The aim of this study was to assess any variation in positive, negative and total affect recorded longitudinally; to compare the results with those from prior transverse or hybrid population studies, based on the same or a different method of mood rating; and to test for any association of mood with cardiovascular, hormonal and geophysical variables monitored concomitantly. The study approach was as follows. A clinically healthy 34-year-old man filled out the positive and negative affective scale (PANAS) questionnaire five times a day for 86 days. Systolic (S) and diastolic (D) blood pressure (BP) and heart rate (HR) were also measured automatically at 30-minute intervals with an ambulatory monitor from May 19 to June 29, 2000, while different endpoints of heart rate variability (HRV) were also determined at 5-minute intervals from beat-to-beat electrocardiogram (ECG) monitoring for 42 days between May 3 and June 14, 2000, with only short interruptions while the subject took a shower and changed ECG tapes. Saliva samples were collected at the times of mood ratings for one month for later determination of melatonin and cortisol concentrations. Intervals of 24 hours of the record of each variable displaced in increments of 24 hours were analyzed by chronobiologic serial section at a trial period of 24 hours to assess the circadian characteristics as they changed from one day to another. Estimates of the midline-estimating statistic of rhythm (MESOR) and circadian amplitude and acrophase obtained on consecutive days were correlated among variables to assess any associations. The findings were as follows. Overall, a circadian rhythm was demonstrated for all variables. A positive association was noteworthy between the circadian amplitude of negative affect and the MESOR of both SBP (r= 0.363; P= 0.029) and DBP (r= 0.389; P= 0.019), suggesting that BP is raised in the presence of large swings in negative affect. Needing further validation was a weak association between the MESOR of negative affect and the circadian amplitude of SBP (r= - 0.272; P = 0.108), suggesting a lowering of the circadian SBP amplitude in the presence of a strong negative affect. Of further interest was the lack of a statistically significant relation between positive and negative affect, not only in terms of the MESOR but also in terms of the circadian amplitude.
