catherine only

AJH 1995; 8M5-453

Insulin Versus Glipizide Treatment in Patients With Non-Insulin-Dependent Diabetes Mellitus Effects on Blood Pressure and Glucose Tolerance Joseph Levy, Melissa Vandenberg, and George Grunberger

Insulin resistance that exists in patients with es- sential hypertension and in those with non-insu- lin-dependent diabetes mellitus (NIDDM) may be the common denominator for the impaired glucose homeostasis and elevated blood pressure (BP) lev- els in patients with NIDDM. Therefore, treatment that improves insulin action may also improve BP levels. Consequently, a four-phase tglipizide v in- sulin) cross-over design study was conducted to determine a better effect of glipizide treatment on insulin sensitivity and the effect this has on BP in 19 NIDDM patients. Patients were subjected to 1 month of diet only (phase I) followed by 3 months of glipizide treatment (phase II), then an addi- tional 1 month of diet only (phase III), and finally 3 months of insulin treatment (phase IV). At the end of phases I, II, and IV oral glucose tolerance tests (OGTT) were performed and plasma glucose, insulin, and C-peptide levels were analyzed. Fast- ing plasma glucose, insulin, total cholesterol, high density lipoprotein cholesterol, low density li- poprotein cholesterol and triglycerides, glycated hemoglobin, fructosamine, and 2-h postprandial

plasma glucose were also analyzed at each phase. Supine and sitting BP levels and body weights were determined biweekly during the study. With the exception of higher plasma insulin and C-pep- tide levels during the OGTT (area under the curve) in phase IV (insulin) v phase II (glipizide) (both P < .051, and higher fasting plasma insulin levels (P < .061, there were no consistently significant meta- bolic differences between phases IV and II. Plasma glucose levels, fasting and during the OGTT, were similar in the two phases and there were no signif- icant differences between phases IV and II in BP levels, body weights, and lipid metabolism mea- sures. Thus, compared with insulin, glipizide treatment in patients with NIDDM is associated with better insulin sensitivity and with lower insu- linemia, but these improvements are not associated with improvements in BP control. Am J Hypertens 1995;8:445453

KEY WORDS: Insulin, glipizide, blood pressure, dia- betes mellitus, insulin resistance.

Received August 3, 1994. Accepted January 9, 1995. From the Division of Endocrinology Metabolism and Hyper-

part in the American Federation of Clinical Research Meeting, Chi- cago, Illinois, November 1993.

tension, Wayne State University School of Medicine, Detroit, Michigan.

Address correspondence and reprint requests to Joseph Levy,

This work was accomplished by a grant 88-R-022 DR from Roerig MD, Wayne State University School of Medicine, University Health Center, 4th Floor, POD H, 4201 St. Antoine, Detroit, MI

Division Pharmaceutical Group, Pfizer Inc. (JL, GG), and National 48201. Institutes of Health grant SO7RRO5384 (JL). It was presented in

0 1995 by the American ]ournal of Hypertension, Ltd. 0895-7061/9X$9.50 0895-7062(95)00052-Q

446 LEVY ET AL AIH-MAY 1995-VOL. 8, NO. 5, PART 1

C onsiderable evidence suggests that essen- tial hypertension is associated with insulin resistance.i4 Insulin resistance is also a hallmark of non-insulin-dependent diabe-

tes mellitus (NIDDM).’ Since the prevalence of hy- pertension among patients with NIDDM is twice as high as its prevalence in nondiabetic, age-matched individuals, it is possible that the underlying insulin- resistant condition in NIDDM predisposes the sub- jects with diabetes to hypertension.2’5 Consequently, improvement in the insulin sensitivity may lower their blood pressure (BP) levels. Therefore, hypogly- cemic agents that decrease the insulin resistance may improve the BP levels in the NIDDM patients.

Treatment of NIDDM patients with insulin may normalize their blood glucose levels, but there is con- troversy to what extent it also improves their under- lying chronic insulin-resistant state.6-8 Some im- provement may be seen secondary to the decrease in blood glucose levels.’ In addition, the association be- tween hyperinsulinemia and hypertension” might suggest that insulin treatment may actually increase the BP in insulin-treated patients. On the other hand, since sulfonylurea agents may improve insulin ac- tion, i1,i2 it is possible that, in addition to their hypo- glycemic effect, they may have an additional benefi- cial effect, that of lowering the BP levels.

To test this hypothesis we designed a four-phase prospective, crossover study aimed at determining the potential BP-lowering effects of treatment with glipizide, as compared with treatment with insulin, in a population of patients with NIDDM.

MATERIALS AND METHODS

Study Population Patients were recruited from the diabetes center at Wayne State University Medical Center. The inclusion criteria for patients were: a) male and female NIDDM patients aged 35 to 70 years; b) prior to inclusion the patients had to be treated with sulfonylurea or diet only; if insulin-treated, the patients had to be off insulin for at least 1 month prior to entry into the study; c) patients had to be less than 150% of ideal body weight; d) their fasting and 60- min stimulated (75 g oral glucose load) plasma C-pep- tide levels had to be higher than 1.0 and 2.0 ng/mL, respectively; e) their screening fasting plasma glucose levels performed on the day of their initial interview had to be less than 250 mg/dL when patients were without any blood glucose lowering pharmacologic therapy. The exclusion criteria included insulin- dependent diabetes, hepatic dysfunction, renal fail- ure, history or presence of significant hematologic, neurologic, cardiac, or gastrointestinal disorders, al- lergy or sensitivity to sulfa drugs, alcoholism, street drugs and nicotine abuse or psychiatric disorder, and, if female patients were enrolled, assurance that they were not pregnant.

