Dennis A. Vanderweele

Analysis of feeding patterns in normal and vagotomized rabbits

Spontaneous meal sizes, intermeal intervals, and 24 hr feeding rhythms were monitored in normal and 60 day recovered vagotomized rabbits fed solid laboratory chow. Mean sizes of meals and intermeal intervals, and the circadian distribution of food intake did not differ between the two groups, but vagotomy was associated with increased frequencies of both smaller and larger than average meals. Positive meal to postmeal interval correlations were evident in intact but not vagotomized animals, whereas vagotomized animals displayed a meal to premeal interval correlation in the light phase that was not present in normal rabbits.

Elevated insulin and satiety in obese and normal-weight rats*

Three experiments were performed to evaluate the effect of elevated insulin levels on feeding. In the first, 22 normal-weight rats were implanted with Minipumps (Alza) which administered 1·92 u regular insulin per 24 h into the peritoneal cavity. These animals showed a reduction in spontaneous food intake which averaged 11·0% over the seven days the pump was actively infusing insulin (suppression averaged 16·8% during the dark). In the second study, dietary-obese rats also showed hypophagia when shifted from the obesity-inducing diet to Vivonex. This hypophagia was primarily accomplished by a reduction in meal size without a change in meal frequency—indicating an elevated satiety ratio—which is the same mode of reduction of ingestion shown by the animals with insulin elevations generated from minipumps. Therefore, we assessed insulin levels in dietary-obese animals and found them elevated (424·0 ± 32·3 μu/ml for obese and 123·6 ± 14·3 μu/ml for control subjects in the unfasted condition). Finally, we hypothesized that the elevated insulin occurring in the dietary-obese animal might account for a reduction in eating and meal size observed when the animal is switched back to a single diet. Obese and normal-weight rats were then given the diabetogenic agent streptozotocin (35–65 mg/kg body weight) to limit insulin release when switched. However, when compared to similarly treated, normal-weight subjects, obese animals showed the same average levels of suppression of food ingestion (16–24%) regardless of insulin levels. In our studies, then, insulin limited weight gain only in the non-obese (non-insulin-resistant) subject.

Pancreatic glucagon, food deprivation and feeding in intact and vagotomized rabbits.

Thirty New-Zealand female rabbits were implanted with hepatic-portal cannulas and six simultaneously underwent bilateral subdiaphragmatic vagotomy. When recovered, all animals received pancreatic glucagon infused at 1.0 cc/min for a total dosage of 12 microgram in 3.0 cc of isotonic saline. On alternate days, isotonic saline alone was infused as a control. Twelve intact and six vagotomized animals received infusions terminating food deprivations of 4, 8, and 24 hr while the remaining animals received the infusions only when free feeding. The feeding behavior of all animals was measured at 0.5, 1 and 2 hr postinfusion. Glucagon significantly suppressed feeding relative to saline only in 0- and 4-hr-food-deprived intact rabbits. Longer deprivations followed by glucagon did not produce suppression, and glucagon was completely ineffective in suppressing feeding in vagotomized animals. Although glucagon infusion in 4-hr food-deprived intact rabbits produced 38% suppression of food intake during the first hr postadministration, glycogen analysis revealed no significant reduction under the behavioral testing paradigm. These results indicate that glucagon can suppress food intake without depletion of liver glycogen. It is suggested that glucagon is not a satiety signal but can probably suppress feeding through initiating glycogenolysis.

The effects of taste adulteration, hypertonic, and hyperoncotic solutions on water ingestion in the gerbil

Mongolian gerbils (Meriones unguiculatus) were examined for components of thirst in the laboratory. The gerbil, reputed to be a desert-adapted rodent, was tested for its response to quinine adulterations of water, hypertonic solutions (1 M sodium chloride, NaCl), and hyperoncotic solutions (40% W/V polyethylene glycol in isotonic saline, PG). Both NaCl and PG were administered intraperitoneally. Gerbils accepted quinine when it was the only solution available and food was ad lib, but preferred distilled water to quinine if both were available. Both PG and NaCl produced increased water intakes in gerbils within the first 2 h, postinjection. PG, however, produced the most reliable and largest increase in water consumption, illustrating that extracellular dehydration is probably a more significant thirst stimulus in gerbils.

