Full text of DOI:10.1207/S15327558IJBM0702_4

INTERNATIONAL JOURNAL OF BEHAVIORAL MEDICINE, 7(2), 143–159
Copyright © 2000, Lawrence Erlbaum Associates, Inc.
Age and Body Make a Difference
in Optimistic Health Beliefs
and Nutrition Behaviors
Britta Renner, Nina Knoll, and Ralf Schwarzer
Nutrition behaviors are governed by health beliefs such as risk perceptions, outcome
expectancies, and optimistic self-beliefs. This study deals with the role that objective
criteria such as age and body weight might play in forming subjective beliefs. The
question is whether they can deter people from forming an overly optimistic judgment
of their health risk. Six kinds of verbal judgments were assessed, namely self-reported
health, vulnerability toward cardiovascular diseases, nutrition outcome expectancies,
nutrition self-efficacy, intentions to change one’s diet, and reported nutrition behav-
iors. In a sample of 1,583 men and women between 14 and 87 years of age, these judg-
ments were statistically related to age and body weight. It was found that people do
take their objective risk status into account, but only to a certain degree. The self-serv-
ing bias continues to exist throughout all age groups and weight levels. Moreover, it
was found that individuals report better intentions to adhere to healthy foods and
better nutrition behaviors as they grow older and gain weight.
Key words: age, body weight, nutrition, self-efficacy, optimism
Weight control and preventive nutrition are considered to be important health be-
haviors that people should maintain, along with physical activity, to attain health,
longevity, and fitness. Following a healthy diet, low in saturated fat and high in fi-
ber, is a popular medical recommendation. According to present medical knowl-
edge, such nutrition helps to prevent cardiovascular disease and other ailments.
Britta Renner, Institut für Psychologie, Ernst-Moritz-Arndt Universität Greifswald, Greifswald,
Germany; Nina Knoll and Ralf Schwarzer, Gesundheitpsychologie, Freie Universität Berlin, Berlin,
Germany.
Correspondence concerning this article should be addressed to Britta Renner, Psychologie,
Ernst-Moritz-Arndt Universität Greifswald, Franz-Mehring-Strasse 47, 17487 Greifswald, Germany.
E-mail: renner@mail.uni-greifswald.de

144
RENNER, KNOLL, SCHWARZER
However, most individuals do not adhere to this health behavior, and many have
not even contemplated adopting it (Ogden, 1996). The Berlin Risk Appraisal and
Health Motivation Study was launched to examine the sociocognitive factors of
health behavior change in various domains, such as physical exercise, smoking, al-
cohol consumption, and preventive nutrition. In this analysis, the focus will be on
the latter, examining to what degree nutrition-related beliefs are based on objective
health indicators such as age and body weight.
Objective health status is related to age because people are more prone to de-
generative diseases and multimorbidity with increasing age (Steinhagen-Thiessen
&
Borchelt, 1999). Also, objective health status is related to body weight because
overweight individuals are at a higher risk of suffering from cardiovascular and
other diseases (e.g., Assmann, 1993; Bray, 1978; WHO Study Group, 1990).
Therefore, health beliefs should account for this continuous decline and disadvan-
tage in some way. Conversely, it has been found that people do not picture the ac-
tual risk they are in, although they do understand that others are at risk. They often
think, “A negative event may happen to others, but it won’t happen to me.” For ex-
ample, asking people how they would judge their risk of having a heart attack,
compared to that of an average person of the same sex and similar age (the “aver-
age” risk), typically yields a “below average” estimate (e.g., Hahn & Renner,
1
998). This bias in risk perception, which has been coined unrealistic optimism or
optimistic bias (Perloff & Fetzer, 1986; Weinstein, 1980), reflects the difference
between the perceived risk of oneself and that of others within the same reference
group and belongs to the broader construct of defensive optimism (Schwarzer,
1
994). Weinstein (1996) reviewed about 200 empirical studies devoted to the issue
of unrealistic optimism. However, most of these studies were based on college stu-
dents or young adults, which raises the question of whether defensive optimism
may be restricted to adolescents and young adults, groups that may be especially
prone to unrealistic optimism, as suggested by some developmental researchers
(Arnett, 1995; Elkind, 1974). One could argue that adolescents as well as young
adults display unrealistic optimism because they have not experienced a decline in
health. This line of reasoning has received indirect support from empirical findings
reported by Weinstein (1980, 1982, 1987). He found that unrealistic optimism de-
creased with illness experience. He argued that in the absence of past illness expe-
rience, people may mistakenly conclude that they are exempt from future risk.
Following this line of reasoning, the elderly should make more realistic inferences
from their illnesses or those of their age group and consequently judge their own
vulnerability in the same manner as that of their peers. Conversely, one could ar-
gue that defensive optimism holds up as a life-long motivational tendency to pro-
tect oneself from threatening thoughts and, by this, to shield mental health (Taylor
&
Brown, 1988). There is no clear empirical evidence to support this assumption
but it is supported indirectly by research on well-being. For example, Staats et al.
1993) reported that optimistic perceptions of age and future quality of life are
(

