Post on 05-Jun-2018
Childrens profile of healthy and unhealthy behaviors: Demographic
characteristics and perceptions of social environment
Yannis Theodorakis a,*, Athanasios Papaioannou b, Antonis Hatzigeorgiadis a,
& Eva Papadimitriou a
a University of Thessaly, Trikala, Greece
b Democtitus University of Thrace, Komotini, Greece
Manuscript submitted: December 5, 2002 Running head: childrens healthy and unhealthy profiles
Address for correspondence: Yannis Theodorakis University of Thessaly,
Department of Physical Education & Sport Science, 42100, Karies, Trikala, Greece.
e-mail: Phone:
ytheo@pe.uth.gr + 30 24310 47001
Childrens profile of healthy and unhealthy behaviors: Demographic
characteristics and perceptions of social environment
Manuscript submitted: December 5, 2002 Running head: CHILDRENS HEALTHY AND UNHEALTHY PROFILES
Childrens profile of healthy and unhealthy behaviors: Demographic
characteristics and perceptions of social environment.
Abstract
Objectives. In recent years health promotion is a matter of great interest
among researchers in the field of social sciences. The aim of the present study was to
identify the profile of children who exhibit healthy and unhealthy behaviors in
relation to demographic characteristics and perceptions of social environment.
Method. Participants were 3640 Greek students from 10 to 16 years of age.
They responded on self-report questionnaires assessing a number of behaviors
(exercise, diet, smoking, and violence), as well as family structure, family income,
perceived family and peer behavior and perceived family support.
Results. Cluster analysis identified four distinct profiles. One including
children adopting healthy behaviors (exercising, healthy eating) and avoiding
unhealthy ones (smoking, violence), a second including children avoiding unhealthy
behaviors, but not adopting healthy ones, a third including children adopting healthy
behaviors, but also taking part in violent incidents, and a fourth including children
adopting smoking and avoiding healthy behaviors. Demographic characteristics
seemed to better explain the adoption or not of healthy behaviors, whereas
perceptions of social environment and age seemed to better explain the adoption or
not of unhealthy behaviors.
Conclusion. The results of the present study indicate that health promotion
programs should take into serious consideration both personal and social
characteristics of the targeted population.
Key words: exercise, nutrition, smoking, violence, children profiles
childrens healthy and unhealthy profiles 1
Childrens profile of healthy and unhealthy behaviors: Demographic
characteristics and perception of social environment.
Introduction
In recent years the interest of health-related scientific and medical organizations
on health promotion is progressively growing. However, despite the increasing focus
on health prevention, results from relevant studies show that behavior patterns are
quite worrying, especially among younger populations (Mota & Queiros, 1996; Pate
et al., 1997; Steptoe et al., 2002). Since the early nineties, the adoption of unhealthy
behaviors like smoking, use of alcohol and drugs, seems to spread rapidly (Torabi &
Nakornhet, 1996). Furthermore, low levels of exercise and poor diet have been
identified that are related to obesity and have been characterized as potentially risk-
factors for individuals health (Muecke, Simons-Morton, Huang & Parcel, 1992).
Thus, the study of health-related behaviors becomes of great importance especially for
younger children and adolescents, since it is at that age when health beliefs are
established (Baranowski, 1997) and healthy/unhealthy habits are adopted (Taylor,
1999). The purpose of the present study was to identify whether patterns of healthy
and/or unhealthy behaviors exist among young population and to examine likely
personal and social factors that may be important in determining the adoption of
healthy and unhealthy habits.
Considerable amount of research during the past and present years has been
dealing with the healthy and unhealthy behaviors of children and the factors that
influence and finally shape them. Researchers claim that healthy behaviors such as
exercise and healthy eating habits are positively correlated (Liang, Shediac -
Ritzkallan, Celantano & Rohde, 1999), and so are unhealthy behaviors, such as the
childrens healthy and unhealthy profiles 2
use of nicotine, drugs and violent behavior (Barnes, Welte, Hoffman & Dintcheff,
1999; Bauman & Phongsavan, 1999; Liang et al., 1999; Friedman, 1998; Griesler &
Kandel, 1998). Of the amount of healthy and unhealthy behaviors the present study
will focus on exercise, healthy eating, smoking, and violence.
There is a number of models which have been developed over the years to
explain health-related behaviors and which recognize the importance of personal and
social factors. According to Rosenstock (1991) the health belief model (Becker, 1974)
has been the most influential and widely tested approach to health behaviors.
According to the health belief model, among the forces that are recognized as
important in shaping behavior are demographic characteristics such as age, gender
education and ethnicity, and socio-psychological factors such as personality, social
environment and social class. In a similar fashion Banduras (1986) social cognitive
theory suggests that behavior is a product of relationships between personal factors
(coming within the individual) and environmental factors (coming from the context in
which the behavior occurs). Finally, Sallis and Hovells (1990) social learning model
stresses that with regard to younger populations, personal factors (e.g. age, gender,
personality) and social influences especially (e.g. family modeling, peer influences,
social support) are significant determinants of health-related behaviors, in particularly
physical activity. Therefore, it becomes evident that the important role of personal and
social factors in shaping health behavior is widely recognized.
In the literature, it has been reported that there are numerous factors influencing
the adoption of behaviors like exercising or smoking, however family and friends are
two of the most important factors proposed to account for such behaviors (e.g.
