Ten-Year-Replicated Circadian Profiles for 36 Physiological,Serological and Urinary Variables in Healthy Men

At 3-hr intervals over a 24-hr span, 36 systemic, serologic and urinary variables were examined in 7 men in their mid 20's in the Spring of 1969, and again in the same 7 men in the Spring of 1979 under a similar chronobiologic protocol, using the same chemical and numerical analytical procedures. The variables examined for rhythms by cosinor were: vital signs—blood pressure (systoliC., diastoliC., pulse pressure and mean arterial pressure), heart rate, intraocular pressure (left and right), oral temperature; serum components—albumin, albumin/globulin ratio, total bilirubin, calcium, carbon dioxide, chlorides, bilirubin, cholesterol, globulin, glucose, potassium, sodium, sodium/potassium ratio, transaminase, triglycerides, total protein, urea nitrogen; and urine components—calcium, calcium/magnesium ratio, creatinine, magnesium, pH, potassium, sodium, sodium/potassium ratio, urea clearance, urea nitrogen, volume and zinc. Although all subjects appeared clinically healthy in 1969 and in 1979, certain inter-study differences were observed in a number of rhythm parameters of different variables. Statistically significant increases in mesor for the group as a whole were observed forserum Ca, cholesterol, Cl, CO₂, K, Na, and while statistically significant mesor decreases for a group as a whole were noted in serum glucose and transaminase. Statistically significant increases in amplitude for the group as a whole were observed in serum chloride and urinary Na/K ratio, while statistically signficiant decreases were observed in amplitude for blood pressure, heart rate, serum albumin, A/G ratio, globulin, glucose, protein, sodium and transaminase. For the group as a whole, a statistically significant advance in acrophase was observed in serum transaminase, while a statistically significant delay in acrophase was observed for serum A/G ratio, globulin, glucose, potassium, protein, sodium and for urinary magnesium. Statistically significant by sign test, but not by cosinor, was a numerical mesor increase for urinary urea clearance, a numerical decrease in mesor for urinary zinc; a numerical amplitude decrease for serum cholesterol; and a numerical delay in acrophase for oral temperature and serum cholesterol, CO², and globulin in all men examined. Only mesor changes in serum cholesterol and urinary Ca/Mg were positively correlated with the change in body size over the 10-year span between studies.

From a circadian chronobiologic perspective, the immense amount of data uniquely reviewed in this report across a 10-year span in seven healthy individuals serves a useful beginning to the study of the effects of normal aging upon commonly measured physiologic and biochemical variables and, more importantly, upon the circadian rhythm characteristics of these variables. A great deal of supposition about what happens to the mesor, amplitude and acrophase of an individual's circadian rhythms in a variety of endpoints has been based upon transverse studies of short duration and relatively few longterm studies. The further accumulation of data such as presented here and similar long-term longitudinal time series can have no adequate substitute for truly understanding whether reproducible age-related changes in circadian rhythms occur as individuals age.

With these qualifications and with the further qualification that the timing of our observations within the aging process (mid-20's and mid-30's) may be suboptimal for conclusions about aging, very interesting trends definitely appear worth comment. There is some evidence in these data that the flattening of circadian rhythms may really accompany advancing age. In grouped data, this fall in amplitude may be secondary to an isolated fall in predictable swing around the mesor or a combination of this and increased variability of the acrophase with or without amplitude changes. The data are not robust enough to be sure of the relative contribution of these two components. In any event, the circadian amplitude of each and every physiologic variable studied demonstrated a tendency to fall between the mid-20's and mid- 30's. This tendency toward a flattening of circadian variability is also a very prominent property of many of the serum chemistries which were measured. The circadian patterns of excretion of substances in the urine change much less between the mid-20's and mid-30's in our subjects. These findings may indicate a separate effect of aging especially upon metabolic hepatic variables and upon nephrologic circadian rhythms. Cardiovascular rhythms seem to change more in parallel with hepatic metabolic rhythms in contradistinction to the kidney-related serum and urinary rhythms.

Further, ongoing statistical analyses may hopefully turn up interesting and relevant cross-correlations among the individual data themselves in each study year and between the 10-year span, as well as with rhythm (mesor, amplitude and acrophase) and other physiologic characteristics of each subject. Planned re-observation of what happens to the circadian time structure of these seven individuals in their mid-40's will prove invaluable to further sorting out of the effects of aging upon circadian time structure.