When weight management lasts. Lower perceived rule complexity
Jutta Mataa,b,*, Peter M. Todda,c, Sonia Lippked
aCenter for Adaptive Behavior and Cognition, Max Planck Institute for Human Development, Lentzeallee 94, 14195 Berlin, Germany
bInternational Max Planck Research School ‘‘The Life Course: Evolutionary and Ontogenetic Dynamics (LIFE)’’, Berlin, Germany
cCognitive Science Program, Indiana University, Bloomington, 1101 E. 10th Street, Bloomington, IN 47405, USA
dDepartment of Health Psychology, Freie Universita ¨t Berlin, Habelschwerdter Allee 45, 14195 Berlin, Germany
Across the Western world, people overeat and often choose
unhealthy foods: almost one in three individuals in the U.S. is
classified as obese (e.g., Baskin, Ard, Franklin, & Allison, 2005). In at
least half of the European Union member states, prevalence levels
of obesity in the population are higher than 20% (Fry & Finley,
2005). In Germany (where our study takes place), this rate is
around 23% (Prugger & Keil, 2007).
Obesity is a major cause of morbidity and mortality (Allison,
Fontaine, Manson, Stevens, & VanItallie, 1999), but still, numbers
are rising (e.g., Ezatti, Martin, Skjold, Van der Hoorn, & Murray,
2006). Attempting to lose weight is a popular remedy; as many as
38% of women and 24% of men in the United States try to lose
weight (e.g., Kruger, Galuska, Serdula, & Jones, 2004), with similar
numbers in Great Britain (Wardle & Johnson, 2002) and Germany
(Westenhoefer, 2001). Yet, few people successfully maintain
weight loss over time (Jeffery et al., 2000), making it vital to
understand how to extend adherence to weight loss and weight
Numerous studies have shown that adherence to changes in
eating behavior is predicted by social-cognitive factors, such as
self-efficacy, intention, and planning (e.g., Gollwitzer & Sheeran,
2006; Kremers, De Bruijn, Schaalma, & Brug, 2004; Luszczynska,
Sobczyk, & Abraham, 2007; Schwarzer et al., 2007), and by
previous behavior such as past diet attempts (Teixeira, Going,
Sardinha, & Lohman, 2005). One aspect that has been little
considered is the role of environment factors, both actual and
perceived environment structure (e.g., Andajani-Sutjahjo, Ball,
Warren, Inglis, & Crawford, 2004; Kurzenha ¨user & Hertwig, 2008;
Wansink, 2006) which food researchers have repeatedly shown to
be significant in other contexts. For instance, Wansink (1996)
showed that larger-sized packages of spaghetti or other foods
lead people to serve themselves more (for an overview of
environmental influences on food choice see Wansink, 2004).
Appetite 54 (2010) 37–43
A R T I C L EI N F O
Received 27 December 2008
Received in revised form 4 September 2009
Accepted 8 September 2009
Weight loss program adherence
Cox hazard regressions
A B S T R A C T
Maintaining behavior change is one of the major challenges in weight management and long-term
weight loss. We investigated the impact of the cognitive complexity of eating rules on adherence to
weight management programs. We studied whether popular weight management programs can fail if
participants find the rules too complicated from a cognitive perspective, meaning that individuals are
not able to recall or process all required information for deciding what to eat. The impact on program
adherence of participants’ perceptions of eating rule complexity and other behavioral factors known to
influence adherence (including previous weight management, self-efficacy, and planning) was assessed
via a longitudinal online questionnaire given to 390 participants on two different popular weight
management regimens. As we show, the regimens, Weight Watchers and a popular German recipe diet
(Brigitte), strongly differ in objective rule complexity and thus their cognitive demands on the dieter.
Perceived rule complexity was the strongest factor associated with increased risk of quitting the
cognitively demanding weight management program (Weight Watchers); it was not related to
adherence length for the low cognitive demand program (Brigitte). Higher self-efficacy generally helped
in maintaining a program. The results emphasize the importance of considering rule complexity to
promote long-term weight management.
? 2009 Elsevier Ltd. All rights reserved.
§Author Note: The authors thank Franziska Hoffmann and Jens Rademacher for
their help in programming the online study; all webpages and internet chat rooms
that allowed us to post the link to our study, especially Lars Penke, Heiko Ernst, and
Susanne Gerlach for their help in this regard; as well as all participants.
* Corresponding author at: Department of Psychology, Stanford University,
Building 420 Jordan Hall, 450 Serra Mall, Stanford, CA 94035, USA.