Friedman & Glassman, 2000; Herrenkohl et al., 2000; Blackson, et al., 1999; Brook,
Mendelberg, Galili, Priel, & Bujanover, 1999; Duncan, Duncan, Biglan, & Ary,
childrens healthy and unhealthy profiles 3
1998). Broadly speaking children 's physical activity and their opinion about it as well
has been found to be in accordance to their parents' (Babkes & Weiss, 1999; Trost,
Pate, Ward, Saunders & Riner, 1999; Kimiecic, Horn & Shurin, 1996; Stucky-Ropp
& DiLorenzo, 1993; Anderssen & Wold, 1992). Parents are said to influence their
children either as models or motivationally (Stucky-Ropp & DiLorenzo, 1993;
Anderssen & Wold, 1992). Vilhjalmlsson and Thorlindsson (1998) also claimed that
father's, friends' and older siblings' involvement in exercise is significantly correlated
to children's behavior. Meta-analysis about the effects of social influences on
individuals attitudes, intention and behavior, revealed that the influence of important
others is stronger than that of family (Carron, Hausenblas & Mack, 1996). However,
relevant study that dealt with parental influence concluded that children who grow up
with one or no parents adopt unhealthier lifestyle behaviors than children who grow
up with both parents (Theodorakis, Papaioannou & Karastogianidou, submitted).
Thus, regarding exercise, the influence of both family and peers is clearly evident.
As far as healthy eating is concerned parental modeling is considered the major
factor in shaping children 's behavior (Lau, Quadrel & Hartman, 1990). Parental
influence is evident in the work of Hooper, Gruber, Munoz and MacConnie (1996)
who studied family and school influence in the modulation of healthy nutritional
habits and concluded that parental and school cooperation was the most effective. It
has been also noticed that financial and social status of the family influence children 's
nutritional habits (Neumark - Sztainer, Story, Perry & Casey, 1999; Hupkens, Knibbe,
Otterloo & Drop, 1998). Children who lived with one of their parents, single and
unemployed people were characterized as at-risk groups (Roux, Le Couedic, Durand-
Gasselin & Luquet, 1999). Finally, nutritional habits are said to be affected by
appearance, weight and peer influence (McLellan, Rissel, Donnelly & Bauman, 1999;
childrens healthy and unhealthy profiles 4
Shisslak, et al., 1998), whereas family and peer influence as models has been also
found to be related to bulimic symptoms of teenagers (Stice, 1998).
Regarding smoking, it has been reported that peer influence is the most
influential factor, and that the closer the friends are the more likely it is for children to
adopt similar beliefs and behaviors (Roosmalen & McDaniel, 1989). Family influence
is also recognized as important (Zhu, Liu, Shelton, Liu & Giovino, 1996; Dusenbury,
et al., 1992), with mothers having the most significant influence (Griesler & Kandel,
1998). In a study by Brook, Mendelberg, Galili, Priel and Bujanover (1999), young
children whose parents were smokers were found to be more tolerant towards
smoking compared to the ones whose parents did not smoke, although they knew its
consequences. Children whose parents and siblings were smokers were also found to
start smoking at an earlier age (Unger & Chen, 1999; Roosmalen & McDaniel, 1989).
Peer influence though is claimed to be stronger than parental (Friedman & Glassman,
2000; Dusenbury et al., 1992) and so is sibling influence (Sugathan, Moody, Bustan
& Elgerges, 1998).
Furthermore, it has been supported that parental monitoring (Barnes et al, 1999;
Duncan et al., 1998) and frequent conflicts between parents and children (Duncan et
al., 1998) are related to increased use of nicotine by children, whereas parental
supportive behavior leads to its decrease (Griesler & Kandel, 1998). According to
Sobeck, Abbey, Agius, Clinton and Harrison (2000) it is more possible for children
who smoke to come from families that face problems and know less about smoking
compared to non-smokers of the same age. Finally, smoking has been found related to
low socioeconomic family condition, poor academic performance and ignorance (Zhu
et al., 1996).
childrens healthy and unhealthy profiles 5
The last behavior that this study deals with is violent behavior. Paetsch and
Bertrand (1997) reported that the most important factor predicting violent behavior
was peer deviant behavior. Sport activities and in a large degree involvement in
amusement activities were positively correlated to such behaviors, whereas
Herrenkohl, et al. (2000) reported violence to be related to poor academic
performance. Blackson, et al. (1999) also supported that peers and family condition
instill deviant behaviors when the environment is tolerable towards such behaviors.
Other factors that violent behavior is likely to depend on during adolescence are
modeling, family malfunctioning, and family separation (Dahlberg, 1998). Family and
peer influences were also supported by Ary, Duncan, Duncan and Hops (1999) who
claimed that family conflicts, congruous family relationship and inadequate children
monitoring by parents play a major role in the adoption of violent behavior.
Summarizing the above-mentioned, it is well documented that family and peer
influence and support are important determinants of children's behavior. The aim of
the study was first to examine how the behaviors of interest, i.e. exercising, smoking,
healthy eating and participating in violent acts, cluster, that is how certain behaviors
relate to each other, and second to identify the profile of children in these clusters of
behaviors, in relation to demographic characteristics and social influences. Given the
exploratory character of the study, no specific hypotheses were formed regarding how
behaviors would cluster. Nevertheless, we expected that age, gender and social
influences (family support and perceived family and peer behavior) would be the most
crucial factors in determining the adoption of healthy and unhealthy behaviors.
childrens healthy and unhealthy profiles 6
Method
Participants and procedures
Participants in this study were 3640 Greek students. Their age ranged from 10
to 16 years. The sample was selected with a random stratified sampling method from
85 classes of 30 elementary schools, grade 6, from 87 classes of 33 junior high
schools, grades 2 and 3, and from 86 classes of 28 high schools, grades 5 and 6. These
schools were elected from 6 urban areas of Greece varying in population from four
millions to seventy-five thousands residents.
