At the time of the study sample selection, on December 31, 1990, 7539 subjects (3132 men and 4407 women) in the city of Oulu were entitled to a special refund of antihypertensive medication. Of these, 443 subjects (268 men and 175 women) were aged 40–44 years, 601 (333men and 268 women) 45–49 years, 757 (400 men and 357 women) 50–54 years and 952 (476 men and 476 women) 55–59 years, totalling 2753 subjects. At that time, 403 000 subjects of the total population had this benefit in the whole country. The control cohort of the present study does not represent the “hypertension-free” part of the population, mostly due to the reimbursement criteria of the Social Insurance Institution. Subjects suffering from mild (diastolic blood pressure below 105 mmHg and systolic below 180 mmHg), uncomplicated hypertension are not entitled to a special refund (instead they receive a basic refund), and they are therefore included in the present control population. Further, 7 men and 14 women in the control cohort were on antihypertensive medication. Throughout the study, these subjects were considered part of the control cohort. All these factors may dilute the potential differences between the control and hypertensive study groups and make it more difficult to draw conclusions from the differences between hypertensives and ”normotensives”, particularly because more than one third of subjects in the male and one fourth of the female control cohort had hypertensive blood pressure values.
The control cohort was randomly selected from the part of the middle-aged population not entitled to a special refund for antihypertensive medication. It is estimated that no more than 10% of the Finnish middle-aged population have this benefit. In spite of this, it is not unlikely that the control cohort would represent the general middle-aged population as a whole, particularly as the participation rate was excellent. Accordingly, the main characteristics of the control cohorts are comparable to those previously reported for the Finnish population (Vartiainen et al. 1994).
The high participation rate and the results which agree with the previous studies suggest that there is no selection bias in the study cohorts and that our conclusions are relevant to the population at large. It should be emphasized that the study subjects were not selected high-risk individuals, but subjects from a random population-based cohort and randomly recruited hypertensive cohort.