Health Promotion International, Vol. 19, No. 1, 21-31, March 2004
© Oxford University Press 2004. All rights reserved
The Dutch Heart Health community intervention ‘Hartslag Limburg’: results of an effect study at individual level
1Department of Health Education and Promotion, 2Department of Methodology and Statistics, 4Department of General Practice and 5Department of Health Education and Promotion, Maastricht University, Maastricht, The Netherlands, 3Department of Public Health of the Regional Public Health Institute Maastricht, Maastricht, The Netherlands and 6Department of Public Health, Erasmus MC Rotterdam, Rotterdam, The Netherlands
Address for correspondence: Gaby Ronda, Department of Health Education and Promotion, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands, E-mail: g.ronda{at}gvo.unimaas.nl
 |
SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
‘Hartslag Limburg’ (Dutch for Heartbeat Limburg), a regional cardiovascular diseases (CVD) prevention program, integrates a community strategy and a high-risk strategy to reduce CVD risk behaviors. The present paper focuses on the effects of the community intervention on fat intake and physical activity. The project was based on community organization principles and health education theories and methods. In order to implement the intervention, nine local Health Committees were set up, each organizing activities that facilitate and encourage people to adopt a healthier lifestyle. A pre-test–post-test control group design with two post-tests was used to evaluate the intervention. At baseline, representative random cohort research samples were selected in the Maastricht region and in a control region. Data on fat intake and physical activity, and on the psychosocial determinants of these behaviors, were gathered by means of mail surveys. The present study indicates that the intervention had a significant effect on fat reduction, especially among respondents aged
48 years (median age). Respondents in the Maastricht region were also more realistic about their fat intake at post-test as compared with the control region. Only a limited effect on intentions to increase physical activity was found. Key words: cardiovascular diseases; community-based prevention; evaluation
|
INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
In 1998, a regional cardiovascular diseases (CVD) prevention program called ‘Hartslag Limburg’ (Dutch for Heartbeat Limburg), integrating a community strategy and a high-risk strategy, was started in the Maastricht region of the province of Limburg. Hartslag Limburg is a joint project of the municipal authorities, the Regional Public Health Institute (RPHI), social work organizations, the community health care organization, general practitioners, the University and University Hospital, and various local organizations, clubs and companies in the Maastricht region. The project aims to establish an organizational structure in which the collaboration between the main partners is given a formal and permanent basis, to ensure long-term attention to the prevention of CVD. In January 2001, the World Health Organization (WHO) selected Hartslag Limburg as one of 12 so-called ‘field projects’, based on its potential to meet pre-established criteria of the WHO project ‘Towards Unity for Health’ (Boelen, 2001
). The major goal of the community intervention is to reduce CVD risk among the 180 000 inhabitants of the region by encouraging behavioral change, i.e. dietary (saturated) fat reduction, increased physical activity and smoking cessation. The project continues to date.
The present article presents the methods and results of the effect study of the community intervention on individual behaviors, i.e. the effects on dietary fat intake and physical activity, as well as the effects on the psychosocial determinants of these behaviors. A description of the full intervention has been published elsewhere (Ruland et al., 1999).
The community intervention program and its evaluation study
The conceptual framework of Hartslag Limburg's community project is based on program planning and evaluation models, and indicates how the program is expected to produce its postulated effects (Table 1) [e.g. (Green and Kreuter, 1999)]. The framework serves as the program planning as well as evaluation framework (Ronda et al., 2003). The main community organization principles included in the project are participation of the community in the planning and implementation of the project, intersectoral collaboration between local organizations, link-up with the current situation (e.g. supporting existing CVD health-promoting initiatives), a social network approach, an environmental strategy (e.g. providing healthier food choices in cafeterias), and a multi-media and multi-method strategy. The specific intervention activities are based on different theoretical concepts, i.e. the Precaution Adoption Model (Weinstein, 1988) and the Attitude–Social influence–Self-efficacy (ASE) model (De Vries et al., 1988; Ajzen, 1991).