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
145
maintained in older persons. Furthermore, recent studies have documented that
older persons are not unhappier than middle-aged or younger persons (e.g., Filipp,
1
996; Mroczek & Kolarz, 1998). Following this theoretical position, one would
expect a systematic underestimation of one’s own health risk as compared to that
of others in any age group.
Furthermore, to more fully elucidate the relation between age and optimistic
health beliefs, it is important to take other objective characteristics, in addition to
age, into account. In the context of nutrition-related beliefs, body weight is cer-
tainly an influential variable. Overweight individuals are expected to take their
disadvantage into account when making health-related judgments. Thus, they
should consider themselves as being at higher risk than people of average weight.
But, on the other hand, they might still persist in distorted social comparisons, dis-
playing unrealistic optimism. Apparently, there are no previous findings of an in-
teraction between age and objective risk status characteristics such as body weight.
This leads to the question of whether body weight adds explanatory value over and
beyond the influence of age. For example, one could assume that people display a
tendency for self-defensive compensation, relying more on positive health charac-
teristics while discounting negative ones.
To initiate and maintain health behaviors, it is not sufficient to perceive a se-
rious risk for oneself. In addition, one should also perceive positive action–out-
come contingencies and believe in one’s own capability to perform the intended
behavior (Schwarzer & Fuchs, 1996; Schwarzer & Renner, in press). This leads
to alternative conceptions of optimism that are adaptive rather than defensive in
nature because they imply instrumental coping behaviors. This functional opti-
mism (Schwarzer, 1994; Taylor, 1989) includes positive outcome expectancies,
but also personal resources, such as perceived self-efficacy (Bandura, 1997).
Outcome expectancies (e.g., “If I stick to a low-fat diet, then I will reduce my
risk of suffering a heart attack”) can be the most influential beliefs in the moti-
vation to change nutrition behavior (Schwarzer, 1992, 1999). Perceived self-effi-
cacy (e.g., “I can manage to stick to healthy food even if something delicious
but unhealthy is served”) portrays individuals’ beliefs in their capabilities to ex-
ercise control over challenging demands and over their own functioning. Such
optimistic self-beliefs shape the goals people set for themselves, what courses of
action they choose to pursue, how much effort they invest in given endeavors,
and how long they persevere in the face of barriers and setbacks. These beliefs
do not simply reflect the present coping competencies of an individual but
slightly overestimate them to become effective. It is assumed that defensive and
functional optimism share a common ground, such as the motive to protect or
enhance self-esteem. Therefore, individuals may display defensive as well as
functional optimism. One could reason that poorer health due to aging or risk
factors such as being overweight will lead to a reduction of defensive and func-
tional optimism: One feels more at risk and becomes doubtful of one’s capabil-

146
RENNER, KNOLL, SCHWARZER
ity of performing the intended health behavior. On the other hand, aging is
associated with more experience in managing daily activities as well as initiating
behavior changes despite various obstacles. This suggests that despite failing
health, individuals may become more capable of coping with obstructions in order
to maintain optimistic self-beliefs. Hence, people may acknowledge their health
risks with increasing age or body weight but would still regard their personal
coping resources in an optimistic way.
This study attempts to elaborate on aging, body weight, and optimistic health
beliefs. The research question is to what degree individuals realistically consider
their age and body weight when making judgments about their health risks, per-
sonal resources, and health behaviors. Initially, it examines whether young adults
are more likely than the elderly to overstate their own invulnerability to harm (de-
fensive optimism) and their personal coping resources (functional optimism). Sec-
ondly, it questions whether body weight is associated with both forms of
optimism. And finally, is considers whether body weight adds explanatory value
over and beyond the influence of age.
METHOD
Procedure
Sixty six percent of the participants were recruited through advertisements placed
in local newspapers in Berlin, Germany. In addition, a letter describing the study
was sent to people insured with the Technician’s Health Insurance Agency
(Techniker Krankenkasse) who lived near the four study locations (two universities
and two city halls). Upon arrival, they were informed that the study was part of a
community-wide effort to collect health information from the population at various
locations of the city. Those who agreed to participate completed a brief question-
naire that included five items assessing subjective health and risk perception. After-
wards, they were thanked and received a more detailed questionnaire that included
2
0 items assessing positive outcome expectancies, perceived self-efficacy, behav-
ioral intentions, and nutrition behavior. This one was completed at home and sent
back in an enclosed envelope.
Measures
Self-reported health status. The following question was asked to assess
subjective health status: “Would you say your health in general is excellent, very
good, good, fair, or poor?” This measure was formulated in line with the health
status question of the National Health Interview Survey (Schechter, 1993).