The study was conducted with the permission of the ministry of education. Ten
trained research assistants were employed in the data collection process. Participants
were informed that questionnaires were anonymous and signed consent forms. In their
class environment, they responded on questionnaires assessing the examined
behaviors, family and peer influence, perceived family support, family structure and
family income. After completing the forms children were asked to place it in a poll.
Measures
Self-report past behaviors. The examined behaviors were four and were
assessed by self-reported measures. The scales were adopted from Kimiecik (1992)
and Steptoe, et al. (2002) and complied with Ajzens (2002) guidelines for
measurement of behavior. In particular, students were asked to indicate frequency of
the examined behaviors on six-point scales. For exercising, how many times you
exercised intensively during the last month (none to more than twenty times).
Participants were instructed that intensive exercise meant taking part in physical
activities, which cause increased heart rate and sweating for more than 30 minutes
(e.g. football, basketball, aerobic). For smoking and eating fruits, how many
cigarettes/fruits you smoked/ate during the last week (none to more than twenty). For
childrens healthy and unhealthy profiles 7
participating in violent acts, how many times you got involved in violent acts during
the last month (none to more than ten times).
Perceived family and peer behavior. Family and peer influence was estimated
in terms of modeling. The questionnaire used was based on prior work by Wang,
Fitzhugh, Westerfield and Eddy (1995). Students answered 16 items regarding
perceived parents' and peer's behavior towards the examined behaviors. Four items for
each behavior were used. More specifically students were asked what they believed
about their parents, siblings and best friends: How often do you think your mother/
father/ siblings/best friend exercised/smoked/ate fruits/ participated in violent action
during the previous month. Responses on these items were rated on a 7-point scale
(never to all the time).
Perceived family support. Perceived parental support was estimated by a
questionnaire based on prior work by Wickrama, Lorenz and Conger (1997). It
consisted of 10 items assessing students' perception of their parents' behavior towards
them. For example, students were asked how often during the previous month their
parents illustrated their real interest for them, expressed their love and affection to
them, were angry with them and so on. Reponses on these items were rated on a 6-
point scale (never to all the time). Cronbachs alpha was .86.
Family structure. To assess family structure, participants were asked to
indicate whether they live with both parents, with their mother only, with their father
only, with their grandparents, or alone. They also had the choice to state other.
Perceived family income. Finally, regarding family income, participants were
asked to indicate on a five-point scale whether they perceived their family income to
be significantly above average, a little above average, on average, a little below
average or significantly below average. They could also state that they did not know.
childrens healthy and unhealthy profiles 8
Data analysis
Cluster analysis was chosen to answer the main research question, that is whether
identifiable subgroups or profiles of children would emerge based on variations
regarding the behaviors of exercising, smoking, eating fruits and participating in
violent acts. A nonhierarchical clustering method was employed (SPSS Quick
Cluster) with the squared Euclidean distance used as the similarity measure. Before
submitting the data to the cluster procedures, all variables were converted to z scores
in order to standardize the measurement scales and to allow the easier interpretation
of the results. A z score value of +/-.50 was used as a criterion for interpreting
whether individuals scored relatively higher or lower compared to their peers on each
of the four variables.
Results
Cases with missing values on the variables of main interest (behavioral
variables) were deleted. This resulted in a sample of 3307 children. Descriptive
statistics for all variables are presented in Table 1. Mean scores indicate that children
scored moderately on exercising and eating fruits and low in smoking and
participating in violent incidents. Furthermore, they scored moderately high on
perceived family support and perceived family and peer exercise behavior,
moderately low on perceived family and peer smoking behavior, moderately high on
perceived family and peer eating fruits behavior, and low on perceived family and
peer violent behavior.
Exercise was negatively, but lowly correlated with smoking (r= -.11), and
positively but again lowly correlated with eating fruits (.20) and participating in
violent acts (.11). Furthermore, smoking correlated moderately (r= .36) with
childrens healthy and unhealthy profiles 9
participating in violent acts. Family support correlated negatively with smoking (r= -
.32) and participating in violent acts (r= -.29). Finally, childrens behavior correlated
with perceived family and peer behavior for all corresponding behaviors (.23 < r <
.46). All correlations were significant at p< .001 level.
Cluster analysis
Results from the cluster analysis revealed four distinct children profiles. Three
and five-cluster analyses were also examined, however the four-cluster solution was
the most meaningful. Z scores, unstandardized means and standard deviations for
each of the key variables on which participants were classified into subgroups are
presented in Table 2.
Cluster one comprised 221 children. The unique characteristic of this cluster
was high scores on participating in violent acts. Children in this cluster scored high on
exercising and eating fruits, and low on smoking. Cluster 2 comprised 1272 children.
The main characteristic of this cluster was the low scores on exercising. Children in
this cluster also scored near-zero on smoking and participating in violent acts, and
moderately low on eating fruits. Cluster three comprised 382 children. The unique
characteristic of this cluster was the high scores on smoking. Children in this cluster
also scored low on exercising, and participating in violent acts, and moderately on
eating fruits. Finally, cluster four comprised 1432 children. The main characteristic of
this cluster was the high scores on exercising in combination with near-zero scores on
smoking and participating in violent acts. Children in this cluster also scored
moderately high on eating fruits.
