Eight around-the-clock surveys of corneal cell division (first graph on left, a) are summarized as averages with the uncertainties involved in the second graph from the left in the time domain and in the cosinor display on the top right in the phase domain. One of the surveys had a damped circadian amplitude which fell outside the lower 5% limit of a 90% prediction interval, computed on the basis of the other 7 surveys, as seen on the right in the lower graph. That particular study happened to be done on a day with a magnetic storm. All original data (left) from the late Lawrence E. Scheving. 

Context in source publication

Context 1

... work based on spotchecks suggested that pulse pressure is a strong risk factor for coronary events in untreated hypertensive men, whereas stroke is best prevented by mean BP (Millar et al. 1999). By 1991, we were long engaged in aligning more and more chronomes of the environment with those in living matter (Halberg et al. 1991), analyzing associations among them, an endeavor known as “Chronomics” (Halberg 2000, Halberg et al. 2001d). At the time of this writing, the point estimate of a cycle’s τ in FH’s systolic BP is about (~) 11 years, while that in his diastolic BP is ~15 years and in his HR ~30 years. The “conservative” ordering CIs (95% confidence intervals) (Marquardt 1963) of the τ s do not overlap each other. This non-overlap is so great that it may mean a true difference, whether or not the CIs are very much wider. It is true that the assumptions for computing CIs often are not met, and we use CIs for ordering purposes only. Fortuitously perhaps, the cycles found objectively in FH’s three sets of data each correspond in length to a different solar activity cycle, namely to the decadal Horrebow (Thiele 1859)-Schwabe (1844) cycle in relative sunspot numbers, Makarov and Sivaraman’s (1989) global solar cycle, and the Brückner (1890, cf. Stehr and von Storch 2000, Halberg et al. 2009c)-Egeson (1889)-Lockyer (1901) climate cycle (dubbed BEL) found in changes of the length in the sunspot cycle. It may also be happenstance that the BEL cycle is reflected in our analyses of 2,556 years of international battles (Wheeler 1951), in hundreds of years of military political affairs (Chizhevsky 1971), and of economics (Goldstein 1988), as well as in health and disease (Halberg et al. 2008d), where we also find a whole spectrum of new cycles known in physics, such as ~16-month and ~5-month cycles reflected further in sudden cardiac death (Halberg et al. 2006a, b) and in cardiac arrhythmia (Cornélissen et al. 2007). The fact that there are often environmental counterparts to cycles in biology and physiology (as in the case of FH’s data) is an argument supporting the validity of the three different τ s found in the same circulation. Happenstance is not ruled out, but it can be rendered unlikely by a statistical significance test of congruence just developed by one of us (PG). Beyond associations gauged by congruence in τ , defined by overlapping if not overlying CIs of the reciprocal cosmic-biospheric τ s, there can also be acongruence in phase, congruence in τ notwithstanding, Figs 2 and 3, a challenging topic in its own right. Partial congruence, i.e., different time courses of phases of 4 variables, Figs 2A to 2C, with global congruence of τ s, the latter not shown, supports the requirement that a complete description of aeolian behavior involves both global and time-varying analyses. We can also let the cosmos do a remove and/or replace approach, to find that the ~16-month cycle in FH’s systolic BP is dampened when the solar wind loses that component or buries it in noise so that it is no longer detectable (Halberg et al. 2006a). We also find in 39 years of terrorism no yearly component but a ~16-month component that appears intermittently in solar wind speed and in geomagnetism. Its appearance, defined by gaining statistical significance, happens to be followed with a lag by the same component’s amplification in terrorism, where this cycle lasts for a long time after the triggering and driving environmental ~16-month component has lost statistical significance (Halberg et al. 2008d), Figs 3A and 3B. Fig. 4 shows physiological associations with solar activity for the case of ~7-day rhythms in HR (Cornélissen et al. 1996), complementing those for an ~1.3-year component in FH’s systolic BP (Halberg et al. 2006a), or in terrorism (Halberg et al. 2008d). Other projects deal with the already noted basic topics, such as the findings in our circulation of an entire spectrum of new cycles, also found in solar flares, in the solar wind and in other gauges of solar and interplanetary dynamics. The nearly forgotten ~30-year BEL cycle, Fig. 5 (Halberg et al. 2008e), and transyears (that have τ s longer than a year) in the circulation and in 17-ketosteroid excretion, Fig. 6, among others, such as the cis-half-year, are just a few examples. Some of these cycles are discussed in two conference proceedings that are available free of charge online (Halberg et al. 2008e, 2009e). Transdisciplinary endeavors include the monitoring of cycles putatively associated with societal