The baseline evaluation included complete history and physical examinations, complete blood cell count and differential, measurements of serum glutamic ox- alic transaminase (SGOT), lactic dehydrogenase (LDH), alkaline phosphatase, bilirubin, creatinine, electrolytes and blood urea nitrogen (BUN), urinaly- sis, fasting plasma glucose and insulin, glycated he- moglobin, plasma fructosamine, and a plasma lipid profile, ie, total cholesterol (TC), high density li- poprotein cholesterol (HDL-C), low density lipopro- tein cholesterol (LDL-C), and triglycerides.

Twenty-seven patients who fulfilled the inclusion criteria entered the study but only 19 completed all four phases. (One patient had primary sulfonylurea failure, 3 were lost to follow-up, and 4 patients with- drew from the study because they found that it re- quired too great a time commitment). Since this was a four-phase cross-over study, patients who did not complete all four phases were not analyzed. Analyz- ing the data from the patients who dropped out would have been particularly difficult since patients dropped at different stages of the study. Of the 19 patient who completed the four phases of the study, 10 had normal BP levels (defined as BP < 140/90) and the remaining nine were mildly hypertensive (BP > 140190 but <180/100; Table 1) and were on different regimens of antihypertensive treatments (calcium channel blockers, P-blockers, angiotensin converting enzyme inhibitors, or diuretics). In the treated pa- tients, the antihypertensive medications were not changed (neither in dose nor in the type of antihy- pertensive agent) during the whole study. Since the study was designed as a crossover study, any effects on BP levels different from those induced by the gly- cemic control treatment modalities are expected to be the same in all four phases. Consequently, compari- sons between the effects of the different glycemic control treatment modalities on prevailing BP levels can be done. Since the size of the patient sample was not large and since there was no difference between the hypertensive and the nonhypertensive NIDDM patients with regard to the effects of the two hypo- glycemic therapies on all the parameters of BP control tested in phase IV 21 phase II (supine systolic BP, su-

TABLE 1. BASELINE CLINICAL CHARACTERISTICS -

Values Range

Height (cm) 173.6 ? 10.7 150-196 Weight (kg) 97.9 2 19.6 63.2-144 BMI 31.5 2 4.1 27.1-34.5 Supine systolic BP (mm Hg) 144.6 2 16.5 116-177 Supine diastolic BP (mm Hg) 82.0 e 11.2 60-100 Sitting systolic BP (mm Hg) 147.7 r+_ 15.5 120-178 Sitting diastolic BP (mm Hg) 85.2 + 10.3 64-100 Pulse rate (beatsimin) 77.4 k 10.2 57-100 Respiration (breathsimin) 14.7 ? 2.3 12-20

Results are gwen as mean i- SD.

A/H-MAY 1995-VOL. 8, NO. 5, PAR7 1 EFFECT OF GLIPIZIDE AND INSULIN ON BLOOD PRESSURE 447

pine diastolic BP, sitting systolic BP, sitting diastolic BP, all P > .lO), the data from all 19 patients were pooled for the analysis.

Of the 19 patients, five (26.3%) were women and 14 (73.7%) were men. Four (21.1%) were black and 15 (78.9%) were white. Duration of the diabetes was 8.5 ? 5.9 years (mean ? SD), with a range of 1.1 to 23.0 years. The range of ages of the patients was 33 to 67 years, with a mean of 52.4 ? 9.1 years.

Study Design A four-phase crossover study was designed. The four phases of the study included: Phase I, baseline, 1 month washout in which all hy- poglycemic medications were discontinued and the patients were on diet only; Phase II, 3 months of treatment with glipizide (Glucotrol; Roerig, Pfizer Pharmaceuticals, NY); Phase III, 1 month of washout (diet only, no hypoglycemic medications); Phase IV, 3 months of treatment with insulin (Lente, regular or 70/30-N.P.H/Regular, Novolin insulin, NOVO, Prince- ton, NJ).

Patients monitored their capillary blood glucose (CBG) 4 times a day, ie, before each meal and before bedtime, and brought their records on their biweekly visits. Every 2 weeks they were evaluated for their BP, pulse rate, weight, fasting (FPG) and 2-h post- prandial plasma glucose (2 PPG) levels, plasma fruc- tosamine, compliance with the diet and treatment regimens, and their overall well-being. The dosages of insulin and glipizide were adjusted every 2 weeks according to their CBG, FPG, 2 PPG, and fruc- tosamine levels. Treatment with glipizide was started at 5 to 10 mgiday and was titrated up to achieve eu- glycemia (FPG 5 140 mg/dL) or until maximum dose of glipizide (40 mgiday) was used. Average glipizide dose was 30.2 ~fr 13.2 mgiday (mean ? SD), with a range of 7.5 to 40 mgiday. Insulin doses ranged be- tween 10 and 105 units (43.2 ? 25.7 units). To assure similar level of glycemic control” during the glipizide and insulin treatment, insulin dose in each patient was titrated to achieve a glycemic control similar to that achieved during the preceding glipizide treat- ment. Both insulin and glipizide were given in di- vided doses, an AM and a PM dose, taken Z-h before breakfast and supper. All the patients were put on an American Diabetes Association diet (ADA diet) con- sisting of 55 to 60% carbohydrates and 0.85 g/kg body weight protein; the rest of the calories (up to 30%) came from fat. Total cholesterol was calculated to be ~300 mgiday. l3 The total amount of calories recom- mended took into account the level of physical activ- ity and height and weight of the patients. The range of the total calories was between 1600 and 2400 cali day. All the patients were nonsmokers, and did not drink alcohol except at occasional social events.