|
In order to implement the intervention, nine local Health Committees were set up: one in each of the four smaller municipalities, one in each of four underprivileged Maastricht neighborhoods, and one to coordinate and implement regional activities. Each Health Committee organizes, within its district or municipality, activities that facilitate and encourage people to adopt a healthier lifestyle. Prior to the official start of the project, in the spring of 1998, community analyses were conducted to introduce the project in the communities, to achieve early community involvement, and to assess which people, organizations and community sectors should be approached for participation in the Health Committees. Throughout the rest of 1998, the Health Committees were set up. Each Health Committee consists of
10 members, who represent local organizations that may play a key role in healthy behavior-promoting activities, such as health organizations, women's associations and associations of the elderly. The Health Committees meet 10 times a year.
The community intervention was officially started in October 1998, with a regional campaign to promote physical activity among individuals >55 years of age (NOC*NSF, 1999). At about the same time, the Health Committees officially started their activities. In order to facilitate the work of the Health Committees, the RPHI encourages and assists them in organizing as many healthy behavior-promoting activities as possible, which have been carefully selected on the basis of earlier studies (Ronda and Van Assema, 1997). Moreover, each Committee is supported by a health educator, a social worker and a civil servant of the municipality. At 32 months into the project, 293 activities had been registered. One hundred and sixty-six of these activities concerned nutrition, 84 physical activity and 15 smoking, and 28 activities were more general and targeted more than one risk behavior. Specific activities were tested in controlled preliminary studies. Examples of such (ongoing) activities include computer-tailored nutrition education (Brug et al., 1998), nutrition tours in supermarkets (Van Assema et al., 1998), a regional daily television program ‘Heartbeat on the Move’ to promote physical activity (Ronda et al., 2001a), walking and cycling months (Ruland et al., 2001), a regional smoking cessation campaign (Ruland et al., 2001), and a non-smoking campaign for the parents of children in playgroups (Ruland et al., 2001). In addition, there are ongoing activities trying to draw attention to the project and its specific activities, such as commercials on local television and radio, newspaper articles, and posters and pamphlets.
Evaluation measures were developed for all stages of the intervention and evaluation framework (Table 1), except for stage A (health), since no detectable effects at this level can be expected within a limited number of years. The evaluation study consisted of an effect study at individual and organizational levels, and a process study.
The research questions for the present paper focused on the individual level and were: (i) has there been a positive change in dietary fat intake and physical activity levels in the Maastricht region compared with the control region?; and (ii) has there been a positive change in the psychosocial determinants related to decreasing dietary fat intake and increasing physical activity in the Maastricht region compared with the control region?
|
METHODS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Design and sample
To assess differences between the Maastricht and control region (where there was no community intervention), a pre-test–post-test control group design was used, with two post-tests (Cook and Campbell, 1979
). The baseline measurement was conducted in May 1998, the first post-test in May 2000 and the second post-test in May 2001. At baseline, representative random cohort research samples were selected in both regions. In the Maastricht region, a stratified random sample of 1450 inhabitants (aged 
14 years), based on the number of inhabitants in each of the five municipalities, was selected from the population registers of each of the five municipalities. A stratified random sample of 1200 inhabitants (aged 14 years), based on the number of inhabitants in each municipality, was selected from the population registers of six municipalities of a comparable control region. These sample sizes were based on a power of 0.95 to detect a priori hypothesized effect sizes at a significance level of 0.05. The control region was comparable with respect to the incidence and prevalence of CVD, number of inhabitants, number of municipalities and degree of urbanization.
The questionnaire
Data on dietary fat intake and physical activity level, and on the determinants of these behaviors, were gathered by means of mail surveys. Non-responders were reminded once at baseline and twice at follow-up.
