Euglycemic progression: worsening of diabetic retinopathy in poorly controlled type 2 diabetes in minorities

AIMS

In type 2 diabetes, early effects of strict near-normalization of glucose control on macrovascular and microvascular disease are still uncertain. We evaluated the effects of early dramatic improvement in glycemia on retinal disease in poorly controlled diabetes.

METHODS

A retrospective, case-control study in public hospital patients with type 2 diabetes, who had annual retinal imaging as part of a case management program or standard diabetes care. Patients included had ≥2 two retinal images ≥1 one year apart, and at least 3 HbA1C measurements. Retinal images were graded using a modified Scottish Diabetic Retinopathy grading scheme. An 'intensive' group (n=34) with HbA1C decrease >1.5% was compared with randomly chosen patients (n=34) with minimal HbA1C changes.

RESULTS

Mean HbA1C (±SEM) over two years was similar in intensive (8.5 ± 0.21%) and control groups (8.1 ± 0.28%, p=NS). However, the intensive group had higher baseline HbA1C and a mean maximal decrease of 4.0 ± 0.41% in contrast to the control group (0.2 ± 0.11%). Retinopathy grade progressed +0.7 ± 0.25 units from baseline in the intensive group (p=0.015), a 22.6% worsening. The control group changed minimally from baseline (0.03 ± 0.14 units, p=NS). Change in retinopathy grade was significantly different between groups (p=0.02). More eyes worsened by ≥ 1 retinal grade (p=0.0025) and developed sight-threatening retinopathy (p=0.003) in the intensive group. Visual acuity was unchanged.

CONCLUSIONS

Diabetic retinopathy significantly worsened in poorly controlled type 2 diabetes after early intensification of glycemic control and dramatic HbA1C change. Retinal status should be part of risk-factor evaluation in patients likely to experience marked reductions in HbA1C in poorly controlled diabetes.

Pathophysiology of ketoacidosis in Type 2 diabetes mellitus

AIMS

Despite an increasing number of reports of ketoacidosis in populations with Type 2 diabetes mellitus, the pathophysiology of the ketoacidosis in these patients is unclear. We therefore tested the roles of three possible mechanisms: elevated stress hormones, increased free fatty acids (FFA), and suppressed insulin secretion.

METHODS

Forty-six patients who presented to the Emergency Department with decompensated diabetes (serum glucose > 22.2 mmol/l and/or ketoacid concentrations > or = 5 mmol/l), had blood sampled prior to insulin therapy. Three groups of subjects were studied: ketosis-prone Type 2 diabetes (KPDM2, n = 13) with ketoacidosis, non-ketosis-prone subjects with Type 2 diabetes (DM2, n = 15), and ketotic Type 1 diabetes (n = 18).

RESULTS

All three groups had similar mean plasma glucose concentrations. The degree of ketoacidosis (plasma ketoacids, bicarbonate and anion gap) in Type 1 and 2 subjects was similar. Mean levels of counterregulatory hormones (glucagon, growth hormone, cortisol, epinephrine, norepinephrine), and FFA were not significantly different in DM2 and KPDM2 patients. In contrast, plasma C-peptide concentrations were approximately three-fold lower in KPDM2 vs. non-ketotic DM2 subjects (P = 0.0001). Type 1 ketotic subjects had significantly higher growth hormone (P = 0.024) and FFA (P < 0.002) and lower glucagon levels (P < 0.02) than DM2.

CONCLUSIONS

At the time of hospital presentation, the predominant mechanism for ketosis in KPDM2 is likely to be greater insulinopenia.

