Abstract

The aim of this study was to evaluate a community-based health promotion programme in terms of changing: (i) attitudes with respect to a healthy lifestyle; (ii) behaviour with respect to access to health-related information; and (iii) attitudes and health literacy regardless of socio-economic status. In this programme, 20 people are selected every 2 years in each municipality from the lay people of the community, and they are designated as members of a ‘community leaders' committee’ by the Mayor. They, as a group, have opportunities to gain knowledge about and skills in healthy lifestyles, and undertake voluntary activities to serve the community. A programme intervention sample (INT group) was selected from programme participants from 13 municipalities in the greater Tokyo area. A questionnaire survey was carried out with the INT group and a general population group (REF group). The data obtained for female respondents, aged 30–59 years, in the two sample populations (n = 662 and 1361, respectively) were analysed using the χ2 test, the Kruskal–Wallis test and multivariate log-linear methods. Another questionnaire was given to female programme participants (n = 200) to identify any changes since the start of their participation. The results showed that the people in the INT group were pursuing healthier lifestyles than those in the REF group; current non-smokers who performed physical exercise and who ate meals regularly paid more attention to a healthy lifestyle and were more interested in the relationship between food and health. From the INT and REF groups, 22 and 4% of people, respectively, frequently obtained information from health professionals, and 29.8 and 10.8%, respectively, were satisfied with their access to health-related information. Results of multivariate log-linear analysis showed that significantly more people in the INT group were doing exercise, eating meals regularly, paying attention to nutritional balance and to food additives, were interested in health, and were satisfied with access to health information, after excluding the effects of age and socio-economic factors (p < 0.05). The results also showed positive changes after the implementation of the programme. These findings indicated that the people in the INT group were significantly more likely to pursue a healthier lifestyle and to have greater health literacy than those in the REF group, regardless of socio-economic status. In conclusion, this community participation approach, employing a committee style, was effective in improving health-related behaviour and in promoting health literacy while overcoming socio-economic variation.

INTRODUCTION

Interest in the development of effective health promotion programmes for the purpose of changing health-related behaviour in people's everyday lives has been increasing, and such programmes have been developed in various ways (MRFIT Research Group, 1982; Lefebvre et al., 1988; Silvstri and Flay, 1989; Susser, 1995; George et al., 1996; Kegeles et al., 1996; Baxter et al., 1997), because they are thought to provide a useful means for preventing diseases and promoting health.

The influence of dietary habits on health status is well established in health promotion, particularly in terms of the major chronic diseases in developed countries (Shaper, 1987). Therefore, we paid particular attention to health education programmes focusing on improving nutrition, encouraging a healthy diet, and trying to increase health literacy in terms of choosing healthy and safe foods.

According to studies carried out in Japan (Japan Food Marketing Information Center, Inc., 1996), women of 30–59 years of age are a key group within the population with respect to caring for family health in the community through diet. They generally take the initiative regarding diet variation, food preferences, and family food habits.

Improving health literacy, defined broadly in the Ottawa Charter as ‘developing personal skills’ (WHO, 1986), is thought to be of central importance in establishing health education programmes. As Fleming et al. pointed out, lack of information on healthy eating is a chief obstacle to behavioural change (Fleming et al., 1997). Lack of access to the appropriate technical information is of concern when trying to change people's health behaviour to improve their health status. The development of health communication tools using new technology could provide a variety of ways in which people may access health-related information (Yach, 1998).

Several studies have mentioned that the impact of health education differs between social groups. Calnan demonstrated that a health education programme had more impact on middle-class groups than on working-class groups (Calnan, 1984). It has also been reported that social class influences recruitment to health education programmes focusing on diet (Baghurst et al., 1994; Smith et al., 1995). Socio-economic differences in awareness and knowledge in community-based public health projects have also been described (Göransson et al., 1996). Less educated groups are likely to show little positive change with respect to risk factors (Jackson et al., 1991).

Inequity in health among the different social groups within a community is one of the major urban public health concerns. The varying ability to access information and the differences in impact of health education programmes for different social groups should be considered when assessing health education programmes.

