Our results indicate that chronic consumption of the viscous, non-fermentable fiber HPMC can decrease diabetic wasting, improve insulin resistance and reduce the development of fatty liver in a model of obesity with type 2 diabetes. The HPMC-containing diets delayed the absorption of glucose by the intestine, as indicated by the decreased postprandial glucose curve after a meal. Since it has been shown that viscous fibers decrease the postprandial glucose curve and possibly ameliorate insulin resistance in normal mice and hamsters [33, 34], we assessed their effect in a model of obesity with diabetes, the ZDF rat.
The consumption of HPMC by ZDF rats slowed the progression of the diabetic phenotype as evidenced by improved glucose control and decreased insulin resistance as well as reducing other associated conditions such as oxidative stress and glycosuria. The decreased plasma glucose tAUC during the oral glucose tolerance test and percent glycated hemoglobin in both HPMC groups, coupled with an increased QUICKI index in the HV-HPMC group, indicate greater glucose control and less insulin resistance and demonstrate that a high intestinal contents viscosity limits the progression of insulin resistance in this animal model. Moreover, there was a significant viscosity-dependent decrease in relative kidney weight during consumption of the HPMC diets when compared to the obese control. During the initial stage of diabetes, renal hypertrophy occurs proportional to glycemic control , which is considered an early stage in the development of diabetic nephropathy. The only marker not to indicate a statistically significant improvement in glucose control with viscous fiber was the iAUC. However, the values were highly variable, due in part to large differences in fasting plasma glucose. Thus, the iAUC may not capture differences in insulin resistance as well as other measures. In this model of advanced type 2 diabetes, it appears that the tAUC may be a better predictor of insulin resistance. Thus, in the ZDF rat, consumption of a viscous fiber greatly improves glycemic control and reduces insulin resistance, and appears to do so in proportion to intestinal contents viscosity.
Paradoxically, the HV-HPMC group had a greater final body weight but lower food intake than the obese control group. Others have reported treatments in ZDF rats that decreased food intake but led to either increased  or no change in body weight . As indicated by the food efficiency ratio of the four groups, the HV-HPMC group was able to more efficiently use the energy consumed compared to the obese control group. The food efficiency ratio was not different during the first week but the obese control was significantly lower than the HV-HPMC groups in weeks 2–5. This could be due to either increased energy expenditure, decreased intestinal absorption of macronutrients, or increased excretion of energy in the obese control group. It is not apparent how HPMC treatment would decrease energy expenditure or increase absorption of macronutrients relative to the obese control group. Therefore it is more likely that the obese control group lost more energy in the urine. This may be explained by the progression of insulin resistance resulting in increased excretion of glucose and ketone bodies in the urine in the obese control group, a known result of untreated type 2 diabetes. The difference in food efficiency ratio did not result in a body weight difference in weeks 2 and 3 but the obese control and HV-HPMC groups gradually separated and were significantly different in the last two weeks. This suggests that the HV-HPMC had decreased diabetic wasting and experienced normalized body growth while the obese control group could not maintain a normal growth curve. As expected, the increased small intestinal viscosity from HPMC was inversely related to the 24-hour urinary excretion of glucose (r=−0.54, p=0.001, as logarithm of viscosity vs. glucose excretion) and β-hydroxybutyrate (r=−0.41, p=0.02, as logarithm of viscosity vs. β-hydroxybutyrate excretion) in the ZDF groups. However, given the magnitude of the difference in food intake, other factors, such as differences in physical activity, may be involved.
A decreased postprandial glucose response will reduce the plasma insulin response, which may lead to reduced tissue lipid accumulation by decreasing lipogenesis or increasing fatty acid β-oxidation. Differences in visceral fat pad weight, while statistically significant, were small and would have contributed little to differences in body weight. However, the HV-HPMC group, which was significantly heavier than the obese control, had the lightest visceral fat pad weight as a percent of final body weight, indicating a change in body composition. Similarly, Syrian hamsters on a high fat diet supplemented with HPMC also had reduced abdominal fat with no change in body weight, further supporting an effect of HPMC on reducing adiposity .
The circulating concentration of NEFA has been postulated to play a role in muscle insulin resistance, possibly through oxidative stress and mitochondrial dysfunction . Neither fasting nor fed NEFA levels differed among the three ZDF groups despite large differences in insulin resistance, suggesting that the increased insulin sensitivity may not be directly related to plasma NEFA in this model, but rather by other factors such as circulating adipokines. One adipokine, leptin, is typically positively correlated with fat mass , however in the ZDF model of extreme insulin resistance and a defective leptin receptor, this correlation is lost , a finding confirmed in the present experiment (r=0.0703, p=0.65). Adiponectin is another circulating adipokine that correlates well with whole-body insulin sensitivity  and is decreased in subjects with type 2 diabetes [41, 42]. In a cross-sectional study, cereal fiber intake associated with higher levels of plasma adiponectin in diabetic men  and Zucker rats consuming soluble cocoa fiber had higher levels of adiponectin compared to rats on a diet containing only cellulose . In this study, the HV-HPMC group had the highest concentration of adiponectin and greatest insulin sensitivity. Adiponectin is thought to inhibit hepatic gluconeogenesis and increase fatty acid oxidation in the muscle through increased AMP kinase and PPARα activity , yet we saw only a tendency for a difference in hepatic expression of G6Pase, a trend for a decrease in PEPCK and no difference in CPT-1β expression or acylcarnitine concentration in the muscle.
