Serum 25-hydroxyvitamin D and insulin resistance in people at high risk of cardiovascular disease: a euglycaemic hyperinsulinaemic clamp study

CONTEXT

In observational studies, low serum 25-hydroxyvitamin D (25-OHD) concentration is associated with an increased risk of type 2 diabetes mellitus (DM). Increasing serum 25-OHD may have beneficial effects on insulin resistance or beta-cell function. Cross-sectional studies utilizing suboptimal methods for assessment of insulin sensitivity and serum 25-OHD concentration provide conflicting results.

OBJECTIVE

This study examined the relationship between serum 25-OHD concentration and insulin resistance in healthy overweight individuals at increased risk of cardiovascular disease, using optimal assessment techniques.

METHODS

A total of 92 subjects (mean age 56·0, SD 6·0 years), who were healthy but overweight (mean body mass index 30·9, SD 2·3 kg/m(2) ), underwent assessments of insulin sensitivity (two-step euglycaemic hyperinsulinaemic clamp, HOMA2-IR), beta-cell function (HOMA2%B), serum 25-OHD concentration and body composition (DEXA).

RESULTS

Mean total 25-OHD concentration was 32·2, range 21·8-46·6 nmol/l. No association was demonstrated between serum 25-OHD concentration and insulin resistance.

CONCLUSIONS

In this study using optimal assessment techniques to measure 25-OHD concentration, insulin sensitivity and body composition, there was no association between serum 25-OHD concentration and insulin resistance in healthy, overweight individuals at high risk of developing cardiovascular disease. This study suggests the documented inverse association between serum 25-OHD concentration and risk of type 2 DM is not mediated by a relationship between serum 25-OHD concentration and insulin resistance.

Comparison of effects of combined ACE inhibitor and low-dose thiazide diuretic with ACE inhibitor alone on insulin action in patients with hypertension and Type 2 diabetes: a double-blind crossover study

AIMS

To establish the safety in terms of insulin sensitivity of a low dose thiazide/ACE inhibitor combination.

METHODS

We examined the effects on insulin sensitivity of captopril either alone or in combination with low-dose bendroflumethiazide (1.25 mg) in 15 hypertensive Type 2 diabetic patients. Insulin action was assessed using an isoglycaemic hyperinsulinaemic clamp in a double-blind, randomised, crossover study after a 6-week placebo run-in and following two 12-week treatment periods with captopril (C) (100 mg) alone or in combination with bendroflumethiazide (CB) (1.25 mg).

RESULTS

Blood pressure was lower following CB compare to C (138/83 vs. 144/85 mmHg; P < 0.05) and both were lower than baseline (153/92 mmHg; P < 0.01). CB resulted in a significant increase in fasting plasma glucose compared to C (9.6 +/- 2.6 vs. 8.5 +/- 1.6 mmol/l; P < 0.05). Exogenous glucose infusion rates required to maintain isoglycaemia during hyperinsulinaemia were lower after CB compared to C (25.1 +/- 13.3 vs. 34.2 +/- 16.8 micromol/kg/min; P < 0.01) as were isotopically determined glucose utilisation rates (29.0 +/- 12.4 vs. 36.6 +/- 17.3 micromol/kg/min; P < 0.05). There was no significant difference in fasting endogenous glucose production between treatments (CB 9.3 +/- 3.3 vs. C 8.6 +/- 1.6 micromol/kg/min), nor between suppression following insulin (CB 4.0 +/- 2.1 vs. C 4.3 +/- 3.1 micromol/kg/min).

CONCLUSIONS

Combination of low-dose bendroflumethiazide with captopril lowered blood pressure but resulted in deleterious effects on insulin action compared to captopril alone.

The effect of manipulation of basal pulsatile insulin on insulin action in Type 2 diabetes

AIM

Disordered insulin pulsatility is associated with insulin resistant states including Type 2 diabetes. However, whether abnormal basal insulin pulses play a role in the pathogenesis of insulin resistance or are simply an associated feature remains undetermined. We investigated this relationship further by studying the effect of overnight (10 h) pulsatile insulin infusion on subsequent insulin sensitivity.