Genome scan for blood pressure in Dutch dyslipidemic families reveals linkage to a locus on chromosome 4p

Genes contributing to common forms of hypertension are largely unknown. A number of studies in humans and in animal models have revealed associations between insulin resistance, dyslipidemia, and elevated hypertension. To identify genes contributing to blood pressure (BP) variation associated with insulin-resistant dyslipidemia, we conducted a genome-wide scan for BP in a set of 18 Dutch families exhibiting the common lipid disorder familial combined hyperlipidemia. Our results reveal a locus on chromosome 4 that exhibits a significant lod score of 3.9 with systolic BP. In addition, this locus also appears to influence plasma free fatty acid levels (lod=2.4). After adjustment for age and gender, the lod score for systolic BP increased to 4.6, whereas the lod score for free fatty acid levels did not change. The chromosome 4 locus contains an attractive candidate gene, alpha-adducin, which has been associated with altered BP in animal studies and in some human populations. However, we found no evidence for an association between 2 intragenic alpha-adducin polymorphisms and systolic BP in this sample. We also observed suggestive evidence for linkage (lod=1.8) of diastolic BP to the lipoprotein lipase gene locus on chromosome 8p, supporting a finding previously observed in a separate insulin-resistant population. In addition, we also obtained suggestive evidence for linkage of systolic BP (lod=2.4) and plasma apolipoprotein B levels (lod=2.0) to a locus on proximal chromosome 19p. In conclusion, our genome scan results support the existence of multiple genetic factors that can influence both BP and plasma lipid parameters.

Glucose entrainment of high-frequency plasma insulin oscillations in control and type 2 diabetic subjects

Regular high-frequency oscillations of insulin secretion are characteristic of normal beta-cell function. These oscillations are easily entrainable to an exogenous rhythm by small changes in glucose concentration in vitro. We tested whether high-frequency insulin oscillations in vivo would also be entrainable by glucose and whether a lack of entrainment would characterize the diabetic beta-cell. We tested 13 control subjects and 11 patients with type 2 diabetes. Subjects underwent serial blood sampling at 1-min intervals for 60-120 min in the basal state or with small (15 mg/kg) boluses of glucose injected intravenously at exact 29-min intervals. Time series analysis was carried out using spectral analysis. Oscillations of basal plasma glucose concentrations were observed in both control and type 2 diabetic subjects, with a mean period of 11.3 +/- 3.1 and 11.6 +/- 2.0 min, respectively. These oscillations were entrained to mean periods of 15.0 +/- 0.6 and 14.2 +/- 0.9 min, respectively, by exogenous glucose. Regular high-frequency insulin oscillations were observed in control subjects; the mean period of basal plasma insulin oscillations was 10.7 +/- 1.2 min and was entrained to exogenously injected glucose, with a period of 15.2 +/- 0.1 min. In contrast, in the type 2 diabetic subjects, spontaneous insulin oscillations were unchanged by the glucose rhythm; the mean periods were 10.0 +/- 1.0 min during the basal period, and 10.1 +/- 0.0 min during glucose injections. These results demonstrate that spontaneous high-frequency insulin oscillations can be successfully entrained by glucose in control subjects. However, these oscillations in type 2 diabetic subjects are not similarly entrained. We conclude that loss of entrainment of spontaneous high-frequency insulin oscillations in type 2 diabetes is a highly sensitive manifestation of beta-cell secretory dysfunction.

Counterregulatory hormones oscillate during steady-state hypoglycemia

During hypoglycemia, the magnitude of the counterregulatory response depends on the extent of plasma glucose reduction. However, our clinical observations during steady-state hypoglycemia indicate that symptom severity can change independently of plasma glucose concentrations, i.e., symptoms appeared to fluctuate despite stable glucose levels. This study was therefore designed to test the hypothesis that hormonal and symptomatic responses to hypoglycemia are pulsatile. Seven healthy subjects had serial blood sampling at 3-min intervals during 90 min of insulin-induced hypoglycemia. Mean +/- SE plasma glucose levels plateaued at 62 +/- 3 mg/dl. Counterregulatory hormones were significantly elevated (P < 0.05-0. 01, except norepinephrine) and strikingly pulsatile. Cluster analysis revealed pulses of large magnitude in plasma glucagon, epinephrine, and norepinephrine concentrations. Amplitudes were, respectively, 72 +/- 4, 64 +/- 8, and 48 +/- 3% of the mean. Interpeak intervals were 27 +/- 7, 19 +/- 4, and 25 +/- 5 min, respectively. Symptom score and cardiovascular responses were also pulsatile; their peaks were found to coincide with epinephrine peaks. We conclude that hormonal and symptomatic counterregulation in hypoglycemia, while critically driven by plasma glucose levels, is also influenced by an endogenous pulsatility that exists despite steady-state glucose concentrations.

