Читать книгу Clinical Obesity in Adults and Children - Группа авторов - Страница 109

Among the most important challenges for biological and biomedical scientists is to identify the appropriate level of complexity at which to address a problem. In obesity research, approaches range from the simple energy balance frameworks (“energy in/energy out” and “eat less/exercise more”), to complex systems models that integrate many biological and environmental factors across multiple scales [9,14–16]. There is no general answer as to which is the optimal level of complexity required because this depends on the system under consideration, the question being asked, and the goal of the study – e.g. whether it is to understand the mechanisms, predict the behavior of the system, or manage it for particular outcomes. An important general principle, however, is that attributed to Albert Einstein: “Things should be as simple as possible … but no simpler.”

But how for a given problem does one decide on what is “as simple as possible,” but not too simple? In the early 1990s, biologists were grappling with this problem, not in obesity research but in the ecological sciences. Ecologists study the interactions among organisms and their environments, and the most ubiquitous and important of such interactions involves diet: eating and being eaten. Understanding the drivers of food and diet selection is thus fundamentally important in ecology, but it is also potentially very complex. To begin with, animals need, and foods contain many nutrients and other components such as indigestible fiber and toxins, and their requirements and tolerances for each vary – for example with age, health, reproductive state, and activity levels. Further, nutrients do not act independently but interact in complex ways, such that the levels in the diet of some influence the effects of others, and therefore their ratios (i.e. balance) can be as, or more important than the individual amounts and concentrations in the diet. Such interactions can involve several nutrients. For example, calcium balance is influenced by dietary phosphorus, vitamin D, and some carbohydrates, such as lactose [17]. To further complicate things, for almost all species, very few, if any, foods contain all the required nutrients in the right ratios to satisfy nutrient requirements and even if they did, no food (with the exception perhaps of mammalian milk) changes its composition to oblige the changing nutrient needs of its consumer. Animals therefore need to combine several, often many, different foods in their diets in the right proportions to satisfy their complex and dynamic nutrient needs. And these are just a few of the complexities that nutrition entails.

In the 1990s, there were three dominant frameworks for understanding the role of diet in ecology, all of which were founded on highly simplified models of nutrition [18]. Two of these, Optimal Foraging Theory and Classical Nutritional Ecology, assumed that animals in the wild are disproportionately influenced by a single dietary component, energy, and protein, respectively, without regard for nuances such as those discussed above. The third framework, ecological stoichiometry, does emphasize the importance of dietary ratios, but makes the simplifying assumption that all nutrients can be represented by chemical elements – principally carbon, nitrogen, and phosphorus. By avoiding the complexities of nutrition, these frameworks enabled diet to be integrated into ecological theory and, in this respect, have made important contributions. However, there was no telling what had been lost to ecological theory by omitting essential features of nutrition not captured by these simplified frameworks. At the other end of the complexity spectrum, the applied and mechanistic nutritional sciences were amassing substantial detail about the chemistry and physiology of nutrition, with little attempt to integrate this detail into theory or draw on theory to synthesize and apply the information [13].

Against this background, the question arose what is “enough but not too much” nutritional complexity for ecological models? The most direct way to answer this is through discovering how animals themselves deal with the complexity of nutrition. There are, in theory, benefits from regulating the dietary intake of all nutrients with great precision, but there are also constraints and costs [19]. For example, the computational machinery required to solve such a high‐dimensional optimization problem is immense, and brains cannot be dedicated to nutrition alone. Even if they could perform the required integration beyond a certain level of dietary perfectionism, the gains from foraging will run into diminishing returns, and time would be better spent on other activities, such as predator avoidance, mating, and sheltering. On the other hand, animals that are too cavalier in their nutritional regulation will be driven to extinction. For these reasons, we might expect animal nutrition to achieve a balance between complexity and simplicity, and identifying that balance will inform the appropriate level of nutritional complexity to incorporate into ecological models.