METHODS

We studied 17 Type 2 diabetic patients who underwent one of two protocols. In protocol A (10 patients) on two separate nights we infused insulin 0.1 mU/kg/min either in a constant infusion or in pulses every 13 min. Octreotide (0.43 microg/kg/h) was given to suppress endogenous insulin secretion and physiological replacement of glucagon (30 ng/kg/h) administered. Insulin sensitivity was measured using a hyperinsulinaemic euglycaemic clamp (2 mU/kg/min) next morning. In protocol B (seven patients), we employed the same experimental procedure but used a basal insulin infusion rate of 0.09 mU/kg/min in 7-min or 13-min pulses.

RESULTS

Appropriate pulse patterns were confirmed in each protocol. In protocol A, after overnight infusions, glucose infusion rates required to maintain euglycaemia at steady state hyperinsulinaemia were similar (33.9 +/- 5.2 vs. 31.2 +/- 4.1 micromol/kg/min; P = NS). In protocol B, after overnight infusions the glucose infusion rates required during hyperinsulinaemia were significantly lower during 7-min pulses (39.9 +/- 5.7 vs. 44.7 +/- 5.6 micromol/kg/min; P < 0.05).

CONCLUSION

There was no demonstrable priming effect derived from overnight pulsatile insulin compared with constant insulin infusion on subsequent insulin sensitivity in Type 2 diabetic subjects. The failure of 7-min pulses to exhibit an advantageous effect over 13-min pulses raises questions about the natural frequency of basal insulin pulses and their biological effect.

Comparison of the priming effects of pulsatile and continuous insulin delivery on insulin action in man

Insulin is normally secreted in man in regular pulses every 5 to 15 minutes. Disordered pulsation has been demonstrated in several insulin-resistant states and it is unclear whether this represents a primary beta-cell defect contributing to impairment of peripheral insulin action or rather is a consequence of insulin resistance. Basal or near basal insulin administration by pulsatile infusion augments hypoglycemic effect and improves insulin-mediated glucose uptake compared with insulin by continuous infusion. To date no study has examined whether normal basal insulin pulsatility is required to preserve subsequent insulin sensitivity during hyperinsulinemia. We studied the effect of overnight pulsatile versus continuous basal insulin on a subsequent hyperinsulinemic euglycemic clamp. Nineteen normal volunteers (male:female ratio, 17:2; mean age +/- SEM, 26.1 +/- 2.3 years) were studied on 2 occasions each. Endogenous insulin secretion was inhibited by octreotide (0.43 microg kg(-1). h(-1)) and replaced overnight at 5.4 mU kg(-1). h(-1) either by continuous infusion or in 2-minute pulses every 13 minutes (n = 10) or every 7 minutes (n = 9). Glucagon was replaced at physiological concentration by continuous infusion (30 ng. kg(-1). h(-1)). Venous plasma glucose overnight was not significantly different between the pulsatile and continuous protocols. After discontinuing the overnight insulin infusion, insulin action was assessed during a hyperinsulinemic euglycemic clamp (1 mU kg(-1). h(-1)). Glucose infusion rates at steady-state during the hyperinsulinemic clamp were similar between continuous and both frequencies of pulsatile infusion (continuous 44.6 +/- 4.3 micromol. kg(-1). min(-1) v 13-minute pulsatile 41.7 +/- 5.9 micromol. kg(-1). min(-1), P =.27; continuous 34.6 +/- 2.5 micromol. kg(-1) min(-1) v 7-minute pulsatile 41.4 +/- 3.2 micromol. kg(-1). min(-1), P =.08). We conclude that overnight pulsatile compared with continuous insulin administration has no different effect on subsequent peripheral insulin-mediated glucose uptake. A priming effect cannot therefore explain the previously demonstrated association between endogenous insulin pulse frequency and peripheral insulin action.