Erratic oscillatory characteristics of plasma insulin concentrations in patients with insulinoma: mechanism for unpredictable hypoglycemia

Patients with insulin-producing tumors may have hypoglycemic symptoms at unpredictable times. This study evaluated whether plasma insulin oscillations, known to occur in normal individuals but not explored in patients with insulinomas, could be an underlying mechanism for such events. Nine normal subjects and five patients with proven insulinomas were studied in the fasting state. Serial sampling of arterialized blood over 80-100 min, at 2- or 3-min intervals was performed. In normal subjects, mean plasma glucose and insulin concentrations were 5.3 +/- 0.1 mmol/L and 58 +/- 9 pmol/L, respectively. Regular, low-amplitude plasma insulin oscillations were observed, with a period of 10-17 min. The subjects with insulinomas had lower mean plasma glucose and higher insulin concentrations than controls, 3.6 +/- 0.3 mmol/L (P = 0.01) and 150 +/- 42 pmol/L (P = 0.01), respectively. They also had insulin oscillations that appeared unstable as a result of variability in duration and amplitude compared with controls. The insulin pulses were irregular, and interpeak intervals varied between 4-54 min in different subjects; in some subjects, the amplitude was also variable, with sudden spontaneous pulses as high as 565 pmol/L, with an associated glucose decrement. We conclude that large spontaneous bursts of insulin secretion occur in patients with insulinomas as part of an erratic pattern of oscillatory insulin secretion, and these can account for unpredictable occurrences of hypoglycemia.

An overnight insulin infusion algorithm provides morning normoglycemia and can be used to predict insulin requirements in noninsulin-dependent diabetes mellitus

Initial insulin requirements in noninsulin-dependent diabetes mellitus (NIDDM) are difficult to estimate because of individual variability in insulin sensitivity and secretion. We evaluated a simple, nurse-managed algorithm for overnight delivery of insulin, for its ability to provide morning near-normoglycemia and as a means to predict initial insulin requirements in NIDDM. Twenty-seven patients with poorly controlled NIDDM were studied on 30 occasions. A 12-h iv insulin infusion was begun at 2000 h, and bedside blood glucose concentrations were measured at hourly intervals. The rate of insulin infusion was adjusted according to blood glucose levels. We estimated the preprandial insulin dose requirement for the following day in 16 patients based on overnight insulin requirements to maintain normoglycemia. Preprandial insulin doses were adjusted for prevailing blood glucose concentrations. At 2000 h, the mean (+/-SEM) blood glucose concentration was 265.7 +/- 10.8; at 0300 h, it was 122.8 +/- 3.4; and at 0700 h, it was 123.8 +/- 5.1 mg/dL. On the next day, mean blood glucose levels (before and 2 h after a meal) were: breakfast, 102.5 +/- 5.9 and 177.3 +/- 19.2; lunch, 138.9 +/- 15.5 and 136.3 +/- 11.4; dinner, 105.7 +/- 7.2 and 178.1 +/- 15.7 mg/dL. There was no significant difference between mean calculated and administered total insulin dosage the next day (84.2 +/- 7.0 vs. 78.2 +/- 8.2 U). Thus, a weight-based algorithm for iv insulin infusion induced near-normoglycemia in NIDDM and successfully predicted the insulin dose requirement. We conclude that initiating insulin therapy in NIDDM patients can be achieved rapidly and efficiently based on a nurse-managed overnight insulin infusion.