It is widely recognized that lower socio-economic status is related to the lower health status of a population (Kunst et al., 1990; Rodriguez and Lemkow, 1990; Wing et al., 1992). Lifestyle behaviour influencing health, i.e. smoking habits, physical activity and nutritional intake, has also been found to be closely related to and influenced by socio-economic status (Osler, 1993). It has been demonstrated that the lower the socio-economic group, the higher the prevalence of indicators of an unhealthy lifestyle (Hulshof et al., 1991). The differences in health behaviour between people of different socio-economic classes should be considered when health promotion programmes are designed, in order for those programmes to be effective.

Health education programmes are aimed at changing personal lifestyles to those conducive to better health. Those that are effective for lower socio-economic groups may reduce the gap in health status in a population.

Various health promotion programmes that involve community participation and that are initiated and driven by community members, i.e. lay people rather than professionals, have been tried. The advantages of adopting a community participation strategy have been widely reported (Mahler, 1986; WHO, 1986; McBeath, 1990). Studies have demonstrated the power of this approach in creating effective programmes (Tonon, 1980; Minkler, 1985; Eng et al., 1990; Green and Kreuter, 1991; Kegeles et al., 1996; Neuhauser, 1998). Community participation, establishing a community leaders' committee and placing importance on voluntary activity are characteristic features of this type of health promotion programme.

A health promotion programme establishing a community leaders' committee was initiated in 1961 in Japan. Leaders are nominated from lay individuals by the community members and are trained to encourage healthy lifestyles. Those recruited individuals, mostly women, are then expected to lead in conducting community- and neighbourhood-based health promotion activities, and are designated health promotion leaders by the Mayors of the individual municipalities. Ten to 20 people are recruited per term per municipality, each with a population ~100 000. The initial term of the programme was set to 1 or 2 years, with it being possible for leaders to be nominated for another term. The unique characteristic of this programme is that the designated health promotion leaders are not health professionals but lay people. Once designated, the appointed leaders complete an initial programme held by the municipality, upon completion of which they are expected to conduct their own activities in the neighbourhood where they live. They are provided with a small fund to support their activities in the community, where they carry out various activities such as health festivals, cooking classes and sports festivals.

Areas of activities have undergone change according to the transition of public health concerns in Japan: tuberculosis control and promotion of infant nutrition in the earlier years, and hypertension, obesity control and assistance for the elderly in recent years. The Japanese Ministry of Health and Welfare requires local governments to conduct this programme; however, the content and focus of the activities varies in response to the needs of the individual communities. The recruited members have the opportunity to learn the general principles of health promotion, and the knowledge and skills required for a healthy lifestyle, including knowledge of nutrition and food safety in the initial phase. They also learn how to conduct neighbourhood-based activities and have the means to obtain information relating to health. This initial phase of the programme is organized and conducted by the municipal health centres, whereas the other phases of the programme are collaborative actions of leaders in the community.

The uniqueness of this programme has been reported (Takahashi, 1996); however, the outcome of this programme has not yet been evaluated systematically. The purpose of the present study was to evaluate the effectiveness of this programme among middle-aged women. The programme involved training committee members to become health promotion leaders, following an initial period of leadership by local government. We took the aforementioned Japanese programme as a model, and compared the lifestyles and health literacy of those currently involved in the programme with those of the general population. The number of years of participation in the programme ranged from one to six in the former group, because the membership changes over the years. This study, therefore, allowed us to evaluate the effects of intervention by this model programme.

We first evaluated whether those involved in the programme had changed and begun to pursue a more healthy lifestyle, and whether they showed improved health literacy. The former evaluation focused on eating, smoking and exercise habits, and the latter evaluation on attitudes with respect to a healthy lifestyle, and behaviour with respect to access to health information. We then analysed whether this programme was effective, regardless of the socio-economic status of the participants.

METHODS

Subjects

We carried out a questionnaire survey in 1995–1996 in the greater Tokyo area, which is defined as being within a 50-km radius of central Tokyo (Takano and Nakamura, 1990). This area of 8610 km2 represents only 0.96% of the whole country. The population size in this area in 1985 was 28 287 256, which accounted for 23.5% of the total population of Japan. Subjects were selected both from those involved in the model programme and from members of the general population. We refer to the former as the intervention group (INT), and the latter as the reference group (REF). Stratified random sampling was performed for the selection of both groups. First, 13 municipalities out of 197 were selected as areas for study of the model programme, then another 13 municipalities were selected as areas for study of the general population. In total, 2050 questionnaires were sent to the municipal governments of the first 13 areas, to be completed by the INT group. In the case of the latter 13 municipalities, a second random selection was performed to obtain a subject population of females aged between 30 and 59 years. Two thousand individuals were selected as subjects for the REF group.