E-mail addresses: firstname.lastname@example.org (J. Mata), email@example.com
(P.M. Todd), firstname.lastname@example.org (S. Lippke).
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/appet
0195-6663/$ – see front matter ? 2009 Elsevier Ltd. All rights reserved.
Kurzenha ¨user and Hertwig (2008) found that perceived environ-
ment structureinfluencedfood choice, particularlythose cuesseen
as most salient in the cafeteria environment they studied,
including variety of foods offered and queue length to get different
foods. Finally, coming closer to our focus in this paper, a
community survey on barriers to weight maintenance found that
26% of young Australian women cited not having enough
information about healthy nutrition as an important barrier,
indicatingproblemswith theirinformationenvironmentrelated to
eating (Andajani-Sutjahjo et al., 2004).
Here we extend the exploration of environment structure
influences on health behaviors to cognitive aspects of the environ-
ment, investigating how the perception of the cognitive complexity
of rules to be followed affects health behavior change in a weight
management context. In contrast to previous studies which
emphasized the influences of either the objective or the subjective
environment, here we attempt to compare both the actual
environment structure and how it is perceived. The cognitive
environment we study consists of the eating rules of weight
that has to be processed to follow a weight management program.
People make decisions in many domains using simple rules
(Gigerenzer, Todd, & the ABC Research Group, 1999; Kahneman,
Slovic, & Tversky, 1982), including decisions about what to eat
(Scheibehenne, Miesler, & Todd, 2007; Wansink, 2006). Schei-
behenne and colleauges showed that a heuristic based on one
decision cue predicted individual food choices as accurately as a
more complex algorithm that took nine different decision-relevant
factors into account. This suggests that human food choices might
be based on very simple rules. Related results emerged from a
recent ‘‘diet challenge’’ in which participants were encouraged to
simplify their eating rules by making small environmental
changes, including eating from smaller plates or not serving meals
family style (Wansink, 2006; Wansink, Just, & Payne, 2009). In this
case, a large proportion of participants who made simple changes
to theirenvironmentlost moreweight thanacontrol groupdoinga
face-to-face intervention with more complex behavior changes
(Wansink,2008).Inspired by theseresultsandthe assumptionthat
it may be easier to adhere to simple rather than complex eating
rules (e.g., Lally,Chipperfield, & Wardle, 2008), in thispaper we ask
whether the complexity of eating rules prescribed by weight
management programs negatively affects adherence to those
programs. We investigate cognitive complexity of two different
popular weight management programs, both objectively and
subjectively, and study its influence on how long individuals stay
on a given program.1
The complexity of a task has been defined as the result of the
demands imposed by the structure of the task’’ (Robinson, 2001, p.
29). Research on concept learning has demonstrated that more
complex concepts are harder to learn (Feldman, 2003). Robinson
further showed that an objectively more complex task was also
rated as subjectively more complex by participants. Cognitive
resource demands arising from following eating rules may be
especially pressing for people trying to lose weight or facing diet
restrictions, who have been found to perform worse on cognitive
tasks than do non-dieters (Kemps, Tiggemann, & Grigg, 2008;
Kemps, Tiggemann, & Marshall, 2005; Vreugdenburg, Bryan, &
Kemps, 2003). According to some studies, trying to lose weight
seems to selectively impair working memory, affecting perfor-
mance on mental arithmetic (Vreugdenburg et al., 2003) and word
length tasks (Shaw & Tiggemann, 2004). In turn, such mental
arithmetic could be involved in keeping track of calories in weight
management programs, while the ability to recall word lists after
some delay could tap into the same memory processes that people
on a weight management program have to use. Thus, trying to lose
weight may deplete the very resources needed to follow complex
eating rules—and may consequently make the rules seem more
Is a weight management program’s cognitive complexity a
significant predictor for the duration of adherence to the program,
beyond motivational factors of behavior change? We hypothesize
that if eating rules are perceived as too complicated from a
cognitive point of view, people will be less likely to remember and
process the information required by the rules, have more difficulty
thus to abandon the subjectively-judged cognitively complex
weight management program as a whole.
This hypothesis will be tested with two different approaches:
first, an objective rule analysis based on the bestselling diet books
for the two diets and second, a subjective longitudinal online
study. We focused on two of Germany’s most frequently used
weight management programs, Weight Watchers and Brigitte.
These two programs were chosen due to their high prevalence and
differing cognitive requirements: Weight Watchers assigns point
values to every food and instructs participants to eat only a certain
number of points per day; points can be compensated for with
physical activity or partly saved up to offset a special occasion.