At the end of each phase the following assessments were performed: 1) pulse rate, weight, BP (left hand-

supine and sitting positions), 2) glycated hemoglobin (Glyc Hgb) and fructosamine levels, 3) 3-h oral glu- cose tolerance test (OGTT) in which plasma samples for glucose and insulin levels were obtained at 0, 30, 60, 90, 120 and 180 min postingestion of 75 g of glu- cose. Samples obtained at 0, 60, and 120 min were also analyzed for C-peptide levels. Tests were done at 9 AM, before the daily hypoglycemic medication was taken.

During phases I, II (glipizide treatment), and IV (insulin treatment), fasting plasma levels of triglycer- ides, total cholesterol, and HDL-C were measured every 2 weeks. Glyc Hgb levels were measured monthly.

All the participants in the study gave informed con- sent. The study protocol was approved by the Hu- man Investigation Committee of Wayne State Univer- sity. BP and pulse rate measurements were measured after X-h of rest in the supine and sitting positions using the appropriate cuff (according to degree of obesity) and an electronic monitoring device (Dina- map, portable vital signs monitor model 8100T, Cri- tikon Johnson and Johnson, Tampa, FL). In compli- ance with the American Heart Association recom- mendations, all BP measurements were done on the nondominant arm, which turned out to be the left arm in all members of our study population. The mean error for BP measurements of this device is k2.4 to 4.1 mm Hg and accuracy for pulse rate mea- surements is +3.5%.

Analytical Methods To assess level of obesity, body mass index (BMI) was calculated as the weight in ki- lograms divided by the height in meters squared (Ta- ble 1). A BMI of 25 is generally considered to be the upper limit of normal.14

Glucose was analyzed using the glucose oxidase method’” and capillary blood glucose was monitored using the One Touch glucometer (Lifescan Inc., Johnson and Johnson, Milpitas, CA). Plasma insulin levels were measured using a ‘251-insulin radioimmu- noassay (RIA) kit (Incstar, Stillwater, MN) with an intraassay variation of 4.2% and interassay variation of 8.5%.16 C-peptide was analyzed using the same RIA kit, with an intraassay variation of 6.1% and in- terassay variation of 14.3% .I7 Plasma fructosamine was analyzed using the Rotag fructosamine assay (Roche Diagnostic Systems, Nutley, NJ), with an in- traassay variation of 1.9% and interassay variation of 2.4%i8 (normal range, 1.5 to 2.7 nmol/L). Glyc Hgb was measured using the Glyc-Affin GHb kit, (Isolab, Akron, OH), with intra- and interassay variations cal- culated to be ~3% (normal range, 4% to 8%). The rest of the tests were performed in the Detroit Medical Center Clinical Chemistry Laboratories using SMA-12 measurements (Technicon Instruments, Tarrytown, NY). Triglycerides, TC, HDL-C, and LDL-C were

448 LEVY ET AL A]H-MAY 199%VOL. 8, NO. 5, PART 1

measured and calculated using previously described methodologies ig,” (Ektachem, Eastman Kodak, Rochester, NY). Areas under the OGTT curves were calculated as previously described.*r

Statistical Methods Paired t test wasused to com- pare between pairs of phases for the variables of in- terest. This statistic is based on the intrasubject dif- ferences and is appropriate for studying the differ- ences before and after an intervention in a group of patients. ** The comparisons between patients receiv- ing and not receiving antihypertensive medication were made using a one-way analysis of variance (ANOVA). The Pearson’s correlation23 test was used to analyze correlations between insulin and glipizide dosages and BP values (supine systolic BP, supine diastolic BP, sitting systolic BP, sitting diastol- ic BP) and correlation between insulin dosage and the delta BP values (changes between BP levels in phase IV ZI phase III). All the tests including correlation eval- uations were performed using BMDP personal com- puter software (BMDP Statistical Software Inc., Los Angles, CA.). A P < .05 was considered statistically significant. Results are given as the mean 5 SEM un- less specified otherwise.

