of community adults
participating in a dietary change program aimed at reducing their dietary fat intake. The goal of
this study was to determine whether participant responses indicative of key components of the
transtheoretical model including: self-efficacy, decisional balance, experiential processes, and
behavioral processes would predict baseline dietary fat and dietary fat reduction. The study also
addressed the role of stress as a direct predictor of dietary fat and as a moderator of the
relationship between the components of the TTM and dietary fat. Results suggested a modest
relationship between the elements of the TTM and dietary fat with limited evidence that stress
may moderate this relationship.
A general overview of the results will be presented below followed by a discussion of each
hypothesis and the pertinent findings. Research question #1 addressed predictors of dietary fat at
intake and the results revealed that participants’ self-reported use of experiential processes was
significantly related to dietary fat intake at baseline. Participants who reported more use of
experiential processes had lower dietary fat intake and this relationship was stronger at levels of
high stress rather than low stress. For example, when experiencing higher numbers of weekly
stressors, individuals who used strategies such as ‘I tell myself I can choose to eat a low fat diet’
or ‘I consider articles I have seen about lowering the amount of fat in my diet’ tended to have
lower levels of dietary fat than those who used less of these strategies. The association between
baseline variables of the TTM and dietary fat reduction, as addressed in research question #2, was
less clear. Participants in the intervention group significantly decreased their average dietary fat
intake and increased their use of behavioral and experiential processes over the six-month dietary
change program. Despite these changes, none of the elements of the TTM proved to be significant
predictors of dietary change in the original analysis. However, a follow-up analysis with a larger
42
sample and adequate power revealed a modest relationship between TTM variables and dietary fat
reduction. Furthermore, less use of experiential processes at baseline was related to greater
reductions in dietary fat intake at the six-month interval. These findings will be discussed below
in greater detail.
Predictors of Dietary Fat Intake
The investigation of psychosocial predictors of dietary fat intake revealed that the
combination of variables associated with the TTM, BMI, minor stress and interactions between
minor stress and the TTM variables accounted for an approximately 15% of the variance in dietary
fat intake for participants entering a dietary change program. Although this is a modest amount of
variance accounted for, it is comparable to other models investigating psychosocial factors and
diet (Glanz, et al., 1998). The significant relationship found between higher dietary fat intake and
less use of the experiential processes is difficult to directly compare with previous research
because most previous studies examine the association of TTM constructs to stage of change,
rather than to changes in a specific health behavior. However, extrapolating from previous
literature, if one considers that use of experiential processes tends to be lower in individuals that
have higher self-reported dietary fat, then the present results would seem to be consistent (Ounpuu
et al., 2000). In short, those who are thinking more (i.e., making commitments, increasing
awareness about high fat foods) about reducing dietary fat are more likely to have lower dietary fat
intake. The relationship between experiential processes and dietary fat was moderated by minor
stress. Individuals who continued to remind themselves of their commitment to dietary change or
continued to be aware of the negative impact of dietary fat when experiencing minor stressors
were more likely to have lower levels of dietary fat consumption. The reverse was also true, those
who did not use the processes as frequently were more likely to have higher dietary fat intake. It
may be that those individuals who are using the experiential processes have a higher level of
43
commitment to behavior change and during times of increased stressful events they are able to use
these strategies to their benefit. On the other hand, for some individuals, more stressors may lead
to a reduction in the use of these strategies as a result of competing demands (Stetson et al., 1997).
For example, individuals may have difficulty eating a low fat diet when dealing with many
stressful events and subsequently “give-up” on the diet or use less experiential processes during
that time frame. Since this present study does not test the causality of these behaviors, further
examination of these experiential processes would be necessary to determine the temporal time
frame between stressful events and negative cognitions.
The question remains, why was the experiential processes variable the only significant
predictor of dietary fat intake? The hypothesis that greater use of the behavioral processes would
predict lower dietary fat intake was not supported. Although individuals had not begun the
intervention, one would expect that those who actively engaged in behavioral strategies to reduce
dietary fat intake would actually have lower dietary fat consumption. One possibility is that the
composite behavioral processes variable may overlook the unique contribution of each individual
process. The behavioral processes variable in this study was composed of five different processes
(e.g., stimulus control, counterconditioning) and it may be that individual processes are more
predictive of actual fat intake than the total score of all behavioral strategies. Regarding selfefficacy and decisional balance, it is unclear as to why these variables were not predictive of
dietary fat intake. Higher self-efficacy was related to lower dietary fat intake at the univariate
level, but the relationship did not hold in the multivariate analysis. Furthermore, decisional
balance was not significantly related to dietary fat intake at the univariate or multivariate level.
Minor stressful events did not independently predict dietary fat intake and except for the
experiential processes variable, minor stress did not moderate the relationship between the
elements of the TTM and percentage of dietary fat intake. There was a limited range of stressful
44
events in this sample since a majority of participants endorsed average levels of stressful events.
While this supports that the number of events experienced by the participants is similar to other
investigations involving the WSI, it may be that higher levels of stress are required to capture the
relationship between stress and dietary fat intake. Additionally, recent researchers examining the
relation between perceived stress and fatty food intake in adolescents found a modest relationship
between stress and increased fat intake (Cartwright et al., 2003). The authors suggest that
everyday stress may be more diffuse and therefore likely to have weaker relationships with food
intake than more acute stressors (e.g., exam weeks) often used in dietary studies (Cartwright et
al., 2003). Given the weak relationships between stress and dietary fat intake and between many
the TTM variables and dietary fat intake, it is not surprising that minor stress was not a moderator
of the TTM-dietary fat relationship.
