Reproducibility of the cold pressor test: studies in normal subjects.

The reproducibility of the cold pressor test was studied in healthy subjects. A non-invasive method was utilized for estimating beat-to-beat arterial blood pressure (BP) and heart rate (HR). The study population of 17 healthy volunteers consisted of two groups. In the first group (Group 1, n = 11), a 1-min test was performed three times during the same day. In the second group (Group 2, n = 6), a 2-min test was repeated at the same time of the day on three consecutive days. In both groups, the test response was defined as the 46- to 60-s mean, minus the prestimulus 15 s baseline mean. In Group 1, a fair test-retest reliability was observed for the systolic BP response (intraclass correlation coefficient R = 0.57). Large intraindividual HR and diastolic BP variabilities were found. The intraindividual testretest difference in Group 1 ranged from -8 to 11 beats/min (SD = 4.3, R = 0.49) for the HR, from -16 to 13 mmHg (SD = 6.3) for systolic BP, and form -21 to 20 mmHg (SD = 9.7, R = 0.23) for diastolic BP. Even larger variability was observed when the test was repeated on different days (Group 2). Thus, the maxim that the response pattern to the cold pressor test is fairly constant for each individual may not be true. It does not seem to be advisable to use the results from one solitary cold pressor test. The use of replicated measurements and large sample sizes in comparative studies to compensate for the low to moderate reliability of the cold pressor test is recommended.