One of the difficulties in managing type 1 diabetes (T1D) stems from the simple fact that exogenous insulin injections or pumped insulin does not and cannot mimic normal pancreatic beta-cell insulin secretion. The autoimmune attack characteristic of T1D leads to destruction of the glucose-sensing beta-cells and their pulsatile insulin, amylin, and zinc secretion. This results in loss of control of glucagon secretion in the neighboring alpha-cells and a diminution of insulin reaching the liver to control hepatic glucose production. I think knowing this is helpful to persons with T1D who are conscientiously caring for their condition so as not to become frustrated at the difficulty of keeping blood glucose in the normal range even with a low carbohydrate high fat ketogenic (KLCHF) diet.
In this blog post, I take a closer look at the cells of and hormones produced in the pancreatic islets of Langerhans to give you a more complete understanding of why exogenous insulin is not enough to restore normal glycemic control. The following figure is from this review of alpha-cell function in T1D.
There are five different endocrine cells in the islets of Langerhans: the alpha-cells produce glucagon, the beta-cells produce insulin, amylin, and zinc, the delta cells produce somatostatin, the epsilon cells produce ghrelin, and the pancreatic polypeptide-producing cells produce, you guessed it, pancreatic polypeptide. The control of nutrient disposition and blood glucose is also influenced by the brain, autonomic nervous system, adipocytes (leptin), and intestinal incretins (e.g., glucose-dependent insulinotropic peptide, glucagon-like peptide-1, peptide YY). Today, I’ll be focusing on the pancreatic islet cells.
The Alpha and Beta Islet Cells
Starting with the beta-cell, you likely know that insulin is an anabolic hormone that directs the body to utilize glucose over fatty acids and helps to store excess glucose in muscle and liver in the form of glycogen and in adipose tissue (fat cells) as triglycerides. Beta-cells are glucose-sensing cells that secrete insulin in response to the glucose concentration surrounding the cell. Insulin is released in pulses throughout the day and night (basal insulin secretion) and in larger amounts in response to meals. The figure below shows insulin secretion for 24 hours in normal and obese persons from this study.
Note that half of the total daily insulin secretion occurs after the three meals (daily calories distributed as 20% breakfast, 40% lunch, 40% dinner) and half during the periods between meals and overnight. The macronutrient composition of the meals in this study consisted of 50% carbs, 35% fat, and 15% protein. Also note that the amount of insulin secreted after breakfast was similar to that after lunch and dinner despite the fact that breakfast had half the calories of lunch and dinner. This may be due to nocturnal surges in growth hormone secretion which leads to increased early morning hepatic glucose production (dawn phenomenon). Although obesity is not discussed in this blog post, the study showed that the obese subjects had a larger insulin response to meals and a higher rate of basal insulin secretion. This would be consistent with insulin resistance which is a common feature of obesity. The elevated insulin levels contribute to obesity since insulin stimulates the entry of fatty acids into adipocytes via lipoprotein lipase and inhibits the exit of fatty acids from adipocytes via hormone sensitive lipase. I’m showing you this study because it is typical of persons with T1D on a 50% carbohydrate diet to require half their insulin as mealtime insulin and half as basal insulin, similar to non-diabetics.
Amylin (hormone) and zinc (mineral) are cosecreted with insulin by the beta-cells. Amylin and zinc inhibit glucagon secretion in the neighboring alpha-cells, and amylin delays gastric emptying and acts as a satiety agent. This is important because after a meal when insulin and amylin secretion increases in response to rising blood glucose levels, the alpha-cells will be instructed to decrease glucagon secretion which in turn will instruct the liver to stop making glucose via both glycogenolysis and gluconeogenesis. This inhibition of glucagon secretion is greatly impaired in persons with T1D often resulting in postprandial hyperglycemia. This is because the concentration of insulin surrounding the alpha-cells is about 100 times higher in healthy persons than that in persons with T1D injecting insulin and amylin is nonexistent. In addition, the concentration of insulin in the blood leaving the pancreas and arriving at the liver through the portal vein is about 3 times higher in healthy persons than in persons with T1D injecting insulin. So both the lack of inhibition of glucagon and the lower concentration of insulin arriving to the liver contributes to hyperglycemia via both glycogenolysis and gluconeogenesis as well as to activation of free fatty acid oxidation and production of ketones. Another consequence of the loss of beta-cells in T1D is the impaired response to hypoglycemia. In healthy persons, beta-cells respond to hypoglycemia by decreasing insulin secretion. The neighboring alpha-cells respond by increasing glucagon secretion to raise blood glucose via both glycogenolysis and gluconeogenesis in the liver. Whereas in T1D, there is no reduction in insulin around the alpha-cells and thus little to no increase in glucagon to help correct the hypoglycemia. This defective glucose counter-regulation can be ameliorated by the sympathetic nervous system, but as discussed in blog post #12, the sympathetic response to hypoglycemia can also become impaired in T1D.
Some investigators, like Dr. Robert Unger, argue that it is glucagon excess, rather than insulin deficiency, that is the primary feature of diabetes. His conclusions are largely based on animal studies. There are significant differences in the number and distribution of pancreatic islet cells between different species, however. This hypothesis would predict that inhibitors of glucagon secretion would improve glycemic control in T1D. One such inhibitor of glucagon secretion is the amylin analog, pramlintide.