Brigitte is a recipe-based weight management program designed
by oneof Germany’smostpopularwomen’smagazinesofthe same
name, which provides recipes and shopping lists for every meal,
thus requiring participants to simply follow the provided meal
The rule complexity of the Brigitte and Weight Watchers
programs were evaluated via the corresponding bestselling books
according to www.amazon.de (retrieved 22 June 2006; see Dost,
2004; Gerlach, Klosterfelde-Wentzel, & Khaschei, 2003). All rules
from each book were individually assessed, counting how many
values, amounts of food, or quantities of drinks (e.g., number of
points, portions of fruit, or glasses of water) participants had to
count and keep track of over the course of 1 day, representing
arithmetic processing demands.
The online study took the form of an Internet questionnaire
(programmed in HTML; data administration with the software
dynQuest—Rademacher & Lippke, 2007) administered to each
participant at three measurement points, each 4 weeks apart. After
the third measurement point, or after stopping their weight
management program, participants received individualized feed-
back on their Body Mass Index (BMI), planning and coping
strategies, and nutrition knowledge.
In contrast to other weight management studies where all
participants start a program at the same time and evaluations of
adherence are done after 6 months or a year, our participants did
not have a common starting point for their programs. Rather, we
studied a time window of roughly 8 weeks within the course of
1This is different from what is assessed in other weight loss research studying
low-complexity cognitive style in dieting, including aspects such as dichotomous
thinking (Byrne, Cooper, & Fairburn, 2004). Cognitive complexity in our analysis
does not mean, for example, whether participants classify foods strictly into ‘‘good’’
and ‘‘bad’’ categories. It rather refers to the perception of how difficult it is to
remember or calculate information relevant to the weight management program,
such as how many ‘‘points’’ a food has and keeping track of the number of points
eaten over the day.
J. Mata et al./Appetite 54 (2010) 37–43
participants’ individual weight management efforts. The advan-
tage of this procedure is that we captured participants in the
natural course of their weight management and thus were able to
get a wider range of timing data for adhering to or quitting a
This method of studying participants in their natural weight
management context (cf. Foster et al., 2003), rather than assigning
them to a particular regimen, provides an especially strict test of
our hypothesis, because if participants in our study self-selected
their programs according to their cognitive abilities, this would
reduce the effects of perceived rule complexity on adherence
behavior, making the expected effect more difficult to observe.
We assessed the following measures in our study. If not
mentioned otherwise, the measure values reported all refer to the
first measurement point.
Perceived complexity. We measured four items assessing the
participant’s perceived complexity of (1) processing the informa-
tion necessary to eat according to their weight management
program, (2) deciding how much of a food one is allowed to
consume, (3) knowing how to replace one food with another, and
Watchers diet, how difficult do you find it to remember how many
points you have consumed in the course of a day?’’). Items were
assessed on a 5-point scale from very easy to very difficult, with
Cronbach’s a = .80. Participants only evaluated the complexity of
the program they were currently following.
Weight management goals. Participants indicated how long they
planned to stick to the rules of their current weight management
program in number of days, weeks, months, years, ‘‘permanently,’’
or ‘‘until I reach my goal weight.’’
Social-cognitive variables. Participants rated statements adapted
from Schwarzer et al. (2007), on a 4-point scale, from strongly
disagree to strongly agree, as follows: self-efficacy used three
statements such as ‘‘I am sure that I can stick to my weight
management program even if I need a long time to develop the
necessary routines’’; Cronbach’s a = .63. Intentions used three
statements such as ‘‘I intend to eat in accordance with my weight
management program’s rules’’; Cronbach’s a = .74. Planning used
six statements such as ‘‘I have planned in detail the occasions
during which I will stick to my weight management rules’’;
Cronbach’s a = .74.
Previous weight loss attempts. Participants stated how many
times they had followed a weight management program before.
Demographics. Participants provided their age, sex, extent of
school and professional education, height, and current weight.
Time on current weight management program. We obtained two
different measures of time spent so far on the current weight
management program at the third measurement point: we asked
how long participants had been following their weight manage-
mentprogram intotal,and weadded the timepassed betweenfirst
and third measurement to the length of time on the weight
management program stated at the first measurement point. The
measures did not differ (t(133) = ?1.14, p = .25). For those who had
stopped their weight management attempt during the course of
our study,we usedtheir estimate of how long they had been on the
weight management program in total (‘‘Approximately how long
did you stick to your weight management regimen?’’), with
possible answer units in days, weeks, months, and years. Time on
current weight management program is our main dependent
variable because it is an alternative means of measuring weight
loss success: higher attrition—equivalent to less time spent on a
weight management program—goes along with lower weight loss
(Teixeira et al., 2005).