Pseudohypohyperparathyroidism: A Case Study of Differential Resistance to Parathyroid Hormone

OBJECTIVE

To present a rare case of a combination of hyperparathyroidism and hypoparathyroidism in a 30-year-old woman.

METHODS

We review the laboratory, radiographic, and pathologic findings in a healthy-appearing woman who sustained a patellar fracture after a simple fall at home.

RESULTS

Our patient had features of hypoparathyroidism--that is, tetanic crises and hypocalcemia--and also hyperparathyroidism--fracture of the patella, multiple bone cysts, and confirmed osteitis fibrosa cystica on bone biopsy specimens. These features were associated with a high serum level of intact parathyroid hormone (PTH) and parathyroid hyperplasia. A lack of response of nephrogenic cyclic adenosine monophosphate (cAMP) to PTH stimulation was observed along with subnormal serum cAMP responses. In contrast, urine phosphate excretion increased after administration of PTH.

CONCLUSION

These results demonstrate a state of renal PTH resistance in a patient with osteitis fibrosa cystica.

Applications of segmented regression models for biomedical studies

In many biological models, a relationship between variables may be modeled as a linear or polynomial function that changes abruptly when an independent variable obtains a threshold level. Usually, the transition point is unknown, and a major objective of the analysis is its estimation. This type of model is known as a segmented regression model. We present two methods, Gallant and Fuller's (J Am. Stat. Assoc. 68: 144-147, 1973) method and Tishler and Zang's (J. Am. Stat. Assoc. 76: 980-987, 1981) method, using nonlinear least-squares techniques for estimating the transition point. We give the following three examples: a hypoglycemia study, a testosterone study, and an estimate of age-cortisol relationship. Simulation techniques are used to compare the two methods. We conclude that these models provide useful information and that the two methods studied produce essentially equivalent results. We recommend that both methods be used to analyze a data set if possible to avoid problems due to local minima and that if the results do not agree, then evaluation of the likelihood function in the range of the estimates be used to determine the best estimate.

Very-low-energy diets alter the counterregulatory response to falling plasma glucose concentrations

A consequence of short-term very-low-energy diets (VLEDs) in lean subjects is reactive hypoglycemia. We therefore tested the responses of overweight women on prolonged (14 d) VLEDs. Subjects lost 4.8 +/- 0.2 kg (mean +/- SEM, n = 13, P < 0.001). Group A (n = 6) was challenged with an oral-glucose-tolerance test (OGTT) and group B (n = 7) with an oral-sucrose-tolerance test (OSTT) on days 1 and 14. In group A, mean nadir plasma glucose after the OGTT was lower on day 14, 3.75 +/- 0.16 vs 4.7 +/- 0.19 mmol/L (P < 0.01), because of an accelerated rate of glucose decline (RGD, 26.7 +/- 3.3 vs 17.2 +/- 3.9 mumol.l-1.min-1, P < 0.05) late in the OGTT. Plasma insulin was also lower (P < 0.03) and the VLED suppressed two growth hormone (GH) peaks on day 14 (P < 0.05 for each). In group B on day 14, a greater RGD was also observed late in the OSTT, 16.9 +/- 4.1 vs 6.5 +/- 2.0 mumol.L.min-1 (P < 0.03). GH peaks were also significantly suppressed. We conclude that a VLED results in altered glucose regulation late after carbohydrate loading, characterized by an accelerated decline in plasma glucose and GH suppression. Patients on a VLED may be at risk for abnormally low plasma glucose concentrations when ingesting high carbohydrate loads.

Accuracy of plasma glucose measurements in the hypoglycemic range

OBJECTIVE

This study was designed to evaluate three different enzymatic methods for glucose measurement in plasma samples with special emphasis on glucose concentrations in the hypoglycemic range.