In this review of three human clinical trials using the amylin analog, pramlintide, with meals along with insulin for T1D found that the HbA1c was reduced by 0.4% over 26 weeks compared to a 0.1% reduction in those with T1D receiving placebo and insulin. The placebo group required a 5% increase in insulin dose to achieve that HbA1c reduction, whereas the pramlintide group experienced a less than 1% reduction in insulin dose. Regarding change in weight, the placebo group gained about 0.8 kg, whereas the pramlintide group lost about 1.1 kg in weight by 26 weeks. Severe hypoglycemia occurred about 2.5 times more often in the pramlintide group compared to placebo when mealtime insulin doses were not adjusted after adding pramlintide. When mealtime insulin dose was preemptively reduced by 50%, the rate of severe hypoglycemia was quite low (0.05 events/patient year). The only other significant side-effect of pramlintide reported was nausea.
My Thoughts on Pramlintide
As mentioned above, the beta-cell hormone, amylin, is no longer produced in persons with T1D due to autoimmune destruction of the beta-cells. Amylin has three effects: it inhibits glucagon secretion in the neighboring alpha-cells, it slows gastric emptying which has the effect of slowing the rate of nutrient absorption (glucose, amino acids, fatty acids), and contributes to satiety. Because a KLCHF diet is very low in glucose and high in fat, it actually mimics the latter two actions of amylin. Pramlintide targets the rise in postprandial glucose concentrations after high carbohydrate meals presumably used in the studies mentioned above. So in my opinion in those following a KLCHF diet, there would be little value in adding pramlintide to an insulin regimen for T1D. The KLCHF diet reduces the postprandial glucose rise which is what pramlintide is designed to address. The KLCHF diet also is known to reduce appetite (increase satiety) although the mechanism of this phenomenon is not exactly known. The KLCHF diet results in a significant reduction in insulin requirements due to the replacement of dietary carbohydrates with dietary fat. The majority of the insulin dose reduction is in mealtime insulin. For example, 90% of the reduction in my insulin dose after starting my KLCHF diet was in mealtime insulin, i.e. the mealtime insulin dose decreased from 28.9 to 9.7 IU/day, whereas basal insulin dose decreased from 23.3 to 21.3 IU/day. Finally, a KLCHF diet accomplishes a larger reduction in HbA1c without the additional cost of an injectable medication. In this study, 23 of 48 (48%) persons with T1D adherent to a 75 gram/day dietary carbohydrate restriction, the mean HbA1c was at start, at 3 months, and 4 years 7.7 ± 1.0%, 6.4 ± 0.9%, and 6.4 ± 0.8% respectively. That’s a reduction of 1.3% in HbA1c compared to 0.4% with pramlintide.
The Other Three Islet Cells
Although less is known about the roles of somatostatin, ghrelin, and pancreatic polypeptide in the pancreatic islets, they are briefly discussed below for completeness sake.
Somatostatin is secreted by pancreatic islet delta-cells and by extraislet neuroendocrine cells. Somatostatin receptors have been identified on alpha- and beta-cells, and exogenous somatostatin inhibits insulin and glucagon secretion, consistent with a role for somatostatin in regulating alpha- and beta-cell function. Somatostatin plays a role in nutrient-induced suppression of glucagon secretion. However, the specific intraislet function of delta-cell somatostatin remains uncertain. Further information can be found here.
Although ghrelin is primarily secreted by stomach cells, it is also secreted by epsilon cells in the pancreatic islet and could act as a paracrine inhibitor of insulin secretion. Some reports have shown stimulatory effects, however. Additional information about ghrelin can be found here.
Pancreatic polypeptide (PP) exhibits a rapid increase after food ingestion peaking at 15-30 min which is followed by a lower sustained phase that lasts 4-5 hours in humans. Protein is the most potent stimulus followed by fat with glucose being the least effective stimulus. The vagal nerve is a major stimulator of PP secretion. Somatostatin inhibits PP release in the pancreatic islets. Further information can be found here.
There are multiple islet cells producing multiple endocrine and paracrine hormones along with intestinal incretins and nervous system signals to tightly regulate blood glucose during both fasting and fed states that cannot be precisely mimicked with exogenous insulin injections in those with T1D. However, the addition of the KLCHF diet to exogenous insulin analogs has a significant positive effect in helping to regulate blood glucose in persons with T1D. When dietary carbohydrate (i.e. glucose) comprises a large percentage of caloric intake, it is difficult, if not impossible, to accurately balance the resulting glucose load with exogenous insulin. By limiting dietary carbohydrate, the KLCHF diet results in improved (near-normal) blood glucose with less glucose variability. The second benefit that results from the KLCHF diet is nutritional ketosis that has the potential to provide an alternate fuel (ketones) to the brain during those inevitable times when iatrogenic (exogenous insulin-induced) hypoglycemia occurs. This prevents or minimizes the symptoms of hypoglycemia because the brain (and body) are not totally dependent on glucose for energy and thus there is no apparent emergency that needs to be addressed by the sympathetic nervous system which when activated causes the symptoms of hypoglycemia. I made a cautionary note in blog post #12 that some persons with T1D may be experiencing decreased symptoms of hypoglycemia due to hypoglycemia unawareness as part of hypoglycemia-associated autonomic failure. This is a potentially dangerous condition that all persons with T1D should be familiar with so I encourage you to read blog post #12. However even in those with T1D who choose to follow a KLCHF diet, there will be unpredictable glucose excursions with exogenous insulin for the reasons outlined above. The strategies I use to deal with this include frequent blood glucose measurements, use of insulin analogs (Lantus and Humalog), regular exercise to maintain insulin sensitivity resulting in lower insulin requirements, consistency in the macronutrient content of my ketogenic meals, occasional correction doses of rapid-acting insulin (Humalog) for hyperglycemia, and occasional doses of glucose tablet(s) for hypoglycemia.
The next blog post will be my monthly (January 2015) update on my blood glucose results.