COMMUNITY EDUCATION FOR CARDIOVASCULAR HEALTH

1192

Community Medicine COMMUNITY EDUCATION FOR

CARDIOVASCULAR HEALTH

JOHN W. FARQUHAR

NATHAN MACCOBY

PETER D. WOOD JANET K. ALEXANDER HENRY BREITROSE BYRON W. BROWN, JR WILLIAM L. HASKELL ALFRED L. MCALISTER ANTHONY J. MEYER JOYCE D. NASH MICHAEL P. STERN Sta nford Heart Disease Prevention Program,

Stanford

University, Stanford, California 94305, U.S.A. determine whether community health education can reduce the risk of cardiovascular disease, a field experiment was conducted in three northern California towns. In two of these communities there were extensive massmedia campaigns over a 2-year period, and in one of these, face-to-face counselling was also provided for a small subset of high-risk people. The third community served as a control. People from each community were interviewed and examined before the campaigns began and one and two years afterwards to assess knowledge and behaviour related to cardiovascular disease (e.g., diet and smoking) and also to measure physiological indicators of risk (e.g., blood-pressure, relative weight, and plasma-cholesterol). In the control community the risk of cardiovascular disease increased over the two years but in the treatment communities there was a subtantial and sustained decrease in risk. In the community in which there was some face-to-face counselling the initial improvement was greater and health education was more successful in reducing cigarette smoking, but at the end of the second year the decrease in risk was similar in both treatment communities. These results strongly suggest that mass-media educational campaigns directed at entire communities may be very effective in educing the risk of cardiovascular disease.

Summary

To

INTRODUCTION

CIGARETTE smoking, high plasma-cholesterol concentrations, and high blood-pressure are important risk faccors for premature cardiovascular disease. 1 2 In 1972 we uegan a field experiment in three northern California communities to attempt to modify these risk factors by community education. The mass media, face-to-face instruction, or combinations of the two may be used in community education campaigns. The habits influencing cardiovascular risk factors are complex and longstanding ones, are often einforced by culture, custom, and continual commercial advertising, and are unlikely to be strongly influenced by mass media alone.3-6 Face-to-face instruction and exhortation also have a long history of failure, particularly when aimed at producing permanent changes in diet7 and smoking habits.8 The disappointing results of a very limited attempt to reduce cardiovascular risk with direct mail and lecture campaign reinforced pessimism about the possibility of changing health behaviour rough public education.9

After considering the powerful cultural forces which reinforce and maintain the health habits that we wished to change, and in view of the failure of past health education campaigns, we decided to use an untested combination of an extensive mass-media campaign plus a considerable amount of face-to-face instruction. We also included three elements often ignored in health campaigns : (1) the mass-media materials were devised to teach specific behavioural skills, as well as offering infor. mation and affecting attitude and motivation; (2) both the mass-media approaches and, in particular, the faceto-face instruction used established methods of achieving changes in behaviour and self-control training principles ; and (3) the campaign was designed after analysis of the knowledge deficits and the media-consumption patterns of the intended audience. Our goal was to de. velop and evaluate methods for achieving changes in smoking, exercise, and diet that would be both costeffective and applicable to large population groups. RESEARCH PROCEDURE

Three roughly comparable communities in northern Catifornia were selected for study. Tracy was selected as a control because it was relatively distant and isolated from media in the other communities. Gilroy and Watsonville, the other two communities, share some media channels (television and radio), but each town has its own newspaper. Watsonville and Gilroy received fundamentally similar health education over two years through a mass-media campaign. Additionally, in Watsonville high-risk people received intensive face-to-face instruction. Two-thirds of this group was randomly assigned to the intensive-instruction treatment group (W-I.I.) and onethird which received health education through the media only was used as a control group (w-R.c.). Data were gathered from a random (multi-stage probability) sample of 35-59-year-old men and women through interviews conducted in a survey centre set up in each of the three communities. These annual interviews were designed to measure both knowledge about heart-disease and individual behaviour related to cardiovascular risk. Knowledge was measured by a 25-item test of factors associated with coronary heart-disease. Daily intake of cholesterol, saturated and polyunsaturated fats, sugar, and alcohol were estimated,’° and the daily rate of cigarette, pipe, and cigar smoking was recorded. Plasma-thiocyanate assay" indicated that only about 4% of those reporting abstinence may have given inaccurate reports. Coincident with each annual interview, we also measured plasma total cholesterol and triglyceride concentrations, sysTABLE I—DEMOGRAPHIC CHARACTERISTICS AND