Risk behaviors
Dietary fat intake was assessed using a validated questionnaire covering 19 (groups of) food products that contribute most to the consumption of saturated fat in The Netherlands. A fat consumption score ranging from 0 to 80 points was calculated (Van Assema et al., 2001). Fat scores of 15 points for women and 18 points for men are the approximate upper levels of recommended dietary intake. Physical activity levels were assessed with a validated short questionnaire (Koremans et al., 2000; Ronda et al., 2001b) that allows a physical activity score to be calculated in minutes and in sessions per week. In addition, respondents were asked to indicate directly on how many days a week they were physically active for at least 30 min. Respondents who had a physical activity score of at least 150 min per week and at least five sessions per week, and who indicated that they were physically active for at least 30 min a day on at least 5 days a week, were classified as meeting the recommended target.
Psychosocial determinants
Important determinants of the risk behaviors, i.e. attitudes, social support, self-efficacy expectations and intentions related to decreasing dietary fat intake and increasing physical activity, as well as self-rated dietary fat intake and physical activity level were assessed, mostly with single items, in order to assess intermediate intervention effects (De Vries et al., 1988; Weinstein, 1988; Ajzen, 1991; Brug et al., 1994; Ronda et al., 2001b). Respondents were asked to evaluate their attitudes towards the different behaviors on a ‘bad–good’ scale, and for physical activity also on an ‘unpleasant–pleasant’ scale (e.g. how good or bad do you think it would be for you to be more physically active/eat less fat). Afterwards, one score for attitude towards increasing physical activity was calculated by computing the mean score of the two items (r = 0.52). Furthermore, the questionnaire asked respondents if they intended to change each of the two behaviors in the future (intention), how confident they were about their ability to make such changes (self-efficacy), and if they experienced support from important others to change the behaviors (directly perceived social support). All items, with the exception of perceived social support (yes/no), could be answered on bipolar five-point scales.
Self-rated dietary fat intake and self-rated physical activity were assessed by asking respondents whether they rated their own fat intake and physical activity level as low or high (on a bipolar five-point scale). Respondents whose fat consumption levels were low, but who rated them (rather) high were classified as unaware of their intake level. Respondents whose fat consumption levels were not low, but who rated them as intermediate or low, were also classified as unaware of their intake level. The remaining respondents were classified as realistic. A similar procedure was used to classify respondents on awareness of physical activity. Furthermore, the first and second post-test included a question measuring respondents' familiarity with a health project in their municipality. Finally, respondents were questioned about their sex, age and education.
Statistical analysis
Descriptive statistics (frequencies and means) were used to describe demographic characteristics, dietary fat consumption, physical activity level, and familiarity with a health project. Multiple logistic regression analyses were conducted to identify potential dropout bias (with attendance versus dropout as the dependent variable, and the baseline values for gender, age, education, fat consumption, physical activity level and condition as independent variables) and potential baseline differences between the Maastricht region and the control region (with study region as the dependent variable and the baseline values for gender, age, education, fat consumption, and physical activity level as independent variables). Only respondents who completed all surveys were included. These preliminary analyses were performed using the SPSS 10.0 statistical package (SPSS, 2000). Variables identified as being statistically different (p < 0.05) between respondents who participated in all three surveys and dropouts, and between respondents from the Maastricht region and the control region, were included as independent variables in subsequent regression analyses of differences between the Maastricht region and the control region.
To study differences in risk behaviors and their psychosocial determinants between the Maastricht and the control region, multi-level regression analysis was used for all outcome variables (continuous, dichotomous and ordinal) using the MlwiN (Rasbash et al., 1999) statistical package and the MIXOR statistical package (Hedeker and Gibbons, 1996). Multi-level regression analysis was used to take into account possible dependencies among individuals within the same municipality (Hedeker et al., 1994; Snijders and Bosker, 1999). In the cases in which multi-level analysis revealed that the individuals within municipalities could be regarded as independent (i.e. all cases, except with fat scores as the dependent variable), analyses were repeated using ‘ordinary’ regression within the SPSS statistical package. Results in SPSS (SPSS, 2000) were consistent with the results found with ‘one level’ models in MlwiN (Rasbash et al., 1999) and MIXOR (Hedeker and Gibbons, 1996).