In relative terms, children in cluster one were those participating in violent acts,
who however maintain a satisfactory level of exercise, comparable to that of children
in cluster four who were characterised by the adoption of healthy behaviors, and
childrens healthy and unhealthy profiles 10
absence of unhealthy ones. Children in cluster two, like those in cluster four were
characterised by absence of unhealthy behaviors, however their level of exercise was
very low, and they also had the lowest scores on eating fruits. Finally, children in
cluster three were those who were smokers, who also had comparatively moderate
levels of participating in violent acts and low levels of exercising.
Cluster profiles in relation to demographic characteristics
Statistics regarding demographic characteristics of participants falling into each
cluster are presented in Table 3. Regarding within gender differences, the first cluster
(high violence, high exercise) included higher percentage of boys than girls. In
particular, 11,6% of the boys and 2,3% of the girls fall in this cluster. The opposite
was evident for cluster two (low exercise, low smoking, low violence), which
included 29,6% of the boys and 46,2% of the girls. Gender representation in clusters
three and four was comparable.
Regarding family structure, because the vast majority of children were living
with both parents, children were regrouped to form two groups, one including
children living with both parents (n= 2835) and one including children living with one
or no parents (n= 409). Representation in clusters one and two was similar for
children living with both parents and children living with one or no parent. In cluster
three (high smoking) there was a higher percentage of children living with one or no
parent. In particular, cluster three included 22,7% of children living with one or no
parents and 9% of children living with both parents. Finally, cluster four (high
exercise, low smoking, low violence) included 44,8% of children living with both
parents and 33,5% of children living with one or no parent.
In relation to grade, in cluster one there was no specific pattern regarding
representation of different grades. In contrast, obvious patterns could be observed in
childrens healthy and unhealthy profiles 11
cluster two, three and four. In particular, in cluster two (low exercise, low smoking,
low violence) and cluster three (high smoking) percentage of membership increased
in direct relation with grade. In cluster four (high exercise, low smoking, low
violence) the pattern was opposite with membership decreasing as grade increased.
Finally, in relation to family income, observable differences were revealed in
cluster three (high smoking), where the percentage of children coming from families
with the lowest income was high (26.5%) compared to the total (12.4%), and cluster
four (high exercise, low smoking, low violence) where the percentage of children
coming from families with the lowest income was low (26.5%) compared to the total
(42.6%).
Cluster profiles in relation to perceptions of social environment
Analysis of variance was subsequently calculated in order to examine
differences in family support and perceived family and peer behavior among
participants falling into each cluster. The results of the analysis and mean scores for
these variables in each cluster are presented in Table 4. One-way ANOVA was
calculated to test for differences in family support. The analysis revealed significant
univariate effect (F3,2118= 88.16, p
childrens healthy and unhealthy profiles 12
contrast, differences between clusters two/four and clusters one/three were larger (ES
ranging from .45 to 1.01).
One-way MANOVA was calculated to examine differences between
participants in each cluster in perceived family and peer behavior. The analysis
revealed significant multivariate effect (F12,5604= 52.18, p
childrens healthy and unhealthy profiles 13
Discussion
The adoption of healthy and unhealthy behaviors is a matter of great
importance in the study of contemporary lifestyle. Children and adolescents are a
population of particular interest, since health beliefs adopted in early years are
indicative of later behavioral patterns. The present study explored patterns of behavior
among adolescent Greek population and examined personal and social factors as
likely determinants of such behavioral patterns.
Cluster profiles in relation to demographic characteristics
The first aim of this study was to identify profiles of the Greek student
population regarding healthy and unhealthy behaviors, according to demographic
characteristics. The cluster profile results indicate considerable variability among
children as far as the examined behaviors are concerned. These differences were
associated, at least partly, with age, gender, family structure and income. In relation to
gender, the results of the present study give some interesting information regarding
the adoption of healthy and unhealthy habits. Comparing cluster two to cluster four it
becomes evident that among children that avoid unhealthy behaviors, boys are more
involved in exercising than girls. Similar findings have been reported by Anderssen
and Wold (1992) and Mota and Queiros, (1996) who found that girls are significantly
less active than boys. Comparison of clusters one and four reveals that among
children that exercise regularly, there is a greater number of boys getting involved in
violent acts. This result incorporates findings from Paetsch and Bertrand (1997) who
found involvement in sport activities being associated to violent behavior and findings
from Herrenkohl et al. (2000) who reported male gender to be a significant predictor
of violence. Regarding smoking and eating behavior there were no differences
childrens healthy and unhealthy profiles 14
between boys and girls, in contrast to Lau et al. (1990) who reported that girls adopt
healthier eating habits than boys.
In relation to age, the patterns that were identified are quite worrying. It seems
that as children grow older they abandon healthy behaviors and adopt unhealthy ones.
Similar results regarding smoking and alcohol use have been reported by Johnston,
OMalley and Backman (2000), who found that during high school rates of smoking
and alcohol use increase rapidly with age, whereas similar conclusions have been
drawn by Botvin and Kantor (2000).
Family structure was another factor that seemed to influence the adoption of
healthy and unhealthy behaviors, especially smoking and exercising. A relatively
large percentage of children living with one or no parents are smokers and non-
exercisers, and respectively a relatively small percentage of those children are regular
exercisers. Similar patterns emerged for family income. However, overall, family
income did not seem to be a very crucial factor.
