To investigate the changes in blood pressure and their causes in an elderly population.


Orthostatic blood pressure measurements were performed in randomly in four birth cohorts (years 65–, 75–, 80– and 85, n=773) at 5-year intervals.


Both systolic and diastolic blood pressures decreased in both genders and all age groups. The falls in blood pressure related closely to initial blood pressure values and to the thickness of the left ventricular posterior wall of the heart. The changes in supine (r=0·118,P=0·007) and standing systolic blood pressure (r=0·123,P=0·005), as well as supine (r=0·148,P<0·001) and standing diastolic blood pressure (r=0·186,P<0·001) correlated with changes in body weight. Changes in supine diastolic blood pressure also related to changes in serum cholesterol (r=0·207,P=0·002) and triglycerides (r=0·160,P=0·016). Changes in supine and standing systolic and standing diastolic blood pressures also related to changes in dehydroepiandrosterone sulphate (r=0·161,P<0·05; r=0·205,P<0·01; r=0·140,P<0·05, respectively). Changes in blood pressure also correlated with self-estimated poor health after 5 years (r for supine systolic blood pressure=−0·133,P<0·001, for standing systolic blood pressure=−0·135,P<0·001, for supine diastolic blood pressure=−0·111,P<0·002). Patients who were institutionalized during the follow-up or who had severely impaired performance capacity after 5 years, displayed the most marked decline in blood pressure. Declining systolic blood pressure was also related to impaired survival prognosis. According to the logistic regression analyses the baseline blood pressure, antihypertensive drugs and changes in dehydroepiandrosterone sulphate and cholesterol explained over 30% of the changes in blood pressure. In the logistic models, declining blood pressure was associated with baseline blood pressure, antihypertensive treatment, poor health after 5 years, and decreasing cholesterol and triglycerides.


Decline in blood pressure in old age is associated with deteriorating health and is only partly explained by the use of antihypertensive drugs.


Correspondence: Professor Reijo S. Tilvis, MD, Division of Geriatrics, Department of Medicine, University of Helsinki, Haartmaninkatu 4, FIN-00290 Helsinki, Finland.