RESULTS

Characteristics of the patients that participated in the study are summarized in Tables 1, 2, and 3. As a group, the study population tended to be overweight and their BP levels ranged between normal and mild hypertensive values. Levels of their fasting and 2-h postprandial plasma glucose when diet treated only were: fasting 233.4 + 15 mg/dL; postprandial, 287.8 ? 19.1 (phase I), and 250.7 ? 17.1 and 308.8 5 21.5 (phase III) (mean + SE), respectively. While our cri- terion for conclusion into the study was a fasting

plasma glucose ~250 mg/dL, this measurement was done only once, on the day of the interview. How- ever, the fasting plasma glucose levels as they appear in Table 2 reflect averages of repeated FPG measure- ments (every 2 weeks) during each treatment phase. Consequently, these numbers differed on certain oc- casions from the initial value, which was ~250 mg/ dL. Therefore, in some patients the average FPG at the baseline period ended up being higher than 250 mg/dL. Both glipizide and insulin improved blood glucose levels (Tables 2 and 3), compared with the corresponding preceding periods in which the pa- tients were on diet therapy only. There were im- provements in CBG and 2 PPG during phase II com- pared with phase I, and in CBG, FPG, and 2 PPG during phase IV compared with phase III. Statistical significance was not reached for the parameters that reflect longer periods of glycemic control (fruc- tosamine and Glyc Hgb), possibly because the peri- ods of baseline-I (1 month) and washout-III (1 month) were not long enough to notice appreciable compar- isons using these parameters. However, glycemic control during the glipizide (II) and insulin (IV) phases was identical by all the parameters (Tables 2 and 3) used. Body weight, plasma total cholesterol, triglycerides, HDL-C, and LDL-C were similar throughout the four phases of the study. In both phase II and phase IV, glucose levels during OGTT were lower than those at phase I (P < .05 for all timepoints) (Figure 1A). However, in spite of similar glucose values, the insulin values (Figure 1B) during the OGTT in the insulin treatment phase were higher than the insulin values in the glipizide phase (phase IV) and were similar to those seen at phase I in which higher glucose values were obtained. This suggests improved insulin sensitivity during the glipizide phase, compared with the insulin phase. Also the

TABLE 2. GLYCEMIC CONTROL. BP LEVELS, PULSE, AND BODY WEIGHTS DURING THE STUDY

Phase Type

Phase I (Baseline)

Phase II Phase III (Glipizide (Washout Treatment) Phase)

Phase IV (Insulin

Treatment)

Changes

Phase II-I Phase IV-III

CBG (mg/dL) 201.7 k 11.2 136.1 2 21.4 185 2 23.2 86.5 i- 28.7 -63.5 +- 17.8 FPG (mg/dL) 233.4 2 15 203.5 k 14.5 250.7 + 17.1 178 * 11.5 -25.5 * 14.2 FPI ($J/mL) 15.4 k 1.6 17.5 k 2.8 18.6 k 4.1 24.1 t 2.8 2.3 * 2.4 2 PPG (mg/dL) 287.8 k 19.1 246.8 -+ 17.5 308.8 k 21.5 238.9 L 14.9 -40.6 2 11.1 Glyc Hgb-(n = 4-8) (%) 11.6 ? 0.6 11.0 -+ 0.5 11.4 k 0.5 10.7 f 0.5 -0.87 f 0.4 Fructosamine-(n = 1.527) (nmolk) 3.7 zk 0.1 3.6 2 0.2 3.3 k 0.1 3.1 * 0.1 1.3 f 1.4 Area under glucose curve (3 h) 854 ? 51 735 ” 48.0 701 -r- 40.2 -115 k 28.5 Area under insulin curve (3 h) 99.4 t 10.0 101 + 9.0 125 ir 15.1 6.35 k 8.5 Area under C-peptide curve (2 h) 674.7 ? 75.6 675 2 65.3 867.6 2 112.5 22.5 t 56.6 Supine systolic BP (mm Hg) 145.8 t 5.2 144 -t 4.1 138.8 2 2.8 142.4 -c 4.2 -2.0 r+_ 4.36 Supine diastolic BP (mm Hg) 77.8 k 2.7 73.3 k 2.8 75.9 2 2.7 76.6 2 2.3 -4.11 2 2.9 Sitting systolic BP (mm Hg) 146.4 k 4.0 145.2 t 4.1 146.8 -+ 2.5 149.7 2 3.8 -1 2 3.4 Sitting diastolic BP (mm Hg) 82.6 -c 3.0 84 rt 2.5 83.6 + 2.8 87.2 2 2.46 -2.2 + 3.3

CBG, capillay blood glucose; FPG, fasting plasm glucose; FPI, fasting plasm insulin; 2 PPG, 2 h postprandial plasma glucose.

-93.3 5 30.0 -72.7 + 15.3

6.9 -+ 4.3 -75.2 k 18.8 -0.55 ? 0.4 -0.27 k 0.1

4.6 5 4.2 0.89 2 2.8 4.28 k 3.8 3.72 -1- 3.2

AIH-MAY 1995-VOL. 8, NO. 5, PART 1 EFFECT OF GLIPIZIDE AND INSULIN ON BLOOD PRESSURE 449

TABLE 3. COMPARISON BETWEEN PHASES (LEVEL OF SIGNIFICANCE, P VALUES)*

Phase IV u Phase III Phase IV v Phase II

Phase II v Phase I (Insulin Treatment (Insulin Treatment (Glipizide Treatment Phase v Washout Phase v Glipizide

Phase v Baseline) Phase) Treatment Phase)