Regarding aspects of family structure and its relation to childrens behavior, the
findings of this study are in accordance with the existing literature suggesting that
children living in single-parent families are at high risk as far as unhealthy behaviors
are concerned (Roux et al., 1999). Shisslak et al. (1998) reported that girls having
separated or divorced parents show at risk-levels of weight control behaviors, whereas
Sobeck et al. (2000) found that it is more possible for smokers to come from families
that are not congruous. Similar results have been reported regarding drug use
(Friedman & Glassman, 2000). Finally, in relation to the present findings, Dahleberg
(1998) reports that among the factors that increase the possibility of violent behavior
are family malfunctioning and family disruption. Similar views have been expressed
childrens healthy and unhealthy profiles 15
by Herrenkohl et al. (2000) and Blackson et al. (1999) who claim that family conflicts
lead to violent behavior.
Overall, the demographic results of the present study indicate that regarding
healthy behaviors, older children, children from disrupted families, and mainly girls
can be characterized as at greater risk groups, whereas regarding unhealthy behaviors
older children, children from disrupted families, and mainly boys can be characterized
as at greater risk groups. However, it is also notable that younger children,
irrespective of other variables, exhibited healthier profiles. To our view, this finding
stresses the need to direct our attention to earlier ages where healthy habits are still
dominant and try to improve maintenance of healthier life-style.
Cluster profiles in relation to perceptions of social environment
Researchers claim that the effect of social factors on the adoption of healthy and
unhealthy behaviors are of particular importance. More specifically, it has been
supported that childrens behavior regarding exercise, healthy eating, smoking, and
violent behavior can be predicted by social factors, mainly family and peer influences
(Friedman & Glassman, 2000; Herrenkohl et al., 2000; Blackson et al., 1999; Brook et
al., 1999; Duncan et al., 1998). The second aim of the present study, was to identify
profiles of the Greek student population regarding healthy and unhealthy behaviors, in
relation to perceived family support and perceived family and peer behavior.
According to the results of the present study, the role of family support seems to
be influential in shaping behavior. A more careful examination reveals that family
support looks more crucial in relation to the adoption of unhealthy, rather than
healthy, behaviors. In particular, children who scored higher on family support were
those who do not smoke and do not take part in violent acts (cluster four and cluster
two). However, children in cluster two had the lowest scores on exercising and eating
childrens healthy and unhealthy profiles 16
fruits. Moreover, children who scored lower on family support were those who
engage in unhealthy behaviors, such as smoking and violent acts (cluster one and
cluster three), even though they scored higher in exercising and eating fruits
compared to children in cluster two.
Similar patterns of relationships were revealed regarding perceived family and
peer behavior. Thus, it seems that perceptions of social environment are more
influential in relation to the adoption or not of unhealthy, rather than healthy,
behaviors. In relation to these results, in the literature, there are evidence that parental
monitoring (Barnes et al., 1999; Duncan et al., 1998) as well as frequent conflicts
between parents and children (Duncan et al., 1998) relate to increased smoking among
children. Moreover, Grisler and Kandel (1998) suggested that supportive parental
behavior invokes reduction of smoking. Finally, Blackson et al. (1999) claim that
negatively perceived parental control and psychological dominance are responsible
for violent behavior, while Ary et al. (1999) reported loose parental control to be also
related to deviant behavior.
Overall, what seems of great importance to us is some of the patterns that
emerged. In particular, even though clear connections between smoking and non-
exercising were detected, the opposite was not evident, that is, no clear pattern
between non-smoking and exercising were revealed. This lead us to believe that
avoidance of unhealthy behaviors is not necessarily connected to adoption of healthy
behaviors. Furthermore, the relationship between exercising and violence indicates in
addition, that adoption of healthy behaviors is not necessarily connected to avoidance
of unhealthy behaviors.
Considering the evidence regarding prevailing lifestyles around the world (e.g.
Johnston et al. 2000, for the USA; Steptoe et al. 2002, for Europe, Leslie et al. 1999,
childrens healthy and unhealthy profiles 17
for Australia) the need to develop health promotion programs is universally
recognized. In accordance to health behavior theories personal and social
characteristics are amongst the important determinants of health-behavior. The results
of the present study can prove helpful in developing more efficient health promotion
programs. Demographic characteristics and descriptions of social environment can
help identify intervention targets. Furthermore, the way behaviors cluster can help
schedule more specific intervention programs in relation to specific groups. For
example, a program with emphasis on the importance of exercising in maintaining
healthy life style would be more appropriate for individuals who are not smokers but
are not physically active, whereas a program emphasizing long-term consequences of
unhealthy life-style and benefits of healthy habits would be more appropriate for
inactive individuals who also smoke and a program designed to promote fair-play,
respect for team-mates and opponents and avoidance of violence would be more
appropriate for exercisers with violent behavior. Finally, with regard to social
environment, this should also become part of the intervention programs with
educational programs for parents about personal behavior and parental guidance.
childrens healthy and unhealthy profiles 18
References
Ajzen, I. (September, 2002). Constructing a TpB Questionnaire: Conceptual and
Methodological Considerations. Retrieved October 1, 2002, from http:/www-
unix.oit.umass.edu/~aizen/
Anderssen, N. & Wold, B. (1992). Parental and peer influences on leisure-time
physical activity in young adolescents. Research Quarterly for Exercise and
Sport, 63, 341-348.
Ary, D. V., Duncan, T. E., Duncan, S. C. & Hops, H. (1999). Adolescent problem
behavior: the influence of parents and peers. Behavior Research and Therapy,
37, 217-230.
Babkes, M. L. & Weiss, M. R. (1999). Parental influence on children s cognitive and
affective responses to competitive soccer participation. Pediatric Exercise
Science, 11, 44-62.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive
theory. N.J.: Prentice-Hall.