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
147
Risk perception. Risk estimates were obtained for each of the following
four cardiovascular diseases: hypertension, heart disease, stroke, and heart attack.
To assess the perceived absolute own risk, participants were asked to estimate the
likelihood of experiencing each particular health problem, for example, “How high
do you think is your risk of becoming hypertensive during your life time?” (see
Perloff & Fetzer, 1986). The same question was asked for an average target person
of one’s own age and sex (absolute other’s risk; e.g., “How likely do you think is
it that someone else of your sex and age will become hypertensive during his or
her life time?”). Responses were made on 7-point Likert scales ranging from –3
(extremely unlikely) to 3 (extremely likely).
Positive outcome expectancies were measured by seven items. Participants
were asked: “What do you think will be the consequences for yourself if you adopt
a low-fat diet?” Following this header, responses were elicited to seven more spe-
cific questions: “If I stick to a low-fat diet, then …” (a) “I will feel physically more
attractive,” (b) “I will feel better mentally,” (c) “I will have no (or fewer) body
weight problems,” (d) “I will lower my cholesterol level,” (e) “I will lower my
blood pressure,” (f) “I will be healthier,” and (g) “I will reduce my risk of suffering
a heart attack.” Responses were made on 4-point scales ranging from 1 (strongly
disagree) to 4 (strongly agree). Internal consistency was satisfactory with
Cronbach’s α = .84.
For the assessment of perceived self-efficacy, four different items were used
(Cronbach’s α = .74). The general stem for all items was, “How certain are you
about being able to overcome the following barriers? I can manage to stick to
healthy food . … ” Afterwards, specific barriers were presented: (a) “even if some-
thing delicious but unhealthy is served,” (b) “even if this will be more expensive,”
(c) “even if I am short of time to do my shopping and preparation,” and (d) “even if
this means I cannot eat everything I crave for.” Responses were made on 4-point
scales ranging from 1 (strongly disagree) to 4 (strongly agree).
The intention to adopt healthy nutrition behaviors was measured with four
items, namely (a) “I intend to eat only a very low amount of fat (such as animal fat,
cheese, butter) over the next months,” (b) “I intend to live a healthier life,” (c) “I
intend to eat healthy foods over the next months,” and (d) “I intend to invest more
into my health.” Responses were made on 7-point scales ranging from 1 (“I don’t
have this intention at all”) to 7 (“I do have this intention”). Cronbach’s alpha was
considerably high, with α = .89.
Nutrition behavior was assessed with four items related to a low-fat diet: (a) “I
follow a low-fat diet,” (b) “When I eat milk products or drink milk, I choose
low-fat products (such as low-fat milk or yogurt),” (c) “When I eat cheese, I
choose high-fat products (such as Gouda or Emmentaler),” and (d) “I avoid foods
with cholesterol.” Responses were made on 4-point scales ranging from 1
(strongly disagree) to 4 (strongly agree). The negatively poled item was recoded
before an overall score was created (Cronbach’s α = .74).

148
RENNER, KNOLL, SCHWARZER
Weighing, height measures, and Body Mass Index (BMI). Weighing
and height measures were standardized (calibrated scales, regular clothing). The
BMI was calculated as the weight in kilograms divided by the square of the height
in meters. According to international standards (Bailey & Ferro-Luzzi, 1995; Bray,
1
978; WHO Study Group, 1990), participants with a BMI over 24 were classified
as being overweight, whereas those with a BMI below this cutoff point were classi-
fied as being of average weight.
Sample
A total of 1,583 inhabitants of Berlin between 14 and 87 years of age volunteered to
participate in the study. Seventy-one percent of the participants (n = 1,019) com-
pleted both the first and the second questionnaire.
RESULTS
Details of the Resulting Sample
Participants recruited through advertisements were slightly heavier than those who
were invited through a personal letter, with a mean BMI of 25.1 (SD = 3.7) versus
2
4.3 (SD = 3.6); t(1, 1581) = 3.11, p = .002. Furthermore, gender differences be-
2
tween the two groups were significant, χ (1) = 43.7, p < .001. Women participants
were mostly recruited through advertisements, whereas men participants were pre-
dominantly motivated by the personal letter. Age differences among the two
groups, however, were not significant, t(1, 1581) = 1.52, p = .129.
Dropout analysis showed that those who completed both questionnaires were on
average 4 years older than those who did not; 40 years (SD = 15.3) versus 36 years
(SD=14.0), t(1, 581)=5.38, p<.001. Furthermore, theyexhibitedaslightlybutsig-
nificantlyhigherBMIthanthedropoutgroup,24.6(SD=3.6)versus23.9(SD=3.4);
t(1, 581) = 3.74, p < .001. There was no significant gender difference between the
2
two samples, χ (1) = .33, p = .56. All subsequent analyses are based on the reduced
sample of n = 1,019 participants who had completed both questionnaires.
2
Of this final sample, 53% were women and 47% were men, χ (1) = 4.40, p =
.
035. As Table 1 shows, 53% were younger than 41 years of age, and 61% had a
BMI lower than or equal to 24.
Self-Reported Health
As expected, the percentage of participants reporting excellent or very good health
status declined steadily with increasing age (Figure 1). About 65% of the younger
participants (younger than 31 years of age) reported their overall health as being

