Relationship between corneal confocal microscopy and markers of peripheral nerve structure and function in Type 2 diabetes.

AIMS: To investigate changes in corneal nerve morphology in Type 2 diabetes and to establish relationships between in vivo corneal confocal microscopy and markers of peripheral nerve structure and function. PARTICIPANTS AND METHODS: We recruited 57 participants with Type 2 diabetes and 26 healthy controls of similar age and sex distribution. We also recruited a disease control group of 54 participants with Type 1 diabetes. All participants were assessed for distal symmetrical polyneuropathy using the Total Neuropathy Score. In vivo corneal confocal microscopy was used to assess corneal nerve fibre length, corneal nerve fibre density, corneal nerve branch density and inferior whorl length. Peripheral nerve structure was assessed using median nerve ultrasonography. Large fibre function was assessed according to median nerve axonal excitability. Small fibre function was assessed using SudoscanTM and the Survey of Autonomic Symptoms. RESULTS: Corneal nerve fibre length, fibre density and branch density and inferior whorl length were significantly lower in individuals with Type 2 diabetes compared to controls (P<0.001 for all). In the Type 2 diabetes cohort, correlations were observed between neuropathy severity and corneal nerve fibre density (P=0.004), corneal nerve branch density (P=0.003), corneal nerve fibre length (P=0.002) and inferior whorl length (P=0.01). Significant correlations were observed between corneal confocal outcomes and axonal excitability measurements. No association was found between corneal confocal microscopy and median nerve cross-sectional area, Sudoscan measurements or the Survey of Autonomic Symptoms. CONCLUSIONS: This study demonstrated significant changes in corneal nerves in individuals with Type 2 diabetes. Reductions in corneal nerve measures correlated with increasing neuropathy severity. Associations were found between corneal confocal microscopy and markers of voltage-gated potassium channel function.

Associations between acute glucose control and peripheral nerve structure and function in type 1 diabetes.

AIM: To examine the associations between continuous overlapping net glycaemic action (CONGA), percentage time in hyperglycaemia (%HG) or normoglycaemia (%NG) and peripheral nerve structure and function in type 1 diabetes. METHODS: Twenty-seven participants with type 1 diabetes underwent continuous glucose monitoring followed by corneal confocal microscopy and nerve excitability assessments. CONGA, %HG (> 10.0 mmol/l) and %NG (3.9-10.0 mmol/l) were correlated against corneal nerve fibre length and density in the central cornea and inferior whorl region, corneal microneuromas, and a nerve excitability score while controlling for age, sex, diabetes duration and HbA1c . RESULTS: An increase in CONGA [median 2.5 (2.0-3.1) mmol/l] or %HG (mean 46 ± 18%) was associated with a worse nerve excitability score (r = -0.433, P = 0.036 and r = -0.670, P = 0.0012, respectively). By contrast, greater %NG (51 ± 17%) correlated with better nerve excitability scores (r = 0.672, P = 0.0011). Logistic regression revealed that increasing %HG increased the likelihood of abnormal nerve function [odds ratio (OR) 1.11, 95% confidence interval (CI) 1.01-1.23; P = 0.037). An increase in CONGA and %HG were associated with worsening nerve conduction measures, whereas longer %NG correlated with improved nerve conduction variables. CONGA and %HG were associated with inferior whorl corneal nerve fibre length (r = 0.483, P = 0.034 and r = 0.591, P = 0.021, respectively) and number of microneuromas (r = 0.433, P = 0.047 and r = 0.516, P = 0.020, respectively). CONCLUSIONS: Short-term measures of glucose control are associated with impaired nerve function and alterations in corneal nerve morphology.

Sonographic assessment of nerve blood flow in diabetic neuropathy.