Insulin action and skeletal muscle blood flow in patients with Type 1 diabetes and microalbuminuria

Our objective was to determine whether Type 1 diabetic patients with microalbuminuria are less sensitive to the effects of insulin on glucose metabolism and skeletal muscle blood flow, compared to those with normal albumin excretion, after careful matching for confounding variables. We recruited 10 normotensive Type 1 diabetic patients with microalbuminuria and 11 with normoalbuminuria matched for age, sex, body mass index, duration of diabetes and HbA(1c). Peripheral and hepatic insulin action was assessed using a two-step euglycaemic hyperinsulinaemic clamp (2 h at 0.4 mU x kg(-1) x min(-1), 2 h at 2.0 mU x kg(-1) x min(-1)) combined with isotope dilution methodology. Skeletal muscle blood flow was determined by venous occlusion plethysmography. During the clamps, glucose infusion rates required to maintain euglycaemia were similar in the microalbuminuric subjects and controls (step 1, 8.2+/-1.4 (SE) vs 9.2+/-1.3 micromol x kg(-1) x min(-1): step 2, 30.9+/-2.7 vs 32.0+/-3.8 micromol x kg(-1) x min(-1)), as was hepatic glucose production basally and at steady state in step 1. In step 2, hepatic glucose production was lower in the microalbuminuric group (2.9+/-0.9 vs 6.4+/-0.7 micromol x kg(-1) x min(-1), P=0.005). During step 2, skeletal muscle blood flow increased significantly above baseline in the normoalbuminuric group (4.1+/-0.5 vs 3.2+/-0.4 ml x 100-ml(-1) x min(-1), P=0.01) but not in the microalbuminuric group (2.4+/-0.3 vs 2.3+/-0.4 ml x 100-ml(-1) x min(-1)). In conclusion, microalbuminuria in Type 1 diabetes was found to be associated with impairment of insulin-mediated skeletal muscle blood flow, but not with insulin resistance.

Insulin action and insulin secretion in polycystic ovary syndrome treated with ethinyl oestradiol/cyproterone acetate

Polycystic ovary syndrome (PCOS) is associated with abnormalities of insulin action and insulin secretion. Ethinyl oestradiol/cyproterone acetate is a common agent used to treat the symptoms of PCOS, but its effects on insulin action and insulin pulsatility have not been examined. We investigated the relationship between insulin action and insulin secretion in 11 patients with PCOS, at diagnosis and after 3 months of treatment with ethinyl oestradiol/cyproterone acetate, and in 13 controls. Insulin action was assessed using the euglycaemic hyperinsulinaemic clamp (2 mU/kg/min for 2 h). Insulin pulsatility was examined over 90 min by 2 min sampling. Short-term insulin pulses were identified using PULSAR. Treatment with ethinyl oestradiol/cyproterone acetate resulted in significant reductions in testosterone (3.3+/-0.7 vs. 1.9+/-0.2 nmol/l, p<0.05), free androgen index (10.2+/-0.7 vs. 1.2+/-0.2, p<0.05) and LH/FSH ratio (2.6+/-0.5 vs. 1.0+/-0.2, p<0.05). During hyperinsulinaemic clamps, the glucose infusion rate (GIR) required to maintain euglycaemia was lower in PCOS compared to controls (33.6+/-2.7 vs. 45.1+/-3.5 micromol/kg/min, p<0.05) but similar in PCOS before and after treatment (33.6+/-2.8 vs. 33.6+/-2.7 micromol/kg/min, p=0.9). Numbers of pulses identified in PCOS and controls were similar and unaltered by ethinyl oestradiol/cyproterone acetate. There was no correlation between GIR and frequency of insulin pulses in PCOS before or after treatment (r=0.2, p=0.6; post r=-0.5, p=0.1) unlike controls (r=-0.6, p=0.04). Despite considerable improvement in androgen profile, treatment with ethinyl oestradiol/cyproterone acetate did not alter insulin action in PCOS, and this insulin resistance does not appear to be determined by insulin pulse frequency.

Insulin resistance and insulin pulsatility in essential hypertension

OBJECTIVE

Studies in normal humans and in patients with type 2 diabetes mellitus have demonstrated a close inverse relationship between peripheral insulin sensitivity and the frequency of short-term insulin secretory pulses in the systemic circulation. Our objective was to study this relationship in essential hypertension.

DESIGN

Study of insulin sensitivity and insulin pulse characteristics in hypertensive subjects and normotensive controls using well-established techniques.