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
149
TABLE 1
Absolute and Relative Frequencies for “Age Group” and “Body Weight”
(Longitudinal Sample)
Younger
Than 31
Older
Than 60
31–40
40–50
51–60
Total
Average weight
BMI < 25 (%)
Overweight
BMI ≥ 25 (%)
Total
328
32
43
4
111
11
62
66
7
96
75
7
112
11
187
(18%)
36
4
90
616
61
403
39
1,019
(100%)
6
9
9
371
173
(17%)
162
(16%)
126
(12%)
(37%)
Note. Twelve individuals were older than 74 years; BMI = Body Mass Index.
FIGURE 1 Very good or excellent self-reported health by age group.
very good or excellent. In contrast, only 24% of the age group 61 or older stated that
2
they were in very good health, χ (1) = 62.22, p < .001. Women and men were as
likely as to claim that their health was very good or excellent, 49% versus 52%;
2
χ (1) = 0.16, p = .690. In addition, participants seemed to take their risk status into
account. Those with average weight were more likely to report very good or excel-
lent health than those who were overweight. Only 38% of the overweight respon-
dents said that their health was excellent or very good, whereas 58% of the partici-

150
RENNER, KNOLL, SCHWARZER
2
pants with average weight viewed their health as excellent or very good, χ (1) =
7.30, p < .001. In conclusion, participants obviously considered objective factors
3
such as age and body weight when judging their overall health status. These results
are remarkably consistent with the existing survey literature on self-reported health
status (e.g., Siegel, 1994) and corroborate reports of similar associations between
self-reported health status, gender, age, and BMI.
According to these results, aging and being overweight are associated with a
decrease in subjective health. Does this mean that aging and being overweight are
also associated with an increase in perceived risk, followed by a decline in defen-
sive optimism? The ensuing question is whether perceived vulnerability to cardio-
vascular diseases is also associated with age and risk status. A 2 × 5 × 2 × 2 mixed
model multivariate analysis of variance was conducted, with the two risk judg-
ments (self vs. other) as levels of the within-subjects factor “Target.” The five age
groups (younger than 31, 31–40, 41–50, 51–60, older than 60), gender (men vs.
women), and weight group (average weight vs. overweight) constituted the levels
1
of the three between-subjects factors. The analyses yielded a significant main ef-
fect for the within-factor “Target,” F(1, 999) = 57.75, p < .001. Figure 2 displays
the mean absolute risk judgments for self and others separately for each age group.
Subsequent analysis showed that self-ratings increased with age, F(4, 999) =
1
3.34, p < .001, as well as absolute risk ratings for an average peer of the same age
and sex, F(4, 999) = 4.71, p = .001.
Hence, personal risk perceptions reveal relative accuracy, because older partic-
ipants gave higher risk ratings than younger ones. However, at the same time, evi-
dence for defensive optimism emerged. Participants expected comparatively more
negative outcomes to happen to other people than to themselves. Further analysis
showed a significant discrepancy between self-ratings and ratings for an average
member of the same age group and gender within each age group, all F(1, 999) val-
ues greater than 9, with p < .001. The observed effect of aging on perceived risk
leads to the conclusion that people acknowledged a higher risk with increasing age
and declining health. Nevertheless, they still assumed that they were less vulnera-
ble than their peers. In other words, aging did not curb unrealistic optimistic risk
perceptions.
1All the figures and reported results are based on full factorial models that meet standard require-
ments of analysis of variance (ANOVA). The effects were estimated by the regression method for parti-
tion sums of squares. That means that each effect is adjusted for all the other effects in the model. For an-
alyzing risk perceptions, a mixed between-/within-subjects ANOVA had been used. Therefore, the total
sum of squares (SS) is divided into a source attributable to the three between-subjects factors of the de-
sign (body weight, age, and sex), and a source attributable to the within-subjects factor named “Target”
(risk perception for oneself and for an average peer). In addition, factor level differences were examined
by constructing simple main effects that mean to compute the effect of one factor within the level of an-
other factor (e.g., the effect of age within overweight participants).

