Health Education Research Advance Access published online on April 4, 2008
Health Education Research, doi:10.1093/her/cyn013
A review of similarities between domain-specific determinants of four health behaviors among adolescents
1 Graduate School of Teaching and Learning, Universiteit van Amsterdam, 1018 HJ Amsterdam, The Netherlands
2 TNO Quality of Life, PO Box 2215, 2301 CE Leiden, The Netherlands
3 Netherlands Institute for Health Promotion and Disease Prevention, 3440 AM Woerden, The Netherlands
Correspondence to: * Correspondence to: L. W. H. Peters. E-mail: louk.peters{at}tno.nl
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
Schools are overloaded with health promotion programs that, altogether, focus on a broad array of behavioral domains, including substance abuse, sexuality and nutrition. Although the specific content of programs varies according to the domain focus, programs usually address similar concepts: knowledge, attitudinal beliefs, social influences and skills. This apparent conceptual overlap between behaviors and programs provides opportunities for a transfer-oriented approach which will stimulate students to apply the knowledge and skills they have learned in one domain (e.g. skills for resisting tobacco use) to other domains (e.g. alcohol, sex). A requirement for such an approach is that behaviors share at least some determinants. This review addresses this issue by examining similarities between domain-specific determinants of smoking, drinking, safe sex and healthy nutrition among adolescents. Recent empirical studies and reviews were examined. The results show that the following determinants are relevant to all four behaviors: beliefs about immediate gratification and social advantages, peer norms, peer and parental modeling and refusal self-efficacy. Several other determinants have been found to relate to at least two behaviors, e.g. health risk beliefs and parental norms. These results can be used for the development of a transfer-oriented school health promotion curriculum.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
Health-compromising lifestyles such as smoking, binge drinking, unsafe sex and insufficient intake of fruit and vegetables are widely prevalent among young people in western societies [1, 2]. Numerous health education programs have been, and continue to be, developed to promote healthful behaviors among adolescents. The majority of adolescent health promotion programs are designed for use in schools and are often supplementary to the regular school curriculum. With a few exceptions, such as substance abuse programs, most projects focus on a single health-related behavior. Altogether, these single health promotion programs may overload the school curriculum and teaching staff [3, 4].
Transfer: looking for similarities
On a conceptual level, many classroom health education programs seem to address similar psychosocial constructs, such as factual knowledge, attitudinal beliefs, social influences and refusal skills [5–7]. The specific content of these constructs varies with the specific behavioral focus of individual programs as consequences, meanings and contexts of behaviors differ. However, the apparent conceptual overlap between health education programs provides opportunities for more integrative approaches, such as one that is oriented toward promoting transfer [8]. In a transfer-oriented approach, students are stimulated to apply the knowledge, attitudes and skills they have learned in one domain (e.g. refusal skills with respect to smoking) to other behavioral domains (e.g. refusing alcohol or unsafe sex). The teaching content thus focuses on building bridges between various behavioral domains, by identifying general principles and considering whether and how they can be applied in other domains. This does not mean that domain-specific issues are neglected. On the contrary, the transfer approach is about connecting domain-specific issues to general principles and vice versa. It requires alternate processes of contextualization (learning new skills in one context), decontextualization (deducing a general principle) and recontextualization (examining its application in other contexts) [9]. Thus, domain-specific issues may very well be addressed as contextualizations of general principles. Beliefs are most predictive of a given behavior when they specifically apply to that behavior [10], and new, meaningful knowledge can be attained only within the context of specific behavioral contexts.
In theory, a transfer-oriented curriculum can integrate and replace several domain-specific curricula and can produce effects on several behaviors simultaneously while keeping time and effort spent by schools and teachers at an acceptable level. Transfer effects have been reported in various subject domains in the education sector [11, 12] but, to our knowledge, they have not yet been examined in health education. We aim to fill this gap by developing and empirically testing a transfer-oriented approach in classroom health education in secondary education. The present literature review is one of the first steps in our project and has been conducted to examine opportunities for a transfer-oriented approach and more specifically to identify determinants to be included in a transfer-oriented program. A transfer-oriented approach to different lifestyles is only possible if these lifestyles have at least some determinants in common. Therefore, the purpose of this review is to examine similarities between determinants across several lifestyles. Determinants of various individual health-related behaviors have been studied extensively but, until now, no review has systematically examined which determinants are shared by several behaviors.
Four target behaviors were selected beforehand for this review: smoking, alcohol abuse, safe sex and healthy nutrition. These behaviors were selected because (i) they are among the ones most frequently addressed in Dutch secondary schools [13] and (ii) we expect there to be differences in the strength of relations between these behaviors, which may influence the occurrence or ease of transfer effects. We have reviewed studies of relations between the four behaviors elsewhere [14] and will address this issue in our empirical study. It is sufficient to mention that the strong clustering relation between tobacco and alcohol use that has often been reported [14] might lead to better transfer effects between these two behavioral domains than between domains that are not strongly related.
Since transfer-oriented learning is about discovering general issues in specific factors across domains, the focus of this review is on similarities between domain-specific determinants. The content of domain-specific factors varies with the behavioral domain in question. For instance, attitudinal beliefs about smoking are different from beliefs about condom use, because the behavioral consequences and circumstances of smoking and condom use differ. Domain-specific factors, such as attitudinal beliefs, are commonly addressed in categorical intervention programs. Despite their domain-specific content, such factors may share common ground on a more general level. For instance, the types of behavioral consequences may be similar for several behaviors: immediate physiological consequences, health consequences and social consequences. This common ground creates opportunities for teaching for transfer.
The focus on domain-specific determinants in this review does not mean that general determinants are insignificant in affecting various behaviors simultaneously. On the contrary, general factors, such as demographic, personality or parenting factors or general social or cognitive skills, are also very important. However, they were not the focus of this review as they have been previously addressed elsewhere [14].
Research question
Which domain-specific determinants correlate with two or more of the following behaviors: smoking, alcohol abuse, safe sex and healthy nutrition?