Goal attainment. Only those participants who stopped their
program during the study (reported at the second and third
measurement point) were asked ‘‘Did you reach your desired
weight?’’ (with five answer options ranging from ‘‘No, I have lost
muchless weightthanIplanned’’ to‘‘Ihavelostmuch moreweight
than I planned’’) and ‘‘Did you stick to the weight management
program for as long as you had planned?’’ (with analogous
answers; adapted from Berry, Danish, Rinke, & Smiciklas-Wright,
Participants were recruited from 23 German-language Internet
chat rooms dealing with weight management, a German Internet
portal for psychological studies online, and via the websites of a
Characteristics of participants.
BrigitteWeight WatchersDifference: Brigitte?Weight Watchers
Adherence total % (t1, t2 and t3)
Attrition total % (t2 and t3)
Weight at t1 (BMI); mean (SD)
Weight loss (BMI; t3?t1); mean (SD)
Age; mean (SD)
?10-year school certificate
13-year school certificate
?3-year vocational training
Profession at the time
In university or professional education
Home (housewife, unemployed, retired)
Weight management goals
Note. t1, t2, and t3 refer to the first, second, and third measurement point, respectively. Remaining percentages in rows that do not add up to 100% correspond to other
categories not shown here because they represent a very small part of the sample.
J. Mata et al./Appetite 54 (2010) 37–43
large-circulation German women’s magazine and a popular
science magazine on psychology. Followers of Weight Watchers
and Brigitte weight management regimens provided a sufficient
number of participants (N = 390) to yield reliable results. Male
participants (1.0% of the sample) and participants who reached
their weight management goals (time or weight) during the study
period (5.6%) were excluded because there were too few people in
these categories for a separate reliable analysis. Furthermore,
participants who reached their weight management goals and
were thus considered successful (in contrast to those who stopped
their diet without reaching their goal) were not the main focus of
the study and so were also excluded.
Participants in this sample were slightly overweight and rather
well-educated (Tables 1 and 2). The average attrition rate at 39.7%
was lower than that of other Internet studies on health behavior
(cf. 59% reported by Schwarzer et al., 2007). The 63 Brigitte (45.3%)
and 80 Weight Watchers (31.8%) followers who did not respond at
either the second or third measurement point (and so were
dropped through attrition) did not differ from those that did
respond in age, BMI, family status, education, time on weight
management program, number of previous weight loss attempts,
or social-cognitive factors.
The two groups, Brigitte and Weight Watchers, differ sig-
were on average older. This age difference is also reflected in the
fact that Brigitte participants had a lower proportion of current
university students and a higher proportion of university
graduates than Weight Watchers. Brigitte participants were more
likelyto haveaweightgoalfor theirweight managementprogram;
Weight Watchers more often had a time goal. Women on Weight
Watchers reported a higher number of previous diet attempts.
concrete planning than Brigitte participants.
Statistical analyses were performed using SPSS version 16.0.
Differences in perceived complexity and adherence between
different diet programs were calculated with Analyses of Variance.
Our main statistical analyses were Cox hazard regression analyses.
Cox regressions model the time that passes until a specified event
occurs, such as quitting a weight management program, based on
predictor variables.Here we reporteffects of predictorson dropout
rate as odds ratios—the greater the ratio, the more likely dropout is
to occur. Cox regressions assume that predictors influence the
outcome equally at all time points; this proportionality of hazards
assumption was fulfilled in our data, meaning our predictors
equally influenced program adherence independent of how long
participants had been eating according to their weight manage-
ment rules at the time of the study. A major advantage of Cox
regressionsisthat theymakeuseofvaryinglengths oftimeperiods
in longitudinal studies: data of participants with an as-of-yet
in their weight management program at the last measurement
point, are nonetheless considered in the odds ratios (Tabachnick &
Comparison of diets regarding their objective and perceived
The objective complexity measures from the book analysis
indicate that the Weight Watchers program is more cognitively
complex than Brigitte (28 units vs. 2 units necessary to calculate
and remember, respectively). Given the differences in cognitive
processing required by the two programs, we also expected
differences in perceived complexity, with Weight Watchers
being higher. However, from the ratings of study participants,
Brigitte and Weight Watchers rules do not differ significantly
in this regard (t(385) = 1.67, p = .10; see Table 2 for means).