The main independent variable included in all analyses was condition (Maastricht region versus the control region). Furthermore, several other independent variables were included, on the basis of their known or suspected relation with the outcome variables. First, we controlled for the pre-intervention scores on the relevant effect indicators. Furthermore, gender, age, education and familiarity with a health project, and the inter-action terms of these variables with condition, were included as independent variables in all analyses. Also, the intention to change behavior at baseline, self-rated behavior at baseline, and their interaction terms with condition were included as independent variables in the analyses of behavioral change. In the cases for which significant (p < 0.05) interactions were found, subsequent ‘simple effects analyses’, i.e. subgroup analyses, were done. For subgroup analyses concerning age, this variable was dichotomized at the median (48 years).
All analyses were done in a so-called ‘top down’ procedure, in which all variables were first included in the model, after which non-significant variables (with the exception of condition) were excluded from the model in a hierarchical and stepwise procedure.
All of the analyses used a two-sided significance level of 5% (with the exception of random cluster effects, for which a one-sided significance level of 5% was used, since these random parameters are 0 or > 0).
|
RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Respondents
Response to the baseline survey was 55.5% [n = 1444; 820 (57.5%) in the Maastricht region and 624 (52.9%) in the control region (46 questionnaires were undeliverable)]. Fifteen cases were excluded from analysis because of incomplete data, leaving 809 respondents in the experimental region and 620 in the control region. Attrition rate from baseline (T0) to second post-test (T2) was 37.2% in the total group. Because of 92 undeliverable questionnaires in 2000 and 28 undeliverable questionnaires in 2001, the net attrition rate from T0 to T2 was 31.5%. Overall, 897 respondents completed all three questionnaires: 505 in the experimental region and 392 in the control region. Attrition did not differ between the two regions. However, dropouts were found to be younger and less educated than those who participated in all three measurements. Table 2 shows baseline demographic characteristics, dietary fat intake and physical activity level at baseline, first and second post-test, and familiarity with a health project at first and second post-test.
|
Dietary fat intake and psychosocial determinants
Table 3 shows the parameter estimates for dietary fat intake and its determinants at the first (T1) and second (T2) post-test. There was an overall condition effect on the fat score at T1. Respondents in the Maastricht region had a lower fat score at T1 than respondents in the control region. Furthermore, there was an interaction effect between age and condition on the fat score at T1. Further exploration of this interaction showed that the overall condition effect on the fat score was fully attributable to the younger half (<49 years) of the population studied. Respondents younger than 49 years in the Maastricht region had a lower fat score than respondents of that age in the control region (B = -1.207; p = 0.003). A similar interaction effect between age and condition on the fat score was detected at T2, and again respondents younger than 49 years in the Maastricht region had a lower fat score than respondents of the same age in the control region (B = –0.992; p = 0.026). A further interaction effect between familiarity with a health project and condition on the fat score was present at T2. Respondents who were familiar with a health project in the Maastricht region had a lower fat score than respondents who were familiar with a health project in the control region, but this effect was not statistically significant (p = 0.078).
|
Respondents who were familiar with a health project in their community reported more social support towards decreasing their fat intake at T2 than those who were not familiar with such a health project [odds ratio (OR) = 1.463; p = 0.045]. Besides, there was an overall condition effect on self-efficacy in the opposite direction at T2. Respondents in the control condition had higher self-efficacy expectations towards decreasing their fat intake than respondents in the Maastricht region (OR = 1.31; p = 0.037). There was also an overall condition effect on awareness of one's own fat intake at T1. Respondents in the Maastricht region were more realistic about their fat intake at T1 than respondents in the control region (OR = 1.401; p = 0.034). Finally, there was an interaction effect between gender and condition at T1 on awareness of one's own fat intake. Further exploration of this interaction effect showed that the overall condition effect on awareness of one's own fat intake was fully attributable to the women in the population studied. Women in the Maastricht region were more realistic about their own fat intake at T1 than women in the control region (OR = 1.869; p = 0.003).