Theoretical model
Many theories have been formulated to predict health-related behaviors, which altogether have led to a broad array of determinants (see [15] for a comprehensive overview). We used the theory of triadic influence [16], which integrates insights from many theories, as a framework for organizing determinants of health behaviors [14]. Figure 1 shows a simplified version of this theory and our framework. It categorizes determinants in three streams (intrapersonal, interpersonal and cultural) and at three levels of influence (proximal, distal and ultimate). The ultimate level of influence includes determinants that are thought to be predictive of multiple behaviors but are almost unmodifiable, e.g. personality characteristics or the broader sociocultural environment. Their influence is mainly indirect, via determinants at the distal and proximal levels. Distal-level and especially proximal-level determinants have better predictive value, but most are specific to one behavior. In addition, intentions and previous experiences with the behavior are assumed to have the most direct influence, whereas barriers with regard to accessibility and availability may undermine intentional behavior. Although the figure only indicates within-stream influences from the ultimate level to the proximal level, we and others [16] assume that there are also interstream influences. The model also includes feedback loops which are indicated in the figure by the broken lines: experiences from performing a behavior give people feedback regarding, for instance, some of its consequences [16].
|
Given our focus on domain-specific determinants, the determinants discussed in this review are, for the most part, but not exclusively, proximal determinants, such as attitudinal, social normative and self-efficacy beliefs.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
Sample of studies
The databases Medline and PsycINFO were used to generate the sample of studies. Searching this combination of databases meets criteria for a comprehensive search, as stated in a quality assessment tool for reviews [17] and is an efficient way for locating studies relevant to health promotion [18]. We used the following key words for determinants: risk-taking, risk factors, risk perception, psychosocial factors, psychology, intention, motivation, personality (characteristics), personality correlates, predisposition, knowledge, attitudes and practice. We performed searches for every behavior and for multiple behaviors. For every search, we added key words specific to that behavior. For tobacco and alcohol: tobacco, smoking, cigarette, substance use, substance abuse, drug use, drug abuse, alcohol, alcoholic, drinking, binge drinking, alcohol drinking patterns and alcohol drinking attitudes. For safe sex: safe sex, contraception behavior, condoms, AIDS/prevention and control, aids prevention, sexual risk taking, psychosexual behavior and AIDS attitudes. For nutrition: food preferences, diets, feeding practices, eating attitudes, food intake, fruit, fat, vegetables, adolescent nutrition and food habits. For multiple behaviors: generalization learning, transfer learning, health compromising behavior, lifestyle, health behavior, problem behavior, risk behavior and behavior problems. In addition, backward searches were conducted by scanning reference lists.
Inclusion criteria
Studies were included if they met the following criteria:
- (i) Studies on behavior-specific correlates of a measure of (self-reported) behavior or intention with respect to smoking, drinking, sexual behavior or healthy nutrition.
- (ii) Correlates were measured at ages 10–18.
- (iii) Data collection was carried out in western countries.
- (iv) Publications were written in English and published in journals from the Social Science Citation Index list.
- (v) Empirical and review studies were considered. Reviews had to be published between 1995 and 2003 and empirical studies between 2000 and 2003. Because there were so few studies that addressed nutrition, we included empirical studies on nutrition from 1995 to 2003.
- (vi) Because of the large numbers of longitudinal studies on tobacco and alcohol use, we included only longitudinal studies for these behaviors.
- (ii) Correlates were measured at ages 10–18.
The publication year criterion for empirical studies was strict because of the quantity of material on the four behavioral domains. Reviews were included to account for results of older studies.
Eighty-seven studies were found to satisfy the inclusion criteria: 14 were on multiple behaviors, 26 on smoking, 10 on alcohol use, 17 on safe sex and 20 on nutrition. Some of the studies also discussed other behaviors in addition to the ones of interest here, but results for these additional behaviors were not recorded.
Coding and synthesis
The studies were divided into three groups which were coded by three reviewers: smoking and alcohol use (LP), safe sex and multiple behaviors (CW) and nutrition (FH). Although each behavioral domain was assessed by one reviewer only, several procedures were used to ensure comparability of coding. Firstly, all reviewers were familiar with conducting literature reviews and with research in all four behavioral domains. Secondly, standardized assessment forms (available from the first author) were used for systematically recording study characteristics. Thirdly, coding of studies was discussed in several meetings and any doubts or problems with coding were resolved through discussion after all reviewers had read the relevant portions of the paper in question. For empirical studies, the following aspects were recorded: study design (longitudinal, cross-sectional), sample size, participant characteristics (age or grade, gender, ethnicity, socio-economic status, country of residence), measurement of determinants (questionnaire, interview; specific measures recorded; yes/no validated), measurement of behavior or intention (questionnaire, interview, observation, biomedical, other; specific measure recorded; yes/no validated), theoretical basis, statistical analyses used (correlation, regression, other, none) and the relation between each determinant and behavior (positive, negative or null; for total sample or subgroup; strength of relationship in correlation, beta weight or odds ratio). Determinants recorded for focus group studies (only in the domains of safe sex and nutrition: study numbers 77, 90, 95 and 97 in Table II) mostly pertained to aspects that, according to the authors of the study, were mentioned frequently in discussion groups. A separate assessment form was used for review studies which contained information on type of review (meta-analysis, narrative), characteristics of included studies (number of studies, study designs, sample sizes, participant characteristics), review authors’ judgment of quality of study designs and instruments and conclusions about relations between determinants and behavior.
|
After initial data were extracted, determinants were further organized in several steps, which is explained in Table I.
|
Coding of study characteristics was descriptive and studies were not rated for overall methodological quality. However, in data synthesis, the type of study was taken into consideration. Results of longitudinal studies were generally rated as being stronger than those of cross-sectional studies because a longitudinal design has better predictive value. Review studies were treated with more caution than empirical studies in our synthesis, especially when evidence was mainly from reviews or when evidence from reviews conflicted with that from empirical studies. This caution is warranted, as using review results may have some disadvantages. Because of their second-hand nature, review results may be less insightful than empirical results. Results of some empirical studies may be overrepresented, as they are perhaps discussed in several reviews. Also, reviews vary in the specificity of the outcome measure and in the number and quality of studies included and sometimes study design or quality is not addressed. Moreover, some reviews only discuss positive findings and do not mention null findings.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
Study characteristics
Table II presents an overview of the characteristics of the studies included. Studies are grouped according to the criterion behavior(s) and according to study design.
The behavioral focus of both the empirical and review studies on multiple behavior was mainly on alcohol and tobacco use. Sexual and nutrition behaviors were only addressed in some of these studies. Of the eight empirical multiple behavior studies, four were longitudinal and four cross-sectional. As for studies that examined only one behavior, empirical studies on safe sex and nutrition were almost exclusively cross-sectional; only one longitudinal nutrition study was located. In the tobacco and alcohol domains, the longitudinal design was much more prevalent, which had led to the decision to include only longitudinal studies for these domains.
Most studies were conducted in the United States. The majority focused on both males and females and on samples with various ethnic composition, with some exceptions especially among safe sex studies (e.g. black females). The age of the respondents in the empirical studies ranged from 7 to 21 years, with a bottom end mean of 12.7 and a top end mean of 16.6 years (overall mean age 14.7 years). Safe sex studies generally examined somewhat older samples, with a mean age range of 13.7–18.3 and an overall mean age of 16.0 years.
