Introduction
Diet, or the consumption of calories from the external environment, is an obligatory task of all metazoans. Yet the effects of the nutritional environment are not simply a binary fed/not-fed switch. The type and density of a nutrient source can have profound secondary effects. In a medical sense, the dietary components can be considered good or bad, either preventing or enhancing the onset of disease due to both caloric load and the presence of auxiliary chemicals that can be beneficial or toxic to cells and organ systems. However, from a broader perspective, the dietary composition can also provide essential information about the state of other attributes of the environment. These factors may have shaped the life history characteristics and behavioral responses of all organisms. For instance, the ripeness (amount of sugar) in fruits can provide seasonality information. The availability of food may also, directly or indirectly, signal the potential presence of predators, competitors, or mates.
In this chapter, I will attempt to summarize our current, albeit limited, understanding of the relationship between dietary factors, internal disease state, and sleep behavior in nonhuman animals, with a particular emphasis on the invertebrate model systems where we can leverage the power of genetics to move forward quickly. I encourage the reader to investigate several excellent recent reviews on the topic of sleep in less complicated organisms, particularly the genetics of sleep in Caenorhabditis elegans and Drosophila (Allada & Siegel, 2008; Cirelli, 2009; Crocker & Sehgal, 2010). Given the depth and quality of these reviews, I will attempt to instead focus specifically on the interconnected relationship between diet, sleep, and disease and highlight major areas where more work is desperately needed.
How Do We Know that the Animal Is Sleeping?
Before we launch into a discussion of how diet affects sleep in animals vastly different from ourselves, it is important to consider the characteristics of sleep. How do we know if an animal is sleeping? This remains a somewhat controversial issue. In 1913, Pieron proposed behavioral criteria that hold up today, including (1) a typical body posture and site, (2) a behavioral state of quiescence, (3) an elevated arousal threshold or reduced responsiveness to external stimuli, and (4) rapid state reversibility (to distinguish sleep from coma, injury, or death). Later researchers added the criteria of a homeostatic response to deprivation and responsiveness of the sleep periods to the circadian rhythm (Hendricks et al., 2000). In humans, electrophysiological correlates of sleep have become invaluable both to positively distinguish sleep from quiet wakefulness and to assess the organization of sleep stages throughout a period of sleep. However, one tricky aspect of this analysis is that occasionally most, but not all, signs of sleep will be present, leading to an ambiguous situation that becomes even more unclear as we assess the impact of environmental variables. As we shall see, rules are meant to be broken. For instance, the bullfrog Rana catesbeiana is notable for its daily pattern of rest with no change in arousal threshold (meeting criteria 1, 2, and 4) (Hobson, 1967). Marine mammals, particularly dolphins, show electrophysiological correlates of sleep but these are only unihemispheric (one side of the brain) and often associated with stereotyped circular motions of the body (meeting criteria 1 and 4) (Lyamin, Manger, Ridgway, Mukhametov, & Siegel, 2008). Similarly, three-toed sloths, some cats, and many birds show electrophysiological correlates of sleep during active waking, and sleep-deprived humans will also show evidence of âsleepâ while behaviorally active (Campbell & Tobler, 1984). It seems clear that a completely rigid set of criteria cannot be applied to all animals and special consideration must be used when factoring in the relationship between sleep behavior and diet. Are all of these animals âtrulyâ sleeping? Likely not. From the perspective of the reductionist, it may not matter or even be beneficial. The reductionist will study each piece of a complex behavior in the organism that is most amenable to study. This approach has been remarkably successful for seemingly intractable problems such as memory, neuronal excitability, and cell biology and is being increasingly applied to complex behaviors and social interactions.
Different Ways to Evaluate Sleep
When considering an analysis of the environmental effects on sleep behavior, it is useful to consider not only the total daily sleep duration but also other characteristics of the sleep patterns, as these may impact the overall âqualityâ of the sleep experience. Some, but not all, of the characteristics may be affected by the dietary environment and disease state. These additional characteristics include the organization of the sleep behavior relative to the circadian day, the transition probability either into or out of sleep, the pattern of sleep states, and the number of sleep periods in the day (pure monophasic nighttime sleep appears to be a feature unique to simians). Furthermore, there are environmentally induced periods of sleep such as the rebound response to prior sleep deprivation and postprandial slowdowns that can share important characteristics with sleep. When considering the potential harm caused by disrupted sleep, there is both a concern regarding the overall long-term health status and the ability to safely complete waking tasks. For instance, a change in the probability of falling asleep (as is seen in narcolepsy) may not alter total daily sleep but would greatly impair safety and lead to loss of independence in a human. The organization of sleep states, such as slow wave and paradoxical sleep, within a given sleep period can also massively impact the quality of sleep. However, because evidence for the existence of sleep states in invertebrate model systems is scant (van Alphen, Yap, Kirszenblat, Kottler, & van Swinderen, 2013; van Swinderen, Nitz, & Greenspan, 2004), this chapter will focus on the analysis of behavioral patterns as indicators of the sleepâwake relationship.
One very useful broad generalization to consider when evaluating behavioral patterns is the reciprocal tradeoff between exploration and exploitation that characterizes behavior patterns and search strategies across a wide range of organisms. These alternating states of movement (exploration to seek resources) and relative inactivity (exploitation of the resources in a given area) are the foundation of reinforcement learning theory. The exploitation phase can comprise active feeding, mating, or sleeping. In all cases, there is a behavioral switch that turns off the exploration drive in order to promote dwelling, with sleeping being a potential extreme case of the dwelling phase where arousal is at a minimum. This relationship between exploration and exploitation is best characterized biochemically in terms of the âroverâ and âsitterâ phenotype in Drosophila larvae, where polymorphisms in a single gene, foraging, a cyclic guanosine monophosphate (cGMP)-dependent protein kinase, can tip the balance between the propensity for ...