The liver contains the only store of glycogen that is readily available for release into the circulation as glucose. For this reason, estimation of the hepatic concentration of glycogen in healthy and diabetic persons is of considerable importance. By combining ¹³C-MRS measurements of net hepatic glycogen content with magnetic resonance imaging (MRI) measurements of liver volume, rates of net hepatic glycogenolysis can be determined. The first accurate measurement on healthy young
humans demonstrated a glycogen concentration of 251 ± 30 mmol/L of liver after an overnight 15-hour fast (6), and other studies with similar nutritional preparation have observed similar values (7). Subjects with poorly controlled type 2 diabetes (mean fasting blood glucose 14.6 mmol/L) studied after a longer overnight fast were observed to have a significantly lower hepatic glycogen concentration than age- and weight-matched nondiabetic controls (97 ± 18 vs. 169 ± 43 mmol/L) (Fig. 12.5) (8). In subjects with type 1 diabetes, under fairly poor control, fasting concentration of liver glycogen was only slightly lower than in matched nondiabetic subjects (Fig. 12.6) (9).

During the overnight period, the hepatic glycogen stores of healthy subjects decrease to around 60% of the level achieved 4 hours after the evening meal (7,9). Direct measurement in two separate groups of healthy subjects has demonstrated that hepatic glycogen concentration decreases in a linear fashion during the night at an average rate of about 6 μmol/kg body weight per minute (10,11). Higher rates of glycogenolysis were associated with higher initial hepatic concentrations of glycogen. In the overnight-fasted state, about 45% of the isotopically measured rate of whole body glucose production can be accounted for by net hepatic glycogenolysis (Fig. 12.7).

Assessment of this process whereby the three-carbon substrates lactate, pyruvate, and glycerol are converted into glucose is beset with problems. Using hepatic venous catheterization to measure the uptake of gluconeogenic amino acids, pyruvate, and lactate across the splanchnic bed, Wahren et al. (12) estimated that the contribution of gluconeogenesis to total glucose production was only ∼20% during the first 12 to 14 hours of a fast. Consoli et al. (13,14) used [2-¹⁴C] acetate to calculate the ¹⁴C-specific activity of the precursor of glucose produced by gluconeogenesis (phosphoenolpyruvate) to estimate rates of hepatic gluconeogenesis and concluded that hepatic glycogenolysis accounted for 29% ± 2% of overall glucose production after an overnight fast. However, acetate is metabolized in extrahepatic tissues, invalidating this approach (12,15,16,17). Using [¹⁴C]lactate, Pimenta et al. (18) estimated that glycogenolysis (from glycogen synthesized by the direct pathway) accounted for ∼45.7% ± 4.7% and gluconeogenesis from lactate for a minimum of 18.6% ± 2.4% of overall hepatic glucose production. However, measurements of gluconeogenesis by the incorporation of label from [¹⁴C]lactate are complicated by the fact that the specific activity of intracellular lactate cannot be estimated from the specific activity of [¹⁴C]lactate in plasma (19).

Rates of gluconeogenesis can be calculated by combining ¹³C-NMR measurement of liver glycogen concentration with magnetic resonance imaging of liver volume and isotopic measurement of hepatic glucose production (HGP). Net hepatic glycogenolysis was determined by sequential measurement of hepatic glycogen concentrations by ¹³C-NMR, and rates of net hepatic glycogenolysis were calculated by multiplying this decrease in liver concentration over time by liver volume. Subtraction of these rates from rates of whole-glucose production determined by tracer turnover techniques yields rates of gluconeogenesis (6). Since the labeled carbon in position C6 of glucose is lost during carboxylation of pyruvate into oxaloacetate and during equilibration with dicarboxylic acids, contributions of glucose cycling between glucose and glucose-6-phosphate or
between fructose-6-phosphate and fructose-1,6-bisphosphate would not be included in the calculated rate of whole-body glucose production derived from this isotope (20). This approach, therefore, yields a minimum estimate for the rate of whole-body glucose production compared with [2-²H]glucose and [3-³H]glucose and a maximal estimate for the contribution of net hepatic glycogenolysis to the rate of whole body glucose production. Seven healthy volunteers were studied over a 68-hour period, starting with a 1,000-kcal liquid test meal at 5 p.m. Liver glycogen concentration fell from 396 ± 29 mmol/L at 4 hours following the meal to 251 ± 30 mmol/L at 15 hours and to 42 ± 9 mmol/L at 64 hours. Over the first 22 hours of fasting, the rate of decrease of liver glycogen concentration was linear at around 4 mmol/L per minute. After 22 hours of fasting, gluconeogenesis accounted for a mean of 64% of hepatic glucose output (Fig. 12.7).