The operationalization of the behavioral criterion variables differed considerably. Tobacco use measures included long-term smoking trajectories (e.g. [34, 44]), established smoking (e.g. 100 cigarettes lifetime) [35, 36], daily smoking (e.g. [42]), ever smoking (e.g. [22]) and both experimental and regular smoking (e.g. [43]). Alcohol studies generally examined heavy use or binge drinking. Studies of safe sex commonly addressed (intended) use of condoms or risky sexual behavior in general, but two multiple behavior studies focused on sexual experience. Studies of nutrition behavior showed the largest variation in behavioral outcomes. Some focused on more or less specific outcomes such as consumption of raw vegetables, of selected foods or of fruit and vegetables in general, whereas others assessed nutrient or food intake or its quality or even eating behavior in general (e.g. [98]). Many studies used generally established outcome measures, but specific information about validity and reliability of measures was often not provided.
The operationalization of determinants also showed a high level of variation. Nearly every empirical study used its own measures and some did not give specific accounts of these. Most empirical studies reported on reliability (internal consistency), but information about validity was largely absent. Reviews generally did not go into details of the measures used.
Results of studies
The process of combining the domain-specific determinants into meaningful categories led to a total of 86 determinants. Of these 86 determinants, the majority (51) had been examined for only one behavior and a minority had been examined for two behaviors (25), three behaviors (4) or for all four behaviors (6). Table III presents the 35 determinants that were examined for more than one behavior. Since our interest is in discovering similar determinants across different behaviors, we will mainly focus on the results in this table. In line with our theoretical model (see Fig. 1), the 35 determinants in Table III were categorized as: 4 behavioral factors, 1 barrier/availability factor, 23 proximal factors, 5 distal factors and 2 ultimate factors. The table indicates, for each study, the direction of the determinant–behavior relationship that was found (positive or negative influence or null findings). It does not provide information about the strength of the relationships. Unfortunately, such information was insufficient in many papers (e.g. only significance levels or group means reported) and totally absent in most reviews.
|
Table IV displays the 51 determinants that have been measured in one domain only. This table is included to complete the overview of all determinants but will not be addressed frequently.
|
Behavioral and availability factors
As for behavioral factors, similarities between the tobacco and alcohol domains exist since these behaviors are predicted by positive experiences with the substance, previous use of the substance in general and early onset of use. The latter finding corresponds to the evidence in the sexuality domain that lower age of first intercourse correlates negatively with safe sex behavior. Behavioral factors that were only examined for one behavior (see Table IV) mainly pertained to situational characteristics in the nutrition domain and are not discussed here further.
Availability/accessibility factors have been examined recently only in the nutrition and tobacco domains. Evidence in the nutrition domain, mostly from focus group studies and reviews, consistently suggests that such factors impact nutrition behavior. Correspondingly, in the tobacco domain, there is some evidence that accessibility of cigarettes is related to smoking.
Proximal factors
Attitudes.
As expected, most domain-specific factors examined were proximal, consisting mainly of attitudinal and social normative beliefs. General, mixed or unspecified measures of attitudes have been found to relate positively to all four behaviors, although some studies reported null findings. The specific attitudinal beliefs examined pertained mainly to health, physiological and psychological gratification, appearance, performance and social contact. Health-related beliefs have been studied for all behaviors, although there is only one such study on alcohol. Positive associations with the health behaviors prevail, although many studies, including the alcohol study, reported null findings. Personal risk beliefs appear to be better predictors than general risk beliefs, but correlations were predominantly weak and some studies reported negative associations. Therefore, health beliefs seem to be relevant, though minor, determinants of safe sex, healthy nutrition and non-smoking.
The evidence for the relevance of beliefs related to physiological and psychological gratification is more consistent. Beliefs that the unhealthy behavior contributes to an immediate positive sensation or that the healthy behavior would obstruct this are related to unhealthy lifestyles in each of the domains studied but especially in those of nutrition and safe sex. The belief that smoking relaxes or helps reduce negative feelings is a consistent predictor of tobacco use; such belief in the relaxing effects of alcohol has also been reported. Image-related beliefs have only been reported in reviews on tobacco (e.g. smoking makes you feel rebellious, see Table IV) and are therefore not discussed here further.
Whereas most beliefs about gratification were in favor of unhealthy behavior, anticipated regret about a hangover or drunken behavior had a negative association with binge drinking; this regret was not related to smoking.
Beliefs related to physical appearance have only been examined in the nutrition and tobacco domains. The belief that smoking has a favorable effect on weight management is negatively associated with non-smoking, as was reported consistently by one longitudinal study and five reviews, whereas the association between weight management beliefs and healthy nutrition behavior tends to be positive. The evidence in the nutrition domain is weaker than that found in the smoking domain since it is based on one longitudinal study with positive results and one cross-sectional study with null findings. Such contrasts have also been found for performance-related beliefs. The belief that healthy behavior promotes physical or athletic performance is associated positively with healthy nutrition and non-smoking. However, a review in the alcohol domain reported positive alcohol expectancies for mental and motor performance among children of alcoholics who are at risk of developing alcohol or drug problems.
There are relatively few studies on beliefs about social consequences which is surprising, given that social norms and especially modeling behavior have been studied extensively (see below). Nevertheless, beliefs that the unhealthy behavior has social advantages have been found for tobacco and alcohol use and safe sex, although for tobacco use also null findings were reported. A somewhat comparable finding in the nutrition domain was the belief that certain social situations such as parties are not conducive to making healthy food choices (see Table IV). Only one finding in the category of social consequences was in the opposite direction: the belief that too much alcohol intake can lead to bad conduct (see Table IV).
Social norms.
Social normative beliefs have been studied in relation to several reference groups but mostly peers and parents. Peer norms have been found to have an effect on all four behaviors. However, results in the alcohol domain are inconsistent, with one longitudinal and two review studies reporting the absence of an association and, in the tobacco domain, much of the evidence stems from reviews. The findings for parental norms are more consistent, at least in the domains of smoking and drinking. Social norms in the sex domain were only examined in one study [23]. It has been found that use of tobacco and alcohol is stimulated when these products are offered. However, adolescents do not feel overtly pressurized by others to engage in substance use. Rather, peer pressure is reported to be more internalized: adolescents want to do what (they see or think) others do [28, 52].
Self-efficacy.
Self-efficacy has been studied less frequently than other proximal factors. General or unspecified measures, mainly used in the nutrition domain, have consistently shown positive correlations with healthy behavior. Refusal self-efficacy has been examined in only a few studies, but there are positive results for all four domains. Other self-efficacy beliefs have been studied, predominantly in the safe sex domain, with the main focus on skills for using and discussing condoms.
