Suppression of insulin secretion by protein deprivation in obesity

Driver versus navigator causation in biology: the case of insulin and fasting glucose

BACKGROUND

In biomedicine, inferring causal relation from experimental intervention or perturbation is believed to be a more reliable approach than inferring causation from cross-sectional correlation. However, we point out here that even in interventional inference there are logical traps. In homeostatic systems, causality in a steady state can be qualitatively different from that in a perturbed state. On a broader scale there is a need to differentiate driver causality from navigator causality. A driver is essential for reaching a destination but may not have any role in deciding the destination. A navigator on the other hand has a role in deciding the destination and the path but may not be able to drive the system to the destination. The failure to differentiate between types of causalities is likely to have resulted into many misinterpretations in physiology and biomedicine.

METHODS

We illustrate this by critically re-examining a specific case of the causal role of insulin in glucose homeostasis using five different approaches (1) Systematic review of tissue specific insulin receptor knock-outs, (2) Systematic review of insulin suppression and insulin enhancement experiments, (3) Differentiating steady state and post-meal state glucose levels in streptozotocin treated rats in primary experiments, (4) Mathematical and theoretical considerations and (5) Glucose-insulin relationship in human epidemiological data.

RESULTS

All the approaches converge on the inference that although insulin action hastens the return to a steady state after a glucose load, there is no evidence that insulin action determines the steady state level of glucose. Insulin, unlike the popular belief in medicine, appears to be a driver but not a navigator for steady state glucose level. It is quite likely therefore that the current line of clinical action in the field of type 2 diabetes has limited success largely because it is based on a misinterpretation of glucose-insulin relationship. The insulin-glucose example suggests that we may have to carefully re-examine causal inferences from perturbation experiments and set up revised norms for experimental design for causal inference.

Effect of diet composition on insulin sensitivity in humans

Diet composition has a marked impact on the risk of developing type 2 diabetes and cardiovascular disease. Prospective studies show that dietary patterns with elevated amount of animal products and low quantity of vegetable food items raise the risk of these diseases. In healthy subjects, animal protein intake intensifies insulin resistance whereas plant-based foods enhance insulin sensitivity. Similar effects have been documented in patients with diabetes. Accordingly, pre-pregnancy intake of meat (processed and unprocessed) has been strongly associated with a higher risk of gestational diabetes whereas greater pre-pregnancy vegetable protein consumption is associated with a lower risk of gestational diabetes. Population groups that modify their traditional dietary habit increasing the amount of animal products while reducing plant-based foods experience a remarkable rise in the frequency of type 2 diabetes. The association of animal protein intake with insulin resistance is independent of body mass index. In obese individuals that consume high animal protein diets, insulin sensitivity does not improve following weight loss. Diets aimed to lose weight that encourage restriction of carbohydrates and elevated consumption of animal protein intensify insulin resistance increasing the risk of developing type 2 diabetes and cardiovascular disease. The effect of dietary components on insulin sensitivity may contribute to explain the striking impact of eating habits on the risk of type 2 diabetes and cardiovascular disease. Insulin resistance predisposes to type 2 diabetes in healthy subjects and deteriorates metabolic control in patients with diabetes. In nondiabetic and diabetic individuals, insulin resistance is a major cardiovascular risk factor.

Syndrome-AC: non-insulin-dependent diabetes mellitus and the anabolic/catabolic paradox

When an organism is starving, infected or injured, the body initiates a catabolic response that, among other things, creates insulin resistance. While many possible mechanisms and numerous loci of insulin resistance have been studied, a theme of inappropriately activated components of catabolic chemistry emerges. Patients with non-insulin-dependent diabetes mellitus (NIDDM) have elevated blood levels of glucagon, cortisol, fatty acids, protein, glucose and possibly acute phase reactants. Recent studies hint that this can occur with meals, resulting in a multilevel, multiorgan interference with glucose handling. This implies NIDDM is the result of the dietary activation of catabolic chemistry simultaneously with that of anabolic chemistry. We review the possibility that this is caused by the consumption of body tissue, the substance ordinarily endogenously released in starvation or injury. Activating the catabolic pathways when eating creates a hormonal paradox, forcing exaggerated insulin levels to compensate. Five case studies are reviewed.