The health literacy of middle-aged women is considered to have an influence on change in dietary behaviour in the family. To match the subjects better in terms of sex and age, for further comparison, the responses of females aged 30–59 years from both groups were used for the analysis.

The questionnaire

The questionnaire asked the subjects questions on the following: age, sex, years of education, monthly household expenses, lifestyle features concerning health (current smoking status, amount of physical exercise and regularity of meals), attention to a healthy lifestyle in terms of food (nutritional balance, atmosphere of the dining room and food additives), interest in the relationship between food and health, sources of health-related information, satisfaction with their access to health-related information, and other related characteristics.

With respect to smoking, we asked whether each respondent was a current smoker, an ex-smoker or a non-smoker. In the analysis, an ex-smoker was taken to be a current non-smoker. With respect to physical exercise, we asked whether respondents performed exercise at least once a week. With respect to eating habits, we asked whether they ate meals three times a day regularly. With respect to the source of health-related information, we asked which health information sources they relied most heavily on, from among ‘television and radio’, ‘magazines and newspapers’, ‘books’, ‘friends and family members’, ‘health centre’ and ‘others’. Finally, with respect to respondents' satisfaction with the information they received, the response categories were: satisfied, fair and unsatisfied.

An in-depth survey

We performed an in-depth survey of the INT group in 1999, studying changes in their lifestyle, satisfaction with their access to health-related information, and satisfaction with care of the health of people in their neighbourhood since their participation in the programme. We asked 200 subjects with >6 months of experience in the programme to report these changes subjectively. This questionnaire helped us to study their acceptance of the programme.

Statistical analysis

We used the χ2 and Kruskal–Wallis tests to assess the differences in attitudes towards health, and the sources of and satisfaction with health-related information between the two groups. Statistical significance was considered to be p < 0.05. SPSS version 6.1 for Macintosh was used for statistical calculations.

In order to clarify the association between intervention through the model programme and the subjects' healthy lifestyles as well as their health literacy, multivariate log-linear methods were used for the purpose of excluding the effects of age and socio-economic status (Bishop et al., 1975; Haberman, 1979). Smoking, physical exercise and regularity of meals were used individually as dependent, health-related behaviour variables. Attention to a healthy lifestyle in terms of attention to food, interest in the relationship between food and health, and satisfaction with access to health-related information were set individually as dependent health literacy variables. The respondents' ages, socio-economic factors (monthly household expenses, educational backgrounds) and involvement in the model programme (the INT group or the REF group) were taken as independent variables.

RESULTS

From the INT group, 1252 responses were obtained (154 males, 1098 females, aged 55.2 ± 11.9 years). Six-hundred-and-sixty-two responses from females aged 30–59 years were used for further analysis. In the REF group, 1361 responses from females aged 30–59 years (44.4 ± 8.1 years) were obtained and used for the analysis.

Table 1 shows the characteristics of the study subjects. The subjects in the INT group were significantly older and had a lower educational background than those in the REF group.

Table 2 shows the lifestyle features, attention to a healthy lifestyle in terms of attention to nutrition and eating habits, and interest in the relationship between food and health of the respondents. The percentage of people in the INT group who were current non-smokers, who performed physical exercise at least once a week, who ate meals regularly and were paying attention to nutritional balance was greater than that in the REF group. More attention was paid to the atmosphere while dining, and to the knowledge of food additives than in the REF group. Also, the percentage of people in the INT group who were interested in the relationship between food and health was greater than that in the REF group (p < 0.05). The percentage of those who had quit smoking in the INT group (51.2%) was higher than that in the REF group (35.1%).

Table 3 shows the results concerning the sources of health-related information people relied mostly on. No answer was obtained in 33 and 45% of cases in the REF and INT groups, respectively. These were excluded in the process of comparison of percentages between the two groups. The people in the REF group obtained information more from mass media such as television, radio, magazines or newspapers, whereas the INT group relied more heavily on information from health professionals.

Table 4 shows the results concerning satisfaction with access to health-related information in the two groups. The people in the INT group were more satisfied with their current access to health-related information than those in the REF group (p < 0.05).