This may be because people who are already following one of
the programs feel that the regimen is acceptably difficult—
otherwise they would not be on it. To test whether this active
regimen participation makes a difference in perceived complex-
ity judgments, we asked an additional 99 women (age: M = 36.5
years, SD = 11.2; body mass index: M = 27.5 kg/m2, SD = 6.9) who
had never followed either Brigitte or Weight Watchers to assess
the complexity of both (using the same four measurement items;
Cronbach’s a = .79 and a = .88 for Brigitte and Weight Watchers,
respectively); these unbiased participants rated Brigitte’s per-
ceived rule complexity to be significantly lower than that of the
Weight Watchers program (M = 2.73 SD = 0.86 vs. M = 3.21
SD = 0.69; respectively, t(98) = ?4.26, p < .001), matching our
objective complexity comparison. Furthermore, among another
51 women who had done Weight Watchers and Brigitte, and thus
had self-selected into both programs, they again perceived no
difference in complexity between the diets (on an individual
basis), just like the women in our main study following one diet
or the other (compared at the group level; N = 51; M = 2.81
SD = 0.77 vs.
M = 2.58
SD = 0.80;
p = .14).
t(50) = 1.52,
Adherence to different weight management programs
No significant differences were found in mean adherence times
between the programs (t(359) = 0.48, p = .64). This is largely due to
the great variance in adherence times within each program (see
Table 2 for means and SDs). Furthermore, given the very small
differences in mean perceived complexity ratings between
followers of the two diets, we would consequently not expect
mean adherence rates to differ between them. Instead, the critical
Weight management program performance of participants.
BrigitteWeight WatchersDifference: Brigitte?Weight Watchers
Time adhered to diet (weeks)
Number previous diet attempts
Perceived complexity (t1)
Note. N for each cell: see Table 1.
J. Mata et al./Appetite 54 (2010) 37–43
point seems to be the individual differences in adherence across
followers within each weight management program, which we
Perceived cognitive complexity as a predictor for the length
of program adherence
To investigate the impact on adherence of individual differ-
ences between participants within the same weight management
program, we ran hierarchical Cox hazard regressions with
variables entered sequentially. Number of previous weight loss
attempts was entered in the first step, followed by self-efficacy,
intention, and planning in the second step. Lastly, perceived
complexity of the weight management rules was added (see
Table 3 for correlations between variables, and Table 4 for results
on all predictors in the third step).
For participants on the Brigitte program, a higher number of
previous weight loss attempts increased the odds to quit by 11%,
and self-efficacy decreased these odds by 53%. Concrete planning
decreased the probability to quit by 47%, which is marginally
significant. Neither intention nor perceived cognitive complexity
was a significant predictor. For participants on Weight Watchers,
higher perceived complexity raised the probability of quitting by
None of the other factors was a significant predictor. To test
whether these results might be influenced by outliers, we also ran
all analyses again, omitting participants whose time on the current
weight management program differed more than three standard
deviations from the mean. These analyses showed no different
outcomes from those including all participants.
In summary, as predicted, higher perceived rule complexity
greatly increased an individual’s likelihood of quitting the Weight
Watchers regimen, a weight management program relying heavily
on computation. It did not however matter for the Brigitte
program, where objective cognitive complexity was low. In
agreement with the literature, higher values on social-cognitive
factors decreased the odds of stopping either weight management
program unsuccessfully, whereas a higher number of previous
weight loss attempts slightly increased the odds.
Taken together, results from the rule analysis and the Cox
rules (a) differs objectively between programs, and (b) has a
negative subjective effect on the time participants adhere to a
computationally demanding program. At the same time, perceived
complexity was also rated as different for the two programs by
women who had never followed either program, but as the same
whenrated by women whohad already started one program or the
other. It is possible that how complexity is perceived is influenced
by not only the objective complexity of the rules, but also by non-
environmental factors regarding the participants themselves,
including social-cognitive characteristics, age, education, experi-
ence (i.e., number of previous weight management programs), or
employment status (e.g., active vs. retired). In our analyses, only
the social-cognitive factors were significantly associated with
perceived complexity, and none particularly strongly (self-efficacy
rBrigitte= ?.21, rWW= ?.18; all ps < .05; see Table 3).
