Minutes of the Columbia University Seminar on Appetitive Behavior(#529)Date: November 6th, 2008 Seminar Title: A Research Symposium Honoring Theodore VanItallie, MD Celebrating 150 Years of Healing at St. Luke's Hospital Speaker's Name and Affiliation:
Speakers:
Presiding Chair: Harry R. Kissileff, Ph.D. Rapporteur: Kathleen L. Keller, Ph.D. Attendees and their Affiliation:
Summary: Barry Levin, MD, NJ School of Medicine and Dentistry
Dr. Levin's talk reviewed new data on Mayer's pioneering glucostatic theory of food intake. Dr. VanItallie and his co-authors commented on this hypothesis in the New England Journal of Medicine 249:13,1953 by noting ¡°In the normal aging animal, variations in blood glucose level exert a regulatory effect on food intake.¡± This theory suggests that a fall in blood glucose (by 6-8%) precedes intake of a meal. Mayer himself noted that hypothalamic cells ¡°sense fluctuations in glucose through the passage of potassium and phosphate into glucoreceptor cells.¡± Despite gaining support from a variety of other labs, including Jacques Le Magnen and Art Campfield, this theory lost support when it was noted that if you manipulate the glucose levels in humans, they do not report hunger until glucose levels are extremely low, well below those seen during even fasting. If the glucostatic theory were correct for physiological conditions, one would have predicted that hunger reports would have occurred at least at levels seen during fasting. Dr. Levin's laboratory has been revisiting the glucostatic theory by studying neurons that sense glucose, simultaneously identified in 1964 by Anand (Am J of Physiol, 207:1146,1964) and Oomura (Science, 143:483,1964). In these studies, two types of neurons were identified: Glucose Excited (GE) neurons which increase firing with increased glucose levels and Glucose Inhibited (GI) neurons which decrease firing with increased glucose levels. These neurons are located throughout areas in the brain that are known to be associated with the control of food intake (ie. arcuate and ventromedial hypothalamic nuclei, lateral hypothalamic area, etc). They are able to sense very small changes in glucose. Glucokinase (GK) is the gatekeeper for neuronal glucosensing and blocking its function ablates the ability of neurons to sense glucose. What role does hypothalamic neuronal sensing of glucose play in feeding? The primary function of these neurons is likely to moderate responses to low glucose levels. The evidence that these neurons play a significant role in feeding is minimal. Reduction of VMH GK by 70% showed little effect on feeding, body weight, or adiposity. Further, increasing VMH GK resulted in no increase in spontaneous or glucoprivic feeding. Thus, while hypothalamic neurons exist that can responds to small fluctuations in glucose levels, the primary function of these neurons is likely to protect organisms from the effects of low glucose levels. Q. Ted's hypothesis was that glucose utilization (not levels) affected feeding. Where are
we with this?
Q. There is evidence that with aging, brain utilization of glucose decreases and this can
play a role in diseases like Alzheimer's.