AIMS: To undertake sonographic assessment of nerve blood flow in people with Type 2 diabetes and correlate the findings with neuropathy severity scores and electrophysiological measurements. METHODS: Median and tibial nerve ultrasound scans were undertaken in 75 people with diabetes and 30 aged-matched controls without diabetes, using a high-resolution linear probe at non-entrapment sites. Nerve blood flow was quantified using power Doppler techniques to obtain the vessel score and the maximum perfusion intensity. Neuropathy severity was assessed using a total neuropathy score. RESULTS: Diabetic nerves had higher rates of nerve blood flow detection (28%) compared to the control group (P < 0.0001). Significant correlations were found between nerve blood flow measurements and nerve size (P <0.001), reported sensory symptoms (P < 0.05) and neuropathy severity scores (P < 0.001). The cohort with diabetes had significantly larger median (8.5 ± 0.3 mm2 vs 7.2 ± 0.1 mm2 ; P < 0.05) and tibial nerves (18.0 ± 0.9 mm2 vs 12.8 ± 0.5 mm2 ; P < 0.05) compared with controls. CONCLUSION: Peripheral nerve hypervascularity is detectable by ultrasonography in moderate to severe diabetic neuropathy with prominent sensory dysfunction.

Circulating fatty acids, non-high density lipoprotein cholesterol, and insulin-infused fat oxidation acutely influence whole body insulin sensitivity in nondiabetic men.

Circulating lipids and tissue lipid depots predict insulin sensitivity. Associations between fat oxidation and insulin sensitivity are variable. We examined whether circulating lipids and fat oxidation independently influence insulin sensitivity. We also examined interrelationships among circulating lipids, fat oxidation, and tissue lipid depots. Fifty-nine nondiabetic males (age, 45.4 +/- 2 yr; body mass index, 29.1 +/- 0.5 kg/m(2)) had fasting circulating nonesterified fatty acids (NEFAs) and lipids measured, euglycemic-hyperinsulinemic clamp for whole body insulin sensitivity [glucose infusion rate (GIR)], substrate oxidation, body composition (determined by dual energy x-ray absorptiometry), and skeletal muscle triglyceride (SMT) measurements. GIR inversely correlated with fasting NEFAs (r = -0.47; P = 0.0002), insulin-infused NEFAs (n = 38; r = -0.62; P < 0.0001), low-density lipoprotein cholesterol (r = -0.50; P < 0.0001), non-high-density lipoprotein cholesterol (r = -0.52; P < 0.0001), basal fat oxidation (r = -0.32; P = 0.03), insulin-infused fat oxidation (r = -0.40; P = 0.02), SMT (r = -0.28; P < 0.05), and central fat (percentage; r = -0.59; P < 0.0001). NEFA levels correlated with central fat, but not with total body fat or SMT. Multiple regression analysis showed non-high-density lipoprotein cholesterol, fasting NEFAs, insulin-infused fat oxidation, and central fat to independently predict GIR, accounting for approximately 60% of the variance. Circulating fatty acids, although closely correlated with central fat, independently predict insulin sensitivity. Insulin-infused fat oxidation independently predicts insulin sensitivity across a wide range of adiposity. Therefore, lipolytic regulation as well as amount of central fat are important in modulating insulin sensitivity.

Expression of genes involved in lipid metabolism correlate with peroxisome proliferator-activated receptor gamma expression in human skeletal muscle.

Peroxisome proliferator-activated receptor gamma (PPAR-gamma) activation in adipose tissue is known to regulate genes involved in adipocyte differentiation and lipid metabolism. However, the role of PPAR-gamma in muscle remains unclear. To examine the potential regulation of genes by PPAR-gamma in human skeletal muscle, we used semiquantitative RT-PCR to determine the expression of PPAR-gamma, lipoprotein lipase (LPL), muscle carnitine palmitoyl transferase-1 (mCPT1), fatty acid-binding protein (FABP), carnitine acylcarnitine transferase (CACT), and glucose transporter-4 (GLUT4) in freeze-dried muscle samples from 14 male subjects. These samples were dissected free of adipose and other tissue contamination, as confirmed by minimal or absent adipsin expression. Between individuals, the messenger ribonucleic acid concentration of PPAR-gamma varied up to 3-fold, whereas LPL varied up to 6.5-fold, mCPT1 13-fold, FABP 4-fold, CACT 4-fold, and GLUT4 up to 3-fold. The expression of LPL (r2 = 0.54; P = 0.003), mCPT1 (r2 = 0.42; P = 0.012), and FABP (r2 = 0.324; P = 0.034) all correlated significantly with PPAR-gamma expression in the same samples. No significant correlation was observed between the expression of CACT and PPAR-gamma or between GLUT4 and PPAR-gamma. These findings demonstrate a relationship between PPAR-gamma expression and the expression of other genes of lipid metabolism in muscle and support the hypothesis that PPAR-gamma activators such as the antidiabetic thiazolidinediones may regulate fatty acid metabolism in skeletal muscle as well as in adipose tissue.

