Mortality in participants and non-participants of a multifactorial prevention study of cardiovascular diseases: a 28 year follow up of the Helsinki Businessmen Study.

Abstract

Objective-To investigate pretrial risk factors and long term mortality in participants and nonparticipants of a multifactorial primary prevention trial. Design-A prospective study among 3313 initially healthy businessmen. During the 1960s (1964 onwards), 3490 healthy male business executives born between 1919 and 1934 participated in voluntary health checks at the Institute of Occupational Health in Helsinki. From that period cardiovascular disease (CVD) risk factors were available in 3313 men. In the beginning of the 1970s these men were invited to join a multifactorial primary prevention trial of CVD. Six groups were formed: (I) healthy participants in a high risk intervention group (n = 612), and (II) their randomised control group (n = 610); (III) a non-participant low risk group (n = 593); (IV) an excluded group with signs of CVD (n = 563); (V) a refused group (n = 867); and (VI) dead (n = 68). Groups I and II participated in the five year prevention trial which started in 1974. Other groups were followed up through registers, with no personal contact. (1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)) mortality (per 1000) was 79.3, 106 6, 155-2, 179-9, and 259 3 in the low risk, control, intervention, refused, and excluded groups, respectively (P < 0.001). In the whole population of 3313 men, the 28-year total (n = 577) and coronary deaths (n = 199) were significantly predicted by smoking, blood pressure, and cholesterol; cancer deaths (n = 163) by smoking only; and violent deaths (n = 83) by none of the risk factors. One-hour postload glucose was significantly associated with total mortality in the intervention group only. When the intervention and control groups were included in the same model, the effect of group on total mortality tended to be dependent on the 1 h blood glucose value (P = 0*06 for the group by 1 h glucose interaction term). Conclusion-The traditional risk factors (smoking, blood pressure, and cholesterol) are significantly associated with 28-year mortality in this high social class population with previous health education. Conversely, a "clustering" of low risk factors predicted low total, coronary, and cancer mortality. The findings on 1 h blood glucose suggest that factors related to glucose tolerance explain in part the excess mortality in the intervention group compared with the control group. (Br Heart_j 1995;74:449-454)

Comparison of Mortality Between Respondents and Non-Respondents in a Mail Survey.

“…Some studies (not using mail questionnaires) have compared mortality between participants and non-participants in community-based surveys with screening examinations which aimed mainly at clarifying risk factors for cardiovascular disease [11][12][13][14] Most studies showed that non-participants had higher total mortality than participants 11-14) Our results are in agreement with these studies and suggest that non-respondents tend to be less healthy than respondents. Because more non-respondents than respondents may have been ill at the beginning of the survey, we recalculated age-adjusted rate ratios for all causes of death between respondents and non-respondents, excluding data from the first two years of the study.…”

“…The above mentioned community-based surveys with screening examinations had similar findings [12][13][14]. In particular, our study showed that the difference in mortality between non-respondents and respondents was much greater for cardiovascular disease than for cancer.…”

Effect on Trend Estimates of the Difference between Survey Respondents and Non-respondents: Results from 27 Populations in the WHO MONICA Project

“…However, many studies have shown that non-respondents are different from respondents [1,[3][4][5][6][7][8][9][10]. When respondents and non-responders differ considerably, the 70% response rate is not enough to ensure accurate and precise populations estimates and response rate of 90% or higher may be needed [18].…”

“…When the unit and item non-response are completely random, estimators of prevalence or association among variables based on respondents are unbiased but the precision is reduced [2]. Several studies have shown, however, that non-respondents differ from respondents in relation to socio-economic status and health profile [3][4][5][6][7][8][9][10]. These differences between respondents and non-respondents may bias estimators obtained from respondents' data especially if the response rate is low and/or these differences are large.…”

Life-style intervention at the worksite - reduction of cardiovascular risk factors in a randomized study

“…However, so far not many randomized intervention studies have been published on the prevention of cardiovascular disease (CVD) at the worksite (9)(10)(11). It is also still unclear which method of screening for cardiovascular risk factors should be recommended in this setting, as traditional screening programs (eg, with blood sampling on a broad basis) are often costly, time-consuming, and of limited value (12).…”