SURVEY-RESPONSE RATES IN EACH OF THREE COMMUNITIES

1193 tolic and diastolic blood-pressure, and relative weight. Blood was collected into disodium E.D.T.A. ’Vacutainers’ after a fast of 12-16 h. Plasma-total-cholesterol concentrations were determined by the procedures of the Lipid Research Clinics Programl2 and were adjusted for systematic variation in bloodsampling method.13 Blood-pressure was determined by means of a standard mercury manometer with the cuff on the right arm and the patient sitting with the arm at heart level. Two measurements were recorded after the subject had been seated for several minutes and the second, taken approximately one minute after the first, was used for analysis. A different person measured blood-pressure in each community and all staff were trained in a standard manner. Other measurements made included plasma-renin and urinary sodium.14 15 Results were sent

participants and their physicians. The overall risk of coronary heart-disease developing within 12 years was estimated by a multiple logistic function incorporating the person’s age, sex, plasma-cholesterol concentrato

and stories, billboards, posters, and printed material to participants. A campaign was also created for the sizeable population of Spanish speakers. The media campaign began two months after the initial survey and continued for nine months in 1973, was withheld for three months during the second survey, and then continued for nine more months in 1974. Two-thirds (113) of the Watsonville participants whom we identified as being in the top quartile of risk of coronary heartdisease’6 were randomly selected for counselling. 107 attended counselling sessions and 77 high-risk individuals and 34 spouses completed all three interviews and examinations. These individuals, and their physicians, were informed by letter of their relatively high risk of coronary heart-disease (the letter was regarded as part of the "treatment"). They and their ments

posted

tion, systolic blood-pressure, relative weight, smoking-rate, and electrocardiographic findings. 16 The mass media and counselling campaigns were designed to produce awareness of the probable causes of coronary disease and of the specific measures which may reduce risk and to provide the knowledge and skills necessary to accomplish and maintain recommended behaviour changes. Dietary habits recommended for all participants were those which, if followed, would lead to substantial reduction of saturated fat, cholesterol, salt, sugar, and alcohol intake. We also urged reduction in body-weight through caloric reduction and increased physical activity. Cigarette smokers were educated on the need and methods for ceasing or at least reducing their daily rate of cigarette consumption. The mass-media campaign in Gilroy and Watsonville consisted of about 50 television spots, three hours of television programming, over 100 radio spots, several hours of radio programming, weekly newspaper columns, newspaper advertise-

TABLE II—COMPOSITION AND TREATMENT OF GROUPS IN COMMUNITIES

9 PARTICIPANT

*

s=Surveying and feedback of results (annual), M.M.=mass media, L.I=Intensive instruction programme. t Participants in the initial survey at Watsonville, Gilroy, and Tracy whose examination results placed them in the top quartile of risk of coronary heart-disease according to a multiple logistic function of risk factors.

t To correct for bias resulting from exclusion of intensively instructed subjects-i.e., high-risk persons and their spouses-means for remain-

ing subjects

in high-risk and lower-risk groups were weighted to compensate for the differential numbers of excluded subjects in the two

t

risk strata. Resulting weighted means were called means of the reconstituted sample-i.e., sample reconstituted after exclusion of intenstvety instructed subjects.

values and percentage change in selected variables after two years in control (shaded) or treatment (dark) groups. baseline or diffSee table ii for definition of groups. a=P<0.05 for erences in percentage change of control versus treatment. b=P<0.05 for differences in percentage change within treatment G versus W (total) or WR. c=P<0.05 for differences in percentage change within

Fig. 1—Absolute baseline

treatment w-R.c. versus W-1.1.

1194 spouses were invited to participate in the instruction programme that was launched six months after the first baseline survey and was conducted intensively over a 10-week period through group classes and home cdunselling sessions. In the summer months of the second year, at a less intensive level, individuals were counselled about special problems (e.g., smoking and weight-loss) and were encouraged to maintain previous changes. The counsellors were graduate students in communication, physicians, and specialist health educators trained in behaviour modification techniques. Pre-tested protocols were

used. 17 The intensive instruction programme was designed’8-2’ to achieve the same changes that were advocated in the media campaign. The strategy was to present information about the behaviour which influences risk of coronary heart-disease, to stimulate personal analysis of existing behaviour, to demonstrate desired skills (e.g., food selection and preparation), and to guide the individual through practice of those skills and

gradually withdraw instructor participation. RESULTS

Baseline values were remarkably uniform. Both the media and media plus face-to-face instruction had significant positive effects on all variables except relative weight after the two years of campaigning (fig. 1). (However, relative weight was significantly lower among TOTAL PARTICIPANTS

