Tobacco Documents Online

Participants in control factories of the WHO Trial were handled mucfi dif£erently from the "usual care"' participants in the MRFIT. In the MRFIT, all "usual care" subjects were invited for annual checkups to assess their risk factor levels. Thus, they knew that they were at high risk, and were reminded of this fact every year. On the other hand, in the WHO Trial, at initial recruitment and screening„ a 10% random sample of men from the control factories was selected and screened for risk factor levels. Then, every two years, this same 10% was reassessed. At the sixth year of follow-up, all control participants were asked to come to a final examination to assess their risk factor levels. Therefore, until the final year of the WHO Trial, the majority of the control participants were unaware they were part off a study. As in the intervention group, monitoring, of CHD incidence and death was continuous. The published data on risk factor changes are given in terms of net percentage changes -- i.e., the risk factor changes in the intervention subjects after subtracting changes seen in the control subjects. In the MRFIT, the absolute changes in risk factor levels were given for both groups of subjects, rather than a single number representing the net change for the intervention group.

THE WORLD HEALTH ORGANIZATION EUROPEAN COLLABORATIVE:TRIAL Under the auspices of the World Health Organization, a randomized' controlled intervention trial was initiated in the early 1970's. The trial, conducted in five European countries -- the United Kingdom, Belgium, Italy, Poland (Cracow and Warsaw) and Spain -- is called the World Health Organization European Collaborative Trial in the Multifactorial Prevention of Coronary Heart Disease (CHD). It was undertaken for two purposes: (1? to assess the extent to which several major risk factors for coronary heart disease can be modified, and (2) to assess possible effects of such modification on CHD mortality and incidence. The inter- vention focused on the following risk factors: elevated plasma cholesterol, high blood pressure, cigarette smoking, excess body weight and physical inactivity. Six year mortality and' morbidity follow-up is complete for the U.K., Belgium, Italy and Cracow, Poland, and the data have recently been published. A summary of the current status of this WHO project is described here. Background and Procedure All participants were middle-aged male factory workers. The 18,210 subjects in the U1.K. were recruited from 24 factories. Belgium recruited 19,409 subjects from 30 factories. Italy C C.? -1 CJ ` F+ 4+

contributed 6,027 participants from 4 factories. In Cracow, Poland, 6,135 workers were studied, from 8 factories. The procedural aspects of the WHO European Collabora- tive Trial are discussed here in relation to those of the U.S. Multiple Risk Factor Intervention Trial (MRFIT). Whereas in the MRFIT, individuals were randomly assigned to intervention and "control" groups, in the WHO Trial factories were randomly assigned. That is, matched pairs of factories were selected for study; one of each pair was selected randomly for intervention efforts, and the other served as a control. In the intervention factories, an attempt was made to recruit all men aged 40 to 59, not just "high risk" men, as was the case in the MRFIT. The reason for not restricting the participants to high risk men was, according to the investigators, primarily to increase the practical relevance of the trial, because a general public health policy in regard to risk factors should be based on research which is population based. At the time of recruitment, the men in the WHO Trial ~ ~ intervention group were screened for CHD risk factors. Risk W 4%b scores were then calculated from a multiple logistic function ~j (MLF) using coefficients derived from the European cohorts of the Seven Countries Study.l In the MRFIT, the Framingham risk function was used.z

3. Rose, G., et,a1., "UK Heart f)isease Prevention Project: Incidence and Mortality Resilts," Lancet I: 1062-1066, May 14, 1983. r- 4. Kornitzer, M., et al., "Belgian Heart Disease Prevention Project: Incidence and Mortality Results," Lancet Io- 1066- 1070, May 14, 1983. 5. World Health Organization European Collaborative Group, "Multifactorial Trial in the Prevention of Coronary Heart Disease: 3. Incidence and Mortality Results," Eur Heart 14: 141-147, 1983. July, 1983

small and statistically unreliable overall effect, as well as the inconsistent effects among the individual countries, do not provide strong evidence that risk factor intervention resulted in any substantial reduction in CHD in the WHO Trial population. In a sense, the overall results of the WHO Trial are similar to those of the MRFIT, in that the observed fatal CHD reductions in intervention subjects were not statistically sig- nificant and were of similar magnitude (7.1% for the MRFIT; 7.4% for the WHO Trial)~. Despite this, results of the WHO Trial may not appear to challenge as strongly as the MRFIT results the risk factor reduction hypothesis. Two considerations are important here. First, the WHO investigators pointed out that the reductions in risk factor levels in their intervention group as a wholie were "generally much less"5 than those reported in the MRFIT. Thus, the reported failure to observe a statistically significant decrease in fatal CHD may not appear as striking in the WHO Trial as it did in the MRFTT, where a lack of beneficial effect was reported d'espite large relative decreases in risk factor levels. Second, unlike the MRFIT, the WHO Trial permitted arguably valid interverntion/control comparisons involving large 0 W ~ C.: ~ F+ ~ -7-