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
151
FIGURE 2 Mean absolute risk judgments for self and others to cardiovascular diseases as a
function of age.
Moreover, participants of average weight considered themselves at lower risk
than overweight participants, –.67 versus –.12; F(1, 999) = 10.30, p = .001. With
respect to the assumed risk faced by an average peer, –.05 versus .16; F(1, 999) =
0
.11, p = .74, both weight groups displayed similar ratings. Comparing the mean
judgment for the self and the average peer revealed a significant unrealistic opti-
mism for the average weight group, F(1, 999) = 139.13, p < .001, and for the over-
weight group, F(1, 999) = 18.54, p < .001.
Finally, compared to men, women showed a lower risk perception for them-
selves, –.67 versus –.20; F(1, 999) = 12.98, p < .001. However, they did not differ
in the perception of an average peer, –.06 versus .08; F(1, 999) = 0.06, p = .81.
Nevertheless, men, F(1, 999) = 29.35, p < .001, as well as women, F(1, 999) =
1
24.91, p < .001, displayed unrealistic optimism.
Theseresultpatternsfurthersupporttheassumptionthatpeopleconsidertheirac-
tual risk status when judging their vulnerability: Older and overweight participants
and men admitted that they were at greater risk, but even so they still considered
themselvesaslessvulnerablethananaveragepeer. Hence, peopletaketheirageand
riskstatusintoaccountwhengaugingtheirrisk, butthisrealismisstillaccompanied
bydefensiveoptimism:Onefeelslessvulnerablethanothers. Thisleadstotheques-
tionofwhetherthereisasimilarrelationbetweenfunctionaloptimism(positiveout-
come expectancies and perceived self-efficacy), aging, and being overweight.

152
RENNER, KNOLL, SCHWARZER
Positive Outcome Expectancies
The 2 × 2 × 5 ANOVA yielded a significant main effect for age, F(4, 999) = 4.20, p
=
.002; weight, F(1, 999) = 30.12, p < .001; and gender, F(1, 999) = 15.82, p < .001.
Furthermore, the two main effects of age and body weight were qualified by a sig-
nificant interaction, F(4, 999) = 2.53, p = .04. Overall, the effects accounted jointly
for 12% of the variation in outcome expectancies.
Compared to men, women displayed more positive outcome expectancies, fol-
lowed by adopting a healthy diet (23.3 vs. 22.0). Moreover, as Figure 3 displays,
overweight participants harbored more positive outcome expectancies in compari-
son to average weight participants within each age group. Even at young age, over-
weight participants anticipated clear positive outcomes for themselves. Hence,
within the overweight group, aging was not associated with a gain of expected
benefit from a healthy diet, F(4, 999) = 0.59, p = .67. In contrast, the analysis
within the average weight group revealed a significant positive trend for outcome
expectancies with increasing age, F(4, 999) = 6.01, p < .001. Additionally, Scheffé
post hoc contrasts within the average weight group showed that individuals aged
5
0 years or younger exhibited significantly less positive outcome expectancies
than older ones.
FIGURE 3 Mean positive outcome expectancies as a function of age and weight.

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
153
Perceived Self-Efficacy
The 2 × 2 × 5 ANOVA yielded a significant main effect for age, F(4, 999) = 21.78,
p < .001, and a significant interaction between age and weight group, F(4, 999) =
3
.92, p = .004. No other effect reached statistical significance. Overall, 11% of the
self-efficacy variance was accounted for by these two effects.
In accordance with the previous results, older participants perceived higher
self-efficacy than younger ones. This holds for average weight, F(4, 999) = 15.40,
p < .001, as well as overweight participants, F(4, 999) = 8.25, p < .001. As Figure 4
displays, within both groups there was a visible positive linear trend for perceived
self-efficacy with increasing age. However, within the overweight group the effect
is more pronounced. Scheffé post hoc contrast analysis within overweight partici-
pants showed that (with only a few exceptions) an increase of self-efficacy from
the youngest to the oldest group took place among the average participants. How-
ever, Scheffé post hoc contrast analysis within average weight participants con-
firmed this trend only partially. Up to the age of 50 there was no significant
increase in self-efficacy (all Scheffé post hoc contrasts with p > .05). But beyond
this age, self-efficacy increased with age.
FIGURE 4 Mean perceived self-efficacy as a function of age and weight.