Physical activity and psychosocial determinants
Table 4 shows the parameter estimates for physical activity and its determinants. There were no overall condition effects on physical activity and its determinants. At T1, the attitude (OR = 1.397; p = 0.029) and self-efficacy (OR = 1.742; p = 0.001) towards increasing physical activity were more positive among respondents who were familiar with a health project in their community than among those who were not familiar with such a health project. Furthermore, there was an interaction effect between age and condition on the intention to increase physical activity in the future at T1. The intention towards increasing physical activity level was more positive among respondents younger than 49 years in the Maastricht region than among respondents of the same age in the control region (OR = 1.543; p = 0.015). There was also an interaction at T2 between familiarity with a health project and condition on awareness of one's own physical activity level. Respondents who were familiar with a health project in the Maastricht region were more realistic about their physical activity level than respondents who were familiar with a health project in the control region, but this effect was not statistically significant (p = 0.056).
|
|
DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
The present study found significant and thus promising community intervention effects on dietary fat intake and physical activity intentions, especially among respondents younger than 49 years. Respondents in the Maastricht region were also more realistic about their fat intake. Lower self-efficacy expectations towards decreasing their fat intake were found in the Maastricht region, but this may not be incompatible with the decline in fat intake, because self-efficacy was measured in relation to behavior change (i.e. ‘eating less fat’) and not in relation to present behavior (i.e. ‘eating a low fat diet’).
As we mentioned in the Introduction, most of the registered activities in the Maastricht region were aimed at nutrition, and this might be an explanation for the positive results on dietary fat intake. Furthermore, preliminary analysis of the process evaluation data reveal, for example, that there is a growing participation of local organizations and community members, and intersectoral collaboration between local organizations in the Maastricht region, so longer-term measures of individual behaviors may show greater changes.
It is difficult to compare the effects of the Hartslag Limburg community intervention with those observed in other Dutch or foreign community projects, because effect measurements differed between projects, different behaviors were studied, and effects were measured over different time periods. Nevertheless, some comparisons can be made. Like the present study, the most recent community-based CVD prevention programs found few significant program effects [e.g. (Winkleby et al., 1997)]. Some earlier projects studying the same behaviors as the present study, however, reported some positive effects on dietary behavior or reported an increased prevalence of regular physical activity [e.g. (Van Assema et al., 1994; Browson et al., 1996; Tudor-Smith et al., 1998)]. Possible explanations frequently mentioned for the modest or absent intervention effects in community CVD prevention programs are: secular trends, i.e. it might be difficult to generate enough additional exposure in experimental communities to exceed secular trends in control communities; smaller than expected net differences in behavior change; ineffective interventions; and inaccurate survey data and/or measures [e.g. (Carleton et al., 1995; Browson et al., 1996; Winkleby et al., 1997)]. Although the present study found that the familiarity with a health project in the Maastricht region exceeded that in the control region, these data also suggest a substantial perceived exposure to health promotion interventions in the control region. Furthermore, preliminary analysis of the process evaluation data of the Hartslag Limburg Community Intervention revealed that although familiarity with nutritional as well as physical activity and smoking cessation activities was higher in the Maastricht region, actual participation in physical activity and smoking cessation activities was higher in the control region. Unfortunately, baseline values with regard to familiarity and participation are unknown. Also, the net differences in behavior change found in the present study were smaller than expected. Although there are several indications of effective individual components of the intervention program (Brug et al., 1998; Van Assema et al., 1998; Ronda et al., 2001a), process data also showed that actual participation of the general population in activities did not exceed 15%, indicating that the intervention did not reach a