Harry R. Kissileff, PhD, St. Luke's Roosevelt Hospital, New York Obesity Research Center
The goals of this presentation were 1) to demonstrate that by looking at the finer details of eating behavior, one could obtain insights into its control, 2) to show how animal research has led to important findings about eating behavior in humans. 3) to honor Ted Vanitallie's contributions to the field of ingestive behavior. A historical and theoretical approach to topic was used to show how the microstructure of eating is inextricably linked to the physiological control of food intake. Nearly 80 years ago, Curt Richter (1) was the first person to recognize the potential of recording patterns of eating behavior and linking them to physiological controls. With a cage resting on tambours with different compartments for food and activity, Richter was able to record the temporal occurrence, but not the quantity of eating and activity rhythms. He hypothesized that these patterns were associated with rhythms of gastric motility, and although he displayed a schematic showing how he thought the gastric rhythm was coupled with feeding and activity, he never succeeded in simultaneously recording all three. Almost 25 years later, Hill and Stellar, succeeded in recording individual licks by rats (2) and provided the first demonstration that licking patterns were sensitive to motivational variables such as water deprivation (3). The importance of recording patterns of intake as the basis for the physiological control of food intake articulated by Brobeck (4) who stated ¡°that the total amount of food eaten is always the product of two factors, the number of meals multiplied by the intake of the average meal¡Any procedure altering food intake does so through some change in one or both of these. The regulation of food intake, then, is based upon the regulation of feeding behavior - the behavior peculiar to the beginning and ending of a typical feeding period.¡± The guiding principle of ingestive behavior, thus became the search for signals which start and stop meals i.e. physiological controls of food intake. Three main types of studies form the backbone of research on the physiological controls of food intake: 1) manipulating a physiological system and examining its effects on food intake, 2) measuring physiological variables before during and after eating and correlating their changes with changes in intake, and 3) measuring physiological variables and correlating them with the onset or termination of eating. My work has used all three approaches, but initially focused on animals. In parallel with methods for recording patterns of eating in animals, an effort was under way at St. Luke's Hospital where Sami Hashim and Ted VanItallie developed a feeding machine (5) based on the classical Skinner box. The experimental subject was a human who lived on a metabolic ward and obtained all his food by pressing a button which delivered 7 ml of a liquid diet for oral consumption through a tube. Patterns of button pushing coalesced into 4 to 5 meals a day at intervals spaced in a similar way to those seen in rats. I was recruited to St. Luke's to apply my skills in analyzing meal patterns in rats to understanding these records of humans. However, I realized that this system could be improved to collect data on rates of eating similar to what had been done in animals. Consequently with the support of Ted Vanitallie and the help of a grad student, Gary Klingsberg, I (6) was able to develop and test a universal eating monitor that could be used to measure instantaneous eating rates of either solid or liquefied foods. The monitor consisted of an electronic balance placed beneath a false panel in a table covered with a cloth and therefore hidden from the subject's view. A container placed on the panel held the food and weights were recorded every three seconds and transmitted to a digital computer in an adjacent room from which the subject could be observed by closed circuit TV. We showed that solid foods were eaten slower than liquids, and that more was eaten after 6 h deprivation than after 3 h, although the difference was small and not significant. With John Thornton, I tested several different equations for fit to the curves of cumulative intake (7). We demonstrated that the curves fit a quadratic function which could be interpreted as the result of two types of influence (8): 1) A constant facilitation, measured by the initial rate of eating, equal to the linear coefficient of the curve and 2) a constantly increasing inhibition (or disinhibition) reflected in the rate of change in eating, i.e. deceleration(or acceleration) of eating rate, measured by half the quadratic coefficient. Parallel studies in animals had shown that CCK reduces intake by increasing the rate of deceleration of eating without changing the initial rate (9), and that animals lacking CCK receptors showed an increase in food intake that was the result of decreasing the rate of eating at a slower rate than wild type animals (10). Therefore it was a surprise when food intake was reduced by CCK infusions in humans, but the rate of eating was not changed (11). Rather intake rate was the same with and without CCK, but duration of eating was reduced. It was conceivable that the difference between the human and animal studies could be attributable to the larger effects in the animal than in the human studies. Typically doses of CCK used in animal studies reduced intake by 50%, while in the human studies, the doses were close to threshold, about 15% reduction. Another similarity to the human studies was the influence of prior consumption on food intake when CCK was given. Antin, et al. (12) demonstrated that prior eating was necessary for CCK's food suppressing effects, and subsequent work suggested that gastric distension was a critical factor. Our lab demonstrated that a similar pattern held true for humans, in that prior food consumption was necessary for significant food intake reduction with a suprathreshold dose (13), and that gastric distension combined with CCK to reduce food intake in humans (14). Currently the laboratory is testing the effects of a CCK receptor agonist, and the results, although preliminary, appear to be promising. The agonist has the advantage of not requiring intravenous infusion and significant reduction in intake was achieved with a 4 mg dose. It is possible that this drug could be effective against binge eating, although it failed to reduce body weight of obese individuals in clinical trials(15) . Q. Does smell and taste affect food intake?