154
RENNER, KNOLL, SCHWARZER
Intention to Adopt Low-Fat Nutrition
There was a significant main effect for age, F(4, 999) = 10.55, p < .001; body
weight, F(1, 999) = 9.44, p = .002; and gender, F(1, 999) = 10.36, p = .001. In addi-
tion, an interaction between age and body weight, F(4, 999) = 3.60, p < .01, reached
significance. Eleven percent of the variation of the intention to adopt low-fat nutri-
tion was accounted for by these effects.
Women reported a higher intention to adopt low-fat nutrition compared to men
(20.4 vs. 19.4). Older as well as overweight participants reported higher intentions
to alter their eating habits according to a low-fat diet than younger ones or partici-
pants with average weight.
Closer inspection of the interaction effect between age and weight revealed a
strongerlineareffectofagewithinparticipantsofaverageweight,F(4,999)=9.09,p
<
.01, as compared to overweight participants, F(4, 999) = 3.28, p = .01. Figure 5
shows that among average weight participants, intentions for preventive nutrition
behaviors increased with age. However, this positive trend was only partly sup-
ported by Scheffé post hoc contrasts. Participants younger than 50 years reported
lowerintentionsthanolderones. Youngeroverweightparticipantsshowedasimilar
positive trend, but the differences were smaller. Therefore, Scheffé post hoc con-
trasts only upheld the notion that, when compared to the older-than-60 group, the
youngest group had lower intentions to adopt a low-fat nutrition.
FIGURE 5 Mean intention to adopt a low-fat diet as a function of age and weight group.

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
155
Preventive Nutrition
As before, the 2 × 2 × 5 ANOVA yielded a significant main effect for age, F(4, 999)
30.70, p < .001; body weight, F(1, 999) = 4.65, p = .03; and gender, F(1, 999) =
=
1
9
7.50, p < .001. Furthermore, an interaction between age and body weight, F(4,
99) = 6.75, p < .001, emerged. These four effects accounted together for 21% of
the preventive nutrition variance.
Compared to men, women reported healthier nutrition behaviors (10.4 vs.
1
1.3). Overweight participants reported a healthier diet than average weighted
participants, and aging was associated with healthier nutrition habits. Figure 6
shows the mean consumption of healthy foods as a function of age and body
weight.
The positive effect of aging on nutrition could be observed within the average
weight group, F(4, 999) = 29.66, p < .001, as well as within overweight partici-
pants, F(4, 999) = 10.73, p < .001. Scheffé post hoc contrasts within average
weight participants supported the finding that aging secured preventive nutrition.
For overweight individuals, the effect was somewhat weaker. Therefore, Scheffé
post hoc contrasts suggest that mainly the youngest and the oldest age groups dif-
fered significantly from the remaining groups.
FIGURE 6 Mean consumption of low-fat food as a function of age and weight group.

156
RENNER, KNOLL, SCHWARZER
DISCUSSION
The results have documented important associations between two objective
health-related factors and six kinds of health beliefs that deal with preventive nutri-
tion. People take their age and body weight into account when they judge their
health status. On average, the older and the heavier they are, the less frequently they
regard their health as “very good or excellent” (see Figure 1). This is realistic and
not surprising. It also demonstrates that individuals do not merely make judgments
in terms of social comparisons within a narrow reference group of similar peers, but
that they use overarching criteria, for instance, objective health status or compari-
sons across the life-span.
The same realism is expressed when people gauge their vulnerability toward
cardiovascular health threats. The older they become, the higher they calibrate
their risk (see Figure 2).
However, the most interesting aspect is the phenomenon of how they deal with
their reference group. Within each age group, they discount their risk compared to
that of others. This demonstrates the existence of an optimistic bias throughout life,
although participants were not indiscriminately optimistic, and tended to adhere
partly to the constraints of reality. The results suggest that the strong defensive opti-
mismeffectwasdeterminedbythetendencytoseeothersasmoreatriskfornegative
events than oneself. The fact that individuals harbor pessimistic biases for others
mayrepresentamechanismbywhichtheymaintainacomparativelyoptimisticout-
look for themselves, despite realizing that health-related risks do increase with age.
This might satisfy a need for accuracy by acknowledging more objective risk at an
absolute level, but serve self-protective needs by maintaining a pessimistic view of
others at the same time (Armor & Taylor, 1998; Whitley & Hern, 1991). The social
network, be it real or imagined, can be seen as a convoy that moves through the life
course and serves as a backdrop for self-serving social comparisons.
Relative invulnerability may not only be achieved by adapting average risk to
one’s own risk status. Overweight participants showed higher risk perceptions for
themselves than average weight participants, but they did not differ with respect to
their perceived average risk. Because both groups displayed unrealistic optimism,
defensive optimism appears to be a matter of degree. Similar findings were ob-
tained in a study among college students (Rothman, Klein, & Weinstein, 1996).
People take their objective disadvantages into account when making judgments,
but they discount them to some extent. This leads to a practical conclusion: Risk
communication that provides only information about the individual’s risk may
have less impact in comparison to a risk communication that provides additional
information about the risk faced by an average peer. People may need both kinds
of information to locate their risk status more accurately.
An important question is whether defensive optimism poses a barrier against
the adoption of indispensable health behaviors. Why should someone adopt a diffi-