Effects of dietary protein restriction on glucose and insulin metabolism in normal and diabetic humans

We determined whether the amount of protein in the diet can affect insulin requirements in subjects with diabetes mellitus and glucose metabolism in normal subjects. Seven normal-weight volunteers with uncomplicated, intensively controlled, type I (insulin-dependent) diabetes and 12 similar nondiabetic subjects were studied on a metabolic ward before and after consuming a maintenance-energy but protein-free diet for 10 days. Blood glucose levels of diabetic subjects were measured seven times daily in response to insulin administration by continuous subcutaneous infusion. The plasma glucose appearance rate (Ra) was measured in seven normal subjects and all diabetic subjects using a primed-continuous infusion of D-[6,6-2H2]glucose. After adaptation to the protein-restricted diet, diabetic subjects experienced a 30% decrease in average preprandial and average daily blood glucose concentrations (P < .01); this occurred despite a concurrent 25% decrease in both basal and bolus insulin dosages (P < .001). Protein restriction decreased the postabsorptive glucose Ra (P < .05) and insulin concentrations (P < .01) of normal subjects by 20%, and increased their fasting glucagon concentrations by 24% (P < .01). We conclude that severe protein restriction decreases insulin requirements in type I diabetes and fasting hepatic glucose output and basal insulin levels in normal subjects. This effect appears to be mediated in part by decreased hepatic gluconeogenesis, but a contributory influence of increased insulin sensitivity is not ruled out.

Endocrine aspects of obesity

As is obvious from the previous discussions, obesity is associated with a wide variety of changes in endocrine parameters (Table 1). Some of these changes, such as the reduction in SHBG without change in serum free testosterone levels, reflect merely laboratory abnormalities that may influence interpretation of diagnostic tests but have no important physiologic relevance. Other abnormalities have major clinical impact, such as hyperestrogenemia-endometrial carcinoma and hyperlipidemia-coronary artery disease. In some cases, endocrine changes in obesity are beneficial--that is, hyperestrogenemia leading to lower incidence of osteoporosis. In other cases, such as the profound suppression of growth hormone output in obesity, the physiologic relevance is unknown. Several endocrine changes in obesity, such as the impaired response of many hormones (growth hormone, prolactin, vasopressin, corticotropin) to insulin-induced hypoglycemia and elevated endorphin levels, suggest hypothalamic dysfunction. Furthermore, the failure of all of these abnormalities to be normalized after weight reduction raises the possibility of an underlying disorder leading to both endocrine dysfunction and obesity, rather than the endocrine dysfunction being simply a consequence of the obesity. Successful elucidation of the pathogenesis of obesity, which might then lead to much needed specific treatment modalities, may be advanced if we can solve some of these puzzles.

Suppression of basal, but not of glucose-stimulated insulin secretion by human insulin in healthy and obese hyperinsulinemic subjects

To evaluate the suppressive effect of biosynthetic human insulin (BHI; 2.5 U/m2 . h) on basal and glucose-stimulated insulin secretion in healthy and obese hyperinsulinemic subjects, the plasma C-peptide response was measured during maintenance of euglycemia and hyperglycemia by means of the glucose clamp technique. In five healthy subjects in whom arterial insulin concentration was increased to 94 +/- 8 microU/mL, but euglycemia was maintained at the fasting level. C-peptide concentration fell from 1.3 +/- 1.0 ng/mL by 21 +/- 8% (P less than 0.05). When hyperglycemia of 7 mmol/L above basal was induced by a variable glucose infusion, the C-peptide response was similar in the control (5.0 +/- 0.6 ng/mL) and BHI experiments (4.7 +/- 0.6 ng/mL) and was paralleled by an identical increase in plasma insulin above the prevailing insulin concentration. In seven obese patients plasma C-peptide fell from 3.5 +/- 0.4 to 2.8 +/- 0.5 ng/mL (P less than 0.05) when BHI was infused at the same rate of euglycemia maintained as in the lean subjects. As in healthy subjects, however, the plasma C-peptide response to the hyperglycemic stimulus (8.7 +/- 0.9 ng/mL) was not altered by BHI (7.9 +/- 0.8 ng/mL). Glucose utilization as determined by the glucose infusion rate necessary to maintain the desired glucose level was reduced by half in the obese patients compared with that of normal subjects. From these data we conclude that in healthy as well as obese hyperinsulinemic subjects, insulin at concentrations capable of suppressing its basal secretion fails to suppress its glucose-stimulated secretion.