Distal factors
Distal domain-specific determinants generally pertain to knowledge and modeling behavior. Knowledge of behavior risks has mostly been studied in the safe sex domain, where results of cross-sectional studies and reviews are conflicting. Positive associations between knowledge and healthy behavior have been reported mainly in reviews, whereas cross-sectional studies have shown null findings or negative associations. Reviews in the domains of nutrition and tobacco indicate that knowledge of behavior risks does not seem to relate directly to behavior; in these reviews, correct information is suggested to be a prerequisite for healthy behavior.
Modeling behavior has received much attention in determinant research, especially in the domain of smoking. Perceived health behavior of peers or friends seems to relate positively to adolescents’ own health behavior in all four domains, although the absence of such a relation was also found for all behaviors. The influence of friends may be overrated in studies, especially cross-sectional ones, as selection and projection processes appear to account for at least a part of the correlation [28]. Nevertheless, in the domain of substance use, not only perceived but also actual peer use relates to adolescents’ own use of tobacco or alcohol, although correlations with actual use are generally lower than those with perceived use [28].
Perceived health behavior of parents has been related to adolescents’ own behavior in all four domains, with the most and firmest evidence coming from the tobacco and alcohol domains and least evidence from the sex domain.
Ultimate factors
At the ultimate level, only two behavior-specific factors were identified: media influence and genetic factors. Two reviews on nutrition reported that the media had a negative influence on healthy nutrition. In the tobacco domain, evidence for negative media influence is very weak, although one longitudinal study found a negative influence of susceptibility to advertising for cigarettes. As for genetic factors, four reviews in the alcohol domain consistently reported that a genetic component to at least one type of problem drinking has been identified. In the tobacco domain, the evidence for genetic factors is less consistent. One review concluded that there is only weak evidence for a genetic influence on smoking. Another review discussed studies that reported substantial heritability but was unclear about the strength of the evidence.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
Similarities between behavior-specific determinants
This review has focused on similarities between behavior-specific determinants of four health-related behaviors: smoking, (binge) drinking, safe sex and healthy nutrition. To allow comparison across different behaviors, the determinants were carefully categorized, where possible, to a higher, non-behavior-specific level. Thirty-five determinants were identified that have been studied for more than one behavior.
Several determinants were found to be relevant for all four behaviors: beliefs that the unhealthy behavior will lead to immediate gratification and to social advantages, peer norms, peer and parental modeling behavior and refusal self-efficacy. Moreover, the direction of each determinant's relationship with behavior (i.e. as a risk or protective factor) was consistent across the four domains. These determinants appear to be the most relevant ones to include in a transfer-oriented program.
For the remaining determinants that have been examined for multiple behaviors, the direction of their influence is in most cases the same across behaviors. A negative influence on multiple behaviors was found for previous experience with the unhealthy behavior (tobacco and alcohol), early onset of unhealthy behavior (tobacco, alcohol and sex), availability or accessibility of unhealthy products (nutrition and tobacco), school acceptance of substances (tobacco and alcohol) and offers of unhealthy products (tobacco and alcohol). A protective influence on multiple behaviors was found for perceived personal health risk (sex, nutrition and tobacco), strict parental norms and rules (nutrition, tobacco and alcohol) and strict sibling norms (nutrition and tobacco). The influence of several factors was inconsistent across behaviors or was unclear for weight management beliefs (risk factor for smoking, inconsistent findings for nutrition), performance beliefs (protective factor for nutrition and smoking, risk factor for alcohol), knowledge of behavior risks (inconsistent findings for safe sex, unimportant for tobacco and nutrition) and media portrayals and commercials (risk factor for nutrition, very weak evidence for tobacco).
Out of a total of 86 determinants, 51 could not be classified meaningfully to a higher level or have only been studied for one behavior. This may be partly due to our conservative categorization process. For some determinants, their uniqueness may be due to their behavior-specific relevance. For instance, the perceived risk of pregnancy is only directly relevant for sexual behavior; we could not think of a meaningful category that would include similar beliefs for other behaviors. Other determinants, however, may be relevant for all domains but may not have been examined for them all. For instance, in the alcohol domain, only one study had examined health-related beliefs.
In addition to this paper's main focus on overlap across domains, it presents a broad overview of research results in four domains. Researchers in a particular domain can use the results of this review to look beyond the boundaries of their own domain to generate ideas from results in other domains.
Implications for interventions
A prerequisite for developing interventions that are tailored to multiple behaviors is that these behaviors have some predictors in common. After all, if a factor is predictive of several behaviors, an intervention that can impact that factor may contribute to changes in all related behaviors. In a recent review, we found evidence that several general, non-domain-specific factors (e.g. self-esteem, warm and strict parenting style) are predictive of all four behaviors that were also examined in the present review [14]. Interventions that affect such factors have thus the potential to lead to changes in all four behaviors. The Child Development Project and the Seattle Social Development Project are examples of such an approach in the primary school setting [106, 107].
The present review concentrated on domain-specific predictors or correlates. These predictors are mostly proximal factors, comprising attitudinal, social normative and self-efficacy beliefs, and are the typical focus of educational interventions. Research in social psychology and health promotion has shown that such beliefs are most predictive of a specific behavior when they are formulated specifically in terms of that behavior [10]. It is not likely that, without extra effort, changes in such factors in one domain will lead to changes in similar factors in other behavioral domains. Research in education has shown that transfer of learning—e.g. from the school context to the private or work setting or from one situation or problem to another—does not happen by itself but must be actively promoted [108]. The issue of transfer has been raised from different theoretical points of view, mainly from cognitive psychology and situated perspectives, which have different implications for promoting transfer [109]. Situated perspectives emphasize that abstract schooling does not make sense to young people [110]. Knowledge and skills should be meaningful in the context of the students’ personal objectives in order for it to be carried over to a similar problem or behavior domain [110]. The perspective of cognitive educational psychology is relevant to the finding of this review that various behaviors have similar determinants. To achieve transfer, the teaching content should not only focus on domain-specific issues but also invite students to decontextualize these issues into general principles and to examine and practice their application in various other behavioral domains (e.g. [9]). For instance, learning how to refuse a cigarette by understanding general refusal skills can help students to refuse alcohol use or unsafe sex. Application to other domains should be specific and should include relevant domain-specific knowledge, beliefs and circumstances as well as an assessment of the similarities and dissimilarities between domains. In the case of recontextualizing refusal skills from the tobacco to the alcohol domain, students could be invited to act out a situation involving alcohol. They would then assess what the situation entails, look at the ways it is comparable to or different from a tobacco situation, examine whether the response options are comparable and discuss which specific response could be used. By practicing this in several domains, students may learn to use their knowledge and skills flexibly, thus increasing the chance that they will use them in domains they have not rehearsed.