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
157
cult behavior, such as restrained eating, when no risk is perceived in comparison to
others? Again, this is a matter of degree. In this sample, increasing age and body
weight are indeed related to risk perception, although they are slightly discounted.
The other health beliefs that were assessed in this study document an inclination
toward preventive nutrition. This is, for example, expressed in increased consump-
tion of low-fat food with advancing age in both weight groups (see Figure 6),
which is in line with three related health beliefs, namely (a) nutrition outcome ex-
pectancies (Figure 3), (b) nutrition self-efficacy (Figure 4), and (c) the intention to
consume low-fat foods (Figure 5). This is cumulative evidence for risk-related rea-
sonable health behavior. Delving further into these data, it was discovered that age
and functional optimism were more closely interrelated within average weight
participants. With advancing age, they acknowledged more positive consequences
and greater confidence to carry out the intended behavior, expressed higher inten-
tions to adopt a low-fat nutrition, and displayed healthier nutrition habits. In over-
weight participants, there was a less positive trend associated with age. To some
extent this could be due to a ceiling effect because even young overweight partici-
pants exhibited a high level of positive outcome expectancies and intention to stick
to a healthy nutrition. This may reflect greater self-relevance and vulnerability
within overweight young people. In conclusion, these differential effects indicate
that the participants were at different points in the behavior change process de-
pending on their objective risk status. Hence, different interventions and informa-
tion may be needed to move people closer to action. Older and overweight people
may benefit less from risk communication than from resource communication,
emphasizing the effectiveness of precautions, and their coping abilities.
On the other hand, health risk perception is nothing but a starting point for
health behavior change. It sets the stage for subsequent cognitions such as out-
come expectancies and self-efficacy, while individuals progress through a motiva-
tional phase that concludes with an explicit behavioral intention (e.g., the intention
to consume low-fat, high-fiber foods). In the subsequent postintentional phase,
they plan and execute the critical behavior with different degrees of success. This
view has been put forward in the Health Action Process Approach (Schwarzer,
1
992, 1999) and other process theories of health behavior change.
These findings shed some light on the changes in optimistic health beliefs
brought on by the aging process, but several caveats must be considered in inter-
preting the findings. First, cross-sectional data do not allow the examination of an
aging effect as opposed to a cohort effect. It may be that the current older genera-
tion of adults harbors greater optimism than future generations. Second, the results
may stem from a self-selection bias. It is possible that older men and women with a
more pessimistic view of their health-related capabilities did not volunteer to par-
ticipate in the study. This would have led to a sampling bias. Another possibility is
that people with a more pessimistic self-view may be at greater risk for serious
health problems, making morbidity and mortality more likely. This would create a