Q. Can you speculate as to why you didn't see a dose response relationship with your CCK
agonist?
Reference List
Blandine Laferrère, MD, St. Luke's Roosevelt Hospital New York Obesity Research Center
Up to 30% of patients presenting for bariatric surgery have type 2 diabetes (T2DM). Weight loss surgery generally results in a loss of 50¨C70% excess body weight and diabetes remission in 77% of patients. The rapidity in the onset and the magnitude of the effect of Rouxen- Y gastric bypass surgery (RY-GBP) on diabetes remission has thus far baffled clinical scientists. Contrary to diet, bariatric surgery results in 30-40% weight loss, often sustained overtime. With this significant weight loss, the resolution of co-morbidities occurs. The effect of bariatric surgery on diabetes remission is rapid in onset and of great magnitude. T2DM goes into remission in 80% of cases after GBP and in 47.9% - 73% of cases after gastric banding. The pathophysiology of T2DM is complex, with insulin resistance and beta cell defects that result in insulin deficiency, increased glucagon and decreased incretins. The incretins are gut peptides secreted in response to meals, which enhance insulin secretion. The incretin effect is responsible for 50% of the insulin secreted after meal absorption. The two main incretins are Glucagon-like peptide 1 (GLP-1) and Glucose-dependent insulinotropic peptide (GIP). The incretin effect, defined as the augmented response of insulin levels after oral glucose compared to matched IV glucose, is blunted in patients with T2DM. The possible mechanisms by which GBP improves T2DM include decreased food intake and caloric restriction, possible malabsorption, weight loss with decreased insulin resistance, and favorable changes in the levels of gut peptides implicated either in the meal-to-meal regulation of food intake (eg. ghrelin, CCK, PYY and GLP-1) or in insulin secretion (eg. incretins, GIP & GLP-1). Reports of increased incretin levels after various types of bypass surgeries started in the late 1970's early 1980's. GLP-1 is almost always reported increased after jejuno ileal bypass, biliopancreatic diversion or gastric bypass. The results on changes of GIP levels are less consistent with either elevated or decreased levels after the same types of surgeries. These studies had some gaps, as data on fasting rather than stimulated incretin levels were reported, studies were cross-sectional and compared results obtained from different surgical methods, and most studies reported data on either GLP-1 or GIP without the measure of the incretin effect on insulin secretion. Moreover, most studies were done in patients without diabetes. Therefore, our goal was to investigate the role of incretins as the mechanism for rapid improvement of glucose control in obese patients with T2DM after GBP. We planned a prospective study of patients with T2DM before and up to 4 years after surgery and included measures of total and active GIP and GLP-1 levels under fasting and stimulated conditions. In addition, we studied whether the changes in incretin levels resulted in a change in incretin effect on insulin secretion and tried to separate the role of weight loss per se from that of the anatomical changes of the gut secondary to bypass surgery. In Study 1, we tested if the blunted incretin effect in patients with T2DM will improve after GBP. Results show that 1) At 1 month after GBP, after10 kg weight loss, patients are off their diabetes medications 2) there is a significant decrease in fasting glucose, fasting insulin, HOMA-IR and in post-prandial glucose levels 3) there is a significant increase in incretin levels (both GIP and GLP-1) in response to oral glucose after GBP 4) the increase in incretin levels results in an improvement of the incretin effect on insulin secretion after GBP. In study 2, the long term incretin changes after GBP surgery were measured. In our research cohort, one year after GBP, we observed normalization of fasting glucose, insulin and HOMA-IR, decreased peak, 120' and 180' glucose levels, healthier pattern of insulin levels in response to oral glucose with recovery of the first phase insulin response and decreased late phase, sustained elevation of GLP-1 and GIP levels in response to oral glucose, and normal incretin effects. In study 3, we investigated the role of weight loss in the incretin changes after GBP. Previous data