the Watsonville intensive-instruction group after one year.) Thus, for risk-factor knowledge, saturated-fat intake,cigarette use, plasma-cholesterol, and systolic blood-pressure there were slight-to-moderate changes in the expected direction. When the last three variables (and relative weight) were incorporated into the risk equation the net difference in estimated total risk between control and treatment samples was 23-28%. This difference is in part attributable to the fact that the greatest change occurred in individuals with the highest plasma-cholesterol and blood-pressure. Face-to-face intensive instruction (Watsonville 1.1.) in high-risk subjects increased knowledge gain and the extent of reduction of smoking but not other variables. Reasons for the changes in blood-pressure are not yet clear but include a probable interaction between weight-loss and enhanced adherence to antihypertensive medications, the latter being an unintended consequence of exposure to health education. 22 Changes in knowledge and risk factors produced in the first year in Watsonville and Gilroy were not only maintained, but improved further during the second year (fig. 2). Groups which received counselling show a greater decrease at the end of the first year than that in the media-only groups. But in the second year the groups exposed only to the media show further substantial gains and the apparent difference between the effects of media and media plus face-to-face instruction is reduced. The parallel relation of the two media-only groups (Gilroy and Watsonville reconstituted) indicates that the media campaigns had similar effects on risk in the two treatment communities.

DISCUSSION

subjects of previous studies of giving up smoking8 and weight reduction,’ our participants were randomly selected from open populations, thus providing a better basis for generalisations about future publicUnlike the

HIGH-RISK PARTICIPANTS

from baseline (0) in risk of coronary heart-disease after 1 and 2 years of health education among participants from three communities.

Fig. 2-Percentage change

a

Groups are defined in table n. Cardiovascular multiple logistic function of risk factors.

risk is measured

by

health education efforts. But since we were able to recruit only about two-thirds of the total samples of eligible participants selected in the three surveys, extrapolation of our results may be limited. Also we were not able to help participants learn to achieve sustained weight-loss. In general the changes in knowledge, behaviour, and physiological endpoints that were observed in the first year of treatment were maintained, and even improved in the second year of study. Intensive face-to-face instruction and counselling seem important for changing refractory behaviour such as cigarette smoking and for inducing rapid change of dietary behaviour. But we must also learn how to use these methods to correct obesity, and to employ them effectively with limited resources (e.g., by training volunteer instructors). Mass media are potentially much more cost-effective than faceto-face education methods. Our results show that mass media can increase knowledge and change various health habits, but we believe that the power of this instrument could be considerably enhanced if we can find ways to use mass media to stimulate and coordinate programmes of interpersonal instruction in natural communities (such as towns and factories) and to deliver forms of specialised training and counselling about weight-loss and smoking avoidance.

1195

Prevention of the premature cardiovascular disease epidemic of industrialised countries will require national purpose, planning, and action. It seems that part of this

effort-i.e., persuading people to alter their life stylescan be

achieved at reasonable cost.

Stunkard, A. J. Psychosom. Med. 1975, 37, 195. Bernstein, D. A., McAlister, A. L. Addict. Behav. 1976, 1, 89. Aronow, W. S., Allen, W. H., De Cristofaro, D., Ungermann, S. Circulation, 1975, 51, 1038. 10. Fetcher, E. S., Foster, N., Anderson, J. T., Grande, F., Keys, A. Am. J. clin. Nutr. 1967, 20, 475. 11. Butts, W. C., Kuehneman, M., Widdowson, G. M. Clin. Chem. 1974, 20, 7. 8 9.

1344.

This investigation was supported by grant HL 14174 (Stanford Specialised Center for Research in Arteriosclerosis) and contract NIH 71-2161-L (Stanford Lipid Research Clinic) from the National Heart, Lung and Blood Institute.

Requests for reprints should be addressed to J. W. F., Stanford Heart Disease Prevention Program, Stanford University, Stanford, Cahfornia 94305, U.S.A. REFERENCES 1. American Heart Association:

Intersociety

Commission for Heart Disease

Resources, Circulation, 1970, 42. 2. Blackburn, H. in Progress in Cardiology, (edited by P. Yu and J. Goodwin); vol. III, p. 1. Philadelphia, 1974. 3. Griffiths, W., Knutson, A. Am. J. publ. Hlth, 1960, 50, 515. 4. Bauer, R. A. Am. Psychol. 1964, 19, 319. 5. Cartwright, D. Hum. Relat. 1951, 4, 381. 6. Robertson, L. S., Kelley, A B., O’Neill, B., Wixom, C. W.,

Eisworth, R. S.,

Haddon, W. Am. J. publ. Hlth, 1974, 64, 1071.