Predictors of Dietary Fat Reduction
There was not enough power to adequately test the planned analysis for the research
question investigating predictors of dietary fat reduction. Therefore, an additional analysis was
conducted that slightly modified the original question in order to have sufficient power to detect
significant differences. The research question addressed the same predictors of dietary fat
reduction but excluded the decisional balance questionnaire. This allowed for investigation of
predictors in the entire intervention group, as originally planned, instead of restricting the analysis
to only those participants who completed the decisional balance questionnaire.
Participants in the intervention group reported significant decreases in dietary fat from 31%
to 20%. With sufficient power, the combination of constructs associated with the TTM, BMI,
minor stress, and the interactions between minor stress and the TTM variables accounted for 8.4%
of the variance in dietary fat reduction. The experiential processes variable was a unique
contributor to dietary fat reduction for participants in the intervention group and the relationship
45
indicated that those participants who endorsed less use of the experiential processes prior to
beginning the intervention tended to have greater dietary fat reductions. This was contrary to the
hypothesis that those with more use of experiential processes at baseline (i.e., more awareness,
more frequent self-commitment) would have greater reductions in dietary fat. One possibility for
these results is that those who had difficulty making self-commitments or who did not pay
attention to information about dietary fat prior to the intervention actually benefited more from the
structure and support of an intervention.
Another possibility is that these participants had greater room to change their dietary fat
intake. The initial cross-sectional analysis on predictors of dietary fat intake demonstrated that
those who used less experiential processes tended to have higher dietary fat intake. Given the use
of the dietary fat change score as the dependent variable in this analysis, those with higher levels
of dietary fat could have the appearance of greater change than those who were already consuming
lower levels of dietary fat. For example, someone who was consuming 40% of their calories from
dietary fat at baseline and was not using experiential processes such as ‘paying attention to
information on dietary fat’ may be able to reduce their fat intake to 28%, whereas someone who
was eating 25% dietary fat at intake and successfully reduces to the dietary goal of 20% has the
appearance of less reduction, although they did achieve the study goal. Many of the reasons cited
in response to the first research question about the lack of relationship between the other TTM
variables apply to this analysis. However, it is interesting that despite increases in use of
behavioral processes over the six-month interval of the intervention, behavioral processes were not
predictive of dietary fat reduction. It may be that the TTM theory is not as effective a model for
dietary behaviors compared to other health behaviors. Further discussion of the adequacy of the
model will be detailed in a later section. However, it might be interesting to examine the change
in TTM constructs in relation to the change in dietary fat. Previous research conducted in primary
46
care clinics indicated the pattern of cross-sectional associations between psychological factors and
dietary fat intake was not the same as that observed longitudinally (Steptoe et al., 2000). The
author’s determined that self-efficacy and constructs similar to decisional balance did not predict
change in dietary fat but there were significant relationships between changes in the variables.
Change in TTM Constructs over Six Months
The hypotheses that elements of the TTM (self-efficacy, behavioral processes, and
experiential processes) would change over the course of the intervention was partially supported
by the hypothesized increase in behavioral processes over the six months. The behavioral
processes include common strategies for changing eating habits such as stimulus control, use of
distraction, and social support. Although the dietary intervention did not target the specific
behavioral processes construct, it did address and encourage use of many of the strategies. The
results suggest that use of behavioral strategies increased for individuals participating in a dietary
fat reduction intervention, lending some validity to the behavioral processes construct. It would
be interesting to follow the individuals over a longer period of time to determine if relapse in
dietary fat change coincided with a subsequent decrease in use of behavioral processes.
Self-efficacy increased for those who participated in the intervention, but it did not approach
significance due to multiple comparisons. It was hypothesized that increases in self-efficacy
would occur over the course of the intervention but further research would have been necessary to
determine the temporal nature of the relationship since one could argue that successful dietary
change could lead to increased self-efficacy rather than increased self-efficacy leading to dietary
change.
The increase in experiential processes over the six-month intervention was contrary to
expectation. In the smoking literature, the most well-researched area of the TTM model, levels of
experiential processes are highest prior to the initiation of behavior change (Prochaska et al.,
47
1991). However, discrepancies between health behaviors (e.g., smoking cessation, exercise, diet)
on the patterns of the processes of change were recently reported (Rosen, 2000). Specifically,
smoking cessation is best characterized as the termination of a negative behavior and exercise is
the initiation of a positive behavior. Dietary fat reduction is a combination of the cessation of a
negative behavior (e.g., eating less fatty foods) and initiation of a positive behavior (e.g., eating
healthy foods in place of the fatty foods). Rosen (2000) posits that change in dietary fat requires
continued use of both processes of change and that there is no taper in the use of the experiential
processes. Since decisions about dietary fat intake likely occur several times a day, continued
reminders to oneself about issues such as commitment to change and consequences of high dietary
fat intake may be necessary to sustain dietary change, lending support to the increase in
experiential processes over the course of the intervention.
Although this study was not a formal test of the TTM model, it calls into question the
relationship between the TTM constructs and actual health behavior. As stated previously, a
majority of the research in this area examines the TTM constructs across the stages of change.
This is somewhat problematic given the often-cited critique of the model that the stages of change
are tautological (Kristal et al., 1999). Specifically, the definition for each stage of change is based
on the health behavior itself. For example, to be in the action stage one must be actively eating
lower fat foods. Excluding the difficulties associated with estimating personal dietary fat intake,
one would expect the action stage of change to be associated with lower dietary fat. Therefore, it
would seem more useful to identify factors other than stage of change that would be associated
with specific health behaviors. Similarly to this study, two studies in the smoking literature have
attempted to examine the TTM constructs as predictors of smoking cessation (i.e., an actual health
behavior) independent of stage of change (Abrams, Herzog, Emmons, Linnan, 2000; Carlson,
Taenzer, Koopmans, Casebeer, 2003). The results of these studies also found minimal predictive
48
value in elements of the TTM above demographic variables and generated significant discussion
amongst researchers. Future research examining the TTM constructs in direct relation to health
behaviors appears warranted.