METHODS

Twelve subjects with essential hypertension and 12 age- and sex-matched normotensive controls were recruited. Insulin action was measured using the glucose clamp technique combined with isotope dilution methodology. Insulin pulsatility in the peripheral circulation was assessed by sampling every 2 min for 90 min after an overnight fast Pulses were identified using the computer program Pulsar.

RESULTS

Insulin sensitivity index (glucose infusion rate/ serum insulin) was lower in the hypertensive patients (P= 0.01) and fasting insulin was increased (P= 0.008) compared to controls. The frequency and amplitude of insulin pulses were similar in the two groups. Insulin pulse frequency and insulin sensitivity were inversely related in the normotensive group (r= -0.68, P= 0.015), but not in the hypertensive group (r= -0.23, P= 0.48). Insulin clearance was reduced in the hypertensive group (P= 0.03), and was inversely related to insulin pulse frequency in the two groups combined (r = -0.51, P= 0.01).

CONCLUSIONS

Insulin action was not related to insulin pulse frequency in essential hypertension, in contrast to the situation in normal man.

Comparison of effects of captopril used either alone or in combination with a thiazide diuretic on insulin action in hypertensive Type 2 diabetic patients: a double-blind crossover study

AIMS

It has been suggested that the adverse metabolic effects of antihypertensive therapy offset some of the benefits of blood pressure reduction. It has also been suggested that angiotensin converting enzyme (ACE) inhibitors reduce insulin resistance and that, if used together with thiazide diuretics, the adverse effects of thiazides on insulin sensitivity may be eliminated. We examined the effects on insulin sensitivity of captopril either alone or in combination with bendrofluazide in 11 hypertensive Type 2 diabetic patients.

METHODS

Insulin action was assessed using an isoglycaemic hyperinsulinaemic clamp in a double-blind, randomized, crossover study after a 6-week placebo run-in and following two 12-week treatment periods with captopril (C) (100 mg) alone or in combination with bendrofluazide (CB) (2.5 mg).

RESULTS

Blood pressure was lower following CB compared to C (128/82 vs. 144/ 88 mmHg; P<0.005) and both were lower than baseline (162/101 mmHg; P < 0.001). CB resulted in a significant increase in fasting plasma glucose compared to C (9.7+/-0.8 vs. 8.5+/-0.6 mmol/; P < 0.05). Exogenous glucose infusion rates required to maintain isoglycaemia during hyperinsulinaemia were lower after CB compared to C (22.3+/-2.4 vs. 27.4+/-4.2 mol x kg(-1) x min(-1); P < 0.05). Suppression of endogenous glucose production was reduced after CB compared to baseline (4.0+/-0.6 vs. 2.4+/-0.5 mol x kg(-1) x min(-1); P< 0.05).

CONCLUSIONS

Combination of bendrofluazide with captopril lowered blood pressure but resulted in deleterious effects on insulin action compared to captopril alone.

Captopril does not improve insulin action in essential hypertension: a double-blind placebo-controlled study

OBJECTIVE

To compare the effect of captopril with that of placebo on peripheral and hepatic insulin action in essential hypertension, in light of evidence that insulin resistance is associated with cardiovascular risk.

DESIGN

Randomized, double-blind, placebo-controlled, crossover trial, with 8 week treatment periods of captopril and placebo preceded and separated by 6 weeks of placebo.

SETTING

Belfast teaching hospital.

PATIENTS

Eighteen Caucasian nondiabetic patients (10 males), aged under 65 years, with essential hypertension, recruited from general practices in the greater Belfast area.

INTERVENTIONS

Captopril at 50 mg twice a day or placebo twice a day for two 8 week treatment periods.

MAIN OUTCOME MEASURES

Peripheral and hepatic insulin sensitivity assessed by glucose clamps.