The results of multivariate log-linear analysis are presented in Table 5. Results relating to non-smoking and attention to the atmosphere while eating were excluded due to the failure to obtain statistically satisfactory models according to the results of the goodness of fit test (goodness of fit is 34.4 for non-smoking and 39.2 for attention to the dining atmosphere). After excluding the effects of age, education and economic status of individuals, involvement in the model programme (being in the INT group) was significantly associated with pursuing a healthy lifestyle in terms of physical exercise, eating habits, paying attention to nutritional balance and food additives, and having an interest in the relationship between food and health. It was also associated with higher health literacy in terms of access to health-related information, regardless of age and socio-economic status. Overall, the people in the INT group were more satisfied. A better educational background was associated significantly with an interest in the relationship between food and health, attention to nutritional balance, and food additives.

Regarding the results of the in-depth study, 118 responses were obtained. The percentages of people who responded that they had changed and were pursuing a healthier lifestyle, that they were more satisfied with their access to health-related information, and that they were taking greater care of the health of people in their neighbourhood were 61.0, 78.0 and 81.3%, respectively.

DISCUSSION

The present study showed that the people in the INT group were pursuing a healthier lifestyle than those in the REF group, and the in-depth study indicated that their lifestyle had changed after the intervention. This evaluation indicated that this health education programme was effective in inducing behavioural change towards greater attention to health in middle-aged women in the greater Tokyo area. We considered that the slight differences in response rates between the two groups would not have influenced the results obtained. Self-selection to participate in this study might indicate that responders in both groups have more interest in these issues than non-responders. One of the main factors considered as contributing to the success of the present programme is that lay people took action voluntarily. This aspect characterized the programme, and we call it a ‘leaders' committee style’ programme. The style of this programme is recognized as a community participation fostering style programme, not a ‘top-down instructive style’ programme.

In addition, as shown by the in-depth survey, the people in the INT group thought that they themselves were taking more care of the health of people in their neighbourhood than they did before the intervention. This indicates that they had become more active in promoting health and were taking care of not only their own health, but also the health of others by providing health knowledge to members of the community. It appears that their attitude plays an important part in promoting health in the community.

We considered two reasons why this programme was effective. First, as shown by Korhonen et al., interpersonal health communication such as communication with doctors, nurses, family members and friends has a strong impact in terms of changing health behaviour (Korhonen et al., 1998). Also, Flay et al. have shown that mass media are not effective enough to change health behaviour alone, i.e. when not combined with other channels (Flay et al., 1989). On the basis of these findings, we consider that the ‘leaders' committee style’ programme worked to increase interpersonal health communication, and facilitated sharing of experiences among the participants by involving community members. Secondly, the communication of health information, which is a key component of health education (WHO, 1998), was considered to contribute to the positive outcome of the present programme. As pointed out by Winkleby et al., people with high exposure to health media, i.e. good access to health-related information, have been found to be more likely to make a positive change in health behaviour (Winkleby et al., 1994). The results showed that the percentage of people who obtained most of their health-related information from health centres was significantly greater in the INT group than in the REF group. This finding indicated that the people in the INT group had better access to more accurate and reliable health-related information. They also used information resources as they wished and had enough skills to satisfy their demand for health information. This implies that those in the INT group had higher health literacy.

Although mass media have become a major source of information for the general public, it is still important to have access to health professionals to obtain reliable information. We consider that access to reliable information is one of the most important factors in changing health behaviour.

Both the results of multivariate analysis and the results of our in-depth study demonstrated that the INT group showed improved satisfaction with their access to information after the programme. We concluded that the respondents' reported need for health-related information before the intervention was satisfied following the intervention. It was probable that satisfaction was also attributable to the enhanced personal skills required for effective use of the information obtained.

The results of the multivariate log-linear analysis clearly showed that the intervention had positive effects in terms of changing people's lifestyle and changing the amount of attention paid to healthy food and diet, regardless of socio-economic status. It has been thought previously that differences in socio-economic status may widen the inequality in health across different socio-economic groups. Whether health programmes that aim to achieve behavioural change towards a healthier lifestyle and improvement of health literacy can overcome socio-economic differences has also been questioned. This programme, however, showed it is possible to lead people to improve their health behaviour by improving health literacy, regardless of educational background.