rWW= ?.23; planning
In this study we extended previous research on how
characteristics of the environment influence eating behavior by
considering both the structure and the perception of the cognitive
environment, specifically testing the effect of a weight manage-
compared the influence of perceived environment structure with
that of social-cognitive factors on leaving a weight management
program prematurely, and found that—in the Weight Watchers
program with complex eating rules and thus heightened cognitive
demands—perceptions of rule complexity had the greatest impact.
we found, were comparable to or larger than other observational
studies on eating behavior using Cox hazard regressions. For
instance, Dalle Grave et al. (2005) studied which factors predicted
leaving a weight management program at 6 and 12 months and
reported increases in odds ratios between 1.12 and 1.42, compared
to ours ranging up to 1.54. In contrast, perceived complexity does
not play much of a role for individual adherence to Brigitte, which
also is in line with the findings from our rule analysis showing that
this recipe-based program places little cognitive demand on
Atthesame time,the twoperceivedcomplexitymeasures differ
little on average between the groups of Weight Watchers and
Brigitte followers. This lack of mean perceived difference could
arise if people tend to self-select their weight management
program such that it fits their (cognitive) abilities, leading to a sort
of equilibrium in perceived difficulty at which people are able to
stick to their eating rules for at least some amount of time. Our
comparison of women who had been on both of the diets (thus had
self-selected them, possibly based on their cognitive abilities) with
Correlation table for predictors included in the Cox proportional hazard regression.
Note. All predictors refer to the first measurement point.
Cox proportional hazard regression predicting dropout for individual programs.
interval of the
Previous weight loss attempts
J. Mata et al./Appetite 54 (2010) 37–43
individuals who had not followed either diet supported this idea.
The rule complexity perception of individuals inexperienced with
both Weight Watchers and Brigitte reflected the results of our
objective rule analyses showing that Weight Watchers rules are
more complex. However, dieters who had followed both programs
showed no difference in their perceived complexity of each. These
findings suggest that on average people seem to choose diets that
they perceive as not being too difficult. This also means that our
study design was not likely to pick up differences in perceived
complexity between diets due to this self-selection effect. We
would expect to find a difference in perceived complexity between
diets in a randomized controlled trial, where people are randomly
assigned to different diets with rules of varying degrees of
complexity and therefore cannot select a diet program that best
suits their abilities—matching what we found with the inexper-
ienced group who rated the diets without ever having followed
them. However, this possibility still awaits further empirical
An additional explanation for not finding perceived complexity
differences between diets could be that social-cognitive factors
affected the perception of rule complexity in a way that resulted in
comparable difficulty perception between the
management programs. We did find small but significant
correlations between the social-cognitive factors and perceived
rule complexity, raising the concern that perceived complexity is
not measuring anything additional to those factors. However,
perceived complexity does make an additional contribution to
length of adherence to a weight management program after those
factors were controlled for in the Cox hazard regressions, showing
the unique importance of perceived complexity on dieting
Strength of intentions did not have a significant impact on
program adherence, presumably because the intention to manage
one’s weight is a necessary precondition to start a weight
management program. Because participants in our study were
already attempting to lose weight, they had very high intention
scores (90% of participants rated their intention strength at 3 or 4
on a 4-point scale); thus, with this little variance (i.e., a ceiling
effect), intention could not be a meaningful predictor here.
While one limitation of our study is reliance on self-report
measures, we designed it to reach a large, diverse audience of
participants attempting to lose weight longitudinally, and the
advantage is that the participants involved were motivated to take
their self-reporting seriously and provide useful data. However,
the important findings should certainly be followed up using other
methods (e.g., controlled laboratory studies and randomized
trials). This would also overcome the fact that participants on
the Brigitte regimen in our study differed from those on Weight
Watchers in a few aspects (though largely less important ones). To
test the importance of perceived complexity across different
subsamples (e.g., people with different diet experience, self-
efficacy, or time vs. weight goals), future research could compare
the roleofperceivedcomplexityfordietadherenceacross matched
subsamples of dieters along with experimental manipulations of
diet rules. Our study does not have enough power to detect
differences in matched subsamples; for example, when we
analyzed subgroups of dieters depending on how long they have
been on the diet (e.g., comparing people who just started the diet
with those that adhered to their diet longer and have more
experience), the magnitude of odds ratios is comparable to those
using the entire sample, but the effects are not significant.
Furthermore, the two diet regimens also differ in types of
additional support they provide to participants, such as weigh-
ins or group meetings (which in our additional online study we
found that approximately a third of the dieters make some use of).
This has to be considered when interpreting the results.