RESULTS

Fourteen patients completed the study. Mean (+/- SEM) levels of fasting glucose, fasting insulin and postabsorptive hepatic glucose production were similar after captopril and placebo (5.4+/-0.1 versus 5.4+/-0.1 mmol/l, 10.6+/-2.2 versus 9.5+/-1.1 mU/l, 11.2+/-0.6 versus 11.0+/-0.5 mmol/kg per min, respectively). During hyperinsulinaemia, hepatic glucose production was suppressed to comparable levels after both treatments (4.8+/-0.6 versus 4.3+/-0.6 mmol/kg per min) and exogenous glucose infusion rates required to maintain euglycaemia were also similar (30.0+/-2.6 versus 30.3+/-2.6 mmol/kg per min).

CONCLUSION

Captopril therapy in uncomplicated essential hypertension has no effect on peripheral or hepatic insulin sensitivity.

Effects of combination therapy with an angiotensin converting enzyme inhibitor and thiazide diuretic on insulin action in essential hypertension

OBJECTIVE

To determine whether combination of an angiotensin converting enzyme inhibitor with a high dose of thiazide diuretic avoids adverse metabolic consequences of thiazide diuretics.

DESIGN

Double-blind randomized crossover study of two 12-week treatment periods with captopril (up to 100 mg/day) either alone or in combination with 5 mg bendrofluazide given after a 6-week placebo run-in period. Treatment periods were separated by a 6-week placebo washout period.

SETTING

Outpatient clinics in greater Belfast.

PATIENTS

Fifteen white non-diabetic essential hypertensives (seven male) aged < 65 years recruited from general practices in greater Belfast.

MAIN OUTCOME MEASURES

Systolic and diastolic blood pressures and peripheral and hepatic insulin action.

RESULTS

Two patients failed to complete the study. Blood pressure was lowered (139/89+/-18/7 mmHg combination versus 160/97+/-21/7 mmHg captopril; P < 0.001). Fasting insulin level was raised (7.9+/-3.6 mU/l combination versus 6.2+/-3.2 mU/l baseline; P < 0.001). There were no differences between treatments for glucose, urate, cholesterol and triglyceride levels. Serum potassium level was lowered (3.8+/-0.4 mmol/l combination versus 4.2+/-0.4 mmol/l captopril, P < 0.05). Postabsorptive endogenous glucose production was raised (10.8+/-1.7 micromol/kg per min combination versus 10.0+/-1.5 micromol/kg per min captopril; P < 0.01) and was greater than baseline (9.7+/-2.1 micromol/kg per min, P < 0.05). Suppression of glucose production by insulin was similar with both treatments. Exogenous glucose infusion rates required to maintain euglycaemia did not differ (32.4+/-7.6 micromol/kg per min captopril, 32.7+/-6.2 micromol/kg per min combination, 31.5+/-7.2 micromol/kg per min baseline).

CONCLUSIONS

Combination therapy increased glucose production (compared with captopril alone), indicating hepatic insulin resistance. It cannot be assumed that combined preparations with angiotensin converting enzyme inhibitors will ameliorate adverse effects of high doses of thiazide diuretics on insulin action.

Skeletal muscle blood flow is not a determinant of insulin resistance in essential hypertension

OBJECTIVE

To investigate the haemodynamic effects of insulin and their relationship to insulin resistance in essential hypertension.

DESIGN

Group comparison between patients with essential hypertension and normal controls.

SETTING

Outpatient clinics serving the greater Belfast area.

PATIENTS

Eleven patients with essential hypertension and eight age-, sex- and weight-matched control subjects were recruited. Administration of all antihypertensive agents to the hypertensive patients was stopped 6 weeks prior to the study.

METHODS

Leg blood flow was measured using venous occlusion plethysmography. Insulin action was assessed using the hyperinsulinaemic euglycaemic clamp technique.

RESULTS

The hypertensive subjects were insulin-resistant compared with the normal controls. Insulin infusion resulted in similar increases in calf blood flow in the two groups. There was no correlation between calf blood flow and measurements of insulin sensitivity in either group.

CONCLUSIONS

Differences in whole-body glucose uptake in hypertensive and control subjects are not likely to be related to differences in insulin-induced stimulation of muscle blood flow.