In conclusion, the present study has demonstrated a positive outcome in a health education, community leaders' committee style programme. This programme helped to promote health communication in the community and to improve health literacy by providing easier access to health information. This programme led the participants, who were being trained as community leaders, to change their behaviour towards health and then to lead in promoting health in their own communities.

Table 1:

Socio-demographic characteristics of the subjects

 INT group n (%) REF group n (%) p < 0.01 
Both the INT and REF groups were selected from females in the greater Tokyo area. The total numbers in the INT and REF groups were 662 and 1361, respectively. Where indicated, the difference between the REF group and the INT group was significant at p < 0.01 for the χ2 test comparison. 
Age (years)   Yes 
30–49 260 (39) 850 (62)  
50–59 402 (61) 511 (38)  
Monthly household expenses, in yen    
<200 000 65 (11) 164 (13)  
200 000–299 999 141 (25) 280 (22)  
300 000–399 999 152 (27) 336 (27)  
400 000–499 999 114 (20) 220 (17)  
≥500 000 96 (17) 266 (21)  
No answer 94 95  
Years of education   Yes 
≤9 70 (11) 108 (8)  
10–12 335 (53) 594 (45)  
13–15 174 (27) 404 (30)  
≥16 55 (9) 224 (17)  
No answer 28 31  
 INT group n (%) REF group n (%) p < 0.01 
Both the INT and REF groups were selected from females in the greater Tokyo area. The total numbers in the INT and REF groups were 662 and 1361, respectively. Where indicated, the difference between the REF group and the INT group was significant at p < 0.01 for the χ2 test comparison. 
Age (years)   Yes 
30–49 260 (39) 850 (62)  
50–59 402 (61) 511 (38)  
Monthly household expenses, in yen    
<200 000 65 (11) 164 (13)  
200 000–299 999 141 (25) 280 (22)  
300 000–399 999 152 (27) 336 (27)  
400 000–499 999 114 (20) 220 (17)  
≥500 000 96 (17) 266 (21)  
No answer 94 95  
Years of education   Yes 
≤9 70 (11) 108 (8)  
10–12 335 (53) 594 (45)  
13–15 174 (27) 404 (30)  
≥16 55 (9) 224 (17)  
No answer 28 31  
Table 2:

Lifestyle features, attention to a healthy lifestyle and interest in the relationship between food and health

 INT group n (%) REF group n (%) p < 0.01 
The total numbers in the INT and REF groups were 662 and 1361, respectively. Where indicated, the difference between the REF group and the INT group was significant at p < 0.01 for the χ2 test comparison.  
aThose who answered ‘I've never smoked’ and ‘I used to smoke before, but I don't smoke now’.  
bThose who answered ‘I eat a meal three times a day regularly’.  
cThose who answered ‘I pay much attention to it’.  
dThose who answered ‘I'm interested in the relationship between food and health’. 
Lifestyle features    
Current non-smokersa 631 (95.3) 1139 (83.7) Yes 
Perform exercise at least once a week 359 (54.2) 441 (32.4) Yes 
Eat a meal regularlyb 592 (89.4) 1047 (76.9) Yes 
Attention to a healthy lifestyle of food    
Nutritional balancec 260 (39.3) 367 (27.0) Yes 
Atmospherec 97 (14.7) 144 (10.6) Yes 
Additivec 286 (43.2) 401 (29.5) Yes 
Interest in the relationship between food and health    
Interestedd 617 (93.8) 1117 (82.7) Yes 
 INT group n (%) REF group n (%) p < 0.01 
The total numbers in the INT and REF groups were 662 and 1361, respectively. Where indicated, the difference between the REF group and the INT group was significant at p < 0.01 for the χ2 test comparison.  
aThose who answered ‘I've never smoked’ and ‘I used to smoke before, but I don't smoke now’.  
bThose who answered ‘I eat a meal three times a day regularly’.  
cThose who answered ‘I pay much attention to it’.  
dThose who answered ‘I'm interested in the relationship between food and health’. 
Lifestyle features    
Current non-smokersa 631 (95.3) 1139 (83.7) Yes 
Perform exercise at least once a week 359 (54.2) 441 (32.4) Yes 
Eat a meal regularlyb 592 (89.4) 1047 (76.9) Yes 
Attention to a healthy lifestyle of food    
Nutritional balancec 260 (39.3) 367 (27.0) Yes 
Atmospherec 97 (14.7) 144 (10.6) Yes 
Additivec 286 (43.2) 401 (29.5) Yes 
Interest in the relationship between food and health    
Interestedd 617 (93.8) 1117 (82.7) Yes 
Table 3:

Comparison of the most important source of health-related information for individuals

 INT group (total n = 362) (%) REF group (total n = 915) (%) p < 0.05 
The percentage of respondents providing no answer was 33% for the REF group and 45% for the INT group, and they were then excluded from the calculation of percentages. Where indicated, the difference between the REF group and the INT group was significant at p < 0.05 for the χ2 test comparison. 
Television and radio 37 42  
Magazines and newspapers 17 28 Yes 
Books 10 12  
Friends and family members 11  
Health centre 22 Yes 
Others  
 INT group (total n = 362) (%) REF group (total n = 915) (%) p < 0.05 
The percentage of respondents providing no answer was 33% for the REF group and 45% for the INT group, and they were then excluded from the calculation of percentages. Where indicated, the difference between the REF group and the INT group was significant at p < 0.05 for the χ2 test comparison. 
Television and radio 37 42  
Magazines and newspapers 17 28 Yes 
Books 10 12  
Friends and family members 11  
Health centre 22 Yes 
Others  
Table 4:

Comparison of satisfaction with the access to health-related information

 INT group n (%) REF group n (%) 
The total numbers in the INT and REF groups were 655 and 1348, respectively. Statistical significance between the INT and the REF group was demonstrated with a p value of <0.05 using the Kruskal–Wallis test. 
Very much satisfied 136 (20.8) 145 (10.8) 
Fair 502 (76.6) 1136 (84.3) 
Unsatisfied 17 ( 2.6) 67 ( 5.0) 
 INT group n (%) REF group n (%) 
The total numbers in the INT and REF groups were 655 and 1348, respectively. Statistical significance between the INT and the REF group was demonstrated with a p value of <0.05 using the Kruskal–Wallis test. 
Very much satisfied 136 (20.8) 145 (10.8) 
Fair 502 (76.6) 1136 (84.3) 
Unsatisfied 17 ( 2.6) 67 ( 5.0) 
Table 5:

Factors associated with a healthy lifestyle and health literacy: results of multivariate log-linear analysis (n = 2024)

Health-related behaviour Health literacy 
 Physical exercise Odds ratio (95% CI) Regularity of meals Odds ratio (95% CI) Attention to nutritional balance Odds ratio (95% CI) Attention to food additives Odds ratio (95% CI) Interest in the relationship between health and food Odds ratio (95% CI) Satisfaction with information Odds ratio (95% CI) 
95% CI = 95% confidence interval. 
Involvement in the model programme (INT/REF) 2.3 4.8 1.8 1.7 3.3 1.9 
 (1.9–2.9) (2.0–11.1) (1.4–2.2) (1.4–2.2) (2.2–5.0) (1.5–2.5) 
Age (middle/elder) 1.5 1.6 1.8 1.6 2.2 1.8 
 (1.2–1.8) (1.0–2.8) (1.4–2.2) (1.3–2.0) (1.6–3.0) (1.4–2.3) 
Educational backgrounds (high/low) 1.1 1.0 2.1 1.4 2.5 1.1 
 (0.9–1.3) (0.6–1.7) (1.7–2.7) (1.1–1.7) (1.9–3.4) (0.8–1.7) 
Monthly household expenses (high/low) 1.5 2.0 2.2 2.0 2.3 1.0 
 (1.0–2.1) (1.0–3.9) (1.4–3.3) (1.3–2.9) (1.4–3.8) (0.7–1.4) 
Monthly household expenses (middle/low) 1.2 3.0 1.9 1.6 1.7 1.2 
 (0.9–1.7) (1.7–5.3) (1.3–2.7) (1.2–2.3) (1.1–2.4) (0.9–1.5) 
c2 20.1 12.0 24.7 18.4 24.5 20.5 
Health-related behaviour Health literacy 
 Physical exercise Odds ratio (95% CI) Regularity of meals Odds ratio (95% CI) Attention to nutritional balance Odds ratio (95% CI) Attention to food additives Odds ratio (95% CI) Interest in the relationship between health and food Odds ratio (95% CI) Satisfaction with information Odds ratio (95% CI) 
95% CI = 95% confidence interval. 
Involvement in the model programme (INT/REF) 2.3 4.8 1.8 1.7 3.3 1.9 
 (1.9–2.9) (2.0–11.1) (1.4–2.2) (1.4–2.2) (2.2–5.0) (1.5–2.5) 
Age (middle/elder) 1.5 1.6 1.8 1.6 2.2 1.8 
 (1.2–1.8) (1.0–2.8) (1.4–2.2) (1.3–2.0) (1.6–3.0) (1.4–2.3) 
Educational backgrounds (high/low) 1.1 1.0 2.1 1.4 2.5 1.1 
 (0.9–1.3) (0.6–1.7) (1.7–2.7) (1.1–1.7) (1.9–3.4) (0.8–1.7) 
Monthly household expenses (high/low) 1.5 2.0 2.2 2.0 2.3 1.0 
 (1.0–2.1) (1.0–3.9) (1.4–3.3) (1.3–2.9) (1.4–3.8) (0.7–1.4) 
Monthly household expenses (middle/low) 1.2 3.0 1.9 1.6 1.7 1.2 
 (0.9–1.7) (1.7–5.3) (1.3–2.7) (1.2–2.3) (1.1–2.4) (0.9–1.5) 
c2 20.1 12.0 24.7 18.4 24.5 20.5 