Occasional

Survey

IMPACT OF NEW DIAGNOSTIC METHODS ON THE DIFFERENTIAL DIAGNOSIS AND TREATMENT OF SECONDARY AMENORRHŒA R. P. SHEARMAN

I. S. FRASER

Department of Obstetrics and Gynaecology, University of Sydney, Australia The development of homologous prolacSummary tin assays, multiple pituitary stimulation, tomography, and computerised axial tomography permit more detailed investigation of patients with secondary amenorrhœa than was formerly possible. 39% of 90 patients with secondary amenorrhœa had hyperprolactinaemia. 10 patients (11% of total) had pituitary tumours. 8 of these women had galactorrhœa (27% of those with galactorrhœa). For patients with hyperprolactinæmia but no tumour, bromocriptine is the treatment of first choice rather than clomiphene or human gonadotrophins. The best treatment for patients with detectable tumour is controversial, particularly when the tumour is confined to the sella turcica. Whether or not these tumours are true neoplasms remains to be determined. Clinically, a history of secondary amenorrhœa with or without galactorrhœa following withdrawal of oral contraceptives remains the commonest

presenting syndrome. INTRODUCTION

DURING the past several years there have been substantial technical developments which might help the investigation of patients presenting with secondary amenorrhœa. One major new method of treatment has become available. Relatively recent investigational techniques include homologous prolactin assays, multiple pituitarystimulation, pituitary tomography and computerised axial tomography (C.A.T. scan). The major new therapeutic agent is 2-bromo-.x-ergocryptine (bromocrip-

tine) .

Department of Health, Education and Welfare (N.I.H.) 75-628. Manual of Laboratory Operations, Lipid Research Clinics Program, Lipid and Lipoprotein Analysis, Bethesda, 1974. 13. Stern, M. P., Farquhar, J. W., Maccoby, N., Russell, S. H. Circulation, 1976, 54, 826. 14. Lucas, C. P., Holzwarth, G. J., Ocobock, R. W., Sozen, T., Stern, M. P., Wood, P. D. S., Haskell, W. L., Farquhar. J. W. Lancet, 1974, ii, 1337. 15. Lucas, C. P., Holzwarth, G. J., Ocobock, R. W., Sozen, T., Stern, M. P., Wood, P. D. S., Haskell, W. L., Farquhar, J. W. Angiology, 1975, 26, 31. 16. Truett, J., Cornfield, J., Kannel, W. J. chron. Dis. 1967, 20, 511. 17. Meyer, A. J., Henderson, J. B. Prev. Med. 1974, 13, 225. 18. McGuire, W. J. in Handbook of Social Psychology (edited by G. Lindzey and E. Aronson); p. 136, Menlo Park, California, 1969. 19. Bandura, A. Principles of Behaviour Modification; New York, 1969. 20. Thoreson, C., Mahoney, M. Behavioural Self-Control. New York, 1974. 21. McAlister, A. L., Farquhar, J. W., Thoreson, C. E., Maccoby, N. Hlth Educ. Monogr. 1976, 4, 45. 22. Curry, P. J., Haskell, W., Stern, M. P., Farquhar, J. W. CVD Epidemiology Newsletter no. 20, p. 48. Council on Epidemiology, American Heart Association, January, 1976. 12.

In order to determine what effect these developments have had on the differential diagnosis and treatment of patients with secondary amenorrhcea we have studied a consecutive series of 90 patients with secondary amenorrhoea seen in a period of 22 months since January, 1975, and we compare our findings with those of an earlier series of 230 patients fully investigated with older technology.1 PATIENTS AND METHODS

Patients

The 90 women included in this series gave a history of at least 12 months’ secondary amenorrhoea. The only patients excluded were those few who declined complete investigation. No patient with oligomenorrhoea, anovulatory cycles, or inadequate luteal function was included. Galactorrhcea was diagnosed if there was secretion of true milk either spontaneously or with gentle manual expression. Methods

investigations of patients with described previously,23 each of the women in the present series had multiple pituitary stimulation. This involves basal measurements of follicle-stimulating hormone (F.S.H.), luteinising hormone (L.H.), thyroid-stimulating hormone (T.S.H.), prolactin, human growth hormone (H.G.H.), and cortisol, followed by assay of the response to a bolus containing gonadotrophin-releasing factor (G.R.F.) 100 µg, thyrotrophin-releasing factor (T.R.F.) 200 µg, and insulin 0.1 In addition to the routine

secondary amenorrhoea

unit/kg body-weight.4 Patients with either abnormal cone views of the pituitary fossa or hyperprolactinaemia had 1 mm tomograms taken of the pituitary fossa. Among non-pregnant women in this unit, the upper limit of normal for prolactin in peripheral venous blood is 22 ng/ml. If tomography was abnormal this was followed by arteriography, pneumoencephalography, and, more

recently, c.A.T. scanning. Methods used for

measurement

of steroidal hormones, gon-

adotrophins, and H.G.H. have been described previously.j Plasma prolactin and T.S.H. levels were estimated by homologous radioimmut.oassay. The double-.antibody technique was used with Calbiochem anti-human-prolactin and T.S.H. antibodies, Calbiochem human prolactin standard, and T.S.H. standard prepared locally. The intra-assay variation was less than 5’c, and inter-assay variation was 12%.