suggest that incretin levels increase in response to a test meal, after a dietinduced 18.8 kg weight loss. Our working hypothesis was that the increase in incretin levels and effect will be greater after GBP surgery than after equivalent weight loss by diet. To address this question, we designed a prospective study with a surgical group studied before and 1month after GBP and a matched diet group studied before and after a diet-induced equivalent weight loss. In summary, study 3 showed that 1) diet-induced weight loss does not result in changes in incretin levels or effect and 2) The changes in incretins after GBP occur as a consequence of the surgery independently of weight loss (Laferrère et al J Clin Endo Metab 2008 Jul;93(7):2479-85. 2008). In summary, bariatric surgery is a revolutionary treatment of morbid obesity and type 2 diabetes. It is the only intervention that results in sustained weight loss and diabetes remission. The effect of GBP on diabetes remission is not only weight loss mediated but possibly the result of favorable gut hormone changes. Although the significant changes of incretin levels and effect after GBP could explain part of the mechanism by which T2DM goes into remission after the surgery, it is likely that many other factors, such as the adipokines, may be implicated. Patients after GBP surgery and after GB experience great improvement of their appetite control, resulting in change of behavior and sustained weight loss. Other peptides, such as ghrelin, PYY, and CCK may also play a role in these changes. Finally, more research needs to be done to explain why a small percentage of patients fail to respond to the surgery and regain weight overtime. Q. Is there any evidence that the increase in incretins form in the distal small intestine after surgery?
Xavier Pi-Sunyer, MD; New York Obesity Research Center, St. Luke's Roosevelt Hospital, Columbia University College of Physicians and Surgeons
The endocannabinoid system (ECS) consists of receptors, their endogenous ligands, and the enzymes that synthesize and degrade these ligands. Physiologically, the ECS plays a role in modulating energy balance, feeding behavior, hepatic lipogenesis, and glucose homeostasis. Cannabinoid 1 (CB1) receptors are present in the central nervous system and peripheral sites, including adipose tissue, liver, gastrointestinal tract, pancreas, muscle and the cardiovascular system. CB1 receptor activation is important in a number of ways. It modulates energy balance, feeding behavior, and relieves stress. There are a number of endogenous ligands to these receptors, the best characterized are anandamide and 2-arachidonylglycerol (2- AG). These compounds are synthesized from phospholipids precursors when needed and have a very sort half-life. They are retrograde messengers that are produced and released from postsynaptic neurons and bind to pre-synaptic neuron receptors. In the pre-synaptic neuron, the activation of the CB1 signal reduces adenylyl cyclase, modulates ion channels, and reduces intracellular Ca++, inhibiting the release of neurotransmitters into the synapse. Endocannabinoids are thought to be taken up into cells and quickly degraded there. These receptors belong to the G-protein coupled receptor family. ECS overactivation is associated with increased food intake, obesity and dyslipidemia. The ECS normally functions to increase food intake and energy storage through coordinated central and peripheral mechanisms. Injection of endocannabinoids into the hypothalamus or mesolimbic region stimulates food intake. The ECS seems to have two ingestive behavior actions in the brain. The hypothalamus is involved in hunger/satiety mechanisms whereas the mesolimbic system is involved with the motivational aspects of eating, such as sensory appeal or reward properties. In animal models of diet-induced obesity, CB1 receptor expression is increased, as are endocannabinnoid levels. Endocannabinoids are potent appetite stimulators and animals with a genetic deletions of CB1 have a lean phenotype and are resistant to diet-induced obesity. Endocannabinoids stimulate AMPK activity in the hypothalamus and inhibit it in liver and adipose tissue. Antagonism to the CB1 receptor has been studied as a way to treat obesity. Rimonabant is a selective CB1 antagonist/reverse agonist. By selectively blocking the CB1 receptors both centrally and peripherally, rimonabant