Baranowski, T. (1997). Families and health actions. In D. S. Gochman (Ed.)
Handbook of health related research, Vol. 1: Personal and social
determinants. N.Y.: Plenum.
Barnes, G. M., Welte, J. W., Hoffman, J. H. & Dintcheff, B. A. (1999). Gambling and
alcohol use among youth. Influence of demographic, socialization and
individual factors. Addictive Behaviors, 24, 749-767.
Bauman, A. & Phongsavan, P. (1999). Epidemiology of substance use in adolescence:
prevalence, trends and policy implications. Drug and Alcohol Dependence, 55,
187-207.
http:/www
childrens healthy and unhealthy profiles 19
Becker, M. H. (1974, winter). The health belief model and personal health behavior.
Health Education Monographs 2, 404-419.
Blackson, T. C., Butler, T., Belsky, J., Ammerman, R. T., Shaw, D. S. & Tarter, R. E.
(1999). Individual traits and family contexts predict sons' externalizing
behavior and preliminary relative risk ratios for conduct disorder and
substance use disorder outcomes. Drug and Alcohol Dependence, 56, 115-131.
Botvin, G. J. & Kantor, L. W. (2000). Preventing alcohol and tobacco use through life
skills training: Theory, research and empirical findings. Alcohol Research &
Health, 24, 250-257.
Brook, U., Mendelberg, A., Galili, A., Priel, I. & Bujanover, Y. (1999). Knowledge
and attitudes of children towards cigarette smoking and its damage. Patient
Education and Counseling, 37, 49-53.
Carron, A. V., Hausenblas, H. A. & Mack, D. (1996). Social influence and exercise: a
meta-analysis. Journal of Sport and Exercise Psychology, 18, 1-16.
Dahlberg, L. L. (1998). Youth violence in the United States. Major trends, risk factors
and prevention approaches. American Journal of Preventive Medicine, 14,
259-272.
Duncan, S. C., Duncan, T. E., Biglan, A. & Ary, D. (1998). Contributions of the
social context to the development of adolescent substance use: a multivariate
latent growth modeling approach. Drug and Alcohol Dependance, 50, 57-71.
Dusenbury, L., Kerner, J. F., Baker, E., Botvin, G., James-Ortiz, S. & Zauber, A.
(1992). Predictors of smoking prevalence among New York Latino youth.
American Journal of Public Health, 82, 55-58.
childrens healthy and unhealthy profiles 20
Friedman, A. S. (1998). Substance use / abuse as a predictor to illegal and violent
behavior. A review of the relevant literature. Aggression and Violent
Behavior, 3, 339-355.
Friedman, A. S. & Glassman, K. (2000). Family risk factors versus peer risk factors
for drug abuse. A longitudinal study of an African American urban community
sample. Journal of Substance Abuse Treatment, 18, 267-275.
Griesler, P. C. & Kandel, D. B. (1998). Ethnic differences in correlates of adolescent
cigarette smoking. Journal of Adolescent Health, 23, 167-180.
Herrenkohl, T. I., Maguin, E., Hill, K. G., Hawkins, J. D., Abbot, R. D. & Catalano,
R. F. (2000). Developmental risk factors for youth violence. Journal of
Adolescent Health, 26, 176-186.
Hooper, C. A., Gruber, M. B., Munoz, K. D. & MacConnie, S. (1996). School-based
cardiovascular exercise and nutrition programs with parent participation.
Journal of Health Education, 27, 32-39.
Hupkens, C. L. H., Knibbe, R. A., Otterloo, A.H. & Drop, M. J. (1998). Class
differences in the food rules mothers impose on their children: A cross-
national study. Social Science & Medicine, 47, 1331-1339.
Johnston, L. D., OMalley, P. M., & Bachman, J. G. (2000). Monitoring the Future:
National Survey Results on Drug Use, 1975-1999: Volume I. Secondary
School Students. MD: National Institute on Drug Abuse.
Kendzierski, D. & Lamastro, V. D. (1988). Reconsidering the role of attitudes in
exercise behavior: A decision theoretic approach. Journal of Applied Social
Psychology, 18, 737-759.
childrens healthy and unhealthy profiles 21
Kimiecik, J. (1992). Predicting vigorous physical activity of corporate employees:
Comparing the theories of reasoned action and planned behavior. Journal of
Sport and Exercise Psychology, 14, 192-206.
Kimiecik, J., Horn, T. S. & Shurin, C. S. (1996). Relationships among children s
beliefs, perceptions of their parents beliefs, and their moderate-to-vigorous
physical activity. Research Quarterly for Exercise and Sport, 67, 324-336.
Lau, R. R., Quadrel, M. J. & Hartman, K. A. (1990). Development and change of
young adults preventive health beliefs and behavior: influence from parents
and peers. Journal of Health and Social Behavior, 31, 240-259.
Leslie, E., Owen, N., Salmon, J., Bauman, A., Sallis, J. & Lo, S. K. (1999).
Insufficiently active Australian college students: Perceived personal, social,
and environmental influences. Preventive Medicine, 28, 20-27.
Liang, W., Shediac - Ritzkallah, M. C., Celentano, D. D. & Rohde, C. (1999). A
population - based study of age and gender differences in patterns of health -
related behaviors. American Journal of Preventive Medicine, 17, 8-17.
McLellan, L., Rissel, C., Donnelly, N. & Bauman, A. (1999). Health behavior and the
school environment in New South Wales, Australia. Social Science &
Medicine, 49, 611-619.