Limitations
Several methodological issues must be taken into account when interpreting the results of this
study. First, the intervention was not developed to test the transtheoretical model and the
constructs were not specifically targeted in the intervention. Furthermore, a majority of studies
examine the TTM variables through the stage of change concept, which also makes it difficult to
compare these results with previous research in the dietary fat area. Therefore, cautious
interpretation is required, particularly regarding hypotheses about the utility of the TTM variables
as predictors. Second, recent criticisms of percent dietary fat intake as an outcome have
implications for understanding and evaluating factors associated with dietary fat intake. Shepherd
(2002) argues that percentage of dietary fat is an outcome composed of many different behaviors
and is not a behavior in and of itself. For example, one must decrease dietary fat in several food
groups and also change food preparation behaviors in order to decrease the percentage of calories
from dietary fat. It is possible that one may decrease their dairy and meat sources of fat but may
continue to eat high fat snack foods. Although it would be very cumbersome to measure
individual dietary behaviors, the variety of behaviors may make it difficult for questionnaires to
identify useful predictors of dietary fat intake and reduction. Shepherd raises an interesting point
that needs to be addressed before prediction models of dietary fat intake can be improved. The
TTM model may need to be adjusted for dietary behavior in order to increase its usefulness in
understanding and predicting dietary change.
49
Conclusion
In summary, there was modest support of variables associated with the transtheoretical
model in explaining dietary fat intake and reduction in this study. More frequent use of
experiential processes under higher levels of stress was related to lower dietary fat intake
suggesting that reminding oneself about commitment to change and other cognitive processes may
be important during stressful times. Although less use of the experiential processes at baseline
was predictive of greater dietary fat reduction, there was little evidence that minor stress was a
useful moderator of the model in explaining dietary fat change. Finally, individuals who
participated in the intervention demonstrated increased use of the experiential and behavioral
processes.
Given the previously reported low success rates for dietary interventions in healthy
populations, it is promising that dietary fat intake was significantly reduced at six months in this
study. The success further supports the need to understand and identify factors related to dietary
fat intake and reduction. Research on dietary fat intake as it relates to the transtheoretical model is
still in the early phases. At this point future research may need to focus on the measurement
characteristics of the TTM constructs to ensure that they are adequately capturing the nuances of
health behavior change. Determining whether the constructs are associated with objective
behavior change rather than subjective motivations would be beneficial in order to adequately test
the model. Further research needs to be conducted to determine if the measurement of dietary fat
intake is the best way to assess TTM factors associated with interventions. Despite being
cumbersome, examination of individual behaviors (e.g., reducing dairy fat or eating lean meats)
may be more helpful in understanding how participants decrease their dietary fat intake. Finally,
future studies need to establish stronger relations between the TTM constructs and dietary fat
change, moving beyond cross sectional research to focus on experimentally increasing self50
efficacy, decisional balance, and processes of change in interventions in order to determine the
utility of the model in predicting behavior change.
51
REFERENCES
Abrams, D.B., Herzog, T.A., Emmons, K.M., & Linnan, L. (2000). Stages of change versus
addiction: A replication and extension. Nicotine & Tobacco Research, 2, 223-229.
AbuSabha, R. & Achterberg, C. (1997). Review of self-efficacy and locus of control for nutrition
and health related behavior. Journal of the American Dietetic Association, 97, 1122-1132.
Aiken, L.S. & West, S.G. (1991). Multiple regression: Testing and interpreting interactions.
Newbury Park, CA: Sage.
Anderson, J.W., Garrity, T.F., Wood, C.L., Whitis, S.E., Smith, B.M., & Oeltgen, P.R. (1992).
Prospective, randomized controlled comparison of the effects of low-fat and low-fat plus
high-fiber diets on serum lipid concentrations. American Journal of Clinical Nutrition, 56,
887-894.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-215.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman and Co.
Bandura, A. (1992). Exercise of personal agency through the self-efficacy mechanism. In R.
Schwarzer (Ed.), Self-efficacy: Thought control of action (pp. 3-38). Washington, DC:
Hemisphere.
Barnard, N.D., Akhtar, A., & Nicholson, A. (1995). Factors that facilitate compliance to lower fat
intake. Archives of Family Medicine, 4, 153-158.
Bellisle, F., Louis-Sylvestre, J., Linet, N., Rocaboy, B., Dalle, B., Cheneau, F., L’Hinoret, D., &
Guyot, L. (1990). Anxiety and food intake in men. Psychosomatic Medicine, 52, 452457.
Birch, L.L. (1999). Development of food preferences. Annual Review of Nutrition, 19, 41-62.
Bovbjerg, V.E., McCann, B.S., Brief, D.J. Follette, W.C., Retzlaff, B.M., Dowdy, A.A., Walden,
C.E., & Knopp, R.H. (1995). Spouse support and long-term adherence to lipid-lowering
diets. American Journal of Epidemiology, 141, 451-460.
Brantley, P.J., Catz, S.L., & Boudreaux, E. (1997). Daily stress inventory. In C.P. Zalaquett &
R.J. Wood (Eds.), Evaluating stress: A book of resources (pp. 97-112). Lanham, MD:
The Scarecrow Press.