Dietary self-selection in cycling and neonatally ovariectomized rats.

Dietary self-selection patterns were assessed in normally cycling (four-day estrous cycle) rats with separate provisions of fat (olive oil), carbohydrate (sucrose), and protein (casein). There was no fluctuation in daily caloric intake across the estrous cycle, but animals did demonstrate changes in consumption of specific macronutrients during the 24-h period in which estrus occurred. At estrus, animals exhibited increased carbohydrate intake and decreased fat intake in comparison with the remaining three days of the cycle. There was also a somewhat decreased (non-significant) protein consumption at estrus. Rats that had been ovariectomized on day 5 postnatally and then tested as adults in our paradigm ingested approximately the same quantity of daily kilocalories as observed in intact animals; but ovariectomy produced changes in consumption of specific macronutrients. In comparison with the intact animals, ovariectomized animals displayed increased fat intake and decreased carbohydrate intake. Thus, the fat and carbohydrate self-selection patterns of our ovariectomized rats reflected the elimination of estrus behavior.

Insulin suppresses intake without inducing illness in sham feeding rats

Insulins effects on consumption were investigated in rats feeding with open gastric cannulas. Insulin produced small but statistically reliable decreases in sham intake. Intake was reduced during the one-hour sham feeding interval by 11 to 18%, however, the second, third and fourth 15-min intervals were reduced by 20 to 33%. An initial increase (first 15 min of sham feeding) following insulin injection offset somewhat insulins suppressive effects on sham intake. All reductions were obtained by comparing sham intake following 0.1 to 0.75 U insulin per rat to intake following saline. When insulin or saline injections were paired with alternatively flavored sweetened condensed milk during sham feeding, subjects preferred the insulin-paired flavor in a subsequent two-bottle test. After just 6 complete pairings, insulin-paired, flavored milk was preferred at a ratio of almost six to one. The results indicate the effectiveness of insulin in inhibiting intake in a situation in which the actions of other potential satiety mechanisms are minimized. Additionally, insulin accomplishes this without inducing a flavor aversion for the milk sham consumed.

Cholecystokinin and estradiol synergistically potentiate satiety in rats

The ability of octapeptide cholecystokinin (CCK), in interaction with ovarian steroid conditions, to decrease 1-h feeding was studied in 5-h food-deprived Sprague-Dawley rats. In Experiment 1, intact, unilaterally ovariectomized, and bilaterally ovariectomized females and intact males were given IP injections of 0, 0.25, 0.50, and 1.0 microgram/kg b.wt. CCK and were assessed for food ingestion at 1, 3, and 19 h. Food intake at 1 h was suppressed in animals receiving CCK compared to saline (0 dose); the threshold dosage was 1 microgram/kg b.wt. (p < 0.01) in this paradigm. No significant sex difference was observed between the four groups; however, animals with decreased ovarian steroids (intact males and bilaterally ovariectomized females) suppressed ingestion less than animals with greater ovarian steroid levels (intact and unilaterally ovariectomized females) at both the 0.25 and 1.0 microgram/kg b.wt. dosages (p < 0.01). Therefore, in a second experiment, sensitivity to CCK was compared in females in early metestrus, when estrogen levels are decreased, and during late diestrus, when estrogen levels are high, using dosages of 0, 0.25, 1, and 2.5 micrograms/kg b.wt. A statistically significant difference was found between sensitivity at early metestrus and late diestrus at the 2.5 micrograms/kg b.wt. dose only, with food ingestion more reliably depressed during periods of increased estrogen (p < 0.05). These results suggest that estradiol and CCK can have a synergistic effect on satiety.