Mota, H & Queiros, P. (1996). Children s behavior. physical activity regarding
parents perception vs. children s activity. Int. Rev. f. Soc. of Sport, 31, 173-
180.
Muecke, L., Simons-Morton, B., Huang, I. W. & Parcel, G. (1992). Is childhood
obesity associated with high-fat foods and low physical activity? Journal of
School Health, 62, 19-23.
childrens healthy and unhealthy profiles 22
Neumark - Sztainer, D., Story, M., Perry, C. & Casey, M. A. (1999). Factors
influencing food choices of adolescents: findings from focus - group
discussions with adolescents. Journal of the American Dieting Association, 99,
924-934.
Paetsch, J. J. & Bertrand, L. D. (1997). The relationship between peer, social, and
school factors, and delinquency among youth. Journal of School Health, 67,
27-32.
Pate, R. R., Trost, S. G., Felton, G. M., Ward, D. S., Dowda, M. & Saunders, R.
(1997). Correlates of physical activity behavior in rural youth. Research
Quarterly for Exercise and Sport, 68, 241-248.
Roosmalen, E. H. & McDaniel, S. A. (1989). Peer group influence as a factor in
smoking behavior of adolescents. Adolescence, 24, 801-816.
Rosenstock, I. M. (1991). The health belief model : Explaining health behavior
through expectancies. In K. Glanz, F. M. Lewis & B. K. Rimer (Eds.) Health
behavior and health education: Theory, research, and practice. San
Francisko : Jossey-Bass Publishers.
Roux, C., Le Couedic, P., Durand - Gasselin, S. & Luquet, F. (1999). Consumption
patterns and food attitudes of a sample of 657 low - income people in France.
Food Policy, 25, 91-103.
Sallis, J. F. & Hovel, M. F. (1990). Determinant of exercise behavior. Exercise and
Sport Sciences Reviews, vol. 18, (pp. 307-330). Baltimore: Williams &
Wilkins.
Shisslack, C. M., Crago, M., McKnight, K. M., Estes, L. S., Gray, N. & Parnaby, O.
G. (1998). Potential risk factors associated with weight control behaviors in
childrens healthy and unhealthy profiles 23
elementary and middle school girls. Journal of Psychosomatic Research, 44,
301-313.
Sobeck, J., Abbey, A., Agius, E., Clinton, M. & Harrison, K. (2000). Predicting early
adolescent substance use: do risk factors differ depending on age of onset?
Journal of Substance Abuse, 11, 89-102.
Steptoe, A., Wardle, J., Cui, W., Bellisle, F., Zotti, A., Baranyai, R. & Sanderman, R.
(2002). Trends in smoking, physical exercise, and attitudes toward health in
European university students from 13 countries, 1990-2000. Preventive
Medicine, 35, 97-104.
Stice, E. (1998). Modeling of eating pathology and social reinforcement of the thin -
ideal predict onset of bulimic symptoms. Behavior Research and Therapy, 36,
931-944.
Stucky-Ropp, R. C. & DiLorenzo, T. M. (1993). Determinants of exercise in children.
Preventive Medicine, 22, 880-889.
Sugathan, T. N., Moody, P. M., Bustan, M. A. & Elgerges, N. S. (1998). Age patterns
of smoking initiation among Kuwait university male students. Social Science
& Medicine, 47, 1855-1858.
Taylor, S. E. (1999). Health psychology. Boston: McGrow, Hill.
Theodorakis, Y., Papaioannou, A., & Karastogianidou K. (submitted). The effect of
family structure on attitudinal and behavioral factors related to healthy
lifestyle.
Torabi, M. & Nakornhet, N. (1996). Smoking among teenagers: educational
recommendations and resources. Journal of Health Education, 27, 40-43.
childrens healthy and unhealthy profiles 24
Trost, S. G., Pate, R. R., Ward, D. S., Saunders, R. & Riner, W. (1999). Correlates of
objectively measured physical activity in preadolescent youth. American
Journal of Preventive Medicine, 17, 120-126.
Unger, J. B. & Chen, X. (1999). The role of social networks and media receptivity in
predicting age of smoking initiation. A proportional hazards model of risk and
protective factors. Addictive Behaviors, 24, 371-381.
Vilhjalmsson, R. & Thorlindsson, T. (1998). Factors related to physical activity: a
study of adolescents. Social Science & Medicine, 47, 665-675.
Wang, M. Q., Fitzhugh, E. C., Westerfield, C. & Eddy, J. M. (1995). Family and peer
influence on smoking behavior among American adolescents: an age trend.
Society for Adolescent Medicine, 16, 200-203.
Wickrama, K. A. S., Lorenz, F. O. & Conger, R. D. (1997). Parental support and
adolescent physical health status: a latent growth-curve analysis. Journal of
Health and Social Behavior, 38, 149-163.
Zhu, B., Liu, M., Shelton, D., Liu, S. & Giovino, G. A. (1996). Cigarette smoking and
its risk factors among elementary school students in Beijing. American
Journal of Public Health, 86, 368-375.
childrens healthy and unhealthy profiles 25
Table 1.
Descriptive statistics for the examined behaviors, perceived family support and
perceived family and peer behaviors.
Variables N Mean SD
Exercise behavior 3307 2.47 1.70
Smoking behavior 3307 .61 1.52
Eating behavior 3307 2.81 1.51
Violence behavior 3307 .46 1.23
Perceived family Support 2866 4.70 .95
Perceived family and peer exercise behavior 2532 3.26 1.18
Perceived family and peer smoking behavior 2567 2.51 1.29
Perceived family and peer eating behavior 2287 4.71 1.28
Perceived family and peer violence behavior 2560 1.63 1.13
childrens healthy and unhealthy profiles 26
Table 2.