Improved carbohydrate tolerance and stimulation of carbohydrate oxidation and lipogenesis during short-term carbohydrate overfeeding

The carbohydrate intake of seven healthy men was increased from 220-265 g/d to 620-770 g/d for 17 days, while protein and fat intake remained constant. Carbohydrate loading did not affect the preprandial plasma glucose levels after an overnight fast, but reduced the postprandial increment in plasma glucose levels after 5, 11, and 17 days of overfeeding. Preprandial plasma insulin levels were slightly increased during carbohydrate overfeeding, but no increase in the postprandial rise in insulin levels was found until 11 days after the start of carbohydrate loading. Whole-body rates of carbohydrate oxidation and of glucose conversion to fat were estimated by indirect calorimetry. Basal carbohydrate oxidation rate was increased by 95% at the end of 17 days of overfeeding, but there was no potentiation in the stimulation of the carbohydrate oxidation rate induced by a meal. There was no net fat synthesis from glucose before carbohydrate loading; carbohydrate overfeeding produced nonprotein respiratory exchange ratios greater than 1.00, suggesting net fat synthesis from glucose. Meals did not stimulate net lipogenesis from glucose, either before or after overfeeding. These results indicate that the improvement in carbohydrate tolerance associated with short-term carbohydrate loading does not appear to depend on elevated insulin levels. Increased carbohydrate oxidation and lipogenesis elevated carbohydrate disposal is more than necessary to account for the improvement in carbohydrate tolerance.

Preservation of glucose tolerance and insulin secretory response to repeated glucose levels by the feeding of minimal glucose during prolonged fasting

The facilitation of glucose disposal (Staub-Traugott effect) and potentiation of serum insulin (IRI) concentration normally occurring after closely spaced intravenous glucose loads, are known to disappear after prolonged starvation. To study the effects of minimal amounts of glucose during fasting upon the insulin response and disposal of repeated intravenous glucose tolerance tests, obese volunteers were fasted for a mean of 25 +/- 2 days, while receiving either 8 or 16 gm of oral glucose every 6 hr, and compared to totally fasted subjects without glucose supplementation. Weight loss rate and the fall in basal IRI and glucose levels were similar to those of totally fasted subjects. However, the Staub-Traugott effect and insulin secretory dynamics after stimulation by repetitive intravenous glucose loading were preserved by this glucose modified fast, while baseline serum glucagon levels (IRG) were significantly lower, and the basal IRI/IRG ratios were thus unchanged from the fed state. IRG and free fatty acid suppression were similar in the fed and glucose modified fasted states. Lactic acid levels increased as expected after the repeated glucose injections in the fed state, but failed to do so after the prolonged modified fast until the second and third repetitive glucose loads, in which a significant rise coincided with accelerated glucose disposal. It is suggested that minimal amounts of carbohydrate during fasting preserve the insulin potentiating action of glucose, preferentially sparing a delayed releasable pool of insulin, while protecting the glucose utilization mechanisms, including increased glycolysis, responsible for the Staub-Traugott effect.

Dietary control of pathogenesis in C57BL/KsJ db/db diabetes mice

Weanling C57BL/KsJ homozygous diabetic (db/db) and normal littermate (+/+ or +/db) mice were maintained for 5 mon on isocaloric diets containing either 60% sucrose, 23% casein, 8% corn oil (diet C) or 0% sucrose, 83% casein, 8% corn oil (diet B). Diabetic homozygotes consumed more diet C than normals, but ate control amounts of diet B. Diabetic mice fed diet C exhibited 57% mortality between 4 or 5 mo of age. All diabetic mutants fed the carbohydrate-free diet B appeared healthy at 6 mo of age; mutant females were normoglycemic and mutant males were only moderately hyperglycemic. Histological examination of pancreatic islets confirmed the absence of islet degeneration. In diet B maintained mutants, increased carcass fat composition, plasma and pancreatic content of insulin and glucagon, and thymidine incorporation into islets, all established that the db gene was being fully expressed. These results indicate that dietary protein stimulates islet growth and function in db/db mice, while high levels of refined carbohydrate in the diet predispose islet beta cells to undefined changes that culminate in necrosis. Restricting mutants' intake of a carbohydrate-containing diet to one-half the caloric intake of normal mice failed to block onset of beta cell necrosis. Thus, dietary composition rather than total caloric intake appears to be critical in the induction of islet necrosis and atrophy in this animal model of genetically transmitted diabetes.