REFERENCES

Baghurst, K. I., Baghurst, A. P. and Record, S. J. (
1994
) Demographic and dietary profiles of high and low fat consumers in Australia.
Journal of Epidemiology and Community Health
 ,
48
,
26
–32.
Baxter, A. P., Milner, P. C., Hawkins, S., Leaf, M., Simpson, C., Wilson, K. V., Owen, T., Higginbottom, G., Nicholl, J. and Cooper, N. (
1997
) The impact of heart health promotion on coronary heart disease lifestyle risk factors in schoolchildren: lessons learnt from a community-based project.
Public Health
 ,
111
,
231
–237.
Bishop, Y. M. M., Frieberg, S. E. and Holland, P. W. (1975) Discrete Multivariate Analysis: Theory and Practice. MIT Press, Cambridge, MA.
Calnan, M. (
1984
) Maintaining health and preventing illness: a comparison of the perceptions of women from different social classes.
Health Promotion International
 ,
1
,
167
–177.
Eng, E., Briscoe, J. and Cunningham, A. (
1990
) Participation effect from water projects on EPI.
Social Science and Medicine
 ,
30
,
1349
–1358.
Flay, B. R., Gruder, C. L., Warnecke, R. B., Jason, L. A. and Peterson, P. P. (
1989
) One year follow-up of the Chicago televised smoking cessation program.
American Journal of Public Health
 ,
79
,
1377
–1380.
Fleming, S., Kelleher, C. and D'Connor, M. (
1997
) Eating patterns and factors influencing likely change in the workplace in Ireland.
Health Promotion International
 ,
12
,
187
–196.
George, M. A., Green L. W. and Daniel, M. (
1996
) Evolution and implications of P.A.R. for public health.
Promotion and Education
 ,
3
,
6
–10.
Göransson, M., Hanson, B. S., Lindbladh, E. and Östergren, P. (
1996
) Using socio-economic differences in knowledge and attitudes to shape community alcohol programmes: experiences from the Kirseberg Project.
Health Promotion International
 ,
11
,
95
–103.
Green, L. W. and Kreuter, M. W. (1991) Health Promotion Planning: An Educational and Environmental Approach. Mayfield, CA.
Haberman, S. J. (1979) The analysis of qualitative data. Vol. 2. Academic Press, New York.
Hulshof, K. F., Lowik, M. R., Kok, F. G., Wedel, M., Brants, H. A., Hermus, R. J. and ten Hoor, F. (
1991
) Diet and other life-style factors in high and low socio-economic groups (Dutch Nutrition Surveillance System).
European Journal of Clinical Nutrition
 ,
45
,
441
–450.
Jackson, C., Winkleby, M. A., Flora, J. A. and Fortmann, S. P. (
1991
) Use of educational resources for cardiovascular risk reduction in the Stanford Five-City Project.
American Journal of Preventive Medicine
 ,
7
,
82
–88.
Japan Food Marketing Information Center, Inc. (1996) Yearly Food Consumption Statistics '96–'97 Survey. 254.
Kegeles, S. M., Hays, R. B. and Coates, T. J. (
1996
) The Mpowerment project: a community-level HIV prevention intervention for young gay men.
American Journal of Public Health
 ,
86
,
1129
–1136.
Korhonen, T., Uutela, A., Korhonen, H. and Puska, P. (
1998
) Impact of mass media and interpersonal health communication on smoking cessation attempts: a study in North Karelia, 1989–1996.
Journal of Health Communication
 ,
3
,
105
–118.
Kunst A. E., Looman, C. W. and Mackenbach, J. P. (
1990