Mean scores, standard deviations and z-scores for the four clusters.
Cluster 1 (n= 221) Cluster 2 (n= 1272) Cluster 3 (n= 382) Cluster 4 (n= 1432)
mean s.d. z mean s.d. z mean s.d. z mean s.d. z
Exercising 3.38 1.57 .53 1.11 .95 -.80 1.93 1.82 -.31 3.68 1.13 .71
Smoking
0.25 .70 -.24 0.01 .41 -.34 4.66 .65 2.64 0.01 .26 -.38
Eating fruits 3.15 1.64 .23 2.09 1.20 -.47 2.63 1.77 -.12 3.44 1.37 .42
Participating in violent incidents 3.53 1.25 2.43 0.01 .30 -.33 1.58 1.98 .87 0.01 .20 -.45
childrens healthy and unhealthy profiles 27
Table 3.
Cluster profiles in relation to demographic characteristics.
Cluster1 Cluster 2 Cluster3 Cluster 4
Gender (n= 3212)
Males 172 (11.6%) 439 (29.6%) 190 (12.8%) 680 (45.9%)
Females 39 (2.3%) 800 (46.2%) 173 (10%) 719 (41.5%)
Family structure (n= 3244)
Both parents 180 (6.3%) 1105 (39%) 279 (9.8%) 1271 (44.8%)
One or no parents 32 (7.8%) 147 (35.9%) 93 (22.7%) 137 (33.5%)
Grade (n= 3303)
Elementary school 6th grade 33 (5.3%) 177 (28.5%) 11 (1.8%) 400 (64.4%)
Junior high school 2nd grade 57 (8.2%) 215 (39.8%) 39 (5.6%) 388 (55.6%)
Junior high school 3rd grade 72 (9.7%) 279 (37.8%) 63 (8.5%) 325 (44%)
High school 2nd grade 49 (6.7%) 339 (46.2%) 136 (18.5%) 210 (28.6%)
High school 3rd grade 8 (1.6%) 261 (51.2%) 133 (26.1%) 108 (21.2%)
Family income (n= 2425)
Much below average 3 (8.8%) 13 (38.2%) 9 (26.5%) 9 (26.5%)
Below average 7 (5.8%) 50 (41.7%) 17 (14.2%) 46 (38.3)
On average 61 (5.8%) 478 (45.4%) 113 (10.7%) 402 (38.1%)
Above average 52 (6.3%) 297 (36%) 91 (11%) 385 (46.7%)
Much above average 36 (9.2%) 93 (23.7%) 71 (18.1%) 192 (49%)
childrens healthy and unhealthy profiles 28
Table 4.
Mean scores on family support and perceived family and peer behavior for the four clusters.
Cluster 1 Cluster 2 Cluster3 Cluster 4 F
Family support (n= 2866) 4.23 2,3,4 4.76 1,3,4 3.94 1,2,4 4.89 1,2,3 88,16**
Perceived family and peer behavior (n= 1873)
Exercise 3.53 2,3 2.98 1,3,4 2.71 1,2,4 3.70 2,3 74.54**
Smoking 3.03 2,3,4 2.351,3 3.63 1,2,4 2.291,3 76.80**
Eating fruits 4.524 4.63 3,4 4.16 2,4 5.051,2,3 37.14**
Participating in violent incidents 2.76 2,3,4 1.36 1,3 2.26 1,2,4 1.43 1,3 100.66**
*p < .05, **p
childrens healthy and unhealthy profiles 29
childrens healthy and unhealthy profiles 30
childrens healthy and unhealthy profiles 31
Yannis Theodorakis University of Thessaly Department of Physical Education and Sport Science 42100, Karies, Trikala, Greece E-mail ytheo@pe.uth.grFax +302431 47042 Tel +302431 470001, Trikala, 05 December 2002 To: Professor Stuart Biddle, Loughborough University, Department of PE, Sports Science & Recreation Management, Loughborough, Leics., LE11 3TU,
Dear Stuart
Please find enclosed 4 copies of the paper titled " CHILDRENS PROFILE OF
HEALTHY AND UNHEALTHY BEHAVIORS: DEMOGRAPHIC
CHARACTERISTICS AND PERCEPTIONS OF SOCIAL ENVIRONMENT " which
I submit for publication into Psychology of Sport and Exercise.
I am looking forward to hearing from you. Sincerely yours,
Yannis Theodorakis,
Professor
mailto:ytheo@pe.uth.gr
childrens healthy and unhealthy profiles 32
e-mail:Phone:ytheo@pe.uth.gr+ 30 24310 47001AbstractIntroductionMethodParticipants and procedures
MeasuresSelf-report past behaviors. The examined behaviors were fourFamily structure. To assess family structure, participants wData analysisCluster analysisCluster profiles in relation to perceptions of social enviroDiscussionThe adoption of healthy and unhealthy behaviors is a matter Cluster profiles in relation to demographic characteristics
Variables
Cluster 1 (n= 221)Cluster 2 (n= 1272)Cluster 3 (n= 382)Cluster 4 (n= 1432)Cluster1Gender (n= 3212)Males
Family structure (n= 3244)Grade (n= 3303)Family income (n= 2425)Much below averageOn average
Cluster 1Family support (n= 2866)Perceived family and peer behavior (n= 1873)Exercise
Yannis Theodorakis