Brantley, P.J. & Jones, G.N. (1993). Daily stress and stress-related disorders. Annals of
Behavioral Medicine, 15, 17-25.
Brantley, P.J., Jones, G.N., Boudreaux, E. & Catz, S.L. (1997). Weekly stress inventory. In C.P.
Zalaquett & R.J. Wood (Eds.), Evaluating stress: A book of resources (pp. 405-420).
Lanham, MD: The Scarecrow Press.
52
Brownell, K.D. & Cohen, L.R. (1995). Adherence to dietary regimens 2: Components of
effective interventions. Behavioral Medicine, 20, 155-164.
Brug, J., Hospers, H.J., & Kok, G. (1997). Differences in psychosocial factors and fat
consumption between stages of change for fat reduction. Psychology and Health, 12, 719727.
Brunner, E., White, I., Thorogood, M., Bristow, A., Curle, D., & Marmot, M. (1997). Can dietary
interventions change diet and cardiovascular risk factors? A meta-analysis of randomized
controlled trials. American Journal of Public Health, 87, 1415-1422.
Burkholder, G.J. & Evers, K.A. (2002). Application of the transtheoretical model to several
problem behaviors. In P.M. Burbank and D. Riebe (Eds.), Promoting exercise and
behavior change in older adults. New York: Springer Publishing Co.
Cannon, W.B. (1914). The interrelations of emotions as suggested by recent physiological
researches. American Journal of Physiology, 25, 256-282.
Carlson, L.E., Taenzer, P., Koopmans, J., & Casebeer, A. (2003). Predictive value of aspects of
the Transtheoretical Model on smoking cessation in a community–based, large group
cognitive behavioral program. Addictive Behaviors, 28, 725-740.
Carmody, T.P. (1989). Affect regulation, nicotine addiction, and smoking cessation. Journal of
Psychoactive Drugs, 21, 331-42.
Carmody, T.P., Matarazzo, L.D., & Istvart, J.A. (1987). Promoting adherence to heart healthy
diets: A review of the literature. Journal of Compliance and Health Care, 2, 105-115.
Cartwright, M., Wardle, J., Steggles, N., Simon, A.E., Croker, H., & Jarvis, M.J. (2003). Stress
and dietary practices in adolescents. Health Psychology, 22, 362-369.
Chrisler, J.C. (1997). Adherence to weight loss and nutritional regimens. In D.S. Gochman
(Ed.), Handbook of Health Behavior Research (pp. 323-333). New York: Plenum Press.
Cohen, S. & Lichtenstein, E. (1990). Perceived stress, quitting smoking, and smoking relapse.
Health Psychology, 9, 466-478.
Cooper, M.L., Russell, M., Skinner, J.B., Frone, M.R., & Mudor, P. (1992). Stress and alcohol
use: The moderating effects of gender, coping, and alcohol expectancies. Journal of
Abnormal Psychology, 101, 139-152.
Curry, S.J., Kristal, A.R., & Bowen, D.J. (1992). An application of the stage model of behavior
change to dietary fat reduction. Health Education Research, 7, 97-105.
Delongis, A., Coyne, J.C., Dakof, G., Folkman, S., & Lazarus, R.S. (1982). Relationship of daily
hassles, uplifts, and major life events to health status. Health Psychology, 1, 119-136.
53
DiClemente, C.C. & Hughes, S.O. (1990). Stages of change profiles in outpatient alcoholism
treatment. Journal of Substance Abuse, 2, 217-235.
DiClemente, C.C. & Prochaska, J.O. (1982). Self-change and therapy change of smoking
behavior: A comparison of processes of change in cessation and maintenance. Addictive
Behavior, 7, 133-142.
DiClemente, C.C., Prochaska, J.O., & Gibertini, M. (1985). Self-efficacy and the stages of
change. Cognitive Therapy and Research, 9, 181-2000.
Dougall, A.L. & Baum, A. (2001). Stress, health, and illness. In A. Baum, T.A. Revenson, &
J.E. Singer (Eds.), Handbook of Health Psychology (pp. 235-247). New Jersey: Lawrence
Erlbaum Associates.
Drewnowski, A. (2002). Taste, taste preferences, and body weight. In C.G. Fairburn & K.D.
Brownell (Eds.), Eating disorders and obesity (pp. 50-54). New York: Guilford Press.
Eertmans, A., Baeyens, F., & VandenBergh, O. (2001). Food likes and their relative importance
in human eating behavior: Review and preliminary suggestions for health promotion.
Health Education Research, 16, 443-456.
Ehnholm, C., Huttunen, J.K., Pietinen, P., Leino, U., Mutanen, M., Kostiainen, E., Pikkarainene,
J., Dougherty, R., Iacono, J., & Pushka, P. (1982). Effect of diet on serum lipoproteins in
a population with a high risk of coronary heart disease. New England Journal of Medicine,
307, 850-855.
Federation of American Societies for Experimental Biology, Life Sciences Research Office.
(1995). Third report on nutrition monitoring in the United States, Vol. 1-2. Washington,
DC: U.S. Government Printing Office.
Feunekes, G.I.J., de Graaf, C., Meyboom, S., & van Staveren, W.A. (1998). Food choice and fat
intake of adolescents and adults: Associations of intakes with social networks. Preventive
Medicine, 27, 645-656.
Field, A.E., Barnoya, J., & Colditz, G.A. Epidemiology and health and economic consequences of
obesity. In T.A. Wadden & A.J. Stunkard (Eds.), Handbook of Obesity Treatment (pp. 318). New York: Guilford Press.