Association between insulin secretory pulse frequency and peripheral insulin action in NIDDM and normal subjects

Abnormalities of both insulin secretion and insulin action occur in NIDDM. It is not clear, however, which is the primary defect. Recently, it has been suggested that the frequency of insulin pulses is an important factor regulating insulin action in normal humans. We examined the relationship between pulsatile insulin secretion and insulin action in eight NIDDM subjects and eight health matched control subjects. Insulin action was assessed prevailing fasting glucose levels before and after hype insulinemia (2-h insulin infusion at 2.0 mU / kg / min). Pulsatility of insulin was assessed by sampling every 2 min for 90 min after an overnight fast and identifying insulin pulses using the computer program Pulsar. Fasting plasma glucose and postabsorptive endogenous glucose production were both greater in diabetic subjects compared with control subjects (10.1 +/- 1.2 vs. 5.4 +/- 0.1 mmol/l, P < 0.01; 11.8 +/- 0.8 vs. 9.9 +/- 0.4 micromol / kg / min, P < 0.05). During the 2.0 mU insulin infusion, glucose clearance was lower in the diabetic subjects (3.6 +/- 0.7 vs. 6.9 +/- 0.5 ml / kg / min), P < 0.05), whereas endogenous glucose production was suppressed to a similar degree in both groups (4.5 +/- 0.8 vs. 3.6 +/- 0.7 micromol x kg(-1) x min(-1), NS). The frequency of insulin pulses and glucose clearance were negatively correlated in both diabetic subjects (r = -0.75, P < 0.05) and normal control subjects (r = -0.82, P < 0.01). This negative correlation was also present in both groups taken together (r = -0.72, P < 0.001). There was no correlation between insulin pulse frequency and endogenous glucose production either in the fasting state or during hyperinsulinemia. We concluded that the frequency of insulin pulses and peripheral insulin sensitivity are closely linked in NIDDM and normal subjects.

A comparison of the effects of low- and conventional-dose thiazide diuretic on insulin action in hypertensive patients with NIDDM

In conventional doses, thiazide diuretics impair glucose tolerance and decrease insulin sensitivity, making them an unpopular choice for treating diabetic patients with hypertension. However, use of low-dose thiazide diuretics may avoid the adverse metabolic effects seen with conventional doses. In a double-blind, randomised crossover study we assessed peripheral and hepatic insulin action in 13 hypertensive non-insulin-dependent diabetic patients after a 6-week placebo run-in and following two 12-week treatment periods with either low (1.25 mg) or conventional (5.0mg) dose bendrofluazide. There were no differences between doses in their effects on systolic and diastolic blood pressure. Bendrofluazide 1.25 mg had significantly less effect on serum potassium, uric acid, fasting glucose and HbA1C concentrations than the 5.00 mg dose. Exogenous glucose infusion rates required to maintain euglycaemia were significantly different between doses (p < 0.05) with conventional-dose bendrofluazide worsening peripheral insulin resistance compared to baseline (23.8 +/- 2.9 vs 27.3 +/- 3.5 mumol.kg-1.min-1, p < 0.05) and low-dose bendrofluazide producing no change compared to baseline (26.8 +/- 3.6 vs 27.3 +/- 3.5 mumol.kg-1.min-1, p = NS). Postabsorptive endogenous glucose production was higher on treatment with bendrofluazide 5.0 mg compared to 1.25 mg (11.7 +/- 0.5 vs 10.2 +/- 0.3 mumol.kg-1.min-1, p < 0.05) and suppressed to a lesser extent following insulin (4.0 +/- 0.7 vs 2.0 +/- 0.4 mumol.kg-1.min-1, p < 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Failure of complete suppression of endogenous glucose production by euglycaemic hyperinsulinaemia in normal humans