Examples of other general principles that seem relevant in the light of the findings of this review are understanding the mechanisms of social influences, exploring and questioning expected consequences of the target behavior, exploring alternative behaviors that have similar immediate gratification or social advantages but are less health compromising and considering and weighing various behavioral options and their consequences (decision making and problem solving). However, since we do not know of any examples of explicit transfer-oriented learning in health promotion, it is not altogether clear what level of generalization would work best. Moreover, whatever level of generalization is chosen, domain-specific components will always be necessary. After all, young people will have to learn basic domain-specific knowledge and skills.
Limitations
This review fulfills generally acknowledged criteria for systematic reviews [111]: identification of the review question in advance, comprehensive literature search, use of explicit inclusion/exclusion criteria, application of established standards for appraising study quality and explicit methods of extracting and synthesizing study findings. The following limitations should be discussed.
Because of our broad focus on four health-related behaviors, we had to limit our search and may thus have missed relevant studies. Optimal use of restricted resources was made by searching a medical and a social science database [18], by searching empirical as well as review studies and by backward search. Reviews were included to account for results of older studies but, as was mentioned above, this may have some disadvantages, such as the danger of overrepresentation of certain results.
There was considerable variation across the four behavioral domains in the design of the empirical studies. Whereas nearly all studies of safe sex and healthy nutrition had a cross-sectional design, all empirical studies of tobacco and alcohol use were longitudinal. In terms of causality, the findings on smoking and alcohol abuse are thus more robust than the findings on safe sex and nutrition. Although this may hamper comparison of results across different domains, the results in each behavioral domain can be considered to reflect available evidence and current study quality standards within that domain.
Within behavioral domains, and especially in the nutrition domain, there was great variation in outcome measures. We included all measures and thus looked at broad behavioral domains, since there is no consensus as to which specific outcome measures in these domains are most relevant.
Definition of determinants was in some cases unclear, as most reviews and some empirical studies did not give specifications of measures. Therefore, as stated earlier, we categorized the determinants conservatively. If we were not sure that determinants addressed the same content or concept, they were treated as separate determinants. Placement under the same heading indicates that there is at least some similarity between determinants. In addition, studies that examined multiple behaviors measured the determinants for each of the behaviors in the same way. In these studies, the results did not differ from studies that examined only one behavior.
This review could even have been stronger if, in addition to type of study, we had included other methodological aspects for weighing study results. Such aspects may include validity and reliability of measures, level of respondent representation and appropriateness of statistical analyses.
Although use of stricter or alternative review methodology might have led to other specific results for some factors or behaviors, it is not likely that the main finding of this review would be different, i.e. that there are similarities between domain-specific determinants across behavioral domains. Despite the inclusion of studies with a variety of designs, measures and analyses, the results for most of the determinants examined for multiple behaviors in this review point in the same direction: most determinants are either a risk factor or a protective factor across different behavioral domains. This main finding implies that an important precondition for a transfer-oriented approach to adolescent health promotion can be met. Such an approach is new to this field but seems promising. The determinants that were found to be relevant to all four behaviors are the primary candidates for consideration in a transfer-oriented program.
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
The Netherlands Organization for Health Research and Development (4005.0006).
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
None declared.
|
Footnotes |
|---|
4 F. Hoekstra is currently employed by Health Service Amsterdam, The Netherlands.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
We thank three anonymous reviewers for their helpful comments.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflict of interest statementAcknowledgementsReferences |
|---|
1. Currie C, Roberts C, Morgan A, et al, eds. Young People's Health in Context. Health Behaviour in School-Aged Children (HBSC) Study: International Report from the 2001/2002 Survey (2006) Copenhagen, Denmark: World Health Organization Regional Office for Europe.
2. Eaton DK, Kann L, Kinchen S, et al. Youth risk behavior surveillance—United States. 2005. MMWR Surveill Summ (2006) 55:1–108.[Medline]
3. Lee A, Keung MW, Tsang CKK. Difficulties faced by health educators in Hong Kong: from barriers to strategies. J Prim Care Health Promot (2004) 1:17–22.
4. Leurs M, Jansen M, Schaalma H, et al. The tailored Schoolbeat-approach: new concepts for health promotion in schools in The Netherlands. In: The Health Promoting School: International Advances in Theory, Evaluation and Practice—Clift S, Jensen BB, eds. (2005) Copenhagen, Denmark: Danish University of Education Press. 89–107.
5. Botvin GJ, Schinke S, Orlandi MA. School health promotion: substance abuse and sexual behavior. Appl Prev Psychol (1995) 4:167–84.[CrossRef][Web of Science]
6. Schaalma HP, Abraham C, Gillmore MR, et al. Sex education as health promotion: what does it take? Arch Sex Behav (2004) 33:259–69.[CrossRef][Web of Science][Medline]
7. Summerfield LM. School health education. In: Handbook of Research on Teaching—Richardson V, ed. (2002) Washington, DC: American Educational Research Association (AERA). 472–90.
8. Ten Dam GTM. Effectiveness in health education. In: Education and Health in Partnership: A European Conference on Linking Education with the Promotion of Health in Schools—Young I, ed. (2002) Woerden, The Netherlands: Netherlands Institute for Health Promotion and Disease Prevention/International Planning Committee. 17–22.
9. Elshout-Mohr M, Van Hout-Wolters BHAM, Broekkamp H. Mapping situations in classroom and research: eight types of instructional learning episodes. Learn Instr (1999) 9:57–75.[CrossRef]
10. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process (1991) 50:179–211.[CrossRef][Web of Science]
11. Alexander PA. Psychology in Learning and Instruction (2006) Upper Saddle River, NJ: Pearson Education.
12. Mayer RE, Wittrock MC. Problem-solving transfer. In: Handbook of Educational Psychology—Berliner DC, Calfee RC, eds. (1996) New York: Simon & Schuster Macmillan. 47–62.
13. Dafesh ZJA, Buijs GJ, Hekkink C, et al. Regioscan GGD en Gezondheid op Scholen: Onderzoek naar de Stand van Zaken rond de Ondersteuning van Scholen door GGD'en op het Terrein van Gezondheid (2006) Woerden, The Netherlands: Netherlands Institute for Health Promotion and Disease Prevention.