Galavotti, C., Cabral, R.J., Lansky, A., Grimley, D.M., Riley, G.E., & Prochaska, J.O. (1995).
Validation of measures of condom and other contraceptive use among women at high risk
for HIV infection and unintended pregnancy. Health Psychology, 14, 570-578.
Garrett, V.D., Brantley, P.J., Jones, G.N., & McKnight, G.T. (1991). The relation between daily
stress and Crohn’s disease. Journal of Behavioral Medicine, 14, 87-95.
Glanz, K., Kristal, A.R., Tilley, B.C., & Hirst, K. (1998). Psychosocial correlates of healthful
diets among male auto workers. Cancer Epidemiology, Biomarkers, & Prevention, 7, 119126.
54
Glanz, K., Lewis, F.M., & Rimer, B.K. (1997). Linking theory, research, and practice. In K.
Glanz, F.M. Lewis, B.K. Rimer (Eds.), Health behavior and health education (pp. 19-36).
San Francisco: Jossey-Bass Publishers.
Glanz, K., Patterson, R.E., Kristal, A.R., DiClemente, C.C., Heimendinger, J., Linnan, L., &
McLerran, D.F. (1994). Stages of change in adopting healthy diets: Fat, fiber, and
correlates of nutrient intake. Health Education Quarterly, 21, 499-519.
Glanz, K., Patterson, R.E., Kristal, A.R., Feng, Z., Linnan, L., Heimendinger, J., & Hebert, J.R.
(1998). Impact of work site health promotion on stages of dietary change: The Working
Well trial. Health Education and Behavior, 25, 448-463.
Gorder, D.D., Bartsch, G.E., Tillotson, J.L., Grandits, G.A., & Stamler J. (1997). Food group and
macronutrient intakes, trial years 1-6, in the special intervention and usual care groups in
the Multiple Risk Factor Intervention Trial. American Journal of Clinical Nutrition, 65 (1
Suppl), 258S-271S.
Green, S.B. (1991). How many subjects does it take to do a regression analysis? Multivariate
Behavioral Research, 26, 499-510.
Greene, G.W. & Rossi, S.R. (1998). Stages of change for reducing dietary fat intake over 18
months. Journal of the American Dietetic Association, 98, 529-534.
Greene, G.W., Rossi, S.R., Rossi, J.S., Velicer, W.F., Fava, J.L., & Prochaska, J.O. (1999).
Dietary applications of the stages of change model. Journal of the American Dietetic
Association, 99, 673-678.
Greeno, C.G. & Wing, R.R. (1994). Stress-induced eating. Psychological Bulletin, 115, 444-464.
Griffin, K.W., Friend, R., Eitel, P., & Lobel, M. (1993). Effects of environmental demands,
stress, and mood on health practices. Journal of Behavioral Medicine, 16, 643-661.
Grunberg, N.E. & Straub, R.O. (1992). The role of gender and taste class in the effects of stress
on eating. Health Psychology, 11, 97-100.
Holahan, C.K. & Holahan, C.J. (1987). Life stress, hassles, and self-efficacy in aging: A
replication and extension. Journal of Applied Social Psychology, 17, 574-592.
Holmbeck, G.N. (2002). Post-hoc probing of significant moderational and mediational effects in
studies of pediatric populations. Journal of Pediatric Psychology, 27, 87-96.
Holmes, T.H. & Rahe, R.H. (1967). The social readjustment rating scale. Journal of
Psychosomatic Research, 11, 213-218.
Janis, I.L. & Mann, L. (1977). Decision-making: A psychological analysis of conflict, choice,
and commitment. New York: Free Press.
55
Kanner, A.D., Coyne, J.C., Schaefer, C., & Lazarus, R.S. (1981). Comparison of two modes of
stress measurement: Daily hassles and uplifts versus major life events. Journal of
Behavioral Medicine, 4, 1-39.
Kavanagh, D.J., Gooley, S., & Wilson, P.H. (1993). Prediction of adherence and control in
diabetics. Journal of Behavioral Medicine, 16, 509-522.
Kelsey, K.S., Kirkley, B.G., DeVeelis, R.F., Earp, J.A., Ammerman, A.S., Keyserling, T.C.,
Shannon, J., & Simpson, R.J. (1996). Social support as a predictor of dietary change in a
low-income population. Health Education Research, 11, 383-395.
Kristal, A.R., Glanz, K., Curry, S.J., & Patterson, R.E. (1999). How can stages of change be best
used in dietary interventions. Journal of the American Dietetic Association, 99, 679-684.
Kristal, A.R., Glanz, K., Tilley, B.C., & Li, S. (2000). Mediating factors in dietary change:
Understanding the impact of a worksite nutrition intervention. Health Education and
Behavior, 27, 112-125.
Kristal, A.R., Hedderson, M.M., Patterson, R.E., & Neuhauser, M.L. (2001). Predictors of selfinitiated, healthful dietary change. Journal of the American Dietetic Association, 101,
762-766.
Kumanyika, S.K., Van Horn, L., Bowen, D., Perri, M.G., Rolls, B.J., Czajkowski, S.M., & Schron,
E. (2000). Maintenance of dietary behavior change. Health Psychology, 19, 42-56.
Laitinen, J., Ek, E., & Sovio, U. (2002). Stress-related eating and drinking behavior and body
mass index and predictors of this behavior. Preventive Medicine, 34, 29-39.
Law, M.R., Wald, N.J., & S.G. Thompson (1994). By how much and how quickly does reduction
in serum cholesterol concentration lower risk of ischaemic heart disease? British Journal
of Medicine, 308, 367-373.
Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.
