Endocrine Physiology
The goal of this 1.5-day workshop will be to gather clinical and basic science investigators who are interested in diseases of the exocrine and/or endocrine pancreas and in achieving an understanding of how the two compartments interact in disease. This workshop will provide an opportunity for investigators in exocrine diseases to come together with those studying islets in diabetes as a means to foster interdisciplinary discussion and identify areas for advancement.
Endocrine Physiology
Numerous problems can occur in the endocrine system. These can be considered as excessive or deficient hormone production. Endocrine organs are also prone to tumours (adenomas) which can over produce hormones. See Metabolic and Endocrine Disorders
Endocrine disruptors are chemicals that can interfere with endocrine systems at certain doses. These disruptions can cause cancerous tumors, birth defects, and other developmental disorders. Any system in the body controlled by hormones can be derailed by hormone disruptors. Specifically, endocrine disruptors may be associated with the development of learning disabilities, severe attention deficit disorder, cognitive and brain development problems deformations of the body that includes breast cancer, prostate cancer, thyroid and other cancers; sexual development problems such as feminizing of males or masculinizing effects on females, etc.[4]
We use a variety of multidisciplinary approaches in our research, from stem cells and cellular/molecular biology, to comparative physiology, to translational work in humans. We are able to provide such diversity in research opportunities by pulling together endocrine researchers from all over campus. We have 49 faculty trainers drawn from 15 departments across 5 schools and colleges, with a variety of backgrounds including PhD, DVM/PhD, MD, and MD/PhD. Our program has helped facilitate a marked increase in human translational studies by multidisciplinary and interdisciplinary teams (MD and PhD) over the last several years. In addition, we have seen a dramatic increase in MD and MD/PhD trainers interested in reproduction as well of the study of related adult diseases, which themselves may lead to complicated pregnancy. Details on our training program and environment can be found here.
Studying physiology and pathophysiology over a broad population for long periods of time is difficult primarily because collecting human physiologic data can be intrusive, dangerous, and expensive. One solution is to use data that have been collected for a different purpose. Electronic health record (EHR) data promise to support the development and testing of mechanistic physiologic models on diverse populations and allow correlation with clinical outcomes, but limitations in the data have thus far thwarted such use. For example, using uncontrolled population-scale EHR data to verify the outcome of time dependent behavior of mechanistic, constructive models can be difficult because: (i) aggregation of the population can obscure or generate a signal, (ii) there is often no control population with a well understood health state, and (iii) diversity in how the population is measured can make the data difficult to fit into conventional analysis techniques. This paper shows that it is possible to use EHR data to test a physiological model for a population and over long time scales. Specifically, a methodology is developed and demonstrated for testing a mechanistic, time-dependent, physiological model of serum glucose dynamics with uncontrolled, population-scale, physiological patient data extracted from an EHR repository. It is shown that there is no observable daily variation the normalized mean glucose for any EHR subpopulations. In contrast, a derived value, daily variation in nonlinear correlation quantified by the time-delayed mutual information (TDMI), did reveal the intuitively expected diurnal variation in glucose levels amongst a random population of humans. Moreover, in a population of continuously (tube) fed patients, there was no observable TDMI-based diurnal signal. These TDMI-based signals, via a glucose insulin model, were then connected with human feeding patterns. In particular, a constructive physiological model was shown to correctly predict the difference between the general uncontrolled population and a subpopulation whose feeding was controlled.
Human physiology, as a science, aims to understand the mechanical, physical, and biochemical functions of humans; moreover, because human dynamics transpire both on multiple spatial scales, ranging from molecular (e.g., genetics), to cell (e.g., metabolism), to organ (e.g., the heart [1]), to collections of organs (e.g., the circulatory system) and on multiple time scales ranging from fractions of a second to decades, it is likely that complete models of human functioning will consist of highly complex models whose scales interact in complex ways (e.g., via nonlinear resonance) [2]. In this context, population physiology aims to understand medium to long time scales of human physiology and pathophysiology where a population of humans is required to construct or discover a signal (metaphorically, population physiology is to physiology as climatology is to weather). Moreover, once a signal is constructed, the goal is to use this signal to understand human dynamics by both understanding the sources of the signals and then use that information to stratify the population into meaningful classes (e.g., phenotypes) according to the different signals. Consequently, population physiology, as we conceive it, has two broad features: data analysis consisting of the construction and analysis of population scale physiological signals, and the mechanistic modeling that can explain and rationalize those signals. The hope is that, through the use of EHR data, physiology can eventually be used by clinicians in the same way that physics is used by engineers. Thus, here we will employ diverse populations in an attempt to verify that an EHR-data-derived signal can be used to resolve first-order physiologic dynamics.
The mathematical modeling of physiological systems on the cellular and organ scales has a long history (cf., [3] and [4] for a wonderful introduction), while the modeling of larger scale organ structures is just beginning [5]. Fundamental to mathematical modeling of physiology is a concrete connection to real data; as is the case for other basic sciences, mathematical physiological modeling is always tested against physiological data collected in rigorously controlled circumstances. Nevertheless, there are at least two elements missing from modern physiological analysis, analysis over large populations and analysis over long time periods. The former is important because human beings have diverse reactions to different inputs (e.g., drugs, foods, etc.), and those differences have their roots in physiology. The latter is important because many differences amongst human reactions to input occur on a slow time-scale; for instance, some smokers develop cancer while others do not. The problem with using the classical physiology framework with its rigorously controlled conditions to study a large population over a long time period is that it is too expensive, intrusive, and dangerous to collect physiologic data for a large population over a long time period. Thus, it is likely that the lack of availability of population scale, long term data is the primary reason why wide-population, long term, physiologic studies to not exist.
We begin with a materials and methods section that has three distinct components. In subsection 0.3 we discuss endocrine physiology and introduce the mechanic model we use in this paper. We then introduce electronic health record data in general and the data we use in particular in subsection 0.4. The materials and methods section concludes with a discussion of the nonlinear time series analysis techniques we use (subsection 0.5). We then work through the results (section 0.5) and discussion (section 0.9) sections.
To interpret the results, it will help to abstract the physical mechanisms to a control system. In particular, the regulation of glucose can be thought of as an intra-body feedback control system where the body has a goal of maintaining a constant concentration of glucose and attempts to achieve this goal via various physiological mechanisms [21]. Broadly, when glucose levels are high, insulin is released by the pancreas leading to glucose being stored in the liver faster than it is released and the rate at which glucose is metabolized by the body is increased. Similarly, when glucose levels are low, glucagon is released by the pancreas, allowing for an increase in the rate glucose is released from the liver as well as a decrease in the rate glucose is metabolized by the body. This contrasts with, for example, the kidneys and their relation with creatinine, which can be grossly thought of as a filtering system instead of a control system aiming at maintaining a particular level of glucose. (Note, there are parts of the kidney that do behave as a control system [22]). It is worth mentioning that the above description of the endocrine system is greatly simplified, (for a more detailed view, cf. [23] [24]).
The end goal of population physiology is twofold: (a) we want to derive population-scale, data-based signals over medium to long time-scales in a way that can be connected to constructive, mechanistic models to further the understanding of human physiology; and (b) we want to be able to use these verified, constructive, mechanistic models to affect the health of human beings via clinical care. In this paper, we have demonstrated (a) but not (b), primarily because glucose/insulin modeling is not yet at a stage were it can be applied to affect clinical care in a direct manner. Nevertheless, we have begun one of the necessary steps for implementing (b): we have demonstrated that a mechanistic model of endocrine dynamics can accurately represent humans over the longer time scales that are relevant to clinical outcomes. 041b061a72