CBG (mg/dL) ,003 .006 .08 FPG (mg/dL) .097 .002 .166 FPI ( pU/mL) .41 .27 .06 2 PPG (mg/dL) .002 .002 .539 Glyc-Hgb (%) .089 .247 .520 Fructosamine (nmol/L) ,408 ,073 .277 Supine systolic BP (mm Hg) .660 ,296 .693 Supine diastolic BP (mm Hg) ,183 ,761 .134 Sitting systolic BP (mm Hg) ,777 ,287 ,297 Sitting diastolic BP (mm Hg) ,532 .273 .358 Area under glucose curve (3 h) ,002 .399 Area under insulin curve (3 h) ,494 .031 Area under C-peptide curve (2 h) .72 .02

*P based on paired t test; CBG, capilhry blood glucose; FPG, fasting plasma glucose; FPZ, fasting plasma insulin; 2 PPG, 2-h postprandial plasma glucose. P < .05 was considered significant.

calculated areas under the glucose curves (Figure 2A) were lower during the glipizide and insulin phases, compared with those during phase I. However, the calculated areas under the curve for insulin and C-peptide during the glipizide phase (II) were lower than during the insulin phase (IV), and similar to those during phase I (Figures 28, C), confirming that the treatment with glipizide was associated with im- provement in insulin sensitivity. Since both plasma C-peptide and plasma insulin levels were higher dur- ing the insulin treatment phase, this would indicate that the insulin levels were higher not because of dif- ferences in clearance rates of insulin in the different treatment phases, but rather secondary to increased pancreatic secretion due to changes in insulin sensi- tivity. Observations of lower fasting plasma insulin levels in the glipizide phase compared to the insulin phase (17.5 ? 2.5 v 24.1 + 2.7 PUlmL, P < .06) were in agreement with that conclusion.

BP levels (both systolic and diastolic, supine and sitting positions) were not different throughout the study (Figure 3). Further, there was no correlation between the insulin dosages (10 to 105 units/day, mean, 43.2 -+ 25.7 units) and any of the BP values, or between insulin dosages and the delta of the BP val- ues (difference between values at phase IV and phase III; Table 4). There was also no correlation between the daily glipizide dosages (7.5 to 40 mg) and BP levels,

DISCUSSION

Hyperinsulinemia and insulin resistance may play a role in the athogenesis of hypertension, NIDDM, and obesity. P f5,10 In conditions in which all these clin- ical manifestations exist, treatment modalities that

decrease the insulin resistance and hyperinsulinemia may have beneficial effects on these three clinical en- tities. In agreement with these assumptions is the common belief that measures such as diet and exer- cise, which decrease the insulin resistance, also im- prove glucose homeostasis, BP levels, and the degree of obesity of NIDDM patients.2s3s

Therefore, it could be anticipated that sulfon lurea (glipizide) therapy that improves insulin action L may also lower BP levels in the NIDDM patients.5,10 How- ever, our results do not confirm these expectations. There was no change in any of the BP parameters in the NIDDM patients when treated either with glipi- zide or with insulin. These results differ from those of Randeree et a1,39 who reported that changing the treatment of NIDDM patients from oral hypoglyce- mic medications (sulfonylurea or biguanide) to insu- lin is associated with an increase in their BP. How- ever, in that retrospective study the treatment of the diabetic patients with insulin was associated with a marked improvement in their blood glucose levels and with significant increase in their body weights. The observed reduction in the blood glucose levels following the initiation of insulin treatment in the NIDDM patients in that study, from 18.36 to 10.4 mmol/L, is expected to reduce markedly the glycos- uria of the diabetic patients and increase their blood volume, which thus could lead to the increase in their BP.39,40 In the current study, the glycemic levels and body weights remained similar during the glipizide and insulin treatment phases. This similar glycemic control was achieved because we selected only pa- tients with sufficient endogenous insulin secretion who responded to sulfonylurea therapy and could maintain blood glucose levels similar to those ob-

450 LEVY ET AL A]H-MAY 1995-VOL. 8, NO. 5, PART 1

(4 350 + Basal

T T -47 Gllplzlde m-o-. hS”lul

0 30 60 90 120 150 180

Time (min)

0 30 60 90 120 1.50

Time (min)

FIGURE 1. Glucose and insulin curues at the end of phases 1 (basal), II (glipizide), and IV finsulini during 75-g OGTT. A) Levels of plasma glucose at phase I were higher than levels at phases 11 and IV (all P < .05). B) Levels of plasma insulin at 90, 120, and 180 min during the insulin phase were higher compared with levels during the glipizide and basal phases.

tained during the insulin treatment phase. In contrast to Randeree et al, Schmitt and Moore found that changing the treatment of NIDDM patients from chlorpropamide to insulin was associated with an ac- tual decrease in their BP levels.41 It was also reported that BP during glyburide treatment was higher than during insulin treatment in NIDDM patients and that glyburide and tolbutamide may increase BP in animals.42,43

However, metformin treatment, which improves insulin sensitivity, lowered arterial BP levels in spon- taneously hypertensive rats.& Likewise, troglitazone, a new potent oral hypoglycemic agent which de- creases insulin resistance in NIDDM patients and in- sulin-resistant animal models, restores the increased BP levels in fructose-induced insulin-resistant hyper- tensive rats back to norma1.45 Thus there is no uni- formity in the reports with regard to the effect of oral

1000

750

500

250

a

150

loo

50

0

loo0

750

a

T

Basal Glipizide Insulin

l PC005

-I- 1

T

Basal Glipizide Insulin

k pd.05

Basal Glipizide Insulin

FIGURE 2. Areas under the glucose, insulin, and C-peptide curves given in arbitrary units. A) Areas under the glucose curves during phases 11 and IV were similar and significantly lower than at phase 1. B) Areas under the insulin curves were higher during the insulin phase compared with the glipizide phase. C) Areas under the curve of the C-pepfide levels. Values were higher during the insulin phase compared with the glipizide phase.