large enough fraction of the population to show greater changes. Finally, the possibility of inaccurate survey data and/or measures can not be neglected, and therefore it is important to note some limitations of the present study. First, the response at baseline was 55.5%. This is certainly not optimal, although regarded as quite acceptable for a mail survey (Miller, 1994). Nevertheless, the response rate in the present study may indicate low external validity because the representativeness of the sample is not assured (Windsor et al., 1994). Compared with the general Dutch population, more highly educated respondents were somewhat (5%) over-represented in the Maastricht region, while low education levels were somewhat (7%) over-represented in the control region (Statistics Netherlands, 1997). Secondly, there was a substantial dropout in the sample from baseline to endpoint (31.5%), which is an intrinsic feature of the longitudinal nature of the study. Dropouts were found to be younger and less educated than those who remained in the sample. However, our analyses controlled for potential confounding or interaction with condition by the above-mentioned factors. A further disadvantage of the cohort design is repeated questioning, implying the possibility that the survey itself may have an effect on the behavior of participants, especially when there is differential re-examination bias in the intervention and control condition (Salonen et al., 1986; Koepsell et al., 1992). In addition, the results are based on self-reports. Although the relative validity and reliability of the instruments used in the present study have been found to be better than or comparable to those used in other studies (Block et al., 1989; Pols et al., 1996), self-reports are often biased. Respondents in the intervention condition in particular may be inclined to over-report desirable health behaviors and under-report undesirable behaviors (Koepsell et al., 1992; Windsor et al., 1994). To reduce this possible differential bias, questions about health-promoting activities were included in the questionnaire in both regions, suggesting to both intervention and control participants that they had been included in the intervention condition.
A further measurement issue is the use of single-item assessments for the determinants of the CVD risk behaviors, since single-item assessments may be not sensitive enough to detect small intervention effects. However, practical limitations (survey questionnaire size) made us decide to use these single items for assessment of determinants. Secondary analysis showed that these assessments had significant and quite strong associations with intentions, which is a clear indication of the predictive validity of the assessments of the psychosocial factors.
The study included only one intervention and one control region, and allocation to either was not random. Cost and feasibility considerations usually limit the intervention and evaluation to a small number of communities, and allocation to treatment groups is often done beforehand by funding agencies or the communities themselves (Koepsell et al., 1992). Most community interventions have used the ‘one-group-per-condition’ design and this design is well accepted in community intervention research [e.g. (Carleton et al., 1995; Tudor-Smith et al., 1998; Vartiainen et al., 1998)]. In an attempt to overcome this design weakness to some extent, the pre-intervention score of the outcome variables, and baseline differences between the Maastricht and control region were included as independent variables in the effect analyses.
The ability to adjust for baseline differences is one of the major advantages of the cohort design (Salonen et al., 1986). Furthermore, multi-level analysis was used to take into account possible dependencies among individuals from the same municipality.
|
CONCLUSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Although there were few significant intervention effects, the decrease in dietary fat intake over time in the Maastricht region was a promising result. However, the results of the present study should be seen against the background of the results of the process evaluation and the results of the effect evaluation at the organizational level, to allow the effectiveness of the community intervention as a whole to be judged.
|
ACKNOWLEDGEMENTS |
|---|
This study was supported financially by the Netherlands Heart Foundation.
|
REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Ajzen, I. (1991) The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.[CrossRef][ISI]
Block, G., Clifford, C., Naughton, M. D., Henderson, M. and McAdams, M. (1989) A brief dietary screen for dietary fat intake. Journal of Nutrition Education, 21, 199–207.
Boelen, C. (2001) Towards Unity for Health: status report. Towards Unity for Health, 3, 5–6. WHO, Geneva.
