Date Published: August 22, 2011
Publisher: Impact Journals LLC
Author(s): Steven A. Brown, Karen Schmitt, Anne Eckert.
The relationship between aging and daily “circadian” behavior in humans is bidirectional: on the one hand, dysfunction of circadian clocks promotes age-related maladies; on the other, aging per se leads to changes and disruption in circadian behavior and physiology. For the latter case, recent research suggests that changes to both homeostatic and circadian sleep regulatory mechanisms may play a role. Could hormonal changes be in part responsible?
Increasing evidence suggests that disruption of circadian clock function – either genetically or environmentally – can exacerbate a wide range of age-related pathologies, ranging from cataracts to cancer. An excellent review on this subject was published recently in this journal . Equally relevant, however, and even more painfully obvious, is the impairment of circadian function that occurs as a natural process of aging. The German language has invented the term “senile Bettflucht” (literally, senile bed evacuation) to describe the difficulty that elderly individuals have in sleeping at night, and the early hour at which they rise. Indeed, one in four aged persons reports regular use of a prescribed sleep medication . Since such medications treat only the symptoms and are also potentially addictive, the origins of this sleep disturbance are an important public health question.
Exploiting the fact that human circadian clocks are conserved in most cell types, Pagani et al. examined the circadian properties of primary fibroblasts from older and younger individuals. Although the cells from both groups showed identical circadian properties (period, amplitude, entrained phase) when cultured identically, inclusion of serum from older individuals resulted in a shortening of circadian period and an earlier entrained phase in either cell type. This change was likely due to a substance in the serum of older individuals, because heat treatment gave older persons’ sera the circadian properties of younger persons’ sera, but did not change the properties of younger persons’ sera (Figure 1, bottom) .
In principle, sleep is regulated by two separable processes: a circadian one, which pushes diurnal species such as humans to sleep preferentially at night; and a homeostatic one, by which sleep drive increases with increasing time spent awake .
Purely theoretically, based upon the mechanisms outlined above, numerous hypotheses can be advanced to explain the disruption of sleep in the elderly. Let us consider the two features of this disruption separately. The earlier bedtime and waking time of elderly individuals could be a result of a shortening of endogenous circadian period. It could also arise from a change in the way the clock changes phase in response to light: anything that resulted in a net gain of morning light or loss of evening light would result in an earlier phase. Finally, the early phase of elderly individuals could arise from homeostatic effects: an increased sleep need would advance bedtime or delay wake time, and a decreased sleep need would advance wake time or delay bedtime.
Although numerous behavioral studies have been conducted over the past decade to address these hypotheses, no clear picture has emerged. Evidence to support and contradict each of them exists: Circadian period length: Although a shortened behavioral period length as a consequence of age has been observed in some animals , careful studies of older humans under conditions of “forced desynchrony” show no hint of such changes. In these experiments, subjects were kept under photoperiods so long (28h) that their endogenous circadian clocks could not adjust. Circadian period was determined under these “free-running” conditions by measuring rhythmic expression of the hormone melatonin, or diurnal variation in body temperature . Pagani et al. showed shortening of period in human fibroblasts, but only in the presence of blood serum from older individuals .
So far, little evidence exists to suggest that the period length of circadian behavior is changed in elderly individuals. Moreover, although studies suggest that homeostatic sleep is affected in fundamental ways in older individuals, these observations are likely insufficient to explain the marked circadian changes observed. In forced desynchrony studies that showed increased sleep fragmentation, investigators also observed an earlier sleep onset relative to the phase of the hormone melatonin . Similarly, under constant routine studies under constant dim light, the phase of gene expression in blood cells is still advanced . Since the light cycle in these studies was not systematically affecting the circadian clock, these results imply that changes in circadian phase are unlikely to be explained via strictly homeostatic mechanisms affecting light choice. Of course, homeostatic sleep mechanisms might also have more direct effects upon the circadian oscillator [31, 32], but these mechanisms remain to be explored.
This Perspective has confined itself (mostly) to discussion of specific theories about the interplay among aging, sleep, and the circadian clock. Already, mutation and gene profiling studies have implicated specific clock genes in the ageing process [1, 33, 34]. In fact, however, the best evidence for any model comes in the form of detailed mechanisms, and here is undoubtedly where future research will be directed. For example, in rodent models, aging is correlated with losses of specific classes of neurons (orexin and CRH) that could affect sleep architecture . Experiments to address whether these changes are necessary and sufficient to explain fragmented sleep — and whether similar changes are observed in the aged human brain that correlate with sleep disturbance – will reinforce homeostatic models. Similarly, it is well-known that human aging is accompanied by large alterations in hormone balance, both in the hypothalamic-pituitary-adrenal axis and elsewhere . If Pagani et al. wish to suggest that a hormone is in part responsible for age-related circadian dysfunction, then the best evidence in their favor would be identification of the suspected factor and characterization of its effects.