AJH-MAY 2995-VOL. 8, NO. 5, PART 1 EFFECT OF GLIPIZIDE AND INSULIN ON BLOOD PRESSURE 451

EI N I n nr IV

Systolic Diastolic I II “I N

Systolic I II IIC N

Diastolic

FIGURE 3. Supine and sitting b/ml pressure levels (systolic and diastolic) during the four phases of the study. The BP levels did not differ during the study and were similar at the insulin and glipizide phases.

hypoglycemic medications on BP levels in NIDDM patients and in insulin-resistant animal models. It seems, therefore, that different hypoglycemic agents may have different effects on the BP levels in insulin- resistant states and this BP-lowering effect may depend on the extent and the mechanism by which these agents improve the peripheral insulin resistance.1’,45,46

Although we show that the glipizide treatment was associated with some improvement in insulin sensi- tivity, as determined by the lower insulin and C-pep- tide levels during the OGTT and the lower fasting plasma insulin levels in the face of similar glucose levels, this improvement was not sufficient to cause a decrease in BP levels in the NIDDM patients. How- ever, it is possible that other oral hypoglycemic agents with a more potent effect in reducing the pe- ripheral insulin resistance might improve the BP also in the NIDDM patients.45,46

A potential explanation for why we did not observe

a difference in BP levels between the glipizide and insulin phases may be that, because it has been sug- gested that insulin itself may have a vasodilatory ef- fect, 47-50 it is possible that such an effect compen- sated for the higher insulin resistance in the insulin treatment phase; therefore, no difference in BP levels between the two phases could be appreciated. How- ever, since the insulin vasodilatory effects are acute, 48,49 and as it has been shown that insulin loses its ability to increase peripheral blood flow in NIDDM patients,51-54 such an explanation will be unlikely. Alternatively, a possible explanation for the dissoci- ation of the effects of glipizide on insulin sensitivity and on BP may stem from current recognition that the term insulin resistance includes heterogeneous enti- ties that involve different pathways. It is possible, therefore, that while the glipizide treatment im- proved the insulin-induced glucose utilization at the target cell 1evel,*1,12~24 it did not affect other aspects of the insulin resistant condition which might be of more significance for BP levels in the NIDDM pa- tients. It has been suggested that decreased muscle blood flow contributes to the insulin resistance of obesity and diabetes.51’52,54 The same defect may in- crease the peripheral vascular resistance and the level of the BP in the NIDDM patients.55 While sulfonyl- urea can increase glucose transport at the cellular level, *1,24 there is no evidence that these drugs im- prove peripheral blood flow, which might be more relevant for regulation of the BP level. Our results do not support the notion of direct vascular dilatory properties of glipizide, since BP levels remained the same during the glipizide and insulin treatment phases. However, some subclinical effect cannot be ruled out.

In summary, glipizide treatment of NIDDM pa- tients improves their insulin-induced glucose utili- zation. However, it does not result in improved BP control.

TABLE 4. CORRELATION OF INSULIN DOSAGE WITH BLOOD PRESSURE VALUES AND DELTA BLOOD PRESSURE SCORES

Correlation Probability Variable n Mean SD Coefficient (P Value)

LSBP (mm Hg) 19 142.36 18.49 0.00626 NS LDBP (mm Hg) 19 76.58 7.79 - 0.10748 NS SSBP (mm Hg) 19 149.68 16.60 - 0.10162 NS SDBP (mm Hg) 19 87.15 10.56 - 0.42781 .08 ALSBP (mm Hg) 18 4.6 18.15 0.08034 NS ALDBP (mm Hg) 18 0.8 12.19 0.04923 NS ASSBP (mm Hg) 18 4.2 16.5 0.08257 NS ASDBP (mm Hg) 18 3.7 13.9 - 0.15882 NS INSDOSE (units) 19 43.15 25.7

~-~-___ LSBP, supine systolic blood pressure; LDSP, suptne diastolic blood pressure; SSBP, sitting systolic blood pressure; SDBP, sitting diastolic blood

pressure; INSDOSE, insulin dosage. ABP was calculated by subtracting BP values at phase 111 from corresponding phase IV BP levels.

452 LEVY ET AL

ACKNOWLEDGMENTS

We appreciate the work of Dr. Marcia Levenstein for her help in the statistical analysis of the data and thank Ms. Lois Wiggins for her excellent secretarial help.

17

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

REFERENCES

Ferrannini E, Buzzigoli G, Bonadonna R, et al: Insulin resistance in essential hypertension. N Engl J Med 1987;317:350-357.