In normal subjects, endogenous glucose production (EGP) is usually assumed to be completely suppressed during euglycaemic clamp studies performed at high insulin levels (> 100 mU L-1). However, this assumption is based on non-steady-state tracer measurements of EGP which are prone to negative errors. We have used purified [6-(3)H]glucose in an optimal tracer infusion protocol to assess the suppression of EGP during 4 h euglycaemic clamps in eight normal men. An insulin infusion rate of 5 mU kg-1 min-1 was chosen to achieve supraphysiological (> 500 mU L-1) plasma insulin concentrations. Using a labelled exogenous glucose infusion, plasma glucose (mean +/- SEM 5.3 +/- 0.1 mmol L-1) and glucose specific activities (mean 100 +/- 3% of basal) were maintained constant from 80 to 240 min. During hyperinsulinaemia, isotopically determined glucose appearance rates (Ra) were greater than glucose infusion rates (GIR) throughout the euglycaemic clamp period (P < 0.001) and EGP (Ra-GIR) was always greater than zero. In seven of the eight subjects studied EGP was partly suppressed but showed a wide variation (EGP 5 to 91% of basal at 80-120 min and 12 to 87% of basal at 200-240 min) while in one subject EGP rose above basal (by 72% at 80-120 min and 49% at 200-240 min). We conclude that EGP is not completely suppressed during euglycaemic clamps at high insulin levels.

Effects of low dose versus conventional dose thiazide diuretic on insulin action in essential hypertension

OBJECTIVE

To see whether low dose thiazide diuretics given to patients with essential hypertension might avoid the adverse metabolic consequences seen with conventional doses.

DESIGN

Double blind randomised crossover study of two 12 week treatment periods with either low dose (1.25 mg) or conventional dose (5.0 mg) bendrofluazide given after a six week placebo run in period.

SETTING

Outpatient clinics serving the greater Belfast area.

SUBJECTS

16 white non-diabetic patients (9 male) under 65 with essential hypertension recruited from general practices within the greater Belfast area.

MAIN OUTCOME MEASURES

Systolic and diastolic blood pressure and peripheral and hepatic insulin action.

RESULTS

One man failed to complete the study. There were no differences between doses in their effects on systolic and diastolic blood pressure. Bendrofluazide 1.25 mg had substantially less effect on serum potassium concentration than the 5.0 mg dose. There were no intertreatment differences in fasting glucose, insulin, cholesterol, and triglyceride concentrations. Bendrofluazide 5.0 mg significantly increased postabsorptive endogenous glucose production compared with baseline (mean 10.9 (SD 1.2) v 10.0 (0.8) mumol/kg/min), whereas bendrofluazide 1.25 mg did not. Postabsorptive endogenous glucose production was significantly higher with bendrofluazide 5.0 mg compared with 1.25 mg (10.9 (1.2) v 9.9 (0.8) mumol/kg/min) but was suppressed to a similar extent after insulin (bendrofluazide 5.0 mg 2.8 (1.5) mumol/kg/min v bendrofluazide 1.25 mg 2.2 (1.5) mumol/kg/min). Exogenous glucose infusion rates required to maintain euglycaemia were not significantly different between doses and were similar to baseline.

CONCLUSIONS

Bendrofluazide 1.25 mg is as effective as conventional doses but has less adverse metabolic effect. In contrast with conventional doses, low dose bendrofluazide has no effect on hepatic insulin action. There is no difference between low and conventional doses of bendrofluazide in their effect on peripheral insulin sensitivity.

The effect of cortisol on glucose/glucose-6-phosphate cycle activity and insulin action

Increased glucose/glucose-6-phosphate (G/G6P) substrate cycle activity may be an early marker of disordered hepatic glucose metabolism. To investigate the effects of glucocorticoids on G/G6P cycle activity and insulin resistance, we studied eight normal subjects using the euglycemic glucose clamp technique with high pressure liquid chromatography-purified [2(3)H]- and [6-3H]glucose tracers at insulin infusion rates of 0.4 and 2.0 mU/kg.min after 24-h cortisol (2 micrograms/kg.min) and saline infusions. Endogenous glucose production ([6-3H]glucose) was greater after cortisol than saline in the postabsorptive state (13.3 +/- 0.5 vs. 12.2 +/- 0.5 mumol/kg.min; P < 0.05) and during 0.4-mU insulin infusion (10.5 +/- 0.7 vs. 5.0 +/- 0.8 mumol/kg.min; P < 0.005). During 2.0-mU insulin infusion, endogenous glucose production was suppressed similarly (5.1 +/- 0.4 vs. 4.1 +/- 0.5 mumol/kg.min), but glucose disappearance was less after cortisol than saline (38.7 +/- 3.5 vs. 64.6 +/- 4.3 mumol/kg.min; P < 0.001). G/G6P cycle activity after cortisol and saline was similar in the postabsorptive state and during 0.4 mU insulin. During 2.0 mU insulin, cycle activity was greater after cortisol than saline (3.6 +/- 0.9 vs. 0.8 +/- 0.5 mumol/kg.min; P < 0.005). In conclusion, cortisol induces hepatic insulin resistance without significantly changing G/G6P cycle activity. At high glucose turnover rates, G/G6P cycle activity is increased by cortisol; however, reduced glucose disappearance is the main cause of impaired insulin action.