Insulin and satiety from feeding in pancreatic-normal and diabetic rats

Insulins role in food ingestion and satiety was investigated in streptozotocin-diabetic and pancreatic-normal rats. Markedly diabetic rats (60-65 mg/kg b.wt. streptozotocin with mean glycemia of 380 mg/dl) were observed for daily food/water intake and body weight changes and meal pattern in a standard operant chamber. Diabetic animals showed an immediate hypophagia (days 1-4) by decreasing meal size (MS). As animals lost weight, a significant hyperphagia appeared, accomplished by an elevation in MS. Treatment with insulin infused via osmotic minipump at a stable rate (6.0 U/24 h) reduced polyuria and glycemia, eliminated glycosuria, and restored weight gain. MS did not, however, return to baseline immediately. In a second experiment with milder diabetes (20-22 mg/rat streptozotocin, which produced glycemia ranging from 148 to 304 mg/dl), significant hyperphagia appeared, again attributable to an increase in MS. Minipumps infusing 2.4-4.0 U insulin/24 h reversed this hyperphagia. In pancreatic-normal rats, pumps infusing insulin at 1.2 or 2.4 U/24 h produced a modest hypophagia accomplished by a primary decrease in nocturnal intake (11% suppression in dark feeding). Subdiaphragmatically vagotomized rats showed an attenuation of this suppression (no significant effects on food or water intake produced by insulin infused). Insulin appears to participate in satiety by limiting MS, without significantly shortening latency to take the next meal.

Glucagon, satiety from feeding and liver/pancreatic interactions

In an attempt to assess pancreatic glucagons efficacy at repeatedly reducing food ingestion during differing circadian periods, three groups of 8 rats each were randomly assigned to 4-hr food deprivations beginning at 0800, 1200 or 1600 with light off at 2000. Subjects were then refed following injections of pancreatic glucagon (400 micrograms/kg b.wt. dissolved in DMSO) or vehicle alone every third day (no injection on intervening day). Food intake was measured at 1 and 20 hr following each injection. Following 3 cycles of the above procedure, each animal was again food deprived at the appropriate time, stunned and sacrificed by decapitation. The liver was sampled and glycogen determinations were made. Glucagon suppressed food intake when injected at 1200 (49.6%) and at 1600 (43.1%) but not when given at 2000 (-2.2%). Glycogen content measured after similar deprivation ending at these times was 5.6, 3.9 and 2.0%, respectively. With repeated glucagon injections, the hormone lost its ability to reduce food intake. In a second study, designed to evaluate the role of insulin in glucagons action, three groups of 6 rats each were given atropine plus glucagon or glucagon or atropine injections alone; food ingestion was then measured one hr later. Atropine alone somewhat decreased eating, however, in combination with glucagon (given 10 min following atropine), no significant decrements in ingestion were achieved. Glucagon injected after saline produced a significant reduction in food intake (62.5%). Since glucagon stimulates insulin release and hyperglycemia; perhaps insulin release is necessary for glucagons satiety effect.

Lithium chloride and inescapable, unsignaled tail shock differentially affect meal patterns of rats

Several motivational states, such as malaise, fear, and satiety, reduce spontaneous food ingestion by rats, and differentiation of these states is often desirable. The present study used the spontaneous meal pattern to this end. The illness-inducing toxin lithium chloride delayed initiation of the first postinjection meal, and that meal was smaller and eaten more slowly (Experiments 1A and 1B). Rats exposed to tail shock also subsequently took longer to initiate meals, but meals were eaten faster and were slightly larger relative to control conditions (Experiment 2). These changes in meal patterns are different from those produced by satiety-related hormones, such as CCK or insulin, which solely reduce meal size in paradigms designed to assess physiological regulation of food intake. Taken together, these findings attest to the ability of meal patterns to distinguish malaise, fear, and satiety from one another.