Molnar, George D.

Abstract A glucose oxidase-peroxidase method for continuous automated monitoring of blood glucose has been developed. The response is linear over the range 0-800 mg/100 ml. Sensitivity can be maintained for 24 hr or longer and can be restored by rinsing the analytic system with sulfuric acid to permit studies of longer than 48 hr in duration. A precision of ± 1% can be maintained between rinses for samples containing 100-600 mg of glucose per 100 ml. This method is satisfactorily specific for glucose: The response with other sugars is less than 1% of the response obtained with the same concentration of glucose. Ascorbic acid causes no significant inhibition of the response to glucose. The inhibition by uric acid has been reduced fifty-fold compared to that in other methods. Transit through the sampling catheter and analytic system requires 15 min. Timed from the first detectable response to a change in concentration, 25% of total response is achieved in 30 sec and 90% in 80 sec. Fifty percent of an oscillation with a half-period of 45 sec can be detected; no oscillations this short were observed in records of human blood glucose. Applicability and feasibility of this method have been demonstrated in over 2000 hr of repeated blood glucose recordings in 12 diabetic and 6 normal subjects.

Eighteen diabetic patients representing a spectrum from hyperlabile to stable forms of this disease were investigated under metabolic ward conditions. Hyperlabile patients had the greatest daily and hourly variation in glycosuria and the most marked ketonuria. In our hyperlabile patients, multiple daily injections of short-acting insulin proved to be the only therapeutic measure that provided significant improvement in diabetic regulation when dietary and general health measures were already optimal. In more stable patients, longer acting insulin injected less frequently was found to be satisfactory therapy if careful adjustments to individual needs were made. Blood glucose and serum ketone-body and free fatty acid concentrations as well as qualitative estimates of glucose and ketone bodies in small samples of freshly voided urine were found to be suitable indicators of the immediate state of metabolic regulation. These parameters correlated with one another to a satisfactory degree, but they proved to be poorly correlated with measurements of urinary glucose and ketone bodies during periods of five to twenty-four hours. Quantitative analysis of individual blood samples or qualitative estimates of glucose or ketone bodies in freshly voided urine specimens proved to be unreliable indices of the patients' long-term metabolic compensation except in the most stable patients. Quantitations of urinary glucose and ketone bodiesin five, nine, and twenty-four-hour collections of urine, on the other hand, give satisfactory data to assess diabetic regulation, both as to accuracy of compensation and stability.

Insulin binding by insulin antibodies was determined serially in eight menstruating and in twelve pregnant diabetic subjects. No significant variation was noted in the insulin binding values before, during or after menstruation or during or after pregnancy. This constancy of insulin binding was observed despite major and minor changes in insulin needs. No causal relationship could be demonstrated between insulin binding by insulin antibodies and changing insulin requirements. If there is a relationship between the hormonal changes and the alterations in insulin needs during menstruation and pregnancy, the mechanism of this relationship is not mediated through the insulin binding capacity of circulating insulin antibodies.

Exercise studies were performed on six intact and seven pancreatectomized, unanesthetized dogs to examine the conditions under which exercise decreases the concentration of glucose in blood. No major change in blood glucose concentration was demonstrated during exercise experiments on normoglycemic intact dogs. In the pancreatectomized dogs, the blood glucose concentration curves during exercise followed either one of two patterns: Some rose with exercise and some fell. In dogs with baseline blood glucose concentrations less than 400 mg. per 100 ml., a sustained decline in blood glucose occurred with exercise, while in those dogs with an initial level greater than 400, a rise occurred. No effect of exercise was evident in a pancreatectomized dog with blood glucose concentration near the normal range during continuous infusion of insulin. Exercise produced a decrease in blood glucose concentration in an intact dog with hyperglycemia induced by glucose infusion, but this dog's blood glucose level increased during rest periods following exercise.

Three normal, three stable diabetic, and eight unstable diabetic subjects were investigated, with continuous automated blood glucose analysis for forty-eight-hour periods, during metabolic balance studies of six days' duration under near normal conditions (fed and ambulatory) after the attainment of clinically optimal diabetic regulation. The studies on unstable diabetics were performed during constant dietary intake, with one or two daily injections of intermediate-acting insulin and then repeated with four daily injections of short-acting insulin. A characteristic of blood glucose behavior, the mean amplitude of glycemic excursion (MAGE), was measured. MAGE was small for normals (range, 22 to 60 mg./100 ml.), larger for stable diabetics (67 to 82 mg./100 ml.), and largest for unstable diabetics (119 to 200 mg./100 ml.). Associated with a significant decrease (p = 0.004) in the mean diurnal glycemic level (from 146 to 244 mg,/100 ml. to 101 to 152 mg./100 ml.) achieved through the use of four daily doses of short-acting insulin there was a significant increase (p = 0.006) in hypoglycemic episodes (from 0 to 4/48 hr. to 3 to 6/48 hr.) without significant change in MAGE. The persistently large value of MAGE despite therapy with multiple injections of short-acting insulin appears to be a characteristic of unstable diabetes.

The responses of glucagon, growth hormone, and insulin secretion to the oral administration of glucose and to the intravenous infusion of saline, arginine, and insulin were measured in seven patients who had stable diabetes, eight who had unstable diabetes, and seven healthy volunteers. Hyperglycemia suppressed secretion of glucagon in normal subjects but not in diabetics. The oral glucose and arginine infusion tests demonstrated partial preservation of insulin-secretory ability in stable diabetics and its virtual absence in unstable diabetics. Glucagon responses to arginine infusion were similar in all three groups. In response to hypoglycemia induced by insulin infusion, the concentrations of plasma glucagon increased in normal subjects and, to a lesser extent, in stable diabetics but increased in only two of the unstable diabetics. The impairment in glucagon response during hypoglycemia in diabetics correlated positively with the degree of diabetic instability and insulin deficiency during glucose and arginine testing. The severity of the insulin deficiency also correlated with the degree of diabetic instability. These findings support the hypothesis that inherent abnormalities of insulin and glucagon secretion may account for many of the clinical characteristics of unstable and stable diabetic patients.

Cellular energy substrates and metabolites, electrolytes, oxygenation, and acid-base balance, in arterial and coronary sinus blood of seven adult patients who died after valvular replacements, were studied and compared with these measurements from five patients having aortic replacement who survived. Six of the patients were in New York Heart Association class IV. All but one died within 24 hr of acute circulatory failure, with excessive bleeding being a contributing factor. Plasma potassium and osmolality remained higher in the first postoperative hours in nonsurvivors. Ketone bodies, growth hormone, and lactate levels rose precipitously as shock became severe. Increased arterial blood osmolality and potassium, plus a still-rising lactate concentration, were evident 2 hr after operation in patients who died in the next 24 hr. Low cardiac output, hyperosmolality, and the degree of lactic acidosis were the only early changes predicting the fatal outcome.