Other studies have shown that the ease of making changes in
eating behavior influences compliance with and maintenance of
those changes and any subsequent weight loss, especially for
changes related to the physical environment (e.g., Wansink, 2006).
Wansink et al. (2009) suggested that instead of vigorously
monitoring one’s calories or other units of food intake, making
small environmental changes that are easy to realize—such as
using smaller plates, not eating while watching TV, or limiting
snacks to three bites—could be very powerful contributors to
successful weight control. Therefore,future studies comparingdiet
adherence should consider not only the role of cognitive rule
complexity as explored in this paper—the ease of the cognitive
environment—but also the ease with which weight-control related
changes can be realized in a person’s physical (or social)
The long-term success of different weight management
programs should be measured not just in terms of direct weight
loss, but also, as here, in terms of how long people stick to their
program. The underlying assumption is that staying longer on a
healthful weight loss or weight management program increases
the likelihood of controlling weight or achieving generally better
health: short-term diets do not generate lasting effects (Mann
et al., 2007), while successful weight management programs
involve long-lasting lifestyle change, including nutritional changes
(Powell, Calvin, & Calvin, 2007). Thus, designing weight manage-
ment rules that can be adhered to for a long period or an entire
lifetime—including by making rules that are not perceived as being
too complex—could help limit the spread of overweight and
Allison, D. B., Fontaine, K. R., Manson, J. E., Stevens, J., & VanItallie, T. B. (1999). Annual
deaths attributable to obesity in the United States. Journal of the American Medical
Association, 282, 1530–1538.
Andajani-Sutjahjo, S., Ball, K., Warren, N., Inglis, V., & Crawford, D. (2004). Perceived
personal, social and environmental barriers to weight maintenance among young
women. A community survey. International Journal of Behavioral Nutrition and
Physical Activity, 1, 15.
Baskin, M. L., Ard, J., Franklin, F., & Allison, D. B. (2005). Prevalence of obesity in the
United States. Obesity Reviews, 6, 5–7.
Berry, M., Danish, S., Rinke, W., & Smiciklas-Wright, H. (1989). Work-site health
promotion. The effects of a goal-setting program on nutrition-related behaviors.
Journal of the American Dietary Association, 89, 914–920.
Byrne, S. M., Cooper, Z., & Fairburn, C. G. (2004). Psychological predictors of weight
regain in obesity. Behaviour Research and Therapy, 42, 1341–1356.
DalleGrave,R.,Calugi,S.,Molinari,E.,Petroni,M.L.,Bondi, B., Compare,A., etal.(2005).
Weight loss expectations in obese patients and treatment attrition. An observa-
tional multicenter study. Obesity Research, 13, 1961–1969.
Dost K. (2004). Weight Watchers. Der 4 Wochen Power Plan. [Weight Watchers. The 4-
week power plan]. Munich, Germany: Graefe & Unzer.
Ezatti, M., Martin, H., Skjold, S., Van der Hoorn, S., & Murray, C. J. L. (2006). Trends in
national and state-level obesity in the USA after correction for self-report bias.
Analysis of health surveys. Journal of the Royal Society of Medicine, 99, 250–
Feldman, J. (2003). The simplicity principle in human concept learning. Current
Directions in Psychological Science, 12(6), 227–232.
Foster, G. D., Wyatt, H. R., Hill, J. O., McGuckin, B. G., Brill, C., Mohammed, B. S., et al.
(2003). Randomized trial of a low-carbohydrate diet for obesity. The New England
Journal of Medicine, 348, 2082–2090.
Fry, J., & Finley, W. (2005). The prevalence and costs of obesity in the EU. Proceedings of
the Nutrition Society, 64, 359–362.
Gerlach, S., Klosterfelde-Wentzel, M., & Khaschei, K. (2003). Brigitte Ideal-Dia ¨t. [Brigitte
ideal diet]. Munich, Germany: Wilhelm Goldmann.
Gigerenzer, G., Todd, P. M., & the ABC Research Group, (1999). Simple heuristics that
make us smart. New York: Oxford University Press.
Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achieve-
ment. A meta-analysis of effects and processes. Advances in Experimental Social
Psychology, 38, 69–119.
Jeffery, R. W., Drewnoswski, A., Epstein, L. H., Stunkard, A. J., Wilson, G. T., & Wing, R. R.
(2000). Long-term maintenance of weight loss. Current status. Health Psychology,
19(Suppl 1), 5–16.
