20. Salonen JT, Kottke TE, Jacobs DR Jr, et al: Analysis of community-based cardio- vascular disease prevention studies-. Evaluation issues in the North Karelia.
Are People More Health Conscious? A Longitudinal Study of One Community Minco Nknian, PhD, R Craig Lefebvre, PhD, and Richard A. Carleton, MD
Introducton Communication campaigns designed with the objective of improving the health status of a target population have become more prevalent in the past two decades. 1-8 Many of these campaigns have been based on the assumption that people will behave in prescribed ways given sufficient information and motivation to support such decisions. An understanding of how these national campaigns are currently affecting awareness, knowledge and behavior related to cardiovascular disease (CVD) and its risk factors is important for studies that will examine risk factor changes and morbidity/mortality in study populations. The setting for this work is the Pawtucket Heart Health Program.5.9 In this paper, the question of whether residents of the comparison community have become more health conscious-especially with respect to CVD-is the focus.
Metods The Pawtucket Health Heart Program conducts a series of biennial surveys to assess relevant behaviors and cardiovascular risk factors in the study communities. Each of these surveys includes interviews and physiologic assessments of approximately 1,250 citizens between the ages of 18 and 65 in each of two New England cities, and spans the entire year in order to control for seasonal differences in physiologic risk factors. Households (defined as in the United States Census) are randomly sampled in each city using current listings known to be at least 95 percent complete. A single respondent from age-eligible members of the household is selected using tables adapted from Kish10 and Deming."1 A 15-20 minute interview includes questions about diet, exercise, smoking, CVD knowledge, availability of CVD-related programs, and sociodemographic characteristics. Measurements of height and weight, pulse, and blood pressure, and a 30 ml blood sample for later analysis of total cholesterol, triglycerides and HDL-cholesterol, are also obtained.
The survey methods and their reliabilities are discussed in detail elsewhere.12 The data presented in this paper were collected in the four cross-sectional surveys conducted in the comparison city in 1981-82, 1983-84, 1985-86, and 1987488. In addition, respondents to survey 1 were resurveyed after a lapse of five years: 61.3 percent completed the second examination. Each respondent was reexamined within one calendar month of the date of their first examination. The percentage of participants answering favorably to questions were considered as dependent variables for surveys. Physiologic variables will not be analyzed until the final surveys are completed. Potential covariates employed in the analyseswere: age, gender, education, language spoken at home, ethnicity, marital status, employment status, and household income level. The objectives of this report are to document awareness, knowledge and behavior changes over time and determine trends. To measure the linear trend, the slope was calculated by weighted linear regression for each dependent variable, and tested with the Chi-square statistic.13,14 To adjust for potential covariate effects, eight groups were constructed. The cutpoints were investigated carefully so that: (a) each group had enough observations, and (b) the differences between distribution of surveys were preserved. The linear trend was calculated for each group and then an adjusted overall measure of linear trend across all groups was taken as the weighted average of the eight trends.15 From the Department of Community Health, Brown University Program in Medicine (Niknian, Lefebvre); and the Department of Medicine, Memorial Hospital of Rhode Island, and Brown University Program in Medicine (Carle-
ton). Address reprint requests to Minoo Niknian, PhD, Biometrics Department, WarnerLambert Company, 2800 Plymouth Road, P.O. Box 1047, Ann Arbor, MI 48106-1047. This paper, submitted to the Journal January 26, 1990, was revised and accepted for publication October 10, 1990.
American Journal of Public Health 205
Public Health Briefs
Results A total of 5,070 persons participated in one ofthe cross-sectional surveys in the comparison city. Preliminary analyses were performed for each covariate. Gender distnbution and marital status were similar across the four surveys with females comprising 58 percent ofthe overall sample; approximately 60 percent of respondents were married. Significant differences among the four surveys were noted for age group distribution, years of education and proportion of employed. Survey participants at the last survey tended to be younger, better educated and more often employed compared to the first one. Persons of Portuguese descent comprised about 45 percent of the sample in
206 American Journal of Public Health
each survey, and only 73 percent of the sampled population spoke English. There was no indication of significant changes in these variables across the surveys. Using the weighted regression analyses to detect linear trends in CVD knowledge and behavior, a number of significant effects were found (see Table 1). When each of these linear relationships was adjusted for the covariates (i.e. age, education level and employment status), similar trends were found for each variable.* The expected changes in the population from regression analysis of changes in the cross-sectional surveys were compared with the actual changes found among cohort survey respondents.* The data indicate that, in most cases, the ob-
served change in the cohort exceeded the expected value derived from the crosssectional surveys.
Discussion These results support the conclusion that important secular changes are occurring in cardiovascular health awareness, knowledge and behavior. It appears that such effects are not due to the increased presence of prevention programs either at participants' worksites or religious and social organizations and are likely attributable to programs mediated through electronic and print mass media. *Data available on request to authors.