) Socio-economic mortality differences in the Netherlands in 1950–1984: regional study of cause-specific mortality.
Social Science and Medicine
 ,
31
,
141
–152.
Lefebvre, R. C., Lasater, T. M., Assaf, A. R. and Carleton, R. A. (
1988
) Pawtucket Heart Health Program: the process of stimulating community change.
Scandinavian Journal of Primary Health Care
 ,
1
, Suppl.,
31
–37.
Mahler, H. (
1986
) Toward a new public health.
Health Promotion International
 ,
1
,
1
.
McBeath, W. M. (
1990
) Health for all: a public health vision.
American Journal of Public Health
 ,
81
,
1560
–1565.
Minkler, M. (
1985
) Building supportive ties and sense of community among the inner-city elderly: the tenderloin senior outreach project.
Health Education Quarterly
 ,
12
,
303
–314.
Multiple Risk Factor Intervention Research Group (MRFIT) (
1982
) Multiple risk factor intervention trial: risk factor changes and mortality results.
Journal of the American Medical Association
 ,
248
,
1465
–1477.
Neuhauser, L., Schwab, M., Syme, S. L. and Obarski, S. K. (
1998
) Community participation in health promotion: evaluation of the California Welness Guide.
Health Promotion International
 ,
13
,
211
–222.
Osler, M. (
1993
) Social class and health behaviour in Danish adults: a longitudinal study.
Public Health
 ,
107
,
251
–260.
Rodriguez, J. A. and Lemkow, L. (
1990
) Health and social inquities in Spain.
Social Science and Medicine
 ,
31
,
351
–358.
Shaper, A. G. (
1987
) Environmental factors in coronary heart disease: diet.
European Heart Journal
 ,
8
, Suppl. E,
31
–38.
Silvestri, B. and Flay, B. R. (
1989
) Smoking education: comparison of practice and state-of-art.
Preventive Medicine
 ,
18
,
257
–266.
Smith, A. M., Baghurst, K. I. and Owen, N. (
1995
) Dietary behaviors of volunteers for nutrition education program, compared with a population sample.
Australian Journal of Public Health
 ,
19
,
64
–69.
Susser, M. (
1995
) Editorial: The tribulation of trials-intervention in communities.
American Journal of Public Health
 ,
85
,
156
–160.
Takahashi, S. and Ogino, H. (
1996
) Case study: Training course for community leaders of health promotion activity: Planning involving the participants and the evaluation.
Hokenfu Zassi
 ,
52
,
881
–886.
Takano, T. and Nakamura, K. (1990) Baseline Data for Healthy City Tokyo, Gyosei, Tokyo.
Tonon, M. A. (
1980
) Concepts in community participation: a case of sanitary changes in a Guatemalan village.
International Journal of Health Education
 ,
23
S,
1
–16.
Wing, S., Barnette, E., Casper, M. and Tyroler, H. A. (
1992
) Geographic and socioeconomic variation in the onset of decline of coronary heart disease mortality in white women.
American Journal of Public Health
 ,
82
,
204
–209.
Winkeleby, M. A., Flora, J. A. and Kraemer, H. C. (
1994
) A community-based heart disease intervention: predictors of change.
American Journal of Public Health
 ,
84
,
767
–772.
World Health Organization (WHO) (
1986
) Ottawa Charter for Health Promotion.
Health Promotion
 ,
1
,
iii
–v.
World Health Organization (WHO) (1998) Health Promotion Glossary.
Yach, D. (
1998
) Telecommunication for health-new opportunity for action.
Health Promotion International
 ,
13
,
339
–347.