158
RENNER, KNOLL, SCHWARZER
group of survivors within the older age strata with higher levels of optimism than
those in the younger strata. If so, levels of functional optimism in samples of older
people may be higher simply through attrition of pessimistic people. To determine
whether the present findings reflect a genuine aging phenomenon, a sample would
have to be followed longitudinally from youth to old age. Only such a prospective
study across the life-span would allow a thorough examination of the underlying
processes of health belief changes.
REFERENCES
Armor, D. A., & Taylor, S. E. (1998). Situated optimism: Specific expectancies and self-regulation. In
M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 309–379). New York:
Academic.
Arnett, J. (1995). The young and the reckless: Adolescent reckless behavior. Current Directions in Psy-
chological Science,, 4, 67–71.
Assmann, G. (1993). Lipid metabolism disorders and coronary heart disease. Primary prevention, di-
agnosis and therapy guidelines for general practice (2nd ed.). München, Germany: MMV Medizin
Verlag.
Bailey, K. V., & Ferro-Luzzi, A. (1995). Use of body mass index of adults in assessing individual and
community nutritional status. WHO Bulletin, 73, 673–680.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
Bray, G. (1978). Definition, measurement, and classification of the syndromes of obesity. International
Journal of Obesity, 2, 99–112.
Elkind, D. (1974). Children and adolescents: Interpretative essays on Jean Piaget. New York: Oxford
University Press.
Filipp, S.-H. (1996). Motivation and emotion. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the
psychology of aging (4th ed., pp. 218–235). San Diego, CA: Academic.
Hahn, A., & Renner, B. (1998). Perception of health risks: How smoker status affects defensive opti-
mism. Anxiety, Stress, and Coping, 11, 93–112.
Mroczek, D. K., & Kolarz, C. M. (1998). The effect of age on positive and negative affect: A develop-
mental perspective on happiness. Journal of Personality and Social Psychology, 75, 1333–1349.
Ogden, J. (1996). Health psychology. Buckingham, England: Open University Press.
Perloff, L. S., & Fetzer, B. K. (1986). Self–other judgments and perceived vulnerability to victimization.
Journal of Personality and Social Psychology, 50, 502–510.
Rothman, A. J., Klein, W. M., & Weinstein, N. D. (1996). Absolute and relative biases in estimations of
personal risk. Journal of Applied and Social Psychology, 26, 1213–1236.
Schechter, S. (1993). Investigation into the cognitive processes of answering self-assessed health status
questions (Cognitive Methods Staff, Working Paper Series, No. 2, Office of Research and Method-
ology). Hyattsville, MD: National Center for Health Statistics.
Schwarzer, R. (1992). Self-efficacy in the adoption and maintenance of health behaviors: Theoretical
approaches and a new model. In R. Schwarzer (Ed.), Self-efficacy: Thought control of action (pp.
2
17–242). Washington, DC: Hemisphere.
Schwarzer, R. (1994). Optimism, vulnerability, and self-beliefs as health-related cognitions: A system-
atic overview. Psychology and Health, 9, 161–180.
Schwarzer, R. (1999). Self-regulatory processes in the adoption and maintenance of health behaviors.
The role of optimism, goals, and threats. Journal of Health Psychology, 4, 115–127.

OPTIMISTIC HEALTH BELIEFS AND NUTRITION
159
Schwarzer, R., & Fuchs, R. (1996). Self-efficacy and health behaviors. In M. Conner & P. Norman
Eds.), Predicting health behaviour: Research and practice with social cognition models (pp.
63–196). Buckingham, England: Open University Press.
(
1
Schwarzer, R., & Renner, B. (in press). Perceived self-efficacy governing preventive nutrition: A longi-
tudinal study of a community sample. Health Psychology.
Siegel, P. Z. (1994). Self-reported health status: Public health surveillance and small-area analysis. In S.
Schechter (Ed.), Proceedings of the 1993 NCHS Conference on the cognitive aspects of self-re-
ported health status (pp. 11–27; Cognitive Methods Staff, Working Paper Series, No. 10. Office of
Research and Methodology). Hyattsville, MD: National Center for Health Statistics.
Staats, S., Heaphey, K., Miller, D., Partlo, C. I., Romine, N., & Stubbs, K. (1993). Subjective age and
health perceptions of older persons: Maintaining the youthful bias in sickness and in health. Interna-
tional Journal of Aging and Human Development, 37, 191–203.
Steinhagen-Thiessen, E., & Borchelt, M. (1999). Morbidity, medication, and functional limitations in
very old age. In P. B. Baltes & K. U. Mayer (Eds.), The Berlin aging study: Aging from 70 to 100 (pp.
1
31–166). New York: Cambridge University Press.
Taylor, S. E. (1989). Positive illusions: Creative self-deception and the healthy mind. New York: Basic
Books.
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective on
mental health. Psychological Bulletin, 103, 193–210.
Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of Personality and So-
cial Psychology, 39, 806–820.
Weinstein, N. D. (1982). Unrealistic optimism about susceptibility to health problems. Journal of Be-
havioral Medicine, 5, 441–460.
Weinstein, N. D. (1987). Unrealistic optimism about susceptibility to health problems: Conclusions
from a community-wide sample. Journal of Behavioral Medicine, 10, 481–500.
Weinstein, N. D. (1996). References on perceived invulnerability and optimistic biases about risk or fu-
ture life events. Unpublished manuscript.
Whitley, B. E., & Hern, A. L. (1991). Perceptions of vulnerability to pregnancy and the use of effective
contraception. Personality and Social Psychology Bulletin, 17, 104–110.
WHO Study Group. (1990). Diet, nutrition, and the prevention of chronic diseases (World Health Orga-
nization Tech. Rep. Series 797). Geneva, Switzerland: World Health Organization.