RESEARCH DESIGN AND METHODS

Glucose dehydrogenase (Hemo-Cue analyzer), glucose oxidase (YSI analyzer), and hexokinase (Abbott analyzer) methods were used to measure plasma samples that were obtained during research studies.

RESULTS

Mean glucose concentrations (n = 240) were 5.3 +/- 0.2, 5.4 +/- 0.2, and 5.6 +/- 0.2 mM (95.6 +/- 3.9, 96.7 +/- 3.9, and 101.6 +/- 4.0 mg/dl) using glucose dehydrogenase, glucose oxidase, and hexokinase, respectively (NS). In the hypoglycemic range, mean glucose concentrations with each method retained the same hierarchy of measurements: 2.7 +/- 0.05, 2.8 +/- 0.04, and 2.9 +/- 0.03 mM (48.4 +/- 0.9, 50.6 +/- 0.8, and 52.3 +/- 0.6 mg/dl) by glucose dehydrogenase, glucose oxidase, and hexokinase, respectively (P < 0.005). Individual glucose dehydrogenase measurements (n = 240) correlated well with glucose oxidase and hexokinase, r = 0.99, and were considerably easier to perform at the bedside. The differences between the glucose measurement methods were consistent and similar in low, normal, and high concentration ranges.

CONCLUSIONS

We conclude that any interpretation or comparison of critical clinical and research measurements of glucose in different settings take into account methodological differences, particularly in the hypoglycemic range.

Evidence for pancreatic pacemaker for insulin oscillations in low-frequency range

Circulating insulin concentrations oscillate in regular fashion, with periods that fall into a high-frequency (period of 5-17 min) or low-frequency (period of 50-150 min) range. Only the high-frequency oscillations have so far been reported in vitro, suggesting that these derive from a primary pancreatic source. This study tested whether the low-frequency insulin oscillations could also be identified in vitro. Rat islets of Langerhans were perifused for 20 h using RPMI medium with 5.5 mM glucose. Perifusate fractions were collected at 9.9-min intervals. Mean insulin concentrations at the outset were 21.4 +/- 2.9 microU/ml, increased to 32.5 +/- 4.6 (P < 0.05) between 13 and 17 h after the start of perifusion, and then either leveled off or decreased to baseline. Superimposed on this general trend, we found sustained insulin oscillations with a period of 50-100 min. The mean amplitude was 14.2 +/- 4.2 microU/ml, and the amplitude/mean ratio was 64.6 +/- 12%. Spectral analysis revealed significant peaks at periods that were close to either 50 or 100 min and a smaller peak at 24-37 min. These data, using in vitro methodology and constant glucose concentrations, indicate the presence of sustained, spontaneous, low-frequency, ultradian insulin oscillations in the pancreatic islets. This provides evidence for a pancreatic component that may participate in the previously described in vivo ultradian insulin oscillations. This finding may also provide a mechanism for the apparent escape from glucose entrainment of serum insulin oscillations in non-insulin-dependent diabetes mellitus.

Plasma glucose thresholds for counterregulation after an oral glucose load

Glucose counterregulation (GCR) plays an important role in the transition between exogenous and endogenous glucose delivery after an oral glucose load. This response is initiated when plasma glucose concentrations are decreased below threshold levels, previously defined in studies of insulin-induced hypoglycemia. In this study, we tested the plasma glucose thresholds for activation of the GCR response under more physiologic circumstances, ie, after glucose ingestion. We studied 20 normal subjects for 300 minutes after 75 g of oral glucose. Between 150 and 300 minutes, blood samples and symptom scores were obtained at 10-minute intervals. After oral glucose, individual glucose nadirs were observed over a wide time range (160 to 290 minutes). Mean glucose concentrations decreased from 5.3 +/- 0.2 mmol/L at 30 minutes before the nadir (-30 minutes) to 3.8 +/- 0.2 mmol/L at the nadir (0 minutes). Mean plasma epinephrine concentrations increased from 210 +/- 35 pmol/L, were significantly elevated at -10 minutes (P < .05), and peaked at +20 minutes (1,008 +/- 184 pmol/L, P < .001). Mean plasma glucagon concentrations were significantly increased over baseline (100%) at +10 minutes (P < .001) and peaked at +30 minutes (122% +/- 7%, P < .001). Seven subjects (out of 15 tested) developed symptoms. Quantitative evaluation revealed a peak in the mean symptom score at +20 minutes, an increase from 0.4 +/- 0.3 to 2.6 +/- 0.1 arbitrary units (P < .06).(ABSTRACT TRUNCATED AT 250 WORDS)

