Health Promotion International, Vol. 18, No. 3, 229-236,
September 2003
© Oxford University Press 2003
Structural relationship of factors affecting health promotion behaviors of Korean urban residents
1Department of Preventive Medicine, College of Medicine, Kangwon National University Chunchon, South Korea, 2Department of Preventive Medicine, College of Medicine, Ewha Womans University, Seoul, South Korea and 3Department of Preventive Medicine, Obstetrics and Gynaecology, College of Medicine, Pochon CHA University-CHA General Hospital, Seoul, South Korea
Address for correspondence: Sunhee Lee, MD, PhD, Associate Professor, Department of Preventive Medicine, College of Medicine, Ewha Womans University, 911–1 Mok-Dong, Yangcheon-Ku, Seoul, South Korea, E-mail: lsh0270{at}ewha.ac.kr
 |
SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Voluntary cooperation is crucial in determining the effectiveness of a health promotion strategy. The Triandis model, a theoretical model with the advantages of the Fishbein model, is useful in directly explaining health behavior. The purpose of this study was to perform a comprehensive evaluation of behavioral tendency related to health promotion and to examine the interrelationship amongst factors related to health behavior. This study uses the Triandis model to examine the structural relationship between explaining variables of health promotion. Five-hundred and nine residents were selected by multi-stage random sampling from the province of Kyungi-do, Korea. A structured questionnaire based on the Triandis model was developed. This questionnaire was distributed to the selected residents and their responses were collected via face-to-face interviews. The results showed that expectation of behavioral consequence, social norm and age variables were significantly related to health behavioral tendency. In addition, facilitating factors, affect variable and education level were indirectly related to health behavior. The results suggest that integrating behavioral subsets may be useful for understanding the mechanism of individual health promotion behavioral tendency.
Key words: health behavior tendency; interrelationship; Triandis model
|
INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Workers in the field of public health have long been interested in predicting and understanding health behavior (Gochman, 1997
). As behavioral risk factors of preventable morbidity and premature death have emerged more frequently, public health professionals have tried to reduce the negative factors and induce the positive ones affecting health behavior. Specifically, health promotion efforts have focused on developing strategies to maintain people’s attitudes toward health promotion.
In studying health promotion behavior, we considered two issues: first, how to analyze the targeted health behavior; and secondly, how to apply an appropriate theoretical model to explain the health behavior. To analyze the targeted health behavior, previous studies focused on an individual specific behavior, but most health promotion behaviors have a synergic relationship. Green and Kreuter (Green and Kreuter, 1991) pointed out that subsets of health behavior had a synergistic or multiplier effect on each other, and that an integrated approach is more useful in analyzing behavior patterns. Thus, health promotion practice must be analyzed using a holistic approach in order to understand behavior patterns more comprehensively.
With respect to the second issue concerning the appropriate theoretical model to explain health behavior, several models for explaining and understanding social behavior have been developed in the past decade (Gochman, 1997). The Fishbein model, based on the theory of reasoned action (TRA), is the representative model, which aims to explain individual adherence (Sheeran and Orbell, 1999). According to this model, a person’s intention to perform a behavior is a key predictor of behavioral performance (Fishbein and Ajzen, 1975; Ajzen and Fishbein, 1980). Furthermore, Lierman et al. (Lierman et al., 1990) have indicated that intention is a function of two basic components: a personal factor (attitude towards the behavior) and a social factor (the individual’s perceptions of the social influence and the motivation to comply with those influences—subjective norms).
Evidence indicates that attitudes and subjective norms are good predictors of intentions and, in turn, intentions are a reliable predictor of behavior (Sheppard et al., 1988). Despite these findings, several researchers have found that the Fishbein model does not directly link its components to actual behavior because the model focuses primarily on behavioral intention (Montano, 1986). Efforts to resolve these problems in the Fishbein model led to the development of the Triandis model. This model incorporates the factors that mediate the relationship between behavior intention and behavior by including a proxy variable instead of intention (Triandis, 1980). Additionally, two factors, such as habit and facilitating conditions, were included in the Fishbein model. Montano (Montano, 1986) reported that the Triandis model was superior to the Fishbein model in predicting behavior intention and behavior.