ills (Halberg et al. 2008d), also detected in data originally collected for personalized health care, far beyond the fashion of molecular clocks that also gains from objective quantification (Sothern et al. 2009). In a ~15-year series of mostly daily determinations of urine volume and 17-ketosteroid excretion (Halberg et al. 2008c), a component of the spectral structure – the cis-halfyear of ~5 months (Wolff 1983, Rieger et al. 1984, Cornélissen et al. 2008b) – reflects a beat period of rotations at different solar latitudes, with differences in phase among the spectra of variability in 17-ketosteroid vs. urine volume, that share the same τ of ~5 months further characterizing melatonin (Cornélissen et al. 2008b); those latter results may allow the study in individuals of mechanisms leading to societal diseases. It is now customary in science to have an experiment start at zero time, without further specification. The clock hour, however, along with details of lighting and feeding, may be needed as a minimum for a comparison between two groups that may be characterized by circadian rhythms differing in amplitude, phase or frequency. If this is the case, e.g., as differently phased lighting and feeding of two groups being compared, the only way to avoid blunders is to map the rhythmic variation. This is how chronobiology started (Halberg 1969, Halberg et al. 2003). To be specific, Fig. 7A was an incidental finding, a statistically highly significant and seemingly important difference. Fortunately, it was never published alone. Otherwise, had the results of Figs 7B and 7C been found and published by others (they were also not used in themselves by us), controversy could have ensued (Halberg et al. 2003). Problems were avoided by realizing by 1950 that one dealt with a feeding schedule (in the morning, of a diet restricted in calories) that became the dominant synchronizer over the lighting regimen for one of two groups studied, whereas the other group was allowed to feed freely by night during the daily dark (active) span. This circumstance resulted in a phase difference between the circadian rhythms of the two groups, leading to differing results when sampling was restricted to a single clock hour. Fig. 7D shows how this difference was resolved and how problems could readily be avoided. As infradian biologic rhythms with τ s longer than 28 hours are being documented, many nonphotic in origin, counterparts are found, e.g., in cycles of particle radiation from the sun and the galaxies, broadly galacto-helio-interplanetary-ionosphero- geomagnetics, gravitation and UV flux, among other variables monitored routinely by physicists with results in the public domain (Silverman 1992, Prabhakaran Nayar 2006, Prestes et al. 2006). These environmental data can all be aligned with biospheric information only if the calendar date and geographic coordinates, such as longitude(s) and latitude(s) of each experiment or survey are known. Pertinent information gains in importance and pertains even to experiments limited to a few minutes or hours, in order to check at least retrospectively, for instance, as to whether one dealt with a magnetically disturbed or quiet time. Fig. 8 allows the possibility that a magnetic storm may dampen a circadian rhythm in corneal cell division (Halberg et al. 2006c). Fig. 9 shows why one may not generalize in the cartography of an organism, as in the geography of continents (Halberg et al. 2006a). With a magnetic storm, a damping of the MESOR (Midline Estimating Statistic Of Rhythm) and of the circadian double amplitude (a measure of the extent of predictable within-day change) of melatonin in situ in the rat pineal is associated in the same animals with an increase of both MESOR and circadian amplitude of the melatonin rhythm in the hypothalamus (Jozsa et al. 2005) (cf. Burch et al. 1999, Weydahl et al. 2001). Changes in amplitude, even when they occur in the “normal range”, can indicate abnormality (Halberg and Cornélissen 1995, Otsuka et al. 1996, Cornélissen et al. 1997, Halberg et al. 2001a). For instance, differences in the opposite directions in the circadian amplitude of melatonin distinguish between health and either breast cancer risk, indicated by an increase in amplitude, or the presence of overt cancer, associated with a decrease in amplitude, respectively (Halberg et al. 2001a). As the circadian amplitude of 1-minute time estimation changes with age, it can possibly underlie trends in the opposite directions when measurements are taken at different clock-hours, Fig. 10. As chronobiology, the study of biologic rhythms, will be complemented by the investigation of associations with the cosmos, by chronomics, precautions in indicating not only photic but also invisible yet measurable nonphotic conditions may become indispensable in reports of scientific results. Such investigations will be facilitated by indications of universal time, the form in which physical data are usually provided, and of geographic site(s). Data coding for chronobiological analyses uses as format cnyrmodahrmn, where cn, yr, mo, da, hr, and mn stand for century, year, month, day, hour ...