14. Wiefferink CH, Peters L, Hoekstra F, et al. Clustering of health-related behaviors and their determinants: possible consequences for school health interventions. Prev Sci (2006) 7:127–49.[CrossRef][Web of Science][Medline]
15. Petraitis J, Flay BR, Miller TQ. Reviewing theories of adolescent substance use: organizing pieces in the puzzle. Psychol Bull (1995) 117:67–86.[CrossRef][Web of Science][Medline]
16. Flay BR, Petraitis J. The theory of triadic influence: a new theory of health behavior with implications for preventive interventions. Adv Med Sociol (1994) 4:19–44.
17. Thomas H, Micucci S, Ciliska D, et al. Effectiveness of School-Based Interventions in Reducing Adolescent Risk Behaviours: A Systematic Review of Reviews (2005) Hamilton, Canada: Effective Public Health Practice Project.
18. Peersman G, Harden A, Oliver S, et al. Effectiveness Reviews in Health Promotion (1999) London: EPPI-Centre.
19. Adalbjarnardottir S, Hafsteinsson LG. Adolescents’ perceived parenting styles and their substance use: concurrent and longitudinal analyses. J Res Adolesc (2001) 11:401–23.[Web of Science]
20. Goldberg JH, Halpern-Felsher BL, Millstein SG. Beyond invulnerability: the importance of benefits in adolescents’ decision to drink alcohol. Health Psychol (2002) 21:477–84.[CrossRef][Web of Science][Medline]
21. Maxwell KA. Friends: the role of peer influence across adolescent risk behaviors. J Youth Adolesc (2002) 31:267–77.[CrossRef][Web of Science]
22. Wills TA, Sandy JM, Yaeger AM. Stress and smoking in adolescence: a test of directional hypotheses. Health Psychol (2002) 21:122–30.[CrossRef][Web of Science][Medline]
23. Beal AC, Ausiello J, Perrin JM. Social influences on health-risk behaviors among minority middle school students. J Adolesc Health (2001) 28:474–80.[CrossRef][Web of Science][Medline]
24. La Greca AN, Prinstein MJ, Fetter MD. Adolescent peer crowd affiliation: linkages with health-risk behaviors and close friendships. J Pediatr Psychol (2001) 26:131–43.
25. Maes L, Lievens J. Can the school make a difference? A multilevel analysis of adolescent risk and health behavior. Soc Sci Med (2003) 56:517–29.[CrossRef][Web of Science][Medline]
26. Topolski TD, Patrick DL, Edwards TC, et al. Quality of life and health-risk behaviors among adolescents. J Adolesc Health (2001) 29:426–35.[CrossRef][Web of Science][Medline]
27. Amaro H, Blake SM, Schwartz PM, et al. Developing theory-based substance abuse prevention programs for young adolescent girls. J Early Adolesc (2001) 21:256–93.[CrossRef]
28. Bauman KE, Ennet ST. On the importance of peer influence for adolescent drug use: commonly neglected considerations. Addiction (1996) 91:185–98.[CrossRef][Web of Science][Medline]
29. Belcher HM, Shinitzky HE. Substance abuse in children: prediction, protection, and prevention. Arch Pediatr Adolesc Med (1998) 152:952–60.
30. Fahs PS, Smith BE, Atav AS, et al. Integrative research review of risk behaviors among adolescents in rural, suburban, and urban areas. J Adolesc Health (1999) 24:230–43.[CrossRef][Web of Science][Medline]
31. Scaramella LV, Keyes AW. The social contextual approach and rural adolescent substance use: implications for prevention in rural settings. Clin Child Fam Psychol Rev (2001) 4:231–51.[CrossRef][Web of Science][Medline]
32. Swadi H. Individual risk factors for adolescent substance use. Drug Alcohol Depend (1999) 55:209–24.[CrossRef][Web of Science][Medline]
33. Carvajal SC, Wiatrek DE, Evans RI, et al. Psychosocial determinants of the onset and escalation of smoking: cross-sectional and prospective findings in multiethnic middle school samples. J Adolesc Health (2000) 27:255–65.[CrossRef][Web of Science][Medline]
34. Chassin L, Presson CC, Sherman SJ. The natural history of cigarette smoking from adolescence to adulthood in a midwestern community sample: multiple trajectories and their psychosocial correlates. Health Psychol (2000) 19:223–31.[CrossRef][Web of Science][Medline]
35. Choi WS, Gilpin EA, Farkas AJ, et al. Determining the probability of future smoking among adolescents. Addiction (2001) 96:313–23.[CrossRef][Web of Science][Medline]
36. Choi WS, Ahluwalia JS, Harris KJ, et al. Progression to established smoking: the influence of tobacco marketing. Am J Prev Med (2002) 22:228–33.[CrossRef][Web of Science][Medline]
37. Ellickson PL, McGuigan KA, Klein DJ. Predictors of late-onset smoking and cessation over 10 years. J Adolesc Health (2001) 29:101–8.[Web of Science][Medline]
38. Epstein JA, Griffin KW, Botvin GJ. A model of smoking among inner-city adolescents: the role of personal competence and perceived social benefits of smoking. Prev Med (2000) 31:107–14.[CrossRef][Web of Science][Medline]
39. Hine DW, McKenzie-Richer A, Lewko J, et al. A comparison of the mediational properties of four adolescent smoking expectancy measures. Psychol Addict Behav (2002) 16:187–95.[CrossRef][Web of Science][Medline]
40. Orlando M, Ellickson PL, Jinnett K. The temporal relationship between emotional distress and cigarette smoking during adolescence and young adulthood. J Consult Clin Psychol (2001) 69:959–70.[CrossRef][Web of Science][Medline]
41. Soldz S, Cui X. Pathways through adolescent smoking: a 7-year longitudinal grouping analysis. Health Psychol (2002) 21:495–504.[CrossRef][Web of Science][Medline]
42. Tucker JS, Ellickson PL, Klein DJ. Five-year prospective study of risk factors for daily smoking in adolescence among early nonsmokers and experimenters. J Applied Soc Psychol (2002) 32:1588–603.[CrossRef]
43. Wang MQ. Social environmental influences on adolescents’ smoking progression. Am J Health Behav (2001) 25:418–25.[Web of Science][Medline]
44. White HR, Pandina RJ, Chen PH. Developmental trajectories of cigarette use from early adolescence into young adulthood. Drug Alcohol Depend (2002) 65:167–78.[CrossRef][Web of Science][Medline]
45. Woodruff SI, Candelaria JI, Kaniado-Laborín R, et al. Availability of cigarettes as a risk factor for trial smoking in adolescents. Am J Health Behav (2003) 27:84–8.[Web of Science][Medline]
46. Avenevoli S, Merikangas KR. Familial influences on adolescent smoking. Addiction (2003) 98(Suppl 1):1–20.[Web of Science][Medline]
47. Darling N, Cumsille P. Theory, measurement, and methods in the study of family influences on adolescent smoking. Addiction (2003) 98(Suppl 1):21–36.