Acute effects of high fat and high glucose meals on the growth hormone response to exercise

The health promoting, anabolic effects of physical activity may be mediated, in part, by an exercise-associated increase in GH. However, little is known about the acute effects of diet on exercise-induced GH release. We hypothesized that a single meal could attenuate the GH response to exercise by modulating substances like somatostatin, insulin, or glucose. Eleven healthy young adults performed 10 min of high intensity, standardized cycle ergometry in the morning following an overnight fast. On separate days they ingested a noncaloric placebo liquid meal or an isovolemic, isocaloric liquid meal high in either fat or glucose. Venous blood samples were obtained before and for 90 min after exercise began, whereas gas exchange data were measured breath by breath. Peak mean oxygen consumption (VO2) was, on average, 9-fold greater than preexercise levels in all groups. Although there was no difference in preexercise GH levels, mean peak, postexercise GH was 54% lower after the high-fat meal compared with placebo (P < 0.01). Modest decreases in GH response to exercise after the high-glucose meal were not statistically significant. Mean serum somatostatin was significantly higher after the high-fat meal compared with both high glucose and placebo meals. This study demonstrates that exercise-induced GH release can be significantly attenuated by the contents of a single preexercise meal. The high fat meal increased circulating somatostatin and was associated with an inhibition of the GH secretion. The data provide a possible specific mechanism to explain how diet can acutely modulate the anabolic effects of exercise.

A mathematical model of oscillatory insulin secretion

Insulin is secreted in sustained oscillatory fashion from isolated islets of Langerhans. This finding has led to the assumption of an underlying synchronizing process that coordinates insulin oscillations. This assumption was tested by developing a mathematical model of oscillatory insulin secretion in which we included degree of synchrony as a parameter. We first evaluated insulin oscillations in perifused isolated rat islets, using spectral analysis to determine their regularity and frequency. A parsimonious mathematical model was developed to account for these characteristics. The model postulates a group of secretory units discharging at discrete intervals with the same underlying period. Variation from two sources, phase differences between units (synchrony) and regularity within units, is introduced by adding two normally distributed random variables with standard deviations (Sg and Si, respectively) to the secretory period. Sets of 100 simulations for different values of Sg and Si were run. Results of the simulations suggest that the system tolerates a relatively large degree of asynchrony yet still demonstrates regularity of oscillations on spectral analysis. Comparison with perifusion data suggests that a moderate degree of asynchrony between islets can best account for the pattern of insulin oscillations observed. This model provides a theoretical basis for the study of mechanisms for insulin oscillations.

Oscillations of lactate released from islets of Langerhans: evidence for oscillatory glycolysis in beta-cells

Oscillations in the glycolytic process have been demonstrated in a number of different biological systems. However, their presence has never been demonstrated in insulin-secreting beta-cells. We used lactate as a marker for glycolysis and measured lactate and insulin concentrations in the effluent of isolated perifused rat islets of Langerhans. Sustained regular oscillations in lactate concentrations with an average period of 16-20 min were observed in islets that were perifused with medium containing 5.5 or 16.7 mM glucose. Sustained oscillations of insulin concentrations secreted by the islets were also observed in these experiments, and the average period of oscillation was 14.6 +/- 2.3 min at 16.7 mM glucose. Mean insulin concentrations at 5.5 mM glucose were too low to permit analysis of oscillations. Spectral analysis confirmed the regularity of the lactate and insulin oscillations and showed peaks that were consistent with the average periods obtained using the Clifton program. Moreover, spectral analysis demonstrated marked similarity between the patterns of lactate and insulin oscillation during perifusion with 16.7 mM glucose. Cross-correlation analysis found these oscillations not to be consistently in phase. In conclusion, sustained oscillations in lactate released from islets of Langerhans suggest that the glycolytic process in beta-cells also oscillates. The similarity of the periods of lactate and insulin raises the possibility that oscillations in glycolysis may provide a mechanism for pulsatile insulin secretion.