Lee, C. & Bobko, P. (1994). Self-efficacy beliefs: Comparison of five measures. Journal of
Applied Psychology, 79, 364-369.
Levine, G., Keaney J., & Vita, J. (1995). Cholesterol reduction in cardiovascular disease. New
England Journal of Medicine, 332, 512-521.
Lewis, C.J., Sims, L.S., & Shannon, B. (1989). Examination of specific nutrition/health
behaviors using a social cognitive model. Journal of the American Dietetic Association,
89, 194-202.
Lowe, M.R. (2002). Dietary restraint and overeating. In C.G. Fairburn & K.D. Brownell (Eds.),
Eating disorders and obesity (pp. 88-92). New York: Guilford Press.
Lovejoy, J. C., Champagne, C.M., Smith, S.R., de Jonge, L., Xie, H. (2001). Ethnic differences
in dietary intakes, physical activity, and energy expenditure in middle-aged,
56
premenopausal women: The Healthy Transitions Study. American Journal of Clinical
Nutrition, 74, 90-95.
Marcus, B.H., Rossi, J.S., Selby, V.C., Niaura, R.S., & Abrams, D.B. (1992). The stages and
processes of exercise adoption and maintenance in a worksite sample. Health Psychology,
11, 386-395.
Marcus, B.H., Selby, V.C., Niaura, R.S., & Rossi, J.S. (1992). Self-efficacy and the stages of
exercise behavior change. Research Quarterly for Exercise and Sport, 63, 60-66.
Martin, P.D., Francis, J.L., Wood, K., Rhode, P.C., O’Hea, E., & Lefevre, M. (2001, November).
Stress as a barrier to dietary change. Poster session presented at the annual meeting of the
Association for the Advancement of Behavior Therapy, Philadelphia.
McCann, B.S., Bovbjerg, V.E., Brief, D.J., Turner, C., Follette, W.C., Fitzpatrick, V., Dowdy, A.,
Retzlaff, B., Walden, C.E., & Knopp, R.H. (1995). Relationship of self-efficacy to
cholesterol lowering and dietary change in hyperlipidemia. Annals of Behavioral
Medicine, 17, 221-226.
McCann, B.S., Retzlaff, B.M., Dowdy, A.A., Walden, C.E., & Knopp, R.H. (1990). Promoting
adherence to low-fat, low cholesterol diets: Review and recommendations. Journal of
American Dietetic Association, 90, 1408-1414.
McCann, B.S., Warnick, G.R., & Knopp, R.H. (1990). Changes in plasma lipids and dietary
intake accompanying shifts in perceived workload and stress. Psychosomatic Medicine,
52, 97-108.
McCaul, K.D., Glasgow, R.E., & Schafer, L.C. (1987). Diabetes regimen behaviors: Predicting
adherence. Medical Care, 25, 868-881.
McAuley, E. & Courneya, K.S. (1993). Adherence to exercise and physical activity as health
promoting behaviours: Attitudinal and self-efficacy influences. Applied and Preventive
Psychology, 2, 65-77.
Michaud, C., Kahn, J.P., Musse, N., Burlet, C., Nicolas, J.P., & Mejean, L. (1990). Relationships
between a critical life event and eating behaviour in high-school students. Stress Medicine,
6, 57-64.
Monroe, S.M. (1983). Major and minor life events as predictors of psychological distress:
Further issues and findings. Journal of Behavioral Medicine, 6, 89-205.
Monroe, S.M. & McQuaid, J.R. (1994). Measuring life stress and assessing its impact on mental
health. In W.R. Avison and I.H. Gotlib (Eds.), Stress and Mental Health: Contemporary
issues and prospects for the future. New York: Plenum Press.
National Cholesterol Education Program. Adult Treatment Panel III Report. [online] 2001;
Available from URL: http://www.nhlbi.nih.gov/guidelines/cholesterol/index.html.
57
Nestle, M., Wing, R., Birch, L., DiSorgra, L., Drewnowski, A., Middleton, S., Sigman-Grant, M.,
Sobal, J., Winston, M., & Economos, C. (1998). Behavioral and social influences on food
choice. Nutrition Reviews, 56, S50-S64.
Neuhouser, M.L., Kristal, A.R., & Patterson, R.E. (1999). Use of nutrition labels is associated
with lower fat intake. Journal of the American Dietetic Association, 99, 45-50.
Ni Mhurchu, C., Margetts, B.M., & Speller, V.M. (1997). Applying the stages-of-change model
to dietary change. Nutrition Reviews, 55, 10-16.
Oliver, G. & Wardle, J. (1999). Perceived effects of stress on food choice. Physiology &
Behavior, 66, 511-515.
Ounpuu, S., Woolcott, D.M., & Greene, G.W. (2000). Defining stage of change for lower-fat
eating. Journal of the American Dietetic Association, 100, 674-679.
Patterson, R.E., Kristal, A.R., & White, E. (1996). Do beliefs, knowledge, and perceived norms
about diet and cancer predict dietary change? American Journal of Public Health, 86,
1394-1400.
Peyrot, M., McMurry, J.F., & Kruger, D.F. (1999). A biopsychosocial model of glycemic control
in diabetes: Stress, coping, and regimen adherence. Journal of Health and Social
Behavior, 40, 141-158.
Polivy, J. & Herman, C.P. (2002). Experimental studies of dieting. In C.G. Fairburn & K.D.
Brownell (Eds.), Eating disorders and obesity (pp. 84-87). New York: Guilford Press.
Pollard, T.M., Steptoe, A., Canaan, L., Davies, G.J., & Wardle, J. (1995). The effects of
academic examination stress on eating behaviour and blood lipid levels. International
Journal of Behavioral Medicine, 2, 299-320.