Brownson, R. C., Smith, C. A., Pratt, M., Mack, N. E., Jackson-Thompson, J., Dean, C. G. et al. (1996) Preventing cardiovascular disease through community-based risk reduction: the Bootheel Heart Health Project. American Journal of Public Health, 86, 206–213.
Brug, J., Van Assema, P., Kok, G., Lenderink, T. and Glanz, K. (1994) Self-rated dietary fat intake: association with objective assessment of fat, psychosocial factors, and intention to change. Journal of Nutrition Education, 26, 218–223.
Brug, J., Glanz, K., Van Assema, P., Kok, G. and Van Breukelen, G. (1998) The impact of computer-tailored feedback on fat, fruit, and vegetable intake. Health Education and Behavior, 25, 357–371.
Carleton, R. A., Lasater, T. M., Assaf, A. R., Feldman, H. A. and McKinlay, S. (1995) The Pawtucket Heart Health Program: community changes in cardiovascular risk factors and projected disease risk. American Journal of Public Health, 85, 777–785.
Cook, T. D. and Campbell, D. T. (1979) Quasi-experimentation: Design and Analysis Issues for Field Settings. Rand McNally, Chicago, IL.
De Vries, H., Dijkstra, M. and Kuhlman, P. (1988) Self-efficacy: the third factor besides attitude and subjective norm as a predictor of behavioral intentions. Health Education Research, 3, 273–282.
Green, L. W. and Kreuter, M. W. (1999) Health Promotion Planning: An Educational and Ecological Approach. Mayfield, Mountain View, CA.
Hedeker, D. and Gibbons, R. D. (1996) A computer program for mixed-effects ordinal regression analysis. Computer Methods and Programs in Biomedicine, 49, 157–176.[CrossRef][ISI][Medline]
Hedeker, D., McMahon, S. D. and Jason, L. A. (1994) Analysis of clustered data in community psychology: with an example from a worksite smoking cessation program. American Journal of Community Psychology, 22, 595–615.[CrossRef][ISI][Medline]
Koepsell, T. D., Wagner, E. H., Cheadle, A. C., Patrick, D. L., Martin, D. C., Diehl, P. H. et al. (1992) Selected methodological issues in evaluating community-based health promotion and disease prevention programs. Annual Reviews Public Health, 13, 31–57.
Koremans, G. M., Derks, H., Tuinstra, A., Diederiks, M. and Brug, J. (2000) A quick method to assess physical activity? [Een snelle methode om beweging te meten?]. Humanities Working Papers, 5, 3–13.
Miller, T. I. (1994) Designing and conducting surveys. In Wholey, J. S., Hatry, H. P. and Newcomer, K. E. (eds) Handbook of Practical Program Evaluation. Jossey-Bass, San Francisco, CA, pp. 271–292.
NOC*NSF (1999) 55-plus in Beweging! NOC*NSF, Arnhem.
Pols, M. A., Peeters, P. H., Kemper, H. C. and Collette, H. J. (1996) Repeatability and relative validity of two physical activity questionnaires in elderly women. Medicine and Science in Sports and Exercise, 28, 1020–1025.[ISI][Medline]
Rasbash, J., Browne, W., Goldstein, H., Yang, M., Plewis, I., Draper, D. et al. (1999) A User's Guide to MlwiN. Institute of Education, London.
Ronda, G. and van Assema, P. (1997) Een Voorstudie ten Behoeve van de Ontwikkeling van de Werkboeken voor het Regioproject ‘Hartslag Limburg’: Minder vet Eten, Meer Bewegen, Stoppen/Niet Beginnen met Roken. Maastricht University, Maastricht.
Ronda, G., Steenbakkers, M., Van Assema, P. and Ruland, E. (2001a) Hartslag Beweeg TeeVee: een provinciaal programma voor bewegingstimulering. GGD Nieuws, 3, 21–27.
Ronda, G., Van Assema, P. and Brug, J. (2001b) Stages of change, psychological factors and awareness of physical activity levels in the Netherlands. Health Promotion International, 16, 305–314.