J. Mata et al./Appetite 54 (2010) 37–43
Kahneman, D., Slovic, P., & Tversky, A. (Eds.). (1982). Judgment under uncertainty: Download full-text
heuristics and biases. Cambridge, UK: Cambridge University Press.
Kemps, E., Tiggemann, M., & Grigg, M. (2008). Food cravings consume limited cognitive
resources. Journal of Experimental Psychology: Applied, 14, 247–254.
Kemps, E., Tiggemann, M., & Marshall, K. (2005). Relationship between dieting to lose
weight and the functioning of the central executive. Appetite, 45, 287–294.
Kremers, S. P. J., De Bruijn, G.-J., Schaalma, H., & Brug, J. (2004). Clustering of energy
balance-related behaviours and their intrapersonal determinants. Psychology &
Health, 19, 595–606.
Kruger, J., Galuska, D. A., Serdula, M. K., & Jones, D. A. (2004). Attempting to lose weight.
Specific practices among U.S. adults. American Journal of Preventive Medicine, 26,
Kurzenha ¨user, S., & Hertwig, R. (2008). Food choice in the cafeteria. Environmental and
preferential determinants. Unpublished manuscript.
Lally, P., Chipperfield, A., & Wardle, J. (2008). Healthy habits. Efficacy of simple advice
on weight control based on a habit-formation model. International Journal of
Obesity, 32, 700–707.
Luszczynska, A., Sobczyk, A., & Abraham, C. (2007). Planning to lose weight. RCT of an
and obese women. Health Psychology, 26, 507–512.
Mann, T., Tomiyama, J., Westling, E., Lew, A.-M., Samuels, B., & Chatman, J. (2007).
Medicare’s search for effective obesity treatments. Diets are not the answer.
American Psychologist, 62, 220–233.
Powell, L. H., Calvin, J. E., III, & Calvin, J. E., Jr. (2007). Effective obesity treatments.
American Psychologist, 62, 234–246.
Prugger, C., & Keil, U. (2007). Entwicklung der Adipositas in Deutschland—
Gro ¨ßenordnung, Determinanten und Perspektiven [Development of obesity in
Germany—prevalence, determinants and perspectives]. Deutsche medizinische
Wochenschrift, 132, 892–897.
Rademacher, J. D. M., & Lippke, S. (2007). Dynamic online surveys and experiments
with the free open source software dynQuest. Behavior Research, 39, 415–426.
Robinson, P. (2001). Task complexity, task difficulty, and task production. Exploring
interactions in a componential framework. Applied Linguistics, 22, 27–57.
Scheibehenne, B., Miesler, L., & Todd, P. M. (2007). Fast and frugal food choices.
Uncovering individual decision heuristics. Appetite, 49, 578–589.
Schwarzer, R., Schu ¨z, B., Ziegelmann, J. P., Lippke, S., Luszczynska, A., & Scholz, U.
(2007). Adoption and maintenance of four health behaviors. Theory-guided long-
itudinal studies on dental flossing, seat belt use, dietary behavior, and physical
activity. Annals of Behavioral Medicine, 33, 156–166.
Shaw, J., & Tiggemann, M. (2004). Dieting and working memory. Preoccupying cogni-
tion and the role of the articulatory control process. British Journal of Health
Psychology, 9, 175–185.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston:
Allyn and Bacon.
Teixeira, P. J., Going, S. B., Sardinha, L. B., & Lohman, T. G. (2005). A review of
psychosocial pre-treatment predictors of weight control. Obesity Reviews, 6,
Vreugdenburg, L., Bryan, J., & Kemps, E. (2003). The effect of self-initiated weight-loss
dieting on working memory. The role of preoccupying cognitions. Appetite, 41,
Wansink, B. (2004). Environmental factors that increase the food intake and consump-
tion volume of unknowing consumers. Annual Review of Nutrition, 24, 455–479.
Wansink, B. (2006). Mindless eating:why we eat more than we think.New York: Bantam-
Wansink, B. (2008). The National Mindless Eating Challenge. Dietary profiles predict a
differential effectiveness of interventions. The Journal of the Federation of American
Societies for Experimental Biology, 22, 1096.
Wansink, B., Just, D. R., & Payne, C. R. (2009). Mindless eating and healthy heuristics for
the irrational. American Economic Review, 99, 165–169.
in British adults. International Journal of Obesity, 26, 1144–1149.
Westenhoefer, J. (2001). Prevalence of eating disorder and weight control practices in
Germany in 1990 and 1997. International Journal of Eating Disorders, 29, 477–
J. Mata et al./Appetite 54 (2010) 37–43