To estimate the maximum contribution of net hepatic glycogenolysis to whole-body glucose production earlier in the postabsorptive period, studies were performed 5 to 12 hours after the ingestion of a large evening meal. At this time, hepatic glycogen stores were found to be at their maximum daily level (9). [6,6²H]Glucose was used for measuring rates of whole-body glucose production. Under these experimental conditions, net hepatic glycogenolysis contributed 43% ± 5% to whole-body glucose production during the first 12 hours after a meal. This is consistent with earlier NMR data and implies that gluconeogenesis typically contributes ∼50% to glucose production even under circumstances when hepatic glycogen stores are at their daily maximal concentration (21,22). This is significantly greater than previous estimates (12,13,18). This result has been confirmed by the more recently developed ²H₂O method (23).

Because of the methodologic difficulties in measuring hepatic glycogen content in vivo, few other studies have directly examined the relative contributions of hepatic glycogenolysis to glucose production in humans. Nilsson and Hultman (24) measured hepatic glycogen concentrations by biopsy in nine subjects 12 hours into a fast and once again 4 hours later and found that the rate of net hepatic glycogenolysis was 288 ± 30 μmol/kg wet liver tissue per minute. This rate is ∼30% higher than the MRS-determined rate of 225 μmol/L liver per minute and might be explained by the stress response associated with needle biopsy. These workers, assuming that liver volume was 1.8 L, calculated that net hepatic glycogenolysis contributed ∼57% to 67% to the rate of whole-body glucose production (24). If the data of Nilsson and Hultman (24) are recalculated with a liver volume of 1.5 L, as found by NMR imaging, the contribution of net hepatic glycogenolysis would be 47% to 56% of the rate of whole-body glucose production, a figure similar to the findings by NMR spectroscopy.

Since the glycogen stores in the liver are typically limited to approximately 70 g after an overnight fast, and glucose is utilized at a rate of approximately 7 to 10 g per hour, hepatic glycogen stores are largely dissipated early in the course of a prolonged fast (6). Thus, the initial phase of starvation is characterized by an increase in the relative contribution of gluconeogenesis to meet the ongoing demands for glucose by tissues (mainly from the central nervous system). The importance of glucose production from protein is reflected by an increase in urinary excretion of nitrogen in the early phase of starvation (25). During the first 24 hours of a fast, urinary excretion of nitrogen averages 4 to 7 g/m² in the absence of nitrogen intake (26). This amounts to approximately 50 to 75 g of protein (∼1 g N/6.25 g protein) for the average 70-kg person. Similar results are obtained with isotopic tracer methods for estimating total-body rates of protein oxidation (27). Therefore, regardless of whether simple urinary nitrogen balance is measured or a tracer method is used, the results suggest that about 50 g of protein is oxidized in the 24 hours following a meal. Since the protein content of tissues does not exceed 20% by weight for any tissue, 50 to 75 g of protein translates into 250 to 375 g of tissue or the equivalent of 1% to 1.5% of lean body mass lost on the first day of a fast. In muscle, the release of alanine and other glycogenic amino acids increases in conjunction with an acceleration of proteolysis (27). At the same time, the rate of hepatic conversion of gluconeogenic amino acids into glucose is augmented (6). The increase in glucose synthesis from amino acids is not, however, due solely to increased availability of precursor, since plasma levels of alanine and other glycogenic amino acids decline (28). Taken together, these observations suggest that intrahepatic gluconeogenic mechanisms are stimulated as well, enhancing the efficiency of this process.

The major metabolic changes that occur during the first 60 hours of starvation have been quantified by a combination of MRS and isotopic methods (6,10). Figure 12.7 draws on data from two different studies. The whole-body utilization of glucose falls by 25% from the overnight-fasted state, and the proportion of HGP contributed by gluconeogenesis rises from 55% at 12 hours to 96% at 60 hours.

The metabolic adaptations in the early stages of starvation (namely increased gluconeogenesis, proteolysis, and lipolysis) coincide with and are orchestrated primarily by a decline in insulin secretion to below levels in the overnight fasted condition and by a modest increase in circulating levels of glucagon (6,29). Insulin deficiency promotes all aspects of the metabolic response, whereas the effect of glucagon is confined to the liver. Specifically, the decline in insulin promotes the breakdown of protein and fat and thus the delivery of gluconeogenic amino acids from muscle and glycerol from fat stores to the liver, whereas the activation of intrahepatic gluconeogenic processes is promoted by both the increase in glucagon and the decrease in insulin in the portal circulation. Hypoinsulinemia further contributes to glucose homeostasis by reducing glucose metabolism by extracerebral tissues (e.g., muscle) and by increasing the availability of free fatty acids for oxidative metabolism by muscle and liver and ketones for muscle and the central nervous system.