February, Vol. 81, No. 2
Public Health Briefs
Respondents in the cohort survey demonstrated greater changes in health promotion knowledge and behavior than expected from cross-sectional survey data. This may reflect a self-selection bias among those persons who were reexamined, although the first examination may also have induced more health consciousness among this group. The large proportion of persons of Portuguese heritage is a factor that limits generalizing these data to other sections of the region or countiy. The high proportion of respondents who spoke a foreign language at home raises the possibility that our estimates of the "secular trend" for CVD prevention-related awareness, knowledge and behavior are conservative. In more homogeneous, affluent, and educated communities, even greater effects may have occurred in the six-year period studied here.16 These observations support other studies of changes in awareness, knowledge and behavior among the general public.8 That these changes are occurring in a city that serves as a comparison area for a population-based CVD research and demonstration project may reduce the power to detect the effects of intervention in the intervention community.'7-21 [1
Acknowledgments This work was supported by Grant HL23629 (Pawtucket Heart Health Program) from the National Heart, Lung and Blood Institute of the US Department of Health and Human Services.
February, Vol. 81, No. 2
References 1. Roccella EJ, Ward GW: The National High Blood Pressure Education Program: A description of its utility as a generic program model. Health Educ Q 1984; 11:225-242. 2. Lenfant C: A new challenge for America: The National Cholesterol Education Program. Circulation 1986; 73:855-856. 3. Hersey JC, Klibanoff LS, Lam DJ, et at Promoting social support: The impact of California's "Friends Can be Good Medicine" campaign. Health Educ Q 1984; 11:293-311. 4. Farquhar JW, Fortmann SP, Maccoby N, et at The Stanford Five-City Project: Design and methods. Am J Epidemiol 1985; 122:323-334. 5. Lefebvre RC, LasaterTM, Carleton RA, et at: Theory and delivery of heath programming in the community: The Pawtucket Heart Health Program. Prev Med 1987;
13. 14.
15. 16. 17.
16:80-95. 6. Mittelmark MB, Luepker RV, Jacobs DR, et at Community-wide prevention of cardiovascular disease: Education strategies of the Minnesota Heart Health Program. Prev Med 1986; 15:1-17. 7. Danaher BG, Berkanovic E, Gerber B: Mass media based health behavior change: Televised smoking cessation program. Addict Behav 1984; 9:245-253. 8. Schucker B, Bailey K, Heimbach JT, etal: Changes in public perspective on cholesterol and heart disease: Results from two national surveys. JAMA 1987; 258:35273531. 9. Carleton RA, Lasater TM, Assaf A, et al: The Pawtucket Heart Health Program: I. An experiment in population-based disease prevention. RI Med J 1987; 70:533-538. 10. Kish L: Survey Sampling. New York: Wiley & Sons, 1965. 11. Deming WE: Sample Design in Business Research. New York: Wiley & Sons, 1960. 12. McKinlay SM, Kipp DM, Johnson P, etal:
18.
A field approach for obtaining physiological measures in surveys of general populations: Response rates, reliability, and costs. Health Survey Research Methods: Proceedings of the fourth conference. DHHS Pub. No. (PHS) 84-3346. Washington, DC: Govt Printing Office, 1984. Cochran WG: Some methods of strengthening the Common X2 tests. Biometrics 1954, 10: 417-451. Armitage P: Tests for linear trends in proportions and frequencies. Biometrics 1955, 11:375-385. Fleiss JL: Statistical Methods for Rates and Proportions. New York, John Wiley and Sons, 1980. Rogers EM: Diffusion of Innovations, 3rd Ed. New York: The Free Press, 1983. Blackburn H: Research and demonstration projects in community cardiovascular disease prevention. J Public Health Policy 1983; 4:398-421. Jacobs DR Jr, Luepker RV, Mittelmark MB, et al: Community-wide prevention strategies: Evaluation design of the Minnesota Heart Health Program. J Chronic Dis
1986; 39:775-788. 19. Puska P, Nissinen A, Tuomilehto I, et al: The community-based strategy to prevent coronary heart disease: Conclusions from the 10 years of the North Karelia Project. Annu Rev Public Health 1985; 6:147-193. 20. Salonen JT, Kottke TE, Jacobs DR Jr, et al: Analysis of community-based cardiovascular disease prevention studiesEvaluation issues in the North Karelia Project and the Minnesota Heart Health Program. Int J Epidemiol 1986; 15:176182. 21. Puska P, Salonen JT, Tuomilehto J, et al: Evaluating community-based preventive cardiovascular programs: Problems and experiences from the North Karelia Project. J Community Health 1983; 9:49-64.
American Journal of Public Health 207