Ronda, G., Van Assema, P., Ruland, E., Steenbakkers, M. and Brug, J. (2003) The Dutch Heart Health community intervention ‘Hartslag Limburg’: evaluation design and baseline data. Health Education, 103, 330–341.[CrossRef]
Ruland, E., Harting, J., Van Limpt, P., Ronda, G., Van Assema, P., Van Ree, J. et al. (1999) ‘Hartslag Limburg’: A United Approach in Preventive Care. GGD-ZZL, Maastricht.
Ruland, E., Harting, J., Van Limpt, P., Ronckers, S., Ronda, G., Steenbakkers, M. et al. (2001) ‘Hartslag Limburg’: Annual Report 2000. GGD-ZZL, Maastricht.
Salonen, J. T., Kottke, T. E., Jacobs, D. R. and Hannan, P. J. (1986) Analysis of community-based cardiovascular disease prevention—evaluation issues in the North Karelia Project and the Minnesota Heart Health Program. International Journal of Epidemiology, 15, 176–182.
Snijders, T. A. B. and Bosker, R. J. (1999) Multilevel Analysis. An Introduction to Basic and Advanced Multilevel Modelling. Sage, London.
SPSS for Windows (2000) Rel. 10.0. SPSS Inc, Chicago, IL.
Statistics Netherlands (1997) Survey Working Population 1996 (Enquête Beroepsbevolking 1996). Sdu/uitgeverij/CBS publicaties, 's-Gravenhage, p. 80.
Tudor-Smith, C., Nutbeam, D., Moore, L. and Catford, J. (1998) Effects of the Heartbeat Wales programme over five years on behavioural risks for cardiovascular disease: quasi-experimental comparison of results from Wales and a matched reference area. British Medical Journal, 316, 818–822.
Van Assema, P., Steenbakkers, M., Kok, G., Eriksen, M. and De Vries, H. (1994) Results of the Dutch Community Project ‘Healthy Bergeyk’. Preventive Medicine, 23, 394–401.[CrossRef][ISI][Medline]
Van Assema, P., Brug, J., Glanz, K., Dolders, M. and Mudde, A. (1998) Nationwide implementation of guided supermarkets tours in the Netherlands: A dissemination study. Health Education Research, 13, 557–566.
Van Assema, P., Brug, J., Ronda, G. and Steenhuis, I. (2001) The relative validity of a short Dutch questionnaire as a means to categorise adults and adolescents to total and saturated fat intake. Journal of Human Nutrition and Dietetics, 14, 377–390.[CrossRef][ISI][Medline]
Vartiainen, E., Korhonen, H. J., Koskela, K. and Puska, P. (1998) Twenty year smoking trends in an community-based cardiovascular diseases prevention programme. European Journal of Public Health, 8, 154–159.
Weinstein, N. D. (1988) The precaution adoption process. Health Psychology, 7, 355–386.[CrossRef][ISI][Medline]
Windsor, R. A., Baranowski, T., Clark, N. and Cutter, G. (1994) Evaluation of Health Promotion, Health Education, and Disease Prevention Programs. Mayfield, Mountain View, CA.
Winkleby, M. A., Feldman, H. A. and Murray, D. M. (1997) Joint analysis of three U.S. community intervention trials for reduction of cardiovascular disease risk. Journal of Clinical Epidemiology, 50, 645–658.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. DE NOOIJER, P. VAN ASSEMA, E. DE VET, and J. BRUG How stable are stages of change for nutrition behaviors in the Netherlands? Health Promot. Int., March 1, 2005; 20(1): 27 - 32. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Ronda, P. Van Assema, E. Ruland, M. Steenbakkers, and J. Brug The Dutch Heart Health Community Intervention 'Hartslag Limburg': design and results of a process study Health Educ. Res., October 1, 2004; 19(5): 596 - 607. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||