Insulin action and hepatic glucose cycling in essential hypertension

Peripheral insulin resistance is a feature of essential hypertension, but there is little information about hepatic insulin sensitivity. To investigate peripheral and hepatic insulin sensitivity and activity of the hepatic glucose/glucose 6-phosphate (G/G6P) substrate cycle in essential hypertension, euglycemic glucose clamps were performed in eight untreated patients and eight matched controls at insulin infusion rates of 0.2 and 1.0 mU.kg-1.min-1. A simultaneous infusion of (2(3)H)- and (6(3)H)glucose, combined with a selective detritiation procedure, was used to determine glucose turnover, the difference being G/G6P cycle activity. Endogenous hepatic glucose production (EGP) determined with (6(3)H)glucose was similar in hypertensive and control groups in the postabsorptive state (11.0 +/- 0.3 v 10.9 +/- 0.3 mumol.kg-1.min-1) and with the 0.2 mU insulin infusion (4.9 +/- 0.5 v 4.0 +/- 0.8 mumol.kg-1.min-1). With the 1.0 mU insulin infusion, glucose disappearance determined with (6(3)H)glucose was lower in the hypertensive group (21.8 +/- 2.4 v 29.9 +/- 2.4 mumol.kg-1.min-1, P less than .001). G/G6P cycle activity was similar both in the postabsorptive state (2.2 +/- 0.4 v 2.7 +/- 0.4 mumol.kg-1.min-1) and during insulin infusion (0.2 mU, 2.5 +/- 0.3 v 2.9 +/- 0.4; 1.0 mU, 4.7 +/- 0.3 v 5.3 +/- 1.1 mumol.kg-1.min-1 for hypertensive and control groups, respectively).(ABSTRACT TRUNCATED AT 250 WORDS)

Underestimation of glucose turnover determined using [6-3H]glucose tracer in non-steady state. The role of a tritiated tracer impurity

The use of tritiated glucose tracers may result in underestimation of glucose turnover during hyperinsulinaemic clamps giving paradoxical negative endogenous glucose production rates. While mathematical modelling errors in the analysis of tracer data are major determinants of this underestimate in the non-steady state, the relative importance of tracer contamination under these conditions remains in doubt. We have used high performance liquid chromatography to assess the possible contribution to this problem of a labelled tracer impurity found in [6-3H]glucose. In conventional 4 h hyperinsulinaemic clamps performed in six normal subjects, labelled impurity increased as a percentage of the neutral plasma radioactivity fraction from 5.3 +/- 0.9% after a 2 h equilibration period (0 min) to 13.5 +/- 2.2% at 120 min and 15.4 +/- 2.4% at 240 min, as plasma glucose specific activities fell following the infusion of insulin. Negative endogenous glucose production rates were observed both at 90-120 min (-8.8 +/- 1.6 mumol.kg-1min-1) and at 210-240 min (-8.5 +/- 1.4 mumol.kg-1min-1) implying a persistent underestimate in isotopically determined glucose appearance rate. Using chromatography data to correct for impurity increased glucose appearance rates by 7.9 +/- 2.1% at 120 min and 11.0 +/- 2.5% at 240 min. Purified tracer was then used for a further six clamps. When the conventional protocol was used with unlabelled glucose infusion an obvious negative error persisted only at 90-120 min. In contrast, labelled infusions gave exclusively positive values for endogenous glucose production.(ABSTRACT TRUNCATED AT 250 WORDS)