Povey, R., Conner, M., Sparks, P., James, R., & Shepherd, R. (1999). A critical examination of
the application of the Transtheoretical Model’s stages of change to dietary behaviours.
Health Education Research, 14, 641-651.
Prochaska, J.O. (1984). Systems of psychotherapy: A transtheoretical analysis. (2nd ed.). Pacific
Grove, California: Brooks-Cole.
Prochaska, J.O. & DiClemente, C.C. (1983). Stages and processes of self-change in smoking:
Towards an integrative model of change. Journal of Consulting and Clinical Psychology,
51, 390-395.
Prochaska, J.O., DiClemente, C.C, & Norcross, J.C. (1992). In search of how people change.
American Psychologist, 47, 1102-1113.
Prochaska, J.O., Norcross, J.C., Fowler, J.L., Follick, M.J., & Abrams, D.B. (1992). Attendance
and outcome in a worksite weight control program: Processes and stages of change as
process and predictor variables. Addictive Behaviors, 17, 35-45.
58
Prochaska, J.O., Redding, C.A., & Evers, K.E. (1997). The transtheoretical model and stages of
change. In K. Glanz, F.M. Lewis, B.K. Rimer (Eds.), Health behavior and health
education (pp. 60-84). San Francisco: Jossey-Bass Publishers.
Prochaska, J.O., Velicer, W.F., Guadagnoli, E., & Rossi, J.S. (1991). Patterns of change:
Dynamic typology applied to smoking cessation. Multivariate Behavioral Research, 26,
83-107.
Prochaska, J.O., Velicer, W.F., Rossi, J.S., Goldstein, M.G., Marcus, G.H., Rakowski, W., Fiore,
C., Harlow, L.L., Redding, C.A., Rosenbloom, D., & Rossi, S.R. (1994). Stages of
change and decisional balance for 12 problem behaviors. Health Psychology, 12, 39-46.
Proctor, V.E., Tesfa, A., & Tompkins, D.C. (1999). Barriers to adherence to highly active
antiretroviral therapy as expressed by people living with HIV/AIDS. AIDS Patient CareSTDS, 13, 535-544.
Rabkin, J.G. & Struening, E.L. (1976). Life events, stress and illness. Science, 194, 1013-1020.
Rakowski, W., Fulton, J.P., & Feldman, J.P. (1993). Women’s decision making about
mammography: A replication of the relationship between stages of adoption and
decisional balance. Health Psychology, 12, 209-214.
Roos, E., Lahelma, E., Virtanen, M., Prattala, R., & Pietinen, P. (1998). Gender, socioeconomic
status, and family status as determinants of food behaviour. Social Science & Medicine,
46, 1519-1529.
Rosen, C.S. (2000). Is the sequencing of change processes by stage consistent across health
problems? A meta-analysis. Health Psychology, 19, 593-604.
Rossi, J.S. (1992, August). Common processes of change across nine problem behaviors. Paper
presented at the meeting of the American Psychological Association, Washington, DC.
Rossi, S.R. (1993). Application of the transtheoretical model of behavior change to dietary fat
reduction in a naturalistic environment. Unpublished doctoral dissertation, University of
Rhode Island, Kingston, RI.
Rossi, S.R., Greene, G.W., Reed, G., Rossi, J.S., Prochaska, J.O., & Velicer, W.F. (1994a).
Continued investigation of a processes of change measure for dietary fat reduction
[Abstract]. Annals of Behavior Medicine, 16(suppl), 167.
Rossi, S.R., Greene, G.W., Reed, G., Rossi, J.S., Prochaska, J.O., & Velicer, W.F. (1994b). Cross
validation of a decisional balance measure for dietary fat reduction [Abstract]. Annals of
Behavior Medicine, 16(suppl), 167.
Rossi, S.R., Greene, G.W., Reed, G., Rossi, J.S., Prochaska, J.O., & Velicer, W.F. (1994c).
Development of a self-efficacy questionnaire for dietary fat reduction [Abstract]. Annals
of Behavior Medicine, 16(suppl), 167.
59
Rossi, S.R., & Rossi, J.S. (1993, March). Processes of change for dietary fat reduction. Paper
presented at the annual meeting of the Society for Behavioral Medicine, San Francisco,
CA.
Rossi, S.R., Rossi, J.S., Prochaska, J.O., & Velicer, W.F. (1993, August). Measurement structure
of a decisional balance questionnaire for dietary fat reduction. Paper presented at the
annual meeting of the American Psychological Association, Toronto, Canada.
Roussouw, J.E. (1994). The effects of lowering serum cholesterol on coronary heart disease risk.
Medical Clinics of North America, 78, 181-195.
Sarason, I.G., Johnson, J.H., & Siegel, J.M. (1978). Assessing the impact of life changes:
Development of the life experience survey. Journal of Consulting and Clinical
Psychology, 46, 932-946.
Schlundt, D.G., Rea, M.R., Kline, S.S., & Pichert, J.W. (1994). Situational obstacles to dietary
adherence for adults with diabetes. Journal of the American Dietetic Association, 94, 874876.
Selye, H. (1956). The stress of life. New York: McGraw-Hill.
Selye, H. (1982). History of the stress concept. In L. Goldberger & S. Breznitz (Eds.), Handbook
of stress: Theoretical and clinical aspects (pp. 7-17). New York: The Free Press.
Shepherd, R. (2002). Resistance to changes in diet. Proceedings of the Nutrition Society, 61,
267-272.
Shimakawa, T., Sorlie, P., Carpenter, M.A., Dennis, B., Tell, G.S., Watson, R., & Williams, O.D.