Livers from the obese control group contained considerably more lipid than those of the lean control group, indicating hepatic steatosis, as reported by others in this animal model . Hepatic steatosis is a result of an imbalance in fatty acid uptake or synthesis versus fatty acid oxidation or export via VLDL. It is considered the first step towards development of nonalcoholic fatty liver disease. This ectopic accumulation of lipid has been strongly linked to increased insulin resistance [47–49], as was found in the obese control group in the present study. The HV-HPMC group had significantly less total liver lipids compared to the obese control group, as well as reduced insulin resistance, measured by both the QUICKI index and the glucose tolerance test, suggesting that the reduction in insulin resistance in this group led to a reduction in hepatic lipid accumulation. Low plasma adiponectin concentrations have been linked to insulin resistance [50, 51], and plasma adiponectin concentrations are inversely related to hepatic steatosis [52–54], although some evidence indicates that the effect of adiponectin on hepatic steatosis is independent of insulin resistance [52, 53]. The HV-HPMC group, which had the lowest insulin resistance and least hepatic steatosis, also displayed the highest plasma adiponectin concentrations. The antisteatotic effect of adiponectin, mediated through the AdipoR1 and R2 receptors , appears to be due to activation of AMP kinase, leading to increased fatty acid oxidation . Increases in hepatic CPT1α, the rate-limiting enzyme in β-oxidation , resulting in increased fatty acid oxidation, have been shown to reduce liver TAG in both lean and obese rats . Activation of AMP kinase also decreases expression of hepatic gluconeogenic enzymes such as PEPCK and G6Pase . This is consistent with findings from the present study, in which the HV-HPMC group, with the highest plasma adiponectin, had the lowest hepatic lipid concentration, the highest hepatic expression of CPT1α, and a trend towards a reduction in the gluconeogenic enzymes PEPCK and G6Pase. However, no differences were found in the expression of FAS or of SREBP-1c, a transcription factor regulating expression of lipogenic genes, in the HV-HPMC group compared to the obese control. Others have reported decreased hepatic gene expression of FAS and SREBP-1c in Syrian hamsters fed HPMC . However, these animals were not insulin resistant. Given the lack of differences between the obese control group and the HV-HPMC group in plasma fatty acids (in either the fasted or fed state), in plasma TAG, or in markers of hepatic lipogenesis, coupled with greater expression of CPT1α in the HV-HMPC group, it seems most likely that an increase in hepatic fatty acid oxidation in the HV-HPMC group is responsible for the observed decrease in hepatic lipid concentration in this group.
One current theory of the progression of skeletal muscle insulin resistance in diabetes is that accumulation of intramuscular lipids will disrupt insulin signaling pathways and decrease glucose uptake . It is now believed that it is not the accumulation of triacylglycerols in the muscle tissue that is the cause of insulin resistance, but rather the generation of lipid metabolites such as ceramides, diacylglycerols and acylcarnitines that produces insulin resistance . In the first and rate-limiting step of β-oxidation, fatty acyl-CoAs are attached to carnitine by the enzyme CPT-1β, allowing transport through the mitochondrial membrane . However, if the energy state in the cell is high, enzymes in the electron transport chain may not increase activity sufficiently to compensate for the increased influx of acylcarnitines via CPT-1β . As a result, the concentration of intracellular acylcarnitines increases. This elevated concentration has been proposed as a possible link to insulin resistance . Indeed, higher levels of plasma acylcarnitines, resulting from the intracellular accumulation of acylcarnitines, are associated with insulin resistance in both humans and rodents [32, 64]. Our results show increased short and long chain acylcarnitines in the three ZDF groups compared to the lean control but, surprisingly, the HV-HPMC group, which displayed less insulin resistance compared to the obese control group, as shown by an improved OGTT and a higher QUICKI, did not differ from the obese control in acylcarnitine concentration in the muscle. A previous study associating increased acylcarnitines with insulin resistance compared models displaying very large differences in insulin resistance and obesity . Although acylcarnitine levels appear to increase during insulin resistance and obesity, it may be that they are only elevated as a function of other characteristics of the model, such as increased fatty acid β-oxidation, and may not be directly related to insulin resistance. Although the concentration of muscle acylcarnitines did not differ between the HV-HPMC and obese control despite differences in insulin resistance, it is conceivable that differences in the rate of β-oxidation may exist. To that end, we measured gene expression in muscle of CPT-1β and UCP3, two genes regulating fatty acid oxidation, but found no change with HV-HPMC consumption. However PGC-1α, a transcriptional coactivator linked to lipid oxidation, did trend lower, implying a decrease in fatty acid oxidation. Therefore, these results show that even though muscle acylcarnitines increased in a situation of greatly increased insulin resistance, as seen when comparing the lean and obese control groups, it appears that moderate decreases in insulin resistance, as produced by the HPMC-containing diets, were insufficient to decrease acylcarnitine concentrations.