Fat oxidation, body composition and insulin sensitivity in diabetic and normoglycaemic obese adults 5 years after weight loss.

OBJECTIVE: To investigate whether normal glucose-tolerant and type II diabetic overweight adults differ in response to weight regain with regard to substrate oxidation and metabolic parameters. METHODS: A total of 15 overweight-obese subjects: seven normal glucose tolerant (NGT) and eight with type II diabetes (DM) were restudied 5 y after significant weight loss. Prediet, after 28 days calorie restriction and at 5 y, subjects were characterised for weight, height, waist-to-hip ratio (WHR) and body composition by dual-energy X-ray absorptiometry. Fasting glucose, insulin, leptin and lipid levels were measured and subjects underwent euglycaemic-hyperinsulinaemic clamp (insulin 0.25 U/kg/h for 150 min). Indirect calorimetry was performed resting and in the final 30 min of the clamp. Dietary assessment was by 4-day diet-diary. RESULTS: Both NGT and DM groups regained weight at 5 y and were not different to prediet. Total body fat (%) and WHR were higher at 5 y compared to prediet in both groups. Fasting glucose was increased in NGT subjects at 5 y, and fasting insulin was higher in both groups at 5 y compared to prediet. Insulin sensitivity (GIR) was similar at 5 y compared to prediet, but at 5 y DM subjects were more insulin resistant than NGT subjects. At 5 y, both DM and NGT groups had significantly reduced basal fat oxidation and no significant suppression of fat oxidation with insulin. Clamp respiratory quotient levels at 5 y were significantly higher in NGT compared to DM subjects. CONCLUSION: Reduced basal fat oxidation, and reduced variation in substrate oxidation in response to insulin develop with fat regain and fasting hyperinsulinaemia in both NGT and DM obese adults.

Independent influences of central fat and skeletal muscle lipids on insulin sensitivity.

OBJECTIVE: Insulin resistance is closely associated with two disparate aspects of lipid storage: the intracellular lipid content of skeletal muscle and the magnitude of central adipose beds. Our aim was to determine their relative contribution to impaired insulin action. RESEARCH METHODS AND PROCEDURES: Eighteen older (56 to 75 years of age) men were studied before elective knee surgery. Insulin sensitivity (M/Delta I) was determined by hyperinsulinemic-euglycemic clamp. Central abdominal fat (CF) was assessed by DXA. Skeletal muscle was excised at surgery and assayed for content of metabolically active long-chain acyl-CoA esters (LCAC). RESULTS: Significant inverse relationships were observed between LCAC and M/Delta I (R(2) = 0.34, p = 0.01) and between CF and M/Delta I (R(2) = 0.38, p = 0.006), but not between CF and LCAC (R(2) = 0.0005, p = 0.93). In a multiple regression model (R(2) = 0.71, p < 0.0001), both CF (p = 0.0006) and LCAC (p = 0.0009) were independent statistical predictors of M/Delta I. Leptin levels correlated inversely with M/Delta I (R(2) = 0.60, p = 0.0002) and positively with central (R(2) = 0.41, p = 0.006) and total body fat (R(2) = 0.63, p = 0.0001). DISCUSSION: The mechanisms by which altered lipid metabolism in skeletal muscle influences insulin action may not be related directly to those linking central fat and insulin sensitivity. In particular, it is unlikely that muscle accumulation of lipids directly derived from labile central fat depots is a principal contributor to peripheral insulin resistance. Instead, our results imply that circulating factors, other than nonesterified fatty acids or triglyceride, mediate between central fat depots and skeletal muscle tissue. Leptin was not exclusively associated with central fat, but other factors, secreted specifically from central fat cells, could modulate muscle insulin sensitivity.