Defronzo RA, Ferrannini E: Insulin resistance: a mul- tifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic car- diovascular disease. Diabetes Care 1991;14:17%193.

Swislocki ALM, Hoffman BB, Reaven GM: Insulin re- sistance, glucose intolerance and hyperinsulinemia in patients with hypertension. Am J Hypertens 1989;2: 419423.

Pollare T, Lithe11 H, Berne C: Insulin resistance is a characteristic feature of primary hypertension inde- pendent of obesity. Metabolism 1990;39:167-174.

Levy J, Gavin JR III, Sowers JR: Diabetes mellitus: a disease of abnormal cellular calcium metabolism? Am J Med 1994;96:260-273.

Scarlett JA, Gray RS, Griffin GJ, et al: Insulin treat- ment reverses the insulin resistance of type II diabetes mellitus. Diabetes Care 1982;5;353-363.

Hidaka H, Naqulesparan M, Klimes I, et al: Improve- ment of insulin secretion but not insulin resistance after short-term control of plasma glucose in obese type II diabetes. J Clin Endocrinol Metab 1981;54:217- 222.

Ginsberg H, Rayfield EJ: Effect of insulin therapy on insulin resistance in type II diabetic subjects. Evidence for heterogeneity. Diabetes 1981;30:739-745,

Rossetti L, Smith D, Shulman GI, et al: Correction of hyperglycemia with phlorizin normalizes tissue sensi- tivity to insulin in diabetic rats. J Clin Invest 1987;79: 1510-1515.

Modan M, Halkin H, Almog S, et al: Hyperinsuline- mia: a link between hypertension obesity and glucose intolerance. J Clin Invest 1985;75:809-817.

Beck-Nielsen H, Hother-Nielsen 0, Pedersen 0: Mechanism of action of sulfonylureas with special ref- erence to the extrapancreatic effect: an overview. Di- abetic Med 1988;5:613-620.

Gavin JR III: Dual actions of sulfonylureas and gly- buride: receptor and postreceptor effects. Am J Med 1985;79:3&42.

Capro AA: Dietary management, in Kahn CR, Weir CG (eds): Joslin’s Diabetes Mellitus. Philadelphia, Lea & Febiger, 1994, p 415430.

Flier JS: Obesity, in Kahn CR, Weir CG (eds): Joslin’s Diabetes Mellitus. Philadelphia, Lea & Febiger, 1994, pp 351-371.

Kadish AH, Little RL, Sternberg JC: A new and rapid method for the determination of glucose by measure- ment of rate of oxygen consumption. Clin Chem 1968; 14:1X-131.

Morgan CR, Lazarow A: Immunoassay of insulin two

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

AJH-MAY 199FVOL. 8, NO. 5, PART I

antibody systems: plasma insulin levels of normal, subdiabetic and diabetic rats. Diabetes 1963;12:115 122.

Rendell M: C-peptide levels as a criterion in treatment of maturity-onset diabetes. J Clin Endocrinol Metab 1983;57:1198-1206.

Lloyd D, Marples J: Simple calorimetry of glycated serum protein in centrifugal analyzer. Clin Chem 1984;30:1686-1688.

Finely PR, Shifman RB, Williams RJ, et al: Cholesterol in high-density lipoprotein: use of Mg*+/dextran sul- fate in its enzyme determination. Clin Chem 1978;24: 931-933.

Allain CC, Poon LS, Chan CSG, et al: Enzymatic de- termination of total cholesterol in serum. Clin Chem 1974;20:470475.

Tai MM: A mathematical model for determination of total area under glucose tolerance and other metabolic curves. Diabetes Care 1994;17:152-154.

Comparing two or more sets of data, in BF Ryan, BL Joiner, TA Ryan, Jr. (eds): Minitab Handbook Boston, PWS-Kent, 1992, pp 92-120.

Summarizing data, in Dawson-Saunders B, Trapp RG (eds): Basic & Clinical Biostatistics. Norwalk, CT, Ap- pleton & Lange, 1994, pp 41-63.

Lebovitz HE, Feinglos MN, Bucholtz HK, et al: In- creased insulin effectiveness: an extrapancreatic action of glipizide in man, in Glipizide, A Worldwide Re- view. Princeton, Excerpta Medica, 1992, pp 3045.

Reaven GM: Role of insulin resistance in human dis- ease. Diabetes 1988;37:1595-1607.

Haffner SM, Fong D, Hazuda HP, et al: Hyperinsu- linemia, upper body adiposity, and cardiovascular risk factors in non-diabetics. Metabolism 1988;37:338- 345.

Henry RR, Wallace I’, Olefsky JM: Effects of weight loss on mechanisms of hyperglycemia in obese non- insulin-dependent diabetes mellitus. Diabetes 1986;35: 99c-998.

Zawadzki JK, Bogardus C, Foley JE: Insulin action in obese non-insulin-dependent diabetics and in their isolated adipocytes before and after weight loss. Dia- betes 1987;36:227-235.

Bogardus C, Lillioja S, Mott DM, et al: Relationship between degree of obesity and in vivo insulin action in man. Am J Physiol 1985;248:E286-E291.