Sustained pulsatile insulin secretion from adenomatous human beta-cells. Synchronous cycling of insulin, C-peptide, and proinsulin

The endocrine pancreas secretes insulin in a pulsatile fashion. This rhythm is generated at a site within the pancreas, although its precise location has not been determined. With an in vitro system, we tested the possibility that beta-cells might generate spontaneous pulsatile insulin secretion in the absence of any external influence. Human insulinoma tissue from five patients was perifused for 7-10 h with RPMI-1640 medium and constant concentrations of glucose (5.5 mM). Insulin, C-peptide, and proinsulin were measured in the effluent collected at 3.3-min intervals. All three peptides demonstrated pulsatility of secretion in a similar, synchronous fashion that was sustained throughout each study. The Clifton cycle detection program demonstrated cycling in all five tumors, with an average period for all tumors of 28, 29, and 26 min for insulin, C-peptide, and proinsulin, respectively. Spectral analysis confirmed the regularity and consistency of the hormonal secretory patterns. Mean hormone concentrations secreted by different tumors varied, but insulin and C-peptide were secreted in a nearly 1:1 ratio. This study demonstrates 1) that beta-cells are able to generate spontaneous pulsatile insulin secretory activity, which is independent of innervation or the presence of other islet cells, and 2) proinsulin secretion from the beta-cell also has an inherent pulsatility. The synchrony observed in the cycles of proinsulin and its peptide products confirms their common secretory pathway in the beta-cell. We conclude that the beta-cell may be the originator of insulin cycling.

Catecholamine concentrations in the hepatic portal system: effect of surgical stress upon portal levels

The biological effects of a hormone are dependent on the concentration delivered to the target tissue. This is generally best reflected in the arterial concentration. However, the liver is unique in that it receives substantial additional blood flow from the portal venous system. This may be important in the case of the catecholamines, where extraction or spillover from the splanchnic circulation may occur. In this study we examined portal venous catecholamine concentrations in anesthetized, laparotomized rabbits. We compared the values with simultaneously sampled arterial levels to evaluate the effects of the splanchnic tissues upon a wide range of catecholamine concentrations delivered to the liver during a state of stress. Spillover of norepinephrine and extraction of epinephrine were observed in all rabbits. Mean (+/- SEM) arterial norepinephrine concentrations were elevated, 716 +/- 167 pg/ml (n = 11); mean portal concentrations were 178 +/- 37% higher (p less than 0.01), at 1,425 +/- 301 pg/ml, representing net spillover from splanchnic tissues. In addition, significant extraction of epinephrine was observed; arterial concentrations were 2,144 +/- 580 pg/ml (n = 11). Portal levels were 1,205 +/- 382 pg/ml, 38 +/- 7.45% lower than corresponding arterial concentrations (p less than 0.02). Furthermore, there was a concentration-dependent effect upon norepinephrine spillover; the highest arterial norepinephrine concentrations were associated with the lowest splanchnic spillover. This resulted in a negative correlation between the arterial norepinephrine levels and the percent increase from spillover into the portal vein (r = -0.81, p less than 0.003). We conclude that portal venous catecholamine concentrations are significantly different from arterial levels in anesthetized, laparotomized rabbits over a wide range of concentrations.(ABSTRACT TRUNCATED AT 250 WORDS)