modulates the overactive ECS system. The Rimonabant in Overweight/Obesity (RIO) program comprises 4 Phase III trials assessing the efficacy and safety of rimonabant. Two 2-year trials, RIO-Europe (RIO-EU) and RIO-North America (RIONA) have been completed, as have two one year trials, RIO-Lipids and RIO-Diabetes. The RIO-EU and RIO-NA were multicenter, randomized, double-blind, placebocontrolled parallel-group studies comparing placebo with rimonabant in patients who were overweight or obese (BMI>30 kg/m2 or BMI >27). The primary endpoint was weight loss and secondary endpoints included waist circumference, lipid levels, fasting glucose and insulin, and metabolic syndrome prevalence. Results of RIO-EU indicated that patients treated for one year with rimonabant 20 mg/day lost an average of 6.6 kg compared to 1.8 kg for those on placebo. Moreover, 27.9% of patients on rimonabant lost more than 10 percent of their initial body weight compared to 7.3% of those on placebo. Patients on rimonabant also had an average decrease in their waist circumference of 8.5 cm versus 2.4 cm for those on placebo. In addition to weight loss, a statistically significant improvement in metabolic risk factors with rimonabant 20 mg vs. placebo was also observed. The number of patients with the metabolic syndrome at baseline was 47.7% and it was reduced by almost half (25.3%) after treatment with rimonabant. After the first year of treatment, patients who received rimonabant 5 mg or 20 mg were re-randomized to either the same dose of rimonabant or placebo for an additional 52 weeks (the placebo group remained on placebo). Patients who remained on rimonabant for a second year maintained their weight loss while those who switched to placebo regained their weight. Metabolic measures were also significantly improved in patients treated with rimonabant 20 mg through the 2nd year. HDL-cholesterol increased, triglycerides were reduced, insulin sensitivity improved, and metabolic syndrome was reduced by more than one third. Side effects were present. Overall discontinuation rates for adverse events in the first year of the study were 7.2% and 12.8% in placebo, and rimonabant 20 mg groups. Depressed mood was 1.3% and 2.2%, while anxiety was 0.3% and 1.0%. The discontinuation rates in the second year were not different among groups. No differences were noted in the three groups in scores of the Hospital Anxiety Depression scale. Side effects also included nausea (4.3% and 12.9% for placebo and rimonabant 20 mg respectively), diarrhea (3.0% and 7.2%) and dizziness (4.9% and 8.7%). Only in a very small number of cases did these side effects lead to discontinuation of drug use. In summary, rimonabant is the first of a new class of drugs that block the overactivity of the ECS in overweight and obese patients. Having both a central and a peripheral action, they help not only to control weight, but also to improve lipids and insulin sensitivity in patients with the metabolic syndrome. Other endocannabinoid receptor blockers and/or reverse agonists have been in Phase III clinical trials and their efficacy and safety is similar. As a class, they were thought to have an important role in the treatment of obesity and its metabolic complications. However, the psychiatric side effects of the one drug, rimonabant, that came to marketing were such that it was withdrawn from the market. This class of drugs is thus not likely to be available in the near future. It is possible that different molecular structures may be found that would selectively affect the regulation of food intake without also causing serious psychiatric effects. Q. Is there any data that the level of cannabanoids (endogenous) is related to dietary fat intake?
Q. Do environmental changes upregulate the endocannabanoid receptors?
Q. Can you design a blocker that only blocks the appetite effects (not the mood
effects) of endocannabanoids?
Q. Was there any increase in memory in these trials?
Dr. Theodore VanItallie
Dr. VanItallie continues publishing and researching in the field of obesity at present. His latest publications are on Parkinson disease and sleep and energy balance. He has influenced the content and direction of countless young physicians and researchers at St. Luke's Roosevelt and the New York Obesity Research Center. His celebration at this 150th anniversary of St. Luke's Hospital is fitting and momentous. |