48. Derzon JH, Lipsey MW. Predicting tobacco use to age 18: a synthesis of longitudinal research. Addiction (1999) 94:995–1006.[CrossRef][Web of Science][Medline]
49. DuRant RH, Smith JA. Adolescent tobacco use and cessation. Prim Care (1999) 26:553–75.[Web of Science][Medline]
50. Eissenberg T, Balster RL. Initial tobacco use episodes in children and adolescents: current knowledge, future directions. Drug Alcohol Depend (2000) 59(Suppl. 1):S41–60.[CrossRef][Web of Science][Medline]
51. Flay BR, Hu FB, Richardson J. Psychosocial predictors of different stages of cigarette smoking among high school students. Prev Med (1998) 27:A9–18.[CrossRef][Web of Science][Medline]
52. Kobus K. Peers and adolescent smoking. Addiction (2003) 98(Suppl. 1):37–55.[CrossRef][Web of Science][Medline]
53. Mayhew K, Flay BR, Mott JA. Stages in the development of adolescent smoking. Drug Alcohol Depend (2000) 59(Suppl. 1):S61–81.[CrossRef][Web of Science][Medline]
54. Pletcher JR, Schwarz DF. Current concepts in adolescent smoking. Curr Opin Pediatr (2000) 12:444–9.[CrossRef][Web of Science][Medline]
55. Sasco AJ, Kleihues P. Why can't we convince the young not to smoke? Eur J Cancer (1999) 35:1933–40.[CrossRef][Web of Science][Medline]
56. Tyas SL, Pederson LL. Psychosocial factors related to adolescent smoking: a critical review of the literature. Tob Control (1998) 7:409–20.
57. Wagner EF, Atkins JH. Smoking among teenage girls. J Child Adolesc Subst Abuse (2000) 9:93–110.[CrossRef]
58. Wilcox P. An ecological approach to understanding youth smoking trajectories: problems and prospects. Addiction (2003) 98(Suppl. 1):57–77.[CrossRef][Web of Science][Medline]
59. D'Amico EJ, Metrik J, McCarthy DM, et al. Progression into and out of binge drinking among high school students. Psychol Addict Behav (2001) 15:341–9.[CrossRef][Web of Science][Medline]
60. Ellickson SL, Tucker JS, Klein DJ, et al. Prospective risk factors for alcohol use in late adolescence. J Stud Alcohol (2001) 62:773–82.[Web of Science][Medline]
61. Griffin KW, Scheier LM, Botvin GJ, et al. Ethnic and gender differences in psychosocial risk, protection, and adolescent alcohol use. Prev Sci (2000) 1:199–212.[CrossRef][Medline]
62. Lonczak HS, Huang B, Catalano RF, et al. The social predictors of adolescent alcohol misuse: a test of the Social Development Model. J Stud Alcohol (2001) 62:179–89.[Web of Science][Medline]
63. Poikolainen K, Tuulio-Henriksson A, Aalto-Setala T, et al. Predictors of alcohol intake and heavy drinking in early adulthood: a 5-year follow-up of 15-19-year-old Finnish adolescents. Alcohol Alcohol (2001) 36:85–8.
64. Scheier LM, Botvin GJ, Griffin KW, et al. Dynamic growth models of self-esteem and adolescent alcohol use. J Early Adolesc (2000) 20:178–209.[CrossRef]
65. Johnson PB, Johnson HL. Cultural and familial influences that maintain the negative meaning of alcohol. J Stud Alcohol (1999) 60(Suppl. 13):79–83.
66. Kodjo CM, Klein JD. Prevention and risk of adolescent substance abuse. The role of adolescents, families, and communities. Pediatr Clin North Am (2002) 49:257–68.[CrossRef][Web of Science][Medline]
67. Patton LH. Adolescent substance abuse: risk factors and protective factors. Pediatr Clin North Am (1995) 42:283–93.[Web of Science][Medline]
68. Schor EL. Adolescent alcohol use: social determinants and the case for early family-centered prevention. Family-focused prevention of adolescent drinking. Bull N Y Acad Med (1996) 73:335–56.[Web of Science][Medline]
69. Bachanas PJ, Morris MK, Lewis-Gess JK, et al. Psychological adjustment, substance use, HIV knowledge, and risky sexual behavior in at-risk minority females: developmental differences during adolescence. J Pediatr Psychol (2002) 27:373–84.
70. Ben-Zur H, Breznitz S, Wardi N, et al. Denial of HIV/AIDS and preventive behavior among Israeli adolescents. J Adolesc (2000) 23:157–74.[CrossRef][Web of Science][Medline]
71. Boyer CB, Shafer MA, Wibbelsman CJ, et al. Associations of sociodemographic, psychosocial, and behavioral factors with sexual risk and sexually transmitted diseases in teen clinic patients. J Adolesc Health (2000) 27:102–11.[CrossRef][Web of Science][Medline]
72. Colon RM, Wiatrek DE, Evans RI. The relationship between psychosocial factors and condom use among African-American adolescents. Adolescence (2000) 35:559–69.[Web of Science][Medline]
73. Crosby RA, DiClemente RJ, Wingood GM, et al. Correlates of unprotected vaginal sex among African American female adolescents. Arch Pediatr Adolesc Med (2000) 154:893–9.