C differences were small. In view of the substantially greater reductions in cigarettes per day reported for the U.K., the substantial reduction in total CHa in Belgium (the only country with a statistically significant relative reduction) is striking. In fact, these results suggest an apparent negative correlation between percentage relative reduction in smoking and CHD. On another point, results from the U.K. section of the WHO Trial provided suggestive evidence that intervention may, in fact, have adverse consequences in terms of CHD and death. As noted earlier, there was a slight excess in the intervention group for all three major end points. While it is true that this excess was fairly small and not statistically significant, the relative increased death rates in the intervention group earlier in the trial were much higher (e.g., after the first year of the trial, fatal CHD was 100% higher and all causes death was 78% higher in the intervention group.) These results raise partic- ularly interesting questions when one remembers that U.K. par- ticipants achieved the largest relative reduction in smoking. Moreover, the U.K. results may be most relevant to the U.S. in that there is a similarity of cultural traditions and genetic heritage. The increased death rates in the intervention subjects may or may not be more than chance phenomena, but even the authors O of the report of the U.K. results noted' that these increased C4 W r+ ~ N f,0 -9-

In addition to net changes in individual risk factors, the WHO investigators also provid'ed a measurement of net change in MLF risk, which summarized the overall impact of intervention on risk factor levels. No such overall score was provided for the MRFIT subjects. The primary end points in the WHO Trial for the six year follow-up were: (1) fatal CHD, (2) total CHD (fatal CHD plus nonfatal myocardial infarction) and (3) total mortality. As with the risk factor data, these end point data are presented in terms of net percentages. That is, CHD and mortality rates in the intervention group are expressed as percentage differences from the control group, as contrasted with the MRFIT where the absolute rates were given for both the intervention and "control" subjects. - Due to the differences between the MRFIT'and the WHO Trial in reporting of risk factor and end point data, a direct comparison of these data in the two intervention trials is q,uite difficult. The WHO Trial end point data, in relation to net risk w ~ factor changes in the intervention group, have been recently ~ reported in three papers. Two of the papers deal individually p~. V with the U.K.3 and Belg;ium4 sections, which provided large enough samples to justify separate statistical analyses. The third

paper presents the combined results for the U.K., Belgium, Italy and Cracow.5 Results The overaLl net MLF risk reduction in the WHO Trial was 11.1%, although there was substantial variability between the countries, with Italy and Belgium reporting the largest MLF reductions (28.2% and 15.8%, respectively)~ and the U.K. the smallest (3.9%). The reported changes in cigarette smoking also varied greatly. The U.K. reported the largest net reduction (15.6%) in cigarettes/day, and Belgium the smallest (3.7%). Concerning end points, except for the U1.K., which reported slight, non-statistically significant increases in rates of CHDand death in the intervention men, the separate countries reported decreased rates (although statistically significant only for total CHD and total death in Belgium). The combined results indicated a non-statistically significant 7.4% reduction in fatal CHD in the intervention group. Comments In sum, despite a measurable reduction in overall net MLF risk, the WHO investigators reported' no statistically sig- nificant reduction in CHD risk in the WHO Trial as a whole. The -6-

death rates were a "cause for concern."3 This makes a valid point: hasty acceptance of the risk factor reduction hypothesis may have unanticipated adverse consequences. Conclusion In conclusion, the results of the WHO Trial do not constitute an indictment against cigarette smoking, nor a general demonstration of the value in CHD prevention of reducing risk factor leveLs and, in fact, suggest that caution should be exercised in any effort to modify personal attributes in an attempt to prevent coronary heart disease. * * * * * * * * * * 1. Keys., A., et al., "Probability of Middle-Aged Men Developing Coronary Heart Disease in Five Years," Circulation XLV: 815-828, April, 1972. 2. The Multiple Risk Factor Intervention Trial Group, "Statis- tical Design Considerations in the NHLI Multiple Risk Factor Intervention Trial (Mftb'IT)," J Chronic Dis 30: 261-275, 1977. -10-

A general campaign of risk factor modification (involving posters, brochures, etc.) was initiated,in each of the inter- vention factories.. In addition, those workers assessed as having the highest MLF risk were given more intensive counseling on risk factor modification; this involved' advice to lower cholesterol through dietary change, to quit smoking, to lose weight, to exercise daily, and to use drugs for control of hypertension. The intervention program in the WHO European Collabora- tive Trial was much less intensive than that in the MRFIT. For example, although the WHO Trial (all five countries) involved almost 32,000 men in the intervention factories, only the equiv- alent of 13 full-time medical and'paramed'ical staff were devoted to intervention. On the other hand, in the MRFIT there were a very large number of behavioral scientists, nutritionists, nurses, physicians, general health counselors and other staff involved in the intensive intervention for only about 6400 men. Risk factor changes in the WHO intervention group were monitored on the basis of yearly random samples of at least 5% of the intervention men. The MRFIT involved annual assessment of all intervention men. The WHO sampling technique may not constitute an important difference from the MRFIT, if it is assumed that the 5% random samples were representative of the WHO intervention group as a whole. - -3-