(1994). Dietary intake patterns and sociodemographic factors in the atherosclerosis risk in
communities study. Preventive Medicine, 23, 769-780.
Sorensen, G., Thompson, B., Glanz, K., Feng, Z., Kinne, S., DiClemente, C., Emmons, K.,
Heimendinger, J., Probart, C., & Lichtenstein, E. (1996). Work site-based cancer
prevention: Primary results from the Working Well Trial. American Journal Public
Health, 86, 939-47.
Sporny, L.A. & Contento, I.R. (1995). Stages of change in dietary fat reduction: Social
psychological correlates. Journal of Nutrition Education, 27, 191-199.
Steptoe, A. (1991). The links between stress and illness. Journal of Psychosomatic Research, 35,
633-644.
Steptoe, A., Doherty, S., Kerry, S., Rink, S., & Hilton, S. (2000). Sociodemographic and
psychological predictors of changes in dietary fat consumption in adults with high blood
cholesterol following counseling in primary care. Health Psychology, 19, 411-419.
60
Steptoe, A. & Wardle, J. (1999). Motivational factors as mediators of socioeconomic variations
in dietary intake patterns. Psychology and Health, 14, 391-402.
Steptoe, A., Wijetunge, S., Doherty, S., & Wardle, J. (1996). Stages of change for dietary fat
reduction: Associations with food intake, decisional balance, and motives for food choice.
Health Education Journal, 55, 108-122.
Stetson B.A., Rahn, J.M., Dubbert, P.M., Wilner, B.I., & Mercury, M.G. (1997). Prospective
evaluation of the effects of stress on exercise adherence in community-residing women.
Health Psychology, 16, 515-520.
Stone, A.A. & Brownell, K. (1994). The stress-eating paradox: Multiple daily measurements in
adult males and females. Psychology & Health, 9, 425-436.
Stotland, S. & Zuroff, D.C. (1991). Relations between multiple measures of dieting self-efficacy
and weight change in a behavioral weight control program. Behavior Therapy, 22, 47-59.
Tabachnick, B.G. & Fidell, L.S. (1996). Using multivariate statistics. New York: HarperCollins.
Tilley, B.C., Glanz, K., Kristal, A.R., Hirst, K., Li, S., Vernon, S.W., & Myers, R. (1999).
Nutrition intervention for high-risk auto workers: Results of the next step trial. Preventive
Medicine, 28, 282-292.
United States Department of Health and Human Services- Public Health Service (1995). Healthy
People 2000: Midcourse review and 1995 revisions. Hyattsville: National center for
Health Statistics.
Urban, N., White, E., Anderson, G.L., Curry, S., & Kristal, A.R. (1992). Correlates of
maintenance of a low-fat diet among women in the Women’s Health Trial. Preventive
Medicine, 21, 279-291.
Van Duyn, M.S., Kristal, A.R., Dodd, K., Campbell, M.K., Subar, A.F., Stables, G., Nebeling, L.,
& Glanz, K. (2001). Association of awareness, intrapersonal and interpersonal factors,
and stage of dietary change with fruit and vegetable consumption: A national survey.
American Journal of Health Promotion, 16, 69-78.
Van Horn, L., Dolecek, T.A., Grandits, G.A., & Skweres, L. (1997). Adherence to dietary
recommendations in the special intervention group in the Multiple Risk Factor Intervention
Trial. American Journal of Clinical Nutrition, 65(Suppl 1), 289S-304S.
Van Horn, L. & Kavey, R.E. (1997). Diet and cardiovascular disease prevention: What works?
Annals of Behavioral Medicine, 19, 197-212.
Velicer, W.F., DiClemente, C.C., Rossi, J.S., & Prochaska, J.O. (1985). Decisional balance
measure for assessing and predicting smoking status. Journal of Personality and Social
Psychology, 48, 1279-1289.
61
Wardle, J., Steptoe, A., Oliver, G., & Lipsey, Z. (2000). Stress, dietary restraint and food intake.
Journal of Psychosomatic Research, 48, 195-202.
Weidner, G., Kohlmann, C.W., Dotzauer, E., & Burns, L.R. (1996). The effects of academic
stress on health behaviors in young adults. Anxiety Stress and Coping, 9, 123-133.
Weinberger, M., Hiner, S.L., & Tierney, W.M. (1987). In support of hassles as a measure of
stress in predicting health outcomes. Journal of Behavioral Medicine, 10, 9-31.
Wiebe, D.J. & McCallum, D. M. (1986). Health practices and hardiness as mediators in the
stress-illness relationship. Health Psychology, 5, 425-438.
Winkleby, M.A., Flora, J.A., & Kraemer, H.C. (1994). A community-based heart disease
intervention: Predictors of change. American Journal of Public Health, 84, 767-772.
62
VITA
Jennifer Leanne Francis was born in Baltimore, Maryland. She received her Bachelor of Science
degree in psychology from James Madison University in Harrisonburg, Virginia, in 1996.
Following the accrual of her bachelor’s degree she gained research experience at Western State
Psychiatric Hospital in Staunton, Virginia, under Michael Shutty, Ph.D. In 1997, she began her
clinical psychology graduate training at Louisiana State University in Baton Rouge, under the
direction of David L. Penn, Ph.D. Following Dr. Penn’s departure from Louisiana State
University in 1999 she completed the remainder of her graduate training with Phillip J. Brantley,
Ph.D. She received her internship training at West Virginia School of Medicine in Morgantown,
West Virginia. Currently, she is completing a two-year post-doctoral fellowship at Brown
University in Providence, Rhode Island.
63