74. Dilorio C, Dudley WN, Kelly M, et al. Social cognitive correlates of sexual experience and condom use among 13- through 15-year-old adolescents. J Adolesc Health (2001) 29:208–16.[CrossRef][Web of Science][Medline]
75. Gutierrez L, Oh HJ, Gillmore MR. Toward an understanding of (em)power(ment) for HIV/AIDS prevention with adolescent women. Sex Roles (2000) 42:581–611.[CrossRef][Web of Science]
76. Henderson M, Wight D, Raab G, et al. Heterosexual risk behavior among young teenagers in Scotland. J Adolesc (2002) 25:483–94.[CrossRef][Web of Science][Medline]
77. Hendrickx K, Lodewijckx E, Van Royen P, et al. Sexual behavior of second generation Moroccan immigrants balancing between traditional attitudes and safer sex. Patient Educ Couns (2002) 47:89–94.[CrossRef][Web of Science][Medline]
78. Rosengard C, Adler NE, Gurvey JE, et al. Protective role of health values in adolescents’ future intentions to use condoms. J Adolesc Health (2001) 29:200–7.[CrossRef][Web of Science][Medline]
79. Beckman LJ, Harvey SM. Factors affecting the consistent use of barrier methods of contraception. Obstet Gynecol (1996) 88(Suppl.):65S–71S.[CrossRef][Medline]
80. Gage AJ. Sexual activity and contraceptive use: the components of the decision-making process. Stud Fam Plann (1998) 29:154–66.[CrossRef][Web of Science][Medline]
81. Jemmott JB, Jemmott LS. HIV risk reduction behavioral interventions with heterosexual adolescents. AIDS (2000) 14(Suppl. 2):S40–52.[Medline]
82. Kirby D. Antecedents of adolescent initiation of sex, contraceptive use, and pregnancy. Am J Health Behav (2002) 26:473–85.[Web of Science][Medline]
83. Kotchick BA, Shaffer A, Forehand R. Adolescent sexual risk behavior: a multi-system perspective. Clin Psychol Rev (2001) 21:493–519.[CrossRef][Web of Science][Medline]
84. Rotheram-Borus MJ, Mahler KA, Rosario M. AIDS prevention with adolescents. AIDS Educ Prev (1995) 7:320–36.[Web of Science][Medline]
85. Whaley AL. Preventing the high-risk sexual behavior of adolescents: focus on HIV/AIDS transmission, unintended pregnancy, or both? J Adolesc Health (1999) 24:376–82.[CrossRef][Web of Science][Medline]
86. Backman DR, Haddad EH, Lee JW, et al. Psychosocial predictors of healthful dietary behavior in adolescents. J Nutr Educ Behav (2002) 34:184–93.[CrossRef][Web of Science][Medline]
87. Berg MC, Jonsson I, Conner MT. Understanding choice of milk and bread for breakfast among Swedish children aged 11–15 years: an application of the theory of planned behavior. Appetite (2000) 34:5–19.[CrossRef][Web of Science][Medline]
88. Berg MC, Jonsson I, Conner MT, et al. Relation between breakfast food choices and knowledge of dietary fat and fiber among Swedish schoolchildren. J Adolesc Health (2002) 31:199–207.[CrossRef][Web of Science][Medline]
89. Contento IR, Michela JL, Williams SS. Adolescent food choice criteria: role of weight and dieting status. Appetite (1995) 25:51–76.[CrossRef][Web of Science][Medline]
90. Croll JK, Neumark-Sztainer D, Story M. Healthy eating: what does it mean to adolescents? J Nutr Educ (2001) 33:193–8.[CrossRef][Web of Science][Medline]
91. De Bourdeaudhuij I, Van Oost P. Family members’ influence on decision making about food: differences in perception and relationship with healthy eating. Am J Health Promot (1998) 12:73–81.
92. Gillman MW, Frazier AL, Rockett HR, et al. Family dinner and diet quality among older children and adolescents. Arch Fam Med (2000) 9:235–40.
93. Masu R, Sallis JF, Berry CC, et al. The relationship between health beliefs and behaviors and dietary intake in early adolescence. J Am Diet Assoc (2002) 102:421–4.[CrossRef][Web of Science][Medline]
94. Neumark-Sztainer D, Story M, Resnick MD, et al. Correlates of inadequate fruit and vegetable consumption among adolescents. Prev Med (1996) 25:497–505.[CrossRef][Web of Science][Medline]
95. Neumark-Sztainer D, Story M, Perry C, et al. Factors influencing food choices of adolescents: findings from focus-group discussions with adolescents. J Am Diet Assoc (1999) 99:929–37.[CrossRef][Web of Science][Medline]
96. Neumark-Sztainer D, Hannan PJ, Story M, et al. Family meal patterns: associations with sociodemographic characteristics and improved dietary intake among adolescents. J Am Diet Assoc (2003) 103:317–22.[Web of Science][Medline]
97. O'dea JA. Why do kids eat healthful food? Perceived benefits of and barriers to healthful eating and physical activity among children and adolescents. J Am Diet Assoc (2003) 103:497–501.[Web of Science][Medline]
98. Pirouznia M. The association between nutrition knowledge and eating behavior in male and female adolescents in the US. Int J Food Sci Nutr (2001) 52:127–32.[CrossRef][Web of Science][Medline]
99. Roos EB, Hirvonen T, Mikkilä V, et al. Household educational level as a determinant of consumption of raw vegetables among male and female adolescents. Prev Med (2001) 33:282–91.[CrossRef][Web of Science][Medline]
100. Weber Cullen K, Bartholomew LK, Parcel GS, et al. Measuring stages of change for fruit and vegetable consumption in 9- to 12-year-old girls. J Behav Med (1998) 21:241–54.[CrossRef][Web of Science][Medline]
101. Woodward DR, Cumming FJ, Ball PJ, et al. Adolescents’ reported usage of selected foods in relation to their perceptions and social norms for those foods. Appetite (1996) 27:109–17.[CrossRef][Web of Science][Medline]
102. Birch LL, Fisher JO. Development of eating behaviors among children and adolescents. Pediatrics (1998) 101:539–49.
103. Eertmans A, Baeyens F, Wechsler H. Food likes and their relative importance in human eating behavior: review and preliminary suggestions for health promotion. Health Educ Res (2001) 16:443–56.
104. Koivisto Hursti UK. Factors influencing children's food choice. Ann Med (1999) 31:26–32.[Web of Science][Medline]
105. Story M, Neumark-Sztainer D, French S. Individual and environmental influences on adolescent eating behaviors. J Am Diet Assoc (2002) 102:S40–51.[CrossRef][Web of Science][Medline]
106. Battistisch V, Schaps E, Watson M, et al. Effects of the Child Development Project on students’ drug use and other problem behaviors. J Prim Prev (2000) 21:75–99.[CrossRef]
107. Hawkins JD, Catalano RF, Kosterman R, et al. Preventing adolescent health-risk behaviors by strengthening protection during childhood. Arch Pediatr Adolesc Med (1999) 153:226–34.
108. Perkins DN, Salomon G, Learning transfer, De Corte E, Weinert FE, eds. International Encyclopedia of Developmental and Instructional Psychology (1996) Oxford: Pergamon. 483–7.
109. Tuomi-Gröhn T, Engeström Y. Between School and Work: New Perspectives on Transfer and Boundary Crossing (2003) Amsterdam, The Netherlands: Pergamon.
110. Säljö R. Epilogue: from transfer to boundary-crossing. In: Between School and Work: New Perspectives on Transfer and Boundary Crossing—Tuomi-Gröhn T, Engeström Y, eds. (2003) Amsterdam, The Netherlands: Pergamon. 311–21.
111. Jackson N. Handbook systematic reviews of health promotion and public health interventions. (2005) Melbourne, Australia: Deakin University/Australian Department of Health and Aging/Cochrane Health Promotion and Public Health Field/Victorian Health Promotion Foundation.
Received on September 18, 2006; accepted on February 11, 2008
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||