Understanding the behavior modification mechanism is important in inducing positive health behavior. Previous studies (Hillhouse et al., 2000; Empelen et al., 2001) focused mainly upon the simple relationship between independent and dependent variables, rarely investigating consecutive relationships among independent variables. In our study, we have used the Triandis model to evaluate the behavioral tendency comprehensively and to examine the relationships among factors related to health promotion.
|
METHODS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Theoretical framework
This study employed the Triandis model of the TRA as a theoretical framework. The Triandis model suggests that health behavior can be predicted from affected attitudes, subjective evaluations of a behavioral consequence, social norm as a social influence, past habits and facilitating factors that induce health promotion practice. In particular, unlike the Fishbein model which includes behavioral intention as an independent variable, the Triandis model treats behavioral intention as an endogenous variable explained by the three components, such as affective attitude, evaluations of behavioral consequence and social influences.
Maddux et al. have suggested that past habits can predict actual behavior, but only if the behavior is habitual (Maddux et al., 1995). The habit variable can be included or excluded depending on the characteristics of the relevant behavior. This research includes all of the main variables of the Triandis model, with the exception of the habit variable as behaviors vary. In addition, individual health status, although not included in the Triandis model, was accounted for in our research as well as sociodemographic characteristics, as both can influence health behavior. The final research framework is shown in Figure 1.
|
Measurement of instruments
An interview questionnaire was developed to measure the components of the Triandis model: behavior, affect, expectation of behavioral consequence, social norm, and facilitating factor (Table 1
). Our questionnaire was the modified questionnaire used in Montano’s study (Montano, 1986). After carrying out pilot surveys several times in the field, we revised the questions to reflect the thinking characteristic of local people.
|
Health promotion behavior
Health promotion behavior is defined as a person’s behavior to promote health status. Performance was measured according to the sum of four categories, on scales from 0 to 4. The categories were: physical exercise, healthy diet, regular health check-ups, and weight management. Physical exercise was scored from 0 to 3 according to how much the individual exercised. Healthy diet was measured by two scores; one that checked whether the individual had breakfast, and another that depended on regular eating habits. Both scores were taken from a scale of 0–3. A regular health check-up is a periodic preventive health examination in the absence of illness. The health examination score depended on the frequency of the individual’s check up in a year, and was evaluated on a scale from 0 to 4. Finally, weight management score was determined by whether the individual was under such a regimen. Weight management is defined as the effort to maintain an ideal body weight. These four categories are regarded as typical health promotion behaviors (U.S. DHSS, 1990
). Behavioral tendency for health promotion can be evaluated holistically by the integration of each behavior.
Affect
Affect means an emotional attitude and reaction towards the performance of health behavior. This component corresponds to the direct measure of attitude in the Fishbein model. The affect of the person was measured using three different categories: pleasant, worthless and wise.
Expectation of behavioral consequence
This component measures one’s expectation or belief about outcomes or characteristics of behavior (e.g. live longer, prevent serious disease). This measurement consists of two factors. The first factor, the belief that a behavior will result in an outcome, was scored on a 4-point Likert scale ranging from ‘extremely unlikely’ to ‘extremely likely’. The second factor, the subjective value of the consequence score, was scored from ‘extremely bad’ (1) to ‘extremely good’ (4). The attitude score was calculated by multiplying the belief value and the behavioral consequence value (belief x value of consequence).
Social norm
Social norm is defined as an individual’s perception of other’s expectation of their behavior. This variable consists of two factors. The first factor is the extent of an individual’s perception that a social influence group, such as family or spouse, want him or her to perform the health behavior. The second factor is the degree of one’s motivation to comply with the expectations of such groups. The score of social norm is calculated by multiplying the extent of the individual’s expectation and the degree of his or her motivation.
Facilitating factors
This component measures the individual or environmental characteristics that facilitate or hinder an individual in executing a behavioral intention. Statements on the questionnaire in this category are shown in Table 1
. Each statement was responded to with a score on a scale of 1–4.
Sociodemographic factors
The sociodemographic factors investigated using the questionnaire were age, gender, education, monthly income and occupation.
Data collection and analysis
The study population lived in a mid-size city in Kyung-gi Province, Republic of Korea. Ages ranged from 20 to 59 years. A multi-stage cluster random sampling method was applied to the entire population of the city (218 662 residents). The city was divided into four areas based on gender, age and household economic level, and a cluster was randomly assigned. Finally, 572 persons were selected to participate in the survey. The survey began on 15 January 1999 and lasted for 1 week. Trained interviewers interviewed the selected persons individually and face-to-face using our questionnaire. These surveys were verified in the field. Excluding incomplete or insincere questionnaires, we analyzed 509 complete questionnaires.