Bogardus C, Ravussin E, Robbins DC, et al: Effects of physical training and diet therapy on carbohydrate metabolism in patients with glucose intolerance and non-insulin-dependent diabetes mellitus. Diabetes 1984;33:311-318.

Lamarche B, Despres JP, Pouliot MC, et al: Is body fat loss a determinant factor in the improvement of car- bohydrate and lipid metabolism following aerobic ex- ercise training in obese women? Metabolism 1992;41: 1249-1256.

Paiizkova J: Obesity and its treatment by diet and ex- ercise. World Rev Nutr Diet 1993;72:7891.

Krotkiewski M, Lonnroth P, Mandroukas Z, et al: The effect of physical training on insulin secretion and ef-

A]H-MAY 1995VOL. 8, NO. 5, PART 1 EFFECT OF GLIPIZIDE AND INSULIN ON BLOOD PRESSURE 453

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

fectiveness and on glucose metabolism in obesity and type II (non-insulin dependent) diabetes Mellitus. Di- abetologia 1985;28:881-890.

Numata K: Very low calorie diet-induced weight loss reverses exaggerated insulin secretion in response to glucose, arginine and glucagon in obesity. Int J Obes 1993;17:103108.

Seals DR, Hagberg JM: The effect of exercise training on human hypertension: a review. Med Sci Sports Ex- erc 1984;16:207-215.

Nelson L, Esler MD, Jennings GL, et al: Effect of changing levels of physical activity on blood pressure and hemodynamics in essential hypertension. Lancet 1986;ii:473476.

Krotkiewski M, Mandourkas K, Sjstrom L, et al: Ef- fects of long term physical training on body fat, me- tabolism, and blood pressure in obesity. Metabolism 1979;28:650-658. Bogardus C, Ravussin E, Robbins DC, et al: Effects of physical training and diet therapy on carbohydrate metabolism in patients with glucose intolerance and non-insulin-dependent diabetes mellitus. Diabetes 1984;33:311-318. Randeree H, Omar MAK, Motala AA, et al: Effect of insulin therapy on blood pressure in NIDDM patients with secondary failure. Diabetes Care 1992;15:1258 1263.

Grunberger G, Levy J: Insulin therapy, blood pres- sure, and NIDDM. Diabetes Care 1993;16:954.

Schmitt JK, More JR: Hypertension secondary to chlorpropamide with amelioration by changing to in- sulin. Am J Hypertens 1993;6:317-319. Peuler JD, Johnson BAB, Phare SM, et al: Sex-specific effects of an insulin secretagogue in stroke-prone hy- pertensive rats. Hypertension 1993;22:214-220.

Melander A, Bitzen P-O, Faber 0, et al: Sulfonylurea antidiabetic drugs: an update of their clinical pharmacol- ogy and rational therapeutic use. Drugs 1989;37:58-72.

Morgan DA, Ray CA, Balon TW, Mark AL: Metformin increases insulin sensitivity and lowers arterial pres- sure in spontaneously hypertensive rats (abst). Hy- pertension 1992;20:421.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

Lee MK, Miles PDG, Khoursheed M, et al: Metabolic effects of troglitazone on fructose-induced insulin re- sistance in the rat. Diabetes 1994;43:1435-1439.

Nolan JJ, Ludvik B, Beerdsen P, et al: Metabolic effects of troglitozone in subjects with impaired glucose tol- erance and insulin resistance (abst). Clin Res 1994;42: 88A.

Riddle MC: Different takes on the relationship of in- sulin treatment to blood pressure. Diabetes Care 1993; 16:95>954.

Anderson EA, Hoffman RI’, Balon TW, et al: Hyper- insulinemia produces both sympathetic neural activa- tion and vasodilation in normal humans. J Clin Invest 1991;87:22462252.

Hall JE, Coleman TG, Mizelle HL: Does chronic hy- perinsulinemia cause hypertension? Am J Hypertens 1989;2:171-173.

Anderson EA, Hoffman RI’, Balon TW, et al: Hyper- insulinemia produces both sympathetic neural activa- tion and vasodilation in normal humans. J Clin Invest 1991;87:224&2252.

Laakso M, Edelman SV, Brechtel G, et al: Decreased effect of insulin to stimulate skeletal muscle blood flow in obese man: a novel mechanism for insulin re- sistance. J Clin Invest 1990;85:1844-1852. Baron AD, Laakso M, Brechtel G, et al: Reduced ca- pacity and affinity of skeletal muscle for insulin- mediated glucose uptake in non-insulin-dependent diabetic subjects. Effects of insulin therapy. J Clin In- vest 1991;87:1186-1194.

Anderson EA, Balon TW, Hoffman RI’, et al: Insulin increases sympathetic activity but not blood pressure in borderline hypertensive humans. Hypertension 1992;19:621-627.

Baron AD, Laakso M, Brechtel G, et al: Mechanism of insulin resistance in insulin-dependent diabetes mel- litus: a major role for reduced skeletal muscle blood flow. J Clin Endocrinol Metab 1991;73:637-643.

Baron AD, Brechtol-Hook G, Johnson A, et al: Skeletal muscle blood flow. A possible link between insulin resistance and blood pressure. Hypertension 1993;21: 129-135.