Statistical analysis was performed by SPSS (Statistical Package for the Social Sciences). We performed univariate analysis to examine the simple relationship between the above mentioned components and health behavior. A covariance structure analysis was applied to investigate the causal relationship among components. The covariance structure analysis with the LISREL (linear structural relationship) program version 7.13, widely used in pedagogy, psychology and public health, was used to combine different measurement variables into one theoretical variable (e.g. behavior intention, affect, etc.).
|
RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sociodemographic characteristics of the respondents
Table 2
presents the sociodemographic characteristics of the respondents. Of the respondents, 49.3% were men. Approximately 26.3% of respondents were in their thirties, 19.4% in their forties, and 9.4% in their fifties. The percentage of college or graduate school graduates was 47.7%, while 44.2% were high school graduates. In terms of income, 36.7% of respondents earned ~1.5–2.0 million Won per month (approximately US$1250–1600), and 30.3% had greater monthly income. Forty-one percent were housewives, 26.1% were white-collar workers and 15.9% were self-employed.
|
Level of health promotion practice
Table 3
shows the level of health promotion practice, such as physical exercise, healthy diet, health check-ups and weight management, by Korean urban residents. Both men and women showed low practice levels with respect to physical exercise and weight management, although men tended to exercise more than women, and women tended to manage their weight more than men (both were statistically significant). The men also had more health check-ups, and 60.2% of men received regular check-ups.
|
Comparing health promotion practice according to age group, the older group showed significantly higher levels of physical activity, health check-ups and weight management. Comparing by education level, the higher education group showed greater levels of practice than the other groups, but not with a statistical significance. The group with a medical history had more regular health check-ups, but this group showed no statistically significant difference from the remaining groups in other categories.
Level of the TRA model component
Sociodemographic characteristics such as gender, education level, monthly income level or present health status were not significantly related to TRA model components, indicating the related factors of health behavior (Table 4). However, the older groups and those with medical history showed higher levels of facilitating factors, indicating that they were more motivated to practice health promotion.
|
Relationships among TRA model components
Pearson correlation coefficients among TRA model components are shown in Figure 2
. All correlations were significant in a positive direction. In particular, the expectations of behavior consequence and social norm showed strong correlation, but affect and facilitating factors showed weak correlation.
|
Structural analysis of the factors related to health promotion behavior
Using the LISREL analysis, we analyzed the relationship among the Triandis model components and sociodemographic factors (Figure 3
). The model in this analysis had an excellent fit score [goodness of fit index (GFI) = 0.993]. The analysis revealed that the expectations of a behavioral consequence, social norm and age were significant variables that directly influenced health promotion practice. The facilitating factors influenced the affect and expectations of behavioral consequence significantly. The affect had an indirect effect on health promotion behavior by influencing the expectations of behavioral consequence and social norm. In general, men held more positive affective attitudes. Level of education positively affected the social norm, indirectly influencing health promotion behavior.
|
|
DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study was designed to investigate the information required to change an individual’s attitude towards health promotion. The results of this study show that the Triandis model can be an effective frame for understanding health behavior.
Univariate analysis of sociodemographic characteristics showed a gender-dependent attitude toward health promotion. It is mandatory that Korean employees undergo regular medical examinations. It is likely that men have many more health examinations than women as they hold more of the jobs. Women, however, were found to be far more interested in weight management, under the influence of a social norm in which a thin image is well regarded. The older age group participated in physical exercise, dietary habits and health check-up programs, demonstrating a more positive attitude toward health promotion practices.
Covariance structure analysis revealed that the level of education had a direct effect on social norm and an indirect effect on health promotion practice. This result confirms that education has a positive effect on health concerns, thus indirectly impacting on health behavior. The group with medical history showed a higher level of health check-up and a positive relationship with facilitating factors. Previous health status must be considered when explaining health promotion behavior. An individualized health promotion strategy can be recommended according to personal health status.
In the Pearson correlation analysis, four components of the Triandis model were correlated with statistical significance, suggesting synergic interaction. In other words, a more positive attitude toward health promotion behavior leads to greater expectations of the behavior consequence, as well as to motivation to comply with social influence. A proper health environment meeting individual needs makes it easier to practice health behavior.
The covariance structure analysis showed causal relationships among the components. The more facilitating factors one has, such as being surrounded by people who practice health promotion and being in frequent contact with health promotion information, the more positive affect one has about health promotion. At the same time, the expectation of behavior consequence improves. Additionally, the positive emotional attitude improves expectation of behavioral consequence and leads to a more positive response to social influence. Recognizing the effects or benefits of health promotion is essential to practicing health promotion. Therefore, there should be more health education and various public relation programs to emphasize the benefits of health promotion. Health education programs for family and friends are recommended. A health promotion program for families should consider the family as a primary unit.
The habit variable, which is included in the Triandis model, was excluded from this study; the reason for this was that in our study, a dependent variable was a subset of health behavior and it was difficult to consider each habitual behavior separately. Sheeran and Orbell (Sheeran and Orbell, 1999) pointed out that past behavior has a different influence on health promotion behavior by depending upon various characteristics of behavior intention or model components. The habit variable must therefore be included or excluded when considerating characteristics of behavior.
Although this investigation has limitations because it is a cross-sectional study, which has difficulty showing the causality between relating factors and health promotion practice, it effectively uses the Triandis model to understand health behavior tendency. We also found that various facilitating factors had an indirect influence on behavioral performance by cognitive variables (e.g. affect and expectation). Our results show the relationships among the components of TRA, including facilitating factor. More detailed mechanisms or interrelationships of the behavior modification process should be explored in the future.
|
REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ajzen, I. and Fishbein, M. (1980) Understanding Attitude and Predicting Social Behavior. Prentice-Hall, Englewood Cliffs, NJ.
Empelen, P. V., Schaalma, H. P., Kok, G. and Jansen, M. W. (2001) Predicting condom use with casual steady sex partners among drug users. Health Education Research, 16, 293–205.
Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Addison-Wesley, Boston, MA.
Gillmore, M. R., Morrison, D. M., Lowery, C. and Baker, S. A. (1994) Belief about condoms and their association with intentions to use condoms among youth in detection. Journal of Adolescent Health, 15, 228–237.[CrossRef][ISI][Medline]
Gochman, D. S. (ed.) (1997) Handbook of Health Behavior Research I. Plenum Press, NewYork, NY.
Green, L. W. and Kreuter, M. W. (1991) Health Promotion Planning—An Educational and Environmental Approach, 2nd edition. Mayfield Publishing Company, London.
Hillhouse, J. J., Turrisi, R. and Kastner, M. (2000) Modelling tanning salon behavioral tendencies using apperarance motivation, self-monitoring and the theory of planned behavior. Health Education Research, 15, 405–414.
Jemmott, J. B., Jemmott, L. S. and Fong, G. T. (1992) Reductions in HIV risk-associated sexual behaviors among black male adolescents: effects of an AIDS prevention intervention. American Journal of Public Health, 82, 372–377.
Leone, L., Perugini, M. and Ercolani, A. P. (1999) A comparison of three models of attitude-behavior relationships in the studying behavior domain. European Journal of Social Psychology, 29, 161–189.[CrossRef]
Lierman, L. M., Kasprzyk, D., Young, H. and Powell-Cope, G. (1990) Analysis of Theoretical Models. Paper presented at the Western Society for Research in Nursing Conference, San Diego, CA.
Maddux, J. E., Brawleyl, L. and Boykinm, A. (1995) Self-efficacy and healthy decision-making: protection, promotion, and detection. In Maddux, J. E. (ed.) Self-efficacy, Adaptation, and Adjustment: Theory, Research, and Application. Plenum Press, New York, NY.
Montano, D. E. (1986) Predicting and understanding influenza vaccination behavior: alternatives to the health belief model. Medical Care, 24, 438–453.[CrossRef][ISI][Medline]
Montano, D. E. and Taplin, S. H. (1991) A test of an expanded theory of reasoned action to predict mammography participation. Social Science and Medicine, 32, 733–741.
Sheeran, P. and Orbell, S. (1999) Does the temporal stability of behavioral intentions moderate intention-behavior and past behavior-future behavior relation? Personality and Social Psychology Bulletin, 25, 724–734.
Sheppard, B. H., Hartwick, J. and Warshaw, P. R. (1988) The theory of reasoned action: A meta-analysis of past research with recommendations for modifications and future research. Journal of Consumer Research, 15, 325–343.[CrossRef]
Triandis, H. C. (1980) Values, attitude, and interpersonal behavior. In Howe, H. E. (ed.) Nebraska Symposium on Motivation. University of Nebraska Press, Lincoln, NE.
U.S. Department of Health and Human Services (DHHS) (1990) Healthy People 2000: National Health Promotion and Disease Prevention Objectives. U.S. Government Printing Offices, Washington DC.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Engelhardt, B. C. Ahn, S.-I. Cho, and H. Joung Predictors of interest in nutrition topics and willingness to participate in local nutrition programmes J. Public Health Med., March 1, 2007; 29(1): 9 - 12. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||